100% found this document useful (1 vote)

1K views375 pages

CE122 Reader

Uploaded by

Lala Dakessian

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

100% found this document useful (1 vote)

1K views375 pages

CE122 Reader

Uploaded by

Lala Dakessian

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 375

Abolhassan Astaneh-Asl, Ph.D., P.E.

Professor, Spring 2015 Edition

University of California, Berkeley
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
ii

Copyright © 2003-2015 by Abolhassan Astaneh-Asl., All rights reserved. No part of this

publication may be reproduced, stored in a database or a retrieval system, or transmitted, in any
form or by any means by anyone except under the terms of the Copyright Designs and Patents Act
1988 or under the terms of a license or a permit issued by the author and copyright holder
Abolhassan Astaneh-Asl in writing..

Cover Photo Credits: San Francisco Skyline (Background): Kaveh Astaneh-Asl

(Inset): Hearst Headquarters, New York City, Photo: Courtesy of
WSP Cantor Seinuk, Structural Engineers (www.wspgroup.com).
(Inset): X-Connection of Hearst Headquarters: Photo: Courtesy of
Cives Steel Company (www.cives.com).

ii
CONTENTS iii

Abolhassan Astaneh-Asl, Ph.D., P.E.,

Professor, University of California, Berkeley

Contents:
Chapter 1 - Introduction to Steel Structures and Design
Chapter 2 - Material Properties of Steel
Chapter 3 - Introduction to Loads Applied to Structures
Chapter 4 -Tension Members
Chapter 5 - Compression Members (Columns)
Chapter 6 - Flexural Members (Beams)
Chapter 7 - Members Subjected to Shear
Chapter 8 - Beam-Columns (Members under Combined Loads)
Chapter 9 - Welds
Chapter 10 - Bolts
Chapter 11 - Connections in Steel Structures
Chapter 12 - Eccentrically Loaded Welded and Bolted Connections
Chapter 13 - Shear Connections
Chapter 14 - Moment Connections
Chapter 15 - Introduction to Structural Systems

iii
Behavior and Design of Steel Structures, Copyright  2003-2015 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122N-Design of Steel Structure, CE122L- Steel Design Project, and CE247-Design
of Steel and Composite Structures at UC Berkeley for their personal use and should not be distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
iv

iv
CHAPTER 1-INTRODUCTION 1-1

CHAPTER 1

INTRODUCTION TO STEEL
STRUCTURES AND DESIGN

Soda Hall, UC Berkeley, Photo by A. Astaneh

Objectives:
In this chapter, you will learn:
1. The advantages of steel structures
2. The use of steel in a variety of structures including buildings bridges
transmission towers, reservoirs and other structures
3. Types of steel structural systems used in buildings
4. Failure modes (or limit states) of steel structures
5. Design methods used in design of steel structures, which are the Load
and Resistance Factor Design (LRFD) method, the Allowable Strength
Design (ASD) method, and the Plastic Design (PD) method.

1.1. INTRODUCTION TO STEEL STRUCTURES

Because of its properties and relatively low cost-to-strength ratio, we use steel in a variety of
applications in civil engineering. Focusing on the structural applications, we use steel to construct
structures of buildings, bridges, stadiums, power transmission towers, liquid and gas reservoirs,
industrial facilities and many others. We even use steel in artwork and historical monuments such
as the Statue of Liberty, which has a steel truss skeleton inside it to support its weight and
environmental loads, such as wind, applied to it. Figure 1.1 shows examples of steel structures.
Depending on the type of structure and main material used in it, we need to use a specific
design code. For example, for design of steel structures of buildings and similar facilities, shown
on the top three rows of Figure 1.1, in United States, we use “Specification for Structural Steel
Buildings”(AISC, 2010a), developed by the American Institute of Steel Construction (AISC).

___________________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be distributed
to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
1-2

Sports Facilities,
Sydney Olympic Stadium
Photo by A. Astaneh

Statue of Liberty
Photo: www2.cnrs.fr
Skyscrapers,
Photo: www.SOM.com

Use of Steel as Facade Elements

Cultural Buildings, Bilbao Museum China Central TV Headquarters
Photo: www.artknowledgenews.com
Photo: www.scz.ch

Sculpture made of steel

Reservoirs,
Elevated Water Tank
Photo by: (unknown)

Transmission Towers Industrial Facilities

Photo www. picture- Photo by A. Astaneh
newsletter .com

Higashi-Kobe Bridge-
Photo by A. Astaneh Medium and Short Span Bridges
Sydney Harbour Bridge
Long-Span Bridges, Photo by A. Astaneh Short Span Bridges and
Golden Gate Bridge Overpasses
Photo by A. Astaneh Photo by A. Astaneh

Figure 1.1. Examples of Use of Steel in Buildings, Other Structures, and Bridges
CHAPTER 1-INTRODUCTION 1-3

Another example is bridge structures, shown in the bottom row of Figure 1.1. Bridges are
designed in the United States according to the provisions of the “LRFD Bridge Design
Specifications” (AASHTO, 2010) developed by the American Association of State Highway and
Transportation Officials (AASHTO).
In this document, we will use the AISC Specification for Structural Steel Buildings (AISC,
2010a) and will focus primarily on the design of steel structures for buildings and other structures.
The information presented and the design concepts and equations discussed here are in most parts
similar to those applied to design of steel bridge structures. The main difference between the
design of building and bridge structures is in the design loads. For bridges, the design loads
include the live loads due to the weight of the cars and trains crossing the bridge.

1.2. ADVANTAGES OF STEEL

a. Relatively high strength to weight ratio, which results in smaller and lighter steel
members for the same load than the reinforced concrete. Smaller and lighter
members are particularly advantageous since the lower weight of steel structures
results in smaller forces to be supported by the foundations and supporting soil,
smaller inertia forces generated in the steel structures during earthquakes and for
tall buildings, the lower weight results in smaller columns occupying the useable
floor spaces.
b. Almost constant and relatively high modulus of elasticity, which is very
advantageous in controlling deflections of the members as well as lateral deflection
(drift) of the structure under the effects of wind and earthquake.
c. High yield stress, which results in steel members to remain elastic under relatively
high stresses.
d. Large ultimate strain of about 0.2, which makes steel very ductile to dissipate
energy and to provide high damping in dynamic and cyclic applications such as
blast, impact, and seismic events.
e. Almost symmetric behavior of steel under tension and compression, which
enables steel components to resist load reversal during cyclic events better. In
addition, in case of change in stress distribution within the structure, this property
of steel enables the structure to resist the redistributed stresses, which were not
considered when the original design was done.
f. Durability and hardness of steel against wear and tear, which results in longer
service life for steel structures.
g. Almost no aging, creep, or shrinkage effects, , which results in properties of steel
remaining unchanged during its lifetime.
h. Air tightness of steel, which results in steel being very suitable for water tanks,
gas storage tanks and reservoirs for other liquids and gases especially when
hazardous material are stored.
i. Prefabrication and Speed of Construction, which enables components of steel
structures to be constructed in the shop and then erected at the site using welds and
bolts. This results in ease of transportation and shortens the duration of the

construction. In addition, prefabrication can result in construction of steel

structures in a very cold or hot environment.
j. The “green” nature of steel structures and their recycling, which enables steel
recovered from demolished structures to be recycled and re-used. Large
percentage of steel used in structures today, is produced by melting “scrap” steel
from cars, other machines and demolished structures.

1.3. FAILURE MODES OF STEEL STRUCTURES

Steel structures, when subjected to loads or displacements, can fail in a number of ways. The main
failure modes of steel structures are:
a. Local buckling is a failure mode in steel structures that can result in reducing the
load carrying capacity of the members in bending, compression or combined
bending and compression. Local buckling of steel members can be considered
comparable to the buckling of compressive rebars in reinforced concrete
members. We will discuss local buckling in Chapter 5.
b. Member buckling - steel members, due to their high strength to weight ratio have
smaller cross sections than the comparable reinforced concrete members do. As a
result, overall buckling of a steel member can result in reduction of its
compressive or bending capacity. To increase buckling capacity, bracings are
added to steel columns and beams to reduce their slenderness and to increase their
compressive capacity. In Chapter 5, we will also discuss overall buckling.
c. Global structural stability - again, due to relatively smaller sizes of members in a
steel structure and flexibility of some connections, such as shear connections,
compared to a reinforced concrete structure, the overall stability of a steel
structure, especially under P-∆ effects, needs to be considered in design. We will
discuss this issue in Chapter 8.
d. Fatigue - If steel is subjected to very large number of alternating tensile stress, it
can develop fatigue cracks. Such cracks under further applications of cyclic
tensile stresses can propagate. When fatigue cracks have propagated to a
“critical” length, the crack can suddenly propagate in a very brittle manner and
result in fatigue fracture failure. Fatigue failure is not a common failure mode for
typical steel building structures. However, some elements of the structure, such
as supports of machinery, where large number of alternating tensile stresses can
occur may need to be checked for fatigue. Fatigue failure mode can be a
governing failure mode in bridges where during the life of a given bridge, when
several million trucks and cars passing over the bridge can create alternating
(cyclic) tensile stresses large enough to cause fatigue failure. Both, the AISC
Specification (AISC, 2010a) and the AASHTO Specifications (AASHTO, 2010)
have provisions on fatigue failure mode for steel buildings and bridges
respectively.
e. Fracture - Steel members, especially those with areas of high stress concentration,
can fracture if subjected to impact loads. This failure mode is not a common
failure mode for typical steel building structures subjected to gravity loads.
However, for exposed structures, with constrained connections and sharp changes
in the direction of stresses, this failure mode can be the governing failure mode.
CHAPTER 1-INTRODUCTION 1-5

f. Fire effects - Structural properties of steel, such as yield stress and modulus of
elasticity, generally decrease as the temperature of steel rises. If the temperature
of a steel member continues to rise during a fire, there will be a time when the
reduced strength of the member will be less than the applied forces and the
member will fail. To avoid this situation, fire codes have provisions on adding
fireproofing to the surface of steel structures. More information on this item is in
Chapter 2.
g. Cold temperature effects – fracture toughness and ductility of steel under cold
temperature decreases. This makes a steel element that was acceptable to be used
in one location unacceptable for use in another location, where the steel can be
exposed to very low temperatures. In cold regions such as in Alaska, where steel
can be exposed to below freezing temperatures, it is necessary to use a type of
steel that has relatively high fracture toughness and ductility. More on material
properties is given in Chapter 2.
h. Corrosion - in corrosive environment, unpainted steel corrodes resulting in
decrease of cross sectional areas of structural components. For most structures,
such as typical buildings, one-time painting of steel will prevent corrosion.
However, for exposed steel structures, such as bridges, especially when they are
located over or near salty waters and/or corrosive environment, continuous
painting and maintenance of steel structure may be needed to control and prevent
corrosion. Other solutions such as the use of corrosion-resistant steels,
galvanizing the members or the use of weathering steel can be employed
economically for long-term protection of steel structures against corrosion.
i. Impact and blast effects- both steel and concrete, if exposed to high velocity
impact and blast should be designed to resist dynamic effects of such loads.
Material with high ductility, such as steel, can resist high velocity dynamic loads
better by dissipating energy of the impact through yielding.

1.4. STEEL STRUCTURAL SYSTEMS IN BUILDINGS

Figure 1.2 shows common types of steel structural systems for buildings.

Concentrically Moment Eccentrically Shear Wall Dual System

Braced Frames Frames Braced Frames System

Figure 1.2 Examples of Common Steel Building Framing Systems

1.5. DEFINITION OF DESIGN

Design is defined as satisfying the following basic equation of design for each limit state:

DEMAND < SUPPLY (1.1)

The designer formulates the design requirements, that were discussed earlier, into the above limit state
format and then by selecting proper material, configurations and geometry satisfies the limit state equations.

1.6. DESIGN CRITERIA

The basic equation of design given earlier is DEMAND ≤ SUPPLY. For structural design, the
equation can be written in terms of load effects being the Demand and resistance being the
capacity. Then, basic equation of design takes the following form:

Load effects ≤ Resistance (1.2)

In order to satisfy this equation, the designer will define a series of limit states that the
structure can reach. The limit states are defined based on the design requirements that were
discussed earlier . Since limit states form the backbone of this procedure, it is called "Limit State
Design.”

1.7. SELECTION OF FORMAT FOR DESIGN

CRITERIA
The equation of design or limit state design criteria can be applied in several formats. Common
formats used in design of steel structures are the Allowable Strength Design (ASD), the Load and
Resistance Factor Design (LRFD) and the Plastic Design (PD). These are the only three formats,
which sometimes are called design methods, allowed by the current AISC Specification (2010) for
design of steel structures. In the following, we discuss each of the three methods briefly with an
emphasis on the ASD and particularly the LRFD methods since these two methods are covered
extensively in the AISC Specification (2010). In this course, we will use the LRFD methods. The
Plastic design methods are discussed in graduate courses on plastic design such as the CE248
course here at UC Berkeley.

1.7.a. Allowable Strength Design (ASD) Method

In the Allowable Strength Design (ASD) method, the design Equation 1.2 given earlier as Load
effects ≤ Resistance becomes:

P service ≤ Rallowable (1.3)

Where,
Pservice = load effect due to applied nominal service loads
Rallowable = allowable value of resistance
CHAPTER 1-INTRODUCTION 1-7

The effect of applied loads, Pservice on the left side of Equation 1.3, is calculated by
analyzing the structure subjected to service loads. In this method the structure is analyzed
elastically. The allowable resistance, Rallowable, the right side of Equation 1.3, is defined by the
governing code such as the AISC Specification (2010) as:
Rallowable ≤ Rn / Ω (1.4)
Where;
Rn = Nominal strength (capacity) of the component for limit state under consideration
Ω = Safety factor for the limit state under consideration
The value of nominal capacity, Rn is calculated using the so-called “specified” material and
geometric properties in the equations given by the specifications for limit state under consideration.
The AISC Specification (2010) provides values of the safety factor Ω for each limit state.
In summary, in the Allowable Strength Design (ASD) method of the AISC Specification
(AISC, 2010a):
1. Material properties used in the Allowable Strength Design (ASD) are the “specified,”
also called “nominal” properties. These specified values of material strength are
generally established by the material specifications such as the ASTM-A6 Standards.
For example, for A36 steel, specified (nominal) values of the yield stress, Fy is given
as 36 ksi and the ultimate strength, Fu as 58 ksi. For all structural steel, modulus of
elasticity, E, is 29,000 ksi. National Standards such as the ASTM establish material
properties of steel for use in analysis and design. In Chapter 2, we discuss material
properties of steel.
2. The dimensions of the structure and its components used in the allowable Strength
Design (ASD) are nominal values. These geometric properties include span, height of
floors, area, moment of inertia, radius of gyration, and others. The nominal
dimensions of the structure itself, such as the height of the floors and span lengths are
given on the structural drawing as the center-to-center of the columns and beams. The
AISC Steel Construction Manual (AISC, 2011) has tables in its Chapter 1 listing
nominal dimensions of the cross sections and geometric properties such as area,
moment of inertia, for most common steel shapes.
3. The loads used in the Allowable Strength Design (ASD) are “nominal service” loads.
The nominal service loads and their combinations to be used in the ASD method are
specified by the governing codes. For example, nominal values of the dead, live,
seismic, wind, snow and other design loads are specified in the ASCE-7 (ASCE/SEI,
2010) standard and IBC (2009) for the U.S. Chapter 3 is devoted to loads and load
combinations according to the ASCE/SEI-7 standard (ASCE, 2010).
4. The resistance or “allowable strength” values used in the Allowable Strength Design
(ASD) are the nominal strength divided by a factor of safety Ω. The AISC
Specification (AISC, 2010a) provides values of Ω for each limit state of failure. For
limit states involving yielding or buckling, factor of safety Ω, is about 1.5 to 1.67 and
for limit states that involve fracture of steel, Ω is 2.0.
5. The “Equation of Design” for each limit sate in the Allowable Strength Design(ASD)
is in the form of:

Ps ≤ Rn / Ω (1.5)
Where;
P s = Load effect (such as force, moment, stress and deflection) due to applied nominal
service loads. P s is established by the analysis of the structure subjected to loads.
Rn = nominal strength (capacity) of the component calculated using provisions of the AISC
Specification (AISC, 2010a)
Ω = Safety factor given by the AISC Specification (AISC, 2010a).

1.7.b. Load and Resistance Factor Design (LRFD) Method

As mentioned earlier, there are many uncertainties in establishing loads that are applied to
structures. There are also uncertainties in material properties and dimensions of the structures used
in establishing the capacity or resistance side of equation of design. The LRFD method is
developed to address these uncertainties in a systematic way without making the method too
mathematical and cumbersome to be used in structural design offices.
Considering that both Load Effects and Resistance in the design equation are random
variables, let us consider variations of load effect P and resistance R in the design equation. Both P
and R are random variables with their own distribution as shown in Figure 1.6. Notice that each
point on R curve represents a possible value of resistance. Similarly, each point on P curve
represents a possible value of load effect. In figure below, Pm and Rm are mean values of the
random variable P and R respectively. If an occurrence of load and resistance is represented by
horizontal Line AA, it is clear that, for this case, not only is there no factor of safety but member
has already failed since resistance, RA is much less than the load effect; PA. The shaded area in
Figure 1.6 indicates occurrences of failure since in this region load effects (values on P-Curve) are
greater than resistance (values on R-Curve).For definition of terms used here and on Figures 1.6
and 1.7 see below.
Let us define a new variable Z equal to R-P (R minus P). Since failure is defined as R
(resistance), being, less than P (applied load effect) then by defining Z equal to R-P, we say that
failure will occur when Z becomes zero or negative. The probability density function of random
variable Z is shown in Figure 1.7. The shaded area represents negative realization of Z function
thus representing probability of failure.

P- Curve
Probability (Load Effects) R- Curve
Density (Resistance)
Function

A A

Pm RA PA Rm Load and Resistance

Figure 1.6. Frequency of Occurrence of Load and Resistance

CHAPTER 1-INTRODUCTION 1-9

pdf- Curve for

Probability Density Z ≤ 0, Failure parameter Z
Function

Zm Random Variable Z=R-P

Figure 1.7. Probability Density Function of Z= R-P

The terms used here and on Figures 1.6 and 1.7 are:
P = Random variable representing load effects
Pm = Mean value of random variable P (which represents load effects)
PA = Value of random variable P (which represents load effects) corresponding to Point A
R = random variable representing resistance
Rm = Mean value of random variable R (which represents resistance)
RA =Value of random variable R (which represents resistance) corresponding to Point A
Zm = Mean value of function Z = R-P

Notice that total area under any probability density function is equal to 1.0. As shown in
Figure 1.7, it is the distance from mean point to the origin, denoted as Zm that is the measure of
safety since the larger this distance is, the smaller the negative shaded area, representing probability
of failure will be. The distance Zm divided by the standard deviation of function Z is the “Safety
Index” β and is:
β = Zm / σz (1.6)
Where;
β = Safety Index
σz = Standard deviation of function Z
If random variables R and P have a lognormal distribution, then safety index β is:

nRm − nPm
β= (1.7)
v +v
2 2
R P

Using safety index approach is a way of incorporating probabilistic nature of the

structural properties in design. This approach is a Level 2 probabilistic approach. However, this
___________________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be distributed
to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
1-10

method is somewhat difficult and not ready yet for design office use and for actual design. The
reason is we can use the level 2 approach and calculate safety index β of a given structure or
component but we cannot make a judgment on whether or not the calculated safety index is
acceptable for the design at hand.
Based on the above argument, the profession has moved in the direction of developing a
simpler and approximate method that still is based on probabilistic nature of the structural design
parameters but is not as tedious as Level 2 probabilistic methods. In developing probabilistic
codes, the random nature of resistance and load variables is considered. In these codes, two types
of factors are developed: (1) reduction factors to be applied to the resistance variables to reduce
their values and; (2) load factors to be applied to the loads to increase their values. The name Load
and Resistance Factor Design is derived from this fundamental operation. Load and resistance
factors are established by considering the probabilistic nature of load and resistance variables and
by using the mean and standard deviation of variables.
Following items summarize the material properties, geometries, loads, and Design
Equation used in LRFD method of the AISC Specification (AISC, 2010a):
1. Material properties used in the Load and Resistance Factor Design (LRFD) of steel
structures are the “specified”, also called “nominal”, values. These specified values are
generally established by the material specifications such as ASTM-A6 Standards and
adapted by the AISC Specification (AISC, 2010a). The material properties used in the
LRFD method are the same as those used in the ASD method.
2. Dimensions of the structure and its components used in the Load and Resistance Factor
Design are nominal values. The geometric properties used in the LRFD method are the
same as those used in the ASD method.
3. The loads used in the Load and Resistance Factor Design (LRFD) method are “nominal
service” loads. The nominal service loads and their combinations for design are specified
by the codes and standards such as the ASCE-7 (ASCE/SEI, 2010) and IBC (2009). The
nominal loads used in the LRFD method are the same as those used in the ASD method.
4. The resistance values used in the Load and Resistance Factor Design (LRFD) method are
the nominal strength values multiplied by a resistance reduction factor φ . Values of φ
for each limit state of failure is given by the AISC Specification (AISC, 2010a).
Generally, for limit states involving yielding or buckling, the resistance reduction factor,
φ., is about 0.90 to 1.0 (mostly 0.90) and for limit states that involve fracture of steel, φ is
0.75.
5. “Equation of Design” for each limit state in the Load and Resistance Factor Design
(LRFD) is in the form of:

Pu ≤ φ R n (1.8)
Where
P u = applied factored load effect (such as force, moment or stress) due to applied nominal
loads multiplied by load factors. Pu is established by the analysis
Rn = nominal capacity of the component calculated using provisions of the AISC
Specification (AISC, 2010a). Rn is the same in ASD and LRFD methods.
φ = resistance reduction factor (less than or equal to 1.0) given by the AISC Specification
(AISC, 2010a).
CHAPTER 1-INTRODUCTION 1-11

1.7.c. Plastic Design (PD) Method

This method is an ultimate design method in which ultimate strength is defined as reaching plastic
collapse mechanism. The method is mostly applied to ductile material such as steel. This method
is discussed at length in other courses on inelastic design of steel structures such as the CE248. We
will not discuss this method here in this document beyond this introductory section.
The Plastic Design method is the third method (in addition to the ASD and LRFD) that is
currently allowed by the AISC Specification (AISC, 2010a) to be used in design of steel structures.
The provisions governing this method are in the Appendix 1 –Inelastic Analysis and Design of the
AISC Specification (AISC, 2010a). The appendix has provisions that are specifically for cases
where inelastic analysis and plastic design is performed.
The equation of design in Plastic Design method is the same as in the LRFD method given
by Equation 1.8 earlier, which is repeated here as Equation 1.8:
Pu ≤ φ Rn (same as 1.8)
The terms in the above equation are similar to the terms defined for the LRFD methods in
previous section with the main difference being in the definition of the “Applied Load Pu” which
for Plastic Design should be established using inelastic (plastic) analysis and not by elastic analysis
as usually done in the LRFD methods. The terms in Equation 1.8 above for Plastic Design are:
P u = Applied factored load effect (such as force, moment or stress) due to applied nominal loads
multiplied by load factors. The load factors are the same as used in the LRFD and given by
load standards such as the ASCE/SEI-7-05 Standard (ASCE/SEI, 2010). However, for
Plastic Design Pu is established by the inelastic (plastic) analysis of the collapse
mechanisms. Elastic analysis, which is normally used in the ASD and LRFD methods, is not
allowed in Plastic Design
Rn = Nominal capacity of the component calculated using provisions of the AISC Specification
(AISC, 2010a). Rn is the same in ASD, LRFD, and PD. However, in Plastic Design
members and systems should satisfy more stringent requirements regarding their ductility and
prevention of their instability.
φ = Resistance reduction factor (less than or equal to 1.0) given by the AISC Specification (AISC,
2010a) for the LRFD methods.
Table 1.1 provides a comparison of major features of ASD, LRFD and PD methods, all
three currently recognized by the AISC Specification (AISC, 2010a).

Table 1.1. Parameters Considered in the ASD, LRFD, and PD Methods

Allowable Strength Load and Resistance Factor Plastic Design

Design Method Design Method Method
Item (ASD) (LRFD) (PD)

Design Equation P s ≤ R n /Ω P u ≤ φ Rn P u-PD ≤ φ R n

Loads P s = load effects due to P u = load effects due to elastic P u = load effects due to
used in the analysis elastic analysis of the analysis of the structure inelastic analysis of the
structure subjected to subjected to factored loads structure subjected to
service (nominal) loads factored loads
Load Factors No load factors are used Given by the governing code Given by the governing code
in ASD such as the ASCE-07 (2010) such as the ASCE-07 (2010)

Resistance R n = nominal resistance R n = nominal resistance R n = nominal resistance

Resistance-reduction Ω = factor of safety φ = resistance-reduction φ = resistance-reduction

factors given by the AISC factor given by the AISC factor given by the AISC
Specification (2010) Specification (2010) Specification (2010)

Analysis Method Elastic Analysis Elastic Analysis Inelastic Analysis of

Collapse Mechanism

References for Chapter 1

1. AASHTO, (2010), “AASHTO LRFD Bridge Design Specifications, Customary U.S. Units, 5th
Edition, with 2010 Interim Revisions,” American Association of State Highway and
Transportation Officials, Washington D.C.
2. AISC, (2010a) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-10, an
American National Standard, American Institute of Steel Construction, Chicago.
3. ASCE/SEI, (2010) “Minimum Design Loads for Building and Other Structures,” ASCE/SEI- 7-
10 Standard, American Society of Civil Engineers, Reston, VA.
4. Galambos, T. et al., (1978). “Special Issue of the ASCE Journal of Structural Division with
Papers on Load and Resistance Factor Design Criteria for Steel Structures,” J. of Structural
Division, ASCE, Vol. 104, No. ST9, September.
5. IBC, 2011, International Building Code, International Code Council, Washington, D.C.
CHAPTER 2- MATERIAL 2-1

CHAPTER 2

MATERIAL PROPERTIES
OF STEEL

Photo Credit:WWW.TIMKEN.com

Objectives:
In this chapter, you will learn:
1. Structural shapes used in steel and composite structures.
2. Geometric properties of structural shapes used in design.
3. Types of steel used in structures.
4. Mechanical properties of steel and concrete used in design of steel and
composite structures.

2.1. INTRODUCTION TO STRUCTURAL STEEL

Structural steel is made of iron combined with relatively small amounts of other elements
such as carbon, manganese, copper, aluminum, silicon, nickel, vanadium, molybdenum, sulfur,
phosphorus, and boron. Each of these elements, when added to iron results in changing certain
properties. For example, increasing carbon in steel increases its strength but reduces its ductility.
Ductility is the ability of material to deform inelastically prior to fracture. The increased strength
of steel due to increased carbon content can result in less material used, therefore reducing the
cost. However, significant reduction in ductility of steel, in most structural applications is not
acceptable.

______________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
2-2

Steel mills produce different types and grades of structural steel. In order to design
efficient and economical steel structures, we need to know material properties of steel. These
properties are mechanical properties and chemical composition.
The American Society for Testing and Materials (ASTM) has standards for properties of
engineering material used in the United States, including the structural steel. The ASTM standards
specify chemical composition of each type and grade of steel and how much of the elements such
as carbon, manganese and others need to be added to iron to make a specific type of structural
steel. The ASTM standards also provide Mechanical properties such as yield point and ultimate
strength for each type of structural steel.

Stress
Heat Treated High Strength
Low Alloy Steel

Carbon Steel

Aluminum

Timber

Concrete (in compression)

Strain

Figure 2.1. Typical Stress-Strain Curve for Common Structural Material

Figure 2.2. A Typical Universal Testing Machine and Two Tension

Test Coupons After the Test

One of the most important mechanical properties of steel is the “stress-strain” curve
showing the variation of stress applied to the material versus corresponding deformation expressed
in terms of strain. Figure 2.1 shows schematic stress-strain curves for common structural
materials, which are steel, concrete, aluminum, and timber. Note that the stress-strain curve for
steel is obtained by performing the ASTM standard uni-axial tension tests of standard specimens.
Figure 2.2 shows a typical Universal Testing Machine and tension coupons (specimens).
CHAPTER 2- MATERIAL 2-3

2.2. STRUCTURAL STEEL SHAPES

Figure 2.3 shows common structural steel shapes used in steel and composite structures.
The AISC Manual (AISC, 2011) has tables listing geometric properties of these sections, such as
the weight/ft, cross sectional area, moment of inertia, section modulus and other properties used in
design. Appendix 2-A of this Chapter provides samples of the AISC Manual Tables. The AISC
Tables list geometric properties of the hot-rolled shapes, plates and bars, double-angle and double
channels, and cold-formed shapes shown in Figures 2.3(a) through (d). The geometric properties
of built-up sections, with two examples shown in Figure 2.3(e), are not tabulated in the AISC
Manual since there are almost infinite number of geometries that we can come-up with for built-up
shapes by combining standard single shapes and plates. For built-up shapes, the geometric
properties are calculated using the dimensions of the cross section. The geometric properties of
the single shapes can also be calculated using the dimensions of the cross section.

Photo: Steel Dynamics Inc.

Wide Flange (WF) M or S Shapes Channel (C) Tee (WT) Angle (L) Pipe

(a) Hot Rolled Shapes

Plate Bar Double-Channel (2C) Double-Angle (2L)

(b) Hot Rolled Plates and Bars (c) Double Channels and Double Angles

Rectangular and Round Hollow Steel Sections (HSS) Welded Box Wide Flange + Plates

(d) Cold-Formed Shapes (e) Built-up Sections

Figure 2.3. Common Structural Steel Shapes Used in Steel Building Construction
______________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
2-4

2.3. MECHANICAL PROPERTIES OF STEEL

2.3. a. Mechanical Properties
Common mechanical properties of structural steel are:
1. Stress-strain curve, which is obtained by conducting standard ASTM tension
test. The stress strain curve establishes modulus of elasticity (E), yield stress
(F y ), ultimate strength (F u ) and maximum elongation (ε r ).
2. Coefficient of thermal expansion (ε).
3. Residual stresses (F R. )
4. Poisson’s ratio (ν.)
5. Hardness
6. Fracture toughness (K C.)
7. High cycle fatigue behavior (S-N curves.)
8. Thermal properties for fire protection.
9. Properties of steel under high strain rate applications such as blast cases.
10. Corrosion resistance
11. Low cycle fatigue behavior and cyclic stress-strain curves used in seismic
applications.)
12. Behavior under tri-axial state of stress.
Properties listed as Item 1 above are the ones that are used in day-to-day design of almost
all steel structures. Properties 2 through 9 in the above list are used in special cases where fatigue,
fracture or fire and blast protection is of concern. Property 10 is considered in design when steel is
exposed to corrosive environment. Properties 11 and 12 are not directly used in typical design, but
are incorporated into the provisions of the design codes and specifications.

2.3. b. Stress-Strain Curves

Parameters frequently used in design and analysis of steel structures are Modulus of
Elasticity, E, Yield Stress, F y , and Ultimate Strength, F u. These parameters are established by
conducting standard tension tests on coupons cut from the steel and machined down to have
standard dimensions required by the ASTM Standards.
Figure 2.4 shows typical stress strain curves for structural steels. The curve on the left is
for steels that when the applied tensile stress reaches the yield point, the tension coupon elongates
considerably without almost any increase in the stress. This behavior creates a horizontal segment
in the stress-strain curves called the “Yield Plateau” as shown in Figure 2.4(a).
Modulus of Elasticity, E, is the slope of initial linear elastic region in the stress-strain
curve as shown in Figure 2.4. Modulus of elasticity is used primarily in calculating elastic
stiffness and in establishing elastic buckling capacity. In all design and analysis cases, a constant
value of 29,000 ksi is used for modulus of elasticity of steel.
Yield Stress, F y , is defined as the stress at which steel begins to deform inelastically
without increase in the applied stresses, see “yield Plateau” in Figure 2.2(a). The value of the
stress corresponding to the yield plateau, see Figure 2.4(a), is Yield Stress F y. Some steels do not
have a well-defined yield plateau. A typical stress-strain curve for
CHAPTER 2- MATERIAL 2-5

Yield Strain Hardening

Stress Point
Stress
Start of Necking
Ultimate
Yield Ultimate Start of
Strength
Plateau Strength Necking
Fu
Fu
Rupture
Fy Rupture
Fy

E
Yield E
Yield Strength
Strength Elastic
1 Modulus
1 Elastic
Modulus

εy εs εu εr Strain
εy εs ε ε Strain
ε=0.002
u r

(a) (b)

Figure 2.4. Mechanical Properties obtained from Stress-Strain Curves

these steels are shown in Figure 2.4(b). In these cases, as shown in Figure 2.4(b) a line having a
slope equal to Modulus of Elasticity is drawn from the point of zero stress and 0.002 strain, see
Figure 2.4(b). The intersection of this line with the stress-strain curve is the yield point as shown
in Figure 2.4(b). The stress at this point is called yield strength instead of yield stress. In this
book, we will use yield stress for both cases of yielding with or without yield plateau. This is in
accordance with the AISC Specifications (AISC, 2010a).
Ultimate Strength, F u , is the maximum stress that a steel tolerates before starting to neck
and rupture, Figures 2.4(a) and (b). Ultimate Strain, ε u is the strain corresponding to the point of
Ultimate strength, see Figure 2.2(a), and (b). Maximum Strain ε r is the maximum “permanent
strain” that remains in a steel tension coupon after it fractures, see Figures 2.4(a), and (b).
Maximum strain is established by measuring the permanent elongation of the tension coupon
during the test and just before it fractures and by dividing it by the initial length of the coupon
before test. A simpler method, which is also acceptable by ASTM, is after the test is done and
coupon has fractured to place the two pieces of steel coupon together and measure the total
elongation and divide the total elongation by the original length of the tested specimen.

2.4. TYPES OF STRUCTURAL STEEL

Figure 2.5 shows stress-strain behavior of common types of structural steel used in
buildings and other structures in the United States. The designation of steel as “A123” refers to
the American Society for Testing and Materials (ASTM) Standard number assigned to a particular
______________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
2-6

steel. For example, the ASTM-A514 Standard provides information on mechanical properties,
chemical composition, and other properties of A514 heat-treated alloy steel.
In the following, for each type of steel shown in Figure 2.5 below a summary of its
characteristics and use in the steel structures is provided.
A36 Steel - This steel has a minimum specified yield stress of 36 ksi and an ultimate tensile
strength of 58 to 80 ksi. In design and analysis of steel structures, the minimum value of 58 ksi for
ultimate strength of A36 steel is used. Until the 1994 Northridge earthquake, this steel was one of
the most commonly used types of steel in structures. It was used in the form of structural shapes
and plates. However, during the Northridge earthquake, welded moment connections in a number
of steel moment frames in Los Angeles developed cracks in the connection areas. Figure 2.6
shows an example of such cracks in a moment connection. In the aftermath of the 1994
Northridge earthquake, research indicated that the use of A36 steel in wide flange (W) and Tee
(WT) sections should be discouraged. Instead of using A36 steel W or WT shapes, the steel
industry has developed a new type of steel, ASTM-A992, as the ”preferred” steel for wide flange
shapes. For other shapes such as M, S, C (channels), MC, and L (angles) and all thicknesses of
plates, A36 steel is still the preferred type of structural steel.

A514
-Heat-Treated (Quenched and Tempered) Alloy Steel

Stress σ (ksi) A852

Heat-Treated (Quenched and Tempered) Low Alloy Steel

100 A852
-Heat-Treated High Strength Low Alloy Steel

A242, A572, A588, A618, A847, A913, A992

-High Strength Low Alloy Steels

A36, A53, A500, 501, A529

-Carbon Steels

0
0.0 0.10 0.20 0.30 Strainε, (in/in)

Figure 2.5. Typical Stress-Strain Curves for Steel used in Structures

A53-Grade B- This steel has a minimum specified yield stress of 35 ksi and specified
ultimate tensile strength of 60 ksi. This steel is used in seamless and welded black and hot-dipped
galvanized carbon steel pipes. The nominal sizes of these pipes are up to 26” in diameter. The
nominal wall thicknesses are from 0.068” to 2.344”. The chemical composition and mechanical
properties of this steel for Types E and S, Grade B are very similar to ASTM A36.
CHAPTER 2- MATERIAL 2-7

Welded Moment
Connection
Crack

Figure 2.6. Welded Moment Connection Fractured During 1994 Northridge, California Earthquake

A500-Grade B- This steel is primarily used in producing cold-formed welded and seamless
carbon steel hollow rectangular or round shapes. The chemical composition of this carbon steel is
similar to A36. When used in producing round shape, it has a specified minimum yield stress of
42 ksi and ultimate tensile strength of 58 ksi. When used to produce rectangular shapes, it has a
specified minimum yield stress of 46 ksi and ultimate tensile strength of 58 ksi.
A500-Grade C- This steel, similar to A500-Grade B, is primarily used in producing cold-
formed hollow rectangular or round shapes. However, as Table 2.1 indicates, this steel is not the
preferred steel for pipes and hollow rectangular shapes. Instead, AISC Table 2.1 indicates that the
above A500-Grade B is the preferred steel for these hollow shapes. When used in producing
round shape, it has a specified minimum yield stress of 46 ksi and ultimate tensile strength of 62
ksi. When used to produce rectangular shapes, it has a specified minimum yield stress of 50 ksi
and ultimate tensile strength of 62 ksi. .
A501- This carbon steel is used in production of cold-formed welded and seamless hollow
rectangular and round cross sections. The round pipes are produced in diameters from ½’ to 24
inches. As Table 2.1 indicates, this steel is not the AISC “preferred” steel for these hollow shapes
and are rarely specified for use in steel structures. Its chemical composition is similar to A36 steel
and its minimum yield stress and ultimate tensile strength are 36 and 58 ksi respectively.
A529 Grade 50 and Grade 55 - this steel is used in producing rolled shapes such as W, M,
S, HP, C, MC and L as well as plates up to thickness of 2.5 inches. This steel is not the AISC
“preferred” steel for these rolled shapes or plates. The minimum yield stress and ultimate tensile
strength for A529 Grade 50 are 50 ksi and 70-100 ksi. For A529 Grade 55, the yield stress is 55
ksi and the ultimate tensile strength is 70-100 ksi.
A572 - This steel has five grades, 42, 50, 55, 60, and 65 ksi, which indicate the minimum yield
stress of each grade. Structural shapes are rolled in all grades as Table 2.1 shows, however, the
most common grade used in steel structures has been A572 Grade 50 with minimum yield stress of
50 ksi and ultimate tensile strength of 65 ksi. Since the 1994 Northridge earthquake, the wide

flange sections made of A572 Grade 50 steel are being replaced with A992 steel (described
below).
A618 - This high-strength, low alloy steel is used in production of hollow rectangular and round
shapes, however, it is not the AISC “preferred” steel for these shapes. Instead, according to Table
2.1, the preferred steel for hollow rectangular and round shapes is A500-Grade B.
A913 - This steel is a relatively new addition to the family of quenched and tempered high-
strength low-alloy steels. This steel has four grades, 50, 60, 65 and 70. The grade numbers
represent the minimum yield stress of the grade. The ultimate tensile strength values for these four
grades are 60, 75, 80, and 90 respectively. The steel is intended for welded, riveted, and bolted
construction of buildings, bridges and other structures. The ASTM-A913 Specification (1993)
indicates that due to inherent characteristics of quenched and self-tempered process, the shapes
made of this steel should not be heated beyond 1100 degrees Fahrenheit during forming or
welding. The welding of this steel should also be done using a welding process suitable for the
grade of steel.
A992 - During 1990’s, and particularly after the 1994 Northridge earthquake, an attempt was
made to develop steel that would have more controlled values for the maximum yield stress as
well as for the ratio of yield stress to ultimate tensile strength. The A992 steel is currently the only
type of steel that has controlled value of the ratio of ultimate strength to yield stress. The A992
steel currently is used only for production of wide flange shapes and is the AISC “preferred’
material for wide flanges. The main characteristics of A992 steel is better definition of a
maximum value for yield stress. In all cases of steel discussed above or below, only the
“minimum” yield stress is given and there is no indication on what is the maximum value of yield
stress.
In most applications where strength of the component by itself is the concern, minimum yield
stress is what we need to use to establish the strength. However, in some cases, such as
connections, the yield strength of one component can govern the amount of force transferred to the
neighboring component. A good example is the case of welded moment connections, shown in
Figure 2.6 earlier and their unacceptable performance during the 1994 Northridge earthquake.
The A992 has an upper limit of 65 ksi on yield stress, a minimum tensile strength of 65 ksi, a
specified maximum yield stress to tensile strength ratio of 0.85 and a specified maximum carbon
equivalent of 0.47%. With the availability of A992, the steel industry currently encourages the use
of this steel in wide flanges instead of A36 steel. Certainly, for seismic regions and in applications
such as blast conditions, where inelastic behavior can be reached, the use of this steel is
recommended over other common steels such as A36 and A572. Table 2.4 shows a comparison of
chemical and mechanical properties of A572 Grade 50 and A992 steels.
A242 - This steel is a corrosion-resistant, high strength low alloy steel with minimum yield stress
of 42, 46 and 50 and ultimate strength of 63, 67 and 70 respectively. Grade 42 and 46 are
primarily used in rolled wide flanges while Grade 50 is used in rolling all shapes and plates up to
and including 4-inch thickness. For plates, the Grade depends on thickness of the plate.
A588 - Similar to A242 above, this steel is also a corrosion-resistant high strength low alloy steel
with specified minimum yield stress of 42, 46, and 50 ksi and ultimate tensile strength of 63, 67
and 70 ksi respectively. This steel is a “weathering” steel where under proper atmospheric
conditions, a layer of oxide forms on the surface, which tightly adheres to the steel and protects it
from further corrosion. Preece and Collin (1993) and Hassett (2003) state that the corrosion
resistance of this steel is twice that of carbon steel with copper.
CHAPTER 2- MATERIAL 2-9

A847- This is corrosion-resistant high-strength, low alloy steel, and it is used in production of
hollow rectangular and round shapes. However, it is not the AISC “preferred” steel for these
shapes. Instead, the preferred steel for hollow rectangular and round shapes is A500-Grade B.
A514 - This is a quenched and tempered alloy high strength steel in two grades with yield stress
of 90, 100 ksi, and ultimate strength of 100-130 and 110-130 respectively. This steel is currently
available only as plates, up to, and including 6 inches of thickness. Plates up to 2.5-inch thickness
are grade 100 and plates with thickness of between 2.5 and 6 inches are grade 90 ksi. According
to Preece and Collin (1993) the tempering of this steel at the mill is performed at a rather high
temperature of 1200-1250 degrees Fahrenheit. They also state that any exposure to temperatures
like these, for example during welding, will reduce the strength. In addition, overheating can
change the crystalline structure and turn the steel to stronger but less ductile form. To avoid these
problems, controlled welding should be used for this steel. Since this steel is only available in
plate forms, its use in buildings, where rolled shapes are common, is quite limited. Quite often
this steel is used in tension areas of bridges such as tension members of the trusses or tension
flanges of hybrid plate girders. The 1958 Carquinez Bridge in California was one of the first
bridges where this steel was used in its top tension chords. As for the building applications, this
steel was used in some of the columns of the now-destroyed World Trade Center towers.

2.2. c. Residual Stresses

Residual stresses are defined as those stresses that are present before load application or
after load removal. These stresses are self-balancing; therefore, any compressive residual stress is
associated with some balancing tensile residual stresses. Figure 2.5 shows examples of residual
stresses in a number of steel sections. In most cases, steel members and connections in a structure
have residual stresses, strains, or both. Residual stresses are mainly due to cooling process,
straightening or cold working. Welding can also cause significant residual stresses in metal
structures.

+ + +

+
-
-
+
+

- -
+ +
+

Plates Hot-Rolled Wide Flanges Welded Built-up Sections

Fr ≈ 5 ksi Fr ≈ 10 ksi Fr ≈ 16.5 ksi

Figure 2.5. Values of Fr, Residual Stresses Suggested in the Literature for Hot-Rolled Shapes

Presence of residual stresses has no effect on failure modes due to yielding or fracture of
the steel. However, residual stresses can reduce the capacity of steel member or connection if the
governing failure mode is local or overall buckling of the component. In Chapter 5, when we
discuss compression members (i.e. columns), we will show how and why residual stresses reduce
the compressive capacity of steel columns. In Chapter 6, when we discuss flexural members (i.e.
beams), again will show that, similar to columns; residual stresses reduce bending capacity of the
steel beams as well.

2.2. d. Coefficient of Thermal Expansion

Steel expands when its temperature rises and contracts when temperature drops. Coefficient of
thermal expansion is the strain created in the material due to one degree of temperature change.
The coefficient of thermal expansion, C th-ex , for steel is 6.5x10-6 (AISC, 2011).

2.2. e. Poisson’s Ratio

Poisson’s ratio is another mechanical property used in design. It is the ratio of transverse strain to
longitudinal strain when material is under axial load. Poisson ratio of steel when elastic is 0.30.
When steel is fully plastic, such as when it is on yield plateau, its Poisson’s ratio is 0.50.

2.2. f. Corrosion Resistance

Another property of steel that is important in design of steel structures is the corrosion resistance.
Corrosion causes loss of steel and reduction of the thickness of the material, which in turn results
in increase in the internal stresses. If corrosion is severe, the designer needs to consider extra
material that will be lost to corrosion or better yet to select corrosion resistant steels.

2.4. SELECTION OF TYPE OF STEEL

Pages 2-48, 2-49 and 2-50 of the 14th Edition of the AISC Steel Construction Manual (AISC,
2011), has tables that show yield stress, F y , and ultimate strength, F u of the steel used in steel
structures. Figure 2.4 earlier showed the stress-strain curves for these steels. The AISC tables
also provide useful information on the availability of steel in various shapes as well as which type
of steel is the “preferred” steel for a given cross sectional shape. For example, the preferred type
of steel for hot-rolled wide flanges is A992; Grade 50 Table 2.1 below provides a summary of the
“preferred” steel for various shapes.
Steel mills produce steel shapes and plates used in the structures. When a batch of steel is
produced in steel mills, coupon tests are done on the steel in the mill and mechanical properties as
well as chemical composition of the particular batch called a “heat” are recorded in a special form
called Report of Chemical/Physical Tests or “Mill Certificate.” A sample of Mill Certificate is
shown in Figure 2.6 below. The information in the form includes Heat Number (see third column
from the left in the form), mechanical properties such as yield stress, ultimate strength, and percent
of elongation, (see columns 5, 6 and 7 from the left). In addition, Mill Certificates show chemical
composition of the steel (see bottom of the form in Figure 2.6).
CHAPTER 2- MATERIAL 2-11

Table 2.1. Preferred Type of Steel for Various Cross Sections

Cross Sections Preferred Steel Type Yield Stress Ultimate Strength

F y , (ksi) F u , (ksi)

M, S, MC and C channels and angles 58-80

A36 36 (Use 58 ksi in design)

M, S MC, C L

Plates and bars up to 8-inch thickness 58-80

A36 36 (Use 58 ksi in design)

Plates Bar

Plates and bars more than 8-inch thick 58-80

A36 32 (Use 58 ksi in design)

Plates Bar

Pipes (not Round HSS Shapes)

A53 Gr. B 35 60
Pipe

Round HSS Shapes (not Pipes)

A500 Gr. B 42 58

Round HSS

Round HSS Shapes (not Pipes)

A500 Gr. B 46 58

Rectangular HSS

HP Pile Section

A572 Gr. 50 50 65
HP

Wide Flanges and T-Sections

A992 50 65

W WT

Figure 2.6. A Sample of a “Mill Certificate” Providing Data on Mechanical Properties and Chemical
Composition of Steel

Any piece of hot-rolled steel, such as wide flanges, channels and angles, is sent to
fabricating shop, the steel is accompanied by its mill certificate. After steel pieces are fabricated,
they are stamped with the heat number indicated in the mill certificated and the steel pieces are
sent to the construction site with copies of the mill certificates. The contractor keeps copies of the
mill certificates and sends copies to the owner’s representative (usually the structural engineer of
the record or the inspection agency). The inspectors ensure that the fabricator has used the correct
type of steel as specified by the structural engineer.

References for Chapter 2

1. AISC, (2010a) “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An American
National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2011) “Steel Construction Manual,” 14th Edition, American Institute of Steel Construction,
Chicago.
3. Hassett, P. (2003) “Structural Steel Construction in the New Millenium,” Steel Technical Information
and Products (Steel TIPS) Report, Structural Steel educational Council, Moraga, CA.
4. Preece, F.R and Collin, A.L., (1993) “Structural Steel Construction in the 90’s,” Steel Technical
Information and Products (Steel TIPS) Report, Structural Steel educational Council, Moraga, CA.
5. Tall, L., Editor, (1964). "Structural Steel Design”, John Wiley.
CHAPTER 3-DESIGN LOADS 3-1

CHAPTER 3

INTRODUCTION TO DESIGN LOADS

APPLIED TO STRUCTURES

Seismic Hazard Map of U.S. (Source: USGS)

Objectives:
This chapter is not part of the CE122N syllabus. The loads and
load combinations are covered in CE122L. The main objective
of this chapter is to introduce you to structural loads and their
combinations.

3.1. INTRODUCTION
3.1. INTRODUCTION
In order to protect public safety and have uniformity in design, all buildings are designed
following a standard specification or a code approved by the building officials for the
jurisdiction where building is located. Several codes such as International Building Code
(ICC,2012) and Life Safety Code® (NFPA, 2012) and the ASCE-7 Standard (ASCE-7,
2010) are common in the United States. These codes and standards hsve provisions on
minimum loads that buildings and other structures should be designed for. In the last few
years, the ASCE-7 Standard (ASCE-7, 2010) has been gaining more and more acceptance
within the structrual engineering community and with the building officials. In this
chapter, we will discuss only the load combinations used in design of buildings and
structures as given by the “ASCE-7 Standard: Minimum Design Loads for Buildings and
Other Structures (ASCE-7, 2010.)

3.1.a Scope of the ASCE-7 Standard

The ASCE-7 Standard provides minimum load requirements for the design of buildings
and other structures. It provides nominal values of design loads, load factors and load

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not
be distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
3-2

combinations to be used in both the allowable stress design (ASD) and the load and
resistance factor design (LRFD) methods.

3.1.b Basic Requiements in the ASCE-7 Standard

In its Chapter 1, the ASCE-7 Standard states that in design of structures the following
requirements should be considered:
1. Strength requirement: building and other structures should be designed and
constructed to support safely the loads given in the standard without
exceeding the appropriate strength limit state. This is, in effect, stating that
the design equation discussed in Chapter 1 and given as “Load Effects ≤
Resistance” should be satisfied in design.
2. Serviceability requirement: structural systems and their components, such as
members and connections, should be designed to have adequate stiffness to
limit deflections, lateral drift, vibration or other serviceability functions to
certain tolerable levels. The serviceability problems are not usually a safety
concern but a matter of comfort and functionality of the building.
3. Self-restraining forces: In addition to strength and serviceability
requirements, the ASCE-7 Standard requires that self-restraining forces
resulting from differential settlements of foundations, temperature changes,
shrinkage and creep and similar effects be considered in design and
provisions be provided for such forces. ASCE-7 Standard does not provide
specific values or provisions for these self-restraining forces or how one can
establish them. For steel components, shrinkage and creep is not a design
issue. However, the effects of uneven settlements of foundations can be very
serious for any structure, particularly in multi-story buildings. The problem
should be investigated and resolved by the combined efforts of the structural
engineering and geotechnical engineering teams. The effects of change of
temperature should be considered in design, especially for relatively long
strures or when steel structure is used in “exposed” state without being
covered by non-structural elements.
4. Analysis requirements: The ASCE-7 Standard provides only general
informaton on modeling and analysis requirements. However, it requires that
the method of structural analysis used “should take into account equilibrium,
general stability, geometric compatibility, and both short- and long-term
material properties (excerpt from ASCE-7,2005)”. It also requires that in
members with repeated loads which can accumulate residual deformations,
the eccentricites expected to occur during the life of the member should be
included in the analysis.
5. Counteracting Strctural Action: the ASCE-7 Standard requires that all
structural systems and components as well as non-structural systems and
components (such as cladding) should be designed to resist forces due to wind
and earthquake with consideration of overturning, sliding and uplift. It also
requires that for these lateral loads (wind and earthquake) continueous load
paths should be provided to transmit these lateral forces to the foundations.
CHAPTER 3-DESIGN LOADS 3-3

3.1.d. Loads included in the ASCE-7 Standard

The ASCE-7 Standard provides information and specific nominal values for thirteen
different environmental loads. These loads are listed in Table 3.2 below.
Table 3.2. Loads Included in the ASCE-7 Standard (ASCE-7, 2005)
Load Symbol
Load
Dead load D
Weight of ice Di
Earthquake E
Fluid load (for fluids with well-defined pressures and maximum heights F
Flood Fa
Load due to lateral earth pressure, ground water pressure, or pressure of bulk material H
Live load L
Roof live load Lr
Rain load R
Snow load S
Self-straining force T
Wind load W
Wind-on-ice load (determined in accordance with Chapter 10 of the ASCE-7 Standard Wi

3.2. COMBINATION OF LOADS

ASCE-7 Standard has load combinations for both ASD and LRFD formats. The load
combinations are given in three categories of Basic Combinations, Load Combinations
Including Flood Loads and Load Combinations Including Atmospheric Ice Loads. Table
3.1 below summarizes only the Basic Combinations of loads given for the LRFD methods
in the ASCE-7 Standards. For other two combinations involving flood and ice, the reader is
referred to the ASCE-7 Standard (ASCE, 2010).

Table 3.1. Basic Load Combinations in the ASCE-7 Standard (ASCE-7, 2010)

Basic Load Combinations:

1. 1.4(D + F)
2. 1.2(D + F + T) + 1.6(L + H) +0.5(L r or S or R)
3. 1.2D + 1.6(L r or S or R) + (L or 0.8W)
4. 1.2D + 1.6W + L + 0.5(L r or S or R)
5. 1.2D + 1.0E + L + 0.25S
6. 0.9D + 1.6W + 1.6H
7. 0.9D + 1.0E +1.6H
Notes:
1. In combinations 3, 4 and 5 the load factor on L is permitted to be equal to 0.5
if nominal live load is less than or equal to 100 psf with the exception of
garages and areas of public assembly.
2. In combinations 6 and 7 load factor for H should be zero if H counteracts with W or E.
3. In combinations 2, 4 and 5 the snow load S should be taken as the flat roof
snow load or the sloped roof snow load.

For structures that are not subject to flood or atmospheric ice loads, instead of load
combinations 5, 6 and 7 in Table 3.1 that involve E, new load combinations, denoted as
5*, 6* and 7* , given in Tables 3.2 below for LRFD should be used.

Table 3.2. Basic Load Combinations,

Including Seismic Effects, in the ASCE-7 Standard (ASCE, 2005)

5*. (1.0+0.14S DS ) D+ H + F + 0.7ρQ E

6*. D + H + F + 0.75 (W or 0.7E) + 0.75L 0.75(L r or S or R)

7*. (0.6-0.14S DS )D + 0.7ρQ E +H

8. 0.6D + 0.7E + H

References for Chapter 3.

1. ASCE, (2010). “Minimum Design Loads for Buildings and Other Structures.” SEI/ASCE
7-10, American Society of Civil Engineers ( www.asce.org ).
2. ICC, (2012). “International Building Code.”, International Code Council
(www.iccsafe.org)
CHAPTER 4 - TENSION MEMBERS 4-1

CHAPTER 4

TENSION MEMBERS

Objectives:
In this chapter, you will learn:
1. Definition of tension members
2. Tension members in typical steel structures
3. Cross sections commonly used in tension members
4. Basic categories of tension members
5. Behavior of a typical tension member
6. Failure modes (or limit states) of steel tension members
7. Establishing tensile strength of a given tension member
8. Designing a tension member to carry a given tension load
9. Application of concepts and equations to actual design of tension
members using numerical examples provided throughout the
chapter.

4.1. INTRODUCTION
Tension members are those members of a structure that primarily resist tension forces acting
through their centroid. If the member is subjected to tension combined with bending, shear, or
torsion, it should be designed as a member under combined forces discussed in Chapter 8.
Steel tension members are commonly used in trusses, braced frames and other
applications, when the primary force in the member is tension. Figure 4.1 shows examples of
structures with tension members in them.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-2

Tension Members

Tension Members A. Astaneh

(b) - Braced Frames

(a)-Trusses

Tension
Tension Members
Members

Photo: A. Astaneh
http://4.bp.blogspot.com
(c) –Transmission Towers (d)- Braced Towers

Chords and other

Truss Members

(e)- Trusses in Bridges

Photo by A. Astaneh

Stay Cables

(f) – Cables in Cable-Stayed Bridges

Main Cables

Suspenders

(g) – Cables in Suspension Bridges

Photo by A. Astaneh

Figure 4.1. Examples of Structures with Tension Members

CHAPTER 4 - TENSION MEMBERS 4-3

4.2. CATEGORIES OF TENSION MEMBERS

(Ref.: AISC Specification, (AISC, 2010a), Chapter D, Page 16.1-26))
The AISC Specification (2010a) divides steel tension members into following five categories:
1. Rolled shapes, tubes, pipes, plates, bars and built-up shapes made of these
sections, Figure 4.2.
2. Pin-connected members, Figure 4.3(a).
3. Eye-bars, Figure 4.3(b).
4. Cables, Figure 4.3(c).
5. Threaded rods, Figure 4.3(d).
In this chapter, we focus on the first category of tension members, shown in Figure 4.2(a)
through 4.2(i). These are the most common tension members. The AISC Specification (AISC,
2010) has specific provisions for design of Pin-connected Members, Eye-bars, Cables, and
Threaded Rods. Design of these less common tension members are not part of this chapter.

(a) Wide flange (b) Channel (c) Angle (d) Tee (e) Plate

(e) Rectangular Tube (f) Pipe (g) Double Angles (h) Double Channels (i) Built-up Members

Figure 4.2. Examples of Common Steel Tension Members Covered in This Chapter

(a) Pin-Connected Member (b) Eye-bar (c) Cables (d) Threaded Rods

Figure 4.3. Other Steel Tension Members

4.3. GENERAL ASSUMPTIONS MADE IN DESIGN OF

STEEL TENSION MEMBERS
(Ref.: AISC Specification, (AISC, 2010a), Chapter D, Page 16.1-26))
In developing design procedures for steel tension members, the following assumptions have been
made:
a. Members are prismatic, which means they have the same cross section throughout
their length.
b. Only axial tension force is present and is acting through the centroid of the cross
section.
c. The tension force is applied gradually and can be considered static load.
d. No cyclic fatigue stresses are present.

4.4. BEHAVIOR OF STEEL TENSION MEMBERS

4.4. a. Stress Distribution in a Tension Member
When a steel member is subjected to tension force acting through its centroid, the applied tension
force creates uniform tensile stresses on the cross section of the member, see Figure 4.4(a).

(a) Uniform tensile stresses applied to the member

(b) Residual stresses present inside the member

before any load is applied.

σmax

+ =

(c). Total stresses due to summation of residual and applied stresses.

Figure 4.4 Typical Stresses Acting on Cross Sections of a Tension Member

CHAPTER 4 - TENSION MEMBERS 4-5

Net Area

Figure 4.5. Typical Stresses Acting on the Net Section of a Tension Member

In addition to these uniform stresses, as discussed in Chapter 2 (Page 2-10), most steel
members have residual stresses in them. These stresses are created during the hot-rolling or cold-
forming processes. Figure 4.4(b) shows an example of residual stresses acting on the cross
sections of a tension member.
The actual steel members are not usually perfectly straight. The out-of-straightness of the
member also creates some secondary bending stresses in the member. Even though we assume
that the tension force acts through the centroid, in actual structures, the end connections introduce
some eccentricity subjecting the member to axial tension as well as some bending moment.
Because of the presence of residual stresses, out-of-straightness and the eccentricities introduced
by the end connections, the total stress acting on the cross sections of a tension member can be
non-uniform as shown in Figure 4.4(c).
In many cases, tension members are connected to the support by bolts. Prior to the 1940’s
most steel structures had riveted connections. Both bolts and rivets are installed in the bolthole or
rivet holes. In these cases, the cross section of the member at the first row of bolt or rivet in the
connection becomes a critical section. Because of the bolt or rivet holes, the area of this cross
section is reduced. The actual area of cross section is the total area minus the area taken out by the
bolt or rivet holes. The remaining area called the net area develops stresses that are larger than the
stresses on the gross section of the member. In addition, because of the presence of the bolt or
rivet holes, stress concentrations will occur on the net section where the holes are located. A
typical distribution of tensile stresses on the net area is shown in Figure 4.5 above. In the
literature, the net area is also called the net section.

4.4. b. Behavior of Typical Steel Tension Members

Figure 4.6 shows behavior of a typical tension member in terms of applied tensile force versus
elongation of the member. Initially, from Point “O” to Point “A” in Figure 4.6, behavior of the
member is elastic. This means that if the tensile force is removed the member will return to its
original length.
As the tensile force increases, the stresses inside the member, acting on the gross sections
and net sections, shown in Figures 4.4(a) and 4.5 respectively, also increase. At one point, the
maximum stress reaches yield stress of the material and the fiber that has the maximum stress
starts yielding. At this point, shown as Point A in Figure 4.6, the elastic behavior ends, and the
member enters the inelastic range of behavior.

Tension Inelastic behavior

Initial Necking starts here
Force
elastic
behavior Fracture
Tilt
E
D F
Yield
B
C

A
Behavior under
both unloading
and reloading is
elastic

O D’ Elongation
δmax

Permanent elongation
is due to inelastic deformation

Figure 4.6. Typical Behavior of Tension Members

As the tensile load increases, more fibers yield and the force-elongation curve becomes
almost horizontal (Point B in Figure 4.6). In some steels, especially high strength steel produced
by quenching and tempering process such as A514 (see Page 2-9 in Chapter 2), where the material
does not have a clear yield plateau, the tensile load-elongation curve does not show a pronounced
horizontal yield plateau. After tension member develops significant yielding, Point “B” in Figure
4.6, it starts strain hardening (starting at Point C in Figure 4.6).
During the inelastic range (beyond Point A in Figure 4.6), if the tension force is reduced,
the member again becomes elastic as shown in Figure 4.6 with unloading line DD’. Notice that
when member is unloaded during the inelastic range of behavior, it does not return to its original
length. Instead, a permanent elongation remains in the member as shown in Figure 4.6 with
elongation OD’. If after unloading, the member is reloaded in tension, starting at Point D in
Figure 4.6, it behaves elastically until it reaches the point of unloading, Point D in Figure 4.6 until
it reaches the original unloading point (Point D in Figure 4.6). Beyond Point D, the member
behaves in inelastic manner again, Line DE in Figure 4.6. At Point E, the necking of the cross
section starts resulting in drop of the strength from a maximum at Point E to a somewhat reduced
value at Point F in Figure 4.6. At Point F, the tension member reaches its ultimate elongation
capacity and fractures.
The behavior shown in Figure 4.6 above, where significant yielding and inelastic
elongation occurs before the member fractures is called ductile behavior and the member is called
a ductile member.
In addition to tensile yielding in the gross section, tension members have two other failure
modes, also called “limit states,” that need to be considered in design. These are rupture in the net
section and block shear rupture. More information on limit states of tension members is given in
the following section.
CHAPTER 4 - TENSION MEMBERS 4-7

4.5. LIMIT STATES OF TENSION MEMBERS

The limit states of failure for tension members are:
a. Tensile yielding in the gross section, Figures 4.7(a) and 4.8(a)
b. Tensile rupture in the net section, Figures 4.7(b) and 4.8(b)
c. Block shear rupture, Figure 4.7(c) and 4.8(c)
d. Failure of connections, Figure 4.7(d)
The above limit states are shown schematically in Figure 4.7. Figure 4.8 shows actual
failure of test specimens for limit states of yielding in the gross section, rupture in the net section,
and block shear rupture.

Connection
Tensile Rupture Block Shear Failure
Tensile Failure
Yielding

Typical (a) Tensile Yielding (b) Tensile Rupture

Tension in the Gross in the Net Section
Section (c) Block Shear Rupture (d) Failure of
Member
Connections
Figure 4.7. Limit States of Strength in Tension Members

Behavior of tension members when limit state of tensile yielding in the gross
section governs was discussed in previous section and shown in Figure 4.6. A typical tension
force versus elongation curve for limit state of yielding in the gross section is again shown by
Curve A in Figure 4.9. When a tensile member yields, it loses much of its stiffness, but maintains
its strength while continuing to elongate. However, in some cases, especially in tension members
with riveted and bolted connections, it is possible that as the member is loaded in tension, a net
section, which because of holes in it is smaller than the gross section, reaches its maximum
strength and ruptures, Figure 4.7(b). When rupture in the net section occurs, the member loses not
only its stiffness but also its tensile strength. Typical behavior of a tension member when rupture
in the net section is the governing limit state is shown by Curve B in Figure 4.9. If rupture in the
net section happens, usually, the behavior is quite brittle with member showing relatively small
amount of inelastic elongation before it ruptures as shown by Curve B in Figure 4.9.

a. Tensile Yielding
in the Gross
b. Tensile Rupture in
Sections
the Net Section

c. Block shear
rupture

Figure 4.8. Three Limit States of Tension Members

Rupture in the net section is an acceptable failure mode in design for gravity, wind, snow
and other loads. However, rupture in the net section is not acceptable to be the governing failure
mode of tension members used in Special lateral force resisting systems in seismic design. As we
will discuss in Chapters 16 and 17 later, for these systems ductile yielding of the gross section
should be the governing failure mode and not the relatively brittle mode of rupture in the net
section. In addition, in blast-resistant design, which is becoming part of design process in some
cases, if a tension member is a participant in redistribution of gravity load in the event of a blast,
such a tension member needs to be ductile and should not have rupture in the net section as its
governing failure mode.
CHAPTER 4 - TENSION MEMBERS 4-9

Another limit state of steel tension members is Block shear rupture, Figures 4.7(c), and
4.8(c). This failure mode occurs in the connection areas of tension members due to a combination
of tensile fracture and shear yielding or tensile and shear rupture. Curve C in Figure 4.9 shows
typical behavior of a tension member when limit state of block shear rupture governs. Similar to
rupture in the net section, the behavior of tension member when block shear rupture governs is
also relatively brittle (Curve C in Figure 4.9) compared to the very ductile behavior when yielding
in the gross section governs (Curve A in Figure 4.9).
Finally, it is possible that the end connections that connect the tension member to the rest
of the structure fail before the yielding in the gross section, rupture in the net section, or block
shear rupture. Design of connections is not part of this chapter and will be covered in later
chapters. For the remainder of this chapter we will continue focusing on the three failure modes or
limit states of yielding in the gross section, rupture in the net section, and block shear rupture.

Tension Curve A- Ductile behavior when yielding

Force in the gross section governs

Curve B- Brittle behavior when rupture in

the net section governs

Curve C- Relatively brittle behavior when block

shear rupture governs

O Elongation

Figure 4.9. Typical Behavior of Tension Members Depending on Which Failure Mode Governs

4.6. SLENDERNESS LIMITATION FOR TENSION

MEMBERS
(Ref.: AISC Specification, (AISC, 2010a), Sec. D1, Page 16.1-26))
In the past, there was a design requirement that the slenderness ratio, L/R, of the tension members
not to exceed 300. However, in the current AISC Specification (AISC, 2010a) this is not a
requirement but only a preferred limit. It means that it is better to satisfy this limit but if you do
not satisfy it, you have not violated the specification. The terms L and r are the length of the
member and radius of gyration of the cross section respectively.
The preferred limit of L/R ≤ 300 for tension members is there to make the members stiff
enough so that they do not sag under their own weight. In addition, having a certain minimum
stiffness makes tension members easy to handle during the fabrication and erection of steel
structures.

Numerical Example 4.1: Slenderness of Tension Members

Check slenderness of the tension member shown below to find out whether or not it satisfies the preferred
slenderness limit of L /r ≤ 300.

Welds
HSS8x6x5/8
y

x x

20’-0”
y

Center to Center of the

Connections

Given:
1. Cross section of the member: HSS 8x6x5/8
2. Length of the member: 20 feet.

Required:
To check L /r ≤ 300 and to find out if it is satisfied or not.

Solution:
First, we need to establish minimum radius of gyration of the cross section. For an HSS8x6x5/8 section the
minimum radius of gyration, r, is with respect to its y-y axis (see sketch above) and is given as 2.27 inches
in the cross section properties tables. The length of the member, measured as the center-to-center of its
end connections, is given as 20 feet. Therefore:
L / r = (20ft) (12 in/ft) / (2.27 in.) = 106 < 300 O.K.
In this case, the preferred limit for slenderness of tension members is satisfied.

4.7. DESIGN OF TENSION MEMBERS

In design of tension members, all four failure modes, or limit states of (1) yielding in the gross
section, (2) rupture in the net section, (3) block shear rupture and (4) failure of connections should
be considered. In the following three sections, limit states (1), (2) and (3), shown in Figure 4.8, are
discussed. Limit state (4), failure of connections is discussed in later chapters on connections.

4.8. LIMIT STATE OF TENSILE YIELDING IN

THE GROSS SECTION
(Ref: AISC Specification, (AISC, 2010a), Sec. D2, Page 16.1-26))
In design, it is assumed that the distribution of tensile stress on the gross section of the member is
uniform, Figure 4.10. When the uniform stress, acting on the entire cross section, reaches yield
stress of steel, F y , the member has reached limit state of yielding in the gross section, and the
entire length of the member outside the connection areas has yielded, see Figure 4.8(a).
CHAPTER 4 - TENSION MEMBERS 4-11

Applied Tension = Tu (in LRFD)

= T (in ASD)

Gross Area= Ag

Uniform Stress= Fy

Figure 4.10. Yielding in the Gross Section under Uniform Stress, Fy

The equation of design for checking failure mode of tensile yielding in the gross section
according to the AISC Specification (AISC, 2010, Sec. D2, Page 16.1-26) is:

Tu ≤ φt Pn (in LRFD) and T ≤ P n /Ω t (in ASD) (4.1)

Where
T u = required axial tensile strength in LRFD, same as the factored applied tension force.
φ t = resistance factor for tensile yielding in LRFD = 0.90
P n = available nominal axial yield strength of the gross section of the member given by:

Pn = Fy Ag (4.2)

A g = gross area equal to the total cross-sectional area of the member.

F y = specified minimum yield stress.
T = required allowable axial tensile strength in ASD, same as the un-factored (service) applied
tension force.
Ω t = safety factor for tensile yielding in ASD = 1.67

Numerical Example 4.2: Evaluating Strength Based on Tensile

Yielding in the Gross Section
Using the LRFD method and considering only failure mode of Yielding in the Gross Section, calculate
maximum factored applied load (the required axial tensile strength),T u , that can be applied to the double
channel member shown below.

2C10x30, A36 Steel

Given:
1. Cross section of the member: 2C10x30
2. Material of the member: A36 steel (Fy = 36 ksi, and F u = 58 ksi)

Required:
To establish the required axial tensile strength of the member by considering only limit state of tensile
yielding in the gross section.

Solution:
According to the LRFD method, the required axial tensile strength of the member, based on yielding in the
gross section, is:
T u = φ t P n = φ t F y A g , where;
φ t = 0.90, F y = 36 ksi (yield stress for given A36 steel); and
A g = gross cross-sectional area of the member made of two C10x30 channels, where area of each channel
is 8.81 in2. Therefore,
A g = 2 x 8.81in2 =17.6 in2.
T u =φ t P n = φ t F y A g = 0.90 x 36 ksi x 17.6 in2 = 570.3 kips
Answer: T u = φ t P n = 570.3 kips

4.9. LIMIT STATE OF TENSILE RUPTURE IN

THE NET SECTION
(Ref: AISC Specification, (AISC, 2010a), Sec. D2, Page 16.1-26))
As we discussed earlier, in tension members, where bolts or rivets are used, due to presence of bolt
or rivet holes, the cross sectional area of the member is reduced. The reduced cross-sectional area
is called net area, which is smaller than the total area (i.e. the gross area) of the cross section. In
addition, due to stress concentrations, the actual stresses around the boltholes are greater than the
average stress on the net area. As a result, when tension force is applied, the net section starts
yielding prior to yielding of the gross section. However, since the length of the net section is
relatively small, yielding of the net section is not considered a failure mode. As the net section
continues to yield, it strain-hardens and eventually stresses on the net section reach ultimate
strength of the steel, F u , and the net section necks down and ruptures (fractures), Figure 4.8(b).

Applied Tension = Tu (in LRFD)

= T (in ASD)
Uniform Stress= Fu

Net Area= An

Figure 4.11. Rupture in the Net Section under Assumed Uniform Ultimate Stress, Fu
CHAPTER 4 - TENSION MEMBERS 4-13

In design, the stress distribution on the net section is assumed to be uniform, Figure 4.11.
When this uniform stress reaches ultimate strength of material, F u , the net section is considered
ruptured.
The equation of design for checking rupture in the net section is:

Tu ≤ φ Pn (in LRFD) and T a ≤ P n /Ω (in ASD) (4.3)

Where
T u = required axial tensile strength in LRFD, same as the factored applied tension force.
φ = resistance factor for rupture under tension in LRFD = 0.75
T a = required axial allowable strength in ASD, same as the un-factored applied tension force
Ω = safety factor for rupture under tension in ASD = 2.0
P n = available nominal axial strength of the net section of the member given by:

Pn = Fu Ae (4.4)
Where
F u = specified minimum tensile strength of steel.
A e = effective net area given by:

Ae = An U (4.5)

A n = net area of member discussed later in Section 4.10

U = shear lag factor discussed in Section 4.11 after
T a = required allowable axial tensile strength in ASD, same as the un-factored (service) applied
tension force.

4.10. NET AREA, A n

4.10. a. Net Area, A n , When Bolt Holes Are on a Straight Line
(Ref: AISC Specifications, 2010, Sec. D3, Page 16.1-27)
Net area of a tension member, A n is the gross area of the cross section minus the sum of the
projected areas of the bolt or rivet holes. Projected area of a bolthole is calculated as the thickness
of the element along the centerline of the bolt times design diameter of the hole. Design diameter
of a standard bolt or rivet hole is equal to actual diameter of the hole plus 1/16 of an inch. The
reason for adding 1/16 of an inch to actual diameter of the hole is mainly to exclude the areas of
the steel around the bolt hole that might have been damaged during the hole punching process or
lost during the hole drilling.

A n = A g - n (d h + 1/16") t (4.6)
Where;
A n = net area of the member
A g = gross area equal to the total cross-sectional area of the member.
n = number of bolt or rivet holes in the fracture plane
d h = nominal hole diameter, see Figure 4.12 on the next page.
t = thickness of the element, see Figure 4.12.

Rupture in the net section

of the channels

Figure 4.12. Rupture in the Net Section of Bolted or Riveted Member

Numerical Example 4.3: Calculating Net Area, A n

Calculate net area of the tension member shown below.

¾” dia. bolts in 13/16” holes

3 bolt holes on the 2C-15x50

rupture plane

Given:
1. Cross section of the member: 2C-15x50
2. Diameter of the holes: 13/16 inch.

Required:
Calculation of the net area.

Solution:
Properties of one channel section C15x50 are: Area =14.7 in2, t w = 0.716 in.
Calculate A n :
A n =A g – n (t w ) (d h +1/16”) (2), where:
A g = Gross area of two back-to-back C15x50 channels = 2x 14.7 in2 = 29.4 in2
n = Number of bolt holes on the net section of one channel = 3 (see above sketch)
d h = Diameter of bolthole = 13/16 in.
t w = Thickness of the web of one channel = 0.716 in.
A n = A g – n (d h +1/16”) (t w )(2)
A n = 29.4 in2 - 3 (13/16” +1/16”) (0.716“) (2) = 25.6 in2
Answer: A n = 25.6 in2
CHAPTER 4 - TENSION MEMBERS 4-15

4.10. b. Net Area, A n, of Cross Sections with Staggered Holes

In bolted and riveted steel structures, sometimes bolt or rivet holes are positioned on a zigzag or
diagonal line as shown in Figure 4.13. This is done to make the connection more compact as well
as to increase fracture capacity of the net section.
If boltholes are staggered, the net area ends up being greater than the net area if the
boltholes were on a straight line. The reason for the additional net area is that the sloped surfaces
in a staggered net section, Figure 4.13, are greater than the straight section.

Rupture in the net section with

staggered boltholes
Sloped (staggered) surface
Straight surface

Figure 4.13. Fracture of Net Area through Staggered Holes

The net area for staggered net sections is calculated by using the following equation:

A n = A g - n (d h + 1/16") t + Σ (s2/4g) t (4.7)

Where
(s2/4g)t = area added to the net area for each stagger
s = longitudinal center-to-center spacing (pitch) of any two adjacent holes, see Figure 4.13.
g = transverse center-to-center spacing (gage) between fastener gage lines, see Figure 4.13.

Numerical Example 4.4: Calculating Net Area, A n, when Bolts are

Staggered
Calculate net area of the tension member shown below.
¾” dia. bolts in 13/16” holes
Staggered net section

g = 3 in.

s = 2 in. 2C-15x50

Given:
1. Cross section is given as 2C-15x50
2. Diameter of holes is given as 13/16 inch.
3. Dimensions of staggered holes are given as s=2 in. and g= 3 in.

Required:
Calculation of net area considering the staggered holes.

Solution:
Properties of one channel section C15x50 are: Area =14.7 in2, t w = 0.716 in.
Calculate A n :
A n = A g – n (d h +1/16”) t + Σ(s2/4g) t
Where:
A g = gross area of two back-to-back C15x50 channels = 2x 14.7 in2 = 29.4 in2
n = Number of bolt holes on the staggered net section of one channel = 3
d h = Diameter of bolthole= 13/16 in.
t = 2t w where tw = Thickness of the web of one channel = 0.716 in.
s = Longitudinal center-to-center spacing of two adjacent holes = 2.0 in.
g = Transverse center-to-center spacing between fastener gage lines = 3.0 in.
A n = 29.4 in2 – (3 holes) (13/16 in. +1/16 in.) (2x 0.716 in.) + [(2)(2 in.)2 / (4x3 in.)] (2x0.716) = 26.6 in2

In this case, there is an additional net section that we needed to check. This net section is a “straight”
section (not staggered) passing through the first bolt line as shown below.

Straight net section

The net area for this straight rupture line is:

A n = A g – n (d h +1/16”) (2t w )
A n = 29.4 in2 – (1 hole)(13/16 in. +1/16 in.) (2x0.716 in.) = 28.2 in2
Since this net area is larger than the net area of 26.6 in2 for the staggered rupture line (established
above), then the answer, being the smaller of the two, is:
Answer: A n = 26.6 in2

4.10. c. Net Area, A n, of Cross Sections When Staggered Holes are on

Two Perpendicular Planes
In some cases, such as the one shown in Figure 4.14 below for angles, the staggered boltholes are
on different elements of the cross section. The situation can also occur when both the web and
flanges of an I-shaped section are bolted to the support. Usually this type of staggering of bolt-
holes is done to prevent the heads of the bolts on one element, such as the leg of an angle, not to
interfere with the bolt heads on the other leg of the angle.
To establish net section in these cases, we can rotate one element, in this case, one of the
legs of the angle, 90 degrees to have both elements (both legs of the angle) in one plane. Then, we
can treat the resulting case as a single plate element with staggered holes.
CHAPTER 4 - TENSION MEMBERS 4-17

Staggered net section

Four Angles

Figure 4.14. An Example of Staggered Bolt Holes on Different Elements of Cross Section

As shown in Figure 4.15(b), the rotation of one element, the short leg of the angle in this
case, has to be done about Point “c” which is the intersection of the centerlines of the two legs of
the angle. After rotating the short leg 90 degrees, the resulting plate is shown in Figure 4.16 where
all boltholes are in one plane. The net area of this plate can be established exactly the same way
we discussed in previous section. The values of dimensions “s” and “g” for the plate are shown in
Figure 4.16.
t
Staggered net section

t
Point “c”
(Center of Rotation)
a-t/2

a t

6” b-t/2 (b)
b (a)

Figure 4.15. (a) Staggered Bolt Holes on Both Legs of the Angle, and
(b) Rotation of the Short Leg to Convert the Angle to a Flat Plate

s Flat Plate
Flat Plate
Staggered net section
Point “c”
(Center of
Rotation)
g1 a-t/2 a

g2 t

b-t/2 b

Figure 4.16. Fracture of Net Area through Staggered Holes in the Plate Obtained
by Converting the Angle to a Flat Plate
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-18

Numerical Example 4.5: Calculating Net Area, A n, when Staggered

Bolts are in Different Planes
Calculate net area of the double-angle tension member shown below.

Gusset Plate 7/8” diameter bolts

in 15/16” holes

1.5” 2L8x8x1
3” 2L8x8x1 (Member)
3.5”
3.5” 3.5”
3” 3”
1.5”

11@1.5” 3.5” 3”

Given:
1. Member cross section: 2L8x8x1
2. Diameter of the boltholes: 15/16 inch.

Required:
Calculation of net area considering the staggered holes in different planes (i.e., two legs of angles).

Solution:
Properties of one angle section L8x8x1 are: A=15.0 in2, t =1.0 in.
Calculate A n :
A n = A g – n (d h +1/16”) t + Σ(s2/4g)t
Where:
A g = Gross area of two back-to-back L8x8x1 angles = 2x 15.0 in2 = 30.0 in2
n = Number of bolt holes on the staggered net section = 4
d h = Diameter of bolthole = 15/16 in.
t = (2) (thickness of the angle leg which is 1.0 inch) = (2x 1.0 in.)
s = Longitudinal center-to-center spacing (pitch) of any two consecutive holes, established below.
g = Transverse center-to-center spacing (gage) between fastener gage lines established below.
As the sketch (a) below shows, the staggered boltholes are on both legs of the angles and on two different
planes. We need to rotate one leg 90o to convert the angles to flat plates, sketch (b). Then, we can
calculate the s and g values as 1.5 and 6 inches respectively, see sketch (c) below.

1.5” 1.0”
1”
Point “c”
(Center of
a-t/2=7.5”
Rotation)
15”
8”
a-t/2=7.5” Rotate the leg
8” 90 degrees

(a) (b)
CHAPTER 4 - TENSION MEMBERS 4-19

S = 1.5”

Flat Plate
3”
3”
g = 6” 15”
3”
3”

(c)

Using above s and g values, we calculate the net area for staggered bolt pattern:
A n = A g - n(d h +1/16”) (t)(2) ) + Σ(s2/4g) t
A n = 30 in2 - 4(15/16 in. +1/16 in.) (2x1.0 in.) + [(1.5 in.)2 / (4x6 in.)] (2x1.0 in.)
A n = 22.0 in2 + 0.2 in2 = 22.2 (for the staggered net section)
Let us now consider the “straight” net section shown below:

Flat Plate

15”
3”
3” Rotate the leg
90 degrees

(d)

The net area for this straight (non-staggered rupture line) is:
A n = A g – n (d h +1/16”) (t)
A n = 30.0 in2 – 2 (15/16 in. +1/16 in.) (2x1.0 in.) = 26.0 in2
Since this net area is larger than the net area of 22.2 in2 for the staggered rupture line (established
above), then the answer, being the smaller of the two, is:
Answer: A n = 22.2 in2

4.10. d. Net Area, A n, of Cross Sections with Plug or Slot Welds

(Ref: AISC Specifications, (AISC, 2010a), Sec. D2, Page 16.1-27))
Occasionally, tension members are connected to the support by using “plug” or “slot” welds,
Figure 4.18. To do these welds, round holes called plugs or elongated slots are made on one of the
steel pieces to be connected. Then, the two pieces are put together and the welds are placed inside
the plug or slot. Notice that rupture in the net section can occur in tension members where “plug”
or “slot” welds are used at their end connection areas, resulting in reduction of their gross area. We
will have more information on these welds and how they are designed in Chapter 10. Here we are
focusing on if such welds are used to connect a tension member to its support, what would be the
net area of the member?

For plug-welded and slot-welded connections, Figure 4.17, the net area of the member
should be calculated by subtracting the projected area of the plug or slot welds from the gross area.
This is the same as we did for net sections with bolt or rivet holes in them. Therefore, Equations
4.6 and 4.7 given earlier apply to this case with one modification that in this calculation, the actual
diameter of a plug or the width of the slot weld hole is used in Equations 4.6 and 4.7 without the
additional 1/16 inch.
dh
Plug Welds Slot Welds
dh t

An = Net area An = Net area

(a) Plug Welds (b) Slot Welds

Figure 4.17. Net Section in Plug and Slot Welded Tension Members

4.10. e. Net Area, A n, of Rectangular Tube or Pipe Sections Welded to Single

Gusset Plate
(Ref: AISC-Specification, (AISC, 2010a), Sec. B4.3b, Page 16.1-18))
In order to connect a rectangular tube or a pipe tension member to a gusset plate an economical
solution is to cut a slot along the length of the member, as shown in Figure 4.18(a) and after
placing the gusset plate inside the slot, to weld the intersection of the member and the gusset plate,
Figure 4.18(b). The construction process leaves a small open gap between the end of the gusset
plate and the end of the slot in the pipe or rectangular member. This gap makes the cross section of
the slotted member smaller than its gross area resulting in the possibility of rupture in the net
section through this reduced net area.

Single tg
Gusset Plate, Gap
Thickness= tg t

Round HSS or
Pipe Section

Net Area
Net Section
Single
Gusset Plate, tg
Gap
Thickness= tg
t
HSS
Tube Section

Net Area

Net section
(a) (b) (c)

Figure 4.18. Net Section in Slotted Round and Rectangular HSS Section Welded
to a Single Gusset Plate
CHAPTER 4 - TENSION MEMBERS 4-21

The net area, A n , in this case is the gross area minus twice the product of the thickness of
the slotted member and the total width of the material that is removed from the slot, t g , Figure
4.18(c). The net area is:
A n = A g – 2t t g (4.8)
Where
t = “design” thickness of the rectangular tube or pipe, in., Figure 4.18(c). The “design” thickness
for an HSS shape is listed in the AISC Manual tables. This thickness is usually slightly less
than the nominal thickness. For example for HSS6x6x1/2, the nominal thickness is ½ in.,
but, the design thickness in the AISC Manual tables is 0.465 in.
t g = thickness of the gusset plate, in., Figure 4.18(c)
A g = gross area of the member.

Numerical Example 4.6: Calculating Net Area, A n, for HSS Tube

Sections Connected to Single Gusset Plate
Calculate net area of the tension member shown below.
Gap
Gap

HSS 6x5x1/2
Gusset plate
tg = 1 in.

Given:
1. Cross section of the member is given as HSS6x5x1/2
tg = 1”
2. Thickness of gusset plate is given as 1.0 inch.

Required: t =0.465”
Calculation of net area of the tube.

Solution:
Properties of HSS 6x5x1/2: A g = 8.81 in2, t = 0.465 in.
Net Area
Calculate A n :
A n = A g – 2 x t x t g = 8.81 in2 – 2 x 0.465 in. X 1.0 in. = 7.9 in2
Answer: A n = 7.9 in2

4.11. EFFECTIVE NET AREA, A e

(Ref.: AISC Specifications, (AISC, 2010a), Sec. D3, Page 16.1-27))
Earlier, in Equation 4.5 (Page 4-13), the term A e, effective net area, was used to define the portion
of the net section of the tension member that effectively is carrying the tension load. In this
section, we focus on establishing effective net area.
When all elements of a cross section are attached to the support, tensile stresses are directly
transferred to the support by all elements as shown in Figure 4.19(a). For such cases, effective net
area, A e is equal to the actual net area A n. However, if some elements of the cross section are
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-22

attached to the support directly, as shown in Figure 4.19(b), but other elements are not, then the
tensile stresses in the parts of the cross section not attached to the support, have to be transferred to
the attached elements as a shear stress, Figure 4.19(b).
This conversion of the tensile stresses to shear stresses to be transferred to the support is
called “shear lag.” Shear lag is the reason for the elements of cross section not directly connected
to the support not to be fully effective in resisting tension. For the tension members with shear lag,
the effective net are, A e is defined as:
Ae= An U (same as Equation 4.5 given earlier)

In the above equation, U is the “shear lag factor” that depends on the geometry of the
cross section and which parts of the section are not directly attached to the support. The AISC
Specification (AISC, 2010a, Table D3.1, Page 16.1-28) provides values of the shear lag factor, U,
for common cases of tensile members. The AISC table for U values is given in the next page as
Table 4.1. For tension members, with cross sections significantly different from those listed in
the AISC Table D3.1, the shear lag factor U can be established by conducting tests or finite
element analyses.
In using Table 4.1 below, which is the same as Table D3.1 of the AISC Specification
(2010), we need to consider the following as well:
1. For open sections, such as W, M, S, HP , WT, T, C, and L (single or double
angles), U, need not be less than the ratio of the gross area of the connected
element(s) to the member gross area (Ref.: AISC Spec. (AISC, 2010a), Sec. D3,
Page 16.1-27)). This provision does not apply to closed sections, such as HSS
sections, nor to plates (Ref: AISC Spec. (2010), Sec. D3, Page 16.1-28).
2. For bolted splice plates A e = A n ≤ 0.85A g (Ref.: AISC Spec. (AISC, 2010a), Sec.
J4.1, Page 16.1-129).
Following sections provide more information and explanations on Cases 1 through 8 of
the shear lag factor U given in Table 4.1. Discussion of each case of U is followed by a numerical
example, which demonstrates application of the case.

Shear stresses This area is not effective

in resisting tension

(a) Both legs of Angle are connected to (b) Only the vertical leg is directly connected to
the support, therefore no shear lag the support. The horizontal leg is not
exists connected to the support, therefore , has
shear lag

Figure 4.19. Shear Lag Effect in Tension Members

CHAPTER 4 - TENSION MEMBERS 4-23

Table 4.1- Shear Lag Factors, U

4.11. a. Discussion and Application of Case 1 of U in Table 4.1

(Ref.: AISC Specifications, (AISC, 2010a), Table D3.1, Page 16.1-28))
This case applies to the open section (W, M, S, HP, WT, WS, C and L) tension members where all
segments of the cross section are connected to the support, transmitting the tensile stresses to the
support directly with very small or no shear lag effects. Examples of the members in this category
are shown in Figure 4.20. For all of these members, U is equal to 1.0 and the entire net section is
effective. Notice that for welded tension members, if there are no boltholes or other reduced
sections, net area is equal to the gross area.

b. The flange and stem of the Tee are

a. Both flanges and web are connected to
connected to the support
the support

c. Both flanges and web of the Channel are d. Both legs of angles are connected
connected to the support to the support

Angle

Wide Channel
] Flange
HSS

Welds
Welds
Welds

g. Both flanges and the web are welded

e. Both flanges and the web are f. All four sides of the box to the support
welded to the support are welded to the support
Welds Welds

Case 1. When all elements of the cross section are connected to the support:

Ae= An U, where An = Net Area, and U = 1.0

Figure 4.20. Effective Net Area for Bolted and Welded Tension Members Connected
to the Support by All Elements of the Cross Section
CHAPTER 4 - TENSION MEMBERS 4-25

Numerical Example 4.7: Calculating Effective Net Area, A e , for a

Case 1 Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the wide flange
member shown below.

End Plate

Welded all around

W12×58, A992 Grade 50

Given
1. Size of the member: W12x58 (Area =17.0 in2)
2. Material of the W section: A992 Grade 50 (F y = 50 ksi, F u = 65 ksi)
3. Details of the welded end connections indicate that all elements of the cross section are connected to
the support by welds.

Required:
To calculate the required axial tensile strength (in LRFD).

Solution: End plate

support
1. Yielding in the Gross Section:
T u1 = φ t P n = φ t F y A g , where;
Weld all around
φ t = 0.90, F y = 50 ksi (given), and A g = 17 in2 (given)
φ t P n = φ t F y A g = 0.90 x 50 ksi x 17.0 in2 = 765.0 kips
T u1 = 765.0 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Calculate Effective Net Area, A e :
Ae = An U
Since the welded member does not have any holes net area is equal to gross area.
A n = A g = 17.0 in2
Since all elements of the cross section, (i.e. web and both flanges), are welded to the end plates, this is a
Case 1 in Table 4.1 and there is no shear lag effect, therefore, U = 1.0
A e = A n U = (17.0 in2) (1.0) = 17.0 in2
T u2 = φ P n = φ F u A e , where;
φ = 0.75, F u = 65 ksi (given); and A e = 17.0 in2
T u2 = φ P n = φ F u A e = 0.75 x 65 ksi x 17.0 in2 = 828.7 kips
T u2 = 828.7 kips (due to rupture in the net section)
Smaller of the two values of T u1 and T u2 is the governing tensile strength.

Answer: T u = φ t P n = 765.0 kips (due to yielding in the gross section)

4.11. b. Discussion and Application of Case 2 of U in Table 4.1

(Ref.: AISC Specifications, (AISC, 2010a), Table D3.1, Page 16.1-28))
Case 2 in Table 4.1(Page 4.23) applies to tension members that are connected to the support by
some, but not all of their cross section elements. Examples of such tension members are shown in
Figures 4.21 and 4.22. In these members, shear lag is present and the effective net area is less than
the actual net area. The U value for this case is given by:

U = 1- x/L (4.9)
Where
x = a distance shown in Figure 4.21 related to location of C.G. of the shaded areas in Figures
4.21 and 4.22.
L = length of connection, in. For bolted connections, L is center-to-center distance of first and
last bolt, see Figure 4.21. For welded connections, L is the back-to-back length of the weld
measured parallel to applied tension force, Figure 4.22.
In some cases, establishing “x” can be time consuming. Case 7 of the Table 4.1, discussed
later in Section 4.11(g), provides alternative and simplified values for U for W, M, S, and HP
sections without a need to calculate the “x.”
Figure 4.23 shows examples of channel and angle members, where some, but not all, of
the cross section elements are connected to the support.
L x
L

C.G. of shaded T-shaped area made of C.G. of the dark colored L-shaped area
½ of the wide flange section made of ¼ of the wide flange section

(a) Only Flanges of W-shape Are Connected (b) Only Web of W-shape Is Connected

x L
L x

C.G. of the dark colored L-shaped

C.G. of the WT Section area made of ½ of the Tee section

Case 2. For bolted W, M, S, HP, and T-sections cut from them, when only some elements of the cross
section are connected to the support:

Ae= An U, An=Net Area, and U=1−x/ L

Figure 4.21. Effective Net Area for Case 2 of Shear Lag for Bolted W, M, S, HP and T-section
Cut from Them, Where Only Some, But Not All Elements Are Connected
CHAPTER 4 - TENSION MEMBERS 4-27
L x
L

Welds
Welds
C.G. of shaded T-shaped area made C.G. of the dark colored L-shaped area
of ½ of the wide flange section made of ¼ of the wide flange section
(a) Only Flanges of W-shape Are Connected (b) Only Web of W-shape Is Connected

x L x
L

Welds C.G. of the dark colored L-shaped

C.G. of the WT Section area made of ½ of the Tee section

Case 2. For bolted and welded W, M, S, HP, and T-sections cut from them, when only some
elements of the cross section are connected to the support:

Ae= An U, An=Net Area, and U=1−x/ L

Figure 4.22. Effective Net Area for Case 2 of Shear Lag for Welded W, M, S, HP and T-section Cut from
Them, Where Only Some, But Not All Elements Are Connected
x x
L
L

C.G. of One Channel

C.G. of One Angle
x x
L L

Welds
C.G. of One Channel
C.G. of One Angle
Case 2. For bolted and welded channels and angles, when only some elements of the cross section
are connected to the support:

Ae= An U, An=Net Area, and U=1−x/ L

Figure 4.23. Effective Net Area for Case 2 of Shear Lag for Channel and Angle Sections Where Only
Some, But Not All Elements Are Connected
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-28

Numerical Example 4.8: Calculating Effective Net Area, A e , for a Case 2

Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the double angle
member shown below.

2L6×4×¾, A36 Steel

8” Welds 7/8” dia. bolts in

15/16” dia. holes
12”

Welded End Connection Bolted End Connection

Given:
1. Size of the member: 2L6x4x¾ (Area of one angle = 6.94 in2, t = 0.75 in.)
2. Material of the angles: A36 (F y = 36 ksi, F u = 58 ksi)
3. Details of both welded and bolted end connections.

Required:
To calculate the required axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
T u1 = φ t P n = φ t F y A g , where;
φ t = 0.90, F y = 36 ksi (given)
A g =2×6.94 in2 =13.88 in2
φ t P n = φ t F y A g = 0.90 x 36 ksi x 13.88 in2 = 449.7 kips
T u1 = 449.7 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Calculate Effective Net Area, A e : Gusset Plate
a. for Welded End:
Since on the welded end there are no holes, 8”
the net area is equal to gross area. y
A n = A g = 13.88 in2
U = 1- x/L (This is a Case 2 in Table 4.1
since only the vertical legs of the angles x x
are welded to the gusset plate.
x =1.07 in. (Location of C.G. of one angle)
L= 8 in. (given on the connection detail)
U =1- x/L = 1- (1.07 in) / (8 in.) = 0.86. Welds Net Section
y x=1.07”
A e1 = A n U = (13.88 in2) (0.86) = 11.93 in2
b. for Bolted End:
A n = Ag - n (d h +1/16”) (t)
A n = 13.88 in2 - 1(15/16“+1/16“) (2x 3/4“) =12.38 in2
U = 1- x/L
CHAPTER 4 - TENSION MEMBERS 4-29

x =1.07 in. (Location of C.G.) Gusset Plate

L= 12 in. (given on the connection detail) 12” y
U =1- x/L = 1- (1.07 in) / (12 in.) = 0.91.
A e 2 = A n U = (12.38 in2) (0.91)= 11.26 in2 x
x
The governing A e is the smaller of the
values obtained for the welded and bolted
ends:
A e = 11.26 in2 (governed by the bolted end)
y x=1.07”
T u2 = φ P n = φ F u A e , where; Net Section 7/8” dia. bolts in
φ = 0.75, F u = 58 ksi (given); and A e = 11.26 in2 15/16” dia. holes
T u2 = φ P n = φ F u A e = 0.75 x 58 ksi x 11.26 in2
T u2 = 489.8 kips (due to rupture in the net section)
Smaller of the two values of T u1 and T u2 is the governing tensile strength:

Answer: T u = 449.7 kips (due to yielding in the gross section)

4.11. c. Discussion and Application of Case 3 of U in Table 4.1

(Ref.: AISC Specifications, (AISC, 2010a), Table D3.1, Page 16.1-28))
Case 3 in Table 4.1 (given earlier) refers to tension members where tension load is transmitted
only by transverse weld connecting some but not all of the segments of the cross section to the
support as shown in Figure 4.24. In this case, very serious shear lag will be present and as a result,
the effective area in tension is only the area of the directly connected elements of the cross section.

Ae = areas of the
Ae = areas of the Transverse webs only
Transverse welds flanges only welds on the
on the flanges only webs only

Transverse welds Ae = area of the

Transverse welds on Ae = areas of the on the flange only flange only
vertical legs only welded legs only

Case 3. For welded connections, when only some elements of the cross section are
connected to the support using transverse welds:

Ae= An U, An=Area of welded elements only, and U=1.0

Figure 4.24. Effective Net Area when Tension is Transmitted Only by Transverse
Welds on Some Elements
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-30

Numerical Example 4.9: Calculating Effective Net Area, A e , for a Case

3 Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the double angle
member shown below.
Transverse
welds only

2L6× 4×¾, A36

End Connections

Given:
1. Size of the member: 2L6x4x¾ (Area of one angle = 6.94 in2, t = 0.75 in.)
2. Material of the angles: A36 steel (F y = 36 ksi, F u = 58 ksi)
3. Details of welded end connections where only transverse welds are used.
Required:
To calculate the design axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
T u1 = φ t P n = φ t F y A g , where;
φ t = 0.90, F y =36 ksi (given)
A g = 2 × 6.94 in2 =13.88 in2
T u1 = φ t P n = φ t F y A g = 0.90 x 36 ksi x 13.88 in2 = 449.7 kips
T u1 = 449.7 kips (due to yielding in the gross section)

Gusset Plate

2. Rupture in the Net Section: y

Calculate Effective Net Area, A e :
Since only vertical legs are welded to the x x
support using only transverse welds, this is
a Case 3 in Table 4.1 and the net area is
equal to the area of the welded legs:
A n = 2×6 in x ¾ in = 9 in2
and;
U = 1.0.
Transverse Welds Net Section is y
A e = A n U = (9.0 in2) (1.0) = 9.0 in2
adjacent to the
T u2 = φ P n = φ F u A e , where; weld
φ = 0.75, F u = 58 ksi (given); and A e = 9.0 in2
T u2 = φ P n = φ F u A e = 0.75 x 58 ksi x 9.0 in2
T u2 = 391.5 kips (due to rupture in the net section)
Smaller of the two values of T u1 and T u2 is the governing tensile strength:

Answer: T u = 391.5 kips (due to rupture in the net section)

CHAPTER 4 - TENSION MEMBERS 4-31

4.11. d. Discussion and Application of Case 4 of U in Table 4.1

(Ref.: AISC Specifications, (AISC, 2010a), Table D3.1, Page 16.1-28))
Case 4 in Table 4.1 applies to plates subjected to tension where the plate is welded to the support
by only longitudinal welds as shown in Figure 4.25. In this case, depending on the width of the
connected plate, W, and the length of the welds, L w, certain amount of shear lag can be present.
The shear lag can result in reducing the effective net area of the plate. Notice that if the end of the
plate is welded by a transverse weld, there would be no shear lag and the effective area in tension
will be equal to the actual area.
Longitudinal
Welds only

Plate
w

Lw
No Transverse Welds

Case 4. For welded plate with only longitudinal welds:

Ae= An U, An=Ag, and U = as follows:

For Lw ≥ 2w ............................... U = 1.0

For 2.0w > Lw ≥ 1.5w ............................ U = 0.87
For 1.5w > Lw ≥ w ................................. U = 0.75

If w > Lw, the design is not acceptable (i.e. U = 0).

Figure 4.25. Effective Net Area of Plates Connected by Longitudinal Welds Only

Note that if in addition to longitudinal welds, there is a transverse weld also, the case will
be a Case 1 in Table 4.1, and U will be equal to 1.0.

Numerical Example 4.10: Calculating Effective Net Area, A e , for a

Case 4 Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the double plate
member shown below.
16”

2PL10×¾, A36 steel

Longitudinal Welds only

Welded End Connections

Given:
1. Size of the member: 2PL10x¾
2. Material of the plates: A36 steel (F y = 36 ksi, F u = 58 ksi)
3. Details of the welded end connections (with longitudinal welds only).

Required:
To calculate the design axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
T u1 = φ t P n = φ t F y A g , where;
φ t =0.90, F y =36 ksi (given)
A g =2×10 in x0.75 in =15.0 in2
T u1 = φt Pn = φt Fy Ag = 0.90 x 36 ksi x 15.0 in2 = 486.0 kips
Tu1= 486.0 kips (due to yielding in the gross section)

Lw =16” Net Section

2. Rupture in the Net Area: of Plates on
Calculate Effective Net Area, Ae: the Right
Ae = An U
Since the welded plate has no holes:
An = Ag = 15.0 in2 w =10”
The plates in this case are welded to the support
by longitudinal welds only, then this is a Case 4
connection in Table 4.1.
Longitudinal Welds
To establish U, check Lw versus w:
only
Lw = 16 in. and w = 10 in.
Since 2w > Lw ≥ 1.5w, then; U = 0.87 (see Figure 4.25 above)
Ae = An U = (15.0 in2) (0.87) = 13.0 in2
Tu2 = φ Pn = φ Fu Ae, where;
φ = 0.75, Fu = 58 ksi (given); and Ae = 13.0 in2
Tu2 = φ Pn = φ Fu Ae = 0.75 x 58 ksi x 13.0 in2
Tu2 = 565.5 kips (due to rupture in the net section)
Smaller of the two values of Tu1 and Tu2 is the governing tensile strength:

Answer: Tu = 486.0 kips (due to yielding in the gross section)

4.11. e. Discussion and Application of Case 5 of U in Table 4.1

(Ref.: AISC Specifications, (AISC, 2010a), Table D3.1, Page 16.1-28))
Case 5 in Table 4.1 (given earlier Page 4-23) deals with tension members with round hollow
sections (pipes) that are connected to a single gusset plate by longitudinal welds. As discussed
earlier, in these cases, the gusset plate is placed in a slot that is cut in the pipe and welded to it as
shown in Figure 4.26. Notice that when pipe section is connected to a single-gusset plate, as
shown in Figure 4.26, there is a small open gap between the end of the gusset plate and end of the
slot in the pipe. This gap makes the cross section of the slotted member smaller than its gross area
resulting in a possibility of rupture in the net section through reduced net area. In fact, this cross
section is called the “net area” of the pipe. The AISC-Specification (AISC, 2010a) states that net
area, An, in this case is the gross area minus twice the product of the thickness and the width of the
material that is removed from the slot, Figure 4.26.
CHAPTER 4 - TENSION MEMBERS 4-33

Single Gusset
Gap Plate
Gap
Slots cut in
the pipe

Longitudinal welds on both sides

L
Gap Round Pipe D tg

C.G. of t
half pipe
D

x=D/π
Net Area

Case 5. For round HSS sections (or pipes) welded to a single gusset plate by longitudinal
welds only:

Ae= An U, An = Net Area, and values of U are given as:

For L ≥ 1.3D. .. . . .U = 1.0
For D ≤ L <1.3D . . . U=1−x/L
Where, x = D/π

Figure 4.26. Effective Net Area of Round HSS or Pipes with Single Gusset Plate

Values of U for Case 5 in Table 4.1 are given only for cases where L, length of the weld,
is greater than or equal to D, diameter of the pipe, (see inside the box in Figure 4.26). The AISC
Specification does not provide a value of U for connections with L less than D. The use of such
connections is discouraged since in such short connections, significant shear lag will exist
resulting in an inefficient design and much reduced capacity of tension member.

Numerical Example 4.11: Calculating Effective Net Area, Ae, for a

Case 5 Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the pipe member
shown below.

Pipe 12 Standard,
7/8” Gusset Plate A53, Gr. B steel

Given:
1. Size of the member: Pipe 12 std, (Area= 13.6 in2, Design wall thickness t = 0.349 in.)
2. Material of the pipe: A53, Gr. B (Fy = 35 ksi, Fu = 60 ksi)
3. Details of the welded end connections (with longitudinal welds only).

Required:
To calculate the design axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
Tu1 = φt Pn = φt Fy Ag , where;
φt =0.90, Fy = 35 ksi (given)
Ag =13.6 in2 (given)
φt Pn = φt Fy Ag = 0.90 x 35 ksi x 13.6 in2 = 428.4 kips
Tu1 = 428.4 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Ae = An U
Since the pipe in this case is welded to the single gusset plate by longitudinal welds only, then this is a
Case 5 connection in Table 4.1 on Page 4-23. For Case 5:
An = Ag - 2 × t × tg (See sketch below-right showing the net area)
Ag = 13.6 in2 for Pipe 12“Std. From section properties tables. Therefore;
An =13.6 in2 - 2 × 0.349” x 7/8” =13.0 in2

Gap tg =7/8”
t =0.349”
t =0.349”

D=12”
C.G. of
half pipe
Net Area
L=14” Net section x

For Pipe 12” Std.:

Outside Dia =12.8 in.
Therefore, D =12.8 in.
Check L versus D:
L = 14 in. and D = 12.8 in. (Notice that D is the outside diameter of the pipe, which for Pipe 12” Std.
section is equal to 12.8 in.)
Since D ≤ L <1.3D, then U =1−x/L
U =1-x / L= 1- (D/π)/L= 1- (12.8“ /3.1415) / (14“) = 0.71
Ae = An U = 13.0 in2 x 0.71 =9.2 in2
Tu2 = φt Pn = φt Fu Ae, where
φ = 0.75, Fu = 60 ksi (given); and Ae = 9.2 in2
Tu2 = φ Pn = φt Fu Ae = 0.75 x 60 ksi x 9.2 in2
Tu2 = 414.0 kips (due to rupture in the net section)
Smaller of the two values of Tu1 and Tu2 is the governing tensile strength:
Answer: Tu = 414.0 kips (due to rupture in the net section)
CHAPTER 4 - TENSION MEMBERS 4-35

4.11. f. Discussion and Application of Case 6 of U in Table 4.1

(Ref.: AISC Specifications, (AISC, 2010a), Table D3.1, Page 16.1-28))
Case 6 in Table 4.1 (Page 4-23) applies to rectangular or square Hollow Steel Sections (HSS)
connected to single or double gusset plates, as shown in Figures 4.27 and 4.28 respectively.
Hollow structural sections are frequently used in trusses as well as in braced frames. The end
connections of these members are often single gusset plate placed inside a slot cut into the HSS
tube as shown in Figure 4.27. Double gusset plates, shown in Figure 4.28, are also used in many
trusses and in some bracings. In both cases of single or double gusset plates, since only horizontal
elements of the tube cross section are directly welded to the plate, a shear lag factor U needs to be
applied to obtain the effective net area. The U factor for each case is given in Figures 4.27 and
4.28 for single and double gusset plates respectively.

Weld
Single
Gusset Plate
Gap

HSS
Tube Section

Longitudinal welds on
both sides
B tg
Gap t
C.G. of
half HSS

Net Area
L
HSS Tube
x

Case 6 (a) , For connections where a rectangular or square HSS tube section is welded to a single gusset
plate by longitudinal welds:

Ae= An U, An=Ag – 2tg t , U = 1- x/L

Where, x=[ B2 + 2B H] / [4(B + H )]

Notice that for this case L should be equal or greater than H for the connection to be acceptable.

Figure 4.27. Effective Net Area of HSS Tension Members Connected to Single Gusset Plate

Double
Gusset Plate

HSS
Tube Section

B
Longitudinal welds on four
corners of the tube

C.G. of
half tube
H

Net Area = Gross Area

L
HSS
x
Rectangular
Tube

Case 6 (b), For connections where a rectangular or square HSS tube section is welded to double gusset
plates by longitudinal welds:

Ae= An U, An = Net Area, and U = 1- x/L, where, x= (B2) / [4(B+H)]

Figure 4.28. Effective Net Area of Rectangular HSS Tension Members

Connected to Double Gusset Plates

Notice that when HSS tube member is connected to a single-gusset plate, as shown in
Figure 4.27, there is a small open gap between the end of the gusset plate and end of the slot in the
tube. This gap makes the cross section of the slotted member smaller than its gross area resulting
in a possibility of rupture in the net section through reduced net area. In fact, this cross section is
called the “net area” of the tube. The net area, An, in this case, is the gross area minus the product
of the thickness of the gusset plate and the thickness of the tube for each slot. For two slots the
area to be subtracted from the gross area is 2tg t, see Figure 4.27.
In double gusset plates, Figure 4.28, there is no slot cut into the member. As a result, the
net area is the same as the gross area. However, a shear lag factor of U still needs to be applied
since in this case, only the vertical elements of the HSS tube section are directly welded to the
gusset plates.
For Case 6 of Table 4.1, in order to use U =1-x/L to establish shear lag factor, the length
of the welds connecting the HSS section to the single or double gusset plate should be greater than
the overall depth, H, of the sides of the cross section.
CHAPTER 4 - TENSION MEMBERS 4-37

Numerical Example 4.12: Calculating Effective Net Area, Ae, for a

Case 6a Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the rectangular
HSS member shown below.
tg =1” Gusset Plate
Longitudinal Welds
Gap

HSS 8x8x1/2
(Square Tube)
14”
HSS8x8x1/2
Welded End Connection
Given:
1. Size of the member: HSS8x8x1/2, Square Tube, with B = H= 8 in. and t = 1/2 in. Note that this
thickness of ½” is the nominal thickness. In calculations, we should use the “design” value of thickness,
which is 0.465” in this case (see Manual Tables for HSS shapes for “design” values of thickness).
2. Material of HSS member: A500, Gr. B (Fy = 46 ksi, Fu = 58 ksi)
3. Details of the welded end connections (with longitudinal welds only).

Required:
To calculate the design axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
Tu1 = φt Pn = φt Fy Ag, where;
φt = 0.90, Fy = 46 ksi (given)
Ag = 13.5 in2 for HSS 8x8x1/2 from section properties tables.
Tu1 = φt Pn = φt Fy Ag = 0.90 x 46 ksi x 13.5 in2 = 558.9 kips
Tu1 = 558.9 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Ae = An U
Notice that since the HSS section in this case is welded to the single gusset plate by longitudinal welds
only, then this is a Case 6a connection in Table 4.1. For Case 6a:
An = Ag - 2 × t × tg (This is Equation 4.8. See sketch below right for net section)
An =13.5 in2 - 2 × 0.465” x 1” =12.6 in2

Gap B=8” tg =1”

t =0.465”

H=8” H=8”

Net C.G. of Net Area

section x half HSS
L=14”

For this design, Case 6 of “U” in Table 4.1, also shown in Figure 4.27 on Page 4-35.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-38

Check L versus H:
L = 14 in. and H = 8 in.
Since L = 14” > H = 8”, then the connection is acceptable (see Figure 4.27) and,
U = 1-x / L where x = (B2 + 2BH)/[4(B + H)]
U = 1- [(82 +2x8x8) / (4x16) ] / 14“ = 0.78
Ae = An U= 12.6 in2 x 0.78 ×= 9.83 in2
Tu2 = φ Pn = φ Fu Ae, where;
φ = 0.75, Fu = 58 ksi (given); and Ae = 9.83 in2
Tu2 = φ Pn = φ Fu Ae = 0.75 x 58 ksi x 9.75 in2
Tu2 = 427.5 kips (due to rupture in the net section)
Smaller of the two values of Tu1 and Tu2 is the governing tensile strength:

Answer: Tu = 427.5 kips (due to rupture in the net section)

4.11. g. Discussion and Application of Case 7 of U in Table 4.1

(Ref.: AISC Specification, 2010a, Table D3.1, Page 16.1-28)
The above six cases in Table 4.1 provide information on how to establish the shear lag factor U in
common cases of tension members with shear lag. Case 7 in Table 4.1 provides alternative
values for U values for W, M, S and HP sections and T-sections cut from these shapes that can be
used instead of those given in Case 2 of the table (as discussed earlier). The advantage of using U
values given for Case 7 is that in this case, there is no need to locate the center of gravity of the
connected element to establish value of x in U=1-x/L equation..

Numerical Example 4.13: Calculating Effective Net Area, Ae, for a

Case 7 Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the wide flange
member shown below.

Double Plates
1” dia. bolts in 1-1/16” holes

2@3”

2” 4@3” HP14x117,
A572, Gr. 50 steel
CHAPTER 4 - TENSION MEMBERS 4-39

Given:
1. Cross section of the member: HP14x117. (Ag =34.4 in2, tw = 0.805 in.). Notice that quite often HP
piles are used as piles supporting foundations and pile-caps.
2. Material of HP member: A572, Gr. 50 steel (Fy = 50 ksi, Fu = 65 ksi)
3. Details of the connections (with bolts on the web only).

Required:
To calculate the design axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
Tu1= φt Pn = φt Fy Ag, where;
φt = 0.90, Fy =50 ksi (given)
Ag = 34.4 in2 for HP14x117 from section properties tables.
Tu1= φt Pn = φt Fy Ag = 0.90 x 50 ksi x 34.4 in2 = 1548.0 kips
Tu1= 1548.0 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Ae = An U
An =Ag - n(dh+1/16”)(t)= 34.4 in2 -3(1.0625“ +1/16“) (0.805”) = 31.7 in2

¼ of HP
Double Plates section to be used in
calculating location of
1” dia. bolts in 1-1/16” the C.G.
holes

2@3”
C.G. of
¼ shaded zone
of HP section
2” 4@3” HP14x117,
A572, Gr. 50 x

To establish U; this problem is actually a Case 2 of “U” in Table 4.1, which means
Ae = An U, and U = (1-x/L). To calculate x for Case 2, we need to establish location of the C.G. of the ¼ of
the HP section, see sketch below (also in Figure 4.21(b) earlier). To avoid this relatively lengthy
calculation, alternatively, as allowed in Case 7 in Table 4.1, we will use U of 0.70. This value of U is given
in Table 4.1, Page 4-23, for webs connected with 4 or more fasteners in the direction of loading, which is
the case in this problem.
Therefore; U =0.70,
Ae = 0.70x 31.7=22.2 in2
Tu2 = φ Pn = φ Fu Ae, where;
φ = 0.75, Fu = 65 ksi (given); and Ae = 22.2 in2
Tu2 = φ Pn = φt Fu Ae = 0.75 x 65 ksi x 22.2 in2
Tu2 = 1082.3 kips (due to rupture in the net section)
Smaller of the two values of Tu1 and Tu2 is the governing tensile strength:

Answer: Tu = 1082.3 kips (due to rupture of net section)

It is interesting to note that in this case, if we had used a more precise value for U by applying Case 2
equation, the x for the L-shaped part of the HP would be established as 2.68 in., resulting in a value of
U = 1- 2.68”/12”= 0.77. This, more precise value of U, is 10% higher than the approximate value of 0.70
that resulted from using Case 7. Using the approximate Case 7, instead of Case 2, usually results in
design that is more conservative. In practice, when a section is being designed, since the cross section is
not known at the time of design calculating x is not possible. In these cases, we use U values given for
Case 7.

4.11. h. Discussion and Application of Case 8 of U in Table 4.1

(Ref.: AISC Specification, (AISC, 2010a), Table D3.1, Page 16.1-28))
Case 8 in Table 4.1 (Page 4-23) is specifically for single and double angles where one leg
of the angle(s) is connected to the support with fasteners (bolts or rivets). The value of U in this
case depends on the number of bolts used in the connection as given in the following:

a. with two fasteners (bolts or rivets) in direction of load, Figure 4.29(a):

U = 1- x/L
b. with three fasteners in direction of load, Figure 4.29(b):
U = 0.60
If Equation: U = 1- x/L gives larger value of U, the larger value can be used in design.
c. with four or more fasteners in direction of load, Figure 4.29(c ):
U = 0.80
If Equation: U = 1- x/L gives larger value of U, the larger value can be used in design.

(a)

U= 1- x/L for single-angles with 2 fasteners per

line in the direction of loading.
(Note that AISC Spec. does not allow the use
of just one fastener in any structural

(c)

(b) U = 0.80 for single-angles with 4 or more

fasteners per line in direction of loading.
U = 0.60 for single-angles with 3 fasteners
per line in the direction of loading.

Figure 4.29. Effective Area for Single Angles Connected to the Support using Bolts
CHAPTER 4 - TENSION MEMBERS 4-41

Numerical Example 4.14: Calculating Effective Net Area, Ae, for a

Case 8 Tension Member in Table 4.1.
Considering only failure modes of Yielding in the Gross Section and Rupture in the Net Section, calculate
maximum factored load (the design axial tensile strength in LRFD) that can be applied to the single-angle
member shown below.

L6× 4×¾, A36 Steel

7/8” dia. bolts in 15/16”

dia. holes
12”

Bolted End Connection

Given:
1. Size of the member: L6x4x¾ (Area = 6.94 in2, t = 0.75 in.)
2. Material of the angle: A36 steel (Fy = 36 ksi, Fu = 58 ksi)
3. Details of bolted end connections.
Required:
To calculate the required axial tensile strength (in LRFD).

Solution:
1. Yielding in the Gross Section:
Tu1 = φt Pn = φt Fy Ag , where;
φt = 0.90, Fy = 36 ksi (given)
Ag = 6.94 in2
Tu1 = φt Pn = φt Fy Ag = 0.90 x 36 ksi x 6.94 in2 = 224.9 kips
Tu1 = 224.9 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Calculate Effective Net Area, Ae:
An = Ag - n(dh+1/16”)(t)
An = 6.94 in2 - 1(15/16“+1/16“)(3/4“) =6.19 in2
To establish U, this problem can be considered a Case 8 in Table 4.1, since we have a single angle
connected to the support using bolts. In this problem, we have five bolts in direction of loading. According
to Figure 4.29 on previous page, we need at least four bolts in direction of loading to use a U = 0.80 for
Case 8. Since we have five bolts in direction of loading, we can use U= 0.80 in this case.

Notice that we could consider this as a Case 2 of “U” cases in Table 4.1 and calculate a more precise
value for U using U = 1- x/L equation. Let us do this and see what would be a more precise value of U:
U =1- x/L = 1- (1.07 in) / (12 in.) = 0.91.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-42

Gusset Plate
y
12”

x x
C.G. of
angle section
Net Section
7/8” dia. bolts in x=1.07”
15/16” dia. holes y

As mentioned in the text (Sec. 4.11.h above), The AISC Specifications allows the use of the larger of these
two U values. So, now that we have calculated a more precise U value for Case 2 and it happens to be
larger than the approximate value obtained from Case 8, let us use the larger value of U= 0.91. Notice
that we could also use U=0.80 and our design would be correct although not as economical as the design
using U of 0.91.
Therefore,
Ae = An U = (6.19 in2) (0.91) = 5.63 in2
Tu2 = φ Pn = φ Fu Ae, where;
φ = 0.75, Fu = 58 ksi (given); and Ae = 5.63 in2
Tu2 = φ Pn = φ Fu Ae = 0.75 x 58 ksi x 5.63 in2
Tu2 = 244.9 kips (due to rupture in the net section)
Smaller of the two values of Tu1 and Tu2 is the governing tensile strength:

Answer: Tu = 224.9 kips (due to yielding in the gross section)

4.12. LIMIT STATE OF BLOCK SHEAR RUPTURE

4.12.a. Block shear rupture under Tension
(Ref: AISC Specification (AISC, 2010a), Sec. J4.3, and Page 16.1-129))

Tension members, in addition to yielding in the gross section and rupture in the net section, have a
third limit state of failure called block shear rupture. As shown in Figure 4.30, this failure mode
occurs when a “block” of material fails due to shear failure on its side(s) and tension failure on its
end.
Figures 4.30(a) and (b) (on the next page) show two ways that the edges of a “block” can
fail in a block shear rupture mode. In both cases, the left vertical edge of the block, which is under
tension force Ttn, ruptures while the horizontal edges either rupture in shear through the net
section, Figure 4.30(a), or yield in shear through the gross section next to the bolt holes, Figure
4.30(b).
Since at this time, research has not established which one of these two failure modes will
occur in any given case, in design, the capacity corresponding to each of these failure modes is
calculated and obviously the smaller capacity will be governing. This is the approach followed in
the AISC Specification (2010a) as discussed below.
Figure 4.31 again shows the two possible failure modes for “block shear rupture” and the
corresponding areas for the failure surfaces. These areas are used below to establish strength of the
tension member for limit state of block shear rupture.
CHAPTER 4 - TENSION MEMBERS 4-43

Block
Block

Vnv/2 Vgv/2
Tnt Tnt

Vnv /2 Vgv/2

Tensile rupture Tensile rupture

Shear rupture Shear yielding

(a) Tensile Rupture and Shear Rupture (b) Tensile Rupture and Shear Yielding

Figure 4.30. Two Failure Modes for Block shear rupture According to the AISC
Specification (2010a)

Tensile rupture along

this net section with Tension fracture
area = Ant Along this net section,
with area = Ant
Shear rupture along these
net sections with area = Anv Shear yielding along these
gross sections with area = Agv

(a) Failure mode of tensile and shear rupture (b) Failure mode of tensile rupture and shear yielding

Figure 4.31. Two Possibilities for Block Shear Rupture and Corresponding Areas

The equation of design for checking failure mode of block shear rupture according to the
AISC Specification (AISC, 2010, Section J4.3) is:

Tu ≤ φ Rn (in LRFD) and Ta ≤ Rn /Ω (in ASD) (4.10)

Where
Tu = required axial tensile strength in LRFD, same as the factored applied tension force.
φ = resistance factor for rupture under tension in LRFD = 0.75
Ω = safety factor for rupture under tension in ASD = 2.0
Rn = available nominal axial strength for limit state of block shear rupture given by:

Rn = 0.6Fu Anv + Ubs Fu Ant ≤ 0.6Fy Agv + Ubs Fu Ant (4.11)

Where
Anv = net area subject to shear, see Figure 4.31(a)
Ant = net area subject to tension, see Figure 4.31(a) and (b)
Agv = gross area subject to shear, see Figure 4.31(b)
Ubs = a factor that depends on how uniform the tensile stress across the section is. Ubs = 1.0
when the tension stress is uniform and Ubs = 0.5 when tensile stress is non-uniform.
T = required allowable axial tensile strength in ASD, same as the un-factored (service) applied
tension force.
The Commentary to the AISC Specification (AISC, 2010a) provides examples of cases
where stress is uniform or non-uniform, see Figure 4.32 below.

Figure 4.32. Figure from the Commentary to the AISC Specification Shows
Examples for Selection of Ubs
CHAPTER 4 - TENSION MEMBERS 4-45

In a given connection, a number of blocks can fail in block shear. Figure 4.33 shows some
examples of block shear rupture in wide flanges, angles, channels, and plates.

Wide flange sandwiched between two Channels connected by their

plates (only flanges connected) flanges only

Tees connected by stem only

Double or single angles

Coped beams in shear Wide flange or channels

connected by web only

Figure 4.33. Examples of “Blocks” in Block Shear Rupture

Numerical Example 4.15: Calculating Strength of Tension Members

Based on Block Shear Rupture
Considering failure modes of Yielding in the Gross Section, Rupture in the Net Section and Block Shear
Rupture calculate maximum factored load (the design axial tensile strength in LRFD) that can be applied to
the double angle member shown below.

12” 7/8” dia. bolts in

15/16” dia. holes

2L6× 4×¾,
A36 Steel

Bolted End Connection

Given:
1. Size of the member: 2L6x4x¾ (Area of one angle = 6.94 in2, t = 0.75 in)
2. Material of angle sections: A36 steel (Fy = 36 ksi, Fu = 58 ksi)
3. Details of the bolted end connections.

Required:
To calculate the required axial tensile strength (in LRFD)

Solution:
1. Yielding in the Gross Section:
This example is the same as the bolted end of Example 8 done earlier, where we established:
Tu1 = 449.7 kips (due to yielding in the gross section)

L=12” 1.5”
2. Rupture in the Net Section:
In Numerical Example 8 earlier, we also established:
Tu2 = 489.8 kips (due to rupture in the net section)

3. Block shear rupture:

Tu3 ≤ φ Rn 3”
Rn = Smaller of Rn1 and Rn2, where;
7/8” dia. bolts in
Rn1 = 0.6Fu Anv + Ubs Fu Ant 15/16” dia. holes
Rn2 = 0.6Fy Agv + Ubs Fu Ant
Agv = (12”+1.5”) (3/4”) (2) = 20.25 in2
Ant = [3”- (1/2 hole) (7/8”+1/16”+1/16”)] (3/4”) (2 angles) = 3.75 in2
Anv = (12”+1.5-4.5x1.0”) (3/4”) (2) =13.5 in2
Ubs =1.0 (since in this case, stress distribution can be assumed uniform, see Fig. 4.31)
Rn1 = 0.6(58 ksi) (13.5 in2) + (1.0) (58 ksi) (3.75 in2) = 687 kips
Rn2 = 0.6(36 ksi) (20.25 in2) + 1.0(58 ksi) (3.75 in2) = 655 kips
Use Rn = 655 kips (smaller of Rn1 and Rn2)
Tu3 = φ Rn = 0.75×655 kips
Tu3 = = 491.0 kips
Smaller of the three values of Tu1, Tu2 and Tu3 is the governing tensile strength:

Answer: Tu = 449.7 kips (due to yielding in the gross section)

4.12. b. Block Shear Rupture of Blocks with “Staggered” Bolt Holes

In some cases, such as those shown in Figure 4.34 below, because staggered boltholes, the “block”
in the block shear rupture is not a well-defined rectangular block. Information on actual behavior
of this type of irregular blocks in block shear rupture is almost non-existent. A rational approach to
the problem can be to use “s2/4g” concept that was used earlier in net sections with staggered holes
in block shear rupture of connections with staggered holes.

s
s

g g

Irregular “block” on the 2L legs

Irregular “block” on the 2C webs

Figure 4.34. Block shear rupture for Irregular Blocks

CHAPTER 4 - TENSION MEMBERS 4-47

Using this concept, the equation of design for checking failure mode of block shear
rupture for blocks with staggered boundaries is the same as for regular blocks as was given earlier
by Equation 4.10 and;
Tu ≤ φ Rn (in LRFD) and Ta ≤ Rn / Ω (in ASD) (4.12)
where
Tu = required axial tensile strength in LRFD, same as the factored applied tension force.
φ = resistance factor for rupture under tension in LRFD = 0.75
T a = required axial allowable strength in ASD, same as the un-factored applied tension force
Ω = safety factor for rupture under tension in ASD = 2.0
Rn = available nominal axial strength due to limit state of block shear rupture given by Equation
4.11 repeated here:

Rn = 0.6 Fu Anv + Ubs Fu Ant ≤ 0.6 Fy Agv + Ubs Fu Ant (same as Eq. 4.11 on Page 4-44)

However, for staggered blocks, in the above equation, the net area, Ant, is given by:

Ant = [Ag - n ( dh+1/16") t + Σ (s2/4g) t ] (4.13)

Note that the above Equation 4.13 is the same as we had earlier as Equation 4.7 when we
discussed net sections with staggered holes.

Numerical Example 4.16: Calculating Strength of Tension Members

Based on Block shear rupture
Considering all three limit states of Yielding in the Gross Section, Rupture in the Net Section and Block
shear rupture calculate maximum factored load (the design axial tensile strength in LRFD) that can be
applied to the double angle tension member shown below.

2L8x6×1 ,
A572, Grade 50 steel

1.5” 8@1.5=12” 1” dia. bolts in 17/16”

dia. holes

3”

1.5”
2L8x6x1,
A572, Grade 50 steel
Bolted End Connection
Given:
1. Size of the member: 2L8x6x1 (Area of one angle =13.0 in2, t =1.0 in., x = 1.65 in.)
2. Material of the angle sections: A572, Grade 50 (Fy = 50 ksi, Fu = 65 ksi)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-48

3. Details of end connections with staggered boltholes.

Required:
To calculate the required axial tensile strength (in LRFD)

Solution:
1. Yielding in the Gross Section:
Tu1 = φt Pn = φt Fy Ag , where;
φt = 0.90, Fy = 50 ksi (given)
Ag = 2x13.0 in2 = 26.0 in2
Tu1 = φt Pn = φt Fy Ag = 0.90 x 50 ksi x 26.0 in2 = 1170.0 kips
Tu1 = 1170.0 kips (due to yielding in the gross section)

2. Rupture in the Net Section:

Calculate Effective Net Area, Ae:
An = Ag - n(dh+1/16”)(t) + Σ (s2/4g)(t)
An = 26.0 in2 – (2 holes)(17/16“+1/16“)(2x1.0“) + 1.52/(4x3”)(2x1.0”) = 21.9 in2
U =1- x/L = 1- (1.65 in) / (12 in.) = 0.86.
Ae = An U = (21.9 in2) (0.86) = 18.9 in2
Tu2= φ Pn = φ Fu Ae, where;
φ = 0.75, Fu = 65 ksi (given); and Ae = 18.9 in2
Tu2= φ Pn = φt Fu Ae = 0.75 x 65 ksi x 18.9 in2
Tu2 = 921.4 kips (due to rupture in the net section)

1.5”
x
8@1.5=12”

3”

1.5”

C. G. of one angle

An
1” dia. Bolts in 2L8x6×1, A572,
3. Block Shear Rupture: 17/16” dia. holes Grade 50 steel
Tu3 ≤ φ Rn
Rn = Smaller of Rn1 and Rn2, where;
Rn1 = 0.6Fu Anv + Ubs Fu Ant
Rn2 = 0.6Fy Agv + Ubs Fu Ant
Anv = [12”+1.5”- (4.5 holes) (17/16”+ 1/16”)] (2x1.0”) =16.9 in2 (see sketch below)
Ant = [3”+1.5”-(1.5 holes) (17/16”+1/16”)] (2x1.0”) + [1.52/ (4x3”)] (2x1.0”) = 6.0 in2 (see sketch below)
Agv = (12”+1.5”) (2x1.0”) = 27.0 in2 (see sketch below)
Ubs =1.0 (since in this case, stress distribution can be assumed uniform, see Fig. 4.31)
Rn1 = 0.6(65 ksi)(16.9 in2) + (1.0)(65 ksi)(6.0 in2) = 1049.1 kips
Rn2 = 0.6(50 ksi)(27.0 in2) + 1.0(65)(6.0 in2) = 1200.0 kips
Use Rn = 1049.1 kips, which is smaller of the Rn1 and Rn2
Tu3 = φ Rn = 0.75× 1049.1 kips = 786.8 kips
Tu3 = 786.8 kips
CHAPTER 4 - TENSION MEMBERS 4-49

1.5” 1” dia. Bolts in 17/16”

8@1.5=12” x
dia. holes

3”
1.5”

Anv C. G. of one angle

Ant 2L8x6×1, A572,
Agv Grade 50 steel

Smaller of the three values of Tu1, Tu2, and Tu3 is the governing tensile strength:

Answer: Tu = Tu3=786.8 kips (due to block shear rupture)

4.12.c. Block Shear Rupture of Welded Connections

(Ref.: AISC Specification (AISC, 2010a), Sec. J4.3, and Page 16.1-129))
Block shear rupture generally a limit state of bolted tension members. However, in some cases, it
can also occur in welded connections of tension members. Figures 4.35 and 4.36 show that block
shear rupture of a double angle and a pipe section tension members connected to the gusset plate
using fillet welds. Information on the actual behavior and failure of welded tension members
undergoing block shear rupture is limited. However, such failures occur and needs to be checked
in design.
Using principles of mechanics and information on block shear rupture of bolted tension
members, the following equations are suggested for establishing strength of welded tension
members due to block shear rupture.

Tu ≤ φ Rn (in LRFD) and Ta ≤ Rn / Ω (in ASD) (4.15)

where
Tu = required axial tensile strength in LRFD, same as the factored applied tension force.
φ = resistance factor for rupture under tension in LRFD = 0.75
Ω = safety factor for rupture under tension in ASD = 2.0
Rn = available nominal axial strength for limit state of block shear rupture given by:

Rn = 0.6Fy Agv + Fu Agt (4.16)

And
Agv = gross area subject to shear
Agt = gross area subject to tension
φ = 0.75 (for LRFD), Ω = 2.0 (for ASD)

Lw
Welds

Anv = net area in shear

This area is also:
Agv = gross area in shear

Ant = net area in tension

Note: Only the top angle is shown. The bottom angle is similar.

Figure 4.35. Block shear rupture of a Welded Double Angle Tension Member

Lw Welds

Ant = net area in tension ≅ 0.0

Anv = net area in shear

This area is also:
Agv = gross area in shear

Figure 4.36. Block shear rupture of a Welded Pipe Section

CHAPTER 4 - TENSION MEMBERS 4-51

Numerical Example 4.17: Calculating Strength of a Welded Tension

Members Based on Block Shear Rupture
Considering all three limit states of Yielding in the Gross Section, Rupture in the Net Section and Block
shear rupture calculate maximum factored load (the design axial tensile strength in LRFD) that can be
applied to the double angle tension member shown below.

2L6× 4×¾, A36 Steel

8” Welds

Welded End Connection

Given:
1. The size of the member: 2L6x4x¾ (Area of one angle = 6.94 in2, t = 0.75 in)
2. Material of angle sections: A36 (Fy = 36 ksi, Fu = 58 ksi)
3. Details of welded end connections.

Required:
To calculate the required axial tensile strength (in LRFD)
Solution:
1. Yielding in the Gross Section:
This Numerical Example is the same as the welded end of Numerical Example 4.8 done earlier, where we
established:
Tu1 = 449.7 kips (due to yielding in the gross section)

Gusset
Plate
8”
y

2. Rupture in the Net Section:

In Numerical Example 4.8, we also established: x x
U =1- x/L = 1- (1.07 in) / (8 in.) = 0.86.
Ae = An U, and an = 13.88 in2 (given)
Ae = An U = (13.88 in2) (0.86) = 11.93 in2 Net Section
Welds
Tu2 = φ Pn = φ Fu Ae, where; y x=1.07”
φ = 0.75, Fu = 58 ksi (given); and Ae = 11.93 in2
Tu2 = φ Pn = φ Fu Ae = 0.75 x 58 ksi x 11.93 in2
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-52

Tu2 = 519.0 kips (due to rupture in the net section of welded angle)

3. Block Shear Rupture:

Tu3 ≤ φRn
Rn = Smaller of Rn1 and Rn2, where;
Rn1 = 0.6Fu Anv + Ubs Fu Ant
Rn2 = 0.6Fy Agv + Ubs Fu Ant
Anv = Agv = 12.0 in2
Agv = (8”) (3/4”) (2) = 12.0 in2
Ant = (6”) (3/4”) (2) =9.0 in2
Ubs =1.0 (see Fig. 4.32 on Page 4-44)
Rn1= 0.6 (58 ksi) (12.0 in2) + (1.0) (58 ksi) (9 in2) = 939.6 kips
Rn2= 0.6 (36 ksi) (12.0 in2) + (1.0) (58 ksi) (9 in2) = 781 kips  Smaller of Rn1 and Rn2
Use R n = 781 kips≤
Tu3 = φ Rn = 0.75×781 kips = 585.9 kips
Tu3= 585.9 kips

8” Gusset Plate
y
Anv = Agv

x x

2L6x4x3/4,
Welds
A36 steel y Ant
“Block” for block x=1.07”
shear rupture

Smaller of the three values of values of Tu1,Tu2 and Tu3 is the governing tensile strength:

Answer: Tu = 449.7 kips (due to yielding of gross area)

4.13. LIMIT STATE OF FAILURE OF CONNECTIONS

As mentioned earlier, failure of the connections in a tension member should also be considered in
its design. In the current AISC Specification (AISC, 2010a), failure modes of connections are
treated in Chapter J of the Specification. We will also discuss the connection failure modes in later
chapters of this document.

4.14. NUMERICAL EXAMPLES FOR COMPLETE DESIGN

OF STEEL TENSION MEMBERS
This section provides numerical examples that demonstrate application of concepts, design
procedures, and code provisions to complete design and evaluation of strength of steel members
subjected to tension. In the previous section, at the end of each main item of discussion, a
numerical example demonstrated the application of that particular concept. This section provides
CHAPTER 4 - TENSION MEMBERS 4-53

Numerical Examples that show application of all the concepts, limit states, and code provisions in
design of tension members.
Most numerical examples given in this section are on design of tension members, where the
applied loads are given and the problem statement asks for selection of the material and cross
section that can resist the applied loads. A few numerical examples are on the evaluation of
strength of a tension member. In these cases, the cross section and material properties are given
and the objective is to evaluate the member and establish its tensile strength. This was very much
what we did in the Numerical Examples in previous sections. In both cases, the LRFD methods
given in the AISC Specification (AISC, 2010a), which were discussed in this chapter, are used.

Numerical Example 4.18:

Design of a Tension Member in a Truss
Consider the bottom chord member of the truss shown below. The analysis of the truss has resulted in the
un-factored tensile dead and live loads in the member as shown in the figure. Consider all applicable limit
states of tension members and using factored load combination Tu = 1230 kips and the LRFD methods,
design a wide flange section for this truss chord to carry the load Tu. In your design, also satisfy the
preferred slenderness limit of L /r ≤ 300.
Detail L6

6@10’-0”= 60’-0”

Tension chord member to be designed.

The factored combined tension force in the member,
based on the analysis of the truss, is 1230 kips.

1” dia. bolts in
17/16” dia. Holes
c/c of bolts = 3 in.

Chord member to be
designed.

Detail L6
(For location of Detail 6, see Page 4-52)

Given:
1. Geometry of the truss
2. Maximum factored tension force in the member to be designed as: Tu = 1230 kips
3. Details of the bolted truss joint

Required:
1. To design the bottom chord member to resist the applied loads
2. Check the preferred slenderness of the member as per AISC: L/r ≤ 300.

Solution:
Note that this is a Design case in LRFD, where factored tension force Tu in the member is known
and we are to select a steel cross section for the member that has a design tensile strength , φt Pn , greater
than or equal to Tu.
Let us use a hot-rolled wide flange (W) section for the bottom chord, which is a common cross
section used for members of bridge trusses.
To select grade of steel, currently the most appropriate material of steel for hot-rolled wide flange
shapes is A992 Grade 50. We could also use A913 Gr. 65, however, A913 is not as common as the A992
steel. As discussed in Chapter 2, A36 steel is not a preferred selection of steel for hot-rolled wide flange
shapes. So, let us use a wide flange section made of A992 Grade 50 steel.
We will design the cross section considering limit state of yielding in the gross section and then
check it for other two limit states (rupture in the net section and block shear rupture.)

1. Design for Yielding in the Gross Section:

The design equation for yielding in the gross section is:
Tu ≤ φ Pn = φ Fy Ag
If we rearrange this equation, the gross area needed is:
Ag ≥ Tu / (φ Fy), where,
φ = 0.90, Fy = 50 ksi for A992 grade 50
Ag ≥ Tu / (φ Fy )
Ag ≥ 1230 kips /(0.9x50 ksi)
Ag ≥ 27.3 in2

We need to try a section that has a gross area not less than 27.3 in2. Let us try a W12x96 section which
has a gross area Ag = 28.2 in2.
The actual design strength of the selected section in LRFD is:
Tu1 = φPn =φ Fy Ag= 0.90 x 50 ksi x 28.2 in2 = 1269 kips
The above strength is greater than the factored applied load of 1230 kips, therefore as expected W12x96,
A992, Gr. 50 steel section is adequate for limit state of yielding in gross section.
Let us now check the limit state of rupture in the net section:

2. Check Rupture in the Net Section:

To check the net section we need to know details of the connection. For this case, we could use welded or
bolted connections. The design of connection is discussed in later chapters of this book. Let us say that we
have decided to use a bolted connection as shown in the “Detail L6” sketch below.
Now, we have enough information to calculate tensile rupture strength of the net section.

Calculate Effective Net Area, Ae:

An = Ag - n(dh +1/16”) (tf)
For W12x96: tf = 0.90 in.
CHAPTER 4 - TENSION MEMBERS 4-55

An = 28.2 in2 – (4 holes)(17/16“+1/16“)(0.90“) = 24.2 in2

U =1- x/L = 1- (1.13 in) / (8x3 in.) = 0.95.
Ae = An U = (24.2 in2)( 0.95)= 23.0 in2
Tu2 = φPn = φFuAe, where;
φ = 0.75, Fu = 65 ksi (for A992, Gr 50 steel); and Ae = 23.0 in2
Tu2 = φ Pn = φ Fu Ae = 0.75 x 65 ksi x 23.0 in2
Tu2 = 1121.3 kips (due to rupture in the net section)

1” dia. bolts in
17/16” dia. Holes W section
c/c of bolts = 3 in. (All verticals
and diagonals)

Double
Gusset
Chord Plates
member to
be designed.

8@3 in.
x C.G. of shaded area
(i.e. WT section made
Detail L6 of ½ of wide flange)
(For location of Detail 6, see Page 4-52)

Since Tu, the design strength is less than applied factored load of 1230 kips, the design is not acceptable.
We need to try a section with larger Ag.
To select next trial section, we can calculate an approximate value for required Ag of that section as:
Ag = (1230 kips / 1121.3 kips) x 28.2 in2 = 31.0 in 2
Let us try W12x106 (Ag =31.2 in 2, tf = 0.99 in. and x = 1.19 in.)
Calculate Effective Net Area, Ae for new trial section (W12x106):
An = Ag - n(dh+1/16”)(tf)
An = 31.2 in2 – (4 holes)(17/16“+1/16“)(0.99“) = 26.75 in2
U =1- x/L = 1- (1.19 in) / (8x3 in.) = 0.95
Ae = An U = (26.75 in2)( 0.95)= 25.4 in2
Tu2 = φPn = φ Fu Ae, where;
φ = 0.75, Fu = 65 ksi (for A992, Gr 50 steel); and Ae = 25.4 in2
Tu2 =φPn = φ Fu Ae = 0.75 x 65 ksi x 25.4 in2
Tu2 =1238 kips (due to rupture in the net section)

The above strength is greater than the factored applied load of 1230 kips, therefore the trial section,
W12x106, A992, Gr. 50 steel is adequate for both limit states of yielding in gross section and rupture in the
net section.
Let us now check limit state of block shear rupture:

3. Check Block shear rupture of W12x106:

To check block shear rupture, we need to know the length of connection. Therefore, we need to design
the connection first, and then check the block shear rupture. Let us say we have designed the connection
and the dimensions are as shown in the sketch below:

Gusset
8@3=24” 1.5” Plates

2”

Anv W12x106,
All Bolts: 1” dia. in A992
Ant Agv
17/16” dia. holes Gr. 50 steel

Tu3 = φRn
Rn = Smaller of Rn1 and Rn2, where;
Rn1 = 0.6Fu Anv + Ubs Fu Ant
Rn2 = 0.6Fy Agv + Ubs Fu Ant
Anv = [24”+1.5”- (8.5) (17/16”+ 1/16”)] (0.99”) (4) = 63.11 in2 (see sketch above for Anv)
Ant =[2”-(0.5)(17/16”+1/16”)](0.99”)](4) = 5.77 in2 (see sketch above for Ant)
Agv = (24”+1.5”) (0.99”) (4) = 101.0 in2 (see sketch above for Agv)
Ubs =1.0 (since in this case, stress distribution can be assumed uniform, see Fig. 4.32)
Rn1 = 0.6(65 ksi) (63.11 in2) + (1.0) (65 ksi) (5.77 in2) = 2836 kips
Rn2 = 0.6(50 ksi) (101.0 in2) + 1.0(65 ksi) (5.77 in2) = 3405 kips
USE Rn= 2836 kips, which is smaller of the Rn1 and Rn2.
Tu3 = φ Rn = 0.75× 2836 kips = 2127 kips
The above strength is greater than the factored applied load of 1230 kips, therefore the trial section,
W12x106, A992, Gr. 50 steel is adequate for all three limit states of yielding in gross section, rupture in the
net section, and block shear rupture.

4. Check the Slenderness:

L = 10 ft. (Given)
rmin for W12x106 = 3.11 in. (This is ry of W12x106 given in the cross section properties tables).
L / r = (10ft) (12 in/ft) / 3.11 in = 39 < 300
The AISC preferred limit of slenderness ratio for tension members is satisfied.
The design is final. USE W12x106, A992 Grade 50 Steel

Numerical Example 4.19:

Design of a Tension Member in a Braced Frame
Use LRFD methods and design an HSS bracing member subjected to combined factored load of 510 kips.
The member is welded to the gusset plates as shown in the figure.

510 kips 510 kips

Length = 34’-0” Centerline of

the Connection
CHAPTER 4 - TENSION MEMBERS 4-57

Gusset plate
Gap at the end Gap at the
of the slot end of the
slot

HSS section
Fillet welds

Given:
1. Geometry of the member
2. Maximum factored tension force in the member
3. Details of end connections being welded connection to single gusset plate.

Required:
1. to design the HSS bracing member to resist the applied loads
2. Check the preferred slenderness of the member as per AISC: L/r ≤ 300.

Solution:
Note that, like the previous Numerical Example, this is a Design case in LRFD, where factored
tension force Tu in the member is known and we are to select a steel cross section for the member that has
a design tensile strength , φt Pn , greater than or equal to Tu.
In this case, we want to use an HSS section. We can use either rectangular or pipe section. Let us
use a rectangular HSS.
To select grade of steel, currently the most appropriate material of steel for HSS is A500 Grade B (Fy = 46
ksi and Fu = 58 ksi)
We will design the cross section considering limit state of yielding in the gross section and then check
it for two other limit states (rupture in the net section and block shear rupture.)

1. Design for Yielding in the Gross Section:

The design equation for yielding in the gross section is:
Tu ≤ φt Pn = φt Fy Ag,
If we rearrange this equation, the gross area needed is:
Ag ≥ Tu / (φt Fy), where
φt = 0.90, Tu = 510 kips, and Fy = 46 ksi (for A500 Grade B)
Ag ≥ Tu / (φt Fy ) = 510 /(0.9x46 ksi)
Ag ≥ 12.3 in2
Let us try an HSS8x8x1/2, A500, Grade B Steel, with t = 0.465 in. (for design taken from the AISC Manual
Tables for HSS8x8x1/2), Ag =13.5 in2. rmin = 3.04 in. Fy = 46 ksi, and Fu = 58 ksi.

2. Check Rupture in the Net Section:

To check the net section we need to establish U, the shear lag factor. To establish U we need to choose
the details of the connection. For this case, the common economical connection detail is to weld the HSS
member to a single gusset plate as shown in the sketch below. Notice that this detail creates a net section
in the member due to presence of the “gap” between the end of the slot in the HSS and the gusset plate
(see sketch below).

tg =1.0”
Net Section
Gaps

t = 0.465”
H = 8” (for design
calculations)

C.G. of the right

half of the HSS
Weld,
L=14”
x
B =8”

Calculate Effective Net Area, Ae:

Ae = An U
For this design, Case 6 of Table 4.1, given earlier, applies. To establish An, and U, we need to have more
information on the connection. We need thickness of gusset plate and length of welds. Let us assume a 1
in. thick gusset plate will be used and let us take the length of weld to be about 1.5 times the “H” dimension
of the rectangular tube. This means a length of weld equal to 14 inches. With these assumptions we will
design the member and after we designed the gusset plates and welds, if the final values for tg and L are
less than the assumed values, we need to recheck the design of the member using final values of tg and L.
Ae = An U
An = Ag - 2 × t × tg = 13.5 in2 - 2×0.465”x1”=12.6 in2
U =1-x/L where
x= (B2 + 2BH)/ [4(B + H)] = (82 +2x8x8)/ (4x16) = 3 in.
U =1- 3“/14“ = 0.78
Ae = An U= 0.78 ×12.6 in2 = 9.8 in2
T u1 = φt Fu Ae = 0.75× 58ksi ×9.8
T u1 = 427.6 (due to rupture in the net section)
Since T u1 = 427.6 kips, the design strength is less than the applied factored load of 510 kips, the design is
not acceptable. We need to redesign and increase design strength of the net section.
We can increase the net section strength in two ways. One is to change the member section to a heavier
section. The other solution is to keep the HSS8x8x1/2 and increase its net area by welding plates to the
side of the HSS tube. We will use the first solution since this can be more economical compared to
relatively more labor intensive solution of welding plates to the side of the tube.
To select next trial section, we can calculate an approximate value for required Ag as:
Ag = (510 kips / 424 kips) x 13.5 in2 = 16.2 in 2
Let us try HSS10x10x1/2, A500 Gr. B (t =0.465 in., Ag =17.2 in2, r = 3.86 in., Fy = 46 ksi, Fu = 58 ksi.)
Since area of HSS10x10x1/2 is larger than 8x8x1/2, by observation there is no need to check yielding in
the gross section of this new section. However, if we wanted to check, here is the calculation:
φt Fy Ag =0.90 × 46ksi × 17.2 in2 = 712 kips > 510 kips. O.K.

Check Rupture in the Net Section of the New Trial Section HSS10x10x1/2:
Calculate Effective Net Area, Ae for new trial section (HSS10x10x1/2):
Ae = An U,
CHAPTER 4 - TENSION MEMBERS 4-59

An = (Ag - 2 × t × tg) = (17.2 in2 - 2 × 0.465” x 1”= 16.3 in2 ,

For this trial design, still Case 6 of shear lag in Table 4.1 applies.
Since L > H, then: U= 1-x/L = 1-[(B2 + 2BH)/4(B + H)]/L
U =1 - [(102 +2x10x10) / (4x20)] / 14= 0.73
Ae= An U=0.73×16.3 in2 =11.9 in2
Tu2 = φ Pn = φ Fu Ae, where;
φ = 0.75, Fu = 58 ksi (for A500 Gr. B steel); and Ae = 11.9 in2
φ Pn = φ Fu Ae = 0.75 x 58 ksi x 11.9 in2 = 518k (due to rupture in the net section of HSS10x10x1/2)
The above design strength of 518 kips is greater than the factored applied load of 510 kips, therefore the
trial section, HSS10x10x1/2, A500, Gr. B steel is adequate for both limit states of yielding in the gross
section and rupture in the net section.
Let us now check the limit state of block shear rupture:

3. Check Block shear rupture in the HSS10x10x1/2:

We need to redesign and increase design strength of the net section.
T u3 = φ Rn
Rn = Smaller of Rn1 and Rn2, where;
Rn1 = 0.6Fu Anv + Ubs Fu Ant
Rn2 = 0.6Fy Agv + Ubs Fu Ant
Anv = Agv = (14“)(0.50“)(2)(2)= 28 in2
Ant = 0.0 in2

tg =1.0”
Ant =0.0

B = 10”
t =1/2”
B=10”
C.G. of the
right half of
Agv = Anv the HSS

L=14”
HSS10x10x1/2, x
A500, Gr. B steel H=10”

Ubs =1.0 (since in this case, stress distribution can be

assumed uniform, see Fig. 4.31). In any event, since
Ant = 0.0 Ubs has no effect on calculations.
Rn1 = 0.6 Fu Anv = 0.6x 58 ksi x28 in2 = 974 kips
Rn2 = 0.6 Fy Ang = 0.6x 42 ksi x28 in2 = 772 kips
USE Rn = 772 kips, which is smaller of the Rn1 and Rn2.
T u3 = φ Rn = 0.75 × 772 kips = 579 kips
The above strength is greater than the factored applied load of 510 kips, therefore the trial section,
HSS10x10x1/2, A500, Gr. B steel is adequate for all three limit states of yielding in the gross section,
rupture in the net section and block shear rupture

4. Check the Slenderness:

L = 34 ft (given)

rmin for HSS10x10x1/2 is equal to 3.86 in. (Given in the cross section properties tables).
L/r = (34 ft) (12 in/ft) / 3.86 in = 105.7 < 300 O.K.
The AISC preferred limit of slenderness ratio for tension members is satisfied.
The design is final.
USE HSS10x10x1/2, A500, Gr. B Steel

References for Chapter 4

1. AISC, (2010a). “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An
American National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-10,
An American National Standard, Am. Institute of Steel Construction, Chicago.
3. AISC, (2011). “Steel Construction Manual,” 14th Edition, American Institute of Steel
Construction, Chicago.
CHAPTER 4 - TENSION MEMBERS 4-61

PROBLEMS ON TENSION MEMBERS

Problem 4.1. Check members and by using the LRFD methods calculate design strength, φRn of the tension
members shown below. Failure modes of the bolts, welds, and gusset plates are not part of this problem.
Also, check the “preferred” limit for slenderness (L/r ≤ 300).

W21x111, A992 Gr. 50 Length Pipe 10 x-Strong, A53 Gr. B

= 20 ft Length= 18 ft

Welded all around

Welded all around
(a) (b)

22’-6”

3/4” thick gusset plate

All bolts are 1” diameter in 17/16” holes

3”

2” 4@3”=12” 2L6x5x1/2, A36

1.5” 1.5”
(c)

12’-6”

1-1/2” thick gusset plate

All bolts are 1” diameter in 17/16” holes
1”

3”
3” 3”
2” 2”

1” 5@3”=15” 2L8x6x1,
A572, Grade 50
1.5” 1.5”

(d)

12’-6”

PL8x3/4, A36 Gusset Plate

All bolts are 3/4” diameter in 13/16” holes
1”
2”
2½”
1½”
4@3”=12”
2L6x4x1/2, A36
1½” 1½”

(e)

Problem 4.2. Using LRFD methods calculate design strength, φRn of the tension members shown below.
Failure modes of the bolts or welds are not part of this problem.

18’-6”

1” Thick gusset plate

All bolts are 7/8” diameter in 1” holes
1”
2-1/2”
3” 3”
1-1/2” 1-1/2”
1” 5@3”=15”
2L7x4x1/2,
1-1/2” 1-1/2” A572, Grade 50

(a)

Splice Plates: 2PL12x1/2, A572, Grade 50

All bolts are 7/8” diameter in 1” holes
1-1/2” Welds

3” 12”
1-1/2”
1-1/2”

4@3”=12” 6”
W16x57
1-1/2” 1-1/2” A992, Grade 50
(b)
CHAPTER 4 - TENSION MEMBERS 4-63

Splice Plates: 2PL12x1/2, A572, Grade 50

All bolts are 7/8” diameter in 1” holes
1-1/2” Welds

3”
1-1/2”
1-1/2”

4@3”=12 W14x53
” A992, Grade
1-1/2” 1-1/2”
(c)

Splice Plates: 2PL16x1/2, A572, Grade 50

All bolts are 7/8” diameter in 15/16” holes

Welds

2” 5@3”=15” 2” W14x257
A992, Grade 50
2”
3”
3”

(d)

1.0 inch
Thick Gusset
Plate, A572
1.0 inch Thick 7/8” diameter A490X 5/16”
Gusset Plate, 2C12x25
bolts in 15/16” A572 Gr. 50 Fillet Welds E70xx
A36 standard round holes

3”

Tu 3” Tu
18” 3” 12”
18”
3”
3”
3”
4@3 “ =12” 18”
2” 2”

(e)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-64

Problem 4.3. A double channel tension member is connected to the end gusset plates by welds or bolts as
shown below. Check failure modes of the channels and angles (if resent) and calculate factored load φPn that
can be applied to the member.
1” thick gusset plate 1” thick gusset plate
All bolts are 7/8” Welds
diameter in 1” holes

3”

1-1/2”

3@3” 9” 2C15x50,
2C15x50,
A572, Gr. 50 A572, Gr. 50
1-1/2”
(a) (b)

1” thick gusset plate 1” thick gusset plate

All bolts are 7/8”
diameter in 1” holes 2L3-1/2x3-1/2x1/2
2L4x4x1/2 Welds

3”

2C15x50, 2C15x50,
A572, Gr. 50 A572, Gr. 50
8@1-1/2” 12”

1” thick gusset plate

1” thick gusset plate
All bolts are 7/8”
diameter in Welds
15/16” holes

3”

1½”

4@3” 12” 2C15x50, A36

2C15x50, A36

1½
”
1½”

(e) (f)
CHAPTER 4 - TENSION MEMBERS 4-65

Problem 4.4. A double WT tension member is bolted to one end and welded to the other end, see details
below. Check failure modes of the WT sections as well as the plates and find maximum factored load φPn
that can be applied to the member.

PL14x1-1/2”, PL16x1-1/2”,
A572, Gr. 50 A572, Gr. 50
All bolts are 1” dia. Weld
in 17/16” holes

3”
2”

2 WT12x96 14”
6@1-1/2” A992, Gr. 50
2” 2”

Problem 4.5. A wide-flange tension member is attached to gusset plates by four angles as shown. Using
information provided, calculate maximum factored tension load that can be applied to the member. Check
any failure mode of the member and angles that you can. Checking the gusset plate is not part of this
problem. 2L4x4x1/2,
A36 All bolts are 7/8” dia.
in 1” holes
2PL9x1/2x1’-1”

1-1/2”
3”

W14x109,
A992, Gr. 50
1-1/2” 4@3”= 2” 2” 4@3”

1-1/2”

Problem 4.6. Using LRFD methods calculate design strength, φRn of the tension members shown below.
Failure modes of the bolts or welds are not part of this problem.

3/4” Thick gusset plate

All bolts are ¾” diameter in 13/16” holes
2L6x4x1/2, A36
2”
2”
2”
3”
3” 3”
2”
2L6x6x1/2, A36
3@3”

2” 2”
(a)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-66

20’-0”

1” Thick gusset plate

All bolts are ¾” diameter in 13/16” holes
1.5”
2” 2L8x6x3/4, A36
4”
3”
3” 3” 3” 2”
2” 2”
1.5” 10@1.5” 2L8x8x3/4, A36

1.5” 1.5”

(b)

Problem 4.7. Considering all applicable failure modes is the tension member shown below adequate to
carry the factored tension loads? Connection plates, welds, and bolts are not part of this problem.
260 kips 260 kips

2PL12”, A36
All bolts are 7/8” PL 10”x1”x2’-3”, A36
diameter
in 1” holes Welds
1”
No gap here
2”
3”

HSS6x5x1/2”
2” 4@3”=12” 2” 10” A500 Gr. B

1”
(A)

3/4” Thick gusset plate

All bolts are ¾” diameter in 13/16” holes

1½” 2L4x3x5/16, A36

1½”
1½”
2”
2” 2”
1½”
2L4x4x5/16,
5@1½ ” A572,Gr 50

1½” 1½”
(B)
CHAPTER 4 - TENSION MEMBERS 4-67

Problem 4.8. Consider tension members with end connections shown below. Design the members for a
factored load of 350 kips. Select the length of the weld if needed. Notice that design of the welds is not part
of this problem. Also, check the preferred limit of L/r ≤ 300. Length of all members is 16 feet.
1” thick gusset plate
1” thick gusset plate
Welds
Welds
Gap here

Rectangular 2C
or
Square HSS
(A) (B)

1” thick gusset plate 1” thick gusset plate

Welds Welds

Gap here

Standard Pipe
2C

Welds on both flanges

in addition, web Welds connecting the
stem only

W
WT
(E) (F)

Welds connecting the

web only Welds connecting the
stem only

W WT

(G) (H)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
4-68

Welds Welds

2L 2L

(I) (J)

Welds
Welds

2PL
2PL

(K (L)

Welds

2PL

(M)
CHAPTER 5-COMPRESSION MEMBERS
5-1

CHAPTER 5
COMPRESSION MEMBERS

Objectives:
In this chapter, you will learn:
1. Definition of axially loaded compression members or columns
2. Cross sections commonly used as compression members in steel
structures.
3. Behavior of axially loaded steel compression members
4. Failure modes (or limit states) of steel compression members.
5. Design of axially loaded steel compression members.
6. Application of concepts and equations to actual design of compression
members through the use of numerical examples provided throughout
the chapter.

5.1. INTRODUCTION
This chapter is devoted to steel members subjected to axial compressive force. As shown in
Figure 5.1, steel compression members quite often are columns in buildings and piers in bridges,
as members of trusses, as arches, and as bracing members in braced frames.
In axially loaded compression members, the axial force passes through the centroid of the
cross section and no bending moment exists. In reality, in most cases, there can be some moment
present, either due to eccentricity of the applied load or due to eccentricity of the reactions at the
supports. If these moments are small, say less than 10% of bending moment capacity of the cross
section, they are ignored since such small amounts of bending moment does not reduce axial
strength of the column more than a small and negligible percentage. For larger bending moments
acting on a compression member, the member should be considered a “beam-column” subjected to
combined effects of bending and axial load. Beam columns are discussed later in Chapter 8 of this
document.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
5-2

(a) COLUMNS BUILDINGS (b) BRACING MEMBERS

Compression
Members

(d) VERTICAL AND DIAGONAL

(e) ARCHES

(f) TRANSMISSION
TOWERS

Steel Tower

Cable-Stayed Bridges

Steel Tower

Steel Suspension Bridges

(g) TOWERS IN CABLE-SUPPORTED BRIDGES Photo: California Public Works Department, 1934

Figure 5.1. Typical Axially Loaded Compression Members in Steel Structures

CHAPTER 5-COMPRESSION MEMBERS
5-3

5.2. STEEL AXIALLY LOADED COMPRESSION MEMBERS

Steel axially loaded compression members (columns) are usually made of rolled shapes, cold-
formed hollow Steel Sections (HSS), or built-up sections as shown in Figure 5.2. With regard to
behavior, buckling modes and design equations given in the AISC Specification (2010a),
compression member are divided into five categories shown in Figure 5.2. The five categories are:
(a) doubly-symmetric cross sections with two axes of symmetry, Figure 5.2(a)
(b) built-up sections made of two or more separated segments connected to each
other at intermediate points, Figure 5.2(b)
(c) singly symmetric cross sections with only one axis of symmetry, Figure 5.2(c)
(d) unsymmetrical cross sections with no axis of symmetry, Figure 5.2(d)
(e) single angles loaded through one leg, Figure 5.2(e)

In this Chapter, we discuss Categories (a) and (b) which are the most common compression
members. For Categories(c) through (e), the reader is referred to the AISC Specification (AISC,
2010a).

WF HP Built-up HSS Pipe

Built-up Built-up Built-up
Tube
(a) Doubly Symmetric Sections

Batten Plate
Stitch

2-L
2-C 2-C (d) Un-symmetric
2-W
Sections
(b) Built-up Sections with Separated Segments

L
Channel Tee 2-L L
Built-up (no separation) (Equal Leg) (e) Single Angles
Loaded Through
(c) Singly Symmetric Sections One Leg

Figure 5.2. Examples of Steel Cross Sections Used in Compression Members

In developing design procedures for axially loaded compression members, the following
assumptions have been made:
a. Members are prismatic.
b. Only axial compression force is present and is acting through the centroid of the
cross section.
c. The axial force is applied gradually and can be considered static load.

5.3. BEHAVIOR OF AXIALLY LOADED STEEL

COMPRESSION MEMBERS
When a steel member is subjected to compression force acting through its centroid, the applied
force creates uniformly distributed normal stresses on the cross sections, Figure 5.3(a). These
stresses combined with the internal residual stresses, shown in Figure 5.3(b), create a non-uniform
stress distribution, Figure 5.3(c). Residual stresses were discussed in Chapter 2 (see Page 2-8.)

Tension
Comp.
+ Comp.
Comp. = σmax
Comp.

(a) Applied Stresses (b) Residual Stresses (c) Total Stresses

Figure 5.3. Typical Stresses Acting on Cross Sections of a Compressive Member

σmax

Flange Web
Local Buckling Local Buckling
(a) (b) (c)

Figure 5.4. Local Buckling of Cross Section Where Compressive Stresses Were High
CHAPTER 5-COMPRESSION MEMBERS
5-5

Let us consider the short wide flange column shown in Figure 5.4(a) with non-uniform
internal stresses. If we increase the axial load, the internal stresses also will increase. When the
maximum compressive stress, σ max, Figure 5.4(a), reaches a critical buckling stress level, local
buckling of flange shown in Figures 5.4(b) can occur. A similar condition occurs in the web
causing web local buckling, shown in Figure 5.4(c) below. Depending on the geometry of the
cross section and material properties, in some cases, the critical local buckling stress is greater than
the yield stress. In these cases, before the cross section buckles locally, the maximum stress σ max
reaches yield point and section starts yielding prior to local buckling.
If local buckling occurs prior to yielding of any fiber, it will be elastic local buckling. If
cross section does not develop local buckling prior to yielding and continues to accept more
compressive load until σ max reaches yielding, then as the load continues to increase more and more
areas of cross section will yield. After sufficiently large areas of the cross section have yielded,
the cross section can still develop local buckling which in this case will be inelastic local buckling
since it happened after parts of the cross section have yielded.
Local buckling of compressive members is a limit state of failure. The phenomenon of
local buckling is discussed in more details later in this chapter. In dependent of local buckling,
compressive members can reach overall buckling failure mode. The phenomenon of overall
buckling is also discussed in more details later in this chapter.
When compression force is applied to a steel member, first, the member shortens
elastically with an axial stiffness of EA/L where E is modulus of elasticity of steel, A is the gross
area of cross section, and L is the length of the member. When the limit state of local or overall
buckling is reached, the load drops as shown in Figure 5.5. The rate in which the load drops is an
indicator of ductility of the member. When a member reaches local buckling or overall buckling
capacity, if the load is maintained for sometime before it drops, then, the behavior can be
categorized as ductile. This is shown by Curve A in Figure 5.5. However, if after reaching overall
or local buckling capacity, the load drops rapidly and stiffness of the member (i.e. the slope of the
force-displacement curve) becomes negative, as is the case with Curve B in Figure 5.5, then the
behavior will be brittle.
Post-buckling ductility (ability to deform inelastically) of a compression member strongly
depends on slenderness of the member. Stocky and compact members, with small slenderness
ratios, KL/r, undergo inelastic buckling and demonstrate greater post-buckling ductility. More
discussion of slenderness ratio is given later in this chapter.

Compressive Curve A:
Force Ductile Behavior

Curve B:
Brittle Behavior

Axial Deformation

Figure 5.5. Typical Ductile and Brittle Behavior of Compression Members

As always, in design of compression members, in addition to failure modes of local and

overall buckling, failure of connections also needs to be considered a failure mode. Connections
are discussed in later chapters of this document.
Behavior of compression members with riveted end connections is more or less similar to
the behavior of members with welded connections discussed above. If instead of rivets, high-
strength bolts are used, if the bolts are in shear, they can slip at some point while the member is
being loaded resulting in a “slippage” step in the force-displacement curve as shown in Figure 5.6.
The slippage causes a temporary loss of stiffness since the slope of the load-deflection curve, when
slippage is occurring, is horizontal which indicates zero stiffness. The slippage can add to
ductility of the system as can be seen in Figure 5.6. To avoid the slippage, we can use “slip-
critical” bolts. We will discuss these and other types of bolts later in chapter on bolts.

Compressive Welded or Riveted End

Force Connections

Bolted End Connections

(Bolts in Shear Slip)

Slippage of
Bolts

Axial Deformation

Figure 5.6. Comparison of Compression Members with Bolted and Welded Connections

In summary, when a steel member is subjected to ever-increasing axial compressive force,

it can reach one of the following limit states or failure modes. When a failure mode is reached,
depending on the nature of the failure mode, the strength or stiffness of the member or both is
reduced significantly and the member is called "failed.”
The main limit states of failure of an axially loaded compression member are:
1. Local Buckling, Figure 5.7(a)
2. Overall Buckling, Figure 5.7(b)
3. Failure of Connections
In compression members, overall buckling failure mode is preferred to local buckling.
For columns, design specifications such as the AISC Specification (2010a), limit the width-to-
thickness ratio of the elements of the cross sections to certain values to prevent local buckling
prior to overall buckling of the member. More on local buckling failure mode is provided in the
following section.
In compression members, the presence of bolt holes does not affect the buckling capacity.
Therefore, if end connections of a compression member have bolt holes, those bolt holes are filled
with bolts and the compressive area is the gross area of the cross section.
CHAPTER 5-COMPRESSION MEMBERS
5-7

Local Overall
Buckling of Local Buckling of
Flange Buckling of Entire Column
Web

a. Local Buckling of a Cross Section b. Overall Buckling of a Member

Figure 5.7. Buckling Failure Modes of Compression Members

5.4. LIMIT STATE OF LOCAL BUCKLING OF AXIALLY

LOADED MEMBERS
5.4. a. Introduction to Local Buckling

When a steel member is subjected to axial compression, if the applied axial stress in any element
of the cross section reaches a critical stress, σ cr, level, the element can locally buckle, Figure 5.8.

σcr

Column Local Buckling Local Buckling

Local Buckling
of Flanges of the Web
of Both Flanges
and the Web
Figure 5.8. Local Buckling of Flanges and Web under Assumed Uniform Axial Stress
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
5-8

The critical stress in local buckling depends on the width-to-thickness ratio, b/t, of the
elements of the cross section, yield stress F y , modulus of elasticity E, Poisson’s ratio,ν and the
boundary condition of the plate element that can locally buckle. Using plate-buckling theories,
critical stress for elastic local buckling of a plate is established as:

Kπ E
2

σcr =
b 2
12 (1 − ν ) ( )
2
(5.1)
t
Where
K = a coefficient that depends on the boundary conditions of the plate and type of the applied
stress, Figure 5.9.
E = modulus of elasticity of steel = 29,000 ksi.
ν = Poisson’s ratio equal to 0.3 for steel behaving elastically
b = width of the plate
t = thickness of the plate
Values of K in Equation 5.1 for plates with various boundary conditions subjected to
uniform compressive stresses are given in Figure 5.9. Curves for Case (c), with K min = 4.0, and
Curves for Case (d), with K min = 1.277, represent the boundary conditions for local buckling of
web and flanges in an I-shaped column respectively. Values of b and t in the above Equation 5.1,
for common cross sections, are defined by the AISC Specification (2010a) and shown in Figure
5.10.
a
K

Kπ E
2
10
σcr = Fixed
2 b 2 b
12 (1 − ν ) ( ) Fixed
t
Case (a)
Case (a), Fixed-Fixed
Kmin = 6.97

Case (b), Fixed-Simple Kmin = 5.42 Fixed Simple

5 Simple Simple
Case (c), Simple-Simple
Kmin = 4.00
Case (b) Case (c)

Fixed Simple
Case (d), Fixed-Free Kmin = 1.277
Free Free
Case (e), Simple-Free Kmin = 0.425
Case (d) Case (e)

0 1 2 3 4 5

a/b

.
Figure 5.9. Values of K for Elastic Buckling of Plates (Data for the graphs from Gerard and Becker (1957)
CHAPTER 5-COMPRESSION MEMBERS
5-9
b
b
b b t
t
t b
t b
b t
t t
b

(a) Unstiffened Elements

b b b
t
t
tw D
tw tw t
h h t b
b h
t
t b
t

(b) Stiffened Elements

Figure 5.10. Examples of Unstiffened and Stiffened Elements for Local Buckling Considerations

If σ cr given by Equation 5.1 on previous page is less than F y , the yield stress, then local
buckling will occur prior to yielding of the cross section. On the other hand, if σ cr is greater than
F y , when compressive stresses reach F y , the section will yield before it locally buckles. Therefore,
by making σ cr from Equation 5.1 equal to F y and solving for b/t, one can obtain the limiting b/t
ratio that if exceeded local buckling will start prior to yielding of any fiber. This limit, denoted as
λ r , is given as:

E (5.2)
λr =C
Fy

The coefficient C in the above equation depends on the boundary condition and residual
stresses. Values of C have been established for typical column cross sections such as those shown
in Figure 5.2 earlier (wide flanges, angles, channels, tees and built-up shapes) and corresponding
values of λ r are given in Table B 4.1a of the AISC Specification (AISC, 2010a), which is given
as Table 5.1 on the following page.

5.4. b. Classification of Sections for Local Buckling

(AISC, 2010a, Page 16.1-17, Section B4)
As discussed earlier, local buckling is a phenomenon that depends on width-to-thickness ratio, b/t,
of the rectangular elements of the steel cross section as well as yield stress, F y , of steel. The
higher the b/t and F y are, the higher is possibility of local buckling to occur and the lower will be
the critical local buckling stress. Having the same b/t ratio, the steel with higher F y is more likely
to locally buckle than the steel with lower yield stress.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
5-10

Table 5.1. Values of λr

(Table B4.1a of the AISC Specification (AISC, 2010a))
CHAPTER 5-COMPRESSION MEMBERS
5-11

With regard to local buckling, and value of “C” in Equation 5.2 on Page 5-9, elements of a
cross section are divided into unstiffened or stiffened elements. As shown in Figure
5.10(a), unstiffened elements are those elements of cross section that are connected to adjacent
elements only on one side. Examples of unstiffened elements are half –flange elements of I- and
T-shapes, flanges of channels, legs of angles and stems of T-shapes, see Figure 5.10(a). On the
other hand, as shown in Figure 5.10(b), stiffened elements are those elements that are connected to
the adjacent elements on both sides. Examples of stiffened elements are webs of I-shapes and
channels and walls of HSS tubes and built-up box sections. Round pipes do not have flat
elements, and cannot be categorized as stiffened or unstiffened. Currently, the AISC Specification
(2010a) considers pipes as stiffened elements.

Considering local buckling behavior of column sections, the AISC Specification divides
elements of a cross section into two categories of nonslender and slender. If an element of cross
section, like flange or web, is nonslender, the element is not expected to develop local buckling
prior to some or all fibers reaching yield point. Elements that are slender can develop local
buckling while still elastic and before any fiber yields. In order to establish whether or not an
element is slender, the width-to-thickness ratio of the element, λ=b/t is compared to the λ r values
given in Table 5.1 (on previous page). If λ = b/t is greater than λ r , then the element is slender.
Otherwise element is nonslender.

Important compression members, such as columns in buildings, are designed without any
slender elements (i.e. the b/t of all elements of the cross section is less than or equal to λ r given in
Table 5.1). However, in some cases, where a compression member is not critical to overall
strength and stability of the structure, we can use cross sections with slender elements (with b/t
greater than λ r .) In these cases, the available strength of the compression member is established
by including the effects of local buckling. The AISC Specification (2010a) in its Section E7
provides equations and procedures to establish compressive capacity of columns with slender
cross sectional elements. In this document, we will limit our discussion to compressive members
with cross sectional elements not being slender (i.e. their b/t is less than or equal to the λ r given in
Table 5.1.)
In order to increase local buckling capacity of a compressive member, the b/t (or h/t) ratio
of elements of cross section should be reduced. Either to reduce b/t, we can increase t or decrease
b. Decreasing b means making the section multi cell, Figure 5.11(b), or adding stiffeners as shown
in Figure 5.11(c). Making multi-cell members is practical only when the cross section is very
large and the cells are still large enough to enter and do welding inside them. This is quite often
the case in bridge piers and towers. In cases, where making the section multi-cell is not practical
or economical, a solution is to add stiffeners, as shown in Figure 5.11(c). Another solution is to
use thicker plates and reduce the b/t ratio.
b b
b

t t t

(a) Box Compression Member (b) Multi-Cell Box (c) Stiffened Box

Figure 5.11. Reducing b/t by Building Multi-Cell Members or by Adding Stiffeners

Numerical Example 5.1: Establishing Local Buckling Slenderness of a

Wide Flange Column Section - Case 1 and Case 5 in Table 5.1
Check cross section shown below for local buckling and establish whether or not the column section is
slender. Use Table B4.1a of the AISC Specification (2010a).

Given:
1. Cross section of the member: W12x40
2. Material of the member: A992 Gr. 50: F y = 50 ksi

Required:
To check b/t or h/t of the elements for local buckling

SOLUTION:
1. Check flanges:
This is Case 1 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a) see Page 5-
10. In this case, we can check b/t values against λ r to find out if section is “slender” or not.
b / t = (8.01 in) / (2 x 0.515 in.) = 7.78
λ r = 0.56(E/F y ) 0.5 =13.5
Since b/t =7.78 is less than λ r =13.5, flanges are nonslender.
2. Check web:
This is Case 5 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a), see Page 5-
10. Let us find out if web is “slender” or not.
h/t w =33.6 (from Tables in the AISC Manual)
λ r =1.49(E / Fy) 0.5 = 35.9
Since h / t w = 33.6 is less than λ r = 35.9, web is nonslender.

Answer: Considering Local Buckling, the Column Section Is Nonslender

Numerical Example 5.2: Establishing Local Buckling Slenderness of a

Built-up Column Section – Cases 2 and 8 in Local Buckling Table
Consider the cross section below and investigate whether this cross section is “slender” with regard to local
buckling.

Given: y
10” 10”
1. Cross section of the member is given as shown.
2. Material of the plates: A36: (F y =36 ksi)
1”

Required: Element A
x
To check b/t or h/t of the elements for local buckling 24”
Element B 1/2”
SOLUTION:
The cross section has two elements to be checked, 1”
which are denoted as A, and B. Following are the
checks for these elements.

1. Check Element A of the Cross Section:

Element A is a flange of a built-up section which matches the definition for Case 2 of Unstiffened Elements
in Table 5.1 given earlier on Page 5-10, (which is the same as Table B4.1a of the AISC Specification,
2010a). For Case 2, the value of λ r is given as:
λ r = 0.64 (k c E/F y ) (0.5) where k c is given as k c = 4/ (h/t w ) (0.5). Notice that k c cannot be taken less than 0.35
nor greater than 0.76. Let us calculate k c:
CHAPTER 5-COMPRESSION MEMBERS
5-13

k c = 4/ (h/t w ) (0.5) = 4/ (24/0.5) (0.5) = 0.58. Since 0.58 is between 0.35 and 0.76, we will use k c equal to 0.58,
see footnote to Table 5.1 on Page 5-10.
λ r = 0.64 (k c E/F y )(0.5) = 0.64(0.58×29,000 ksi / 36 ksi)(0.5) = 10.5
Since b/t =10 is less than λ r =10.5, then Element A is nonslender.

2. Check Element B of the Cross Section:

Element B is a stiffened element that is a Case 8 in Table 5.1 of this document (Table B4.1a of the AISC
Spec.). For Case 8, the value of λ r is given as λ r =1.49(E/F y ) (0.5). Let us check to see if the element is
slender:
h/t w = 24”/0.5”= 48 , and;
λ r =1.49(E/F y )(0.5) =1.49(29,000/36)(0.5)=42.3
Since h / t w = 48 is greater than λ r = 42.3, then Element B is slender.

Conclusion: Since one element of the cross section is slender, then the cross section is slender.

Numerical Example 5.3: Establishing Local Buckling Slenderness of

Angle Column Sections – Case 3 in Local Buckling Table
Check double-angle cross section shown below for local buckling and establish whether or not the section is
slender. Use Table B4.1a of the AISC Specification.

2L-4x4x3/8
Given:
1. Cross section of the member: 2L-4x4x3/8
2. Material of the member: A36: (F y = 36 ksi.)
Required:
The element
To check b/t or h/t of the elements for local buckling
to be checked.

SOLUTION: Steel is A36

1. Check one of the legs:
This is a Case 3 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a), see Page
5-10.

b / t = 4”/ (3/8”) =10.7

λ r = 0.45(E/F y ) 0.5 = 0.45(29,000/36)0.5 = 12.7
Since b / t < λ r  Leg element is nonslender.

Since all elements of the cross section are nonslender, the 2L column section is nonslender.

Numerical Example 5.4: Establishing Local Buckling Slenderness of

Tee Sections – Cases 1 and 4 in Local Buckling Table
Check double-Tee cross section shown below for local buckling and establish whether or not the section is
slender. Use Table B4.1a of the AISC Specification.

Given:

1. Cross section of the member: 2WT9x79

2. Material of the member: A992 Gr. 50: (F y = 50 ksi.) 2WT9x79

Required:
To check b/t or h/t of the elements for local buckling

SOLUTION:
1. Check flange of Tee: Steel is A992 Gr. 50
Since WT section is produced by cutting wide flange section in half,
the local buckling check for WT is the same as for a wide flange.
Therefore, this is Case 1 of local buckling cases in Table B4.1a of the AISC Specification (2010a), see Page
5-10.
b / t = (11.3 in)/(2*1.44 in) = 3.9
λ r = 0.56(E/F y ) 0.5 = 0.56(29,000 ksi /50 ksi) 0.5 = 13.5
Since b / t < λ r  flange of the WT is nonslender.

2. Check “stem” of Tee:

This is Case 4 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a), see Page 5-
10.
b / t = 9.86”/ 0.81” =12.17
λ r =0.75(E/F y ) 0.5 = 0.75(29,000 ksi / 50 ksi) 0.5 = 18.1
Since b / t < λ r  stem of the WT is nonslender.

Since all elements of the cross section are nonslender, the section is nonslender.

Numerical Example 5.5: Establishing Local Buckling Slenderness of

Channel Sections– Cases 1 and 5 in Local Buckling Table
Check double-channel cross section shown below for local buckling and establish whether or not the section
is slender. Use Table B4.1a of the AISC Specification.
2C15x50
Given:
1. Cross section of the member: 2C15x50
2. Material of the member: A36: F y = 36 ksi.
Required:
To check b/t and h/t of the elements for local buckling

Steel is A36
SOLUTION:
Properties of a C15x50 cross section are: b f = 3.72 in., t f = 0.650 in., d = 15 in., t w = 0.716 in., T =12.125 in.
1. Check flange of the channel:
This is Case 1 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a).
b / t = b f / t f = 3.72 / 0.65 = 5.7
λ r = 0.56(E/F y ) 0.5 = 0.56(29,000/36)0.5 = 15.9
Since b / t < λ r  flange is nonslender.

2. Check web of the channel:

This is Case 5 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a).
h / t w = 12.125”/0.716”= 17.0
λ r = 1.49(E/Fy) 0.5=1.49(29000/36)0.5 = 42.2
Since h / t w < λ r  web is nonslender.

Since all elements of the cross section are nonslender, the 2-C section is nonslender.

Numerical Example 5.6: Establishing Local Buckling Slenderness of a

Tube HSS Column Section – Case 6 in Local Buckling Table
CHAPTER 5-COMPRESSION MEMBERS
5-15

Check tube cross sections shown below for local buckling and establish whether the section is compact,
non-compact, or slender. Use Table B4.1 of the AISC Specification.

Given: HSS 12 x 6 x3/8

1. Cross section of the member: HSS12x6x3/8

2. Material of the member: A500, Gr. B: F y = 46 ksi.
Required:
To check b / t of the elements for local buckling.

SOLUTION:
In this case, we only need to check b/t of the long side of the HSS, which is the vertical side, called “web.”

Check Vertical Element (web):

This is Case 6 of local buckling cases in Table B4.1a of the AISC Spec. (AISC, 2010a), see Page 5-10.
b / t = 31.4
λ r =1.40(E/F y ) 0.5=35.2
Since for vertical element (web), b / t < λ r then web is nonslender

Since all elements of cross section are nonslender, the HSS section is nonslender.

Numerical Example 5.7: Establishing Local Buckling Slenderness of a

Column Cover Plate- Case 7 in Table 5.1
Check cross section shown below for local buckling and establish whether or not the column section is
slender. Use Table B4.1a of the AISC Specification (2010a).

Given: Weld
PL 8x1/2, A572
1. Cross section of the member: W12x72+2 PL8x1/2” Gr. 50 on top & s
2. Material of the member: A992 Gr. 50: F y =50 ksi bottom flanges
3. Material of the plates: A572 Gr. 50: F y =50 ksi
Required: W12x72,
A992, Gr. 50
To check b/t and h/t of the elements for local buckling

SOLUTION:

1. Check flanges:
This is Case 1 of local buckling cases in Table B4.1a:
b / t = (12.0 in) / (2*0.670 in) = 8.99
λ r = 0.56(E/F y ) 0.5 =0.56(29,000/ 50)0.5 =13.5
Since b / t = 8.99 is less than λ r =13.5, flanges are nonslender.
2. Check web:
This is Case 5 of local buckling cases in Table B4.1a , Page 5-10:
H / t w = 22.6 (see AISC Manual Table 1-1 for W12x72)
λ r =1.49(E/Fy) 0.5 =35.9
Since h/t w = 22.6 is less than λ r = 35.9, web is nonslender.

3. Check flange cover plates:

This is Case 7 of local buckling cases in Table B4.1a:
h / t w = 8/ 0.5 = 16
λ r =1.40(E/Fy) 0.5 =1.40(29,000/ 50)0.5 = 33.7
Since h/t w = 16 is less than λ r = 33.7, flange cover plates are nonslender.

Since all elements of cross section are nonslender, the cross section is nonslender.

Numerical Example 5.8: Establishing Local Buckling Slenderness of a

Built-up Box Column Section – Case 8 in Local Buckling Table
Check the welded built-up box section shown below for local buckling and establish whether or not the
column section is slender. Use Table B4.1a of the AISC Specification.

Given:
PL-24x3/4
1. Cross section of the member: Built-up Box as shown.
2. Material of the member: A36 steel: F y = 36 ksi.
PL-20x1/2
Required:
To check b/t or h/t of the element for local buckling

SOLUTION: All Steel: A36

1. Check Horizontal Element (flange):

This is Case 8 of local buckling cases in Table B4.1a of the AISC Spec. (AISC, 2010a).
b / t = (20”-2 x 0.75”) / (0.5”) = 37
λ r =1.49(E/F y ) 0.5 =1.49(29,000 / 36)0.5 = 42
Since for horizontal element (flange), b / t = 37 < λ r =42, then flange plate is nonslender.

2. Check Vertical Elements (webs):

This is also Case 8 of local buckling cases in Table B4.1a of the AISC Spec. (AISC, 2010a).
b / t = 24” / 0.75” = 32
λ r =1.49(E/F y ) 0.5 =1.49(29,000/36)0.5 = 42
Since for vertical element (web), b / t= 32 < λ r =42, then web is nonslender.

Since all elements of cross section are nonslender, the column section is nonslender.

Numerical Example 5.9: Establishing Local Buckling of a Pipe Column

Section – Case 9 in Local Buckling Table
Check pipe cross sections shown below for local buckling and establish whether or not the section is
slender. Use Table B4.1a of the AISC Specification.
Given:
1. Cross section of the member: Pipe 18x1/2
2. Material of the member: A53, Gr. B: F y = 35 ksi.
Required:
R=18”, t=1/2”
To check b/t or h/t of element for local buckling

SOLUTION:

1. Check Pipe:
This is Case 9 of local buckling cases in Table B4.1a of the AISC Specification (AISC, 2010a).
D / t = 36 in. / 0.5 in. = 72
λ r = 0.11E / F y = 0.11x29, 000 ksi / 35 ksi = 91.1

Since D/t = 72 is less than λ r =91.1, the pipe section is nonslender.

CHAPTER 5-COMPRESSION MEMBERS
5-17

5.5. LIMIT STATE OF OVERALL BUCKLING OF

AXIALLY LOADED STEEL MEMBERS

When the strain across the section of a compression member reaches critical
level, the member has reached limit state of Overall Buckling and buckles over
a relatively long portion of its length, Figure 5.12.
Figure 5.13 shows three modes of overall buckling of a column. The
three modes are buckling about x-x axis (a flexural buckling mode), buckling
about y-y axis (also a flexural buckling mode), and buckling about longitudinal
z-z axis (torsional buckling mode). In addition to these three buckling modes, a
column can experience flexural-torsional buckling mode due to coupling of y
flexural modes and the torsional mode.
The AISC Specification (AISC, 2010a) addresses all of the above y
overall buckling modes as well as reduction of capacity due to the effects of
local buckling. In the following, we will discuss overall buckling of columns
that do not have slender cross sectional elements (i.e. b/t ratios of their cross
sectional elements are not greater than λ r given in Table B4.1a of the AISC
Specification (2010a), see Page 5-10).

Figure 5.12. Overall Buckling

about y-y axis

y
x
y
x
Torsional
Buckling Buckling
Buckling
of Entire about z-z
of Entire Column axis
Column
about y-y
about x-x
axis
axis

a. Buckling About X-axis b. Buckling About Y-axis c. Torsional Buckling About Z-axis

Figure 5.13. Three Overall Buckling Modes in Columns

5.6. COMPRESSIVE STRENGTH FOR FLEXURAL

BUCKLING OF MEMBERS WITHOUT SLENDER
ELEMENTS
5.6. a. Background on Column Buckling
Leonhard Euler (1707–1783), a Swiss mathematician, for the first time established buckling
capacity of a column under axial compression, where both ends of the column had pin supports.
The equation, now known as Euler’s Buckling Equation, was given as:

π EI
2
(5.3)
Pcr = 2
Where L
P cr = critical buckling load (i.e. compressive strength)
E = modulus of elasticity of the material (for steel, E = 29000 ksi)
I = smaller of moment of inertia of cross section with respect to its principal axes (x or y).
L = length of the column.

If in Equation 5.3, we replace P cr with F cr A, where F cr is the critical value of uniform

stress applied to the column causing it to buckle, and A is the area, then we will have:

π EI π E
2 2
Pcr (5.4)
Fcr = = 2
= 2
A AL L /( I / A)

The term I/A with r2, where r is the radius of gyration, Equation 5.4 becomes:

π E
2
(5.5)
Fcr = 2
(L / r)
In actual structures, the ends of a column, in many cases, are not pin connected, which was
the case in Euler’s ideal column. Depending on the rotational rigidity of the end supports of a
column, its compressive capacity changes. If the end supports of a column are fixed, which means
rotationally rigid, its capacity will be greater than the same column but with pin end connections.
To incorporate this effect of end conditions on the buckling capacity, a factor called effective
length factor, K, is introduced to Equation 5.5 above and the equation is rewritten as:

π E
2
(5.6)
Fcr = 2
( KL / r )

Figure 5.14 on the next page shows variation of critical stress, F cr , versus KL/r, the
effective slenderness ratio of the column. Note that the value of KL/r should be calculated for both
principle axes of the cross section of the column and the larger of the two KL/r, which is called the
governing effective slenderness ratio of the column, should be used in Equation 5.6 to establish
critical stress, F cr .
CHAPTER 5-COMPRESSION MEMBERS
5-19

π E
2
Fcr
Fcr = (Euler’s Equation)
KL 2
( )
r

Short columns
long columns

KL/r

Figure 5.14. Variation of Critical Buckling Stress, Fcr versus KL/r for Euler’s Equation

In Figure 5.14 above, as KL/r decreases, the capacity rapidly increases. In fact, for very
short columns the capacity is approaching infinity. Of course, we cannot have such a large
capacity since before reaching that capacity, member will yield in compression. Therefore, the
upper value for Equation 5.6 (Euler’s critical buckling stress) is the yield stress, F y . This is shown
in Figure 5.15, where Euler’s curve is terminated where it intersects the horizontal line drawn at
yield stress.
Figure 5.15 also shows approximate locations of data points obtained from actual column
tests. In general, test results are lower than the capacity predicted by Euler’s equation and the
yield stress cut off line. The exception is for very short columns that the critical stress is more
than the yield stress. This increase is primarily due to strain hardening in the very short columns.
The decrease in critical stress for medium and long column capacity is primarily due to presence
of residual stresses and the fact that actual columns are not mathematically straight but have some
out-of-straightness.
In developing design equations that can predict buckling critical stress with reasonable
accuracy, the test results have been used to modify the Euler’s equation. In addition, the cut off
horizontal line passing through the yield stress, is replaced with an empirical curve to fit the test
result as shown in Figure 5.16.

Fcr Fcr = Fy
π E
2

Fcr = (Euler’s
kL 2 Equation)
Fy ( )
r

Test Results
(Approximate Data)

KL/r

Figure 5.15. Variation of Critical Buckling Stress, Fcr versus KL/r for Euler’s Equation

 F
( y )

= 0.658 Fe F
π E 2 Fcr
  y
Fe =
Fcr
Fy kL 2  
Fcr = Fy ( )
r (Design

Fy Fcr = 0.877 Fe
a (Design

Inelastic Buckling Elastic buckling c

KL
KL E = 200 (Preferred KL/r
= 4.71 r
r Fy Maximum KL/r)

Figure 5.16. Comparison of Design Equations to Test Results and Euler’s Equation

The design equation represented by the curve abc in Figure 5.16 above, consists of two
curves. Curve ab, represents inelastic buckling behavior and Curve bc, which is a slightly reduced
version of Euler’s equation, represents elastic buckling region. The value of KL/r, that divides the
inelastic and elastic buckling regions, is set at 4.71 √(E/F y ). The KL/r of columns designed
according to the AISC Specification (AISC, 2010a) is preferred to be less than 200 as shown in
Figure 5.16. In the following, these design equations, which fit the test results better and shown in
Figure 5.16, are discussed.

5.6. b. Design of Columns for Limit State of Overall Flexural Buckling

(Ref.: AISC Specification, (AISC, 2010a), Section E3, Page 16.1-33))

The equation of design for checking overall buckling of a column is:

Pu ≤ φc Pn (in LRFD) and P a ≤ P n /Ω c (in ASD) (5.7)

Where
P u = required axial compression strength in LRFD, same as the factored applied compression
force.
φ c = resistance factor for overall flexural buckling in LRFD = 0.90
P a = required allowable buckling strength in ASD, same as the un-factored (service) applied
compression force.
Ω c = safety factor for overall buckling in ASD = 1.67
P n = available nominal overall buckling strength of the column given by:

P n = F cr A g (5.8)
and
A g = gross area of member
CHAPTER 5-COMPRESSION MEMBERS
5-21

F cr = flexural buckling stress, determined as follows:

KL E
a. when ≤ 4.71 then buckling is inelastic and Fcr is given by:
r Fy

 (
Fy
)
Fcr = 0.658 Fe  Fy
 
(5.9)

KL E
b. when > 4.71 then buckling is elastic and F cr is given by:
r Fy

Fcr = 0.877 Fe (5.10)

Where
F y = specified yield stress of the steel used in the column
E = modulus of elasticity of the steel = 29,000 ksi
KL/r = governing effective slenderness ratio, larger of KL/r with respect to xx and yy axes.
K = effective length factor (discussed in the following Section 5.7 of this chapter)
L = length of the compressive member
r = radius of gyration of cross section with respect to axis of buckling
F e = elastic critical buckling stress determined from Euler’s Buckling Equation below:
2
π E
Fe = 2 (5.11)
 KL 
 
 r 
 

Figure 5.17 below shows Equations 5.9 and 5.10 above that define critical buckling stress
due to overall buckling of a column.

 (
Fy
)

Fcr
Fcr = 0.658 Fe  Fy
 
Fy x y
x y
Fcr = 0.877 Fe

0.44 Fy
Inelastic
Buckling
Elastic Buckling

0
E Effective Slenderness Ratio, KL/r
4.71
Fy

Figure 5.17. Design Equations to Establish Nominal Capacity of Compressive Members

In addition to above failure mode of overall flexural buckling, there are torsional and
flexural-torsional buckling modes, which in some cases can govern. For doubly- and singly-
symmetric sections, usually flexural mode of buckling, as discussed above, governs. However, in
rare cases even for doubly symmetric sections, torsional or flexural torsional buckling mode can
govern. For members with no axis of symmetry, buckling modes involving torsional modes can
be the governing buckling mode and need to be checked. More information on torsional and
flexural torsional buckling is given in Part 2 of this Chapter starting on Page 5-37.

Numerical Example 5.10: Calculating Radius of Gyration of a Cross

Section
Calculate radius of gyration of the welded built-up cross section shown, with respect to its principal axes
xx and yy.

Given: y
Cross section of the member 10” 10”
Required:
To calculate radius of gyration with
respect to xx and yy axes. 1”

x
SOLUTION: x
1. Establish location of the centroid (C.G.): 24”
ē = (ΣA i d i )/ (ΣA i ) 1/2” ē
C.G.
ē = (20×1×25.5+24×0.5×13+12×1×0.5)/(20×1+24×0.5+12×1)
1”
ē = 15.3 in
6” 6”
2. Calculate area of the cross section, A:
A = ΣA i =20×1+24×0.5+12×1=44 in2 y

3. Calculate moments of inertia I x and I y :

I x = Σ(bh3/12+Ad2 ) i
I x =20×13 /12+20×1×10.22 +0.5×243 /12+24×0.5×2.32 +12×13/12+12×1×14.82 =5351 in4
I y = Σ(hb3/12)I =1×203 /12+24×0.53 /12+1×123 /12= 811 in4

4. Calculate radii of gyration r x and r y :

r x = (I x /A) (0.5) = (5351/44) (0.5) =11.03 in. and ry = (I y /A) (0.5) = (811/44) (0.5) = 4.29 in.

5.7. EFFECTIVE LENGTH FACTOR

5.7. a. Approximate Effective Length Factor, K
(Ref.: AISC Spec., (AISC, 2010a), Commentary to Appendix 7, Page 16.1-509))

Effective length factor K represents the effects of end conditions on column buckling. AISC
Specification (AISC, 2010a) provides values of K for columns with idealized end conditions such
as pin ends and fixed ends. These K values are used in design frequently by selecting an idealized
condition that is close to the actual support conditions for the column under consideration. Figure
5.18 shows the six idealized columns with their boundary conditions and associated K values.
CHAPTER 5-COMPRESSION MEMBERS
5-23

Figure 5.18. Values of K for Columns of Buildings with Well-defined End Conditions

Notice that in the table below, two values of K, the “Theoretical K value” and the
“Recommended design value” are given for each case of column buckling. The recommended
design value is used in design.

Numerical Example 5.11: Calculating Governing Effective

Slenderness Ratio Pin about y-y,
Calculate effective slenderness ratios (KL/r) x and Fixed about x-x
(KL/r) y for both stories of the column shown below,
then the larger of the two is the governing effective
slenderness ratio.
Given: 13’
1. Cross section and length of the column
2. support conditions
3. Locations and directions of lateral braces.
Required:
To calculate effective slenderness ratios w.r.t. 18’ W12x120
both xx and yy axis for both stories.
y x
SOLUTION:
1. Establish Radius of Gyration for xx and yy axes:
x y
r x = 5.51 in. and r y =3.13 in. from tables for cross sections.

2. Sketch the Buckled Shape of the Columns w.r.t. xx and yy Axes: Fixed base
3. Select K, the Effective Length Factor from Figure 5.18 on previous page:

In order to pick one of the six buckled shapes in Figure 5.18 of the text that fits our case, we need to draw
the buckled shape. This is done in sketches below for buckling of column with respect to xx and yy axes.
By considering the buckled shapes shown below, effective length factors are:
For first floor:
K x1 = 0.8 (for this case, the buckled shape matches Case (b) of the Table in Figure 5.18.)
K y1 =0.8 (for this case, the buckled shape matches Case (b) of the Table in Figure 5.18.)
Therefore, the KL/r values are:
(KL/r) x1 = 0.8×18×12/5.51=32, and;
(KL/r) y1 = 0.8×18×12/3.13=55.
For the 1st floor, the governing KL/r is the larger of the above two values, which is 55.

Pin about y-y,

Pin about y-y, Fixed about x-x
(Free end rotation)
Fixed about x-x (No end rotation)
x y
y x

x y x y
13’

W12x120
Buckled
Shape
18’ Buckled
Shape

y x
x y

x y
y x
Fixed base
Fixed base Fixed base

Buckling About x-x Buckling about y-y

For second floor:

K x2 = 0.8, (for this case, the buckled shape matches Case (b) of the Table in Figure 5.18.)
K y2 =1.0 (for this case, the buckled shape matches Case (d) of the Table in Figure 5.18.)
(KL/r) x2 = 0.8×13×12/5.51 = 23 and;
(KL/r) y2 = 1.0×13×12/3.13 = 50
For the 2nd floor, the governing KL/r is the larger of the above two values, which is 50.
Governing KL/r values:
For 1st Floor, governing KL/r = 55 (governed by buckling about the yy axis)
For 2nd Floor, governing KL/r = 50 (governed by buckling about yy axis)
CHAPTER 5-COMPRESSION MEMBERS
5-25

5.7.b. More Accurate Values of Effective Length Factor, K

(Ref.: AISC Spec., (AISC, 2010a), Commentary to Appendix 7, Page 16.1-513))
To obtain a more accurate value of K, the effective length factor, we can use a method based on
the rotational stiffness of the end supports of the column. AISC Specification (AISC, 2010a),
which categorizes frames into two categories of frames where sidesway is inhibited and frames
where sidesway is uninhibited. Simply put, in frames with sidesway inhibited, lateral movement of
floor beams is prevented. This is more or less the case in braced frames or frames with shear walls
where due to relatively large stiffness of the bracings or the shear walls, the floors do not move
laterally with respect to each other significantly. Even though in these frames some lateral
movement of floors takes place under lateral wind and earthquake forces, the movement is small
enough to be ignored for establishing K.
In frames where sidesway is uninhibited, the floor beams can move laterally much easier
than the braced frames or frames with shear walls. Moment frames are good examples of frames
where sidesway is uninhibited.
For braced frames, where side sway movement of the structure is relatively small, which
AISC calls a case of sidesway inhibited, the effective length factor, K, for columns varies from 0.5
to 1.0 and can be conservatively taken as equal to 1.0 in design. For frames with sidesway
inhibited (prevented), a more accurate value of effective length factor K, can be obtained from the
following equation:

GAGB (π/K )  GA + GB 
2
π/K  2 tan(π / 2K )
+ 1 − + − 1 = 0.0 (5.12)
4  2  tan(π/K )  π/K

For moment frames, where due to flexibility of the frame, side sway movement of the
frame can take place, K varies from 1.0 to almost infinity. With such large range of variation of K,
for moment frames, it might be necessary to use more accurate methods to establish K. One such
method is to use the following equation:

G AGB (π / K ) 2 − 36 (π / K ) (5.13)
− = 0.0
6(G A + GB ) tan(π / K )

The above two equations are derived assuming that all columns of the frame will buckle
elastically. For more information on derivation of equations, the reader is referred to advanced
mechanics textbooks.
In the above equations, K is not given in a closed-form solution. In order to find it, we
need to solve the equation numerically or graphically. The Commentary section of the AISC
Specification (AISC, 2010a, Page 16.1-513) provides graphical solution to above equations called
alignment charts. The AISC alignment charts, reproduced in Figure 5.19 on the following page,
can be used to find values of the effective length factor, K, graphically for a column in a braced or
a moment frame.
In the above equations, as well as in the alignment chart shown in Figure 5.19, G A, and GB
represent the rotational stiffness of the top and bottom ends of the column. G A and G B are given
by:

G=
∑ ( EI / L)columns (5.14)
∑ ( EI / L) girders
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
5-26

(a) (b)
Figure 5.19. Alignment Charts for Effective Length Factor, K, for Columns in:
(a) Braced Frames and; (b) Unbraced Frames

Numerical Example 5.12: Calculating Effective Length Factor, K,

Using G A and G B
Calculate KL/r for the x-x and y-y direction of the column AB in the third floor of the structure shown
below. All beam-to-column connections as well as the column bases are rigid connections. Use the AISC
Alignment charts to establish K-factors.
W12x96 columns

12' W24x55 A W24x55

14’ W24x76 W27x102
B A
W24x76 W27x102
15'
18'

W12x120 columns ELEVATION B

40' 35'

W30x90
in All Floors
50'

PLAN

Given:
CHAPTER 5-COMPRESSION MEMBERS
5-27

1. cross section of the columns and beams connected to column under consideration.
2. length of the columns and beams connected to column under consideration.
3. support conditions

Required:
To calculate effective length factor, K, w.r.t. both xx and yy axis for Column AB.

SOLUTION:
1. Establish properties of cross sections:
Columns:
W12x96: I x = 833 in4 and I y = 270 in4, W12x120: I x = 1070 in4 and I y = 345 in4.
Girders:
W24x55: I x = 1350 in4, W24x76: I x = 2100 in4, W27x102: I x = 3620 in4, W30x90: I x = 3610 in4,

2. Buckling about Strong Axis (x-x):

Calculate G A and G B for the top and bottom of the column:
G A = Σ(EI/L) col. / Σ(EI/L) gir
G A = (833 / 12’+833 / 14’) / (1350 / 40’ + 1350 / 35’) = 1.78
G B = Σ(EI/L) col. / Σ(EI/L) gir
G B = (833 /14’ + 1070 / 15’) / (2100 / 40’ + 3620 / 35’) = 0.84
Using alignment charts for moment frames: K x = 1.4 (See Alignment Chart on the next page)
3. Buckling about Weak Axis (y-y):
Calculate G A and G B for top and bottom of the column:
G A = Σ(EI/L) col. / Σ(EI/L) girder
G A = (270/12’+ 270/14’) / (3610/50’) = 0.58
G B = Σ(EI/L) col. / Σ(EI/L) girder
G B = (270 / 14’+ 345 / 15) / (3610 / 50’) = 0.59
Using alignment charts for moment frames: K y = 1.17 (See Alignment Chart below)

Step 1: Enter GA =1.78

Step 3: Read Kx =1.4

Step 2: Enter GB=0.84

Step 4: Enter GA=0.58 Step 5: Enter GB=0.59

Step________________________________________________________________________
6: Read Ky =1.17
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
5-28

5.8. BUILT-UP COMPRESSION MEMBERS WITH

SEPARATE SEGMENTS
(Ref.: AISC Spec., (AISC, 2010a), Section E6, Page 16.1-37))

In some built-up compression members made of two or more shapes, the shapes, called segments,
are placed apart from each other, and are connected to each other only at intermediate points by
plates as shown in Figure 5.20.

Stitch Stitch a
a a
Batten
Plates

y y y

y y
y

Figure 5.20. Built-up Columns with Separate Segments

If a built-up member with separate segments buckles about its y-y axis, which is the axis
of separation (see Figure 5.20 above), shear forces will be generated in the connectors (i.e. stitches
and batten plates) as shown in Figure 5.21 on the next page. In this case, shear strength as well as
shear stiffness of the connectors will affect the overall buckling behavior of the column and will
dictate how two separate segments work together to create a single cross section with one neutral
axis. In addition, since the two segments of the cross section are separated, there is an additional
buckling mode in these built-up columns and that is the buckling of the individual segments
between the connectors as shown on the right side of Figure 5.22 on the next page.
In order to prevent segment buckling prior to overall buckling of built-up compression
members, the AISC Specification (AISC, 2010a) , Section E6.2 requires that the slenderness ratio,
K i a/r y , of each segment between the connectors be less than ¾ of the governing effective
slenderness ratio, KL/r , of the member:

(K i a/r i ) segment ≤ (3/4) (KL/r) member (5.15)

The radius of gyration r i in Kia/r i is the least radius of gyration of individual segment with
respect to its own principal axes. The length of the segment, a, is the distance between the
connectors, see Figure 5.20. Even though by following this requirement, the segment buckling is
not expected to be the governing buckling mode, segment buckling can be coupled with overall
buckling and cause overall buckling to occur under smaller compressive loads than if the segments
CHAPTER 5-COMPRESSION MEMBERS
5-29

of the cross section were not separated from each other. To account for the shear strength and
stiffness of connectors as well as coupling of segment buckling with overall buckling in built-up
columns with separated segments, the AISC Specification (AISC, 2010a) requires that the KL/r of
the column be increased to a modified level of (KL/r) m . Then, the modified effective slenderness
ratio is used in Equations 5.9 and 5.10, given earlier on Page 5-21, to calculate critical
compressive stress.

L a
Shear

Figure 5.21. Examples of Built-up Compression Cross Sections Buckling about y-y Axis

Overall
Buckling Segment
Buckling

________________________________________________________________________
Figure 5.22. Overall Buckling and Segment Buckling of Built-up Members
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
5-30

y
y

y
x x x x
x x

y y
y

(a) (b) (c)

y
y

x x x
x

y
y
(d) (e)

Figure 5.23 Examples of Built-up Compression Members with separated Segments

Modification of KL/r should only be done if the axis that separates the segments is the
governing axis of buckling. Figure 5.23 above shows examples of built-up columns where
segments of the cross sections are separated from each other. For cases (a) through (d) the
modification of KL/r should only be done if the KL/r of the y-y axis, which is the axis separating
segments is also the governing axis of buckling. For Case (e) in Figure 5.23, the modification of
KL/r should be done for which ever axis , x-x, or y-y that governs since both axes are separating
the segments.
The equations given by the AISC Specification (AISC, 2010a) to modify KL/r in built-up
members with separate segment are:
(a) for intermediate connections, i.e. stitches or batten plates that are bolted snug tight:

2
 KL   a 
2
 KL  (5.16)
  =   +
 r m  r  o  ri 

(b) for intermediate connections, i.e. stitches or batten plates, those are welded or fully
tensioned bolts:
a
(i) when ≤ 40
ri
 KL   KL 
  =  
 r  m  r o (5.17a)

(ii) when a
> 40
ri
CHAPTER 5-COMPRESSION MEMBERS
5-31
2 2
 KL   KL   K i a  (5.17b)
  =   +  
Where  r m  r o  ri 

 KL  = effective slenderness of the built-up member with respect to the axis

 
 r o

 KL  = modified ( i.e. increased) effective slenderness of the member

 
 r m
K i = 0.50 for angles back-to-back
= 0.75 for channels back-to-back
= 0.86 for all others
a = center-to-center distance of connectors (stitches or batten plates) as shown in Figure 5.20 on
Page 5-28.
r i = minimum radius of gyration of individual segment shown in Figure 5.24 below.
As mentioned earlier, the modification of KL/r is done only on KL/r with respect to the
axis that separates the segments. So in Figure 5.24, for cases (a) through (d), KL/r of yy axis will
be modified but not the xx-axis. Only for case (e), KL/r with respect to both xx and yy axes should
be modified.
It should be mentioned that, with the above modification of KL/r, the requirement of the
AISC Specification that the K i a/r i of each segment between the connectors be less than ¾ of the
governing effective slenderness of the member takes the following form with above (KL/r) m used
as the member effective slenderness ratio:

 Ki a  (5.18)
  ≤ (3 / 4) KL y
 r   r m
 i 

y Axis for ri Axis for ri y

Axis for
ri
x x
x x x

y y y
2-L 2-C
(a) (b) (c)
y y
Axis for ri Axis for ri

x x x x

y
y
2-WF 4-L

(d) (e)

Figure 5.24. Axis for Minimum Radius of Gyration, ri of a Segment

Numerical Example 5.13

Establish Modified Effective Slenderness Ratio, (KL/r) m , for a
Built-up Column with Separated Segments
Calculate the governing effective slenderness ratio, KL/r, for the compressive member shown below for
two cases:
Case A: The two channels are connected to each other using bolted snug-tight stitches.
Case B: The two channels are connected to each other using fillet welds.

¾” thick gusset plates Snug-tight Bolts for Case A , and

at both ends Welds for Case B
Case A: Snug-tight
2C15x50, A36 ¾” Thick, Stitch Plates Spacing=3/4” Bolts

6'-6" 6'-6" 6'-6"

Case B: Welds

Given:
1. cross section of the member is given as 2C15x50.
2. length of member, L, is given as 3x 6.5 ft x 12 in/ft = 234 in.
3. the material is given as A36 (F y = 36 ksi. F u = 58 ksi.)
4. the 2C sections are connected to each other by snug-tight or welded stitches.

Required:
To establish governing effective slenderness ratio for both cases of snug-tight and welded stitches.

SOLUTION
1. Establish properties of the 2C- section: y
Axis
Properties of one C15x50 cross section are: for ri
b f = 3.72 in., t f = 0.650 in., d = 15 in., t w = 0.716 in.,
r x = 5.24 in., r y = 0.865 in., Area = 14.7 in2.
For 2 channels:
x x
A = Area of two channels= 2×14.7= 29.4 in2
I x = 2×404= 808 in4, I y = 2[11+14.7(0.75/2+0.799)2] = 62.5 in4
r x = 5.24 in., (the same as for one channel since xx axis is still y
principle axis for 2C section)
r y = (I y /A) (0.5) = (62.5/29.4) (0.5) =1.46 in
r i = radius of gyration of one channel w.r.t. axis
of buckling (yy axis) = 0.865 in.

2. Establish effective length factors, K x and K y :

K x = 0.65
Since gusset plates have relatively large in-plane stiffness and act as fixed connections for buckling about
x-x axis, a value of K x equal to 0.65 was selected.
K y =1.0
Since gusset plates have very small out-of-plane bending stiffness and act as pin connections for
buckling about y-y axis, a value of K y equal to 1.0 was selected.
CHAPTER 5-COMPRESSION MEMBERS
5-33

3. Establish effective slenderness ratios, (KL/r) x and (KL/r) y :

L= 234 in.
(KL/r) x = 0.65×(234 in)/ (5.24) = 29
(KL/r) y = 1.0× (234 in)/ (1.46) =161  (governs)
Since member is buckling about y-y axis, which is the separating axis for the built-up column, we need to
calculate (KL/r) m for y-y axis. See Page 5-31 for more info on this item. This is done in the following.

4. Calculate modified effective slenderness ratio for Case A, “snug-tight bolt” stitches:
(KL/r) m =[ (KL/r) o 2 +(a/r i )2 ]0.5 ( This is the Equation 5.16 on Page 5-30)
(KL/r) o =161 (see part 3 above)
a = 6’-6”= 78 in.
r i = minimum radius of gyration of one channel with respect to its yy axis= 0.865 in. (from the AISC
Manual Cross Section Tables)
(KL/r) m =[ (KL/r) o 2 +(a/r i )2 ](0.5) =[ (161)2 +(78 in / 0.865 in )2 ](0.5)
(KL/r) m = 185 for Case A with snug tight stitches.

5. Calculate modified effective slenderness ratio for Case B, “welded” stitches:

Calculate a/r i :
a/r i = 78”/0.865” = 90
Since (a/r i ) = 90 is greater than 40, we need to modify (KL/r) o : y
2
(KL/r) m =[ (KL/r) o +(K i a/r i )2 ]0.5 (see equation 5.17b on Page 5-30)
(KL/r) o =161 (see part 3 above) Axis for ri

a = 6’-6” = 78 in. x
x
r i = minimum radius of gyration of one channel
with respect to its yy axis= 0.865 in.
K i = 0.75 for double channels back-to-back h
(see Page 5-30)
(KL/r) m = [ (KL/r ) o 2 +( K i a/r i )2 ](0.5) y
= [ (161)2 +(0.75x78 in / 0.865 in )2 ](0.5)
(KL/r) m = 174 for Case B with welded stitches.

Numerical Example 5.14:

Evaluate Capacity of a Built up Column with Separated
Segments
Calculate capacity (φ c P n ) of the built-up compression member shown. Steel is A36. Stitches are welded.
Consider only local buckling and flexural buckling modes.

Welded Stitch

Spacing=3/4”
2C15x50

10'-6" 10'-6"

Given:
1. cross section of the member is given as 2C15x50.
2. the material is given as A36 (F y = 36 ksi. F u = 58 ksi.)
3. the 2C sections are connected to each other by welded stitches.
Required:
To establish compressive capacity of the member by considering all failure modes which are “local
buckling” and overall “flexural buckling.”

SOLUTION
1. Establish properties of the 2C-15x50 section:
Properties of one channel are:
b f = 3.72 in., t f = 0.650 in., d = 15 in., t w = 0.716 in, r y = 0.865 in., I y = 11 in4 , I x = 404 in4,
For two channels, the properties are:
A g = Area of two channels = 2×14.7 in2 = 29.4 in2
I x = 2×404=808 in4, r x = 5.24 in.,
I y = 2[11+14.7(0.75/2+0.799)2] = 62.5 in4 , r y = (I y /A) (0.5) = (62.5/29.4) (0.5) =1.46 in

2. Check Limit State of Flexural Buckling:

K x = 0.65 (since gusset plates have very large in-plane stiffness and act as fixed end connection when
member buckles about x-x axis)
K y =1.0 (since gusset plates have very small out-of-plane bending stiffness and act as pin connections
when member buckles about y-y axis.)
The equation of design for limit state of flexural buckling is P u ≤ φc Pn where Pn =Fcr Ag.
Let us find Fcr:
(KL/r) x = 0.65×21’×(12 in/ft )/ 5.24” =31
(KL/r) y = 1.0×21’×(12 in/ft)/1.4”6 =173  (governs)
(KL/r) o = governing KL/r = 173
Since this is a built-up column with separated segments and buckling is about the axis that separates the
segments, we need to calculate (KL/r) m for y-y axis.
Since (KL/r) o = 173 is greater than 40, we need to modify (KL/r) o using following equation:
(KL/r) m = [ (KL/r)o2 +(Ki a/ri )2 ]0.5 (see Page 5-30):
(KL/r) o =173 (see above)
a = 10’-6’ = 126 in.
ri = minimum radius of gyration of one channel with respect to its yy axis= 0.865 in.
Ki = 0.75 for double channels back-to-back (see Page 5-30)
(KL/r) m = [ (KL/r)o2 +( Ki a/ri )2 ](0.5) = [ (173)2 +(0.75x126 in / 0.865 in )2 ](0.5) = 204
Check whether Ka/ri is less than or equal to (3/4) (KL/r) m.
KIa/ri = 0.75×126/0.865= 109
Since Ka/ri =109 < (3/4) (KL/r) m = 3/4(204) = 156, then segment buckling is not the governing failure
mode.
Let us consider overall buckling with effective slenderness ratio being (KL/r) m =204:
Since (KL/r) m= 204 > 4.71(E/F y) (0.5) =133, then buckling is elastic and we will use Fcr = 0.877Fe. (see
Page 5-21.)

Fe =π2 E/(KL/r)2 =(3.1415) 2(29000)/(204)2 = 6.87 ksi

Fcr = 0.877×6.87 ksi = 6.1 ksi

Pu = φc Fcr Ag.= 0.90×6.1 ksi × 29.4 in2 =161.4kips.

Capacity Pu =161.4 kips due to flexural buckling about y-y axis

CHAPTER 5-COMPRESSION MEMBERS
5-35

5.9. LIMIT STATES OF CONNECTION FAILURE

Connections of compression members in trusses, braced frames, moment frames and other
structures need to be designed to carry their loads. Failure modes of connections such as gusset
plates and base plates, often used in axially loaded members, are discussed in later chapters.

5.10. OTHER CONSIDERATIONS IN DESIGN OF

AXIALLY LOADED COMPRESSION MEMBERS
In addition to above failure modes that need to be considered in design, it is preferred that the KL/r
of a compression member is less than 200. Previous editions of the AISC Specification had a
requirement that the KL/r of columns should not exceed 200. However, this is no longer the case
and the current AISC Specification does not have such a requirement and only states that: “For
members designed on the basis of compression, the effective slenderness ratio KL/r preferably
should not exceed 200.” (AISC, 2010a, P. 16.1-33). In addition, in built-up members, the AISC
Specification (2010a) in Section E6.2 provides information on “Dimensional Requirements” for
built-up columns. These requirements are for dimensions of stitches, batten plates and other
elements that connect the segments to each other and to end supports of the member.

5.11. NUMERICAL EXAMPLE FOR DESIGN OF A COLUMN

The numerical example given in this section is on design of a column, where the applied loads are
given and the problem statement asks for selection of the material and cross section that can resist
the applied loads. In this design problem, the LRFD methods, given in the AISC Specification
(AISC, 2010a), which were discussed in this chapter, are used.

Numerical Example 5.15:

Design of a Column with Wide Flange (WF) Cross Section
Design the wide flange column shown in the figure. Loads are nominal service loads. Use LRFD methods
and consider only flexural buckling modes. Assume effective length factors, Kx and Ky to be 1.2 and 0.8
respectively.

End Plate y

Welded
x x

DL= 300 k DL= 300 k y

LL= 250 k 18’-6” LL= 250 k

Given:
1. Applied service (unfactored) loads are given as: DL= 300 kips and LL= 250 kips, both compressive forces.
2. The length of member is given as 18’-6”.

3. The effective length factors Kx and Ky are given as 1.2 and 0.8 respectively.
Required:
1. To design the wide flange column using the LRFD methods.
2. To check the AISC requirement for the KL/r to be preferably less than or equal to 200 for columns.

SOLUTION

1. Establish the load:

The load combinations involving DL and LL are 1.2D+1.6L and 1.4D. Therefore, Pu, the factored
compression force acting on the member, also called required strength, will be the larger of these two
combinations:
1.2×300k + 1.6×250k = 760 k and 1.4×300 k = 420 ⇒ Use Pu = 760 k

2. Select Material of the Member:

With reference to the AISC Manual Table for Material (given in Chapter 2), the “preferred” material for wide
flanges is ASTM-A992. We will use A992 (with Fy = 50 ksi and Fu = 65 ksi). In our design we will ensure that
the selected cross section is nonslender from local buckling point of view since columns with slender cross
sectional elements are not as economical as compact or non-compact sections.

3. Design for Limit State of Flexural Buckling:

The equation of design for limit state of flexural buckling is Pu ≤ φc Pn where Pn = Fcr Ag. Therefore, for
design:
Pu ≤ φc Fcr Ag
760 ≤ 0.90 Fcr Ag ⇒ Fcr Ag > 844.4 kips.
Now, we can select a wide flange and then by knowing the cross section we can establish Fcr and check to
see if the above inequality is satisfied in an economical way (i.e. the two sides are relatively close).
Alternatively, we can select an economical value for Fcr and select a cross section that its area will satisfy
above inequality and its rx and ry , used to find governing KL/r, would result in Fcr value only slightly larger
than the assumed Fcr.
In order to end up with an economical section, let us assume Fcr will be about 80% of Fy. This is an arbitrary
choice of Fcr to be able to start with a reasonably economical section. Using this Fcr we have:
Ag > 844.4 kips/ (0.80×50ksi) ⇒ Ag > 21.11 in2. W12 and W14 wide flange sections can be more
economical for axially loaded columns than deeper wide flanges.

Let us Try W12×72, A992, Gr. 50 Steel,

(Ag = 21.1 in2, Fy = 50 ksi, rx = 5.31 in., ry = 3.04 inches and we had other properties as: L = 222 in., Kx =1.2
and Ky = 0.8)
(Kx L/rx) =1.2×222/5.31= 50.1 and (Ky L/ry) = 0.8×222/3.04=58.4 ⇒ Use Ky L/ry =58.4 as governing KL/r.
Since KL/r =58.4 < 4.71(E/F y) (0.5) = 113, then buckling is inelastic and we will use:
Fcr = [0.658(Fy/ Fe)] Fy.

Fe = π2 EI(KL/r) 2 = (3.1415) 2(29000)/ (58.4)2 = 83.9 ksi

Fcr = [0.658(Fy/ Fe)] Fy = [0.658(50/83.9)] 50 = 39 ksi

Since Pu =760 kips > φc Fcr Ag =0.9(39) (21.1) = 740 kips, this design is not acceptable (it missed the
acceptable capacity by less than 2 %!). However, now we can try a slightly larger section, and this second
trial most likely will be the final design.

Let us Try W12×79, A992, Gr. 50 Steel,

(Ag = 23.2 in2, Fy = 50 ksi, rx = 5.34 in., ry = 3.05 inches and we had other properties as: L = 222 in., Kx =1.2
and Ky =0.8)
(Kx L/rx) =1.2×222/5.34=49.9 and (KY L/ry) =0.8×222/3.05=58.2 ⇒ Use Ky L/ry = 58.2 as governing KL/r.
Since KL/r = 58.2 < 4.71(E/F y) (0.5) = 113, then buckling is inelastic and we will use:
Fcr = [0.658(Fy/ Fe) ]Fy .
CHAPTER 5-COMPRESSION MEMBERS
5-37

Fe = π2 E/(KL/r)2 = (3.1415) 2(29000)/(58.2)2 = 84.5 ksi

Fcr = [0.658(Fy/ Fe)] Fy = [0.658(50/84.5)] 50 = 39 ksi
Since Pu = 760 kips < φc Fcr Ag = 0.9(39) (23.2) =814 kips, this design might be acceptable if the section is
nonslender. Let us check b/t and h/tw for local buckling:
b/t = 8.22 (from AISC tables for W12x79. We could also directly calculate this value).)
Since b/t =8.22 < 0.56(E/Fy) (0.5) = 13.5, then flange is nonslender. Notice that this is Case 1 of Table 5.1
above (same as Table 4.1a of the AISC Specification, 2010a).
h/tw = 20.7 (from AISC tables for W12x79. We could also directly calculate this value)
Since h/tw = 20.7 < 1.49(E/Fy) (0.5) = 35.9, then web is a non-slender element. Notice that this is Case 5 of
Table 5.1 on page 5-10 (same as Table 4.1a of the AISC Specification, 2010a).
Since both flanges as well as web are nonslender, then there is no need to consider local buckling. So, for
this problem, and considering only flexural buckling modes:
USE W12×79, A992, Gr. 50 Steel

References for Chapter 5

PROBLEMS ON COMPRESSION MEMBERS

Problem 5.1. Check cross sections shown below for local buckling and establish whether they are
nonslender or slender. Use Table B4.1a of the AISC Specification (2010a). Steel for each cross section is
given under the sketch of each section.

W12x120 2C12x30
L6x6x5/8 WT6x60

(D)
(A) (C)
(B) A992, Gr. 50 Steel
A992, Gr. 50 Steel A36 Steel
A36 Steel

2CMC18x42.7
Plate 8"x3/4"

2L7x4x3/8 W24x146
Plate 8"x 3/8"
Plate 8"x3/4"

(E) (F) (G)

All Plates: A36 Steel A572, Gr. 50 Steel A572, Gr. 50 Steel

10” Plate 20"x1" Plate 10"x1"

Plate 6"x3/4"
1"
2” Plate 4"x1/2"
12” 8"
Plate 5"x1/2"
Plate 8"x 1/2 4"
Plate 28"x1/2" 1"
7/8"
Plate 10"x2" 6"
2" 2"
5” (I)
All Plates: A36 Steel
(H)
All Plates: A572, Gr. 50 Steel 6" 6" 6"
6"

Plate 16"x3/4"

3/4" Plate 24"x3/4" Top &

Plate 4"x1/2" Bott.

3 Plates 32"x 5/8, Plate

Plate 30"x 1/2 A36

Plate 24"x 5/8”

(J)
All Plates: A36 Steel
(K)
CHAPTER 5-COMPRESSION MEMBERS
5-39

Problem 5.2 Using LRFD, check all segments of the column cross section below for possibility of local
buckling and indicate whether the cross section is nonslender or slender. If it is slender, indicate a solution
to make it nonslender. Plates are A36, HSS is A500 Gr. B, and wide flange is A992 Gr. 50 steel.

W14x90
PL36x3/8

HSS8x8x3/16

Problem 5.3. Check cross sections shown below for local buckling and establish whether they are
nonslender or slender. Use Table B4.1 of the AISC Specification (2010a). Steel for each cross section is
given under the sketch of each section.
R=18”, t=1/2” t=1 inch for all
HSS12x6x3/16 plates

4@4 ft
(B)
(A) A53, Gr.B Steel
A500, Gr. B Steel

4@4 ft
(C)
All Plates: A36 Steel
Problem 5.4 For cross sections given in Problem 5.1 calculate Ix, Iy, rx, ry...

Problem 5.5. Calculate radius of gyration of the cross sections shown below with respect to x-x and y-y
axes. y
For all
All Angles: y y plates, t=7/8”
C15x50
L6x6x1/2”

W12x120

x x
1’-8” x
x
Weld all 4@3 ft x
four corners

C15x50
1’-8”
y
y 4@3 ft

(A) (B) (C)

Problem 5.7. Calculate axial load carrying capacity (φcPn) of the columns shown. The material of each
column is indicated below each cross section. For both columns, assume a length of 20 feet and Kx=1.5 and
KY=1.8.
Weld all y Plate 10"x1"
Weld all four four
corners y corners

x x Plate 14"x
x x C15x50
1/2"
Plate 10"x1"
W30x148 y C15x50
W10x112
2-WT15x74 y
All Steel is A572 Gr. 50 Steel (C)
A992 Gr. 50 Steel
(B) All Steel is A572 Gr. 50 Steel

(A) Plate 14"x1"

Plate 14"x 1/2"

(D)
Plate 14"x1"

All Steel is A572 Gr. 50 Steel

Problem 5.9. Establish effective length factor for the segments of the columns shown below.

Top: Pin connection in both

directions Top: Fixed in both Top: Fixed in x-x direction pinned in
directions y-y direction

12’ 14’
18’
12’
10’
12’

y 12’
12’ x 12’ y x
y x

x y x y
x y
Fixed base
(in both directions) Fixed base
(in both directions) Pin Base
(in both directions)

(A) (B) (C)

CHAPTER 5-COMPRESSION MEMBERS
5-41

Top: Pin connection in

Top: Fixed in both Top: Pin connection in both
both directions
directions directions

14’
12’ 14’

12’
14’ 14’

18’ 14’
10’

y 12’ 16’ 14’

x y y x
x

x y x y
x Fixed base y
Fixed base Pin Base
(in both directions) (in both directions)
(in both directions)

(D) (E)
(F)

Problem 5.10. Calculate factored load Pu that can be applied to the columns shown.
12’

Top is pin in both

Pu directions

Pu Pu
`
Top is free
Pu

W12x120 12'-0" 2Pu

12'-6" 2Pu 10'-0"
2Pu
Braces
14'-6"

W14x159 20'-6"
Pu W14x68
18'-0"
17'-6" y x y
x
Steel is x x
A992, y x
Gr.50 y
12'-0" Base is fixed y Base is fixed
y

Base is fixed x Steel is A992, Gr.50 Steel is A572, Gr. 50

(A) (B) (C)

Problem 5.11. Calculate the axial load capacity of the column shown in the figure. Assume Fy= 50 ksi
and Fu= 65 ksi.
Pu Weld all four
Pin Connection y
corners

15’ x x

10’
W30x148
10’ y
2-WT15x74
y 15’ A992 Gr. 50 Steel
x
Fixed base
x y
(A)

Fixed in both Pu
Weld all four corners
directions
y

15’

15’ x x

15’
W33x318
y 15’
x y
2-WT16.5x159

x y A992 Gr. 50 Steel

Fixed in both (B)
directions

Problem 5.12. Calculate capacity (φcPn) of the compression members shown below. Steel for all sections
is A36. Top and bottom supports can be considered fixed. Both columns have the same length of 5’-0”.

S5x10 C4x5.4

5’-0” 5’-0”

(A) (B)
CHAPTER 5-COMPRESSION MEMBERS
5-43

Problem 5.13. Using a wide flange section, design the column shown in Problem 5.9(A) for a factored
applied load of 200 kips. Use A992 Grade 50 steel. The x-x and y-y axes shown are strong and weak axes of
the cross section respectively.
Problem 5.14. Using a double channel section with ¾ spacing between the channels, design the column
shown in Problem 5.9(B) for a factored applied load of 355 kips. Use steel that is preferred by AISC for
channels. The x-x and y-y axes shown are strong and weak axes of the cross section respectively.
Problem 5.15. Using a wide flange section, design the column shown in Problem 5.9(C) for a factored
applied load of 500 kips. Use steel that is preferred by AISC for W sections. The x-x and y-y axes shown are
strong and weak axes of the cross section respectively.

Problem 5.16. Calculate factored load Pu that can be applied to the structures shown. Use alignment charts
to establish K factors in plane of the frame and for out of plane buckling use K=1.0. Notice the column
orientations shown under each column. All beams are W21x57 and all columns are W10x45. Steel is A992
Gr. 50.

P1 2P1 P2 3P2 3P3 P3 2P3

2@ 10 ft

Pin Fix

15 ft 15 ft 2 @15 ft

(A) (B) (C)

Problem 5.17- Check all applicable failure modes of Column AB in the frame and establish whether the
column is adequate to carry its load. The column is axially loaded and has Fy = 65ksi. All loads are factored
loads.

Ky is equal to 1.2. To establish Kx use AISC alignment charts.

Notice the orientation of the column shown in the figure (below column AB)

400 kips

W12x120
15 ft 400 kips

W24x55 W24x55
A

18 ft W12x120
Fixed
B Connection

y y
20 ft 30 ft

Problem 5.18. Design column AB, using one cross section for both first and second floors. All beam-to-
column connections as well as column bases are rigid connections. The loads shown are factored loads. Use
A992, Gr.50 steel. Use AISC Alignment charts to establish K-factors. 100 kips

20' 20' 10' 100 kips

15' 14' 100 kips

W24x55 W24x76
A
In all floors 14'
30’ 100 kips
W30x90
A
18'
B

PLAN B
ELEVATION

Problem 5.19. Consider the corner column of the structure shown below and calculate maximum value of
factored load Pu. All columns are W12x120, all beams are W21x93 and all steel is A992, Gr. 50. Use GA
and GB alignment charts to calculate K values. The base of column is fixed in both directions. All beam-to-
column connections in both directions are rigid (moment) connections.

20'
Pu

40'

12'

12' Column to be
considered.
x y
ELEVATION
Base fixed y
PLAN in both x
directions

Problem 5.20- Design column AB shown in the figure below. Use A992, Grade 50 steel. Use AISC
Alignment charts to establish KL/r for the x-x and y-y direction of the column AB. All beam-to-column
connections as well as column bases are rigid connections. The total axial load in the columns in each floor
is indicated in the 3-D sketch on the right. The loads are factored loads.

30' 30' 20' W27x84 W27x84 A W24x55

14'
W27x84 W27x84 W24x55
W24x55 300k
40' In all floors 10' W30x90 W30x90 W24x76
600k A
14'
W30x90 W30x90 W24x76 900k
18' ELEVATION 1200k
PLAN
B B
CHAPTER 5-COMPRESSION MEMBERS
5-45

Problem 5.21. Consider the corner column of the structure shown below and calculate maximum value of
factored load Pu. All columns are W12x120, all beams are W24x55 and all steel is A992, Gr. 50. Use GA
and GB alignment charts to calculate K values. The base of column is fixed in both directions. All beam-to-
column connections are rigid connections.

40' 20'

40'
Pu

2Pu

PLAN 3Pu

Base fixed in
10' These are total axial both directions
load in the column.

10'

ELEVATION

Problem 5.22. Using LRFD, check the axially loaded corner column in the figure below and find its
capacity Pu. All wide flange sections are A992 Gr. 50 steel. Base connections of all columns are fixed
support. All beam-to-column connections are moment connections.

30' 30'

W24x76
20’

PLAN
Pu kips

W30x90
14'
W12x120

18'
(A)

ELEVATION

30' 30'
W24x76

24’
Pu

PLAN
Pu kips

W30x90
14'
W12x120

18'
(B)

ELEVATION

Problem 5.23. Calculate axial load capacity of the column shown in the figure. Steel is A36.

P
Pin support
In both y 2C15x50,
y directions A36 Steel
x

x y
2C15x50, A36 steel x x
Welded Stitch
At 3 feet spacing
15’-0”
1 inch
Stitches are @ 3 feet c/c
y

15’-0” Fixed base Cross Section

in both directions

Problem 5.24. Calculate capacity (φcPn) of the built-up compression member shown. Steel is A36.
Stitches are welded.

Stitch

2L 8x8x1 Spacing=1"

4'-6" 4'-6" 4'-6"

CHAPTER 5-COMPRESSION MEMBERS
5-47

Problem 5.25. Calculate maximum factored load Pu that can be applied to the truss shown. All connections
are welded and failure modes of connection area are not part of this assignment. All members are made of
two angles diagonally positioned as shown. Each member has one welded stitch at mid-span. All steel is
A36. The truss has lateral braces at location of all joints.

10'-0" 2L-4x4x1/2
10'-0"

P
5'-0" 1”
10'-0"

Stitch

1”

Cross Section of the Truss

Members

Problem 5.26 Consider the compression member below and find maximum factored load capacity Pu.
t= 1”
Fillet welds, L =12” HSS-8x3x3/8
Pu

Pu
16 ft
Cross Section
Problem 5.27. Calculate capacity of the built-up compression member shown below. Steel is A572 Gr 50.
The built-up shape is made of two W16x31 sections connected to each other by welded “batten” plates at 1/3
points as shown below.

Fixed End Center-to-Center 2W16x31 Pinned

Distance=8 inches A913, Gr 65 End

4 feet 4 feet 4 feet

(A)

Fixed Center-to-Center 2W12x120 Pinned End

End Distance=6 A913, Gr 65
inches

3 feet 3 feet 3 feet

(B)

Problem 5.28 Calculate factored axial load capacity of the built-up column shown on the right. The total
length of the column is 12 feet. The center-to-center distance of two adjacent batten plates (stitches) is three
feet. Steel is A572, Grade 50. The two wide flange sections connected to each other are both W14x132. The
batten plates are welded. The center-to-center distance of two wide flanges is 24 inches. The base of the
column is fixed in both x-x and y-y direction of buckling. Top of the column can move horizontally and
rotate when considering buckling about y-y axis. However, when considering buckling about the x-x axis
top of the column should be considered free to rotate but restrained against horizontal movement. Use the
AISC approximate chart to establish effective length factors K.

Pu
See the text for
support condition
here.

x
y

y
x

Problem 5.29. Calculate capacity φcPn of the built-up compression member shown. Steel is A36. Top and
bottom supports are pin connections. Stitches are welded to channels and are spaced every 20 inches.
Lateral bracing is provided at a point 15 feet from the top of the column as shown in the figure.

y
x
x
y
y
Stitches welded to
channels
MC18x58

x x 15 ft
20 ft
Bracing
Stitch placed
20 inches
center to Spacing=1" x y
center y
Cross Section 5 ft
y
x
Lateral Bracing at
mid-height
Member
CHAPTER 5-COMPRESSION MEMBERS
5-49

Problem 5.30. Using LRFD, check all applicable failure modes and calculate maximum factored load Pu
that can b applied to the column. The column is made of two W12x120 built-up cross sections as shown.

Pin about y-y, y

Fixed about x-x

16’
x
2 feet x
Bracing

Welds
16’ y 2W12x120, A572
Batten Plates , Gr. 50 Steel
y (snitches)
x
Fixed base
Cross Section
x
y

Column
Problem 5.31. Using LRFD, check all applicable failure modes and calculate maximum factored load Pu
that can be applied to the column. The column is made of four Round HSS 5.000x0.500, A500 Gr.B pipes.

y
Pu Pin about y-y,
Fixed about x-x
35”

x 25”
x
2@4 ft
Bracing

Welded y
Batten Pipe
Plates
4@4 ft
Cross Section
y x Fixed base in both directions

y
x
Column

CHAPTER 6

FLEXURAL MEMBERS

Getty Images

Steel girders and beams in a steel Structure

Objectives:
In this chapter you will learn:
1. The definition of flexural members (beams) and what type of
cross sections are used as flexural members.
2. Behavior of beams under bending.
3. Failure modes (or limit states) of steel beams.
4. How to design a beam for bending.
5. Application of concepts and equations to actual design of beams
through the use of numerical examples provided throughout the
chapter.

6.1. INTRODUCTION
This chapter covers behavior and design of steel members subjected to bending. Chapter 7
discusses behavior and design of beams subjected to shear. Behavior and design of steel members
under combined bending and axial is in Chapter 8.
Figure 6.1 shows typical applications of steel beams in buildings, bridges and other structures.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-2

Girder

Beam

(a) BEAMS IN BUILDINGS

Steel Box
Girders
Steel Girders

Major Bridge Girders

Overpass Girders Elevated Freeways

(b) BEAMS IN SHORT AND MEDIUM SPAN BRIDGES

Figure 6.1. Beams and Girders in Typical Steel Building and Bridge Structures
CHAPTER 6 – FLEXURAL MEMBERS 6-3

6.2. STEEL FLEXURAL MEMBERS (BEAMS)

The AISC Specification (2010a) in its Chapter F, which is devoted to “Design of Members for
Flexure”, has provisions for design of what it calls “members subject to simple bending”. The
Specification defines the members subjected to simple bending as:
1. Members that are subjected to bending about only one principal axis.
2. The member is loaded in a plane parallel to a principal axis and the load plane passes
through the shear center of the beam and is restrained against twisting at load points
and supports.
In other words, if a member is subjected to bending about both principal axes or subjected
to bending about one principal axis but the load does not pass through its shear center, the design
procedures given in this chapter do not apply to such member.
Figure 6.2 shows cross sections of beams that are considered “members subject to simple
bending” and are covered in the AISC Specification (2010a). We have grouped the cross sections
into nine groups.

y y y
y
x x x x x x
Channels
Doubly I-shaped Sections Bending
Doubly Doubly Bending About About Their
Symmetric Symmetrical
Symmetric Built-up Their Minor Axis Minor Axis
I-Shape 2-C Section
Shapes
Group 1 Group 2

x x
x x x x x x
x x

Square or Singly Singly Round HSS

Box-
Rectangular Symmetric Symmetric or Pipe
shaped
HSS Built-up Shapes
Member I-Shape
Group 3 Group 4 Group 5

x x x x x x x x x
x x x
x x Shear
x x Shear
Center Center
Tees and Double Single Rectangular
Angles Loaded in Angles Bars Un-symmetrical
and Rounds Shapes
Plane of Symmetry

Group 6 Group 7 Group 8 Group

Figure 6.2. Examples of Beam Cross Sections Subjected to “Simple Bending”

In developing design procedures included in the AISC Specification (AISC, 2010a) for
flexural members, the following assumptions are made:
1. Members are prismatic, which means the beam has constant cross section throughout
its length.
2. The point of supports for beams and girders are restrained against rotation about their
longitudinal axis.
3. Members are only subjected to bending about the principal axes of the cross section
and the load passes through the shear center of the cross section.
4. The load is applied relatively slowly and can be considered static load. No inelastic
load reversal or fatigue is present. Elastic load reversals of less than 20,000 cycles
during life of a structure are considered static.

6.3. BEHAVIOR OF STEEL BEAMS IN BENDING

When a steel beam is subjected to bending moment, as shown in Figure 6.3, first it behaves
elastically, develops tensile and compressive strains (Figure 6.3(a)) and stresses (Figure 6.3(b))
normal to the cross section. During the elastic behavior, stress distribution is linear which means
stress can be calculated from Hooke’s Law as the strain times modulus of elasticity (σ = ε E). As
the applied moment increases, the strains and stresses acting on the cross section also increase
until the stress in one or both extreme fibers reaches yield stress and cross section starts yielding,
Figure 6.3(c). The moment that causes yielding of the first extreme fiber is called “yield
moment” and is denoted as M y . If applied bending moment increases beyond the M y , more and
more fibers yield, Figure 6.3(d). Eventually, as more and more fibers yield, the member reaches
its maximum capacity and stress distribution is more or less as shown in Figure 6.3(e).
In design, the stress distribution shown in Figure 6.3(e) is replaced by a much simpler
rectangular stress distribution shown in Figure 6.3(f). The moment resulting from the rectangular
stress distribution, where all fibers have a uniform stress equal to the yield stress, is called “plastic
moment” and is denoted as M p .
-Fy

E. N.A. C
E. N.A.

P. N.A.

+Fy +Fy +Fu +Fy

Beam Side View Strain Stress Stress Stress Actual Assumed

Cross During at Yield During Stress Plastic Stress
Section Elastic Point Inelastic at Maximum Distribution
Behavior Behavior Capacity In Design

(a) (b) (c) (d) (e) (f)

Figure 6.3. Strain and Stress Distribution on a Beam Section as Bending Moment Increases
CHAPTER 6 – FLEXURAL MEMBERS 6-5

In the AISC Specification (2010a), M p is considered maximum moment that a cross

section can develop. Of course, due to other failure modes of beams, such as lateral–torsional
buckling or local buckling of compression flanges or web, which we will discuss shortly, the
bending capacity of a cross section can be less than M p . Notice that during elastic behavior, the
elastic neutral axis (E.N.A.) of cross section, where stress is zero, remains at the centroid (C.G.) of
the cross section. After yielding of cross section starts, for doubly symmetric sections, the neutral
axis continues to remain on the centroid until plastic moment M p is reached. However, for singly
symmetric and unsymmetrical sections, after yielding, the neutral axis does not remain on the
centroid and moves toward the yielded part of the cross section. For example, for singly-
symmetric cross section in Figure 6.3, after yielding, the neutral axis moves downward. As
yielding continues, movement of the neutral axis also continues until the entire cross section has
yielded, Figure 6.3(f). The neutral axis for this case is called “plastic neutral axis” (shown as
P.N.A. on figure 6.3(f)).

Yield Moment, M y : is defined as the bending moment that causes the first fiber (extreme fiber)
of the cross section to yield as shown in 6.3(c) and again Figure 6.4(b) below.

C
x x E.N.A.

Elastic
Neural Axis
Centroid of
My c
the cross section

+Fy
Beam Side View Only this extreme fiber
Cross Section has yielded.

(a) (b)

Figure 6.4. Stress Distribution when Beam Reaches Yield Moment My

Yield moment, M y , can be calculated by taking moment of compression and tension forces
with respect to elastic neutral axis. The stress distribution to be considered at the time of yielding
is the linear distribution of Figure 6.4(b), where extreme fiber has just yielded.

Yield moment of a section is given by:

M y = (F y ) (I x /c) (6.1)
where
M y = yield moment
F y = specified yield stress of steel
I x = moment of inertia of the cross section with respect to axis of bending, which is x-x axis in
Figure 6.4(a).
c = largest distance of from neutral axis to two extreme fibers, see Figure 6.4(b).
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-6

I x /c = elastic section modulus about neutral axis. Denoted as Sx, the I/c values for standard rolled
shapes are given in the AISC Manual Tables (AISC, 2011).
Plastic Moment: is defined as the moment that causes all fibers of a cross section to have a
stress equal to yield stress as shown in Figure 6.3(e) and again in Figure 6.5.

Ac = Comp. Area
-Fy
Plastic
Elastic Neural
Neural Axis Cp
Axis
x x

Plastic Mp
Neural Axis yp

Tp
At = Tension Area
+Fy

Beam Cross Section Side View

(a) (b)

Figure 6.5. Stress Distribution when Bending Moment is Equal to Plastic Moment Mp

When plastic moment is acting on a beam, the internal tension force T p and compression
force C p , see Figure 6.5, should be equal for the cross section to be in equilibrium. To find M p , the
first step is to find the location of “plastic neutral axis”. Plastic neutral axis is an axis
perpendicular to the plane of bending such that it divides the cross section into tension and
compression zones such that the total tension force in the tension zone is equal to total
compression force in the compression zone.
To find the location of plastic neutral axis we can use equilibrium of tension and
compression forces:
Tp= Cp (6.2)

Plastic
Neural Mp
Axis

di
Element “i”
Fy Ai

Figure 6.6. Forces Acting on Elements of Cross Section When Mp is Reached

CHAPTER 6 – FLEXURAL MEMBERS 6-7

where
T p = total tension force due to tension yielding of all fibers on one side of the plastic neutral axis
= AT Fy
C p = total compressive force due to compression yielding of all fibers on the compression side of
the plastic neutral axis = A C F y
A T = area of tension zone of cross section, (see Figure 6.5 on previous page)
A C = area of compression zone of cross section, (see Figure 6.5)
As mentioned earlier, for double symmetric sections the elastic and plastic neutral axes
are the same and pass through the centroid (C.G.) of the cross section, but for singly symmetric
sections, elastic and plastic neutral axis are not necessarily the same and generally are at different
locations. Elastic neutral axis always passes through centroid of the cross section while plastic
neutral axis is an axis that divides the cross section into two tension and compression parts such
that the tension and compression forces acting on these two parts are equal.
After location of the plastic neutral axis is found, plastic moment capacity of the section is
given by:
M p = ΣF y A i d i (6.3)
Where:
M p = plastic moment
F y = yield stress of steel
A i = area of one element of cross section
d i = distance of centroid of area A i from plastic neutral axis, see Figure 6.6 above.

If all elements of cross section are made of the same steel, F y will be the same for all of them and
Equation 6.3 above can be rewritten as:

M p = F y Σ Ai di (6.4)
The term Σ Ai.di, which is a geometric property is called plastic section modulus and is
denoted as Zx:
Zx = Σ Ai di (6.5)
The cross section tables in the AISC Manual (AISC, 2011) provide values of Zx, and Zy,
the plastic section modulus for bending about xx and yy axes respectively for standard rolled
shapes.

Illustrative Example 6.1:

Calculation of Yield and Plastic Moments
Calculate elastic and plastic section modulus as well as yield and plastic moments for the cross section
shown with respect to its x-x axes. All steel is A36.

Given:
1. Cross section of the member
2. Material of steel as A36 with Fy = 36 ksi and Fu = 58 ksi.

Required:
To calculate yield moment My and plastic moment Mp.

SOLUTION:
Establish Location of Centroid (Note that this cross section is the same as the section in
Illustrative Example 5.1 in Chapter 5):
ye = (ΣAi di ) / (ΣAi )= (20”×1”×25.5”+24”×0.5”×13”+12”×1”×0.5”) / ( 20”×1”+24”×0.5”+12”×1”) = 15.3 in.

y
Centroid
10” 10”
Plastic Neutral Axis
Elastic Neutral Axis
1”
x x
24”
ye 1/2”
Yp

1”

6” 6”
Calculate Ix
y
Ix = Σ(bh3 / 12+Ad 2 )i
Ix = 20”×(1”)3 /12+20”×1”×(10.2”)2 +0.5”×(24”)3 /12+24”×0.5”×(2.3”)2 +12”×(1”)3 /12 +12”×1”×(14.8”)2
Ix = 5,351 in4

Calculate Elastic Section Modulus, Sx

Sxc = elastic section modulus w.r.t. compression flange
Sxc = Ix /dc = (5351 in4)/(26”-15.3”) = 500 in3.
Sxt = elastic section modulus w.r.t. tension flange
Sxt = Ix /dt =(5351 in4)/15.3”=350 in3. 20”

P.N.A. 10” 10” 1/2”

1” C1
Calculate Plastic Section Modulus, Zx C2 dc
4”
Locate Plastic Neutral Axis:
T =C 1/2” 24” T1
Yp=21” 20.5” dt
T =(Yp -1”)(0.5”)(Fy )+(1”×12”)(Fy)
C =20”×1”×Fy +0.5”×(25”-Yp)(Fy) 1” T2

T =C results in: 6” 6”
1/2”
Yp = 21 in. 12”
Zx = Mp /Fy =Σ(btd)i
Zx =20”×1”×4.5”+4”×.5”×2”+ 20”×0.5”×10”+12”x1”×20.5”
Zx =440 in3

Calculate My and Mp:

My = yield moment= Minimum of (Sxc Fy) and (Sxt Fy )
My =(350 in3) (36 ksi)=12,600 k-in
Mp = plastic moment = Zx Fy = (440 in3) (36 ksi) =15840 k-in.
CHAPTER 6 – FLEXURAL MEMBERS 6-9

6.4. LIMIT STATES OF FLEXURAL MEMBERS

As the bending moment acting on a cross section increases, it is possible that the beam fails
through other modes before reaching its maximum plastic moment capacity of Mp. When a failure
mode is reached, depending on the nature of the failure mode, the strength and/or stiffness of the
member are reduced significantly and the member is called "failed". The main limit states of
failure of a beam are:
a. Limit State of Bending Capacity.
b. Limit State of Deflection.
c. Limit State of Shear Capacity.
d. Limit States of Connection Capacity.

In this chapter we focus on Limit State (a) in the above list which is reached when the
beam fails in bending. Limit State (b), which is reached when beam deflects beyond permissible
level, is not a strength limit state. Limit State (c), which is reached when a beam fails in shear , is
discussed in Chapter 7. Limit State (d) is related to the failure of connections, which are discussed
in later chapters of this document.

6.5. LIMIT STATES OF BENDING CAPACITY FOR

I-SHAPES
When a beam is subjected to bending, it can fail in five different ways. These five limit states of
bending capacity are shown in Figure 6.7 and are:
(a) Yielding and development of plastic hinge.
(b) Local buckling of the compression flange.
(c) Lateral torsional buckling of the compression flange.
(d) Local buckling of the compression region of the web.
(e) Yielding of tension flange in singly symmetric sections.

The nominal bending strength, Mn , of a beam established based on the limit state of
Yielding and development of plastic hinge is the largest capacity among all limit states. Each of the
above limit states and corresponding bending capacity are discussed in the following sections. The
lowest bending capacity among all limit states is the capacity of the beam. Typical moment-
rotation behavior of beams is shown in Figure 6.8.

6.6. LIMIT STATE OF YIELDING AND DEVELOPMENT

OF PLASTIC HINGE
6.6.a. Yielding and Development of Plastic Hinge in a Beam
This limit state is reached when most areas of the cross section of a beam have yielded either in
tension or compression and have lost their bending stiffness. As a result of loss of stiffness, the
cross section will be free to rotate like a hinge. This is why this limit state is called “Development
of Plastic Hinge”. Notice that the difference between an actual hinge and a plastic hinge is that the
moment in an actual hinge is zero but in a plastic hinge, the moment is equal to plastic moment
Mp.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-10

The actual stress distribution on the cross section where a plastic hinge has formed is close to what
is shown in Figure 6.9 (b), but, as mentioned earlier, the much simplified stress distribution shown
in Figure 6.9(c) is used in design to calculate plastic moment capacity Mp of a beam section.

(a) Yielding of
Cross Section,
Plastic Hinge
Formation

(b) Flange
Local
Buckling

(d) Web
Local
Buckling

(e) Yielding
of Tension
Flange

Figure 6.7. Limit States of Failure of Beams in Bending

Full yielding and development θ

Bending of plastic hinge (Ductile)
Moment
Mp
M
My C

Mcr B Inelastic TFY, LTB, FLB or

WLB (some ductility present)
A Notes
LTB=Lateral Torsional Buckling
Elastic LTB, Elastic FLB or Elastic WLB FLB=Flange Local Buckling
(Brittle) WLB=Web Local Buckling
TFY=Tension Flange Yielding

O Rotation

Figure 6.8. Typical Ductile and Brittle Behavior of Beams in Bending

CHAPTER 6 – FLEXURAL MEMBERS 6-11
-Fy
-Fy

Mp C

P. N.A.
T

Plastic hinge +Fy

+Fy
has formed here

(a) Plastic hinge (b) Actual stress (c) Assumed stress

at midspan distribution when distribution to
Mp is reached Calculate Mp

Figure 6.9. Limit State of Reaching Plastic Moment Capacity

6.6.b. Design for Limit State of Yielding and Development of Plastic Hinge
The equation of design for checking yielding of cross section and development of plastic hinge in
beams bending about their major axis (x-axis) is:

Mu ≤ φb Mn (in LRFD) and Ma ≤ Mn /Ωb (in ASD) (6.6)

where
Mu = required bending strength in LRFD, same as the maximum factored applied bending
moment.
φb = resistance factor for bending (in LRFD) = 0.90
Ma = required bending strength in ASD, same as the maximum un-factored applied bending
moment.
Ωb = factor of safety for bending (in ASD) = 1.67
Mn = nominal bending moment capacity equal to plastic moment capacity, Mp, given by:

Mp =Fy Zx (6.7)
Where
Mp = plastic moment capacity of cross section of the beam
Fy = specified minimum yield stress of steel being used
Zx = plastic section modulus of cross section of the beam about x-axis
M = required bending strength in ASD, same as the maximum service (un-factored) applied
bending moment.
Ωb = safety factor for bending (in ASD) = 1.67

Illustrative Example 6.2:

Evaluating Strength of a Beam Based on Yielding and
Formation of Plastic Hinge
The beam shown below is a W24x162, A992 grade 50 steel. Checking only limit state of yielding and
formation of plastic hinge, calculate factored load Puthat can be applied to the section. Ignore self-
weight of the beam.

W24x162, A992, Gr. 50 steel

Pu Pu y

x x

15 ft 20 ft 15 ft y

Given:
1. Cross section of the member as W24x162
2. The material is given as A992 grade 50 steel with Fy = 50 ksi and Fu = 65 ksi.

Required:
To establish bending capacity of the member in terms of applied load Pu.

SOLUTION:
Properties of the W24x162 are: Zx = 468 in3

Check Limit State of Yielding and Formation of Plastic Hinge:

The equation of design for limit state of yielding is Mu ≤ φb Mn where:
Mn =Mp = Fy Zx .
Mn =50 ksi x 468 in3 = 23,400 k-in
Mu=(15 ft x12 in/ft )(Pu)=1800 Pu k-in.

Mu =1,800Pu k-in

1,80 Pu k-in = 0.9(23400 k-in.)

Pu = 117 kips 15 ft 20 ft 15 ft

Bending Moment-Diagram

6.7. LIMIT STATE OF LOCAL BUCKLING OF

COMPRESSION FLANGE IN DOUBLY SYMMETRIC
MEMBERS
6.7.a. Local Buckling of Compression Flange of Beams
This section provides information and design procedures on local buckling of compression flange
of doubly symmetric , I-shape and double channel members. The beam has compact web which
means the web does not buckle locally. Figure 6.10(a) shows bending stresses acting on a beam
cross section. In this case the top flange is under compression. When compressive stresses acting
on a flange reach a critical value, the flange can locally buckle as shown in Figure 6.10(b).
CHAPTER 6 – FLEXURAL MEMBERS 6-13

(a) Local Buckling of Compression (b) Close-up of Locally

Flange in a Beam Buckled Area

Figure 6.10. Local Buckling of Compression Flange

As we discussed in Chapter 5 for columns, the critical compressive stress that causes
elastic local buckling of a plate subjected to uniform compression is given by:

Kπ E
2

Fcr = (6.8)
b 2
12 (1 − ν ) (
2
)
where t
Fcr = critical value of stress on the beam compressive flange that causes its local buckling.
K = a coefficient related to boundary conditions of the plate and type of the applied
stress. Values of K for typical cases of loading and boundary conditions are shown in
Figure 5.9 in Chapter 5 (Page 5-8).
ν = Poisson’s ratio equal to 0.3 for steel behaving elastically
b = width of the plate element
t = thickness of the plate element
Values of b, and t in the above equation for elements of common cross sections, are
measured as shown in Figure 5.10 in Chapter 5. Again, similar to column flanges, if Fcr , given by
the Equation 6.8 above, is less than the yield stress of steel, Fy, then local buckling of the flange
will occur prior to yielding. However, if Fcr is greater than Fy, then, the compression flange will
yield before it locally buckles.
The critical bending moment that can cause local buckling of the compression flange is
equal to critical stress, Fcr, times the elastic section modulus:
Kπ ES x
2

M n = Fcr S x = (6.9)
b 2
12 (1 − ν ) (
2
)
t

The above equation can be simplified by using a Poisson’s ratio, ν, equal to 0.30 for steel and the
numerical value of π to obtain:
0.9 Ek c S x
Mn = (6.10)
λ2
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-14

where
λ = b/t of compression flange
Sx = elastic section modulus of cross section
4
kc = 0.5
, and should not be taken less than 0.35 nor greater than 0.76.
 h 
 
 t 
h = height of flat portion of the web as shown in the following sketches.
Mn = nominal bending capacity due to limit state of local buckling of compression flange.
The value of Mn cannot exceed Mp , plastic moment capacity of the cross section.
tf k

d h
d h t
t

k
tf In Rolled Sections:
h=d–2k
In Welded Sections:
(Values of k are given in the AISC Manual Tables on
Cross Section Properties)
h = d – 2 tf

Figure 6.11. Dimension “h” for Built-up and rolled I-shaped Sections

Figure 6.12 shows variation of moment capacity, Mn, versus b/t ratio of compression
flange. Figure 6.12 also shows approximate region of test results. When b/t ratio is relatively
large, Equation 6.10 can predict the critical bending moment, Mcr, quite well. However, for smaller
values of b/t ratio, the bending moment capacity predicted by Equation 6.10 is greater than the test
results. For the region that test results are less than predictions of Equation 6.10, the AISC
Specification (2010a) uses a straight line, given by Equation 6.11 below, to define the values of
bending moment capacity, Mn. Also, notice that for very small values of b/t ratio, where test results
are greater than the plastic moment capacity, Mp , the bending moment capacity is taken to be
equal to plastic moment, Mp.

  λ − λ pf 
M n =  M p − ( M p − 0.7 F y S x )  (6.11)
  λ rf − λ pf 

Figure 6.13 shows the equations in the AISC Specification (2010a) that defines the value
of nominal bending moment capacity in terms of λ= b/t ratio of compression flange.
Depending on λ, the b/t ratio of the compression flange, the beam sections are divided into
three categories of compact, non-compact, and slender, Figure 6.13. If b/t ratio of the compression
flange is less than λp given in Table B4.1b of the AISC Specification (2010a), (reproduced on the
previous page) then the flange is compact and cross section is not expected to locally buckle prior
to reaching plastic moment capacity, Mp.
CHAPTER 6 – FLEXURAL MEMBERS 6-15
Table 6.1. Values of λp and λr for Beams
(Table B4.1b of the AISC Specification (AISC, 2010a) )

Mn Local Buckling of
Compression
Flange
Mp 0.9 Ek c S x
Mn = ≤Mp
λ
2

Approximate
region of test
results
Straight line
curve-fitted to the
test results

b/t

Figure 6.12. Variation of Critical Bending Moment, Mn versus b/t of the Compressive Flange

If b/t ratio of compression flange is greater than λp but less than λr given in Table B4.1b of
the AISC Specification (2010a), then the flange is non-compact and cross section is expected to
locally buckle before reaching plastic moment capacity, Mp. Non-compact beams develop inelastic
local buckling which means local buckling of compression flange occurs after some fibers of the
cross section, with compressive residual stresses in them, have already yielded.
If b/t ratio of compression flange is greater than λr given for flexure in Table B4.1b of the
AISC Specification (2010a), then the flange is slender and compression flange is expected to
locally buckle in an elastic mode before yielding of any fiber. Similar to non-compact case, in
this case also, the nominal bending capacity, Mn is less than the plastic moment capacity Mp.

2. Inelastic local buckling of

flange is expected.
Mn 1. No local buckling
of flange occurs.   λ − λ pf 
M n =  M p − ( M p − 0.7 Fy S x ) 
 λ rf − λ pf
Mn =Mp
 
Mp
3. Elastic local buckling of flange
is expected.
0.9 Ek c S x
1. Mn =
Compact
2. Non- λ2
compact
sections sections 3. Slender sections

λp λr b/t of Compression Flange, λ

Note: For values of λpf and λrf which are λp and λr values for the flange, see
Table B4.1b of the AISC (reproduced earlier as Table 6.1).

Figure 6.13. Three Regions of Behavior for Local Buckling of the Compression Flange
as per AISC Spec. (2010a)
CHAPTER 6 – FLEXURAL MEMBERS 6-17

6.7.b. Design for Limit State of Local Buckling of Compression Flange

(AISC, 2010a)
The design procedures given in this section apply only to local
buckling of compression flange of doubly symmetric I-shape and
double channel sections with compact web bending about their major
axis. The equation of design for such cases is:

M u ≤ φb M n (in LRFD) and Ma ≤ Mn /Ωb (in ASD) (6.12)

where
Mu = required bending strength in LRFD, same as the maximum factored applied bending moment.
φb = resistance factor for bending (in LRFD) = 0.90
Ma= required bending strength in ASD, same as the maximum service (un-factored) applied
bending moment.
Ωb = safety factor for bending (in ASD) = 1.67
Mn = nominal bending moment capacity given by following equations and on Figure 6.13 above:

a. When λ ≤ λpf , cross section is compact, no local buckling of flange is expected, and:

Mn =Mp (6.13a)

b. When λpf < λ ≤ λrf , cross section is non-compact, inelastic local buckling of flange is
expected and:
  λ − λ pf 
M n =  M p − ( M p − 0.7 F y S x ) 
 λ rf − λ pf
(6.13b)
 

c. When λ > λrf , cross section is slender, elastic local buckling of flange is expected and:

0.9 Ek c S x
Mn = (6.13c)
λ 2

where
λ = b/t ratio of compression flange
λpf = same as λp , limit of b/t ratio for compact elements given in Table B4.1b of the AISC
Specification (AISC, 2010a) reproduced as Table 6.1 earlier on Page 6-15.
λrf = same asλr , limit of b/t ratio for slender elements given in Table B4.1b of the AISC
Specification (AISC, 2010a) reproduced as Table 6.1 earlier on Page 6-15.
Sx = elastic section modulus of cross section
4
kc = 0.5
, which should not be taken less than 0.35 nor greater than 0.76.
 h 
 
 t 

For values of h and t see sketches at the bottom of Page 6-14.

Illustrative Example 6.3:

Evaluating Strength of a Beam Based on Local Buckling of
Flange
Considering only limit state of local buckling of compression flange, find maximum value of factored
load Pu that can be applied to the beam. Ignore self-weight of the girder.

Pu PL14x1/2”, A36 y
Both Flanges

Weld both
sides
A B
C
8 ft 8 ft
x x

Given: PL 30x3/8, A36

Web
1. Cross section of the member is given
2. The material is given as A36 steel with Fy = 36 ksi and Fu = 58 ksi. y

Required:
To establish bending capacity of the member in terms of applied load Pu.

SOLUTION:
PL14x1/2”, A36 y
Establish Properties of the I-shape Section: Both Flanges
bf =14 in, tf = 0.5 in, tw = 0.375 in, h = 30 in. , d = 31 in.,
Weld both
sides
Calculate A, Ix , Sx , and Zx
A = 2x14”x0.5” + 0.375”x30” = 25.25 in2 x x

Ix = Σ(bh3/12+Ad2 )i
PL 30x3/8, A36
Ix = 0.375”x(30”)3 /12 + 2x14×0.53 /12 + 2x14”×0.5”×(15.25”)2
Web
Ix = 4,050 in4
≈ 0.0
y
Sx = elastic section modulus = Ix /C= 4,050 in4 / 15.5 in.= 264.5 in3.
Zx =ΣAi di = 2x14”×0.5”×15.25” + 2x15”×0.375”×7.5” = 298 in3
Mp = plastic moment = Zx Fy = (298 in3) (36 ksi) =10,728 in.

1. Check Local Buckling of Web:

h/tw = 30/0.375= 80 < λp = 3.76(E/Fy)0.5 = 3.76(29000 ksi / 36 ksi)0.5 = 90
Since h/tw < λp web is compact

2. Check Local Buckling of Compression Flange:

λ = (bf /2) / tf =(14“/2)/ (0.5“) =14
λpf = 0.38(E/Fy)0.5 = 0.38(29000 ksi / 36 ksi)0.5 = 11
λrf = 0.95(kcE/FL)0.5
kc = (4.0) / (h / t )^0.5 = (4.0) / (30”/0.375”) = 0.447.
Since kc is greater than 0.35 and less than 0.76, we will use this value of kc in the calculations.
CHAPTER 6 – FLEXURAL MEMBERS 6-19

FL is equal to 0.7Fy if Sxt / Sxc > 0.7. (see footnote “b” in Table B4.1b on Page 6-15). Sxt and Sxc are
elastic section modulus with respect to tension and compression extreme fibers. For this doubly
symmetric section, they are equal. Since Sxt / Sxc = 1.0 is greater than 0.7 then, FL is equal to 0.7Fy.
λrf = (0.95) [ (0.447 x 29000 ksi ) / (0.7 x 36 ksi) ]0.5 = 21.5
Since (λpf =11) < (λ=14) ≤ (λrf =21), then, compression flange is “non-compact” and local buckling
failure is expected. Mn is given by:
Pu

Bending Moment Diagram  A B

C
8 ft 8 ft

  λ − λ pf 
M n =  M p − ( M p − 0.7 F y S x ) 
 λ rf − λ pf
Mu = (Pu )(16 ft)/4= 4Pu k-ft
 
Mn = 10,728 k-in – (10,728 k-in – 0.7x 36 ksi x 264.5 in3 ) [(14-10.8) / (21.5-10.8)] = 9,513.3 k-in
Using equation of design we can find Pu :
Mu ≤ φb Mn
Mu = 4Pu k-ft = 48Pu k-in.
48 Pu k-in ≤ 0.9(9,513.3 k-in.) ⇒ Pu ≤ 178.3 kips (due to local buckling of comp. flange)

6.8. LIMIT STATE OF LATERAL-TORSIONAL

BUCKLING DOUBLY SYMMETRIC, I-SHAPED
MEMBERS WITH COMPACT WEB
6.8.a. Lateral Torsional Buckling of Beams
This section provides information and design procedures on lateral torsional buckling of doubly
symmetric , I-shape and double channel members, shown in Figure 6.10, with compact web. For a
web to be compact, the h/t ratio of the web should not exceed the λp values given in Table B4.1b
of the AISC Specification (AISC, 2010a), which was reproduced earlier as Table 6.1 on Page 6-
15.
When a beam is bending, the compression flange acts as a column, Figure 6.14(a). In most
applications, like simply-supported beams in floors of buildings, the compression flange is
continuously connected to the floor slab and cannot buckle in horizontal direction. But, if the
compression flange is braced laterally with relatively long length between the brace points, then,
the compression flange acting as a column can buckle laterally between the braced points, Figure,
6.14(b). When compression flange starts buckling laterally, since it is attached to the web and the
tension flange, it cannot freely move in just the lateral direction but it has to rotate as well. That is
why this limit state is called Lateral-Torsional Buckling.
Lateral torsional buckling of compression flange occurs when the maximum stress in the
compression flange reaches a critical buckling value causing the flange to buckle laterally and
twist, developing lateral-torsional deformations, Figure 6.14(b). The critical stress in the
compression flange is related to geometry of the compression flange and part of the compression
zone of the web attached to it and the un-braced length of this T-shaped compression flange,
Figure 6.15(a).
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-20

Lateral Bracings
Floor Slab For Compression

Comp. Flange
Lateral Torsional
Buckling

(a) Floor slab braces compression flange, (b) Unbraced length of compression flange
preventing lateral torsional buckling buckles in lateral torsional mode

Figure 6.14. Limit State of Reaching Lateral Torsional Buckling of Compression Flange

Compression hc /6
“T-shape” area
considered in
lateral torsional C C
buckling hc /2 M M
x x T
T
Elastic Unraced Length
Neural
Axis

(b) An un-braced length of a beam

(a) A typical I-section before and after
subjected to uniform bending moment
lateral torsional buckling.

Figure 6.15. Lateral Torsional Buckling of an I-shaped Section

Let us consider an un-braced length of a beam subjected to equal and opposite bending
moments at its ends, Figure 6.15(b). The equal moments at the ends of the beam cause uniform
bending moments throughout the beam. The uniform bending moment creates constant
compression throughout the length of the compression part of the cross section. As mentioned
earlier, the compression zone of the cross section for lateral torsional buckling is assumed to be
just the dark “T-shaped” area in Figure 6.15(a) instead of the entire compression area above the
neutral axis.
As the moment M applied to the ends of the beam in Figure 6.15(b) increases, the
compressive force C in the T-shaped compressive flange also increases. When applied moment
reaches a critical value, Mcr, the T-shaped compressive flange buckles laterally and twists
resulting in lateral torsional buckling of the un-braced length of the beam. By solving differential
equation of lateral torsional buckling the critical moment, Mcr , which causes lateral-torsional
buckling of a beam subjected to uniform bending moment, is obtained as:

 π2E  
 I y  w +
C GJ
M cr =   L2

 Lb   b π E
2

(6.14)
 
CHAPTER 6 – FLEXURAL MEMBERS 6-21

where
Mcr = critical uniform bending moment acting on the I-beam causing lateral torsional buckling
E = modulus of elasticity of steel= 29,000 ksi.
Lb = un-braced length of the compression flange
Iy = moment of inertia with respect to y-y axis (weak axis).
Cw = warping constant which is a torsional property of the cross section. Values of Cw for typical
steel sections are given in the literature. For doubly symmetric sections: C w = I ho / (2S x )
G = shear modulus = E/[2(1+ ν)]
ν = elastic Poisson’s ratio = 0.30 for steel
J = torsion constant, also called torsional stiffness. Values of J for typical steel sections are given
in the literature (CISC, 2002). For rolled shapes, values of J are given in the AISC Manual
(AISC, 2011)
To obtain critical compressive stress, Fcr, in the flange, we need to divide critical moment
by elastic section modulus, Sx :

π 2E   
Fcr =   I y  C w + GJ  / S x (6.15)
 L   L2 π 2 E 
 b   b 

If we multiply and divide the right side of the above equation by (Lb /rts2) the above
equation takes the form of Equation 6.16:

2
π E I y  Cw   Lb 
2
GJ
Fcr =  + 2 2 2  
2
S x2  rts4  π ES r  rts  (6.16)
 Lb   x ts  
 
 rts 
 
By substituting various terms in the above equation with their values, as given in Figure
6.16 below, we obtain Equation 6.17, which is a simplified version of above equation for Fcr.

2
π 2E Jc  Lb 
Fcr = 1 + 0.078  
 Lb 
2
S x ho  rts 
(6.17)
 
 rts 

y y
bf bf Iy = 2(tf bf 3 /12)= tf bf 3 /6
tf tf

hc /6 I yCw
= rts
2
hc /2
hc /6 Sx
x x
d ho h x d ho h x

tw Sx = (Ix )/(d/2) = [(2tf bf )(d/2)2 +tw d3 /12)]

tw Elastic /(d/2)
Neural Axis
Cw =Iy (ho2 /4)
y y G = E/[2(1+ ν] =E/2.6
Welded Sections Rolled Sections

Figure 6.16. Values of Cross Section Properties used in Lateral Torsional Buckling Equation

Notice the similarity of this equation with Euler’s column buckling equation which is
essentially the first part of the above equation without the square root term. The AISC
Specification (2010a) in a User Note states that the square root term in the above equation can
conservatively be taken as 1.0.
In the above equations:
Fcr = critical bending compressive stress for lateral-torsional buckling
ho = distance between the centroids of top and bottom flanges, Figure 6.16.
c = a coefficient related to cross section being doubly or singly symmetric:
c =1.0 for doubly symmetric sections
rts = radius of gyration of a T-shaped section, made of compression flange plus 1/6 of the web,
with respect to y-y axis. Value of the rts can also be calculated from either of following
equations:
bf
rts =
 1 ht w  (6.18)
121 + 
 6 bf t f 

I y ho (6.19)
rts2 =
2S x
For rolled shapes, values of rts are given in the AISC Tables for cross section properties.
Equation 6.17 defines critical buckling stress for lateral torsional buckling of a beam
subjected to uniform bending. Quite often, bending moment along the beam varies. To incorporate
the effect of variation of bending moment diagram along the length of the beam, Equation 6.17 is
multiplied by a factor, Cb , where Cb represents variation of bending moment along the length of
the beam. It should be mentioned that the case of uniform bending moment, used in derivation of
Equation 6.17, is the most severe case and Cb value for this case is equal to 1.0. For other cases
where bending moment varies along the length of the beam, Cb will be greater than 1.0. Therefore,
for general case of bending moment, Equation 6.17 is written as:

π2E
2
Jc  Lb 
Fcr = C b 2
1 + 0.078   (6.20)
 Lb  S x ho  rts 
 
 rts 

Value of Cb is given by the AISC Specification (2010a) as:

12.5M max
Cb = ≤ 3.0
2.5M max + 3M A + 4 M B + 3M C (6.21)

where
Cb = modification factor for bending moment gradient =conservatively equal to 1.0.
Mmax = absolute value of max moment in laterally un-braced segment, see Figure 6.17.
MA = absolute value of moment at ¼-point in laterally un-braced segment, see Figure 6.17.
MB = absolute value of moment at ½ -point in laterally un-braced segment, see Figure 6.17.
MC = absolute value of moment at ¾ -point in laterally un-braced segment, see Figure 6.17.
CHAPTER 6 – FLEXURAL MEMBERS 6-23

Lateral Lateral
Bracing Bracing

Lb =Unraced Length

Lb /4 Lb /4 Lb /4 Lb /4 Mmax

MC
MA MB

Bending Moment Diagram for Unbraced Length of the Beam

Figure 6.17 Examples of Bending Moment Diagrams and Values of Moments used to Establish Cb

Figure 6.18 shows Fcr given by Equation 6.20 plotted against the length of the laterally
un-braced segment of the beam, Lb. The range of test results is also shown in the figure. For
relatively large values of Lb, Equation 6.20 predicts critical stresses reasonably well. However, for
smaller values of Lb , test values are smaller than the values predicted by Equation 6.20. For the
region that test results are less than predictions of the Equation 6.20, the AISC Specification
(2010a) uses a straight line, shown as Line AB in Figure 6.18, to define the values of critical
lateral –torsional buckling stress, Fcr.

Fcr  1    Lb − L p 
Fcr =Fy
Fcr =  (Cb )  M p − ( M p − 0.7 Fy S x ) 
  L − L 
A  Sx   r p 

Fy
2
π2E Jc  Lb 
Fcr = Cb 2
1 + 0.078  
 Lb  S x ho  rts 
 
B  rts 

Approximate region of test

results

Unraced Length, Lb

Figure 6.18. Theoretical Critical Lateral-Torsional Buckling Stress versus Test Results
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-24

1. No LTB 2. Inelastic LTB

Mn
Mn =Mp   Lb − Lp 
M n = Cb  M p − ( M p − 0.7 Fy S x )  ≤ M p
  Lr − Lp 

Mp 3. Elastic LTB 2
π2E Jc  Lb 
M n = Fcr S x = C b S x 2
1 + 0.078   ≤ M p
0.7Fy Sx  Lb  S x ho  rts 
 
 rts 

Lp Lr
Unraced Length, Lb

Figure 6.19. Three Regions of Behavior for Lateral Torsional Buckling as per AISC Spec. (2010a)

The equation of straight Line AB, see Figure 6.18, fitted to test results is given by:

 1    Lb − L p 
Fcr =  (C b ) M p − ( M p − 0.7 F y S x )  ≤ F y (6.22)
 Sx    Lr − L p 

Notice that for very small values of Lb, critical stress for test results is greater than yield
stress and is conservatively taken as equal to Fy.
Figure 6.19 shows the equations in the AISC Specifications (AISC, 2010a) that define the
value of nominal bending capacity, Mn, for lateral torsional buckling of a beam segment with
laterally un-braced length of Lb.
The terms Lp and Lr in the above equations and in Figure 6.19 are given below. All other
terms in the above equation have been defined earlier in this chapter.
Lp = limiting value of un-braced length separating no Lateral Torsional Buckling and Inelastic
Lateral Torsional Buckling regions (see Figure 6.19). If un-braced length of compression
flange is less than Lp , which is given by the following equation, the beam is not expected to
have lateral torsional buckling before reaching plastic moment capacity.

E
L p = 1.76ry (6.23)
Fy

Lr = limiting value of un-braced length separating Inelastic Lateral Torsional Buckling and Elastic
Lateral Torsional Buckling regions (see Figure 6.19). Lr is given by:

2 2
E Jc  Jc   0.7 Fy  (6.24)
Lr = 1.95rts +   + 6.76 
0.7 Fy S x ho  S x ho   E 

The terms in the above equation were defined earlier.

CHAPTER 6 – FLEXURAL MEMBERS 6-25

6.8.b. Design for Limit State of Lateral Torsional Buckling of Doubly-

Symmetric Beams with Compact Web (AISC, 2010a)
The design procedures given in this section apply only to lateral
torsional buckling of doubly symmetric I-shape and double channel
sections with compact web bending about their major axis. The equation
of design for such cases is:

M u ≤ φb M n (in LRFD) and M ≤ Mn /Ωb (in ASD) (6.25)

where
Mu = required bending strength in LRFD, same as the maximum factored applied bending
moment.
φb = resistance factor for bending (in LRFD) = 0.90
M = required bending strength in ASD, same as the maximum service (un-factored) applied
bending moment.
Ωb = safety factor for bending (in ASD) = 1.67
Mn = nominal bending moment capacity due to lateral-torsional buckling given by following
equations and shown on Figure 6.19 in previous page:

a. When Lb ≤ Lp :
Mn =Mp (6.26a)

b. When Lp < Lb ≤ Lr :
  Lb − L p 
M n = C b  M p − ( M p − 0.7 F y S x )  ≤ M p (6.26b)
  Lr − L p 
c. When Lb > Lr : 2
π2E Jc L  (6.26c)
M n = Fcr S x = C b S x 2
1 + 0.078  b  ≤ M p
 Lb  S x ho  rts 
 
 rts 
The terms in the above equations were all defined earlier in this Chapter.

Illustrative Example 6.4:

Evaluating Strength of a Beam Based on Lateral-Torsional
Buckling
Find maximum value of factored load Pu that can be applied to the beam. The beam is a W21x62, A992
grade 50 steel. The top and bottom flanges of the beam are laterally braced at Points K, L, M and N as
shown below. Ignore self-weight of the beam.

Pu W21x62, A992, Gr 50
Lateral Bracing Pu y

x x
K N
L M 15 ft
15 ft 20 ft
y

Given:
1. Cross section of the member as W21x62
2. The material is given as A992 grade 50 steel with Fy = 50 ksi and Fu = 65 ksi.
3. The un-braced lengths are given as 15, 20 and 15 feet (see figure above).
Required:
To establish bending capacity of the member in terms of the applied load Pu.
SOLUTION:
Properties of the W21x62 are:
bf = 8.24 in, tf = 0.614 in, tw = 0.4 in, h= 18.38 in. , d = 21 in., ry = 1.77 in., Zx =144 in3 , Sx =127 in3,
J = 1.83 in4 , rts = 2.17 in.
1. Check Local Buckling:
bf / 2tf =8.24 / (2x0.614)= 6.71 < λp = 0.38(E/Fy)0.5 = 0.38(29000/50)0.5 = 9.1
h/tw =18.38/0.4=46 < λp = 3.76(E/Fy)0.5 = 3.76(29000 ksi / 50 ksi)0.5 = 90
Section is compact, no local buckling failure is expected.

2. Check Limit State of Lateral-torsional Buckling:

The un-braced length, Lb is equal to 15 ft, 20 ft and 15 ft for segment 1, 2 and 3 respectively (see
bending moment diagram below).

Lb = 15 ft Lb = 20 ft Lb = 15 ft
Segment 1 Segment 2 Segment 3
Lateral
Bracing
Mu=2.14
M-diagram M
N
K L

Mu=15Puk-in

Generally, we need to check all un-braced segments of a beam to establish the capacity. But, with some
observations, we may be able to avoid wasting time on checking some segments which will not govern.
For example, in this case, segments 1 and 3 have the same length of 15 feet, but segment 3 has much
larger maximum moment. Note that the value of Cb for both segments will be the same since both
segments have a triangular bending moment diagram. Therefore, we only will check segments 2 and 3.

Check Segment 2 (from L to M, see Figure in the previous page ):

Lb = un-braced length of segment = 20 ft = 240 in
Lp = limit of un-braced length for not having lateral-torsional buckling given by:

E
L p = 1.76ry
Fy

Lp =1.76 ×( 1.77 in.) ×(29000 ksi / 50 ksi )0.5 = 75 in. = 6.3 ft.
Lr = limit of un-braced length for elastic lateral-torsional buckling

2 2
E Jc  Jc   0.7Fy 
Lr = 1.95rts +   + 6.76 
0.7Fy S x ho  S x ho   E 
CHAPTER 6 – FLEXURAL MEMBERS 6-27

rts = 2.17 inch from the AISC Manual tables on cross section properties.
Note that if this was not a rolled shape, we had to use the following equation to calculate rts. If we wanted
to use:
bf
rts =
 1 ht w 
12
1 + 6 

 bf t f 

8.24
rts = = 2.14"
 1 (18.38" )(0.4" ) 
12
1 + 6 

 (8.24" )(0.614" ) 
The above value of 2.14 inches is almost the same as 2.17 in. given in the AISC Tables.
J = 1.83 in3 from AISC cross section properties Tables.
Note that if this was a built-up section, we would use the equation below to calculate J since there are no
tables for properties of custom-made built-up sections.
J = Σ bi ( ti )3 / 3 Which, would result in:
J = 2x 8.24 in x (0.614 in)3 / 3+18.38 in. x (0.4 in. )3 / 3 =1.66 in4
The above J value is less than the J value in the AISC Manual Tables for this section. The reason for this
discrepancy is that the above equation of J does not include the additional amount of torsion constant due
to effect of the joint where web is connected to flanges. The values in the AISC Tables include this
additional effect.
c =1.0, since this is doubly symmetric section.
h0 = center-to-center distance of flanges = d- tf = 21 in. - 0.614 in.= 20.4 in.

2 2
E Jc  Jc   0.7Fy 
Lr = 1.95rts +   + 6.76 
0.7Fy S x ho  S x ho   E 

2
4  (1.83 in 4 )(1.0)  2
L r = 1.95(2.17" )
29000 ksi (1.83 in )(1.0)
+   + 6.76 0.7(50 ksi) 
0.7(50 ksi) 3  (127 in 3 )(20.4" )   29000 ksi 
(127 in )(20.4" )    

Lr = 218 in.
Since Lb = 240 in. is greater than Lr = 218 in. , then, elastic lateral torsional buckling is expected and the
nominal bending capacity is obtained from:

Mu=2.14P
Segment 2
N
L M
K
MA= 2.14P

MB=6.43P
MC=10.72Pu Mu=15Pu k-in

Mmax =15Pu

2
π 2E Jc  Lb 
Mn = Fcr Sx = CbSx 1 + 0.078   ≤ Mp
2 Sx ho  rts 
 Lb   
 
 rts 
 
Let us calculate Cb for this Segment 2:

12.5Mmax
Cb = ≤ 3.0
2.5Mmax + 3MA + 4MB + 3MC

12.5(15Pu )
Cb = = 1.84
2.5(15Pu ) + 3(2.14Pu ) + 4(6.43Pu ) + 3(10.72Pu )

Since Cb of 1.84 is less than 3.0, we will use Cb = 1.84

2
π E
2
Jc  Lb 
M n = Fcr S x = C b S x 1 + 0.078   ≤M p
2 S x ho
 Lb   rts 
 
 rts 

2
π (29000 ksi )
2
(1.83 in 4 )(1.0)  240" 
M n = (1.84)(127 in )
3
1 + 0.078   = 7074 k − in
2
 240"  (127 in3 )(20.4" )  2.17" 
 
 2.17" 
Let us calculate Mp:
Mp =Fy Zx =50 ksi x144 in = 7,200 kip-in
Since Mn = 7,074 k-in is less than Mp =7,200 k-in , then we will use Mn = 7,074 k-in. as the nominal
bending capacity of this segment due to lateral torsional buckling.
Using equation of design we can find Pu :
Mu ≤ φb Mn
(15Pu k-ft)(12 in/ft) <= 0.9(7074 k-in)

Pu ≤ 35.4 kips (due to Lat. Tor. Buckling of segment 2

Check Segment 3 (from M to N ):

Lb = un-braced length of the segment =180 in= 15 ft

E
Lp = 1.76ry
Fy

Lp = 6.3 ft ( see calculation of Lp done for Segment 2 above. )

Lr = 218 in. = 18.2 ft ( see calculation of Lr done for Segment 2 earlier. )

Since Lb = 15 ft is greater than Lp =6.3 ft and less than Lr = 18.2 ft we have the condition of
Lp > Lb > Lr , then the nominal bending moment capacity of this segment is:
  Lb − L p 
M n = C b M p − (M p − 0.7Fy S x )  ≤ M p
  Lr − L p 

Let us calculate Cb for this segment:

CHAPTER 6 – FLEXURAL MEMBERS 6-29

Mu=2.14P
Segment 3
N
L M
K
MC= 3.75P

MB=7.5P
MA= 11.25P
MMax =15P

12.5M max
Cb = ≤ 3.0
2.5M max + 3M A + 4 M B + 3M C

12.5(15 Pu )
Cb = = 1.67
2.5(15 Pu ) + 3(11.25 Pu ) + 4(7.5 Pu ) + 3(3.75 Pu )

Notice that whenever bending moment diagram on the segment is a triangle with maximum at one end and
zero value at the other end, Cb is equal to 1.67 regardless of values of bending moments.

Since Cb of 1.67 is less than 3.0, we will use Cb = 1.67

  Lb − L p 
M n = C b M p − (M p − 0.7Fy S x )  ≤ M p
  Lr − L p 

Mp = 7200 k-in. (see calculations of Mp for Segment 2 above )

  15 − 6.3 
M n = (1.67) 7200 − (7200 − 0.7(50)(127))  = 8665 k − in
  18.2 − 6.3 
Since Mn = 8,665 k-in is greater than Mp =7,200 k-in, we will use Mn = Mp = 7,200 k-in.
Now, by using equation of design we can find Pu :
Mu ≤ φb Mn
(15Pu k-ft) (12 in/ft) =(0.9)(7,200 k-in)

Pu ≤ 36.0 kips (due to Lat. Tor. Buckling of segment 3

The governing value of Pu is 35.4 kips due to lateral torsional buckling of segment 2.

6.9. LIMIT STATES OF DOUBLY SYMMETRIC BOX-

SHAPED BEAMS
When a square or rectangular beam is subjected to bending, it can fail in three different ways.
These three limit states of bending capacity are shown in Figure 6.20 and are:

(a) Yielding and development of plastic hinge

(b) Local buckling of the compression flange
(c) Local buckling of the compression region of the web

Notice that, unlike I-shape members and double channels, we do not consider lateral torsional
buckling as a limit state for beams with box sections. This is due to the fact that the torsional
properties , J and Cw as well as Iy for closed box sections are very large compared to the
corresponding values for open I-shape or double channel members. This results in lateral-torsional
buckling capacity of box sections to be much larger than Mp and not governing. In very rare cases,
if the box is very narrow and very tall, it might be necessary to check lateral torsional buckling of
such box shapes by using the approximate value of Lp given by Equation 6.27 below. If the un-
braced length Lb is less than Lp , no lateral-torsional buckling is expected.

E
L p = 1.76ry
Fy (6.27)

The AISC Specification (AISC, 2010a) does not recognize lateral-torsional buckling as a
limit state for beams with box sections.

Yielding of Cross
Section, Plastic
Hinge Formation

(a) Flange
Local
Buckling

(b)

Web
Local
Buckling

Figure 6.20. Limit States of Failure of Beams with Box Sections in Bending

6.10. LIMIT STATE OF YIELDING AND DEVELOPMENT

OF PLASTIC HINGE IN BOX SECTIONS
6.10.a. Yielding and Development of Plastic Hinge in a Box Beam
Similar to I-shapes as discussed earlier, for box sections this limit state is reached when most areas
of the cross section have yielded either in tension or compression and a plastic hinge has formed
in the cross section, Figure 6.21(a). The actual stress distribution on the cross section where a
plastic hinge has formed is close to what is shown in Figure 6.21 (b), but, as mentioned earlier, the
much simplified stress distribution shown in Figure 6.21(c) is used in design to calculate plastic
moment capacity Mp of a beam section. This is the same we did for I-shape sections earlier.
-Fy
-Fy

Mp C

P. N.A.
CHAPTER 6 – FLEXURAL MEMBERS 6-31

6.10.b. Design for Limit State of Yielding and Development of Plastic

Hinge in Box Beams (AISC, 2010a)
The equation of design for checking yielding of cross section and development of plastic hinge in
beams with box sections is the same as for I-shape cross sections bending about their major axis
discussed earlier. The equations are:

M u ≤ φb M n (in LRFD) and Ma ≤ Mn /Ωb (in ASD) (6.28)

where
Mu = required bending strength in LRFD, same as the maximum factored applied bending moment.
φb = resistance factor for bending (in LRFD) = 0.90
Ma = required bending strength in ASD, same as the maximum service (un-factored) applied
bending moment.
Ωb = safety factor for bending (in ASD) = 1.67
Mn = nominal bending moment capacity equal to plastic moment capacity:

Mn =Mp (6.29)
where
Mp = plastic moment capacity of box cross section of the beam
Fy = specified minimum yield stress of steel being used
Zx = plastic section modulus of cross section of the beam about x-axis. To calculate plastic
moment capacity of a box section, the same procedures given in Section 6.3 earlier. The
difference is that box section has two webs while I-shape had only one web.

6.11. LIMIT STATE OF LOCAL BUCKLING OF

COMPRESSION FLANGE IN BOX BEAMS WITH
COMPACT WEBS
6.11.a. Local Buckling of Compression Flange of Box Beams
This section provides information and design procedures on local buckling of compression
flange of box beams with compact webs. For box sections, webs are the vertical plates. As
mentioned earlier for I-shape sections, for a web to be compact the h/t ratio of the web should not
exceed the λp values given in Table B4.1b of the AISC Specification (AISC, 2010a). If the webs
of box are not compact, its compression zone can locally buckle and reduce the local buckling
capacity of the compression flange supported by the webs.
Figure 6.22(a) shows a box beam under bending stresses. In this case the top flange is under
compression. When compressive stresses acting on a flange reach a critical value, the flange can
locally buckle as shown in Figure 6.22(b).

(a) Local Buckling of

Compression Flange (b) Close-up of Locally
in a Box Beam Buckled Area

Figure 6.22. Local Buckling of Compression Flange of a Box Beam

6.11.b. Design for Limit State of Local Buckling of Compression

Flange in Box Beams (AISC, 2010a)
The equation of design for checking local buckling of compression flange in box beams with
compact webs is:

Mu ≤ φb Mn (in LRFD) and Ma ≤ Mn /Ωb (in ASD) (6.30)

where
Mu = required bending strength in LRFD, same as the maximum factored applied bending
moment.
φb = resistance factor for local buckling (in LRFD) = 0.90
Ma = required bending strength in ASD, same as the maximum service (un-factored) applied
bending moment.
Ωb = safety factor for local buckling (in ASD) = 1.67
Mn = nominal bending moment capacity given by following equations and on Figure 6.23:
CHAPTER 6 – FLEXURAL MEMBERS 6-33

a. When λ ≤ λpf , cross section is compact, no local buckling of flange is expected, and:
Mn = Mp (6.31a)

b. When λpf < λ ≤ λrf , cross section is Non-compact, inelastic local buckling of flange is
expected and:
 
 b Fy 
M n = M p − ( M p − F y S x ) 3.57 − 4.0  ≤ M p
 t f E  (6.31b)
 

c. When λ > λrf , cross section is Slender, elastic local buckling of flange is expected and:
M n = Fy S eff (6.31c)
where
λ = b/t ratio of compression flange
λpf = same as λp , limit of b/t ratio for compact elements given in Table B4.1b of the AISC
Specification (AISC, 2010a), see Table 6.1 earlier.
λrf = same as λr , limit of b/t ratio for slender elements given in Table B4.1b of the AISC
Specification (AISC, 2010a).
Sx = section modulus of cross section w.r.t. x-axis (axis of bending)
S eff = effective section modulus of cross section w.r.t. x-axis (axis of bending) calculated using an
effective width of flange, beff, given by:

E  
be = 1.92t f 1 − 0.38 E ≤b (6.32)
Fy  b/tf Fy 
 
Other parameters in above equations were defined earlier in this chapter. The
above design equations are shown in Figure 6.23 below.

Mn
1. No Flange
Local Buckling 2. Inelastic Flange Local Buckling
Mn =Mp  
 b Fy 
M n = M p − (M p − Fy S x ) 3.57 − 4.0  ≤ M p
 t f E 
 
Mp
3. Elastic Flange Local Buckling
1. Compact M n = Fy S eff
sections 2.
Non-compact
sections

3. Slender sections

b/t of Compression Flange, λ

λpf λrf
Note: For values of λpf and λrf , which are the same as λp and λr (see Table 6.1 on Page 6-15 ).
Figure 6.23. Three Regions of Behavior for Local Buckling of Compression Flange of a Box
Beam as per AISC Spec. (2010a)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-34

Illustrative Example 6.6:

Evaluating Strength of a Beam Based on Local Buckling of
Flange
Find maximum value of factored load P that can be applied to the beam.
The beam is a built-up welded girder as shown below. The self-weights
of girder and floor are included in the distributed load.

Pu kips y
Pu /12 k/ft PL16x3/4”,
A572, Gr. 50
Top & Bottom
Flanges x x
A B
C
8 ft 8 ft PL 36x3/4,
A572, Gr. 50
Both Webs y

Given:
1. Cross section of the member is given
2. The material is given as A572, Gr. 50 with Fy = 50 ksi and Fu = 65 ksi.
3. The beam has continuous lateral bracing for compression flange since it is connected to floor slab.

Required:
To establish bending capacity of the member in terms of applied load PU.

SOLUTION:
Establish Properties of the Box Section:
bf =16 in., tf = 0.75 in., tw = 0.75 in., h= 36 in. , d = 37.5 in.,
Calculate Zx
Zx =ΣAi di =2x16”×0.75” × 18.375” + 4x18”×0.75”×9” = 927 in3
Notice that we could calculate Zx by subtracting Zx for inner rectangle from Zx for the outer rectangle as
done below. Zx for a solid rectangle is equal to bh2/4.
Zx = 16” x (37.5”)2 / 4 – 14.5” x (36”)2 / 4 = 927 in3 which is the same as we calculated above.
Mp = plastic moment= FyZx = (50 ksi)(927 in3) = 46,350 k-in.

1. Check Local Buckling of Web:

h / tw =36”/0.75”= 48
λp = 2.42(E/Fy)0.5 = 2.42(29,000 ksi / 50 ksi)0.5 = 58 (See Table 6.1 on Page 6-15)
Since h / tw = 48 is less than 58, the web is compact.

2. Check Local Buckling of Compression Flange:

λ = bf / tf =16/(0.75) = 24
λpf = 1.12(E/Fy)0.5 = 1.12(29,000 ksi / 50 ksi)0.5 = 27.0
λrf = 1.40(E/Fy)0.5 = 1.40(29,000 ksi/ 50 ksi)0.5 = 33.7
Since λ = 24 is less than λ pf = 27.0; then compression flange is compact.
Section is compact, no local buckling failure is expected.
Mu = Pu L / 4 + (Pu /12) (L2/8) = Pu (16 ft) / 4 + (Pu /12) (162 / 8) = 6.67Pu k-ft
CHAPTER 6 – FLEXURAL MEMBERS 6-35

Pu kips
Pu /12 k/ft

8 ft 8 ft

M- diagram

Mmax = 6.67Pu k-ft

Using equation of design we can find Pu :

Mmax ≤ φb Mn
(6.67Pu k-ft)(12 in/ft) =0.9(46,350 k-in) ⇒ Pu = 521 k
Pu = 521 kips (due to formation of plastic hinge at midspan

3. Check Limit State of Lateral-Torsional Buckling:

Since this is a box section, limit state of lateral-torsional buckling does not apply. Its compression flange is
continuously braced which will also prevent lateral-torsional buckling.

The governing value of Pu is 521 kips due to formation of plastic hinge at midspan.

6.12. NUMERICAL EXAMPLE FOR DESIGN OF A BEAM

The numerical example given in this section is on design of a beam, where the applied loads are
given and the problem statement asks for selection of the material and cross section that can resist
the applied loads.
Illustrative Example 6.7:
Design of a Beam with Plastic Hinge Formation as the Limit State
Design the wide flange beam shown in the figure. Loads are nominal service loads. Use LRFD methods
and consider only limit state of yielding and plastic hinge formation. The compression flange is
continuously braced in lateral direction preventing lateral –torsional buckling.

DL= 30 k
LL= 10 k y

x x

2@9’-3” =18’-6”
y

Given:

1. Applied service (un-factored) loads are given as: DL= 30 kips and LL= 10 kips, both as a concentrated
load at midspan.
2. Span of the beam is given as 18’-6”.
Required:
To design the wide flange beam using the LRFD methods.

SOLUTION
1. Establish the Factored Load to be used in LRFD:
The load combinations involving DL and LL are 1.2D+1.6L and 1.4D. Therefore, Pu, the factored
concentrated force acting at the midspan, will be the larger of these two combinations:
1.2×30k + 1.6×10k = 52 k and 1.4×30k = 42k ⇒ Use Pu = 52 k

2. Select Material of the Member:

With reference to the AISC Manual Table for Material (summarized on Page 2-11 od Chapter 2), the
”preferred” material for wide flanges is A992. We will use A992 (with Fy =50 ksi and Fu = 65 ksi). In our
design we will ensure that the selected cross section is compact ( i.e. λ = b/t ≤ λp), therefore no local
buckling will occur prior to reaching plastic moment capacity, Mp.

3. Design for Limit State of Yielding of Cross Section and Formation of Plastic Hinge:
The equation of design for limit state of yielding and formation of plastic hinge is Mu ≤ φb Mn where
Mn= Fy Zx . Therefore, for design:
Mu ≤ φb Fy Zx
Mu = maximum factored bending moment = Pu L/4 = 52 k x 18.5 ft x 12 in/ft /4 = 2,886 k-in.
2,886 ≤ (0.90) (50 ksi)(Zx ) ⇒ Zx ≥ 64.1 in3.
Now, we can select a wide flange such that its plastic section modulus , Zx , is greater than the above
required value of 64.1 in3. Notice that the AISC cross section tables has a column for Zx.
Let us try W16x40, which has Zx =64.1 in3.

4. Check Local Buckling:

bf / 2tf = 6.93 < λp = 0.38(E/Fy)0.5 = 0.38(29,000 ksi / 50 ksi) 0.5=9.1
h/tw = 46.5 < λp = 3.76(E/Fy)0.5 = 3.76(29,000 ksi / 50 ksi)0.5=90
Section is compact, no local buckling failure is expected prior to formation of plastic hinge.
Use W16×40, A992, Gr. 50 Steel

References for Chapter 6

1. AISC, (2010a). “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An American
National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-10, An
American National Standard, Am. Institute of Steel Construction, Chicago.
3. AISC, (2011). “Steel Construction Manual,” 14th Edition, American Institute of Steel
Construction, Chicago.
CHAPTER 6 – FLEXURAL MEMBERS 6-37

PROBLEMS ON FLEXURAL MEMBERS

Problem 6.1. For following beam cross sections calculate yield moment My and plastic moment Mp with
respect to horizontal axis. Steel used in each cross sections is indicated on the sketches. Bending is taking
place w.r.t. to horizontal axis. PL20x 3/4, PL10x1, A36
PL10x1/2, A572, Gr. Top & Bott. Flanges
A36 Top &
W18x50 Top & Bottom Bottom Weld both
A992, Flange Flange sides
Grade 50
PL 15x1/2
PL16x1/4, A36 PL 26x3/8, A36
A572 Gr. 50
(A) (B)
(C
(D)
C15x50 )
A36 y 2PL16x2 W24x84,
A36 PL12x3/4, A36
A992,
Weld both Gr. 50 PL16-1/2x 5/8,
sides
2PL 20x1 A36
W18x50 A36 Plate 23x1/2
A992, A572, Gr. 50 PL 16x3/8, A36
Grade 50 y
(H)
(E) (F) (G)

Problem 6.2. For beam cross sections shown in Problem 1 above, check local buckling as well as lateral-
torsional buckling. All cross sections have an unbraced length of 8 feet for lateral-torsional buckling. Steel
used in each section is shown on the figure. Use a value of Cb equal to 1.0. Bending is taking place w.r.t. to
horizontal axis .
Problem 6.3. Calculate maximum factored load Pu that can be applied to the beam shown in the figure.
Beam has lateral support for its flange and lateral-torsional buckling is prevented.
Pu W36x210, A992, Pu
Gr.50 steel

9’-0” 15’-0” 15’-0”

(A)
W30x235, A992,
Pu Gr.50 steel Pu
Pu Pu
P/6 kips/ft

7’-0” 30’-0” 7’-0”

(B) W30x211,
2Pu A992, Gr.50
2Pu
Pu /4 kips/ft

10’-0” 32’-0” 10’-0”

(C)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
6-38

Problem 6.4. Calculate maximum load Pu that can be applied to the beam shown in the figure. Lateral
bracings are provided at the supports and at other locations as shown for each case below. Steel is A992..

Pu / 8 kips/ft
Pu kips

(A)

2@15 ft=30 ft 12 ft W30x235, A992, Gr 50

= Lateral Bracing on Compression Flange

Pu / 10 k/ft Pu Pu /10 k/ft

(B)

10 ft 2@10 ft=30 ft 10 ft

= Lateral Bracing W30x235 A992, Gr. 50 steel

2Pu Pu /4 k/ft
Pu Pu

(C)

10 ft 2@12 ft=24 ft 10 ft

W30x235 A992, Gr. 50 steel

= Lateral Bracing

Pu 2Pu Pu W36x210, A992, Gr. 50

(D)

12’-6” 3@10’-0”=30’-0”

Pu /2 kips/ft

Pu
Pu W30x235,
A992, Gr.50

(E)
4’-0” 15’-0” 4’-0”
CHAPTER 6 – FLEXURAL MEMBERS 6-39

Problem 6.5. Design the following beams. Service (un-factored ) loads are shown on each beam. Steel for
all wide flange rolled shapes is A992 Grade 50. Lateral bracings are provided at locations shown.
LL=10 kips LL=2.5 kips
DL=10 kips DL=2.5 kips

(A)

10 ft 10ft 10 ft

= Lateral Bracing

No lateral DL=2 k/ft No lateral

bracing here LL=1.5 k/ft bracing here

(B)

10 ft 20ft 20ft 10 ft

= Lateral Bracing

LL=5 kips LL= 8 kips LL=10 kips LL=10 kips LL=5 kips
DL=5 kips DL= 8 kips DL=10 kips DL=10 kips DL=5 kips

= Lateral Bracing

LL=5 kips LL= 180 k-ft

DL=6 kips LL=5 kips
DL = 400 k-ft DL=6 kips

(D)
5 ft 16ft No lateral
6 ft
bracing at this
end
= Lateral Bracing

2 kips/ft (1 k/ft DL, 2 k/ft LL) 10 kips (10k DL, 10k LL)

.
(E)

2@15ft = 30 ft 2@15ft = 30ft

Problem 6.6. Design the following beams. Service (un-factored ) loads are shown on each beam. Steel for
all wide flange rolled shapes is A992 Grade 50. Lateral bracings are provided at locations shown.
Each concentrated load = 8kips (3k DL, 5k LL)

(A)

7@10'-0"=70'-0"

LL=5 kips LL=5 kips LL=5 kips

DL=6 kips DL=2 k/ft DL=6 kips
LL=1.5 k/ft DL=6 kips

(B)

10 ft 20ft 20ft 10 ft

= Lateral Bracing

Problem 6.7. Find maximum value of load Pu that can be applied to the beam. The steel used is A572 gr.
50. The locations of lateral braces are shown in the figure by ⊗ sign. Notice that Points C and D do not
have lateral bracing. When checking for bending, check local buckling and lateral-torsional buckling modes.

20”
_Pu k/ft _Pu k/ft y
20 2Pu kips 2Pu kips 20 2”

½”
36”
x x

A B C D E F 2”
20 ft 15 ft 15 ft 15 ft 20 ft
20”
y
Beam Loading Beam Cross Section

Problem 6.8. Calculate maximum factored load Qu that can be applied to the beam shown below. The top
flange of the beam is continuously connected to the floor slab and since cross section is box-shape, lateral-
torsional buckling need not be checked. Check all other possible failure modes of bending. For all plates,
Fy = 36 ksi and Fu = 58 ksi.
PL 32x2 y
Top & Bottom
3.5Qu kips 3.5Qu kips
2.5Qu kips
Weld 48”
x x

. 2PL96x3/4
5 ft 14 ft 14 ft 5 ft 48”

Elevation of the Beam y

Beam Cross Section
CHAPTER 6 – FLEXURAL MEMBERS 6-41

Problem 6.9. Calculate maximum load Qu that can be applied to the beam shown below. Consider all
possible failure modes (i.e. lateral torsional buckling and local buckling of the web and flanges). For all
plates, Fy= 36 ksi and Fu=58 ksi.

y
3Qu kips PL10x3/4
Qu kips Qu kips

Weld both sides

Top & bottom
x
x ho = 48.75”
PL48x1/2
10 ft 15 ft 15 ft 10 ft
PL 10x3/4
= Lateral Bracing on Compression Flange

y
Elevation of the Beam
Beam Cross Section

Problem 6.10. Calculate maximum factored load Pu that can be applied to the beam. Steel is all A36 and
girder is braced by the floor slab and lateral –torsional buckling is not a failure mode.
Top and
Pu Pu Bottom Flanges:
PL 18x1.5

Web:
PL36x3/8
10ft 12ft 10 ft

Problem 6.11. Design a wide flange beam to carry uniformly distributed factored load of 3 kips/ft applied
downward to the beam below. Use A992 Gr. 50 steel. Floor slab and transverse beams provide sufficient
bracing for the beam preventing lateral-torsional buckling failure mode.

3 k/ft

3@10’-0”=30’-0” 3@10’-0”=30’-0”

Problem 6.12. Determine maximum value of factored load Pu that can be applied to the beam shown
below. The top flange is continuously braced against lateral torsional buckling. The bottom flange is braced
only at A, B, C and D. The material is all A36 with Fy=36 ksi and Fu=58 ksi.
y
Pu Pu
24”

3/4”

3/8”
A B C D
x x
15 ft 15 ft 15 ft 60”

= Lateral Bracing on Compression Flange

3/4”
y
24”

Problem 6.13- Calculate maximum load Pu that can be applied to the beam shown below. For all plates,
Fy=36 ksi and Fu=58 ksi.

Pu kips Pu/10 k/ft

y
PL15x1/2
A36

Weld
(A) both sides
top & bottom
x x

PL48x3/8 h=48.5”
A36
18 ft 18 ft 10 ft PL 15x1/2
A36
y
= Lateral Bracing on Compression Flange

Beam Cross Section

Elevation of the Beam

Top & Bottom

Flanges
Pu kips PL 20x1.5
0.5Pu kips Pu kips Pu/6 (kips/ft) 0.5Pu kips

Web:
(B)
PL36x3/8

10ft 36ft 10 ft
CHAPTER 7-DESIGN FOR SHEAR 7-1

CHAPTER 7

DESIGN FOR SHEAR

V
V

Steel girders and beams in a steel Structure

Objectives:
In this chapter, you will learn:
1. Behavior of beams when subjected to shear forces.
2. Failure mode (or limit state) of shear failure.
3. How to design a beam for shear
4. How to apply what you have learned to actual design and
evaluation of beams by using numerical examples
5. provided throughout the chapter.

7.1. INTRODUCTION
In this Part 1, we will discuss shear design of steel beams. Figure 7.1 shows cross sections
that we will consider, which are I-sections, double channel and box built-up sections. In
these cases, the elements of the cross section parallel to the applied shear resist the bulk of
the applied shear force. For I-shaped cross sections with single web, Figure 7.1, the web
resists the bulk of the applied shear. For HSS and box sections, the two vertical walls of
the box, that are parallel to the shear force, resist the bulk of the applied shear. In all of
these cases, flanges resist a small amount of shear, but we ignore this resistance in design.
____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
7-2
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

x x x x x
x x
x x x x x

Figure 7.1. Vertical Elements of Cross Sections Resisting the Applied Shear

7.2. BEHAVIOR OF STEEL SECTIONS SUBJECTED

TO SHEAR
Shear strength of beams primarily depends on the behavior of their web. Consider a girder
with vertical stiffeners subjected to pure shear, Figure 7.2(a). As we will discuss later in
this chapter, adding vertical stiffeners to the web can increase shear capacity of the web.
The shear force applied to a web, Figure 7.2(a), creates compressive stresses in a diagonal
area of the web, called compression field and tensile stresses in the opposite diagonal area
of the web called tension field, Figure 7.2(b). These diagonal tension and compression
fields are similar to an X-brace configuration as shown in the analogous truss of Figure
7.2(c).

Compression Compression
Web Field Tension
Stiffeners Tension Field Diagonal Diagonal
(a) (b)
(c)

Buckling of
Compression Buckling of
Yielding of Field
Tension Field Compression
Yielding of Diagonal
(d) Tension
Diagonal (e)

Figure 7.2 (a) Compressions and Tension Fields, (b) Truss Analogy, (c) Buckling of
Compression Field and Tension in The Tension Field, and (d) Truss Analogy
CHAPTER 7-DESIGN FOR SHEAR 7-3

As the applied shear force increases, since the web is usually relatively thin and slender,
the diagonal compression area can buckle, Figure 7.2(c). This is analogous to buckling of
the compression diagonal member in the analogous truss shown in Figure 7.2(d) and (e).
After buckling of the diagonal compression field, the diagonal tension field of the web
continues to resist additional shear similar to a tension diagonal of a truss. Then as the
tension, diagonal field yields the shear stiffness of the web panel decreases significantly
and at this point, the panel has “failed.” Continuation of the loading will eventually result
in the strain hardening of the tension field and its fracture failure.

Shear V
Force, V
C V
Vmax γ

B Kst Start of Fracture in the

Vy-tf 1 Diagonal Tension Field

Yielding of Diagonal V V
Tension Field γ
A Ktf
1
Vcr
Buckling of Diagonal
Compression Field V V

Elastic Behavior γ

Ke V
V
1
γ
O Shear Distortion Angle , γ

Figure 7.3. Typical Shear Behavior of Web in a Girder Web

Figure 7.3 shows behavior of the web of a typical steel I-beam subjected to shear.
Line OA represents initial elastic behavior. At Point A, the diagonal compression field
buckles. The shear force corresponding to buckling of the compression field is called V cr .
Notice that buckling of the compression field causes reduction of the shear stiffness from
K e, the elastic stiffness, to K tf , the tension field stiffness, Figure 7.3. After buckling of the
diagonal compression field, the web continues to resist additional applied shear by tension
resistance of the diagonal tension field. Line AB represents this region of the behavior,
where tension field is resisting applied shear in excess of the V cr . At Point B, diagonal
tension field yields and shear stiffness of the web is further reduced to K st . Line BC
represents strain hardening of the tension field. At point C, tension field starts fracturing
resulting in significant loss of shear strength and stiffness of the web and failure of the
girder.

____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others
or posted anywhere.
7-4
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Currently, in most applications, especially in buildings, shear capacity of a beam or

a girder is limited to the shear force that causes buckling of the compression field shown
in Figure 7.3 as V cr . In these cases, the additional shear capacity due to tension field
action is not used in design. However, in most bridges, where plate girders with relatively
slender webs, shear capacity of a plate girder is the capacity corresponding to yielding of
the tension field area of the web, shown as V y-tf in Figure 7.3.
The current AISC Specification (2010a) provides provisions for shear design of
beams in two categories:
1. design for shear where tension field action is not utilized and the capacity of the
beam or girder is limited to V cr in Figure 7.3, and;
2. design for shear where tension field action is utilized and the capacity of the beam
or girder is the yield strength of the tension field V y-tf in Figure 7.3.

In this course, we will discuss the first case, where tension field action is not
utilized.

7.3. DESIGN FOR SHEAR WITHOUT UTILIZING

POST-BUCKLING STRENGTH (TENSION FIELD
ACTION)
7.3.a. Introduction
If tension field action is not used in design, shear capacity of a web is limited to its shear
buckling capacity. Figure 7.4 shows a rectangular plate, similar to the web of a girder,
subjected to pure shear. The state of stress acting on the boundaries of the typical finite
elements of the web is also pure shear as shown in Figure 7.4. The pure shear stresses on
the boundary of the elements create normal compressive and tensile stresses under
45o angle diagonal directions. The magnitudes of these diagonal normal stresses are equal
and their magnitude is equal to the applied shear stressτ. As the applied shear stress τ
increases, the diagonal compressive and tensile stresses also increase.

Figure 7.4. A Web Subjected to Pure Shear and Internal Stresses

CHAPTER 7-DESIGN FOR SHEAR 7-5

V V
V
V

Vn V

Vy
Expected
Realistic
Behavior

AISC
1. Shear 2. Inelastic Specification
Yielding of Buckling of Curves
“Compact” Web “Non-compact” 3. Elastic
Without Web buckling of
Buckling
“Slender” Web

1.10 kv E / Fyw 1.37 kv E / Fyw h/tw

Figure 7.5. Shear Capacity of Web, Vn , versus Slenderness of the Web, h/tw

Depending on the width-to-thickness ratio of the web plate subjected to pure

shear, one of the three failure modes can occur. If the web slenderness, h/t w , is relatively
small, the web is compact and will yield under pure shear prior to any shear buckling. If
the h/t w of the web is intermediate, the web is non-compact, some parts of the web,
especially areas with high residual stresses and other imperfections, such as initial out-of-

Compact Web

V V V
Compact Web
B Non-Compact Web
A Non-compact Web V
V
C D E
F
Slender Web
Slender Web
G H
I
V
V
γ
O
γ

Figure 7.6. Shear Behavior of “Compact”, “Non-Compact” and “Slender” Webs

straightness, will yield, and then prior to full yielding, the web will buckle while parts of it
had yielded and other parts are still elastic. This type of buckling is called inelastic web
buckling. If h/t w of the web is relatively large, say larger than about 90 for A36 steel, the
web is slender and prior to any significant yielding, and while the web is almost elastic, it
will buckle. This is called elastic buckling of the web
As mentioned earlier, depending on the behavior of beam web under shear, the webs
are divided into compact, non-compact, and slender. These categories are discussed
below and shown on Figure 7.6.

1. Compact Webs -Beam webs with their h/t w less than or equal to 1.10 kv E / Fyw are
considered by the AISC Specification (2010a) to be “compact.” Compact webs ,
when subjected to shear force, behave elastically until they reach yielding in shear
prior to buckling of diagonal compression field. The behavior of a compact web is
shown by the Curve OAB in Figure 7.6.
The nominal shear capacity of the web in this case, according to the AISC
Specification (2010a), is equal to the area of the web times shear yield strength.
Shear yield strength of steel is taken as 0.6F yw. It should be mentioned that
throughout the AISC Specification (2010a) shear yield stress is specified to be equal
to 0.6F y , which approximates shear yield stress τ y = Fy / 3 = 0.577 Fy as established
by the Von Mises’ yield criterion for elements subjected to pure shear.
The value of h / t w ≤ 1.10 k v E / F yw , as the limit of compact sections has been
established by calculating critical buckling stress of plates subjected to pure shear and
making it equal to the shear yield stress.

The critical elastic buckling stress of a plate subjected to pure shear is given by:

kv π2 E
τ cr = (7.1)
12( 1 − ν 2 )( h / t w ) 2

τcr
where
h = depth of web that is free to buckle as defined in Figure 7.7 for typical sections.
t w = thickness of the web
E = modulus of elasticity of steel = 29,000 ksi
ν = elastic Poisson’s ratio for steel = 0.30
k v = shear buckling factor, a number between 5 and 10, representing the effects of
boundary conditions and aspect ratio a/h of the web on the buckling stress. In
the current AISC Specification( 2010a) values of k v , for typical girders, where
web can be considered simply supported on all four sides are given by
Equation 7.2 below a = spacing of stiffener for stiffened beams. Value of “a”
is taken as infinity (∞) for beams without web stiffeners making value of k v
equal to five for these beams.
CHAPTER 7-DESIGN FOR SHEAR 7-7

d Welded h d h d
h

Bolted h=d
h d
or
Riveted

Figure 7.7. Definition of “a” , “h”, and “d”

5.0
kv = 5.0 + (7.2)
( a / h )2

The above equation is a simplified version of a set of two Equations 7.3a and 7.3b
given below in Galambos (1998). The equations are derived assuming elastic
buckling of plates subjected to pure shear on their boundaries.
5.34 (7.3a)
k v = 4 .0 + ( for a / h ≤ 1.0)
( a / h) 2
h
4.0 (7.3.b)
k v = 5.34 + 2
( for a / h ≥ 1.0)
( a / h)
a

The shear yielding stress according to Von-Mises yield criterion for an element
subjected to pure shear is:

Fy
τy = ≅ 0.6 Fy (7.4)
3
where
τ y = shear yield stress of steel
F y = yield stress of steel subjected to normal tensile stresses
By equating Equations 7.1 and 7.4, we can obtain the limit of h/t w for
“compact” webs. For h/t w smaller than this limiting value, the web will yield
prior to buckling.

k vπ 2 E
= Fy / 3 (7.5)
12(1 − µ 2 )(h / t w ) 2

Simplifying above equation results in an h/t w limit of approximately

1.10 k v E / F yw which is the limit specified in the AISC Specification (2010a) for
compact webs.
3. Non-Compact Webs- When h/t w of a beam web is more than 1.10 k v E / F yw but
less than or equal to 1.37 k v E / F yw , the web is expected to start yielding under
the applied shear, but, before the entire web yields in shear, the partially yielded
web buckles in a diagonal manner as shown in Figure 7.6. Typical behavior of a
non-compact web is shown by the Curve OCDEF in Figure 7.6. From Point O to
C in Figure 7.6, the behavior is elastic. At Point C, some areas of the web start
yielding. As yielding increases, at Point D on the curve, the inelastic web buckles
in a diagonal direction. However, after buckling of the compression diagonal and
from Point D to E in Figure 7.6, the diagonal tension field of the web continues to
resist more shear; although the shear stiffness is somewhat reduced due to
buckling of the diagonal compression field. As the shear applied to the web
increases, the diagonal tension field starts yielding at Point E in Figure 7.6. As
shear force increases, yielding of the diagonal tension field continues, Point E to
F, and some diagonal fibers that have yielded earlier strain-harden. Finally, after
some fibers have reached their ultimate fracture strain, the tension diagonal
reaches its maximum capacity at Point E and fracture of the tension field of the
web starts.

The h/t w limit of 1.37 k v E / F yw is established by assuming that the web plate
will start showing inelastic behavior when shear stress is about 0.8τ y instead of
yielding at τ y . The early yielding of the web can be due to presence of residual
stresses, out-of-straightness and other material or geometric imperfections. If we
equate Equation 7.1 to 0.8τ y and simplify it, we obtain a limit of h/t w equal to
1.37 k v E / F yw , which is the limit in the AISC Specification(2010a) for non-
compact webs.

4. Slender Webs - Webs with h/t w greater than 1.37 k v E / F yw are considered
slender and their diagonal compression field will buckle under relatively small
shear while the web is still almost elastic. Curve OGHI in Figure 7.6 on Page 7-5
earlier shows shear behavior of a beam with slender web. From O to G in Figure
7.6 behavior is elastic. At Point G, compression diagonal field in the web buckles
while the web is essentially elastic. After buckling of the compression diagonal
and from Point G to H in Figure 7.6 on Page 7-5, the tension diagonal field
continues to resist more shear forces. As mentioned earlier, when buckling of the
compression field occurs (Point H) , the shear stiffness is reduced. As the applied
shear increases, diagonal tension field yields, strain-hardens and finally fractures
(Point I in Figure 7.6 on Page 7-5).
CHAPTER 7-DESIGN FOR SHEAR 7-9

7.3.b. Design of Members Subjected to Shear without using

Post-Buckling (Tension Field Action) Strength
(Ref.: AISC Specification, (AISC, 2010a), Chapter G, Page 16.1-67))

The equation for establishing shear strength of double symmetric or single symmetric
beams is:
V u ≤ φ v V n (in LRFD) and V a ≤ V n /Ω c (in ASD) (7.6)

where
V u = required shear strength in LRFD, same as the maximum factored applied shear
force.
φ v = resistance factor for shear (in LRFD) given as two values:
1. φ v = 1.0 for rolled I-shapes with h / t w ≤ 2.24 E / F y
2. φ v = 0.90 for all other conditions

V a = required allowable shear strength in ASD, same as the maximum unfactored applied
shear.
Ω b = safety factor for shear (in ASD) given by one of the two values:
1. Ω b = 1.50 for rolled I-shapes with h / t w ≤ 2.24 E / F y
2. Ω b = 1.67 for all other conditions

V n = nominal shear strength given by:

V n = 0.6F y A w C v
(7.7)
where
V n = nominal shear strength of the beam
F y = specified minimum yield stress of steel being used
A w = area of the web equal to total depth of the web times the thickness of the web. For
total depth see
Figure 7.1 on Page 7-2 earlier.
C v = a factor to reduce shear capacity of the web depending on its slenderness, established
as:
a. for webs of rolled-shaped members with h / t w ≤ 2.24 E / Fy

C v = 1.0 (7.8)
b. for webs of all other doubly symmetric and singly symmetric shapes and
channels, except round HSS, the web shear coefficient C v , is established based
on “3-segment” curve in Figure 7.8 on the next age:

(a) For h / t w ≤ 1.10 kv E / Fy

C v = 1.0 (7.9a)

(b) For 1.10 k v E / F y < h / t w ≤ 1.37 k v E / F y

1.10 k v E / Fy (7.9b)
Cv =
h / tw

(c ) For h / t w > 1.37 k v E / F y

1.51Ek v
Cv = (7.9c)
( h / t w )2 Fy

The value of C v given by the above equation is the same as τ cr /τ y where τ cr is the
critical elastic buckling shear stress for a plate under pure shear given by Equation 7.1 and
τ y is the shear yield stress under pure shear given by Equation 7.4 earlier.
The above Equations 7.9(a), 7.9(b), and 7.9(c) give values of C v . Figure 7.8
shows values of V n = 0.6F y A w C v.
In the above equations:
A w = area of the web in shear = d t w
F y = specified minimum yield stress of the steel
t w =thickness of the web
d = total depth of the beam section
h = depth of the web as shown in Figure 7.7 on Page 7-6.
kv = web plate buckling coefficient established as:

(i) For unstiffened webs with h/t w < 260:

k v = 5.0 (7.10a)

If h/t w is ≥ 260, then the web needs to be stiffened and k v established using
Equations (7.10c) and (7.10d below for stiffened webs.
For the stem of tee sections:
k v = 1.2 (7.10b)

(ii) For stiffened webs:

5.0
k v = 5.0 + (7.10c)
( a / h) 2
CHAPTER 7-DESIGN FOR SHEAR 7-11

2
 260 
kv = 5.0 for a / h > 3.0 or a / h >   (7.10d)
 ( h / tw ) 

Web
Stiffeners

where
a = clear distance between transverse stiffeners, Figure 7.7. On the structural drawings,
usually the center-to-center distance is specified and can be conservatively used as
“a”
h = height of the web as shown in Figure 7.7 on Page 7-6.

V
V
V

Eq. 7.7
with Cv=1.0
Vn Eq. 7.7 (Without Tension Field Action)
using Cv from Eq. 7.9b V

Vy V

Eq. 7.7 (Without Tension Field Action)

using Cv from Eq. 7.9c

h / tw
1.10 kv E / Fyw 1.37 kv E / Fyw

Figure 7.8. Shear Capacity of Web, Vn, versus Web Slenderness, h/tw
(Tension Field Action Not Used)

7.3.c. Design of Transverse Stiffeners When Tension Field Action is

not Used
(Ref.: AISC Specification, (AISC, 2010a), Chapter G, Page 16.1-69))
When tension-field action is not used in design, the role of transverse stiffeners in a girder
is to reduce the value of “a” the spacing between stiffeners. The dimension “a” appears in
the denominator of Equation 7.10c on previous page for the k v factor. A smaller value of
____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others
or posted anywhere.
7-12
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

“a” will result in a higher value for k v which in turn results in a higher value of the shear
buckling strength for the web. Although adding transverse stiffener increases shear
strength of a girder, in many cases, the cost of cutting, welding and inspection of the
welded stiffeners, may not justify the gain in shear strength. In most cases, if extra shear
strength is needed, instead of adding the stiffeners to the web, to increase shear capacity, it
might be more economical to simply use another section with a thicker web or, if the
section is a built-up section, use a thicker plate for the web wherever shear is high.
According to the AISC Specification (2010a), transverse stiffeners are not needed if
h / t w ≤ 2.46 k v E / F y or if the applied shear force is less than the available shear strength
calculated using the above equations with a k v of 5.0.
If stiffeners are used and the “a” distance between the stiffeners is used to
establish a k v greater than 5.0 to obtain higher shear strength for the web, then moment of
inertia of each stiffener about the w-w axis, shown in Figure 7.9, should be more than or
equal to I ww given by Equation 7.11 below.
I ww = at w 3 j (7.11)
where
a = clear distance between transverse stiffeners, Figure 7.7. On the structural drawings,
usually the center-to-center distance of stiffeners is specified and can be
conservatively used as “a” (AISC 16.1-69 Iww=bt w 3j, where b is the smaller of a
and h)
t w = thickness of the web
and
j= (2.5)/ (a/h) 2 -2 ≥ 0.5 (7.12)
The above stiffness requirement is imposed to ensure that the stiffeners are stiff
enough in bending out-of-plane of the web and are able to brace the web and force it to
buckle between the stiffeners. Otherwise, the buckling of the web can take place over a
longer length resulting in reduced shear buckling strength.
The AISC Specification (2010a) has certain limitations in design and detailing of
stiffness. These limitations are listed below and shown in Figure 7.10 on the next page.
1. Intermediate stiffeners are permitted to stop short of tension flanges if bearing is
not needed to transfer compression, Figure 7.10(a) on the next page. Leaving
stiffeners slightly short of tension flange instead of tight fit makes fabrication and
placement of stiffener between the two flanges easier. In bridges, where fatigue
load is generally present, not attaching the intermediate stiffeners to tension flange
has a more important reason, which is to avoid welds that are transverse to the
tension flange. Such transverse welds can initiate fatigue cracks under repeated
load reversals.
2. The welds connecting the intermediate stiffeners to the web should be terminated
not less than 4t w and not more than 6tw from the near toe of the web to flange
weld of the plate girder, Figure 7.10(a) on the next page. The minimum limit is
imposed to prevent high stress concentration in the web in the area free of
stiffeners. The maximum limit is imposed to prevent long stiffener free webs and
to prevent buckling of the web plate in this area.
CHAPTER 7-DESIGN FOR SHEAR 7-13

3. When single stiffeners are used and the compression flange is a rectangular plate,
the stiffeners should be attached to compression flange, Figure 7.10(b) on the next
page. This is done to prevent torsional deformation of the compression flange.
4. When lateral bracing is attached to a single stiffener, or a pair of stiffeners, the
stiffeners should be connected to compression flange to transmit one percent of
the total flange force unless flange is made of only angles.
5. If bolts are used to connect stiffeners to the web, as is the case for angle stiffeners
shown in Figure 7.10 below, the spacing of the bolts should not be more than 12
inches on center, Figure 7.10(c). If intermittent fillet welds are used, the clear
distance between the welds should not be more than 16 t w or 10 inches, Figure
7.10(d) below.

Compression
Flange

Weld
Tension
Flange

s 4 tw ≤ s ≤ 6 tw
(b) Stiffener Welded to Top
(a) Limit on the Gap Between Compression Flange
Stiffener and Tension Flange

Intermittent
fillet welds

Bolts

≤ 10 inches and 16tw ≤ 12 inches

(c) Limit on the Un-welded (d) Limit on the Spacing of

Length of Stiffeners Stiffener Bolts

Figure 7.10. Limitation and Requirements on Intermediate Stiffeners as per AISC Spec.
(2010a)

Illustrative Example 7.1:

Shear Strength of a Beam with Compact Web

The beam shown below is a W24x162, A992 , Grade 50 steel. The top flange of the beam is
laterally braced every 4 feet by a floor joist. Calculate factored load P u that can be applied to the
section. Check shear failure mode only. Tension Field Action is not used.

Pu W24x162,
Pu
A992 Gr. 50

15 ft 20ft 15 ft

Given:
1. Cross section of the member
2. Material of steel as A36 with F y = 50 ksi and F u as 65 ksi.
Required: h=20.7” d=25”
To calculate shear capacity of the member.

SOLUTION:
Properties of the W24x162 are:
b f =13 in, t f = 1.22 in, t w = 0.705 in, h = 20.7 in. , d = 25 in., r y = 3.05 in., Z x =468 in3

Check Shear:
Vu ≤ φv Vn (in LRFD)
V u = P u (see above shear diagram)
V n = 0.6F y A w C v

15 ft 20 ft 15 ft

Shear Pu
Diagram
-Pu

Check web compactness: h / t w ≤ 2.24(E/F y )0.5

Since h / t w = 29.4 < 2.24(29000 ksi / 50 ksi)0.5 = 53.9, web is compact and
C v = 1.0
V n = 0.6F y A w C v = 0.6 x 50 ksi x 25” x 0.705” x 1.0 = 528.7 kips
φ v = 1.00 since this is a rolled I-shape. For other sections such as welded I-shapes, φ v would be
0.90 , (see Page 7-9).
φ v V n = 528.7 kips
P u = 528.7 kips due to shear yielding of web.
CHAPTER 7-DESIGN FOR SHEAR 7-15

Illustrative Example 7.2:

Shear Strength of a Beam with Non-Compact Web
Consider the beam shown below. All plates are A572, Gr. 50. Floor decks laterally brace the top
flange of the beam. Calculate factored load P u that the beam can resist. Check shear failure
mode only. Tension Field Action is not used.

Stiffeners: PL16"x2"
Pu Pu PL5x1/2, A572 Top & Bott.
Gr,50 Flanges
(on one side
of web only) Stiffeners:
PL5x1/2, A572
Gr.50
(on one side
15 ft 5 ft 15 ft PL 40"x 1/2"

Given:
1. Cross section of the member
2. Material of steel is A572, Gr. 50 with F y = 50 ksi and F u as 65 ksi.
Required:
To calculate strength of the beam based on shear failure mode

SOLUTION:
Properties of the cross section are:
b f =16 in, t f = 2 in, t w = 0.5 in, h = 40 in., d = 44 in., Area = 84 in2, Weight per foot= 285 lb.

Check Shear:
Vu ≤ φv Vn (in LRFD)
V u = P u V n = 0.6F y A w C v

Check h/t w :
h/ t w = 40/0.5 = 80
Calculate 1.1(k v E/F y )0.5 :
5.0 5.0
k v = 5.0 + 2
= 5.0 + = 7.22
(a/ h) ( 60" / 40" )2
1.1(k v E/F y )0.5 = 1.1(7.22 x 29000 ksi / 50 ksi)0.5 = 71
Calculate 1.37(k v E/F y )0.5 :
1.37(k v xE/F y )0.5= 1.37(7.22x29000 ksi / 50 ksi)0.5 = 88
Since h/ t w = 80 is greater than 1.1(k v E/F y ) 0.5 and less than 1.37(k v E/F y ) 0.5, then this section is
“non-compact” and C v is given by

1.10 k v E / Fy
Cv =
h / tw

C v = 1.1(7.22x29000 ksi / 50 ksi)0.5 / 80 = 0.89

φ v = 0.9 (for LRFD) since this is not a rolled shape. (For rolled I-shapes, φ v would be 1.0.)
φ v V n = φ v (0.6F y A w C v ) = 0.90 x 0.6 x 50 ksi x 44” x 0.50” x 0.89 = 528.7 kips

P u = 528.7 kips.
Check Stiffeners
Required I ww = at w 3 j

J = (2.5)/ (a/h) 2 - 2 ≥ 0.5

J = (2.5)/ (60/40) 2 - 2 = 0.88 ≥ 0.5 Use J=0.88

I ww = at w 3 j = 60x (0.5)3x0.88 = 6.6 in4
Actual I w = 0.5" x (5") 3 / 3 = 20 in4
Since actual I w = 20 in4 is greater than the required value of 6.6 in4, then design of the stiffeners is
OK.

References for Chapter 7

1. AISC, (2010a) “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An
American National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-
05, An American National Standard, American Institute of Steel Construction,
Chicago.
3. AISC, (2011) “Steel Construction Manual,” 14th Edition, American Institute of Steel
Construction, Chicago.
CHAPTER 7-DESIGN FOR SHEAR 7-17

Problems on Shear Design of Beams

Problem 7.1. Considering shear strength only, calculate φ v V n of the cross sections shown below.
For sections that have stiffeners, check adequacy of stiffener. Shear is applied along the y-y axis.
The spacing of stiffeners for each case is indicated below the cross sections.

y y PL10x1” A36
Top & Bott.
y
Weld
x x x x
both sides
W24x55 x x
W30x99
A992 A992
PL 20x1/2
A36
y y

No Stiffeners Used. y
No Stiffeners Used.
(A) No Stiffeners Used.
(B)
(C)

Plates 10"x1", Plate 16"x2",

A36 A36
Top & Bottom
Plate 12"x 2",
A36

Plate 40"x 3/8"

A36 All steel: A36

(E) No Stiffeners Used

Stiffener spacing=40”
(D) (F)

W12x190, A992,
Gr. 50
W24x84,
A992, Plate 24"x3",
Gr. 50 Plate 12x1 A36, top & bott.
A572, Gr. 50
on both sides
Plate 23x1/2
3 Plates
A572, Gr. 50 on
36"x1½",
both sides
A36
No Stiffeners Used No Stiffeners Used

(G) (H)
No Stiffeners Used
(I)

Problem 7.2. Calculate maximum factored load P u that can be applied to the beam shown in the
figure. Consider only shear failure mode. The cross section of the beam is given below. No
stiffeners are used.
PL6x1”, A36 y
3Pu Pu 2Pu
Weld both sides
x x

PL 9x1/2
9’-6” 3@8’-0”=24’-0” A36
y
(A)
PL18x1”, A36
y
Pu Pu 2Pu
Weld
both
sides x x

PL 24x1/2
2@10’-0”=20’-0” 2@10’-0”=20’-0” A36
y
(B)

Problem 7.3. Design the following beams to carry the loads. Steel for all wide flange rolled
shapes is A992 Grade 50. Lateral bracings are provided at locations shown. The loads shown are
service (un-factored ) loads. Consider only shear failure mode.

1.8 kips/ft 20 kips

(LL=1.0 k/ft , DL=0.8 k/ft) (LL=10 kips DL=10 kips) Lateral bracing provided
throughout the flange

10 ft 2@15 ft=30ft 10 ft

(A)
5 kips
10 kips (LL=3 kips DL=2 kips)
(LL=6 kips DL=4 kips)

10 ft 10ft 8 ft

= Lateral Bracing on Compression Flange

(B)

DL=2 k/ft
LL=1.5 k/ft

10 ft 20 ft 20 ft 10 ft

= Lateral Bracing on Compression Flange

Problem 7.4. Find maximum value of factored load P u that can be applied to the beam. The steel
used is A572 Gr. 50. Consider bending and shear failure modes. The locations of lateral braces of
compression flange are shown in the figure by ⊗ sign. Notice that Points C and D do not have
lateral bracing. When checking for bending, check local buckling and lateral-torsional buckling
modes. Stiffeners are provided at A, B, C, D, E , and F on one side of the web. Also, check
stiffeners. y
20”
Pu /20 k/ft 2Pu kips 2Pu kips Pu /20 k/ft
tf = 2”

tw = ½”
A B C D E F x x h=36”
20 ft 15 ft 15 ft 15 ft 20 ft

tf =2”
Beam Loading 20”

Beam Cross Section

Problem 7.5. Find maximum value of factored load Q u that can be applied to the beam. The steel
used is A36 Gr. 50. Consider bending and shear failure modes. The locations of lateral braces of
compression flange are shown in the figure by ⊗ sign.

PL32”x2” y
3.5Qu 3.5Qu A36
kips 2.5Qu kips kips
Weld 24”
⊗ ⊗ ⊗ x x

. 24”
3PL48”x3/4”
5 ft 14 ft 14 ft 5 ft A36
30”
Elevation of the Beam y
Beam Cross Section

(A)

PL16”x1”
y
2Qu kips 2Qu kips A36
Qu kips

Weld 12”
⊗ ⊗ ⊗ ⊗ ⊗ x x

2PL24”x1/2” 12”
7 ft 5 ft 9 ft 3 ft A36
12”
Elevation of the Beam
y
(B) Beam Cross Section
____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others
or posted anywhere.
7-20
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Problem 7.5. Determine maximum value of factored load P u that can be applied to the beam
shown below. Consider bending (i.e. lateral torsional buckling and local buckling) and shear
failure modes. Lateral bracings are provided shown on each beam. The material is all A36 with
F y =36 ksi and F u =58 ksi.
tf = 1-1/2”
Pu kips y
Pu /10 kips/ft
tw= 3/8”
h =30”
x x

B C D
tf = 1-1/2”
15 ft 15 ft 10 ft y
bf = 24”
Note:
= Lateral bracing
Beam Section
(A)

y
Pu/2 k/ft PL15”x1/2”
Pu kips
Weld
both sides

x x
PL48”x3/8”
PL 15”x1/2”
18 ft 18 ft 10 ft
y
= Lateral Bracing on Compression Flange Beam Cross Section
Elevation of the Beam
(B)

Top Flange:
Pu kips PL 20”x1.5”
0.5Pu kips Pu/6 (kips/ft) Pu kips
0.5Pu kips

Web:
PL36x3/8

5 ft 18 ft 5 ft
Bottom Flange:
= Lateral bracing PL 20”x1.5”

(C)

Top Flange:
2Pu kips PL 16”x2”
Pu kips Pu/3 (kips/ft) 1.5 Pu kips
Pu kips

Web:
PL36x5/8

6 ft 20 ft 6 ft
Bottom Flange:
= Lateral bracing PL 16”x2”

(D)
CHAPTER 8 – BEAM-COLUMNS 8-1

CHAPTER 8

BEAM-COLUMNS

Objectives:
In this chapter, you will learn:
1. Behavior of members subjected to combined axial load and bending.
2. Failure mode (or limit state) of beam columns. These failure modes are
failure modes of beams and columns.
3. How to design a beam-column for combined axial load and bending
4. How to apply what you have learned to actual design and evaluation of
beam-columns by using numerical example provided in this chapter.

8.1. INTRODUCTION
This chapter discusses behavior and design of steel members subjected to combined loads.
The combined loads considered here are axial load (tension or compression) bending (about x-x
and/or y-y axis), shear and torque.
Chapter H of the AISC Specification (AISC, 2010a), which discusses beam-columns, is
organized into three sections to address: (a) Doubly and singly symmetric members subjected to
combined flexure and axial force, (b) unsymmetrical and other members subject to combined
flexure and axial force; and (c) members under torsion and combined torsion, flexure, shear and/or
axial force. In this chapter, we will discuss case (a) above. For Cases (b) and (c) the reader
isreferredto the AISC Specifications (AISC, 2010a).

8.2. GENERAL ASSUMPTIONS

The following assumptions have been made in developing design procedures for members
subject to combined axial load and bending moment about one or both principle axes:
a. Members are prismatic which means cross section of member is the same throughout its
length.
b. Axial load is acting through the centroid of the cross section.
c. Load is applied gradually and can be considered static load.
d. For built-up members in compression, limitations given in Section E6.2 of the AISC
Specifications are satisfied.

8.3. DOUBLY AND SINGLY SYMMETRIC MEMBERS

SUBJECTED TO FLEXURE AND AXIAL FORCE
8.3.a. Doubly and Singly Symmetric Members Subject to Flexure and
Compressive Axial Force

Figure 8.1(a) shows a steel member subjected to bending and axial compression. Under
combined bending and axial force, elastic tensile and compressive stresses develop in the cross
sections as shown in Figure 8.1(b). The maximum values of these stresses are at the extreme
fibers of the cross section. As loading continues to increase, the maximum compressive stress, due
to combined bending and axial compression, increases until it reaches yield point and the extreme
compressive fiber yields. This constitutes yield point of the cross section beyond which the
member will behave in an inelastic manner shown in Figure 8.1(c).

P P

M P
Elastic
M Behavior
σa = P/A
(-)
+ Inelastic
σb = Mc/I Behavior
(-)
+Fy
σtotal = P/A+Mc/I Idealized
(-) Plastic
Behavior -Fy

(a) (b) (c)

Figure 8.1. (a) Member Subjected to Bending and Axial Compression, (b) Combination
of Bending and Axial Stresses , and (c) Elastic , Inelastic and Fully Plastic Combined
Stresses
CHAPTER 8 – BEAM-COLUMNS 8-3
P

Plastic stress P
distribution due +Fy M
to Combined axial (tension)
load and bending -Fy d
moment (compression) b
M
Pure bending +Fy
stresses
-Fy
P

-Fy
Pure axial stresses
a

Figure 8.2. Plastic Stress Distribution

Further increase of the loading beyond yield point will cause yielding of more and more
fibers and strain hardening of the fibers that have already yielded. Eventually the strain in the
extreme compression fiber reaches ultimate strain and strength of the cross section starts to drop.
The above behavior of a steel member under combined bending and axial compressive
load can change if member experiences local or overall buckling. Considering Figure 8.1(b),
when maximum strain in the compression extreme fiber reaches the critical strain for local or
overall buckling, the member can buckle and fail. We will return to buckling issues later in this
section.
In design, the idealized plastic state of stress, shown in Figure 8.1(c) is used to establish
plastic capacity of members subjected to combined axial loads and bending.
As Figure 8.2 above shows, the plastic stress distribution can be divided into two stress
distributions, one creating pure bending and one creating pure axial compression. For a
rectangular shape, the relationship between compression force P and bending moment M at the
time of full yielding of the cross section (i.e. yielding of all fibers) , shown in Figure 8.2, can be
written as:

M = (b / 4)(d 2 − a 2 )( F y ) (8.1)

P = (b)(a )( F y ) (8.2)

By eliminating “a” in the above equations, the relationship between P and M is obtained
as:

2
 P 
  + M = 1.0 (8.3)
 Py  Mp
 
where
a = width of core area resisting axial force, Figure 8.2 on previous page
b = width of rectangular section, Figure 8.2
d = depth of rectangular section, Figure 8.2
F y = yield stress of steel
M = applied bending moment
M p = plastic moment capacity = (bd 2 /4) (F y )
Z x = plastic section modulus= bd 2/4 for solid rectangular shape
P = applied compressive force
P y = yield capacity in pure compression= A F y

Equation 8.3 above is called “interaction” equation for a rectangular member subjected to
combined bending and axial compression force. Similar equations can be derived for wide
flanges, tee sections, angles, and other shapes. Figure 8.3 schematically shows these equations for
a number of typical sections.

P / Py

Tube
1.0 WF Solid
Rectangle

2-L
2-L

M / Mp
1.0

Figure 8.3. Interaction Equations for Combined Compression and Bending

When Only Yielding Failure Mode is Considered

Steel members subjected to bending combined with axial compressive force have failure
modes associated with compression alone and bending alone applied to the member. Therefore,
failure modes of members subjected to combined bending and axial compression are:
1. yielding of gross area
2. local buckling of compressive elements of the cross section
3. formation of plastic hinge
4. lateral torsional buckling of the compression flange
5. overall buckling
6. failure of connections
CHAPTER 8 – BEAM-COLUMNS 8-5

The P-M interaction curve for members subjected to bending and axial compression
strongly depends on the slenderness ratio KL/r of the member and the shape of the bending
moment diagram (Tall, 1974). Based on analysis of the behavior of beam-columns, it has been
established that the more slender the member is the less strength it shows under combined bending
and axial compression. In addition, bending moment diagrams, which cause double curvature in
the member, have the least reduction effect while uniform bending moment diagram, creating a
single curvature, results in largest reduction in the capacity of a beam-column. Curve A in Figure
8.4 shows a typical P-M interaction curve when only yielding of the cross section is considered
and no local or overall buckling occurs. Curve B on the figure shows a typical P-M interaction
curve when overall buckling of the beam column is also included.

P / Py
Curve A
Interaction Curve
Based on Yielding
1.0
WF
Curve B
Pcr / Py Interaction Curve
Based on Buckling

Mcr /Mp 1.0 M / Mp

Figure 8.4. Interaction Equations for Combined Compression and Bending

When Buckling Failure Modes are Also Considered (Curves are for Wide Flange only)

Pr
Pc
Pr 8  M r 
1.0 +   ≤ 1.0
Pc 9  M c 

Pr  M r 
+  ≤ 1 .0
0.20 2 Pc  M c 

Mr
1.0
Mc
Figure 8.5. Interaction Equations for Combined Bending and Axial Compression

Current steel design specification (AISc, 2010a) uses a linear approximation for the P-M
interaction curve to define failure of members subjected to combined bending and axial load.
Axial loads are normalized (divided) by the critical axial strength and bending moments are
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-6

normalized by dividing them by the critical bending moment due to possibility of local or lateral-
torsional buckling. The resulting interaction equations, which are currently in the AISC
Specification (2010a) are shown in Figure 8.5 above and are given as:

Pr 8  M r 
For P r / P c ≥ 0.2 +   ≤ 1.0
(8.4) Pc 9  M c 

Pr  M r 
+  ≤ 1.0
2 Pc  M c 
For P r / P c < 0.2
(8.5)

where
P r = required axial compressive strength using LRFD or ASD load combinations established by
following procedures in the next section (Sec. 8.3.b).
P c = available axial compressive strength assuming bending moment to be zero. (use Chapter 5 to
establish this strength)
P c = φ c P n for LRFD method and P c = P n / Ω c for ASD method
φ c = resistance factor for compression (in LRFD method) = 0.90
Ω c = safety factor for compression (in ASD method)= 1.67
P n = nominal axial compressive strength assuming bending moment is zero. (use Chapter 5 to
establish P n )
M r = required flexural strength about axis of bending established as discussed in the following
section(Sec. 8.3.b).
M c = available flexural strength w.r.t axis of bending assuming axial load is zero.
M c = φ b M n for LRFD method and M c = M n / Ω b for ASD method
M n = nominal flexural strength w.r.t. axis of bending assuming axial load is zero. (use Chapter 6
to establish M n )
φ b = resistance factor for flexure (in LRFD method) = 0.90
Ω b = safety factor for bending (in ASD method) =1.67

Illustrative Example 8.1:

Beam-Column Section Interaction Curves (Without Buckling
Consideration)
For cross section shown below, if applied bending moment is W12x 120
equal to 60% of the plastic moment capacity of the section A992 Grade 50
about x-x axis , how much factored axial load P r can be
applied to the cross section? The member is very short
and bucklings is not a consideration.
x

Given:
1. Cross section of the member as W12x120 y
2. The material is given as A992 grade 50 steel with (B)
F y = 50 ksi and F u = 65 ksi.
Required:
To calculate bending capacity of the cross section when an axial load is also present.
CHAPTER 8 – BEAM-COLUMNS 8-7

SOLUTION:
Properties of the W12x120 are:
b f / 2t f = 5.57 (from Manual Tables) , h / t w = 13.7 in. From Manual Tables) , A g = 35.2 in2 , Z x = 186 in3

Pr
Pc
The interaction equations that we will use
Pr 8  M r 
are those in the AISC Specification (2010a) 1.0 +   ≤ 1.0
shown in the sketch to the right. Pc 9  M c 

Pr  M r 
+  ≤ 1.0
0.20 2 Pc  M c 

Mr
1.0
Mc
We need to establish four terms in the interaction equations, which are: P r , P c , M r , and M c .

1. Establish P r :
P r is the factored axial load applied to the member. In this problem, P r is the unknown. So, we will live it as
Pr .

2. Establish P c :
P c is the axial compressive strength of the cross section based on the LRFD method. To calculate P r , we
need to use information in Chapter 5 , which was on axially loaded columns.

Check Local Buckling Assuming the Cross Section to be a Column:

Since b f / 2t f = 5.57 is less than λ r = 0.56(E/F y )0.5 = 0.56(29000/50)0.5 = 13.5, flanges are not slender.
Since h/t w =13.7 is less than λ r = 1.49(E/F y )0.5 = 1.49(29000 ksi / 50 ksi)0.5 = 35.9, web is not slender.
Since both flanges as well as the web are not slender, the cross section is not slender and can reach overall
buckling capacity prior to local buckling.

Check Overall Buckling Assuming the Cross Section to be a Column:

Since problem statement indicates that buckling is not a concern in this problem, the axial strength will be
equal to axial yield strength:
P c = φ c P n = φ c F y A g = (0.90)(50 ksi)( 35.2 in2 ) = 1,584 kips

3. Establish M r :
M r is the factored bending moment applied to the member. Let us establish M r .
The applied factored moment, M r , is given as 0.60M px
Where
M px = plastic moment capacity of the cross section = F y Z x = 50 ksi x 186 in3 = 9,300 k-in
Therefore,
M r = 0.60 x M px = 0.60 x 9,300 k-in
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-8

M r = 5,580 k-in.
4. Establish M c :
M c = bending moment capacity of the cross section
To establish M c , we need to check limit states of a beam.
Check Local Buckling Assuming the Member to be a Beam
We will check local buckling assuming this is a beam and apply limitations given in Chapter 6 for beams.
Since b f / 2t f = 5.57 is less than λ p = 0.38(E/F y )0.5 = 0.38(29000/50)0.5 = 9.1, flanges are compact.
Since h/t w =13.7 is less than λ p = 3.76(E/F y )0.5 = 3.76(29000 ksi / 50 ksi)0.5 = 90, web is compact.
Since both flanges as well as the web in this beam section are compact, the cross section is compact and
can reach bending capacity prior to local buckling.

Check Lateral-Torsional Buckling Assuming the Member to be a Beam

Since we only have cross section and not the unbraced length of the beam, we cannot check lateral-torsional
buckling. Assuming lateral torsional buckling does not govern, bending strength of the cross section is equal
to φ M px , where:
φ M px = 0.90 x 9,300 k-in = 8,370 k-in.

5. Check the Beam-Column Section for Yielding and Formation of Plastic Hinge due to
Combined Axial Load and Bending Moment
Assuming that P r / P c is greater than 0.20, the interaction to be used is as follows. We will check this
assumption later after we find value of P r.

Pr 8  Mr 
+   ≤ 1.0
Pc 9  M c 

Plugging above values into interaction equation we obtain:

Pr 8  5580 k − in 
+   ≤ 1.0
1584 kips 9  8370 k − in 

P r = 645 kips.
Let us check if our assumption of P r / P c > 0.20 is correct.
P r / P c = 645 kips / 1584 kips = 0.41 which is greater than 0.20, therefore, the value of P r is;

P r = 645 kips.

8.3.b. Calculation of Required Bending Strength M r and Required

Axial Strength P r Buckling is Involved
When a steel member is subjected to combined axial compression and bending, the axial
load acting on the deflections created by bending moment results in additional bending moment,
Figure 8.6(a). This member “P-δ” effect, in the form of additional bending moment, should be
considered in design and added to the moments created in the members by applied loads.
In addition to “member P-δ effects”, if one end of a beam-column is displaced
relative to the other end, as shown in Figure 8.6(b), the resulting displacement times the axial load
creates additional bending moments in the member. These so-called frame “P-∆” effects are due
to deformations of the global framing system and not due to deformation of the member itself,
which creates the P-δ effects discussed above.
CHAPTER 8 – BEAM-COLUMNS 8-9

In the current AISC Specification (AISC, 2010a), the three above-mentioned effects,
namely the effects of member P-δ, the effects of global frame P-∆ and the effects of shape of the
bending moment diagram, are considered in design by conducting a second order analysis.
Currently, there are two types of analysis of frames, First Order and Second Order. First
order analysis means that the analysis is done using a computer model of the structure that has the
initial geometry. Second order analysis means the analysis is done on the deflected geometry of
the structure. The difference between first order and second order analyses is that in the first order
analysis the effects of P-δ and P-∆ are not included while in the second order analysis, these
effects are directly included by analyzing the deflected frame. The AISC Specification (AISC,
2010a) provides two methods for including the P-δ and P-∆ effects in design:
(a) Second Order Elastic Analysis, and;
(b) Amplified First-Order Elastic Analysis.
As mentioned earlier, in general, in a Second Order Analysis both P-δ, and P-∆ second
order effects are directly included in the analysis. In the Amplified First Order Elastic Analysis,
bending moments and axial forces resulting from first order analysis of the structure (without
consideration of P-δ and P-∆ effects) are amplified in an approximate manner to represent the P-δ
and P-∆ effects. In the following, this approximate method, which engineers quite often use in
design offices, is presented. This method is in Appendix 8 of the AISC Specification (2010a).
To do an Amplified First Order Elastic Analysis, we need to separate bending moments in
the members into two parts. One part, called no-translation moments, M nt are bending moments
due to applied loads while horizontal movements of floors are prevented, Figure 8.7(b). The
second part, called lateral-translation moments, M lt are bending moments due to horizontal
movement of the floors. To obtain M nt and M lt , we need to analyze two frames shown in Figures
8.7(b) and 8.7(c). Frame “nt (no translation)” is the original frame but roller supports are added at
each floor level to prevent lateral movements of the floors due to story shear. Frame “lt (lateral
translation)“ is the same frame as the original but is only subjected to story shears by applying the
floor roller support reaction to it, Figure 8.7(c). Superposition of bending moments and axial loads
resulting from analysis of “nt” and “lt” frames, gives us bending moments and axial forces in the
original frame, Figure 8.7(a).
∆

δ
δ

Bending P
Bending moment
P
moment due to
due to P times δ P times ∆

a-Frame with no lateral translation b-Frame with lateral translation

Figure 8.6. (a) Member P-δ Effects and (b) Member P-δ Effects Superimposed on Frame P-∆ Effects , both for
the Left Column
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-10

R3 R3

R2 R2

+
= R1
R1

(a) Original Frame (b) “nt” Frame (c) “lt” Frame

(No Translation Frame) (Lateral Translation Frame)

Figure 8.7. Frames to be Used to Calculate Mnt and Mlt

After separating bending moments and axial forces into two parts, M nt and M lt ,
values of required flexural strength, M r , and required axial strength, P r , can be obtained from
following approximate equations:

M r = B 1 M nt + B 2 M lt (8.6)
P r = P nt + B 2 P lt (8.7)

The first term on the right side of above equations represents the P-δ effects only and the
second term represents the P-∆ effects. The above equations are generally used for moment
frames. For braced frames, since lateral translation of floors is relatively small, the P-∆ effects
are small and the second term on the right side of the above equations can be ignored.
In the above equations:
M r = amplified required bending strength which includes second-order effects
M nt = applied first order moment using LRFD or ASD load combinations, assuming there is no
lateral translation of the frame. The condition occurs in braced frames or in moment
frames when the horizontal translation of the floors is prevented.
M lt = applied first order moment using LRFD or ASD load combinations, caused by lateral
translation of the frame. The condition, more or less, occurs in moment frames. For
braced frames, Mlt is negligible and can be assumed zero.
B 1 = amplification factor to incorporate member P-δ effects given by:

Cm
B1 = ≥ 1.0 (8.8)
( 1 − αPr / Pe1 )
where
C m = a coefficient assuming no lateral translation of the frame, given by:
CHAPTER 8 – BEAM-COLUMNS 8-11

a. For beam-columns not subjected to transverse loading between supports in the plane
of bending:

C m = 0.6 − 0.4( M 1 / M 2 ) (8.9)

where M 1 and M 2 are the end moments of the member calculated from first order
analysis M 1 is the smaller of two end moments. The term M 1 /M 2 is positive if
member has reverse curvature, Figure 8.8(a), and negative if member has single
curvature throughout its length, Figure 8.8(b).

b. For beam-columns with transverse load between their supports, value of C m should
be determined by analysis or taken as equal to 1.0 for all cases.

δ δ
M2
M1 M2 M1
(a) (b)

Figure 8.8. (a) and (b) , Beam-Columns Not Subjected to Transverse Loads

α = 1.00 for LRFD and α = 1.60 for ASD

P e1 = critical elastic buckling resistance of the member in plane of bending given by:
π EI
2

Pe1 = 2 (8.10)
(K 1 L )

E = modulus of elasticity of steel = 29,000 ksi

I = moment of inertia of cross section with respect to the axis of bending
K 1 = effective length factor in the plane of bending, determined based on the assumption of
sidesway prohibited. Conservatively, a K 1 =1.0 can be used.
L = length of the member
B 2 = amplification factor to incorporate frame P-∆ effects given by:
1
B2 = ≥ 1.0 (8.11)
( 1 − αPstory / Pe story )
P story = total vertical load supported by the story, including loads in the columns that are not part of
the lateral load resisting system (see Figure 8.9).

P e story = elastic critical buckling strength of the story in the direction of translation being
considered determined by sidesway buckling analysis or by using following equation:

HL
Pe story = R M (8.12)
∆H
R M = 1- 0.15(P mf / P story )
P mf = total vertical load in columns that are part of moment frame, (see Figure 8.9)
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-12

H = story shear ( sum of the shear forces in the columns of the moment frame)
L = height of story, (see Figure 8.9).
∆ H = first order interstory drift in the direction of translation being considered (i.e. in the plane of
the frame). Where ∆ H varies over the plan area of the structure, ∆ H should be the average
drift weighted in proportion to vertical load or, alternatively, the maximum drift. In early
stages of design, if value of ∆ H is not known, it common to use the drift ratio limits such
as 0.02L for wind and 0.03 for seismic to calculate ∆ H .
Moment Frame

Pstory is sum of axial

L force in all columns
of the story.

Pmf is the sum of axial load in these columns that

are part of moment frame.
H is the story shear resisted by these columns.

Figure 8.9. Establishing Pstory, Pmf and H

Illustrative Example 8.2:

Beam-Columns Where Sidesway is Inhibited (Braced Frames)
Consider the beam-column shown below loaded axially and laterally. The cross section of the beam-column
is shown in the figure. The axial factored load is 500 kips. Calculate factored lateral load Q u . Steel is A572
Gr. 50. Lateral bracings are provided throughout the length and lateral torsional buckling is not a concern.

Qu PL 2” x 18”
y (both flanges)
500 kips
PL 1”x 34”

500 kips x x
20 ft
20 ft Continuous
Welds

y
CHAPTER 8 – BEAM-COLUMNS 8-13

Given:
1. Cross section of the member is given as a welded plate girder.
2. The material is given as A572 grade 50 steel with F y = 50 ksi and F u = 65 ksi.

Required:
To calculate factored lateral load Q u .

SOLUTION:
Calculate Cross Section Properties:
A g = Σ(A) i = 2(18”×2”)+1”×34” = 106 in2
I x = Σ(bh3/12+Ad2 ) i = 2[(18”)(2”)3 /12 + (18”)(2”)(18”)2 ] + (1)(34”)3 /12 = 26,627 in4
I y = Σ(bh3/12+Ad2) i = 2[ (2”)(18”)3 /12 ]+(34”)(1”)3 /12 = 1946 in4
r x = (I x / A g )0.5 = (26,627 in4 / 106 in2)0.5 = 15.8 in.
r y = (I y / A g )0.5 =(1946 in4 / 106 in2)0.5 = 4.3 in.
S x = (I x /C ) = (26,627 in4) / (19“) = 1,401 in3
Z x = Σ(Ad) i = 2[ (2”)(18”)(18”) ]+2(1”)(17”)(8.5”) = 1585 in3.
The interaction equations that we will use are those in the AISC Specification (2010a) shown in the sketch
below.
Pr
Pc
Pr 8  M r 
1.0 +   ≤ 1.0
Pc 9  M c 

Pr  M r 
+  ≤ 1.0
0.20 2 Pc  M c 

Mr
1.0 Mc
We need to establish four terms in the
nteraction equations, which are: P r , P c , M r , and M c .

1. Establish P r :
P r is the factored axial load applied to the member. In this problem, P r is given as 500 kips (see problem
statement sketch)
P r = 500 kips

2. Establish P c :
P c is the axial compressive strength of the cross section based on the LRFD method. To calculate P c , we
need to use information given in Chapter 5.

Check Local Buckling Assuming the Cross Section to be a Column:

Since b f /2t f = 18”/(2x2”) = 4.5 is less than λ r = 0.56(E/F y )0.5 = 0.56(29000 ksi / 50 ksi )0.5 = 13.5, flanges
are not slender.
Since h/t w = 34”/1” = 34 is less than λ r = 1.49(E/F y )0.5 = 1.49(29000 ksi / 50 ksi)0.5 = 35.9, web
is not slender.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-14

Since both flanges as well as the web are not slender, the cross section is not slender and can reach overall
buckling capacity prior to local buckling.

Check Overall Buckling Assuming the Cross Section to be a Column:

Again we need to use information from Chapter 5. The equation of design for limit state of overall buckling of
axially loaded members is P u ≤ φ c P n where P n = F cr A g . Let us calculate F cr :
Find governing effective slenderness ratio:
(KL/r) x = 1.0×40’×12”/15.8” =31
(KL/r) y = 1.0×20’×12” / 4.3” = 56 (governs)
?
Check (KL/r) y ≤ 4.71(E/F y )(0.5)
Since KL/r = 56 < 4.71(29,000 ksi / 50 ksi)(0.5) = 113, then buckling mode is inelastic and F cr is:
F cr = [0.658(Fy/ Fe) ]F y .

F e =π2 E/(KL/r)2 = (3.1415) 2(29000 ksi) / (56)2 = 91.3 ksi

F cr = [0.658(Fy / Fe) ]F y =[ 0.658(50 /91.3) ](50 ksi) = 38 ksi

P c = φ c F cr A g. = 0.90 (38 ksi) (106 in2)
P c = 3,625 kips.

3. Establish M r :
M r = B 1 M nt + B 2 M lt
Since the two ends of this beam-column do not move relative to each other , the beam column can be
considered part of a braced frame. Therefore, B 2 is equal to zero and equation of M r becomes:
M r = B 1 M nt
Where;
Cm
B1 = ≥ 1 .0
( 1− αPr / Pe1 )

C m = 1.0 (since there is a transverse load on the member (see Page 8-11, Eqn 8.8)
2
π EI
Pe1 = 2
(K 1L )
2 4
π ( 29000 ksi )( 26627 in )
Pe1 = = 33,076 kips
( 1.0 )( 480 in.)2

α = 1.0 (for LRFD )

1 .0
B1 = = 1.015
1− ( 1.0 )( 500 kips / 33076 kips )
Qu
M r = B 1 M nt
500
M nt = Maximum applied factored
kips
bending moment = Q u L /4
M nt = Q u x 40’x12” /4 = 120 Q u k-in
M r = B 1 M nt = 1.015 x 120” Q u = 122Q u 20 ft
20 ft 500
kips
CHAPTER 8 – BEAM-COLUMNS 8-15

4. Establish M c :
To establish M c , we assume there is no axial load and member is under pure bending as a flexural member.
We will use information from Chapter 6 , which was on flexural members.

Check Local Buckling Under Pure Bending Assuming that the Member is a Flexural Member with No Axial
Load :
Check flange local buckling:
Since b f /2t f = 18” / (2x2”) = 4.5 is less than λ p = 0.38(E/F y )0.5 = 0.38(29000 ksi / 50 ksi)0.5 = 9.1 , flanges are
compact.

Check web local buckling:

Since h/t w = 34” / 1” = 34 is less than λ p = 3.76(E/F y )0.5 = 3.76(29000 ksi / 50 ksi)0.5 = 90, web is compact.
Since both flanges and the web are compact, cross section is compact and is expected to yield prior to local
buckling.

Check Lateral Torsional Buckling:

Problem statement indicates that there is no lateral torsional buckling.
Therefore, M n = M p , where
M p = Plastic Moment Capacity = F y Z x
M p = (50 ksi) (1585 in3) = 79,250 kip-in (Z x was calculated at the start of solution earlier)
M c = φ b M n = 0.9 x 79,250 k-in = 71325 k-in.
5. Check the Beam-Column Section Subjected to Combined Axial Load and Bending Moment
The interaction Equation is:
Pr 8  Mr 
+   ≤ 1.0
Pc 9  M c 
The values of parameters in the above equation are:
P r = 500 kips. (given factored axial load
Qu
P c = 3,625 kips
500
M r = 122 Q u
kips
M c = 71,325 k-in.
By inserting above values in the interaction 500
equation we obtain: 20 ft ki
20 ft
500 kips 8  122Q u 
+   ≤ 1.0
3625 kips 9  71325 k − in 

Q = 567 kips.

8.3.c. Doubly and Singly Symmetric Members Subjected to Combined

Bending about Both x-x and y-y Axes and Axial Compression Load
If member is subjected to bending about both principle axes and axial compression load,
then Equations 8.4 and 8.5 given earlier can be written to include both bending moments as given
below and shown in Figure 8.10.

Pr 8  M rx M ry 
 ≤ 1.0
For P r / P c ≥ 0.2 + + (8.13)
Pc 9  M cx M cy 

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-16

For P r / P c < 0.2 Pr  M rx M ry 

 ≤ 1.0
+ +
2 Pc  M cx M cy  (8.14)
where 
P r = amplified required axial compression strength calculated using Equation 8.7 given earlier.
P c = available axial compressive strength (use Chapter 5 to establish this strength)
P c = φ c P n for LRFD method and P c = P n /Ω c for ASD method
M rx = required flexural strength about x-x (strong) axis, calculated using Equation 8.6 given
earlier.
M ry = required flexural strength about y-y (weak) axis, calculated using Equation 8.6 given earlier.
M cx = design flexural strength for bending about strong axis
M cx = φ b M nx for LRFD methods and M cx = M nx /Ω b for ASD methods
M nx = nominal bending capacity for bending about x-x axis (use Chapter 6- Flexural Members to
establish M nx )
M cy = design flexural strength for bending about weak axis
M cy = φ b M ny for LRFD methods and M cy = M ny /Ω b for ASD methods
M ny = nominal bending capacity for bending about y-y axis (use Chapter 6- Flexural Members to
establish M ny )
φ c = resistance factor for compression (in LRFD method) equal to 0.90.
φ b = resistance factor for flexure (in LRFD method) =0.90
Ω t = safety factor for compression = 1.67.
Ω b = safety factor for bending (in ASD method) =1.67

Pr
1.0
Pc

1.0
M ry
1.0
M rx M cy
M cx
Figure 8.10. Interaction Equation for Combined Tension and Bending About Both
Principle Axes

8.3.d Members Subjected to Bending and Axial Tension

Figure 8.11(a) shows a steel member subjected to combined tension force and bending. As
the tension force and bending moment increase, elastic tensile stress, due to combined bending and
axial compression, increases until it reaches yield point and the extreme tension fiber yields.
This constitutes yield point of the cross section beyond which the member will behave in
an inelastic manner shown in Figure 8.11(c). Further increase of loading beyond yield point will
cause yielding of more and more fibers. Eventually the extreme tension fiber reaches ultimate
strength, necks, and fractures initiating fracture of the cross section. In design, the above inelastic
stress distribution is idealized as fully plastic condition shown in Figure 8.11(c) where plastic
CHAPTER 8 – BEAM-COLUMNS 8-17

capacity is defined as the condition where all tension and compression fibers across the cross
section have reached yield point either in tension or in compression.
In design, the idealized plastic state of stress, shown in Figure 8.11(c) is used to establish
capacity of members subjected to tension and bending. Steel members subjected to combined
bending and tension force have failure modes associated with tension alone or bending alone
applied to the member. Therefore, failure modes of members subjected to combined bending and
tension axial force are:
1. Yielding of gross area (pure tension failure mode)
2. Fracture of net area (pure tension failure mode)
3. Formation of plastic hinge (pure bending failure mode)
4. Lateral torsional buckling of compression flange (pure bending failure mode)
5. Local buckling of compression flange (pure bending failure mode)
6. Failure of connections (under combined effects of tension and bending)

Elastic Behavior
σa = P/A
+
σb = Mc/I

|| Inelastic Behavior
σtotal = P/A+Mc/I
+Fy
Idealized Plastic
Behavior
-Fy

(a) (b) (c)

Figure 8.11. (a) A Member Subjected to Axial Tension and Bending,

(b) Combination of Axial and Bending Stresses and; (c) Elastic ,
Inelastic and Fully Plastic Combined Stresses

For members subjected to combined bending and tension axial force, current steel design
specifications (AISC, 2010a) uses the same linear interaction equations that were given by
Equations 8.4 and 8.5 earlier for combined bending and axial compression. The equations are
repeated here for the convenience followed by definition of terms. The main difference between
members subjected to bending and tension with members subjected to bending and compression is
that in members subjected to bending and tension no amplification of bending moment or axial
load is needed since in this case, P-δ and P-∆ effects reduce the bending moments and axial forces
in the members and conservatively are ignored.

Pr 8  M r 
For P r / P c ≥ 0.2 +   ≤ 1.0
(8.15) Pc 9  M c 

For P r / P c < 0.2 Pr  M r 

+  ≤ 1.0
2 Pc  M c  (8.16)
where
P r = required axial tension strength using LRFD or ASD load combinations.
P c = available axial tension strength (use Chapter 4 to establish this strength)
P c = φ t P n for LRFD method and P c = P n /Ω t for ASD method
M r = required flexural strength about axis of bending. This is the same as maximum factored
bending moment in the member.
M c = available flexural strength with respect to axis of bending
M c = φ b M n for LRFD method and M c = M n /Ω t for ASD method
M n = nominal bending capacity (use Chapter 6 to establish M n )
φ t = resistance factor for tension (in LRFD method) equal to 0.90 for yielding of gross area and
0.75 for fracture of net area
φ b = resistance factor for flexure (in LRFD method) =0.90
Ω t = safety factor for tension (in ASD method) equal to 1.67 for yielding of gross area and 2.0
for fracture of net area
Ω b = safety factor for bending (in ASD method) =1.67

8.3.e. Doubly and Singly Symmetric Members Subjected to Combined Bending

About Both x-x and y-y Axes and Axial Tension Load

If member is subjected to bending about both principle axes and axial tension load, then
equations 8.15 and 8.16 given earlier can be written to include both bending moments as was
given earlier by Equations 8.13 and 8.14 (repeated here for convenience):

For P r / P c ≥ 0.2 Pr 8  M rx M ry 
+ +  ≤ 1.0
(8.13) Pc 9  M cx M cy 


Pr  M rx M ry 
For P r / P c < 0.2
+ +  ≤ 1.0 (8.14)
2 Pc  M cx M cy 

Where
P r = required axial tension strength using LRFD or ASD load combinations.
P c = available axial tension strength (use Chapter 4 to establish this strength)
P c = φ t P n for LRFD method and P c = P n /Ω t for ASD method
M rx = required flexural strength about x-x (strong) axis, same as maximum bending moment in
the member w.r.t. x-x axis
M ry = required flexural strength about y-y (weak) axis, same as maximum bending moment in the
member w.r.t. y-y axis
M cx = design flexural strength for bending about strong axis
M cx = φ b M nx for LRFD methods and M cx = M nx /Ω b for ASD methods
CHAPTER 8 – BEAM-COLUMNS 8-19

M nx = nominal bending capacity for bending about x-x axis (use Chapter 6- Flexural Members to
establish M nx )
M cy = design flexural strength for bending about weak axis
M cy = φ b M ny for LRFD methods and M cy = M ny /Ω b for ASD methods
M ny = nominal bending capacity for bending about y-y axis (use Chapter 6- Flexural Members to
establish M ny )
φ t = resistance factor for tension (in LRFD method) equal to 0.90 for yielding of gross area and
0.75 for fracture of net area
φ b = resistance factor for flexure (in LRFD method) =0.90
Ω t = safety factor for tension (in ASD method) equal to 1.67 for yielding of gross area and 2.0 for
fracture of net area
Ω b = safety factor for bending (in ASD method) =1.67

References for Chapter 8

1. AISC, (2010a). “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An American
National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-10, An
American National Standard, Am. Institute of Steel Construction, Chicago.
3. AISC, (2011). “Steel Construction Manual,” 14th Edition, American Institute of Steel
Construction, Chicago.
4. Tall, L. (Editor) (1974), “Structural Steel Design”. 2nd Ed., John Wiley and Sons.

Problems on Beam Columns

Problem 8.1. For cross sections shown below, if applied factored axial load is equal to 60% of the yield
capacity of the section, how much factored bending moment M r about x-x axis can be applied to the cross
section? Buckling of the member is not a consideration.

y
PL 10”X1” , A36 W14x211 PL 1-1/2” x 15”
y
(both flanges) A992 Grade 50 (Top & Bott.)

PL 12’X1/2”, A36
x x x x
x PL 3/4”x 20”
Continuous
Welds Continuous Welds

y y y

(B) (C)
(A)

y W24x117
y
2 W12x120 ALL PLATES: A992 ,Grade 50
A992 Grade 50 22”X1/2”, A36
y
x
x
x

WT12x58.5
A992 ,Grade 50
(F)
(D) (E)

PL20x1, A36 PL18x1, A36

Top & Bottom
Flange
PL21x3/4,
A36

PL 20x1/2, A36
PL24x5/8, A36

(G) (H)

Problem 8.2. For cross sections shown above in Problem 8.1, if applied factored bending moment about x-
x axis is equal to 40% of plastic moment capacity M p , how much factored axial compressive force can be
applied to each cross section? Buckling of the member is not a consideration.

Problem 8.3. A W12x120, A992 Grade 50 column supports axial load P u and lateral load P u /2 as shown in
the figure. Length of column is 20 ft and K x = K y =1.4. Note that the column should be considered part of a
moment frame since the top of column can move horizontally relative to the bottom of the column. Find
maximum value of factored load P u . Maximum horizontal deflection of the top of the column is limited to
4.8 inches.
CHAPTER 8 – BEAM-COLUMNS 8-21
Pu Pu
Pu
Pu/2 Pu/2 Pu

20 ft 2@10 ft Pu
Pu/2 3@ 20 ft

(A) (B) (C)

Problem 8.4. Design the beam column to carry combination of bending and either axial compression or
axial tension. All loads are factored loads.
4 kips 4 kips

60 kips Comp 60 kips

100 kips
Tension 100 kips
Tension
4@10'-0"=40'-0"

Continuous lateral brace on tension and comp flange

Problem 8.5. Consider the beam-column loaded axially and laterally. Beam-column has a cross section
shown in the figure. Given the axial factored load as 500 kips, calculate factored lateral load Q u . Lateral
bracings are provided at end supports and under the two concentrated loads.
PL8’x1”, A36
y (both flanges)
500 kips Qu
Qu

Continuous PL24”x3/4”
Welds
500 kips
x x
. 3@10 ft =30 ft

(A) y

2Qu
Qu PL 1” x 8” , A36,
y (both flanges)
Qu
500 kips

PL 14”x3/4”, A36
Continuous
Welds
500 kips x x
3@10 ft =30 ft

(B) y
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
8-22

Problem 8.6. Consider a W 12x120 column subjected to axial load P u and bending moment M u about
strong axis. Steel is A572 Gr. 50 and KL for both x and y direction buckling is 18 feet.. Calculate, and plot
AISC-LRFD M-P interaction curve for the column.

Problem 8.7. Calculate maximum load P u that can be applied to the structure if cross section of the members
is the steel section in the figure below. Steel properties are; F y = 50 ksi and F u = 65 ksi. Maximum
horizontal deflection of the beam is limited to 0.03x20x12= 7.2 inches.
Pu All plates are 10”x½”

Pu
11”
20 ft

a a 10”

10 ft Cross Section a-a of steel Column

and Beam

Problem 8.8. Design the pole for the wind turbine shown below. Use cylindrical cross section and A36 steel.
Maximum horizontal deflection of the top of the pole is limited to 10 inches.
15 kips

20 kips

80 ft

Zomermag.com
homewindmillpower.net

400 kips
450 kips
Problem 8.9. Considering all
applicable failure modes, is the 10 kips
beam column adequate to carry
the factored loads shown? Note
W14x211,
that in both directions, the column W14x211, A992 Gr. 50
is fixed at the base and free at the A992 Gr. 50 Steel
top. Maximum horizontal 14 ft Steel 10 kips
movement of the top of the
column can be obtained from 2@7 ft
equations for deflection of a
cantilever.

Fixed Base Fixed Base

(A)
(B)
CHAPTER 8 – BEAM-COLUMNS 8-23

Problem 8.10. Do problem 8.9 but consider axial load to be tension load.
Problem 8.11. Calculate load P u that can be resisted by the structure shown. Use alignment charts to
establish K factors in the plane of the frame. For out of plane buckling use effective length factor K=1.0.
Notice that the column orientations are shown under each column. All beams are W21x57 and all columns
are W10x45. Steel is A992 Gr. 50. You can use approximate methods to analyze the indeterminate
structures. Maximum horizontal deflection of the beams is limited to 0.02x14x12= 3.3 inches.

Pu Pu Pu
Pu Pu Pu
Pu Pu
Pu/2
Pu/2
14 ft

2@18 ft= 36 ft 2@16 ft= 32 ft

(A) (B)
3Pu

Pu
Pu
Pu/2
Pu/2

14 ft

2@18 ft= 36 ft

(C)

Problem 8.12. Check if the frame shown in the figure is able to carry the loads. All loads are factored
loads. All steel is A 992 Grade 50. All members (beam and two columns) are W10x112. Lateral bracing is
provided only at A, B, C, and E. Maximum horizontal deflection of the beam is limited to 0.02x10x12= 2.4
inches.
10 ft
60 kips
C
B
Pin 45 degrees

15 kips

2 kips D 10 ft

5 ft

A E

Problem 8.13. Calculate maximum factored load P u that can be applied to the structure shown below. Beam
is a W24x104 and each column is 2W12x120. Steel is A992 Grade 50. Maximum horizontal deflection of the
beam is limited to 0.02x10x12= 2.4 inches.

3@6 ft= 18 ft

Pu Pu Pu
Pu
B

C D

Pu Column Section
E 10 ft
5 ft

Pin F
A
Roller
(A)

2Pu
2Pu
3@6 ft= 18 ft

B C Pu Pu

D F
Pu E

3@5 ft
E
Column Section

G
A Roller
Pin

(B)

10 ft
Problem 8.14. Check if the frame 60 kips 60 kips
in the figure is able to carry the
Pin
shown loads. All loads are factored B
2 kips D
loads. All steel is A992 Gr. 50. All
beams and columns are W10x112. o
C 45
Lateral bracing is provided only at
A, B, C, D and E. Maximum 5 kips
horizontal deflection of the beam is
limited to 0.02x10x12= 2.4 inches.
10 ft
5 ft

A E
CHAPTER 8 – BEAM-COLUMNS 8-25

Problem 8.15. The frame shown

in the figure is subjected to
vertical and lateral loads.
Calculate maximum value of P u . Pu 1.5Pu
1.5Pu B
Beam section is a W27x114 and
both columns are W12x87.
Lateral bracing is provided at A,
B, C, D, E, F, G and K. All steel A C D E
is A992, Grade 50. Maximum 9 ft
horizontal deflection of the beam Roller
2 Pu
is limited to 0.02x18x12= 4.3 Support
inches.
F
9 ft
Pin K
G

5ft 2@ 9 ft= 18 ft 5ft

CHAPTER 9

WELDS

Objectives:
In this chapter, you will learn:
1. Types of welds
2. Failure mode (or limit state) of welds.
3. How to design welds.

9.1. INTRODUCTION
Welds, bolts and rivets are the three connectors used in steel structures. This
chapter discusses the behavior and design of welds. Bolts and rivets are in Chapter 10.
Provisions of the latest edition of the Structural Welding Code of the American
Welding Society (ANSI/AWS, Latest Edition) D1.1 governs design of the welds in
buildings. The portions of the AWS code, which are directly applicable to steel building
structures, are included in the AISC Specification (AISC, 2010a) and AISC Seismic
Provisions (AISC, 2010b) as applicable. Provisions from the AWS code applicable to
bridges are in the AASHTO Bridge Specifications (AASHTO, 2011). In this chapter, we
will discuss design of welds and those provisions of the AISC Specification that apply to
welds used in building structures.
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-2

9.2. TYPES OF WELDS

Four different types of welds, shown in Figure 9.1, are:

a. Complete Joint Penetration Groove Welds.

b. Partial Joint Penetration Groove Welds.
c. Fillet Welds.
d. Plug or Slot welds.

Base Metal 1 Plug Weld

PJP Weld
CJP Weld
Base
Slot Weld
Base Base Metal 2
Base
Base Metal 2 Metal 1 Metal 2
Metal 1

Fillet Base
Base
Welds Metal 2
No welds here Metal 1

a. Complete Joint b. Partial Joint c. Fillet Weld d. Plug and Slot

penetration (CJP) Penetration (PJP) Welds
Groove Weld Groove Weld

Figure 9.1. Four Common Types of Welds Used in Steel Structures

9.2.a. Combining Welds

The above weld types can be combined and used in a single joint. The
strength of such joints is the sum of the strength of each individual type of weld
used in the connection.

9.2.b. General Behavior of Welds under Monotonic Static Loads

Weld lines subjected to normal or shear forces develop a very complex state of
stress. This is due to the irregular geometry of the weld and adjacent heat-affected areas as
well as a complex state of residual stresses trapped in the weld area. The residual stresses
are primarily due to very high temperatures of welding and relatively rapid cooling of the
steel (i.e. base metal) and the weld after welding.
When a weld line is subjected to normal or shear forces, initially it responds
elastically until maximum stresses in the weld or in the adjacent base metal reach yield
point and yielding starts. Due to presence of non-uniform residual stresses within the weld
and the heat-affected base metal, yielding continues to spread to other areas of the weld
CHAPTER 9 – WELDS 9-3

and/or heat affected zone. In the meantime, some areas of the weld or base metal that had
yielded earlier continue to strain-harden. As the loading continues to increase, due to the
presence of concentrated strains at the tip of the micro-cracks, a crack propagates and weld
or heat affected zone of the base metal fractures. Tests of groove and fillet welds indicate
that ductility of weld line is related to the type of applied stress, the type of electrode (weld
filler material) used and the failure mode that governs. Welded joints have two failure
modes, as shown in Figure 9.2, which are:
1. Failure of the base metal due to yielding or rupture (fracture), and;
2. Rupture of the weld.

Yielded Yielded
Areas
Areas

1. Yielding of Base Metal in 1. Yielding of Base Metal in Fillet Weld

Groove Weld

Fractured Area Fractured Area

2. Fracture of Base Metal in 2. Fracture of Base Metal in Fillet Weld

Groove Weld

Fracture Fracture Surface

Surface

Fracture Surface
3. Fracture of Groove Weld 3. Fracture of Fillet Weld

Groove Weld (CJP) Fillet Weld

Figure 9.2. Three Failure Modes of Typical Welded Joints

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-4

9.3. DESIGN OF WELDS

(Ref.: AISC Spec., (AISC, 2010a), Section J, Page 16.1-108))

Design of welds in general should consider failure of the weld as well as failure of
the base metal, shown in Figure 9.2 on previous page. The failure modes or limit states of
base metal are yielding or fracture of base metal. For welds, only one failure mode or limit
state is considered and that is fracture of the weld. Yielding of the weld is not considered a
failure mode, since welds usually have relatively small volume compared to the base metal,
and yielding of weld does not create large deformations. The small deformations of in the
welds, due to yielding, are not of concern and are ignored in design and analysis.
Obviously, if the weld ruptures it looses its strength and stiffness, therefore, weld rupture is
the only failure mode considered for the weld itself.
The design equations given by the AISC Specification (AISC, 2010) for the three
failure modes shown in Figure 9.2, in LRFD format, are:

R u ≤ φ F nBM A BM (LRFD format for base metal yielding or rupture) (9.1)

R u ≤ φ w F nw A we (LRFD format for weld rupture) (9.2)

Where
A BM = cross sectional area of the base metal, given in the following sections for each type
of weld: Complete Joint Penetration (CJP), Partial Joint Penetration (PJP), Fillet,
and Plug and Slot welds.
A we = effective throat area of the weld given in the following sections for each type of
weld: CJP, PJP, fillet, and plug and slot welds
F nBM = nominal stress of the base metal given in the following sections for each type of
weld: CJP, PJP, fillet and plug and slot welds
F nw = nominal stress of welds (welds are also called filler metals) given in the following
sections for each type of weld: CJP, PJP, fillet, and plug and slot welds
Ru = factored load in LRFD applied to the “effective throat area” of the weld
φ = resistance reduction factor in LRFD method for base metal given in the following
sections for four types of welds (Complete Joint Penetration, Partial Joint
Penetration, Fillet Welds, and Plug and Slot welds).
φ w = resistance reduction factor in the LRFD method for fracture of the welds given in
the following sections for four types of welds (Complete Joint Penetration, Partial
Joint Penetration, Fillet Welds, and Plug and Slot Welds).

9.4. BEHAVIOR AND DESIGN OF COMPLETE-JOINT-

PENETRATION (CJP) GROOVE WELDS
9.4.a. General Information on Complete-Joint-Penetration (CJP)
Groove Welds
When two pieces of steel are placed next to each other with a groove cut in one or
both pieces and filled with weld material, a groove weld results. If the depth of the groove
is equal to the thickness of the thiner plate, the weld is a “complete joint penetraion (CJP)”
CHAPTER 9 – WELDS 9-5

weld. Several choices of geometries of groove are permitted in complete-joint-penetration,

AWS (2009).

9.4.b. Effective Area of Complete Joint Penetration (CJP) Welds, A we

The effective area of a complete joint penetration (CJP) weld, A we , is the effective length
times the effective throat thickness(AISC, 2010). The effective length is the smaller width
of the parts joined as shown in Figure 9.3. The effective throat thickness is defined as the
thickness of the thinner plate.

Effective Length
of the Weld

Weld axis

Effective Throat Thickness is the thickness of

the thinner plate

Figure 9.3. Effective Length and Effective Throat Thickness for

Complete-Joint-Penetration (CJP) Welds

9.4.c. Area of Base Metal, A BM

The area of base metal, A BM , to be used in design Equation 9.1 for CJP welds is the same
as the effective area of the weld, A we defined in previous section , which is effective length
of weld times effective throat thickness, see Figure 9.3 above.

9.4.d. Weld Metal Requirements for Complete-Joint-Penetration

Groove Welds
If the applied load to a complete-joint-penetration groove weld is tension normal to
the effective area, the AISC Specification (AISC, 2010), requires that the weld electrode be
a matching welds, and satisfies the following requirements. Using matching welds, results
in fracture strength of weld to be greater than the strength of the welded base metal. The
requirements for matching welds are:
1. The electrode used in complete-joint-penetration groove welds subjected to tension
normal to effective area of the weld, should comply with the requirements of
“matching” weld metals as given in AWS-D1.1. The AWS-D1.1 provisions on the
“matching” weld metals are discussed below in Section 9.4.e
2. The weld metal of a specified Charpy V-notch (CVN) toughness of 20 ft-lbs (27
Joules) at 40oF (4oC ) is used for a complete-joint-penetration groove welds in Tee
joint and corner joints where the backing bar is left in place and the weld is
subjected to tension normal to its effective weld area. If the weld metal does not
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-6

have this specified CVN, then the groove weld should be designed as partial-joint-
penetration (PJP) groove weld.
3. For splices of heavy sections with flanges thicker than two inches where the weld
is subjected to tension normal to its effective area, a weld metal of specified
Charpy V-notch (CVN) toughness of 20 ft-lbs (27 Joules) at 40oF (4oC ) is used.

9.4.e. Matching Weld as Defined by the AWS-D1.1

In order to ensure that the welds in complete-joint penetration welds will not fail
prior to the failure of the connected steel, the AWS D1.1 has defined “matching” welds.
The following table summarizes the “matching weld” for steels and electrodes commonly
used in steel construction. For more detailed table see AISC Specification (AISC, 2010,
Chapter J).

Table 9.1 Basic Information on Matching Welds

Base Metal Matching Filler Metal (Electrode)

A36 Steel, with thickness ≤ ¾ in. 60 and 70 ksi electrodes

A36 Steel, with thickness > ¾ in.

70 ksi electrodes
A572, Gr. 50 and 55, A913, Gr. 50, A588, A992

A913, Gr. 60 and 65 80 ksi electrodes

Notes: (a) This table provides basic information on “matching welds. For information, that is
more detailed see AISC Specification.
(b) In joints with base metals of different strengths we should use a filler metal that
matches the higher strength base metal.

9.4.f. Design of Complete-Joint-Penetration (CJP) Groove Welds

The design equations for design of complete joint penetration (CJP) welds are the
same as given earlier by Equations 9.1 and 9.2 (in LRFD) for yielding or fracture of base
metal and rupture of the welds. However, in design of complete-joint penetration (CJP)
groove welds, since “matching welds” are used, the CJP weld is stronger than the base
metal and as a result only failure modes of base metal are considered.
Therefore, the equation of design for Complete Joint Penetration (CJP) welds is the
equation for base metal given earlier as Equations 9.1 and repeated below for convenience.
The terms in the equation were defined earlier on Page 9-4.

R u ≤ φ F nBM A BM (9.1 repeated)

Depending on the direction of applied force, specific values of φ F nBM , the

nominal stress of the base metal in above equation is specified by the AISC Specification
(2010) in its Table J2.5. Table 9.2 on the next page shows the values of φ and F nBM from
the AISC Table J2.5 for complete joint penetration (CJP) welds.
CHAPTER 9 – WELDS 9-7

Table 9.2. Design Strength of Complete Joint Penetration (CJP) Groove Welds

Design Parameters for Complete Joint-Penetration Groove Welds

Limit State to be Nominal
Load Type and Direction Relative to Checked Resistance
Weld Axis Stress Comments on Filler Metal
Factor
F nw or (i.e. Weld Electrode)
φ or φ w
F nBM

weld axis Base Metal, φ = 0.90

F nBM = F y “Matching” Filler Metal should
Tension Yielding be used. For T and corner
joints with backing left in place,
Base Metal, F nBM = notch tough filler metal is
φ = 0.75 required (see Sec. J2.6 of
Tension Rupture Fu
AISC Spec.). [a]
Tension
normal to weld axis No need to be checked since
Weld Rupture
“matching welds” are used
Base Metal, Filler Metal with a strength
Buckling in φ = 0.90 F nBM = F cr level equal to or one strength
Comp. level less than matching filler
metal is permitted.
Compression Weld Rupture No need to be checked since
normal to weld axis “matching welds” are used

Filler Metal with a strength

level equal to or less than
matching filler metal is
permitted.

Tension or compression in parts joined parallel to

a weld need not be considered in design of welds
joining the parts.

Tension or
compression
parallel to weld axis

Base Metal, φ =1.0 “Matching” Filler Metal should

F nBM = 0.60F y be used.
Shear Yielding

Base Metal,
φ = 0.75
Shear F nBM = 0.6F u
Rupture

No need to be
checked since
“matching
Shear parallel or normal Weld Rupture welds” are
to weld axis used.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-8

9.5. BEHAVIOR AND DESIGN OF PARTIAL-JOINT-

PENETRATION (PJP) GROOVE WELDS
9.5.a. General Information on Partial-Joint-Penetration Groove Welds
Similar to complete-joint-penetration groove welds, in partial-joint-penetration
groove welds, two pieces of steel are placed next to each other with a groove between them
where the weld material is deposited. In partial-joint-penetration groove welds, depth of
the groove is less than the complete depth of the thiner piece being connected.

9.5.b. Behavior of Partial-Joint-Penetration Groove Welds

Figure 9.4 shows the stress path inside a typical complete-joint-penetration (CJP)-
and a partial-joint-penetration (PJP) weld. For PJP welds, due to the presence of opening,
stress concentrations are expected to exist as shown in the figure. The stress
concentrations, which are at the notch root of the weld, may result in initiation of cracks in
the weld or heat-affected zone of the base metal. As a result, PJP welds are not expected to
be as ductile as CJP welds.
Opening
(not welded)

Stress
concentration
Stress Lines Stress Distribution at the notch
root of the
Partial-Joint-Penetration Groove Weld (PJP) weld

Stress Lines Stress Distribution

Complete-Joint-Penetration Groove Weld (CJP)

Figure 9.4. Stress Concentration in a Partial-Joint-Penetration Weld

9.5.c. Effective Throat Area, A we , of Partial Joint Penetration Welds

The effective throat area of partial-joint-penetration groove welds (PJP) depends
on welding procedure, type of welding and shape of the groove as given in Table J2.1 of
the AISC Specification (AISC, 2010) given on the next page as Table 9.3.

9.5.d. Design of Partial-Joint-Penetration (PJP)Groove Welds

In design of partial-joint-penetration groove welds, we need to consider three
failure modes (or limit states) which are rupture of the weld and yielding in the base metal
CHAPTER 9 – WELDS 9-9

and rupture in the base metal. Design equations for PJP are the same as given earlier by
Equations 9.1 and 9.2. Table 9.4 below provides values of design parameters used in these
equations when applied to Partial Joint Penetration (PJP) groove welds.
For partial-joint-penetration groove welds, design codes provide a minimum
effective throat thickness based on thickness of the thicker plate joined, Figure 9.5. Table
9.5 on page 9-11 from the AISC Specification (AISC, 2010a) shows minimum effective
throat thickness for partial-joint-penetration groove welds. The AISC Specification (AISC,
2010) also adds that the minimum effective throat thickness does not need to be greater
than the thinner of the two plates joined even when a larger size is required by the
calculated strength.
Weld axis
Effective Length
of Weld

Weld axis

Effective Length Depth of

of Weld Depth of Groove
Groove
Figure 9.5. Effective Lengths and Depth of Chamfer for
Partial-Joint-Penetration Groove Welds

Table 9.3- Effective Throat Thickness for PJP Welds

V, U, and J weld configurations:

V-groove welds U-groove welds J-groove welds

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-10

Table 9.4. Design Strength of Partial Joint Penetration (PJP) Groove Welds

Design Parameters for Partial Joint-Penetration Groove Welds

Limit State to
Load Type and Direction Relative to be Checked Resistance Nominal Stress Comments on
Weld Axis Factor F nBM or F nw Filler Metal
φ or φ w (i.e. Electrode)

Base Metal, φ = 0.90

weld Tension F nBM = F y Filler Metal with a
axis Yielding strength level
Base Metal, equal to or one
φ = 0.75 strength level less
Tension F nBM = F u
Rupture than matching
Tension filler metal is
φ w = 0.80
normal to weld axis Weld Rupture F nw = 0.60F Exx permitted.

Compression
Column to base plates and column Compressive stress need not be considered in
design of welds joining parts --
splices designed per Sec. J1.4(a) of the
AISC Spec., 2010

Compression Base Metal, Filler Metal with a

Comp. Yielding φ = 0.90 F nBM = F y
Connections of members designed to strength level
bear other than columns as described in equal to or one
Base Metal, φ = 0.75
Sec. J1.4(b) of the AISC Spec., 2010. Comp. Failure
F nBM = F u strength level less
than matching
filler metal is
Weld Rupture φ w = 0.80 F nw = 0.60F Exx
permitted.
Base Metal,
Weld axis Comp. φ BM = 0.90 F nBM = F y Filler Metal with a
Yielding strength level
equal to or one
Base Metal, φ BM = 0.75 strength level less
Comp. Failure
F nBM = F u
Compression than matching
Connections not filler metal is
finished-to-bear Weld Rupture φ w = 0.80 F nw = 0.90F Exx permitted.

Weld Base metal needs to resist the Filler Metal with a

Base Metal
tension and compression stresses strength level
axis Yielding and
that are imposed on the base metal equal to or one
Rupture
being part of the cross section.
strength level less
than matching
Tension or compression in parts filler metal is
joined parallel to a weld need not be
Tension or compression Weld Rupture
considered in design of welds joining
permitted.
parallel to weld axis the parts.

Weld axis Weld axis Filler Metal with a

Base Metal,
Shear Yielding φ = 1.00 F nBM = 0.60F y strength level
equal to or one
strength level less
than matching
Base Metal,
Shear Rupture φ = 0.75 F nBM = 0.60F u filler metal is
permitted.

Shear parallel or normal Weld Rupture φ w = 0.75 F nw = 0.60F Exx

to weld axis
CHAPTER 9 – WELDS 9-11

Table 9.5

Note: This table is from the AISC Specification (AISC, 2010). Copyright AISC.

9.6. FILLET WELDS

When two pieces of steel are placed next to each other under an angle (usually 90-
degrees) and the corner is welded, a fillet weld results. Figure 9.6 shows typical fillet
welded joints. Weld Axis
Base Metal -2

Base Metal -2

Weld Axis
Fillet
Weld Base Metal -1
Base Metal -1 Fillet Weld

Weld Axis Weld Axis Weld Axis Weld Axis

(a) (b) (c) (d)

Weld Axis Weld Axis Weld Axis Weld Axis

(e) (f) (g) (h)

Figure 9.6. Typical Fillet Welds
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-12

9.6.a. Behavior of Fillet Welds

Behavior of fillet welds depends on the direction of the applied load. If the applied
load is parallel to the weld axis , which is the longitudinal axis of the weld, Figures 9.6(c)
and 9.6(f), the weld will behave in a weaker but more ductile manner than when the
direction of applied load is perpendicular to the weld, Figures 9.6(a), (b), (d), (e), (g), and
(h). Figure 9.7 shows force deformation behavior of fillet welds with various angles
between the weld line and line of action of the force.

Weld Axis Force Perpendicular to

Weld Axis

Force

Force Parallel to Weld Axis

Deformation

Figure 9.7. Schematic Behavior of Fillet Welds Subjected to Force

9.6.b. Effective Throat Area, A we , of Fillet Welds

The effective area of a fillet weld is defined in Section 2.3 of the AWS Code
(AWS, 2009). In general, the length of a fillet weld is the total length measured along the
centerline of its throat thickness, Figure 9.8(a). The effective throat thickness for fillet
welds is the shortest distance from the joint root to the weld face, Figure 9.8(b). Again, for
more information on the throat thickness of fillet welds, see the AWS Code (AWS, 2006).

Effective
Throat of
Fillet Weld

Effective Length Weld

of Fillet Weld leg size
Weld
Face
Joint Root
(a) Fillet Weld Weld
leg size

(b) Effect Throat

Figure 9.8 Effective Length of Fillet Welds

CHAPTER 9 – WELDS 9-13

9.6.c. Effective Length of Fillet Welds

The minimum effective length of the fillet welds, designed based on strength,
should not be less than four times the size of the weld, or else the size of the weld should
be considered not to exceed 1/4 of the length of the weld.

Approximate stress
distribution in the weld w End loading

Figure 9.9. Approximate Stress Distribution in Long “End-Loaded” Welds

For end-loaded fillet welds, Figure 9.9, the stress distribution along the length is
not uniform. The AISC Specification (AISC , 2010) has introduced a limitation on the
maximum effective length of the end-loaded fillet welds. If the actual length of fillet weld
exceeds 100 times the leg size of the weld, w, the effective length of the weld should be
calculated as:
L eff –w = β L (If L > 100w) (9.5)
Where
L eff-w = effective length of the end–loaded fillet weld
L = actual length of the end-loaded fillet weld
β = a reduction factor given as:

β = 1.2 - 0.002 (L/w) ≤ 1.0 (9.6)

w = size of weld leg

If L/w of the end-loaded fillet weld is greater than 300, a reduction factor β equal
to 0.60 should be used.

9.6.d. Minimum Size of Fillet Weld

Minimum size of fillet welds should be according to Table J2.4 of the AISC
Specification given below as Table 9.6.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-14

Table 9.6

9.6.e. Maximum Size of Fillet Weld

AISC Specification (2010), has the following statement on maximum size of fillet
welds:

9.6.f. Limitations on the Geometry of the Fillet Welds

There are limitations on the geometry of fillet welds. For details of this limitation,
see Page 16.1-111 of the AISC Specification (2010).

9.6.g. Design of Fillet Welds

The failure modes of weld and base metal in a fillet-welded connection are shown in
Figure 9.10 for a typical “T” joint and are:
1. Fracture of weld through its effective throat area, shown in Figure 9.10(b), under
a combined and somewhat complex tension/shear effects
2. Yielding in the gross area of base metal in tension or shear as shown in Figure
9.11(c) for the case of tension yielding, and;
3. Rupture in the net area of base metal in shear or tension, as shown in Figure
9.11(d) for the case of tension fracture.
4. In some fillet weld connections, in addition to failure modes of yielding and
rupture of the base metal, block shear rupture of the connection also needs to be
checked.
CHAPTER 9 – WELDS 9-15

The design equations for fillet welds are the same as given earlier by Equations 9.1
and 9.2. Table 9.7 below provides values of design parameters used in these equations.

Fillet
Weld Base Metal -2
(e.g. Beam flange, Tension-Shear Rupture of
Base Metal -1 Weld Effective Throat
column flange or column
(e.g. gusset plate)
web)

(a) A Typical Fillet Weld (b) Fracture of Weld

Yielding of Gross Area of Fracture of Effective Net

the Base Metal in Area of Base Metal in
Connected Element Connected Element

(c) Yielding of Gross Area of Base Metal (d) Fracture of Effective Net Area of
Base metal
(See Chapter 4 for effective net area
of base metal)

Figure 9.10. Failure Modes of a Typical Fillet Weld

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-16

Table 9.7 Design Strength of Fillet Welds

Design Parameters for Fillet Welds Subjected to Single Load Effect

Resistance Nominal
Limit State Factor Stress
Load Type and Direction Relative Comment on Filler Metal
to be φ F nBM
to Weld Axis (i.e. Electrode)
Checked Or Or
φw F nw

In all cases, weld is considered Base Metal, “Filler metal with a strength
subjected to shear. Examples of Tension or φ = 0.90 [a] level equal or less than
loading: Compression F nBM = F y matching filler metal is
Weld Axis Yielding
permitted.”
Base Metal,
Tension or φ = 0.75
F nBM = F u
Compression
Rupture
Base Metal, φ = 1.00
Shear F nBM = 0.60F y
Yielding

Base Metal, φ = 0.75

Shear F nBM = 0.60F u
Rupture

φ w = 0.75 [b]
Weld Rupture
F nw = 0.60F Exx

Block Shear Rupture-- needs to be checked

according to provisions of Section J.4 of the AISC
Specification. Chapter 4 of this document provides
more information on block shear rupture.

Tension or compression parallel to Tension and Compressive strength need not

the weld axis: be considered in design of welds.

Base metal needs to resist the tension and

compression stresses that are imposed on the
base metal being part of the cross section.

Weld Axis

Footnotes:
[a] The area that the stress should be multiplied to obtain the nominal resistance is defined by the AISC Specification (AISC, 2010) as:
“The effective area of a fillet weld shall be effective length multiplied by the effective throat.” For information on effective length,
effective throat and minimum size of fillet welds refer to Sections J2a, J2b, and Table J2.4 of the AISC Specification.
[b] AISC Specification (AISC, 2010) states that: “ Alternatively, the provisions of J2.4(a) are permitted provided the deformation
compatibility of the various weld elements in considered. Alternatively, Sections J2.4(b) and (c) (of the Specification) are special
applications of J2.4(a) that provide for deformation compatibility.
CHAPTER 9 – WELDS 9-17

References for Chapter 9

1. AISC, (2010a) “Specification for Structural Steel Buildings”, ANSI/AISC 360-10, An
American National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings”, ANSI/AISC 341-10,
An American National Standard, American Institute of Steel Construction, Chicago.
3. AISC, (2006). “AISC Design Guide 21-Welded Connections-A Primer for Engineers” ,
American Institute of Steel Construction, Chicago, pp. 151.
4. AISC, (2011) “Steel Construction Manual”, 14th Edition, American Institute of Steel
Construction, Chicago.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-18

Problems for Chapter 9

Problem 9.1. Check the Complete-Joint-Penetration (CJP) welds and establish if their design is
adequate. All loads are factored loads. All filler metals (electrodes) are E70xx.

PL 1”x 10” PL 1” x 8” PL 1”x 10” PL 3/4” x 8”

A36 steel A36 steel A36 steel A572 Gr. 50 steel

220 kips 220 kips 410 kips 410 kips

(B)
(A)

PL 1”x 8” PL 3/4” x 8” PL 1”x 8”

A36 steel A572 Gr. 50 steel PL 1” x 8”
A36 steel
A36 steel

175 kips 175 kips

200 kips
200 kips

Problem 9.2. Design the Partial Joint Penetration groove welds in the connections shown below.
The electrode is E70xx. Steel material is indicated in the figure for each component. The loads
shown are factored loads.

PL 1”x 10” PL 3/4” x 8” PL 1”x 10” PL 3/4” x 8”

A36 steel A36 steel A36 steel A572 Gr. 50 steel

200 kips 200 kips

200 kips
200 kips

S1=? 60o bevel V-

S1=? 60o bevel V-
S1=? groove weld
groove weld (PJP)
S1=? (PJP)

(B)
(A)
CHAPTER 9 – WELDS 9-19

Problem 9.3. Find maximum load carrying capacity (φR n ) of the connection shown below. Check
all failure modes including failure modes of the bolts and welds. All welds are 3/8” fillet welds with
E70 electrodes.

1” plate, A36 HSS6x6x1/2” A500 Gr.B

φRn φRn

3/8”
3/8”

Problem 9.4. Design the fillet welds for connections shown below. Electrode is E70xx. Steel
material is indicated in the figure for each component. The loads shown are factored loads. In this
case, you need to select the length and size of weld leg. Notice that in some of the cases below, in
addition to checking yielding and rupture of the base metals as required by the weld design, you
need to check block shear failure of the connections as well.

PL 1-1/2x 12” PL 1”x 10” 2PL 1/2” x 10”

2PL 3/4” x 10”
A36 steel A36 steel A572, Gr. 50
A36 steel
steel
380 kips

380 kips 220 kips 220 kips

410 kips
205 kips 240 kips 120 kips

205 kips 120 kips

(A) (B)

Plate 2” x22”
A36 steel 2C15x50
A572 Gr. 50
500 kips 500 kips

1100 kips
1100 kips

2MC18x58 Channel, A36 7”

5/16”
PL 20x1-1/2, A36 steel

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
9-20

Problem 9.5. Design the fillet welds shown below. Select the electrode. Steel material is indicated
on the figure for each component. Loads shown are factored loads.

PL12x1-¼ , A36
WT15x146, A913 Gr. 65

2L6x6x1/2”, A36 Steel 500 500

kips kips

300 kips 2L8x4x1/2, A36

Long Legs Back to Back
½” Fillet Weld, (B)
Design the Required Length on each
side
(A)

PL1 x12, A36

PL 3/4x16, A36
2C15x50, A36
Round HSS 10x0.625
A500 Gr. B.

430 kips 430 kips

500 kips 500 kips

CHAPTER 10

BOLTS

Objectives:
In this chapter, you will learn:
1. What is a structural bolt?
2. Types of bolts and bolt holes
3. Failure modes (or limit states) of bolts, which are rupture under
tension, shear and combination of shear and tension.
4. How to design bolts.
5. How to apply what you have learned to actual design and
evaluation of bolted connections using numerical examples
provided in this chapter.

10.1. INTRODUCTION
This chapter discusses behavior and design of bolts used in connections of steel structures. Welds
were discussed in Chapter 9. Design of fasteners (bolts and threaded parts) is governed by the
provisions of the “Specification for Structural Joints Using High-Strength Bolts” (RCSC, 2011).
The RCSC Specification is developed and maintained by the Research Council on Structural
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-2

Connections (www.boltcouncil.org). A copy of this specification is attached to the AISC Manual

(AISC, 2011) and can be downloaded, free of charge, from the www.boltcouncil.org. In addition,
the book: “Guide to Design Criteria for Bolted and Riveted Joints” by Kulak et al. (2001),
provides extensive and very valuable information on the behavior and design of fasteners in
particular bolts. Part of Chapter J of the AISC Specification (AISC, 2010) is devoted to design of
fasteners (primarily bolts) and bolted joints. In the remainder of this chapter, we will focus on the
basic design of fasteners (bolts and threaded parts) that is included in Chapter J of the AISC
Specification (2010a).

10.1.a. Types of Fasteners

Fasteners, shown in Figure 10.1 are commonly used in steel structures and included in the AISC
Specification (2010), are:
a. A307 grade A carbon steel bolt primarily used in low-rise buildings and for static loads.
b. A325 high strength steel bolt, which is currently the most common bolt used in steel
structures.
c. A490 high strength quenched and tempered alloy steel bolt, which is also used frequently
in structures today.
d. Threaded parts made of A36, A193, A354, A572, A588, and A F1554 steels generally
used in connecting steel to concrete such as anchor bolts, or in connections that require
dimensions for fasteners that are outside the standard bolt dimensions.

A307 Bolt A325 and A490 High Strength Threaded Rod

Bolts

Figure 10.1. Common Types of Fasteners

In order to use fasteners to connect pieces of metals, holes are drilled or punched in all
parts to be connected. To facilitate the placement of the fasteners in the holes, the diameter of the
hole is usually greater than the nominal diameter of the corresponding fastener. The next step is to
bring the parts together, place the bolts in the hole, and tighten the nuts.
CHAPTER 10 – BOLTS 10-3

10.1.b. Types of Bolt Holes

Typical bolt holes are shown in Figure 10.2 with the most common type being the standard hole,
where the diameter of the hole is 1/16 inch larger than the nominal diameter of the bolt. Types of
bolt holes are:
1. Standard Hole- these are most common types of the bolt holes and consist of drilled or punched
round holes. The diameter of the hole is 1/16” greater than the nominal
diameter of the fastener.
2. Oversized Hole- these are round drilled or punched holes with a diameter of about ¼” greater
than the nominal diameter of the fasteners. Oversized holes are permitted for
slip critical bolts but not for bearing bolts.
3. Short-Slotted Hole- these are “elongated” holes which are permitted for both slip-critical and
bearing bolts. However, in bearing bolts, the length of the slot should be
perpendicular to the applied load.
4. Long-Slotted Hole- these are also elongated holes but the length of the hole is much greater
than the width and is equal to 2.5 times diameter of the fastener while the width
of the hole is diameter of the fastener plus 1/16 inch. Long slotted holes are
commonly used in expansion joints or wherever free movement of the fastener
in one direction is desired. Long-slotted holes are permitted for both slip-
critical and bearing bolts. However, in bearing bolts, the length of the slot
should be perpendicular to the applied load.
Table 10.1 below, which is Table J3.3. of the AISC Specification (2010) shows
sizes of different types of holes for a given bolt diameter.
Bolts and threaded parts are subjected to tension, Figure 10.3(a), shear, Figure 10.3(b), or
combination of shear and tension.

Standard Oversized

Short-slotted Long-slotted

Figure 10.2. Types of Bolt Holes used in Steel Structures

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-4

Table 10.1
(of the AISC Spec. (2010) )

(of the AISC Spec. (2010) )

(Note: This table is from the AISC Specification (2010), Copyright AISC. Table is given here for
educational purposes. For actual design, the AISC Specification should be used.

Bolts in Tension
Bolts in Shear

(a) Tee Connection (b) Lap Connection

Figure 10.3. Bolts Subjected to Tension or Shear

10.1.c. Bolts Pretension

To force the connected plate to come into contact and stay in contact, bolts are tightened.
In addition to this construction requirement for bolt tightening, when bolts are not tightened, under
a very small applied shear, since there is a gap between the bolt and its hole, first the bolt will slip
until it bears against the material. Such slippage, if occurs under service load, can impair the
functionality of the connection. Particularly, when connection is subjected to load reversal such as
wind where, back-and-forth slippage can result in damage to non-structural elements and larger
drift and P-∆ effects. Until the 1980’s, in almost all applications of A325 and A490 bolts, these
high strength bolts were tightened to develop a code-specified pre-tension force in the bolt. The
idea was that the pre-tension force in the bolt would clamp the connected pieces together, creating
friction force in the faying surfaces. The friction force will then prevent slippage of the bolt under
shear force due to service loads.
The amount of pre-tensioning force in a bolt is about 70% of the tension capacity of the
bolt. The precise value of the pre-tensioning force for each bolt diameter is in Table J3.1 of the
AISC Specification (AISC, 2010) reproduced below as Table 10.2.
CHAPTER 10 – BOLTS 10-5

Table 10.2

(of the AISC Spec. (2010) )

(Note: This table is from the AISC Specification (2010), Copyright AISC. Table is given here for
educational purposes. For actual design, the AISC Specification should be used.

During 1980’s, cyclic and monotonic tests of bolted connections indicated that in some
applications, the so called “snug-tight” bolts, which had very little tightening, could perform
satisfactorily. An example of application of snug-tight bolts is shear tab connections used as
supports for simply supported beams. In connections of simply-supported beams, it is desirable
that under applied load, the ends of the beam be free to rotate and behave more or less as a pin
connection. If the bolts at the end of the simply-supported beams do not slip, bending moments
will be generated in the end connections of the beam. Such moments need to be transferred
through the connection to the columns. However, if connection bolts slip easily, the connection
will rotate and the amount of moment generated in the connection will be reduced and connection
will act closer to a pin.
The AISC Specification (2010) allows the use of “snug tight” bolts in cases that slippage
of bolt under service shear load is acceptable. For bolts in tension, both tightened and snug tight
bolts can be used.

10.1.d. Limit States of Bolts

Limit states of bolts are:
1. Tensile rupture under pure tension
2. Shear rupture under pure shear
3. Tension/shear rupture under combined tension and shear
4. Bearing failure of the connected material when bolts are subjected to shear or shear plus
tension.
In the following sections the above limit states are discussed and design eqiuations for
each limit state provided.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-6

10.2. DESIGN OF BOLTS SUBJECTED TO TENSION

Figure 10.4. shows typical behavior of bolts subjected to tension. The A307 bolt, being a
relatively low strength bolt, shows a yield plateau and relatively large ultimate elongation.

A490 (Group B) Bolt

Tension Load
A325 (Group A) Bolt

A307 Bolt

Elongation

Figure 10.4. Schematic Behavior of Bolts Subjected to Tension

Figure 10.5. Failure of Bolts Under Tension

The A325 and A490 high strength bolts show less but sufficient ductility. Bolts subjected
to tension are expected to fracture through their threaded areas as shown in Figure 10.5.
The only limit state or failure mode of bolts and threaded parts subjected to tension, is
tensile rupture of the bolt. The design equations for this limit state of bolts and threaded parts, in
LRFD , are:

R u ≤ φ F nt A b (in LRFD) (10.1)

CHAPTER 10 – BOLTS 10-7

where
R u = factored tension force applied to the bolt for use in LRFD methods
F nt = nominal tensile strength of bolt given in Table J3.2 of the AISC Spec. (2010), reproduced on
Page 10-9 as Table 10.3.
A b = nominal area of bolt calculated using nominal diameter.
φ = resistance factor for bolts= 0.75

10.3. BEHAVIOR AND DESIGN OF BOLTS

SUBJECTED TO SHEAR
10.3.a. Introduction
Figure 10.6 shows typical behavior of A325 and A490 bolts subjected to shear. The A325 bolt
shows larger ductility in shear than the A490 bolt. Notice that a portion of the shear deformation
shown in Figure 10.6 is due to slippage of the plates. The slippage depends on the gap between the
bolt hole and bolt shank.
Depending on whether the shear plane passes through the threaded or unthreaded area,
shown in Figure 10.7(a) and (b) respectively, the shear strength of a fastener will be different and
should be calculated based on the actual area that is being sheared. However, many design codes,
including the AISC Specification (2010), assume a constant area throughout the length of the
fastener regardless of the location of the shear plane. Then, to account for the reduced shear
capacity in the threaded areas, when threads are included in the shear plane, a reduced bolt
nominal strength, F nv , is used whenever threads are included in the shear plane.

Shear
Force

A490 Bolt

Slip

(b)

A325 Bolt

Shear Deformation

(a) (c)

Figure 10.6. Schematic Behavior of Bolt Assemblies Subjected to Shear

(Source: Kulak, Fisher, and Struik, 2001)

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-8

Table J3.2 (next page) has two values of F nv for two cases of “threads are not excluded
from shear planes” and “threads are excluded from shear planes”. Since presence of threads in the
shear plane affects design strength, the designer needs to specify what assumption is made in
design by indicating the type of the bolt used as A325X or A490X for threads excluded case and
A325N or A490N for threads not excluded case. The fabricator will then be responsible to use the
correct bolt with a length of the threads that satisfies the designer’s specified type.
For A307 bolts, the values given in Table 10.3 are for both cases of threads not excluded
and threads excluded from shear plane.
Footnote (b) to Table 10.3 below states that if bolted connection is end loaded and the
length of the connection, which is center-to-center distance of first and last bolts, is greater than 38
inches, the shear strength values given in Table 10.3 should be reduced to 83.3% of the values
given in the table.
“Section A3.4” that is referred to in the 8th and 9th row of Table 10.3 is the section in
Chapter A of the AISC Specifications (2010) , Page 16.1-9 and discusses the material of anchor
rods and threaded rods.

shear shear
plane plane

(a) Threads are not excluded (b) Threads are excluded

from shear plane from shear plane

Figure 10.7. (a) Threads are not excluded from shear plane, and;
(b) Threads are excluded from shear plane

Depending on the relative strength of the “fastener” and “connected part,” four failure
modes, shown in Figure 10.8, have been identified for bolts and threaded parts subjected to shear:
1. Shear rupture of bolts and threaded parts, Figure 10.8a.
2. Bearing failure in the connected parts in front of the bolt or threaded part due to
compression exerted on it by the bolt or threaded part shank, Figure 10.8.b.
3. Bolt spacing failure of the connected part in the areas between the two adjacent bolts,
Figure 10.8.c.
4. Edge distance failure of the connected part, Figure 10.8.d.
CHAPTER 10 – BOLTS 10-9

Table 10.3

(of the AISC Spec. (2010)

(Note: This table is from the AISC Specification (2010), Copyright AISC. Table is given here for
educational purposes. For actual design, the AISC Specification should be used.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-10

10.3.b. Limit States of Bolts Subjected to Shear

Edge Distance

b. Bearing Failure

Bolt Spacing

Edge Distance

c. Bolt Spacing Failure

a. Bolt Shear
d. Edge Distance Failure

Figure 10.8. Failure Modes of Fasteners in Shear

10.3.c. Shear Rupture of the Bolts

Design equation for failure mode of shear rupture of a bolt is:
R u ≤ φF nv A b (in LRFD) (10.2)

where
R u = factored shear force applied to the bolt for use in LRFD methods
F nv = nominal shear strength of the bolt given in Table J3.2 of the AISC Spec. (2010), see
Page 10-9.
A b = nominal area of the fastener calculated using nominal diameter.
φ = resistance factor for bolts = 0.75

10.3.d. Bearing Failure of the Connected Parts

When bolts are subjected to shear, depending on the amount of pre-tensioning force in the
bolt, at one point the bolts will slip and the slipped bolt will bear against the wall of its hole,
Figure 10.9(a). Until slippage occurs, transfer of shear is through friction at the faying surfaces,
Figure 10.9(b). After slippage, the primary shear transfer mechanism is by bearing of the bolt
against the wall of the hole as shown in Figure 10.9(c).
After slippage is complete and bolt is bearing against the wall, behavior of the joint
depends on relative strength of the bolt and connected part. For typical civil engineering
CHAPTER 10 – BOLTS 10-11

applications, the bolts are usually high strength, A325 or A490, and the connected part in some
occasions is high strength (F y = 70-90 ksi) and more often moderate strength (F y = 50-65 ksi) or
low strength (F y = 36 ksi). In all of these applications, high strength bolts are stronger than the
connected parts. As a result, when a bolt is bearing against the wall of the bolt hole, the bulk of
yielding will occur in the connected part while the bolt remains essentially elastic.

Shear C
Force

Shear is
A B transferred
by bearing
(b) Before Slippage
Slippage
of bolt
Shear is transferred
by friction

O Relative Displacement
Between the Plates

(a) Behavior under Shear Force

Figure 10.9. Schematic Bearing Behavior of Bolts Subjected to Shear

Bearing
Fracture

Bearing
Yielding

Figure 10.10. (a) Bearing Yielding , and; (b) Bearing Rupture

(Photos: Courtesy of www.AISC.org)

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-12

Considering Figure 10.9(a) above, from Point O to A behavior is elastic and shear force is
transferred by friction between connected plates, as shown in Figure 10.9(b). At Point A, the
applied shear force is equal to static friction force between the plates and bolt slippage starts. At
Point B, bolt slippage is completed and the bolt is in contact with the wall of the bolt hole. At this
point bearing of bolt against the hole also participates in transfer of force. From Point B to Point C
the shear force is resisted by friction and bearing with the latter resisting more and more shear
force, Figure 10.9(c). In design of bearing bolts, the contribution of friction in resisting shear
force is ignored and it is assumed that the shear force is transferred only by bearing of bolt against
the wall of the bolt hole, Figure 10.9(c).
In design, the ultimate shear strength of bolt, Point C in Figure 10.9(a) is used. Figure
10.10 shows yielding and fracture in the areas of connected plate adjacent to the bolt.
Design equation for bearing failure mode of the bolt holes is given as:
R u ≤ φR n (for LRFD) (10.3)
where
R u = Required bearing strength of the bolt hole (same as factored shear force applied to the bolt.
φ = Resistance factor for bolts = 0.75
R n = Nominal bearing strength of the bolt hole given as:
(a) For standard, oversized, short slotted holes irrespective of the direction of the shear and
for long-slotted holes where shear is parallel to slot,
When elongation of the hole under service load is a design consideration:
R n = 1.2 L c t F u ≤ 2.4 d t F u (10.4a)
When elongation of the hole is not a design consideration:
R n = 1.5 L c t F u ≤ 3.0 d t F u (10.4b)

(b) For long-slotted holes where shear is perpendicular to slot,

R n = 1.0 L c t F u ≤ 2.0 d t F u (10.4c)
where
L c = clear distance, in the direction of the force, between the edge of the hole and the edge of the
adjacent hole or edge of the material as shown in Figure 10.11.
t = thickness of the connected part.
F u = minimum specified ultimate strength of the steel (i.e. connected part)
d = nominal bolt diameter.

Lc Lc Lc Lc

Lc Lc Lc
Lc

Lc Lc Lc Lc

Figure 10.11. Values of Lc

CHAPTER 10 – BOLTS 10-13

10.3.e. Minimum Bolt Spacing

As Figure 10.8(c) showed earlier , one of the failure modes of the bolts subjected to shear is failure
of bolt spacing. To prevent this, the AISC Specification (2010) states that the center-to-center
distance of the standard, oversized or slotted holes should not be less than 2.67 times nominal
diameter of the bolts while a spacing of three times diameter of the bolt is preferred. Currently for
ease of fabrication, for bolts with diameter of up to 1-1/8 (included), the bolt spacing is given as
three inches. This satisfies the requirement of bolt spacing not being less than 2.67 times diameter
of bolt. A bolt spacing of 3 inches is very common.

10.3.f. Maximum Bolt Spacing

The longitudinal spacing of the bolts (center-to-center distance of two adjacent bolts) in direction
parallel to applied shear, should not exceed 24 times the thickness of the thinner connected part
nor 12 inches. For weathering steel, which is corrosion resistant steel primarily used in bridges or
exposed steel in buildings, the maximum spacing should not exceed 14 times the thickness of the
thinner connected part nor 7 inches.

10.3.g. Minimum Edge Distance

To prevent edge distance failure, shown in Figure 10.8(d), the AISC Specification (2010a)
provides minimum edge distance in Table J3.4, which is given below as Table 10.4.

10.3.h. Maximum Edge Distance

Maximum distance from the center of any bolt to the edge of the connected part, see Figure
10.8(d), should be 12 times the thickness of the connected part under consideration but not more
than 6 inches. This requirement is to make sure that the edges of the connected plates do not warp
upwards and separated from each other.
Table 10.4
(of the AISC Spec.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-14

10.4. BEHAVIOR AND DESIGN OF BOLTS

SUBJECTED TO COMBINED SHEAR AND
TENSION
In many connections of steel and composite structures, such as the example in Figure 10.12(a),
fasteners are subjected to a combined shear and tension, Figure 10.12(b). Chesson et al.(1965)
tested A325 and A354 bolts under such condition. Using their results, Kulak et al (2001) proposed
the eliptical interaction curve given by Equation 10.11 below for bolts subjected to shear and
tension.
2 2
 fv   ft 
  +   ≤ 1.0 (10.5)
 0.60 Fv   Ft 
Where
f v = applied shear stress on the nominal area of the bolt
f t = applied tensile stress on the nominal area of the bolt
F v = strength of the bolt in pure shear.
F t = strength of the bolt in pure tension.

AISC Spec. (2010),

LRFD format
fv
Fv Prposed by Kulak at
al., 2001:
1.0
2 2
 fv   ft 
T
T 0.75   +   ≤ 1.0
 0.60 Fv   Ft 

V
0.25

ft
1.0
Ft
(a) (b) (c)

Figure 10.12. Bolts Subjected to Combined Shear and Tension Forces

Design equation for failure mode of tension rupture of a bolt subjected to combined shear
and tension is given as (AISC, 2010):
R u ≤ φF’ nt A b (in LRFD) (10.6)

where
CHAPTER 10 – BOLTS 10-15

R u = factored tension force applied to the bolt for use in LRFD methods
φ = resistance factor for bolts= 0.75
A b = nominal area of the fastener calculated using nominal diameter.
F’ nt = nominal tensile strength of the bolt modified to include the effects of shear stress and is
given by:

' Fnt
Fnt = 1.3Fnt − f v ≤ Fnt (for LRFD) (10.7)
φFnv

where
F nt =nominal tensile strength of bolt given in Table J3.2 of the AISC Spec. (2010a), reproduced
earlier as Table 10.3 on Page 10-9
F nv = nominal shear strength of bolt given in Table J3.2 of the AISC Spec. (2010a), reproduced
earlier as Table 10.3 on Page 10-9
f v = required shear stress(i.e. shear stress in the bolt due to applied shear force)

Figure 10.12(c) shows the AISC design Equation 10.6 as well as Equation 10.3 proposed by
Kulak et al. (2001).

References for Chapter 10

1. AASHTO, (2007) “LRFD Bridge Design Specifications”, Am. Assoc. of State Highway and
Transportation Officials.
2. AISC, (2010a) “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An American
National Standard, American Institute of Steel Construction, Chicago.
3. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-05, An
American National Standard, American Institute of Steel Construction, Chicago.
4. AISC, (2011) “Steel Construction Manual,” 14th Edition, American Institute of Steel Construction,
Chicago.
5. Kulak, G.L., Fisher, J. W., and Struik, J.H.A. (2001), “Guide to Design Criteria for Bolted and
Riveted Joints, Second Edition”, A Wiley Interscience Publication reproduced by AISC, Chicago.
(Free download at: www.boltcouncil.org),

6. RCSC, (2009). “Specification for Structural Joints Using A325 or A490 Bolts”, Research
Council on Structural Connections, (Free download at: www.boltcouncil.org),
7. Salmon, C. G. and Johnson, J. E. (1996), “Steel Structures-Behavior and Design”. 4th Ed.,
Harper and Collins.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-16

Problems for Chapter 10

Problem 10.1. A 2-MC channel member is bolted to a gusset plate. Find maximum factored load T u = φP n
that can be applied to the connection. Check only failure modes of the bolts. Channels and gusset plate are
A36 steel.

3”
TU
3” Tu
20”
(A)
3”

3”

2MC18x58 Channel

2” 3” 3” 3” 2”
All bolts are A490X, 3/4” diameter
in standard round holes.
Thickness of gusset plate=1.5”

1” Gusset Plate
1” plate
HSS6x6x1/2” A500 Gr.B
(B) 3”
3”
3”

3”
Two 5/8” splice plates
2” 3@3” 2@2” 3@3” 2@2” 9”
3/8”, E70

Problem 10.2. Check the bolted connections shown below for adequacy to carry the applied loads shown.
Check edge distances and bolt spacing in each case. Also check shear and bearing failure of the bolts.

1 ¼ inch Thick Gusset Plate, A36

Ten ¾” dia. , A325N bolts in standard holes.

3”
3”
300 kips 1.5”
300 kips
1.5”
2”
4”

2 L 8 x 6 x1/2 (Double Angles)

(A) A36 steel
4 @ 3 “ = 12”

2” 2”
CHAPTER 10 – BOLTS 10-17

All bolts are in ¾” dia A490 Bolts inside

1” thick gusset standard holes. Standard Round Holes
1” plate A36 steel
plate
A36 steel
2 WT16.5x120.5
A992 Grade 50
2.5
12” 1.5””
2@3” 1.5”

700 kips 1500 kips 2.5” 1500 kips

700 kips

2MC18x58 Channel
A36 steel
1.5” 3@3” 1.5” 3” 6@2”=12” 3”

1” thick gusset
plate
A36 steel ¾” dia A490 Bolts inside
Standard Round Holes

500 kips
500 kips

2C15x50 Channels
4@3” A36 steel

1.5”
1.5”

(D)

Problem 10.3. Check adequacy or inadequacy of the connection shown below. If you find any component
inadequate to carry the applied load, redesign the component to be adequate. All steel is A572 Gr. 50. All
bolts are 1” dia. A325N bolts in standard round holes. Check all possible bolt failure modes. Checking
failure modes of the members is not part of this problem.
WT 22x 115
1” dia A325N bolts
W14x500
P=600 kips
a

2WT 7x66
a

45o
2” 2”
4@3”
Section a-a 2”

Two rows of nine 1” dia

A325N bolts.

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-18

Problem 10.4. Find maximum load Q u (in tension as well as in compression) that can be applied to the
bracing member shown below. Consider all failure modes of the bolts and welds. Bolts are 3/4” dia A490N
bolts. All fillet welds are on both sides, which are done using E70 electrodes. Failure modes of the HSS
and the gusset plate are not part of this problem.
Qu

9” HSS 7x7x1/2
Two rows of nine
3/4” dia A490SC Gusset Plate, 1” thick, A36
bolts. Fillet Welds
(A) 55o 1-1/4” thick, A36 Plate

2” 8@3” 2” W12x120,
14 11 A992 Grade 50 Steel

HSS 6x6x1/2 Qu
Brace

Two rows of six

3/8”
3/4” dia A325N bolts. 7”
3/8”

Gusset Plate, 3/4” thick

(B) A36
60o 1” thick A36 Plate

9”

5.5”
2” 5@3”
2” 2”

Qu
HSS 8x8x5/8, A500 B
Steel Brace

14”
3/8”
Gusset Plate, 3/4” thick 3/8”
A36

(C) Two rows of nine

3/4” dia A325N bolts.
1” thick A36 Plate
60o

9”
5.5
2”
2” 2”
8@3”
CHAPTER 10 – BOLTS 10-19

Problem 10.5. Given the information below and by checking ALL possible failure modes of the member as
well as the connection, find maximum value of the factored tension force T u that can be applied to the
member. Steel grades are indicated on the figure. Fillet welds are done using E70 electrodes and have a size
of 5/16 inch. Bolts are all ¾-inch diameter A325N bolts inside standard round holes.
3/4” Thick Gusset Plate, 2L5x5x1/2, A36
A36
3/16”
All bolts are ¾” diameter in 13/16” holes
16”
1” WT15x66, A992, Gr. 50

5”

3”
PL9x1/2 Tu
3”
A36
3”
Tu
3”
3”
11” 3@3”

2” 2”
2.5”

Problem 10.6- Consider a tension member made of a W14x257, A992, Grade 50 wide flange section. The
splice for this member is a welded/bolted connection shown below where two WT7x128.5, A992 Grade 50
are used on the flange to transfer the tension force from one side of the splice to the other side. Properties of
bolts and welds are indicated on the figure. Considering all possible failure modes of the wide flanges, WT
sections, bolts and welds, find the governing maximum tension force T u for this connection.
1 ” dia A325N bolts in 1-1/16” dia holes

3” 2” ½”
E70xx
Welds

Tu Tu

6@3”=18” 12”

2” Weld
2” 1”

8”

12@1.5” 12”

2” 2” 1”

________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided only to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to
others or posted anywhere.
CHAPTER 11 – CONNECTIONS 11-1

CHAPTER 11

INTRODUCTION TO STEEL
CONNECTIONS

Objectives:
In this chapter, you will learn:
1. Common types of connections in steel structures
2. General provisions for design of steel connection elements

11.1. INTRODUCTION
From performance, cost and construction points of view, connections are one of the most
important elements in a steel structure. In design of connections, cost considerations are
very important. In addition, the designer needs to consider the ease of fabrication and
erection. This chapter introduces common types of connections used in steel structures and
then presents general procedures for design of connection elements.

11.2. MAIN TYPES OF STEEL CONNECTIONS

We can categorize connections in steel structures as follows:

a. Eccentrically Loaded Connections

b. Shear Connections
c. Moment Connections
d. Partially Restrained (semi-rigid) Connections
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
11-2

e. Gusset Plates Connections

f. Column Base Plates
g. Beam Bearing Plates
h. Column Splices
i. Beam Splices
j. Tee Hanger Connections

Following sections provide a brief discussion of each of the above connection

categories. Chapters 12, 13, and 14 provide more detailed discussions of the behavior and
design of Eccentrically Loaded Connections, Shear Connections, and Moment
Connections, which are the most common types of connections in steel structures.

11.2. a. Eccentrically Loaded Connections

Figure 11.1 shows typical welded and bolted eccentrically loaded connections. In these
connections, the applied load does not pass through the centroid of the welds or bolt group.
The eccentricity of applied load with respect to the centroid of welds or bolt groups created
torque or bending moment on the welds or the bolt group. Chapter 12 discusses design of
these connections.

e = Eccentricity e = Eccentricity
θ
θ

Line of Action
of the Applied Load
Line of Action Pu
of the Applied Load

C.G. of the Welds C.G. of the Weld

(a ) Eccentrically Loaded Weld Lines

e = Eccentricity e = Eccentricity

Line of Action
Pu θ of the Load
Pu
Line of Action
of the Applied Load
C.G. of the
Bolt Group
C.G. of the
Bolt Group

(b ) Eccentrically Loaded Bolt Groups

Figure 11.1 Examples of Eccentrically Loaded Steel Connections

CHAPTER 11 – CONNECTIONS 11-3

11.2. b. Shear Connections

Shear connections usually connect simply supported end of a beam either to a column or to
a girder and transfer reaction of the beam, as a shear force, to the column or to the girder.
Figure 11.2 shows typical steel shear connections used in the United States. Shear
connections have small and, quite often, negligible bending strength. They are sufficiently
flexible to rotate and accommodate end rotations of simply supported beams. Chapter 13
discusses design of shear connections.

Single Double Tee Seat

Plate Angles Section Support

(a) Shear Tab (b) Double Angle (c) Tee (d) Seat

Figure 11.2. Typical Steel Shear Connections

11.2.c. Moment Connections

Moment connections are usually beam-to-column connections. The main role of a moment
connection is to transfer bending moment from a girder to its supporting column. Moment
connections also transfer shear and in some cases axial force as well. Figure 11.3 shows
common types of steel moment connections. Chapter 14 discusses design of moment
connections. Welds
Welds Bolts
Bolts
Welds
Flange Flange
Flange Plate End
Plate Plate
Welds
Welds

(a) Welded Flange (b) Bolted Flange Plate (c) Welded Flange Plate (d) End Plate

Figure 11.3. Typical Steel Moment Connections

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
11-4

11.2.d. Partially Restrained (Semi-Rigid) Connections

Partially restrained moment connections are usually beam-to-column connections. These
connections have bending strength and rotational stiffness less than the moment
connections but more than the shear connections. Research has indicated that partially
restrained connections perform well under monotonic as well as cyclic loading. However,
their use has been limited, probably because the amount of analysis time that is required
for design of this connection is usually more than the time needed for design of rigid
moment connections. Figures 11.4 shows examples of partially restrained steel
connections.

Photo: R. Leon Photo: R. Leon Photo: A. Astaneh

(a) Bolted Flange Angles (b) Bolted Flange Tees (c) Bolted Flange Plates

Figure 11.4. Examples of Partially Restrained Steel Connections

11.2.e. Gusset Plate Connections

One of the most important applications of gusset plates is in trusses to connect the
members to each other, Figure 11.5(a). Gusset plates have similar function to connect

Column

Truss Members Bracing

Gusset Plate

(a) Gusset Plate Beam

(b)
Figure 11.5. Examples of Gusset plate Connections in Trusses
CHAPTER 11 – CONNECTIONS 11-5

the columns and beams in braced frames. The primary force in a gusset plate is axial
tension or axial compression. Small amounts of bending moment and shear in gusset
plates are “secondary" effects and ignored in design.
Figure 11.5 shows gusset plates used in trusses and braced frames. Figure 11.6(a)
shows typical “low-seismic and wind” gusset plate and Figure 11.6(b) shows a “seismic”
gusset plates in a concentrically braced frame respectively.

Fillet Welds

Wide Flange

2L
Bolts
Double Angles Gusset Plate
on Each Flange

Shear Tab

Gusset Plate

(a)
(b)

Figure 11.6 (a) a Typical “Low-Seismic and Wind” Gusset Plate Connection and
(b) a typical “Seismic” Ductile Gusset Plate Connection in Concentrically Braced Frames

11.2.f. Column Base Plate Connections

In the United States, typical column base connections consist of base plates welded to the
bottom of the columns and resting on the R/C foundations or pile caps. The connection is
a “Base Plate Connection.” The anchor bolts connect the base plates to foundations. Two
types of base plate connections, shown in Figures 11.7 and 11.8 are:

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
11-6

a. base plates for columns in braced frames, Figure 11.7, where the applied loads are
primarily axial load (N ) and shear force (V ),
b. base plates in moment frames, Figure 11.8, where base plates act as a rigid
connection and transfer axial load (N ), shear force (V ) and bending moment (M )
from the column to the foundation.

Base Plate

Figure 11.7. Base Plates for Columns in Braced Frames

Figure 11.8. A Typical Base Plate for Columns in Moment Frames

CHAPTER 11 – CONNECTIONS 11-7

11.2.g. Beam Bearing Plates

When a concrete or masonry wall supports a beam, in order to distribute the compressive
force over a larger area, Bearing Plates are used under the beam flange, see Figure 11.9(a).
The connection is quite common in bridge girders supported on reinforced concrete piers,
Figure 11.9(b).

Bearing Plate Bearing Plate

(a) Beam Bearing Plate (b) Bearing Plate in a Bridge Girder

Figure 11.9. A Typical Base Plate for Columns in Moment Frames

12.2.h. Column Splices

Splices are used in steel structures to connect the members. Figure 11.10 shows examples
of splices used in columns. Column splices in general are subjected to shear, axial force
and bending moment.

Bolted-Bolted Splice Welded-Bolted Splice

Figure 11.10. Examples of Column Splice Connections

11.2.i. Beam Splices

Similar to columns, beam splices are used to connect two pieces of a beam to each other.
Generally, beam splices have flange plates as well as web plates, Figure 11.11(a).

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
11-8

(a) Beam splice transferring shear, (b) Beam splice transferring shear and
axial force and moment axial force

Figure 11.11. Examples of Beam and Girder Splice Connections

However, if splice is designed to be at a point of no moment (point of inflection) or

at a point where moment is relatively small, the flange plates can be quite small or even not
needed at all resulting in cost saving, Figure 11.11(b).

11.2.j. Tee Hanger Connections

Tee Hanger connections are used to transfer tension axial force to supporting beams,
girders, or truss cords. Figure 11.12 shows an example of tee hanger connections. In tee
hanger connections, bolts are subjected to tension due to the externally applied load. In
addition, the bolts are subjected to additional tension due to “prying Action” of the tee
hanger flanges.

Tee Hanger
Connection

Figure 11.12. An Example of Tee Hanger Connection

11.3. GENERAL PROVISIONS FOR DESIGN OF

STEEL CONNECTIONS
In design, quite often connections are separated into “elements” such as angles, plates and
other shapes and using equilibrium of the resulting free body diagrams, the forces acting on
each element are established. Then, depending on dominant load acting on each element, it
is considered a tension member, a compression member, a shear element, a beam, or a
beam-column, the latter being subjected to combined loads. In some complex connections,
establishing internal forces on each element of connection by using equilibrium of free
bodies is quite difficult. In these cases, a more rigorous finite element analysis is done to
establish internal stresses in the elements of the connections. Following sections discuss
procedures for design of the elements of steel connections. Chapters 9 and 10 discuss
behavior and design of welds and bolts respectively.
CHAPTER 11 – CONNECTIONS 11-9

11.3.a. Tension Elements of Connections

The provisions and equations for design of tension element of steel connections are the
same as those used for tension members and discussed in Chapter 4. Failure modes are
yielding in the gross area, rupture in the net area and block shear failure.

11.3.b. Compression Elements of Connections

11.3.b.1. Local Buckling of Compression Elements of Connections

As discussed in Chapter 5, local buckling is a phenomenon that depends on width-to-

thickness ratio, b/t, of the rectangular elements of the steel cross section as well as yield
stress, F y , of steel. Similar to steel members, compression elements of steel connections
can also have local buckling. The higher the b/t and F y are, the higher is possibility of
local buckling to occur resulting in lower critical local buckling stress. To check local
buckling of connection elements, similar to columns, Table 5.1 on Page (5-10), can be
used.

11.3.b.2. Overall Buckling of Compression Elements of Connections

The provisions and equations used for overall buckling of compression elements of
connections are the same as we use for design of columns and discussed in Chapter
5, but with one difference. In design of compression elements of connections, if
KL/r is not exceeding 25, the critical buckling stress is equal to yield stress, Fy.
For compression elements of connections the equation of design for overall
buckling is:

Pu ≤ = φ c Pn (in LRFD) and P ≤ P n /Ω c (in ASD)

(11.1)
Where
P u = required axial compression strength in LRFD, same as the factored applied
compression force.
φ c = resistance factor for overall flexural buckling in LRFD = 0.90
P = required allowable buckling strength in ASD, same as the un-factored (service)
applied compression force.
Ω c = safety factor for overall buckling in ASD = 1.67
P n = available nominal overall buckling strength of the column given by:

P n = F cr A g (11.2)
And;
A g = gross area of the connection element
F cr = critical buckling stress, determined as follows:

(a) When KL ≤ 25
r
F cr = F y (11.3)

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
11-10

KL E
(b) When 25 < ≤ 4.71
r Fy

 F
( y )
Fcr 
= 0.658 Fe F (11.4)
  y
 
KL E
(c) When > 4.71
r Fy
Fcr = 0.877 Fe (11.5)
Where
F y = specified yield stress of the steel
E = modulus of elasticity of the steel = 29,000 ksi
KL/r = governing effective slenderness ratio, larger of KL/r with respect to xx and yy axes.
K = effective length factor
L = length of the compressive element
r = radius of gyration of cross section of the compressive element with respect to axis of
buckling
F e = elastic critical buckling stress determined from Euler’s Buckling Equation:
2
π E
Fe = 2 (11.6)
 KL 
 
 r 
 

Equations 11.3, 11.4 and 11.5 above that define critical buckling stress, F cr , are
shown in Figure 11.13. Value of K, the effective length factor for connection elements, is
established using the “six case” AISC Table given on Page (5-23) in Chapter 5. The end
conditions for connection elements in compression need to be established by structural
designer considering the buckled geometry of the element.
Fcr Fcr = Fy
 ( y )
F

Fcr = 0.658 Fe  Fy

 
Fy  

Fcr = 0.877 Fe
0.44 Fy No
Buck- Inelastic
ling Buckling
Elastic Buckling

0 25 E Slenderness Ratio, KL/r

4.71
Fy

Figure 11.13. Design Equations to Establish Nominal Capacity of Compressive Members

CHAPTER 11 – CONNECTIONS 11-11

11.3.c. Bending Elements of Connections

In order to establish available bending strength of a connection element, the limit states of
flexural members discussed in Chapter 6 should be applied. These limit states are flexural
yielding and formation of plastic hinge, local buckling, and lateral torsional buckling. In
addition, for connection elements subjected to bending, limit state of flexural rupture in the
net area should also be considered.
Design equation for limit state of flexural rupture in the net area of a connection
element is:

M u ≤ φ b M n (in LRFD) and M ≤ M n /Ω b (in ASD) (11.7)

where
M u = required factored bending strength in LRFD, same as factored applied bending
moment.
φ = resistance factor in LRFD for bending rupture = 0.75
M = required allowable bending strength in ASD, same as the un-factored (service)
applied bending moment
Ω b = safety factor in ASD for bending rupture = 2.0
M n = available nominal flexural strength of the element given by:

M n = F u Z xn (11.8)

F u = ultimate strength of steel used

Z xn = plastic section modulus of the net area of the element

11.3.d. Shear Elements of Connections

For elements of connections subjected to shear, limit states of shear yielding in the
gross area and shear rupture in the net area need to be checked. The lower strength
obtained from these two limit states is the available shear strength of the element.
Design equation for limit states of shear yielding in the gross area and shear
rupture in the net area of the elements of connections is:

V u ≤ φ V n (in LRFD) and V ≤ V n /Ω (in ASD) (11.9)

where
V u = required factored shear strength in LRFD, same as factored applied shear.
V = required allowable shear strength in ASD, same as the un-factored (service) applied
shear

Values of φ , Ω and V n in the above equations are:

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
11-12

(a) For shear yielding in the gross area:

φ = Resistance factor in LRFD for shear yielding in the gross area of
connection elements = 1.00
Ω = safety factor in ASD for shear yielding in the gross area of
connection elements = 1.50
V n = available nominal shear yield strength given as:

R n = 0.60F y A gv (11.10)

A gv = gross area of the element subjected to shear

F y = specified minimum yield stress of steel

(b) For shear rupture in the net area:

φ = resistance factor in LRFD for shear rupture in the net area of
connection elements = 0.75
Ω = safety factor in ASD for shear rupture in the net area of
connection elements = 2.0
V n = available nominal shear rupture strength given as:

R n = 0.60F u A nv (11.11)

F u = specified minimum ultimate strength of steel

A nv = net area of the element subjected to shear

References for Chapter 11

1. AISC, (2010a) “Specification for Structural Steel Buildings,” ANSI/AISC 360-10, An
American National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” ANSI/AISC 341-05,
An American National Standard, American Institute of Steel Construction, Chicago.
3. AISC, (2011) “Steel Construction Manual,” 14th Edition, American Institute of Steel
Construction, Chicago..
CHAPTER 12 – ECC. LOADED CONNECTIONS
12-1

CHAPTER 12

ECCENTRICALLY LOADED
CONNECTIONS

12.1. Eccentrically Loaded Welds

In Chapter 9, we discussed design of welds. The discussion was based on the assumption
that the applied force passes through the centroid (i.e. C.G.) of the weld line. In this
chapter, we look into cases that the applied load does not pass through the centroid of the
weld line. In these cases, the load is eccentrically applied load. The condition generally
applies to fillet welds and in rare cases to partial joint penetration welds. In this section,
we will focus on fillet weld lines subjected to eccentric loading. The same design
procedures and discussion applies to partial joint penetration welds.
When a fillet weld line is subjected to eccentric loading, Figure 12.1, the eccentric
load can be moved to the center of gravity of the weld lines and a torque or bending
moment is also applied at the center of gravity to satisfy the equilibrium. Then, the weld
group can be designed for the combined effects of direct shear plus torque, Figure 12.1(b),
or direct shear plus bending moment, Figure 12.1(d).
Figure 12.2 shows common types of weld groups subjected to eccentric loading.
These are:
a. Parallel weld lines
b. Box-shaped weld lines
c. L-shaped weld lines
d. C -shaped weld lines

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved. This
document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-2

e = Eccentricity
θ

Line of Action
of the Applied Load

C.G. of the Welds C.G. of the Welds

Pu Pue

(a) Eccentrically Applied Load (b) Load Moved to C.G. of the Welds
and a Torque is Also Applied
Eccentricity = e

Line of Action
of the Load

Pu
Pu
Pu e
C.G. of the
Weld Line
C.G. of the
Weld Line
(c) Eccentrically Applied Load (d) Load Moved to C.G. of the Welds
and a Torque is Also Applied
Figure 12.1. Weld Groups Subjected to Eccentric Loading

Two methods are used in design of weld lines subjected to eccentric loads. One
method is to consider the weld behavior to be elastic and the second method is to consider
inelastic behavior of the weld line. The elastic method ignores redistribution of stresses
inside the welds as the highly stressed areas yield. Elastic method is less economical
design than the inelastic method since following elastic method, we obtain less capacity for
the weld than we can obtain by using inelastic method.

12.1. a. Elastic Method or “Vector” Combination Method

In this method, the elastic stresses due to the applied shear and torque or applied shear and
bending moment are calculated separately assuming elastic behavior and then added up as
vectors. The resulting combined stress is considered the “demand” and is used to check
against the “capacity” of the weld. This method was the only method in use until the 1980
is when the “inelastic” method, discussed in the next section, was incorporated into the
AISC Specification (AISC, 2010) and the Steel Construction Manual (AISC, 2011). The
elastic method still can be used as an alternative to inelastic method. As mentioned
earlier, the elastic method generally results in a less economical design than the inelastic
method.
CHAPTER 12 – ECC. LOADED CONNECTIONS
12-3

In design of eccentrically loaded weld lines using elastic method and LRFD
procedures, the following steps are taken.
1. Establish geometry and location of the centroid (i.e. C.G.) of the weld line.
Assume the same size of weld, D, for all welds in the fillet weld line. The size of
welds will be calculated later.
2. Establish applied load P r acting on the weld line with an eccentricity of e with
respect to the C.G. of the weld line, See Figure 12.3(a).
3. Consider C.G. of weld group to be the origin of the xy coordinate system, see
Figure 12.3(a).
4. Move the applied load P r to the C.G. of the weld line and apply a torque equal to
(P r ) (e) to the C.G., where e is the eccentricity of the load P r from the centroid of
the weld lines, see Figure 12.3(b).
5. Resolve the applied load P r into its x and y components; P rx and P ry , see Figure
12.3(c):
P rx = P r sin(θ) (12.1.a)

P ry = P r cos(θ) (12.1.b)

In the above equations, θ is the angle between applied load P r and the y-axis.
6. Calculate direct shear stress on the welds in x and y directions as:
f vx =P rx / (w) ( ∑ Lw ) (12.2.a)

fvy =Pry / (w) ( ∑ Lw ) (12.2.b)

See Figure 12.3(d) for direct shear stresses. Note that fvx and fvy can be negative or
positive depending on the sign of Prx , and Pry. In the above equations, w is the
throat thickness of the fillet weld equal to D / 2 , where D is the size of fillet
welds.
7. Select a point on the weld that appears to have maximum stress due to combined
effects of direct shear and shear due to torque. Usually the corner points, Points A,
B, C, and D, in Figure 12.3(c) that are farther away from the centroid are likely to
have large combined stresses. Several corner points may need to be considered to
obtain the one that has the largest combined stresses. For the selected point,
calculate shear stress due to torque in x and y directions as:
fTx = (Prx) (ex) (y) / (Ip) (12.3.a)

fTy = (Pu) (e) (x) / (Ip) (12.3.b)

See Figure 12.3(e) for stresses due to torque. In the above equations, x and y are
the distance of the point under consideration from centroid (C.G.) of the welds in x
and y directions respectively. Ip is the polar moment of inertia of the weld lines
with respect to C.G. and is equal to Ix+Iy of the weld lines. Note that fTx and fTy can
be negative or positive depending on the sign of Prx , and Pry, x, and y.
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved. This
document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-4

8. Calculate total horizontal ( along x-axis) and vertical (along y-axis) shear stresses
by adding shear stresses due to direct shear and torque:
fx = fvx + fTx (12.4.a)

fy = fvy + fTy (12.4.b)

9. Combine stresses fx and fy as vectors and obtain total stress in the weld as:

2 2
(12.5)
fr = fx + f y

10. After establishing the largest resultant stress, denoted as fr , use the following
equation to ensure that the applied stress fr does not exceed the available strength
of the weld:
fr ≤ φ Fnw (in LRFD) and fr ≤ Fnw /Ω (in ASD) (12.6)
where
Fnw = nominal strength of weld in shear = 0.60FExx
FExx= nominal strength of weld electrod
φ = reduction factor for rupture in the fillet weld in LRFD = 0.75
Ω = factor of safety for rupture in the fillet weld in ASD = 2.0

y y y θ
e
θ Pry

Pr Pr x x
x
P
Prx
r
Pr e Pr e
e

(a) Force Pr with (b) Force Pr moved (c) Force Pr resolved

eccentricity “e” to C.G. of the weld lines into x and y

y y y

C D

x x x

Pr
B fvx A fPTxr A fTx fP r
vx A
e e e
A
fvy fvy
fTy

fr fTy

(d) Stresses due to (e) Stresses due to (f) Total stresses due to combined
direct shear torque shear and torque

Figure 12.3. Application of Elastic Method

CHAPTER 12 – ECC. LOADED CONNECTIONS
12-5

12.1.b. Inelastic Method or “Instantaneous Center of Rotation”

Method
This method considers inelastic range of behavior when the highly stressed areas
of the weld group yield and redistribute their stress to the adjacent areas of the weld. As
the applied load increases, the yielding spreads to other parts of the weld in the weld group.
When a point on the weld line that has highest stress starts fracturing, the weld has reached
its capacity.
The AISC Specification (AISC, 2010) has equations to establish strength of welds
using inelastic method. For common cases, the AISC Steel Construction Manual (AISC,
2011) provides a series of tables that can be used to establish strength of the eccentrically
loaded welds with common geometries based on using the inelastic method. The tables
provide “Available Strength” of eccentrically loaded weld groups with common
geometries that were shown in Figure 12.2 earlier.

12.2. ECCENTRICALLY LOADED BOLT GROUPS

When a bolt group is subjected to eccentric loading, Figure 12.4, the eccentric load
can be moved to the center of gravity of the bolt group and a torque applied to the center of
gravity of the bolt group to satisfy the equilibrium. Then, the bolt group can be designed
for the combined effects of direct shear and torque.
e

Line of Action
Pu of Load
Pu
Pu e

Centerline of
Centerline of Bolt Group
Bolt Group

(a) Eccentrically Applied Load (b) Load Moved to C.G. of the Bolt
Group and a Torque is Also Applied
e

Pu
Pu

C.G. of C.G. of
Bolt Group Pu e Bolt Group

Figure 12.4. Eccentrically Loaded Bolt Groups

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved. This
document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-6

Figure 12.4 shows bolt groups with one or more vertical bolt lines subjected to
eccentric loading. Similar to welds, there are two design methods for bolt groups: elastic
and inelastic. The methods are explained below. Again, similar to welds, the elastic
method results in less economical design than the inelastic method.

12.2.a. Elastic Method or “Vector” Combination Method

In this method, bolts are assumed to remain elastic until the shear force in one bolt reaches
its shear capacity and the bolt fails in shear. To obtain the total factored shear force
applied to each bolt, shear forces due to direct shear and torque are combined as vectors,
Figure 12.5, and the resulting shear force is compared to the shear strength of one bolt.
Similar to welds, the resulting combined shear force is considered the “demand” and is
used to check against the shear “capacity” of the bolts. Shear capacity of individual bolts
was discussed in Chapter 10. Again, like welds, this method was the only method in use
until the 1980’s when the “inelastic” method, discussed below, was incorporated into the
AISC Specification (AISC, 2010) and the AISC Steel Construction Manual (AISC, 2011).
The elastic method can still be used as an alternative to inelastic method, however, as
mentioned earlier the elastic method generally results in more conservative design than the
inelastic method.

e e
Pu
Pu
Pu

Line of
s Action of
the Load
s Pue Fbu
Fu
g C.G. of Fvu
the Bolt Group

(a) Bolt Group with Applied Force (b) Bolt Group with Shear Forces Acting on Bolts

Figure 12.5. Direct Shear, Fvu, and Shear due to Torque, Fbu Acting on the Bolts

In design of eccentrically loaded bolt groups using elastic method, following steps
are taken:
1. Establish the geometry of the bolt group and its centroid (location of C.G.). Assume
the same diameter and type of bolt for all bolts in the bolt group. The diameter of
bolts will be calculated later.
2. Establish loads acting on the bolt group.

3. Move the applied force Pu to the centroid of the bolt group and apply the torque
(Pu) (e) to the centroid, where e is the eccentricity of the force Pu from centroid of
the bolt group, see Figure 12.5(b).
CHAPTER 12 – ECC. LOADED CONNECTIONS
12-7

4. Calculate direct shear force acting on each bolt by dividing the applied load Pu by
the number of bolts, see Figure 12.5(b):

Fvu =Pu / N (12.5)

5. Calculate shear force due to torque:

Ftu = (Pu) (e) (r)(1.0 in2) / (Ip) (12.6)

6. Consider Fvu and Ftu vectors and find their resultants, Fu for all bolts, see Figure
12.6(b). By observation, we should be able to identify which bolt will have the
largest resultant. After establishing the largest resultant force, Fu, use the following
equation of design to ensure that the applied shear force (the resultant force Fu )
does not exceed the available shear strength of the bolt, φ Rnb :
Fu ≤ φ Rnb (in LRFD) and F ≤ φ Rnb /Ω (in ASD) (12.7)

In the above equations,

Fvu = direct shear force in a bolt = Pu / N
N = number of bolts in the bolt group
Fbu = shear force in a bolt due to torque (Pu) (e) (r) (1.0 in2) (Ab) / Ip
e = eccentricity of applied load Pu
r = radial distance from centroid of bolt group to the center of a bolt
Ip = polar moment of inertia of the bolt group = ∑ ( 1.0 in2) (Ab) (r 2)
Rnb = nominal shear strength of one bolt = (Ab) (Fnv)
Ab = area of one bolt
Fnv = shear strength of bolt established in Chapter 10, (Page 10-7).
Pu =applied factored load
Fu = resultant shear force, acting on one bolt = magnitude of vector Fu = Fvu + Fbu
φ = reduction factor for bolts = 0.75

12.2.b. Inelastic Method or “Instantaneous Center of Rotation”

Method
This method considers inelastic range of behavior when the highly stressed bolts
yield in shear and bearing and redistribute their shear force to the other bolts. As the
applied load increases, the yielding spreads to other bolts in the bolt group. When shear
force in a bolt reaches its available strength, the bolt fractures in shear or fails in bearing
failure. When the first bolt fails, the connection is considered failed.
The AISC Specification (AISC, 2010) has equations to establish strength of bolt
groups using inelastic method. For common cases, the AISC Steel Construction Manual
(AISC, 2011) provides a series of tables that can be used to establish strength of the
eccentrically loaded bolt groups with common geometries based on using the inelastic
method. The tables provide “Available Strength” of eccentrically loaded bolt groups with
common geometries of bolt groups. These tables start on Page (7-30) of the AISC Manual.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved. This
document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-8

References for Chapter 12

1. AISC, (2010a) “Specification for Structural Steel Buildings”, ANSI/AISC 360-10, An
American National Standard, American Institute of Steel Construction, Chicago.
2. AISC, (2010b) “Seismic Provisions for Structural Steel Buildings”, ANSI/AISC 341-10,
An American National Standard, American Institute of Steel Construction, Chicago.
3. AISC, (2011) “Steel Construction Manual,” 14th Edition, American Institute of Steel
Construction, Chicago.
4. Kulak, G.L., Fisher, J. W., and Struik, J.H.A. (2001), “Guide to Design Criteria for
Bolted and Riveted Joints, Second Edition,” A Wiley Interscience Publication
reproduced by AISC, Chicago.
5. Salmon, C. G. and Johnson, J. E. (1996), “Steel Structures-Behavior and Design”. 4th
Ed., Harper and Collins.
CHAPTER 12 – ECC. LOADED CONNECTIONS
12-9

PROBROBLEMS ON ECCENTRICALLY LOADED

WELDS, BOLTS, AND STUDS
Problem 12.1. Calculate Pu , the factored load capacity of bolted or welded connections shown
below. Consider only failure modes of welds, bolts or shear studs as applicable.

5” 9”
5” 12”
Pu
Pu
300

10”
12”

½” E70
5/8” E70
(A) (B)

16”
8”
200
Pu
300
Pu

1.5” 1.5”
3 3/4” thick 3 3/4” thick
3 A36 plate 3 A572 Gr 50
3 3
1.5” 1.5”

6”
(C)
(D)
8”
10” Pu

45o
Pu
2” 5/8” , plate

5/8” , plate 6@3 1” dia

14”
A325N
bolts
2”
C.G. of welds
Fillet welds, D=3/8” W14x61
E70 electrode
5½”
_____________________________________________________________________________
6”
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved. This
(F)
document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
(E)
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
10-10

Problem 12.2. Establish adequacy or inadequacy of the connection shown below. If you find any
component inadequate to carry the applied load, redesign the component to be adequate. All steel is
A572 Gr. 50. All welds are E70xx. Applied load is factored load.

5” Pu =150 kips Pu =90 kips

6” 6”

Wide
Flange
Column 5/8”
5/8” E70
20” 5/8” E70
20”

PL20x8x1”
A572, Gr. 50 Steel 2PL20x16x1”
A572, Gr. 50 Steel
(A)
(B)

Pu =70 kips
3”
30o
Pu = 45 kips
6” 6”

70o
12” 1/2” E70
1/2”

22
PL12x8x1”
A572 Gr 50
Steel
(C) 5/8” E70

2PL22x1”
A572, Gr. 50
(D) Steel

8” 4”
Pu =60 kips

45o

18”

5/8” E70

2PL20x16x1”
A572, Gr. 50
Steel
(E)
CHAPTER 13-SHEAR CONNECTIONS
13-1

CHAPTER 13

SHEAR CONNECTIONS

Objectives:
In this chapter, you will learn:
1. Common types of steel shear connections `
2. Behavior of steel shear connections
3. Failure mode (or limit state) of steel shear connections
4. Design of common steel shear connections

13.1. INTRODUCTION TO SHEAR CONNECTIONS

Shear connections are frequently used in steel and composite structures to connect simply
supported beams to their support. Figure 13.1 shows typical steel shear connections where the
connections are on the column flange or column web. Shear connections are also used to connect
beams and girders to other beams or girders.
The most common types of steel shear connection in the United States is shear tabs, also
called single plate shear connection. In this chapter, we will focus on this shear connection. In
addition, a discussion of double-angle shear connections is provided and followed by design
procedures. For design of shear end plate and seat connections the tables in the AISC Steel
Construction Manual (2011) can be used.

a. Shear Tab b. Double Angles c. Tee d. Shear End Plate e. Seat

(a) Typical Shear Connections on the Flange of WF Columns

a. Shear Tab b. Double Angles c. Tee d. Shear End Plate e. Seat

(b) Typical Shear Connections on the Web Side of WF Columns

Figure 13.1. Typical Steel Shear Connections

13.2. DEFINITION OF SHEAR CONNECTIONS

Beam end connections in steel structures are divided into shear and moment connections. Moment
connections are further divided into two categories of Fully Restrained (FR) and Partially
Restrained (PR) connections (AISC, 2010). This chapter is about shear connections. The Fully
Restrained connections, often called rigid moment connections, or just moment connections, are
discussed in the next chapter.
Figure 13.2 shows a more formal definition of the beam end connections in terms of their
moment-rotation behavior. Shear connections are defined by Salmon and Johnson (1996) as
connections with moment capacity less than 20% of the plastic moment capacity of the connected
beam. This definition of shear connections is used in Figure 13.2 to define the zone of “simple” or
shear connections. Line OAB in Figure 13.2 is the boundary between the zones of partially
restrained and shear connections, while Line OCD is the boundary between the zones of fully
restrained and partially restrained connections. Depending on the location of the moment-
CHAPTER 13-SHEAR CONNECTIONS
13-3

rotation curve of a connection, the connection can be categorized as FR, PR or Simple (shear)
connection. In Figure 13.2 curves OEF, OGH, and OIG show moment-rotation curves for simple,
PR or FR connections respectively. The stiffness term “m” in Figure 13.2 is the ratio of the elastic
rotational stiffness of the connection, K Conn , to the bending stiffness of the connected beam, (EI/L)
b.

Rotation
Moment,
MConn / MPb
Zone of Fully Restrained Connections
Moment
L
G
. 1.0 I
C D

m=18 Zone of Partially Restrained Connections

H Note:
G m = (KConn) / (EI /L)b
m=0.5

A B
0.2
E Zone of (Simple) Shear Connections F

O Rotation, θ

Figure 13.2. Three Types of Beam-to-Column Connections

The terms in Figure 13.2 are:

M conn = moment in the connection
M Pb = plastic moment capacity of the beam
m = ratio of rotational stiffness of the connection to bending stiffness of the beam
m = (K Conn )/(EI/L) b
K Con n = rotational stiffness of connection
(EI/L) b = bending stiffness of the connected beam
E = modulus of elasticity
I = moment of inertia of the beam about strong axis (x-x axis)
L = span of the beam
θ = rotation of the connection , same as rotation of beam end
According to the AISC Specifications (AISC, 2010), a simple ( or shear) connection has
the following properties:
a. Transmits a negligible moment across the connection
b. Has sufficient rotation capacity to accommodate the required rotation determined by
the analysis of the structure. Inelastic rotation of the connection is permitted.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
13-4

In the analysis, simple connections can be assumed to act as “pin” connections and
provide unrestrained rotation between the connected members.
As shown in Figure 13.2, in shear connections, a relatively small bending moment, less
than 20% of the plastic moment capacity of the beam, develops. This relatively small negative
moment acting at the ends of the beam is usually ignored in design of the beam itself and the beam
is designed as a simply supported beam. Doing so satisfies the above AISC specification
requirement (a). However, the relatively small moment at the end of the beam can have
detrimental effects on the design of the connection elements such as plates, angles, welds,
and bolts. Therefore, the small moment in a shear connection should be considered in
design of the connection elements.

13.3. BEHAVIOR OF SHEAR CONNECTIONS OF

SIMPLY SUPPORTED BEAMS
13.3. a. General Behavior
Depending on the location of the applied shear and its eccentricity from various elements
of the connection, such as bolts, welds, angles and plates, these elements will be subjected to a
combination of shear and bending moment. As a result, behavior and design of a shear connection
strongly depend on the location of the point of inflection of the beam where bending moment is
zero and only pure shear exists.
If the location of point of inflection is known, one can calculate bending moment at any
cross section of the connection by simply multiplying the shear acting through the point of
inflection by the eccentricity of the element under consideration from the point of inflection,
Figure 13.3.

Shear
Connection

End
Moment Point of
Inflection
V

Bending
Point of Moment
Inflection Diagram

Eccentricity of Point of Inflection

from the centerline of the support
(a) (b)

Figure 13.3. (a) Location of Point of Inflection and (b) Free Body Diagram of Connection Area

Location of point of inflection in simply supported beams depends on the rotational

stiffness and bending strength of the shear connections. In general, location of point of inflection
is established by tests. This is because shear connections experience inelastic behavior even under
service loads and the inelasticity makes the state of stress quite complex and hard to establish
analytically.
CHAPTER 13-SHEAR CONNECTIONS
13-5

Figure 13.4(a) shows a simply supported beam loaded until plastic hinge formed at
midspan. Figure 13.4(b) shows typical moment-rotation curve for the end connections, which
indicates relatively small moment, develops in shear connections. Figure 13.4(c) is typical shear-
rotation curve for shear connections. Shear-rotation curve is the most important data on
connection behavior as well as its design.

θ
V M

(a) Typical Simply-supported Beam Uniformly Loaded Until Failure

Moment, M
Shear, V
Mp
Moment
Vy
Shear

Rotation

Rotation, θ Rotation, θ

(b Moment –Rotation Behavior (c) Shear –Rotation Behavior

Figure 13.4. (a) A Simply-Supported Beam Loaded Uniformly until Collapse

(b) Moment-Rotation Curve,
(c) Shear-Rotation Curve, and;

13.3.b. Design Criteria for Shear Connections

In design of shear connections the following requirements should be satisfied:
1. The connection should have sufficient shear strength to resist applied forces. Under
gravity load alone, the main forces acting on a shear connection are shear force and a
relatively small bending moment. Under the combined effects of gravity and lateral loads
(earthquake and wind effects), shear connections are subjected to combined shear, a
relatively small bending moment and axial load.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
13-6

2. Shear connections should be sufficiently flexible in rotation such that the end moments
developed in the beam are small and less than 20% of the plastic moment capacity of the
connected beam.
3. Shear connections should have sufficient rotational ductility to tolerate rotations due to
combined effects of gravity and lateral loads.

13.4. SINGLE-LATE (SHEAR TAB) CONNECTIONS

13.4.a. Introduction to Shear Tab Connections
A single plate or shear tab connection consists of a steel plate welded to a column or a girder and
bolted to the web of a beam.
a
Single-Plate (Shear Tab)

Short slotted or round standard

holes

Beam

Column Flange

Figure 13.5. Typical Shear Tab Connection

13.4. b. Behavior of Shear Tabs under Gravity Load Effects

Under gravity load, a shear tab connection of a simply supported beam is subjected to:
1. Shear force, which is the reaction of the beam
2. Relatively small moment, which is usually less than 20% of the plastic moment
capacity of the beam, and;
3. Relatively large rotations on the order of 0.03 to 0.05 radians.
Figure 13.6 shows a shear tab specimen during the tests and after the connection has failed
at the end of the test.
When a shear tab connection is subjected to shear and rotation, the following failure modes are
possible:
1. Yielding of gross area of plate, see Figure 13.6
2. Bearing yielding of bolt holes in the plate and/or beam web, see Figure 13.6
3. Failure of edge distance of bolts , see Figure 13.6
4. Shear fracture of net area of plate
5. Fracture of bolts, see Figure 13.6
6. Fracture of welds
CHAPTER 13-SHEAR CONNECTIONS
13-7

Shear
Yielding in
the Gross
Area

Shear Deformation
Rotation
of the Beam

Bearing
Failure

Bolt Shear
Failure

Edge
Distance
Failure

Figure 13.6. Shear Yielding of Shear Tab (left) and Failure of Bolts and Edge Distance

13.4.c. Location of Point of Inflection in Shear Tab Connections

Tests of shear connections, subjected to realistic shear forces and rotations, have indicated
that as the load in a simply supported beam with shear tab end connections increases, the point of
inflection moves closer to the connection. This is due to yielding in the connection, as shown in
Figure 13.6, which results in shedding of the fixed end moments developed in the connection.
Figure 13.7 shows movement of point of inflection towards the connection as the applied
shear increases.
Currently, the AISC Manual (2011) provides eccentricity of shear force in shear tab
connections in terms of number of bolts. Table 13.1, from the AISC Manual (2011), shows
eccentricities measured from the point of inflection to the bolt line.
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
13-8

Shear, V

Vy
e

Points of
Inflection

Eccentricity, e

Eccentricity used
in design V

Figure 13.7. Variation of Distance of Point of Inflection from the Bolt Line as Shear Increases

In the table below, t p is thickness of the plate, t w is thickness of the beam web, n is the
number of bolts, d is the diameter of the bolts and e is the eccentricity of the shear force from the
bolt line.

Table 13.1
CHAPTER 13-SHEAR CONNECTIONS
13-9

13.5. DESIGN OF SHEAR TABS AS PER AISC

MANUAL PROCEDURES
The AISC Steel Construction Manual (AISC, 2011) has a section on design of Single-Plate Shear
Connections (shear tabs), which includes a procedure for design followed by design tables. The
limit states considered in these procedures are those listed earlier (see P. 13-7).
13.5.a. Design of Single-Plate Shear Connections with
“Conventional Configuration”
Figure 13.8 shows three types of single-plate (shear tab) connections included in the AISC Steel
Construction Manual (AISC, 2011). There are Conventional, Stiffened Extended, and Unstiffened
Extended. The conventional single plate shear connection, Figure 13.8(a) is generally for
connection of a simply supported beam to the flange of a column. The stiffened Extended single-
plate connection, Figure 13.8(b) is used when the simply supported beam is connected to the web
side of a wide flange column. The unstiffened extended single plate shear connection is used to
connect beams to web of the columns, Figure 13.8(c), or the girders, Figure 13.8(d). In this
document, we will discuss design of Conventional and Stiffened Extended single-plate shear
connections as given in the AISC Steel Construction Manual. These two types are the most
common type of single plate shear connections. The AISC Manual has also design procedures for
Unstiffened extended single-plate connections.

Stiffeners
(Top & Bottom)

(a) Conventional (b) Stiffened Extended (c) Unstiffened Extended

(d) Unstiffened Extended

Figure 13.8. Applications of Shear Tab Connections

13.5.b. Single Plate (Shear Tab) Design Using the AISC Manual Tables
The AISC Manual (AISC, 2011) provides a series of tables to establish “Available Shear
Strength” of shear connections, V c . To use the tables to establish available shear strength of a
given shear tab connection, values of d, F y , n, type of bolts (A325 or A490), thread condition
(threads included or excluded from shear plane, and t can be entered and available shear strength
is the value in the table. In the table, d is bolt dia., F y is specified minimum yield stress, n is the
number of bolts and t is the plate thickness. Following illustrative example shows its use.
For design of a shear tab for a given applied shear, a value of shear listed in the value of
shear can be , a value of shear listed in the body of the table, equal or slightly greater than the
applied shear should be selected and the properties of designed shear tab read from the table.
Illustrative Example 13.1
Using the LRFD methods, find factored shear load V u that can be applied to the shear tab
connection shown below.
6 ¾” dia. A325N in
short slotted holes

W21x44, A992,
Gr. 50

¼
¼

Given:
PL 23½ x 6 x 3/8 ,
Details of the shear tab connection are given. A36 steel
3½
Required:
To establish shear strength of the connection.
Solution:

Entering d = ¾”, F y = 36 ksi, n = 6, , type of bolt=A325, thread condition= included, plate

thickness = 3/8 into the Table in Page 13-15, we obtain shear strength of:
V u = 100 kips (in LRFD).

Now we need to check thickness of plate to satisfy AISC requirement for bearing yielding:
t = 3/8” ≤ d/2 +1/16 = (3/4)/2+1/16 = 0.4375” OK

Check thickness of web:

t w = 0.35 ≤ d/2 +1/16 = (3/4)/2+1/16 = 0.4375” OK

Check weld size:

D= 0.25” ≥ ( 5/8)(t ) = 0.234” OK ( The minimum weld size is also shown at the bottom of the Table.

V u = 100 kips
CHAPTER 13-SHEAR CONNECTIONS
13-11

Table 13.2

13.5.c. Shear Tabs Connected to the Web Side of Columns

In many applications, shear tabs are used to connect the beam end to the web side of wide flange
columns, as shown in Figure 13.9(a). In these cases, the shear tabs are widened to be welded to
the web of the column while the bolt holes remain on the extended portion of the shear tab and
outside the column. As shown in Figure 13.9(c), basic equilibrium and free-body diagrams can be
used to establish forces that the elements of connection are subjected to and then design the
elements for such forces. The current AISC Manual of Steel Construction has design procedures
for design of this type of shear tab connections (AISC, 2011)

H/2
H/2
H/2

H/2 H H

L V
L
V V V

H
V
H/2
H
H/2
H/2
Area of Significant
Shear Yielding
H/2
b eb

V = Design Shear Force

H = (V ecw )/ L

(a) (b) (c)

Figure 13.9. (a) Shear Tabs Connected to the Web Side of the Column, (b) Expected Behavior;
and (c) Free-Body Diagram of Connection Elements

13.6. DOUBLE-ANGLE SHEAR CONNECTIONS

13.6.a. Introduction to Double Angle Shear Connections
Double- angle shear connections consist of two angles bolted or welded to a column or a girder
and bolted or welded to the web of a beam, Figure 13.10. Double angle shear connections have
been used frequently in the past in bridges and buildings to transfer shear from the web of the
beams to their support. Examples of double-angle shear connections are shown in Figure 13.11.
The angles can be bolted or welded to the beam and the supporting member. From construction
point of view, connections in which the angles are bolted to the beam web, Figure 13.11(a), (c),
(e), and (g), provide more tolerances in fabrication due to presence of a gap between the bolt hole
and the bolt. When the angles are welded to the web of a beam or flange of a column, Figure
13.11(b), (d), (f) and (h), the back to back dimension of the beam, with angles connected to it,
should be slightly (about ¼”) smaller than the actual face-to-face dimension of the column flanges
or girder webs to enable the erectors to place the beam easily in its location. The gap between the
back of the angles and the face of the column flange or girder web can be filled with shim plate on
one end or both ends. For welded-welded connection shown in Figure 13.11(d) and (e), the back
________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122-Design of Steel Structure at UC Berkeley for their personal use and should not be
distributed to others or posted anywhere.
BEHAVIOR AND DESIGN OF STEEL AND COMPOSITE STRUCTURES
13-14

to back dimension of the beam should be very close to the distance between the face of the
supports since in this case the shims cannot be used between the angle legs and column faces. If
double angles are connected to the column in the shop, then in order to place the beam web
between the two outstanding legs of the angles, the bottom flange of the beam is usually coped,
Figure 13.11(a) and (e).

(a) (b) (c) (d)

Figure 13.10 Typical Double Angle Shear Connections

Column
Beam

Double Double Double

Double Angles Angles Angles
Angles

(a) Bolted-Welded (b) Welded-Bolted (c) Bolted-Bolted (d) Welded-Welded

Girder
Beam

Double Double
Double
Double Angles Angles
Angles
Angles

(e) Bolted-Welded (f) Welded-Bolted (g) Bolted-Bolted (h) Welded-Welded

(Notice that connections can be on both sides of the supporting member)

Figure 13.11. Examples of Double Angle Shear Connections

13.6.b. Design of Double Angle Shear Connections

CHAPTER 13-SHEAR CONNECTIONS
13-15

Currently, design of double angle shear connections are done by simply selecting a connection to
carry the applied load from the tables in the AISC Manual (AISC, 2011). The tables, over the
years, have proven to be easy to use and the connections designed using these tables have
performed quite satisfactorily in the field and in the laboratory as was established by tests. The
AISC Manual (2011) tables provide capacities that are based on the assumption of point of
inflection being along the bolt line on the web of the beam. The studies summarized earlier
confirmed validity of this assumption. The tables are also based on a few other limitations, which
include maximum thickness of the angle as well as the size of the angle used.

REFERENCES FOR CHAPTER 13

1. AISC, (2010A) “Specifications for Structural Steel Buildings,” American Institute of Steel
Construction, Chicago.
2. AISC, (2011) “Steel Construction Manual”, American Institute of Steel Construction, Chicago.
3. AISC, (2005) “Seismic Provisions for Structural Steel Buildings”, American Institute of Steel
Construction, Chicago.
4. Astaneh-Asl, A., S. M. Call, and K. M. McMullin. (1989). “Design of Single Plate Shear Connections.”
Engineering Journal, AISC, Vol. 26, 1st quarter, 21–32.
5. Astaneh-Asl, A., and K. M. McMullin. (1993). “Behavior and Design of Steel Double-Angle Shear
Connections.” J. of Structural Engineering, ASCE, Vol. 119.

Problems for Chapter 13

Problem 13.1. Design a single plate (shear tab ) connection to carry 130 kips factored reaction force of a
beam. Use A6 steel and AISC Manual Tables for Shear Tabs to design the shear tab. Then, using the
equations given for six failure modes of shear tab in this chapter of the Reader, establish strength of the
connection.
W36x135, A992
W27x102, A992 Gr. 50 beam.
Gr. 50 beam.

300 kips
130 kips

(A) (B)

Problem 13.2. Establish shear capacity of the connections shown below using equations given in the
Reader. Edge distances are 2 inches.
24x6x1/2 , A572 Gr. 50

15x6x1/2 , A36 plate 1” dia, A490N Bolts in short

slotted hole
¾” dia, A325X Bolts in
short slotted hole

3/8” 1/2”
3/8” 1/2”

W27x84, A992,
Grade 50 Beam W30x90, A992,
Grade 50 Beam
W12x120, A992,
Grade 50 Column W14x132, A992,
Grade 50 Beam

(A) (B)

Problem 13.3. Design a double angle shear connection to connect a W36x135, A992 Gr. 50 beam to the
flange side of a wide flange column. Double angles are welded to the column flange and bolted to the beam
web.
CHAPTER 14 MOMENT CONNECTIONS
14-1

CHAPTER 14

MOMENT CONNECTIONS

Objectives:
In this chapter, you will learn:
1. Types of steel moment connections
2. Behavior of steel moment connections
3. Failure mode (or limit state) of steel welded and bolted
moment connections
4. Design of common steel moment connections

14.1. INTRODUCTION TO MOMENT CONNECTIONS

This chapter discusses behavior and design of steel rigid or Fully Restrained (FR) moment
connections. These are used in beam-to-column connections of steel moment frames.

14.2. DEFINITION OF MOMENT CONNECTIONS

Figure 14.1 shows three zones of moment-rotation behavior of beam-to-column
connections. The zones, as discussed in previous chapter, are flexible or simple, semi-rigid
or partially restrained (PR) and rigid or fully restrained (FR). Rigid connections are those
that their moment-rotation curve falls within the zone of “rigid” behavior. This means that
the rotational stiffness of the connection exceeds 18EI/L and its moment capacity is greater
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-2

than the plastic moment capacity of the girder. EI/L is the rotational stiffness of the
connected beam.
Moment connections used in lateral load resisting systems should:
a. Have sufficient strength in bending to develop plastic moment capacity of
the connected girder
b. Be strong enough to transfer shear force, and;
c. In High Seismic Applications, have sufficient rotational ductility to tolerate
rotations imposed on them.
End
Moment,
Mconn/MPb Rotation
Zone of Fully Restrained Connections
J
I
Moment
1.0 D
C

m=18 Zone of Partially Restrained Connections

H Note:
m = (KConn)/(EI/L)b
G
m=0.5

A B
0.2
Zone of (Simple) Shear Connections F
E

0 End Rotation, θ

Figure 14.11. Three Types of Beam-to-Column Connections

In the analysis, moment connections are modeled as “fixed” connections. Moment

connections are primarily designed to transfer combined bending moment and shear and in
some cases, axial force as well.
One of the most important uses of moment connections is in the moment frames
that provide lateral strength and stiffness to the building structure. Currently, for wind and
seismic design, the AISC Manual (AISC, 2011) divides buildings into two classification of
“Wind and Low Seismic Applications” or WLSA and “High-Seismic Applications” or
HSA. The Wind and Low Seismic Applications are those where seismic effects are
relatively low and seismic design is done using a Response Modification Coefficient, R
factor, less than or equal to three. In design of WSLA structures, only provisions of the
AISC Specification (AISC, 2010a), needs to be satisfied.
The High Seismic Applications are those that in seismic design of structure, an R
factor greater than three is used. In design of these structures, provisions of both the AISC
Specification (AISC, 2010a) and the AISC Seismic Provisions (2010b) need to be satisfied.
CHAPTER 14 MOMENT CONNECTIONS
14-3

14.3. STEEL MOMENT CONNECTIONS IN WIND AND

LOW SEISMIC APPLICATIONS (R ≤ 3.0)
Figure 14.2 shows common moment connections. Design of moment connections
in Wind and Low Seismic Applications is based on assuming that the applied moment is
resisted only by the flanges and the applied shear is resisted only by the shear connection
on the web of the beam (AISC, 2011). Figure 14.3 shows free body diagram of a bolted
flange plate moment connection where these assumptions are applied. For other moment
connections, similar free body diagrams can be developed.

(a) Welded Connection (b) Welded Connection (c) Welded Flange (d) Welded Flange
on Flange Side of on Web Side of Plate on Flange Plate on Web Side
Column Column Side of Column of Column

Shim
Plate Shim Plate

(e) Bolted Top & Bottom (f) Bolted Top & Bottom (f) End Plate (g) Stiffened End Plate
Flange Plate on Flange Flange Plate on Web Connection Connection
Side of Column Side of Column

Figure 14.2. Examples of Steel Moment Connections

Using free body diagram, shown in Figure 14.3, the elements of connection such as
plates, bolts, and welds can be designed to resist their corresponding forces and moments.
The flange force F f , in Figure 14.3, can be calculated as:

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-4

Ff = Mu / dm (14.1)
where
F f = force in the flange due to factored applied moment M u
M u = factored applied moment
d m = moment arm
Notice that the shear connection on the web and its elements, such as plates, angles,
bolts, and welds, are designed to resist pure shear with no eccentricity. This is because of
flange connections resisting any moment that exists in the connection, including the
moment due to eccentricity of the shear force from the support.

V M

Column

Welds Flange Plates Ff

Ff
V
V V V dm

Ff V M

Ff
Ff

Figure 14.3. Free Body Diagram of a Bolted Flange Plate Moment Connection
CHAPTER 14 MOMENT CONNECTIONS
14-5

14.4. STEEL MOMENT CONNECTIONS IN HIGH

SEISMIC APPLICATIONS (R > 3.0)
Since the 1994 Northridge earthquake, where a number of field-welded moment
connections in steel buildings were damaged, the design of steel moment connections in
High Seismic Applications changed significantly. As a result, currently three standards,
together, govern seismic design of steel moment connections in high seismic areas. These
are the AISC Specification (2010a), the AISC Seismic Provisions (2010b) and the AISC
Prequalified Moment Connections (2010c).
Figure 14.4 shows six common types of steel rigid or Fully-Restrained (FR)
moment connections for high-seismic applications. Connections (a), (e), and (f) require
field welding and connections (b), (c), and (d) are shop-welded filed-bolted connections.

(a) Welded Reduced Beam Section (b) Unstiffened Extended End Plate (c) Stiffened Extended End Plate

(d) Bolted Flange Plate (e) Welded Flange (f) Welded Flange Plate

Figure 14.4. Common Steel Moment Connections

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-6

14.4.a. Rotational Ductility of Moment Connections

One of the most important parameters in moment frames in high seismic
applications, is the rotational ductility of their moment connections, which is defined as the
Inter-Story Drift Angle. The inter-story drift angle is the cyclic rotation that a moment
connection can tolerate before the moment resisted by the connection is dropped below
0.80M p , where M p is the plastic moment capacity of the connected beam equal to F y Z x .
Depending on rotational ductility of connections, moment frames are divided into
Ordinary, Intermediate, and Special. In Ordinary Moment Frames (OMF), there is no
minimum rotational ductility requirement for connections. For moment connections in
Intermediate Moment Frames (IMF), the minimum interstory drift angle is 0.02 radian
before the moment in the connection is dropped below 0.80M p .
Moment connections in Special Moment Frames (SMF), should exhibit an inter-
story drift angle of at least 0.04 radians before the moment is dropped below 0.80M p . In
some tests in the past, instead of reporting interstory drift angle, the researchers have
reported plastic rotation of the connection. The interstory drift angle of 0.04 radians is
almost equivalent to plastic rotation of 0.03 radians.
Using special ductile moment connection in moment frames will result in moment
frames behaving in an inelastic ductile manner during major seismic events without
fracture of the connection elements. The inelasticity in the frame results in increased
damping and reduced stiffness, which in turn result in reduced seismic forces. In addition,
if the frame behaves in a ductile manner and no fracture occurs, the structure most likely
will return to its elastic condition after the earthquake shaking stops and will be able to
continue resisting the gravity loads without much P-∆ effects. For these reasons, seismic
design codes permit the use of smaller seismic forces for Special Moment Frames by
assigning a response modification coefficient of R equal to eight to them.
Steel moment frames are divided into “welded” or “bolted” frames. Figures 14.5
and 14.6, show examples of two moment frame structures with welded and bolted moment
connections respectively.

Photos by A. Astaneh-Asl
Photos by R. Pyle, AISC

Figure 14.5. A Welded Moment Frame Figure 14.6. A Bolted Moment Frame
Structure and Typical Connection Structure and Typical Connection
CHAPTER 14 MOMENT CONNECTIONS
14-7

14.5. SEISMIC BEHAVIOR OF WELDED STEEL

MOMENT CONNECTIONS (R > 3.0)
Welded moment connections are divided here into two categories of “Pre-Northridge” and
“SAC” connections. The Pre-Northridge moment connections refer to field-welded beam-
to-column connection shown in Figure 14.7 where the flanges of the beam were welded to
column in the field with complete joint penetration (CJP) groove welds. Prior to the 1994
Northridge earthquake, these connections were very common particularly in seismic areas
such as the west coast of the United States. However, the Northridge earthquake caused
cracks in the welds and adjacent base metal areas of more than 200 modern field-welded
steel structures built using this type of field-welded moment connections. Three examples
of damage are shown in Figure 14.7.

a
CJP

b
c
d

Figure 14.7. “Pre-Northridge” Welded Moment Connection, and Example

of Seismic Damage

Based on the results of SAC Joint Venture Steel Project and other research efforts,
the main reason for damage to the Pre- Northridge connections were:
1. Relatively low ductility of the CJP welds on the beam flanges
2. Chemical composition and mechanical properties of the base metal, in particular the
relatively high ratio of F y /F u of the base metals
3. Stress and strain concentrations in the connection area due to the geometry of the
connection as well as the weld access holes
4. Relatively large rotational demand imposed on the connections by the seismic event
5. Configuration of the global lateral load resisting system
6. Lack of stringent quality control and inspection in some cases
7. Effects of the ground motion characteristics and the closeness of the ruptured fault
to the structures.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-8

Following are the main recommendations resulting from the SAC Joint Venture
Steel Project and other research projects conducted since the 1994 Northridge earthquake:

1. To use more ductile filler material (i.e. weld electrodes). See the discussion in
Chapter 9 on seismic considerations in design of welds
2. To use steel, such as A992, with F y /F u not exceeding 0.85
3. To use improved access hole configurations
4. To use “Prequalified” connections that have been studied, tested and seismic design
procedures have been developed for them. For Prequalified connections see AISC
(2010c). The Prequalified connections should be able to tolerate cyclic inelastic
rotations of at least 0.03 radian, which is almost equivalent to interstory drift ratio of
0.04 radian
5. Consider using the new and sometimes patented connections that were developed
after the Northridge earthquake and were tested and Prequalified.
6. To better distribute lateral load resisting system in the plan by using moment
connections in more bays in the frames and on more frame lines
7. To provide quality control and quality assurance program according to the
provisions of the AISC Seismic Provisions (AISC, 2010b) and the AWS D1.8.
8. To design moment connections in Special and Intermediate moment frames to be
stronger than the expected –strain hardened bending strength of the beam. This is
to ensure that the connection has sufficient strength to remain almost elastic while
the beam section outside the connection undergoes plastification and strain
hardening. The exception to keeping connection elastic is allowing some controlled
shear yielding in the panel zone of the connection.

In order to ensure that the plastic hinge forms in the beam and not in the
connection, either the connection should be reinforced by adding flange plates and
brackets to it, or the plastic moment capacity of the beam should be reduced by
reducing the width of the flange, Figure 14.8.

(Research and Photo by M. Engelhardt et. al.)

Figure 14.8. Reduced Beam Section Moment Connection

CHAPTER 14 MOMENT CONNECTIONS
14-9

Using the above recommendations, a number of connections were designed and

tested by researchers as part of the SAC Joint Venture Steel Project (FEMA, 2000). The
connections that demonstrated sufficient strength, stiffness and more importantly, rotational
ductility and satisfied Item 8 in the above list, were “prequalified” by the SAC Joint
Venture Steel Project (FEMA-350, 2000) to be used in Special and Intermediate Moment
Frames.
Currently, the “Prequalified Connections for Special and Intermediate Steel
Moment Frames for Seismic Applications”, ANSI/AISC 358- 10 Standard, (AISC, 2010c),
provides provisions and guidelines for seismic design of Prequalified steel moment
connections. Table 14.1 below from this standard shows moment connections that are
currently “Prequalified.” The prequalified connections can be used without any need for
further testing.
Table 14.1

(of the AISC (2010b)

One of the standard prequalified connections is the welded Reduced Beam Section
(RBS) moment connection, which is covered in Chapter 5 of the ANSI/AISC-358-10
(AISC, 2010c). Currently, this connection is frequently used in High Seismic Areas in the
U.S.

14.6. SEISMIC DESIGN OF WELDED STEEL

MOMENT CONNECTIONS (R > 3.0)
Three documents provide information and guidelines on and govern seismic design of
welded Prequalified moment connections. The documents are (a) FEMA-350 (FEMA-
2000) report, (b) ANSI/AISC-358 Standard (AISC, 2010c) and AISC Seismic Provisions
(2010b). As mentioned earlier, Prequalified moment connections are connections that have
been studied and tested under cyclic loading and can be used in Special and Intermediate
moment frames in High Seismic Applications (R > 3.) Other connections that are not
Prequalified by the above documents, need to be tested following a standard testing
protocol , given in the AISC Seismic Provisions (AISC, 2010b) or shown by analysis that
their performance satisfies the requirements on performance of Special or Intermediate
moment connections.
The ANSI/AISC -358 Standard (2010c) has step-by-step procedures for design of
prequalified connections including the RBS connection.
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-10

14.7. SEISMIC BEHAVIOR OF BOLTED STEEL

MOMENT CONNECTIONS
Many existing riveted and bolted steel structures have been shaken by the earthquakes in
the past. No report of structural damage OR collapse of major riveted and bolted moment
resisting frames could be found in the literature. One of the early tests of seismic
performance of riveted steel structures was the 1906 San Francisco earthquake. In the post
earthquake reports and photographs taken in the aftermath of the 1906 quake, it appears
that there was no collapse or serious structural damage to riveted steel structures in
downtown San Francisco.
Figure 14.9 shows three prequalified bolted moment connections listed in Table
14.1

Bolted Flange Plate Bolted Unstiffened Extended Bolted Stiffened Extended

Connection End Plate Connections End Plate Connections

Figure 14.9. Common Bolted Moment Connections used in the United States

Studies of performance of steel structures during the 1994 Northridge and the 1995
Great Hanshin earthquake in Japan indicate very good performance of bolted steel
structures. It should be emphasized that most of the existing riveted and bolted MRFs were
not designed or detailed as Special Moment Frames and can be categorized as Intermediate
or Ordinary MRFs. Therefore it is expected that some of them could experience damage
during future major earthquakes. However, because of the relatively higher quality control
for bolted steel structures than for the field-welded structures, the likelihood of brittle
damage is low. In addition, in bolted moment connections, because of slippage of the bolts
and the gap opening and closing in the connections, bolted steel structures demonstrate a
certain amount of semi-rigidity is developed in the connections.
One of the reasons for good performance of bolted steel structures during
past earthquakes could be the semi-rigidity of the bolted connections. In most cases, the
semi-rigidity increases damping, releases and reduces stiffness, dissipates seismic energy,
isolates the mass of the floors from the accelerations in the columns, and elongates the
period, all of which reduce seismic response of the structure. These beneficial effects are
not present in welded moment connections.
CHAPTER 14 MOMENT CONNECTIONS
14-11

14.8. SEISMIC DESIGN OF BOLTED MOMENT

CONNECTIONS
14.8.a. Bolted Flange Plate Connections
Figure 14.10 shows details of a typical bolted flange plate connection. The top and bottom
flanges are welded to the column in the shop and field-bolted to the beam flange.
Continuity Plate

5/16"
C.J.P.
5/16"
A325 or A490 Bolts in Standard
Round Holes

Flange plates, A36 or A572 Grade 50

Finger shim plates, max. thickness =1/4 inch

Min. Fillet Weld Size = 0.75t ,

A325X or A490X Bolts in Standard Round Holes

Shear Plate, A35 or A572, Gr. 50

3” spacing tp WF Beam, A992 , Gr. 50 (preferred)

preferred

2”
5”
WF Column, A913, Gr.
50 or 65
Satisfy AISC requirements.

Satisfy AISC Requirements.

Figure 14.10. Typical Detail for Flange Plate Moment Connection

14.8.b. Failure Modes of Bolted Flange Plate Moment Connections

Failure modes of bolted flange plate moment connections are divided into categories of
ductile failure modes, failure modes with limited ductility and relatively brittle failure
modes. The failure modes are listed below and shown in Figure 14.11.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-12

Ductile Failure Modes of Bolted Flange Plate Moment Connections

1. Slippage of the flange bolts
2. Yielding of the gross area of the top and bottom flange plates
3. Bearing yielding of the bolt holes in the girder flanges and in the flange plates
4. Yielding of the gross area of the girder flange

Failure Modes of Bolted Flange Plate Moment Connections

(with Limited Ductility):
5. Local buckling of the top and bottom flange plates
6. Local buckling of the girder flanges
7. Shear yielding of the panel zone of the column

Brittle Failure Modes of Bolted Flange Plate Moment Connections:

8. Fracture of the edge distances or bolt spacing in the plate
9. Block shear failure of the flange plates
10. Fracture of the net section of the flange plate
11. Fracture of the edge distances or bolt spacing in the girder
12. Block shear failure of the girder flanges
13. Shear fracture of the flange bolts
14. Fracture of the welds connecting the top and bottom flange plates to column
15. Fracture of the net section of the girder flanges
Shear Yielding of the Fracture of Welds
Panel Zone
Fracture of Net Section of Plate
Yielding of Flange Plates
Fracture of Bolts
Bearing Yielding at Bolt Holes
Fracture of Edge Distance or
Slippage of Bolts Bolt Spacing in the Plate
Yielding of Beam Block Shear Failure of Plate

(a)
Fracture of
Net Section of the Girder

Block Shear Failure of Girder

Flange

Fracture of Edge Distance or Bolt

Spacing in the Flange

Local Buckling of Flange

Figure 14.11. (a) Ductile Failure Modes, (b) Failure Modes with Limited Ductility, and
(c) Brittle Failure Modes
CHAPTER 14 MOMENT CONNECTIONS
14-13

Yielding of Panel Zone Panel Zone Hinge

Plate Hinge
Yielding of Flange Plates and
Bearing Yielding of Bolt Holes Girder Hinge

Yielding of Beam

Drift Angle,
θ
O
Drift Angle,
θ

Figure 14.12. Three Areas of Plastic Hinge Figure 14.13. Three Hinges in
Formation in Bolted Flange Plate Connections the Connection

Seismic design of a bolted flange plate connection is done such that ductile
(yielding) failure modes govern over the brittle ones.
When a Bolted Flange Plate moment connection is subjected to bending, as listed
above, four yield mechanisms can occur and result in very ductile and desirable behavior of
the connection. Each yield mechanism results in formation of an inelastic zone ( partial
plastic hinge) either in the girder, flange plates or in the panel zone as shown in Figure
14.12. These partial hinges are identified as “Panel Zone Hinge,” “Plate Hinge” and
“Girder Hinge.” These three hinges are in a series, Figure 14.13, and rotations created by
these hinges are additive. Therefore, the best seismic design strategy would be to design
the connection such that all three hinges form during cyclic loading creating relatively large
rotational ductility for the connection. However, it must be mentioned that the most
desirable location for inelasticity is in the plate, then in the girder and later in the panel
zone. This is due to the fact that column panel zone is a global element responsible for
transfer of its share of vertical and horizontal forces of the column above to the column
below the connection. The flange plates and the girder are local elements and are only
responsible for transferring girder end moment.

14.10.c. Seismic Design of Bolted Flange Plate Moment Connections

The basic principle in developing seismic design procedures for bolted flange plate moment
connections is to ensure that yielding failure modes, which are ductile, occur prior to any
fracture failure mode. Based on this philosophy of making ductile failure modes governing
over the brittle ones, the ANSI/AISC-358 Standard (AISC, 2010c) in its Chapter 7 has step-
by-step procedures for seismic design of prequalified Bolted Flange Plate moment
connections.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-14

References for Chapter 14

1. AISC, (2010a) “Specifications for Structural Steel Buildings,” American Institute of Steel
Construction, Chicago.
2. AISC, (2011) “Manual of Steel Construction,” American Institute of Steel Construction,
Chicago.
3. . AISC, (2010b) “Seismic Provisions for Structural Steel Buildings,” American Institute of
Steel Construction, Chicago.
4. AISC, (2010c), “Prequalified Connections for Special and Intermediate Steel Moment
Frames for Seismic Applications”, ANSI/AISC 358-05 Standard, published by the
American Institute of Steel Construction, Chicago.
5. Astaneh-Asl, A., (1986), "Field Bolted Moment Connection,” Proceedings, National Steel
Construction Conference, AISC, Nashville, Tenn. , June.
6. FEMA-351A, (2000). “State of the Art Report on Base Metals and Fracture”, Federal
Emergency Management Agency (http://www.nehrp.gov/pdf/fema355a.pdf)
7. Pinkney, R. B. and Popov, E. P., (1967), "Behavior of Steel Building Connections
Subjected to Repeated Inelastic Strain Reversal- Experimental Data", Report No.,
UCB/SEMM 67-31. Dept. of Civil Engineering, University of California, Berkeley.
8. SAC, (2000), "Recommended Criteria, 100% Draft,” Nearing Publication, SAC Joint,
Venture Steel Project , Sacramento.
9. Schneider, S.P. and Teeraparbwong, I, (2000) “Bolted Flange Plate Connections,” Report
#355d, FEMA , (A Report on SAC Project Task 7.09,) Washington.
CHAPTER 14 MOMENT CONNECTIONS
14-15

Problems for Chapter 14

Problem 14.1. Design a prequalified Reduced Beam Section (RBS) welded moment
connection according to ANSI/AISC-358-10. The beam is a W27x102, A992 Gr. 50 and
column is a W12x120, A913 Gr. 65 ksi.
Problem 14.2. Design a prequalified Bolted Flange Plate (BFP) moment connection
according to ANSI/AISC-358-10. The beam is a W27x102, A992 Gr. 50 and column is a
W12x120, A913 Gr. 65 ksi.
Problem 14.3. Design a prequalified Reduced Beam Section (RBS) welded moment
connection according to ANSI/AISC-358-10. The beam is a W30x292, A992 Gr. 50 and
column is a W14x257, A913 Gr. 65 ksi.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
BEHAVIOR AND DESIGN OF STEEL STRUCTURES
14-16
CHAPTER 15-STEEL STRUCTURAL SYSTEMS 15-1

CHAPTER 15

STEEL STRUCTURAL SYSTEMS

Objectives:
In this chapter you will be introduced to:
1. Common Steel Structural Systems

15.1. INTRODUCTION TO STEEL STRUCTURAL

SYSTEMS
Figure 15.1 shows the most common steel structural systems for buildings, which are:

1. Concentrically Braced Frames

2. Eccentrically Braced Frames
3. Moment Frames
4. Steel Shear Walls
5. Dual Systems

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2013 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
20-2
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Steel Concentrically Steel Eccentrically Steel Moment Steel and Composite

Braced Frames Braced Frames Frames Shear Wall System

Figure 15.1. Most Common Steel Structural Systems

Figure 15.2 shows other steel and composite structural systems that are used
most often in the mid- to high-rise buildings. Composite structural systems are
combination of steel and concrete used in members. These are:

1. Tube
2. Core plus Outrigger
3. Tube in Tube
4. Bundled Tube
5. Mega Frame
6. Mega Truss

In the following sections we briefly discuss main features of the relatively less
common steel and composite systems shown in Figure 15.2.

2. Core plus 3. Tube-in-Tube 4. Bundled 5. Mega Frame 6. Mega Truss

1. Tube
Outrigger System
System System Tube System System
System

Figure 15.2. Examples of Common Steel and Composite Building Framing Systems
CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-3

15.2. TUBE SYSTEM

Tube systems consists of a complete perimeter moment frame with closely spaced
columns. Because of closeness of the columns and short spans of the beams, the perimeter
frame acts more-or-less as a tube with windows cut out of it. The lateral stiffness of the
system is primarily due to shear and bending stiffness of the columns and to a greater
extent due to shear stiffness of the relatively short span girders of the perimeter moment
frames. Figure 15.3 shows an example of a framed Tube system. The columns in this
system can be steel or composite, (or R/C) , but the girders are usually steel (or R/C in
R/C tubes). One of the important features of the tube system is that the gravity load in the
tube stabilizes the windward columns against uplift when the structure is subjected to
lateral loads.

nycarchitecture.com

Figure 15.3. Buildings with Tube System

15.3. CORE PLUS OUTRIGGER SYSTEM

This system consists of a core braced frame and horizontal “outrigger” trusses connecting
the braced core to the exterior columns. The role of outrigger trusses is to mobilize the
exterior columns and use their axial stiffness and strength to increase bending stiffness and
strength of the system. Figure 15.4 shows an example of core plus outrigger system. In
recent years, with the advent of steel and composite shear walls, instead of core braced
frame and horizontal outrigger trusses, steel or composite shear walls are used in the core
as well as outriggers.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
20-4
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Figure 15.4. Taipei 101, Which has Core Plus Outrigger Truss System

15.4. TUBE IN TUBE SYSTEM

This system consists of two tubes placed inside each other and connected to each other
with floor framing system. The bending stiffness and strength of the floor framing system
including floor slabs, plays a major role in the lateral stiffness of the system. If the floor
framing system is relatively rigid, the two tubes will work together as a single system and
bend as a beam where plane sections (i.e. floor framing system) remain plane. If the floor
framing system is relatively flexible, the two tubes will act almost independently and
provide their own bending and shear stiffness and strength to the overall system. If
stiffness of floor framing system is not sufficient to force the tubes to bend together with
one curvature, outrigger trusses can be placed between the two tubes. Such trusses in tall
buildings usually are placed in mechanical floors where space interruptions can be handled
much easier than the occupied floors.

15.5. BUNDLED TUBE

This system consists of two or more tube systems connected to each other to form a single
structural system. The bundled tubes share the same common “tube wall” wherever they
are in contact. One of the most important structures, where this system is used is the Sears
Tower in Chicago shown in Figure 15.5.
CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-5

Skyscrapercity.com

Figure 15.5 Sears Tower (Now Willis Tower) Has a Bundled Tube structural System

15.6. MEGA TRUSS FRAME SYSTEM

The system, which has been used in Japan, consists of very large moment frames in which
columns are several floors high and girders are several spans long, see Figure 15.6. The
columns and beams in the mega frame are trusses. The gravity load in the columns of
several floors between the truss girder are supported on the truss girders of the Mega Truss
Frame.
Figure 15.7 shows a case of mega truss frame used in Japan. The chords of the
vertical trusses (i.e. columns of mega truss frame) were steel built-up box section. The
chords of the horizontal trusses (i.e. girders of mega truss frame), as well as all diagonals
were wide flanges. The connections were generally complete joint penetration groove
welds. A residential complex in Kobe Japan has a large number of buildings built using
this system. The 1995 Kobe earthquake fractured several of the vertical truss chord
members (box columns) and caused damage to diagonals of the truss girder as shown in
Figure 15.7.
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
20-6
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Figure 15.6. Mega Truss Frame System

Fracture of the
vertical member

Yielding and local

buckling of diagonals

Photos by Abolhassan Astaneh-Asl

Figure 15.7. Mega Truss Frame System in Japan, Damaged during the 1995 Kobe
Earthquake
(Ref. Kanada and Astaneh-Asl, 1995)

15.7. MEGA TRUSS SYSTEM

This system either has an exterior perimeter truss system, Figure 15.8(a, left), or a 3-
dimensional truss system, Figure 15.8(b, right). In this system, the perimeter or 3-D
trusses carry the gravity loads of the interior columns to the exterior and in most cases to
the corner columns. The very large gravity load in these mega columns stabilizes the
structure against uplift of windward side and overturning.
CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-7

Futurecrue.com
Figure 15.8. High-Rises with Steel (left) and Composite (right) Braced Tube Systems

.
15.8. COUPLED SYSTEMS
Generally, this system consists of two buildings connected (coupled) to each other by one
or more bridges. Coupling the structures results in different and improved response of the
coupled structures than the response of individual structures would be if they were
separate. One of the most important coupled systems is the Petronas Towers in alesia
shown in Figure 15.9. The two towers have reinforced concrete columns and core shear
walls and steel floor framing system.

Figure 15.9. Petronas Towers

15.9. STEEL BEARING WALL SYSTEM

The bearing wall system has been used in several buildings, but its use has not been
widespread. The system consists of an external steel wall, internal core columns and floor
truss joists connecting the exterior steel bearing wall to the interior gravity columns. One
of the most important buildings, where this system was used was the two 110-story World
Trade Center. The towers were attacked by the terrorists on September 11, 1501 and
collapsed. Figure 15.10 shows structure of the World Trade Center.

Steel Bearing Wall System of the WTC

The WTC Towers
Before Collapse Steel Bearing Walls of the WTC

Figure 15.10. The WTC Steel Bearing Wall System

(Astaneh-Asl, 2001-2007)

References for Chapter 15

1. AISC. (2001), “Staggered Framing Truss System”, Steel Design Guide No. 14, AISC,
Chicago.
2. Astaneh-Asl, A., (2001-2007) “World Trade Center Studies”, A research project conducted
by A. Astaneh-Asl, J. Son, C. Heydari, and Q. Zhao.
3. Angell, S. and Faraone, T., (2002), Structural Engineer, November 2002 Issue, Mercor
Media Inc. Alpharetta, Georgia.
4. Beedle, L.S., (1996). “Tall Buildings Structures-A World View”, Council on Tall Buildings
and Urban Habitat, Lehigh University.
5. Kanada, M. and Astaneh-Asl, A. (1995) “Seismic Performance of Steel Buildings during
the 1995 Hanshin Earthquake”, Report No. UCB/CEE-steel-95/02, Dept. of Civil and Env.
Engineering, Univ. of California, Berkeley, May.
6. Smith, B.S. and Coull, A. (1991). “Tall Building Structures”, John Willey & Sons Inc.
CHAPTER 16-CONCENTRICALLY BRACED FRAMES 16-1

CHAPTER 16

CONCENTRICALLY BRACED
FRAMES

Objectives:
In this chapter we discuss:
1. Types of steel concentrically braced frames
2. Design of Special Concentrically Braced Frames for High
Seismic Applications

16.1. INTRODUCTION
The main function of a steel concentrically braced frame system in a building is to
provide lateral strength and lateral stiffness by acting primarily as a vertical truss, Figure
16.1. The name concentrically braced frame implies that in this system, like trusses, the
members are concentric and centerlines of all members joining each other at one node pass
through single point called work point, see Figure 16.1(c).
Concentrically braced frames are one of the most efficient and economical systems
since their stiffness and strength is provided by the axial strength and axial stiffness of their
members. In an axially loaded member, all cross sections over the entire length of the
member are subjected to the same level of stress, therefore, the entire volume of the
material used in the member is participating equally in resisting the loads as well as in
providing stiffness. In flexural members, only a relatively small fraction of the material in
cross sections where bending moment is the highest, is subjected to high stresses and the
rest of the member is subjected to relatively small stresses. Even in cross sections with
maximum moment; only extreme fibers are subjected to maximum stresses and the rest of
the cross section is under much less stress.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003- 2012 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
16-2
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Diagonals of
Bracing
Truss
Members

Work
Point

Chords of Columns
Truss

Beams
Verticals of
Truss

Typical Truss Typical Braced Frame

(a) (b) (c)

Figure 16.1. A Typical Truss Compared to a Typical Concentrically Braced Frame

A typical concentrically braced frame consists of vertical columns, horizontal

beams and diagonal braces as shown in Figure 16.1(b). The columns , which can be steel
or composite, generally are axially loaded columns carrying a combination of gravity and
lateral loads. The beams, also steel section or more often steel sections composite with the
floor decks and acting as a beam column resist a combination of gravity and lateral loads.
The diagonal bracing members, in general, act as axially loaded members. In mega-truss
braced systems, discussed later, diagonals also participate in carrying gravity loads.

16.1.a. Advantages of Concentrically Braced Frames

Some of the advantages of using concentrically braced frame systems are:

1. In general, concentrically braced frame systems are very stiff compared to other
systems such as steel and composite shear walls or eccentrically braced frames and
especially moment frames.
2. Concentrically braced frame systems are very economical in resisting lateral loads.
3. Concentrically braced frame systems have relatively high elastic initial stiffness,
maintaining it up to buckling of compression brace resulting in very desirable drift
control throughout the service life.
4. Compared to reinforced concrete shear walls, the steel concentrically braced frame
is much lighter resulting in less weight to be carried by the columns and
foundations. In addition, similar to steel shear walls, due to reduced mass of the
structure, less seismic forces will be generated in the building CBF during
earthquakes by the self weight of the bracing system itself.
5. Shop-welded, field-bolted concentrically braced frame system are quite common,
especially in non-seismic areas and for wind lateral load resistance. The shop-
welded field bolted concentrically braced frames can result in further saving in the
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-3

cost of lateral load resisting system. In addition, the field bolted structures can be
constructed throughout the year almost independent of the temperatures at the
construction site.
6. The system would also be very practical and efficient for cold regions where
concrete construction is neither economical nor sometimes feasible in very cold
seasons.
Concentrically braced frame systems have been used since the early years of
modern steel skeleton construction and they continue to be used in steel structures from
one story to very tall skyscrapers such as the Bank of China in Honk Kong with a
composite mega truss concentrically braced frame system.
Braced frames to some extent allow door and window openings, Figure 16.2, but
in general, the best place for braced frames is around the elevator shafts and stair wells
where very limited openings in the walls are needed.

Opening Opening

Bracing

a. Stiffened Steel Shear Wall with a. Stiffened Steel Shear Wall with
Opening Opening

Opening Opening

Bracing Bracing

a. Stiffened Steel Shear Wall with a. Stiffened Steel Shear Wall with
Opening Opening

Figure 16.2. Openings in Braced Bays

16.2. CONFIGURATIONS OF CONCENTRICALLY

BRACED FRAMES
The possibilities for configuration of braced frames are unlimited and quite often
the configuration of a braced frame system is driven more by architectural demands and
restraints than from the structural considerations. Figures 16.3 and 16.4 show several
configurations of braced frame systems which are:

a. X-Bracing
b. Chevron Bracing
c. V-Bracing
d. Multi Floor X-Bracing
e. Zipper Bracing
f. Bracing with Outrigger
g. Suspension bracing system
h.Truss frames
i. Mega Frame System
j. Mega Truss System

(a) X-Bracing (b) Chevron-Bracing (c ) V-Bracing (d) Multi-Floor (e) Zipper-

X-Bracing Bracing

Figure 16.3. Types of Concentrically Braced Frame Systems Used in Buildings

(f) Outrigger (g) Suspended (i) Mega Frame

(h) Truss Frames (j) Mega Truss
System Braced Frames System
System

Figure 16.4. Types of Concentrically Braced Frames Systems used in Tall Buildings
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-5

16.2.a. Single Bay-Single Floor X-bracing, Chevron, V-bracing and

Multi-Floor bracing
Shown in Figure 16.3(a), (b),(c) and (d), the single bay concentrically braced
frames are most common types of bracings for low and mid-rise buildings. The braced bay
resists lateral load as well as over tuning moment. As the height of the building increases,
the overturning moment increases. At one point the overturning moment exceeds the
stabilizing moment of the gravity load and the windward columns start uplifting. To
prevent building from toppling because of the uplift, the windward columns need either be
tied-down to foundation and foundation tied down to the ground, or more gravity load be
transferred to the boundary columns of the single bay braced frame.
Tying down the columns to the foundations often becomes a relatively expensive
solution. If building has basements, the columns can be embedded in the basement walls
preventing them from uplift. The solution of transferring gravity load to boundary columns
of braced frame can be done by widening the braced bay and using mega truss system
shown in Figure 16.4(j) or by using mega frame shown in Figure 16.4(i).
In X-bracing, quite often the bay length is longer than the floor height making the
angle of the bracing member with the beam quite small and the bracing less efficient.
Chevron brace system has been quite popular in recent years since it allows opening for the
doors at mid panel as well as makes the angle of bracing members with the beam near 45
degrees. However, chevron braces create relatively large unbalanced vertical force in the
beam when the compression brace buckles.

16.2.b. Effective Length Factor in X-Bracing Members

When an X-braced bay is subjected to story shear, one of the braces of the
X will be in tension and the other will be in compression. If the braces are
connected to each other at the intersection of the X, which is the case in most
applications, the question is what effective length factor , K, should be used for the
bracing members? The effective length depends on how much rotational stiffness
is provided to the compressive diagonal brace by the tensile diagonal brace as well
as rotational stiffness of the brace end connections. To find an answer, we need to
look to in-plane buckling and out of plane buckling of the compressive diagonal
separately.
For in-plane buckling of braces, depending on the end connection of
bracing members, the buckled geometry will be close to one of the two cases
shown in Figure 16.5. Using either the six approximate cases or the more accurate
alignment charts , both discussed in Chapter 5, a reasonable effective length factor,
K, can be established for the bracing members buckling in-plane of the braced
frame.
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
16-6
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

L L

Brace End Connections and Brace End Connections and

X-Intersection Connections are Pin X-Intersection Connections are Rigid
K= 0.50 (Theoretical) and K=0.65 (for Design) K=0.35 (Theoretical) and K=0.40 (for Design)

In-Plane Buckling of Braced Panels

Figure 16.5 Values of Effective Length Factor for Bracing

Members in In-Plane Buckling X-bracings

Currently, in most concentrically braced frames, the end connections of the

braces are single gusset plates that have relatively large in-plane rotational stiffness
providing almost fully rigid end conditions for the brace buckling in plane of the
frame. If small rotation of the gusset plates due to lateral movement of one floor
relative to the floor below is ignored, for in-plane buckling, the ends of the braces
can be assumed to be rigid. However, the connection at the intersection of X-braces
is expected to behave more or less as a pin if only one bolt is used, or behave as a
rigid joint if weld lines or more than one bolt are used at the intersection. In a
conservative way, we can assume the end connections as well as the intersection act
as pin connections, shown in Figure 16.5(a). With this assumption, the theoretical
K for half length of the brace will be 1.0 if half-length of the brace is used as
length of the brace and equal to 0.50 if full length of brace, length L in Figure
16.5(a) is used. The Design K value of 0.65 can be used with full length L of the
brace, Figure 16.5(a).
If the end connections of in-plane buckling braces are assumed to be rigid
and the intersection connection to be pin, which are more realistic assumption than
all connections assumed to be pin, then a Design K value of 0.40 can be used with
full length L of the brace.
For out-of-plane buckling of braces, the effective length factor K in an X-
brace will also depend on the rotational rigidity of the end and intersection
connections. However, in this case, the amount of rotational and translational
stiffness that is provided by the tension bracing member to the compression bracing
member also affects the value of K. Figure 16.6 from Stoman (1988) shows the
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-7

model of X-bracing used to study elastic buckling of X-braced panels and to

establish K for the compression brace member. In this model it is assumed that at
the instant of buckling of compression diagonal under the load P, the tension force
in the tension bracing member is Q. The lengths of the tension and compression
braces are given as L t and L c respectively.

Figure 16.6. (a), (b) and (c), Elastic Model of X-braced Bay and
(d) and (e) ; Modes of Buckling (Source: Stoman, 1988)

Stoman (1988), using the models shown in Figure 16.6 and assuming elastic
behavior for both braces until buckling of the compressive brace occurs, derived
the following equations for K of out-of-plane buckling X-braces where both
tension and compression diagonals have the same length, same cross section and
are made of the same steel material.
(a) If ends of the bracing members are simple, which is applicable to single
gusset plate cases:

(1 − Q / Pe )
K= ≥ 0.50 (16.1a)
2
(b) If ends of the bracing members are rigid, which is applicable to double
gusset plate cases:

(1 − Q / Pe )
K= ≥ 0.35 (16.1b)
8
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
16-8
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

where
K = Effective length factor
Q = Tension force in tension bracing member
P e = Euler’s elastic buckling load for a pin-pin column given by:

π 2 EI (16.2)
Pe = 2
L

One of the important findings of Stoman(1988) was that as long as Q, the force in
the tensile brace, is greater than or equal to 0.2P e for braces with pin end connections and
0.183 for braces with rigid end connections, the compression brace will buckle in second
mode. Buckling of braces in second mode implies that a K value of 0.50 can be used with
a length being the total length of the brace..
In design, for X-braces made of single angle diagonal braces, El Tayem and Goel
(1986) suggest a K value of 0.425 for diagonal braces. The length of the brace is the total
length. For X-bracings made of other cross sections a conservative theoretical value of
0.50 corresponding to a design value of 0.65 can be justified.

Traditional Brace Member Buckling Restrained Brace Member

(a) (b)
Viscous or Friction Dampers Hysteretic Damper

Figure 16.7. Four Types of Concentrically Braced Frames Used in High Seismic Applications
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-9

16.3. TYPES OF BRACED BAYS

In the remainder of this chapter we focus on seismic applications of concentrically
braced frames. Examples of types of braced frame bays for seismic applications are shown
in Figure 16.7. They are:

1. Conventional concentrically braced frame

2. Buckling restrained concentrically braced frame
3. Concentrically braced frame with viscous or friction damping devices
4. Concentrically braced frames with hysteretic damping devices

16.4. CONCENTRICALLY BRACED FRAMES

Currently the most commonly used concentrically braced frame in HSA are
conventional (X, V and Chevron) and buckling restrained braced frames, Figures 16.7(a)
and (b), with the latter replacing the conventional concentrically braced frames in more
and more applications.
In this chapter, the name “concentrically braced frame” refers to the conventions x-
V and chevron concentrically braced frames.
The beams in a concentrically braced frame are modeled as pin-connected to the
columns while columns are modeled as continuous members. As discussed in Chapter 19-
Trusses, the assumption of pin connection for the beam-to-column connections of a
concentrically braced frame with relatively large gusset plates is not realistic. In reality,
the connections are more or less rigid connections or at least semi-rigid connections
discussed in Chapter 15. However, in design office when modeling braced frames, similar
to beams, the ends of the bracing members are modeled as pin connections. In dual
systems, where bracing members are inside a moment frame, making the system dual, the
connections can be modeled as moment connections and the bracing members be designed
as beam-columns with both bending moment and axial load considered in their design.
There are two types of concentrically braced frames: Ordinary and Special. The
main differences between these two systems are in the ductility of their bracing members
and connections and the seismic response parameters, which are Response Modification
Coefficient, R, Overstrength Factor, Ω o and Deflection Amplification Factor C d . Values of
R, Ω o and C d for Special and Ordinary CBFs are 6, 2, 5 and 3.25, 2, 3.25 respectively
(ASCE-7, 2005). By using SCBF instead of OCBF, seismic loads will be cut almost in
half (3.25/6= 0.54). However, the expected deflections (drift) of SBCFs will be about 1.5
times (5/3.25=1.53) the deflections of the OCBFs. Since in braced frames, with their
relatively high lateral stiffness, drift is not usually an issue, the use of Special
Concentrically Braced Frame in most cases will result in more economical design not only

for the frame but also for the supporting foundations and piles (if any). In High Seismic
Applications (HAS), the use of Ordinary Concentrically Braced Frames (OCBF) is very
limited and there are no reports of the use of OCBF in important structures in HSA.
In special concentrically braced frames (SCBF), the connections are designed and
detailed to develop the lesser of the expected yield strength R y A g F y of the brace member or
the force that can be transmitted to the connection by the system. Gusset plates in special
concentrically braced frames should have detailing that will allow free space equal to 2t
beyond the termination of the member for formation of unrestrained plastic hinge in the
gusset plates if the brace is buckling out of plane of the braced frame. Design of gusset
plates for Special Concentrically Braced Frames(SCBF) was discussed in Chapter 17
earlier.
Figure 16.8 shows examples of suggested details for gusset plates of Special
Concentrically Braced Frames from the Steel TIPS report (www.steeltips.org) by Astaneh-
Asl, Cochran and Sabelli (2006). The Steel TIPS has more information on the design and
detailing of gusset plates for SCBFs including more suggested details for gusset plate
connections in SCBF as well as detailing information.

Bolts

Figure 16.8. Suggested Details for Special Concentrically Braced Frames

(From Steel TIPS by Astaneh-Asl, Cochran and Sabelli, 2006 at www.steeltips.org for free download)
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-11

Figure 16.8(Continued). Suggested Details for Special Concentrically Braced Frames

(From Steel TIPS by Astaneh-Asl, Cochran and Sabelli, 2006 at www.steeltips.org for free download)

The above details are for Special Concentrically Braced Frame Systems and only
for cases that the bracing member buckles out-of-plane of the braced frame requiring the
minimum “2t” distance for
formation of plastic hinge in the
gusset plates. If bracing member
buckles in-plane of the braced
frame, there is no need for the
minimum “2t” distance for
formation of plastic hinge in the
gusset plates since the plastic
hinge in this case will form in the
bracing member just outside the
gusset plate. Figure 16.9 shows
an example of such details.

Figure 16.9. An Example of Gusset Plates in Special

Concentrically Braced Frames
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
16-12
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

16.5. DESIGN OF SPECIAL CONCENTRICALLY

BRACED FRAMES
Special Concentrically Braced Frames (SCBF) have the following main elements:

(1) Bracing Members

(2) Connection of Bracing Members to the Gusset Plates
(3) Gusset Plates
(4) Connection of Gusset Plates to beams and columns
(5) Mid-point Connections of Bracing Members in X-braced Frames
(6) Columns
(7) Beams
(8) Beam-to-Column Connections

The design philosophy in Special Concentrically Braced Frames is to design the

bracing member for the applied loads and make sure ductile failure mode of axial yielding
in tension and overall buckling in compression governs. Then, design all other elements of
the braced frame as listed above to have a strength equal or more than the lesser of the
expected strength of the brace or the load that can be delivered to them by the system.
Following this design criteria results in the bracing member to be the ductile fuse in
Special Concentrically Braced Frames.
Following sections provide a summary of design procedures and requirements for
elements of the SCBFs.

16.6. DESIGN OF BRACING MEMBERS IN SCBF

Bracing members in SCBFs are designed to resist axial tension and compression forces
established by the analysis of the structure subjected to combined gravity and seismic loads
as given by the ASCE-7 (2005). The load combinations that include seismic loads are:

(1.2+0.2S DS ) D+ ρQ E + L + 0.2S
(16.3)
(0.9-0.2 S DS ) D + ρQ E + 1.6H (16.4)

where:
D= Dead load
H= load due to lateral earth pressure, ground water pressure, or pressure of bulk material
L= Live load
Q E = Effect of horizontal seismic load
S =Snow load
ρ = Redundancy factor ( see Section 12.3.4 of the ASCE-7)
S DS =The design spectral response acceleration parameter in short period range as
determined from Section 11.4 of the ASCE-7(2010).
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-13

In addition to above load combinations, there are amplified seismic load

combinations. The amplified seismic load combinations are used in strength design of
certain connections and members. These connections and members are specified in the
AISC Seismic Provisions (2010) for each system.
The purpose of using amplified seismic load combinations is to incorporate the
inherent overstrength of the system in strength design of certain connections and elements
so that when the system develops its overstrength, these connections and members still can
resist forces resulting from the system overstrength. The amplified seismic load
combinations are similar to the last two load combinations in the above list but in this
case, the earthquake load Q E is multiplied by the Overstrength Factor, Ω o , instead of the
ρ. The overstrength factors Ω o for common systems are given in the ASCE-7(2010),
Section 12.2. As mentioned earlier, the overstrength factor, Ω o for Special Concentrically
Braced Frames is 2.0.
The amplified seismic load combinations are:

(1.2+0.2S DS ) D+ Ω o Q E + L + 0.2S (16.5)

(0.9-0.2 S DS ) D + Ω o Q E + 1.6H
(16.6)
The terms in the above equations were defined in previous page.

16.7. DESIGN REQUIREMENTS FOR BRACING

MEMBERS IN SCBF
After establishing maximum axial tension and compression forces in a bracing
member, it is designed to resist the forces while satisfying the following requirements.
These requirements are imposed to ensure that the bracing member will be the ductile fuse
in the system and all other elements will remain essentially elastic as the bracing member
undergoes inelastic cyclic axial (push-pull) deformations.
The design requirements for the bracing members of Special Concentrically
Braced Frames are discussed in the following sections.

16.7.a. Lateral Force Distribution of in SCBF

Tension-only braces are not allowed in SCBFs. Special Concentrically Braced Frames
should have tension and compression bracing members in each floor and along all lines of
bracing.

In this context, the line of bracing is defined as a single line of bracing or parallel
lines of bracings if the distance between the parallel lines of bracing is not more than 10%
of building dimension perpendicular to parallel lines of bracing.
Along any line of bracing, tension and compression braces should be designed
such that at least 30% and no more than 70% of the shear transmitted by the braces is
resisted by the tension braces. The exception to this requirement is if the available
strength of each brace in compression, is larger than the required strength to resist
maximum compression resulting from the load combinations in the governing code
including the amplified seismic load.

16.7.b. Overall Buckling of the Bracing Member

Bracing members in SCBF should be designed, following the provisions of the AISC
Specification (2010) to have available buckling strength, φP c , not less than the required
compressive strength resulting from the analysis of the structure using load combinations
given by Equations 16.3 through 16.6 earlier.
The effective slenderness ratio of the bracing member should satisfy:

K / r ≤ 4 E / F y (16.7)

Braces with 4 E / F y < K / r ≤ 200 are allowed if available strength of the

column (in LRFD) is at least equal to the maximum load transferred to the column
considering R y times the nominal strength of the connecting brace elements.
The length of a bracing member is taken as work-point to work-point of the two
ends of the brace. However, the AISC Seismic Design Manual (2006) suggests that “The
Owner’s Designated Representative for Design” (presumably The Engineer of the Record)
can use shorter length for the bracing member if justified.
For in-plane buckling braces, the clear length of the brace between the gusset
plates is a justifiable length since plastic hinges in these braces form just outside the gusset
plates.
For out-of-plane buckling bracings, the distance from the two “2t” zones, where
the plastic hinges form in the gusset plates is a justifiable length. The choice of length of
the bracing member is important since the length of bracing member affects its capacity
with a power of two. Using shorter but justifiable length will result in lighter sections for
bracing members, which in turn will result in all other elements of the braced frame to be
lighter since they are all designed for the expected strength of the brace section.
Built-up members, when used as bracing members in SCBF, should be designed
following the provisions of the AISC Specification (2010), which were discussed in
Chapter 5 (starting on Page 5-8). However, the following additional requirements should
also be met.
(a) If buckling of brace causes shear to develop in the stitches, the slenderness
ratio of individual segments a/r i between the stitches should satisfy:

 a   KL 
  ≤ (0.4) 
 ri   r m
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-15

(16.8)
where
 KL  = Governing effective slenderness of the member
 
 r m
a = Distance between stitches, see Figure 16.10 below
r i = Minimum radius of gyration of individual segment

(b) If buckling of brace causes no shear to develop in the stitches, the slenderness
ratio of individual segments a/r i between the stitches should satisfy:

 a   KL 
  ≤ (0.75) 
 ri   r m (16.9)
(c) The sum of shear strength of the stitches, shown in Figure 16.6 should not be
less than the available tensile strength of each segment. The exception is
where the buckling of braces about their governing axis does not cause shear in
stitches. In braces using single gusset plate, when braces buckle in-plane of
the frame, no significant shear is developed in the stitches.
(d) At least two stitches should be used and stitches should be equally spaced
throughout the length of the brace. No stitch should be placed at mid-length of
the brace. Combining the two latter requirements means that only even
number of stitches equally spaced should be used.

Stitch

Stitch a a
L a

y
y
Shear

y
y
_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2013, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
16-16
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Figure 16.10. Examples of Built-up Bracing Members and Development of Shear

16.7.c.inLocal
StitchesBuckling
of Out-of-Plane Buckling
of the Braces
Bracing Member
(Astaneh-Asl, Goel and Hanson 1982)

The b/t and d/t ratios of bracing members in SCBF should satisfy:
b/t ≤ λ ps (16.10)
Where, λ ps. is the limit for cyclic local buckling given in Table I-8-1 of the AISC
Seismic Provisions (2005). The table was given in Chapter 5 (see Page 5-38).

16.7.d. Tensile Strength of the Bracing Member

Seismic design of tension members was discussed in Chapter 4 starting on Page 5-34.
Following is a summary of that discussion and relevant equations.
Following three limit states should be considered in design of tension bracing members
of SCBF subjected to tension:
As discussed earlier in Chapter 4, steel and composite tension members have four failure
modes:
1. Tensile yielding in the gross section
2. Tensile rupture in the net section
3. Block shear rupture

16.7.d.1- Tensile yielding in the gross section

Bracing members in SCBF should be designed, following the provisions of the AISC
Specification (2010), to satisfy:

Tu ≤ φt Pn (in LRFD) (16.11)

where
T u = Required axial tensile strength in LRFD, same as the factored applied tension force
resulting from the analysis of the structure using load combinations given by
Equations 16.3 through 16.6 earlier.
φ t = Resistance factor for tensile yielding in LRFD = 0.90
P n = Available nominal axial yield strength of the gross section of the member given by:

Pn = Fy Ag
(16.12)

A g = Gross area equal to the total cross-sectional area of the member.

F y = Specified minimum yield stress.

16.7.d.2- Tensile rupture in the net section

CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-17

The available strength of the bracing members for limit state of tensile rupture in
the net section should satisfy:
T u-exp ≤ P n (in LRFD) (16.13)
where
T u-exp = Required axial tensile strength in LRFD given as lesser of the two following
values:

(a) The expected yield strength, in tension, of the bracing member, determined as
R y F y A g (in LRFD format).
(b) The maximum tension force (in LRFD format), indicated by analysis that can
be transferred to the brace by the system.

P n = Expected available nominal axial strength of the member based on limit state of
rupture in the net area given by:

Pn = RtFu Ae (16.14)
where
F u = Specified minimum tensile strength of steel.
A e = Effective net area given by:

Ae = An U (16.15)

A n = Net area of the member discussed in Section 4.10 in Chapter 4 (starting on Page 4-7)
U = Shear lag factor discussed in Chapter 4 (starting on Page 4-10)

16.7.d.3-Block shear rupture

It should be noted that even though the AISC Seismic Provisions (2005) has specific
requirements only for rupture in the net area as discussed above, the limit state of block
shear rupture should also be checked.
For block shear rupture the following should be satisfied:

T u-exp ≤ R n (in LRFD) (16.16)

where
T u-exp = Same as defined in previous page
R n = Expected available nominal axial strength of the member based on limit state of block
shear rupture given by:

R n = 0.6 R t F u A nv + U bs R t F u A nt ≤ 0.6 R y F y A gv + U bs R t F u A nt
(16.17)

and
A nv = Net area subject to shear, (see Section 4.12 in Chapter 4)
A nt = Net area subject to tension, (see Section 4.12 in Chapter 4)
A gv = Gross area subject to shear, (see Section 4.12 in Chapter 4)
U bs = A factor that depends on how uniform the tensile stress across the section is.
U bs = 1.0 when the tension stress is uniform and
U bs = 0.5 when tensile stress is non-uniform, (see Section 4.12 in Chapter 4)

16.8. DESIGN REQUIREMENTS FOR BRACING

CONNECTIONS IN SCBF
16.8.a. Required Tensile Strength
The required tensile strength of the bracing connections in SCBF is the lesser of the
following:

The available tensile strength of the bracing connections, φR n , calculated using

provisions of the AISC Specification (2010) should not be less than the above required
tensile strength.

16.8.b. Required Flexural Strength

The required flexural strength of the bracing connections in SCBF is equal to 1.1R y M p (in
LRFD) of the bracing member.
The available flexural strength of the bracing connections, φM n , calculated using
provisions of the AISC Specification (2010) should not be less than the required flexural
strength.
Bracing connections that satisfy the above requirement for tensile strength (Section
16.9.a) and can accommodate the inelastic rotations associated with brace post-buckling
deformations, need not meet the above flexural strength requirement. To accommodate
inelastic rotations in the gusset plates of an out-of-plane buckling brace, a distance of “2t”
can be left free of the member, as discussed in Chapter 16 and shown in Figure 16.11.
Bracing
2t Member

Gusset Plate (t)

Location of Plastic
Hinge in the Gusset
Plate

Beam
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-19

Figure 16.11. Location of the “2t” Zone in Gusset Plate to Accommodate

Plastic Hinge Rotations
16.8.c. Required Compressive Strength
The required compressive strength of the bracing connections is at least equal to 1.1 R y P n
(in LRFD) where P n is the nominal compressive strength of the bracing member. The
available compressive strength of the bracing connections, φR n , calculated using
provisions of the AISC Specification (2010) should not be less than the above required
compressive strength.

16.9. ADDITIONAL REQUIREMENTS FOR

CHEVRON AND V-TYPE SCBF
In chevron and V-braced frames, bracing members are connected to midspan of the beam.
During seismic event, after the compression brace buckles, the tension bracing member
continues accepting additional loads until it reaches its yield capacity. After the
compression brace buckles, the axial load in compressive and tensile brace will not be
equal. The vertical component of the difference between tension and compression forces is
applied to the mid-span of the beam. To ensure that the beam can resist this force and
remain essentially elastic the following requirements should be met.

(a) For load combinations that do not involve E the earthquake loads, the beam
should be designed to support its loads on its own span by ignoring the support
bracing members provide at midspan of the beam.
(b) For load combinations that involve E the earthquake loads, the beam should be
designed to support, in addition to its loads, the vertical component of the
difference between the tensile and compressive brace forces applied to its
midspan. The tensile and compressive brace forces for this calculation should be
equal to R y F y A g and 0.3R y F cr A g respectively. The critical buckling stress for
compressive brace, F cr , is established using provisions of the AISC Specification
(2010) as discussed in Chapter 5.
(c) Both top and bottom flanges of the beams connected to chevron or V bracing
members should have lateral bracings designed according to the provisions of
Appendix A-6 of the AISC Specification (2010a) which we discussed in Chapter
18. The strength and stiffness requirements for the lateral bracings are:
(1) The required strength of the lateral bracing is:

P rb = 0.008M r C d / h o (for LRFD) (16.18)

(2) The required stiffness of the lateral bracing is:

1  4M r C d 
β br =   (for LRFD) (16.19)
φ  Lb ho 
where
P rb =Required strength of the lateral bracing (i.e. axial force in the lateral
bracing to be used in its design)
M r = Required bending strength of the beam which in this case is the
expected plastic moment capacity of the beam given as R y F y Z y .
C d = 1.0
h o = Center to center distance of the beam flanges.
φ = 0.75
β br = Required Axial stiffness of the lateral bracing
L b = Distance between the lateral braces (i.e. unbraced length of the
compression and tension flange which is equal to ½ span of the beam if
lateral bracings are provided only at mid-span

16.10. COLUMN REQUIREMENTS IN SPECIAL

CONCENTRICALLY BRACED FRAMES
The b/t ratio of the columns in SCBF should be less than λ ps given in Table I-8-1 of the
AISC Seismic Provisions (2010). The table was given in Chapter 5 .

16.11. BUCKLING RESTRAINED BRACED FRAMES

Buckling Restrained Braced Frames(BRBF) have gained popularity and have been
used frequently in recent years. The system, first invented in Japan about 20 years ago and
tested and developed by Nippon Steel Corporation of Japan, is currently being used quite
efficiently in highly seismic areas such as California, Figure 16.12, and Japan, Figure
16.13. The Osaka Int. Convention Center, in Figure 16.13, has some of the largest BRB
elements crossing two floors.

A Typical Pin Connection of BRBF

http://farm4.static.flickr.com/3020/2984089207_12f23f343d.jpg
http://www.structuremag.org/images/0708-is-1.jpg
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-21

Buckling Restrained Braced Frame systems have high stiffness, predictable

strength and excellent ductility and energy dissipating capability. Many concentrically
braced frames designed and built in recent years have used this system instead of the
conventional X, chevron or V braces. Figure 16.14 shows basic components of buckling
restrained braced frames and Figure 16.15 shows cyclic axial load-axial displacement
behavior of BRB element compared to the behavior of a conventional brace member. As
can be seen, the conventional brace member has much less strength in compression and has
less energy dissipating capacity, represented by the area under the curve, than the BRB.

The Top Down Construction of Middle portion

Figure 16.13. The Osaka International Convention Centre uses a BRB System
(Photos and info courtesy of Nippon Steel Corporation)

Figure 22.14 Components of a Typical Buckling Restrained Bracing Member

(Courtesy of Professor Akira Wada).
16-22
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

(a) (b)

Figure 22.15. Cyclic Behavior of (a) Buckling Restrained Brace; and,

(b) Conventional Concentric Brace
(Courtesy Professor Akira Wada).

Figure 16.12 shows two typical details currently used in connecting the buckling
restrained brace to the beam column joint.
More information on buckling restrained braced frames , including numerical
examples of seismic design of this system, can be found in a Steel TIPS report by Lopez
and Sabelli (2004).

(a) (b)
Figure 22.16. Typical Connections in BRBF System
(From: Steel TIPS by Lopez and Sabelli, 2004).
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-23

16.12. DESIGN OF BUCKLING RESTRAINED BRACED

FRAMES
Almost all BRBF system are patented products and are tested and designed by the
manufacturers. The Steel TIPS by Lopez and Sabelli (2004) provides useful information
on this system and its testing and design processes.
The AISC Seismic Provisions (2006) has following requirements for Buckling
Restrained Braced Frames:
(a) The steel core should be designed to resist the entire axial force in the brace.
(b) The brace design axial strength, φP ysc (in LRFD) in tension and compression,
according to the limit state of yielding is given by:

φP ysc = φF ysc A sc (16.20)

where

P ysc = Design axial strength of BRB

F ysc = Specified minimum yield stress of the steel core, or actual
yield stress of the steel core as determined from a coupon
test
A sc = Net area of steel core
φ = 0.90

(c) Plates used in the steel core that are 2 in thick or greater should satisfy the
minimum notch toughness requirements of Section 6.3 of the AISC Seismic
Provisions (2005).
(d) Splices in the steel core are not permitted.

In addition to above requirements regarding the “core” element, AISC Seismic

Provisions (2005) in its Section 16 has requirements on BRB system as well as test
protocol for testing BRB systems.

16.13. ZIPPER CONCENTRICALLY BRACED FRAME

As discussed earlier, in chevron braces, shown in Figure 16.17(a), an unbalanced
vertical force develops at midspan of the beams. To redistribute this unbalanced force
applied to the beam among the stories, zipper braced frame system shown in Figure
16.17(b) can be used. In this system, other than the first floor, all floors have a vertical
member at the point where brace members are connected to the beam. Figure 16.17 shows
a comparison of behavior of typical chevron and zipper braced frames.

Recent research by Leon et al. (2010) has indicated that the system can be
configured to provide better ductility than a chevron brace system as shown in Figure
16.18. A suspended type of zipper braced frame system shows even more improved
ductility, Figure 16.18. The research on this subject is still ongoing (Leon et al., 2010).
Notice that in Figure 16.18, the concentrically braced frame is depicted as Ordinary and
not Special braced frame which would have provided better and more ductile behavior. It
seems that the zipper braced frames, when developed further and included in the design
codes, can be used as a viable ductile braced frame system in High Seismic Applications..

Figure 16.17. Schematic Comparisons of Behavior of

Chevron and Zipper Braced Frames (Khatib et al. 1988).

Figure 16.18. A Comparison of the Pushover Behavior of Ordinary

Concentrically Braced Fame Compared to a Zipper and Suspended
Zipper Braced Frame (From; Leon et al. 2010)
CHAPTER 21- CONCENTRICALLY BRACED FRAMES 16-25

References for Concentrically Braced Frames

Aslani, F., and Goel, S. C. (1991), “Stitch Spacing and Local Buckling in Seismic Resistant Double
Angle Bracing Members,” Journal of Structural Engineering, Vol. 177, No. 8, (August),
ASCE, Reston, VA.
Astaneh-Asl, A., Goel, S. C., and Hanson, R. D. (1986). “Earthquake resistant design of double
angle bracing.” Eng. J., 23_4_.
Black, C., and Makris, N. (2004). “Component testing, seismic evaluation and characterization of
buckling-restrained braces.” J. Structural Engineering, 130(6), 880-894
El-Tayem, A. A. and Goel, S.C., (1986). “Effective Length Factor for the Design of X-bracing
Systems” Engineering Journal, AISC, First Quarter.
FEMA 355C, (2000), State of the Art Report on Systems Performance of Steel Moment Frames
Subject to Earthquake Ground Shaking, Report FEMA 355C, prepared by the SAC Joint
Venture for the Federal Emergency Management Agency, Washington, DC.
Goel, S. C., (1992), “Earthquake Resistant Design of Ductile Braced Steel Structures,” Stability and
Ductility of Steel Structures under Cyclic Loading, pp. 297-308, CRC Press, Boca Raton,
FL.
Goel, S. C. and Lee, S., (1992), “A Fracture Criterion for Concrete-Filled Tubular Bracing Members
under Cyclic Loading,” Proceedings of the 1992 ASCE Structures Congress, pp. 922-925,
ASCE, Reston, VA.
Hassan, O., and Goel, S. C. (1991). “Seismic behavior and design of concentrically braced steel
structures.” Rep. UMCE 91-1, Dept. of Civil and Environmental Engineering, Univ. of
Michigan, Ann Arbor, Mich.
Khatib, I.F., Mahin, S.A., and Pister, K.S. (1988). Seismic Behavior of Concentrically Braced Steel
Frames. Report No. UCB/EERC-88/01. Berkeley: Earthquake Engineering Research
Center. University of California.
Leon, R., Yang, C-S., DesRoches, R., Reinhorn,A., Schachter, M, Stojadinovic, B., Yang,T. Shing,
B., Zhong, W., (2010), Advances in Experimental Structural Engineering, Itoh and Aoki
(editors.).
Stoman, S.H., (1988). “Stability Criteria for X-Bracing Systems”, Journal of Engineering
Mechanics Vol.114, No. 8, pp.1426-1434.
Tang, X., and Goel, S. C. _1987_. “Seismic analysis and design considerations of braced steel
structures.” Rep. UMCE 87-4, Dept. of Civil and Environmental Engineering, Univ. of
Michigan, Ann Arbor, Mich.

Tremblay, R. and Trica, L., (2004), “Behavior and Design of Multi-Story Zipper Concentrically
Braced Steel Frames for the Mitigation of Soft-Story Response”, Proceedings of the 13th
world conference on earthquake engineering.
CHAPTER 17-MOMENT FRAMES 17-1

CHAPTER 17

MOMENT FRAMES

Objectives:
In this chapter, you will be introduced to:
1. Moment Frames
2. Design and Behavior of Special Moment Frames
3. Column-Tree Moment Frames

17.1. INTRODUCTION
In moment frame systems, beams and columns are connected to each other by rigid
moment connections. In general, moment frame systems are more flexible and less
efficient than braced frames in resisting lateral drifts. But, unlike brace frames, where the
bracing can create obstruction in the useable space, in moment frames, the spans are not
obstructed.
The main mechanism in moment frames to resist the load is bending of the beams
and columns. In flexural members, only the extreme fibers are loaded to high level of stress
and only at cross sections where moment is high. As a result, the bulk of material in a
moment frame remains elastic even when the frame reaches its collapse mechanism and its
_____________________________________________________________________________
Behavior and Design of Steel and Composite Structures,
Copyright  2003- 2012 Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE247 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
17-2
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

ultimate strength. This is the primary reason for relatively lower cost efficiency of the
moment frame systems compared to the braced frames.

Column
remains
elastic

Plastic hinges
form in the
beams
Most of beam Yielded
remains Zone
elastic

Typical Moment Frame

Figure 17.1. An Example of Moment Frame and Limited Plastic Area of the Hinge

Figure 17.2 shows examples of moment frame systems in actual buildings. Until
the 1994 Northridge earthquake, most of the moment frames used in seismic areas was
welded moment frames. However, during the Northridge earthquake, the moment
connections in hundreds of welded moment frames in greater Los Angeles were damaged.
As discussed in Chapter 14, the damage was in the form of cracking of welds connecting
the flanges of the girders to columns or fracture of the base metal in girders and columns
adjacent to the welds. A year later, during the 1995 Kobe Japan earthquake, similar
damage was observed in many welded moment frames. More on what caused the damage
and what was done to improve the moment connections was given in Chapter 14 when
discussing moment connections. In this chapter we will focus more on moment frame
systems and not connections.

http://www.anvilsteel.com/images/commercial/foxl1.jpg

Figure 17.2. Examples of Moment Frame Systems

CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-3

17.1.a. Advantages of Moment Frames

Some of the advantages of using moment frame systems are:
1. Moment frames result in unobstructed bays, a desirable feature from architectural
and space use points of view.
2. Moment frames can be shop-welded field-bolted resulting in speed of construction
as well as less field inspection and quality control /quality assurance.

17.2. TYPES OF MOMENT FRAMES

Examples of types of moment frames are shown in Figure 17.3. The main
difference among the moment frames shown is in the location of plastic hinge and the
detailing of the connections. In seismic applications, all moment frame systems should
satisfy the requirement of strong column, weak beam, which results in forcing plastic
hinges to form in the beams and not in the columns.
As mentioned earlier, the most important characteristics of a moment frame is that all the
inelasticity is concentrated in the plastic hinges in the beams while all other components of
the system, including columns and connections remain elastic. The only exception is that
plastic hinges are allowed to form at the bases of columns. Figure 17.4 shows deformed
shape of moment frames shown in Figure 17.3. The rotational demand on the connection
depends on the location of the plastic hinge in the beam.
In moment frames, the bulk of the lateral deformation of the frame results from
rotational deformation of plastic hinge in the beams as can be observed in Figure 17.4.
The beam “stub,” the part of the beam between the plastic hinge and the column, remains
elastic and maintains an angle of 90o with the column. The connection of the beam stub to
the column also remains elastic.
From the geometry in Figure 17.5 and assuming elastic deformations of the
frame members are relatively small, we can observe that:

 L 
θ =   γ (17.1)
 L' 
Where;
θ =Drift angle for the frame
γ = Rotational angle of the plastic hinge
L = Length of span
e = Length of the beam stub between the column face and the plastic hinge in the girder.

_____________________________________________________________________________
Behavior and Design of Steel Structures, Copyright  2003-2012, Abolhassan Astaneh-Asl. All rights reserved.
This document is provided to students enrolled in CE122 at UC Berkeley for their personal use and should not be distributed to others or
posted anywhere.
17-4
BEHAVIOR AND DESIGN OF STEEL STRUCTURES

Column
Beam Beam

Rigid Moment Semi-Rigid or

Connection Shear Connection

Column Rigid Moment

Connection

(a) Rigid Moment Frame (b) Column Tree Moment Frame

Rigid Moment
Rigid Moment Connection Semi-Rigid or
Beam Shear Connection
Connection

Column

Figure 17.3. Four Examples of Moment Frames

Plastic
Hinges

γ γ
θ
θ

L’ γ
L’
L
L

(a) (b)

γ
γ
θ
θ Shear
Deformations

L’ L’

L L

Figure 17.4. Deformed Geometry of Moment Frames with Various Locations of

Plastic Hinge Formation
CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-5

17.3. CONNECTIONS IN MOMENT FRAMES

The beam-to-column connections in moment frames are the most critical
components of the system. In particular, the field-welded moment connections, as the
Northridge earthquake showed, can experience damage if not designed properly. In
Chapter 14, we discussed welded and bolted moment connections currently designated as
“Prequalified” connections for moment frames in seismic applications. Figure 17.5 shows
examples of prequalified connections used in moment frames in seismic applications.

Figure 17.5. Current “Prequalified Moment Connections

Since the 1994 Northridge earthquake, a number of proprietary (i.e. patented)

moment connections have been developed and used in a number of steel moment frames in
seismic regions. Two of these connections are Slotted Web TM connections, Figure 17.6,
and SidePlateTM connections, Figure 17.7. Another system, used by Anvil Steel, is the
Flange Slotted connection shown in Figure 17.8

Proprietary SlottedWebTM Connection

Figure 17.6. Slotted Web Propriety Connection

Photos courtesy of ABKJ

Figure 17.7. SidePlateTM Proprietary Connection

Long
Slots in
the
Flanges

http://www.anvilsteel.com/images/commercial/NBCl01.jpg

17.8. Flange Slotted Bolted Moment Connection

CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-7

17.4. CONFIGURATIONS OF MOMENT FRAMES

Figures 17.9 and 17.10 show typical configurations of moment frame systems,
which are:
1. Space Moment Frame Systems
2. Perimeter Moment Frame System
3. Partial Moment Frame System
4. Tube System
5. Tube in Tube System
6. Bundled Tube System

(b) Perimeter Moment Frames. (c) Partial Moment Frames.

(a) Space Moment Frames. All connections in the perimeter Connections in only some
All connections in both frame only spans are moment connections
directions are moment are moment connections
connections

Figure 17.9. Common Configurations of Moment Frame Systems

Figure 17.10. Other Configuration of Moment Frame Systems

23.5. SEISMIC PERFORMANCE OF MOMENT

FRAMES
One of the important structures that have moment frames is the 49-story
Embarcadero-4 building in San Francisco. The building has instrumentation at various
levels installed by the California Strong Motion Instrumentation Program (CSMIP). The
Loma Prieta earthquake of 1989 shook this building and valuable data was recorded at this
structure in the form of acceleration time histories at various levels. In the aftermath of the
Loma Prieta earthquake, a research project was conducted by Chen, Bonowitz and
Astaneh-Asl (1992) on this building and the recorded data. Figure 17.11 shows inelastic
behavior of moment frames in the E-W direction when building is subjected to a M8+ San
Andreas ground motion.

Figure 17.11. Response of Moment Frames of the Embarcadero Building to

M8+ San Andreas Ground Motions
CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-9

23.6. COLUMN-TREE MOMENT FRAME SYSTEMS

One of the very efficient shop-welded and field bolted systems is the column-tree
system. In a column-tree system short segments of the girders or a built-up short girder,
usually two to four feet long, are welded to the columns in the shop. Then, after the column
trees are erected in the field, the middle segment of the girder is bolted to the ends of the
short girder stubs. Figure 17.12 shows examples of special ductile column-tree moment
resisting frames.

Figure 17.12. Typical Column Tree Moment Frame Systems

In construction projects where field-welding and field inspection are too costly or
cannot be done easily, the use of column-tree system can be more economical than the
other structural systems that require field-welding. In Japan, perhaps due to the high cost
of labor, and the fact that shop-welding is mostly automated, column-tree frames were
almost the only steel moment-resisting systems until recent years when the welded moment
frames of “pre-Northridge” configurations started to being used there (Takanashi, 1994).

17.7. TYPES OF COLUMN-TREE MOMENT FRAMES

Column Tree Moment Frames can be divided into Special, Intermediate and Ordinary
moment frames. Depending on the ductility of the splice connection, the column tree
system will be in one of these three categories. Figure 17.13 shows suggested details for
the splices of column-tree moment frames.

Figure 17.13. Suggested Details for Splices of Column-Tree Moment Frames

17.8. SUGGESTED CRITERIA FOR DESIGN OF

COMPONENTS OF SPECIAL COLUMN-TREE
MOMENT FRAMES
The girder splices in Column-Tree moment frame systems are suggested to be designed to
satisfy the following:
1. Plastic moment capacity of the girder splice does not need to be greater than the
plastic moment capacity of the girder. The plastic moment capacity of the splice,
CHAPTER 20- STEEL AND COMPOSITE SYSTEMS 20-11

M p , should be equal or greater than the larger of: (a) the calculated applied moment
at the location of the splice or 1-(2L’/L) times the plastic moment capacity of the
girder, where L’ is the distance between splices and L is the span length, see Figure
17.14.
Plastic
Hinges

γ
θ

L’

Figure 17.14. Deformations of a Typical Column-Tree Moment Frame

2. The splice hinges should be designed to have available hinge rotation capacity, γ,
equal or greater than θ (L/L’) where L and L’ are as shown in Figure 17.14 and θ is
the maximum design story drift angle allowed for the frame by the governing code.
3. The girder splice should be designed such that the yield capacity of the gross area
of the plates in the splice governs. Other failure modes such as net section failure
or bolt failure should have larger capacity than the yield capacity.
4. The connection of girder stub to the column should have the strength equal or
greater than the expected plastic moment capacity of the girder equal to C p R y Z x F y
where
C p = (F y +F u )/2F y
R y = Parameter for expected yield stress given in Table I-6-1 of the AISC Seismic
Provisions . The table was given in Chapter 2 (see Page 2-20).
Z x = Plastic section modulus of the beam
F y =Specified minimum yield stress of steel
5. The panel zone in the column should have shear strength of C p R y Z x F y /h o where h o
is the center-to-center of beam flanges.
6. The girders and columns should have b/t ratios not exceeding λ ps , limit of seismic
compactness.
7. The fixed base plates should be designed to develop C p R y Z x F y of the column.

References for Chapter 17:

1. AISC, (2010) “Specifications for Structural Steel Buildings,” American Institute of Steel
Construction, Chicago.
2. AISC, (2011) “Steel Construction Manual,” American Institute of Steel Construction,
Chicago.
3. AISC (2005), AISC 341, Seismic provisions for structural steel buildings, American
Institute of Steel Construction Inc., Chicago, IL
4. ASCE (2005), ASCE 7-05, Minimum design loads for buildings and other structures,
American Society of Civil Engineering, Reston, VA
5. AIJ, (1995), "Reconnaissance Report on Damage to Steel Building Structures Observed
from the 1995 Hyogoken-Nanbu (Hanshin/Awaji) Earthquake,” Report, Architectural
Institute of Japan, (in Japanese), May.
6. Astaneh-Asl, A., (1995), "Seismic Design of Bolted Steel Moment-Resisting Frames,
(1995), Steel Tips Report, Structural Steel Educational Council, Moraga, CA, July.
7. Astaneh-Asl, A., Shen, J. H., D'Amore, E., McMullin, K. M., and Modjtahedi, D. (1995),
Seismic Safety of damaged Welded Steel Moment Frames,” Report No. UCB/CE-Steel-
95/01, Department of Civil Engineering, University of California, Berkeley, October.
8. Englekirk, R., (1994), "Steel Structures, Controlling Behavior Through Design,” , John
Wiley and Sons Inc..
9. McMullin, K., Astaneh-Asl, A., Fenves, G. and Fukuzawa, E., "Innovative Semi-Rigid
Steel Frames for Control of the Seismic Response of Buildings,” Report No. UCB/CE-
Steel-93/02, Department of Civil and Environmental Engineering, University of
California, Berkeley.
10. Popov, E. P., and Bertero, V. V. , (1973), "Cyclic Loading of Steel Beams and
Connections," Journal of Structural Division, ASCE, Vol. 99, No. 6.
11. Popov, E. P. and Stephen, R. M., (1972)," Cyclic Loading of Full-Size Steel
Connections," Bulletin No. 21, AISI, New York.
12. SAC, (1994), "Invitational Workshop on Steel Seismic Issues,” Proceedings, Workshop
by SAC Joint Venture held in Los Angeles, September.
13. Youssef, N. F. G., Bonowitz, D., and Gross John L., " A Survey of Steel Moment-
Resisting Frame Buildings Affected by the 1994 Northridge Earthquake,” Report No.
NISTIR 5625 National Institute of Standards and Technology, Washington D.C., April.

Steel Structure Design Guide
100% (1)
Steel Structure Design Guide
9 pages
Applied Structural Steel Design 4 Edition
100% (7)
Applied Structural Steel Design 4 Edition
460 pages
Todd J. D., 1974, Structural Theory and Analysis
100% (5)
Todd J. D., 1974, Structural Theory and Analysis
364 pages
Steel Structures, Design & Behavior 4th Ed.-Salmon, Johnson
100% (1)
Steel Structures, Design & Behavior 4th Ed.-Salmon, Johnson
1,044 pages
Theory of Structures P.1 EL-Dakhakhni1
100% (10)
Theory of Structures P.1 EL-Dakhakhni1
452 pages
Structural Analysis
100% (3)
Structural Analysis
260 pages
Reinforced Concrete Design
100% (1)
Reinforced Concrete Design
132 pages
Steel Gables and Arches PDF
100% (2)
Steel Gables and Arches PDF
52 pages
Qualitative Deflected Shapes - EngineeringWiki
No ratings yet
Qualitative Deflected Shapes - EngineeringWiki
13 pages
Structural Stability Basics
100% (1)
Structural Stability Basics
25 pages
507 33 Powerpoint-Slides Ch8 DRCS
No ratings yet
507 33 Powerpoint-Slides Ch8 DRCS
70 pages
Structural Theory II Study Guide
100% (6)
Structural Theory II Study Guide
204 pages
Theory of Structures P.2 EL-Dakhakhni
83% (12)
Theory of Structures P.2 EL-Dakhakhni
481 pages
Introduction To Shear Resistance of Plain' and Reinforced Concrete Sections
No ratings yet
Introduction To Shear Resistance of Plain' and Reinforced Concrete Sections
11 pages
Steel Beam Design: LRFD-AISC Guide
100% (1)
Steel Beam Design: LRFD-AISC Guide
52 pages
Design of Truss
100% (2)
Design of Truss
29 pages
Theory of Indeterminate Structures - Syed Ali Rizwan
100% (11)
Theory of Indeterminate Structures - Syed Ali Rizwan
470 pages
Theory of Structures, Timoshenko PDF
83% (12)
Theory of Structures, Timoshenko PDF
638 pages
Theory of Structure PDF
No ratings yet
Theory of Structure PDF
547 pages
In Structural Steel: Limit States Design
No ratings yet
In Structural Steel: Limit States Design
229 pages
Reinforce Concrete: Structures
No ratings yet
Reinforce Concrete: Structures
16 pages
Steel Structures Design and Behavior 5th Edition Salmon Fast Access
No ratings yet
Steel Structures Design and Behavior 5th Edition Salmon Fast Access
326 pages
Design of Compression Member
No ratings yet
Design of Compression Member
29 pages
Structural Analysis for Engineers
No ratings yet
Structural Analysis for Engineers
63 pages
Chapter 2 Part 2 - Deflection Unit Load Method
100% (2)
Chapter 2 Part 2 - Deflection Unit Load Method
51 pages
Analysis of Structures
100% (1)
Analysis of Structures
264 pages
Steel Design
100% (3)
Steel Design
180 pages
Influence Lines For Statically Indeterminate Structures Lectures Notes
100% (1)
Influence Lines For Statically Indeterminate Structures Lectures Notes
43 pages
Structural Engineers' Amplification Guide
100% (1)
Structural Engineers' Amplification Guide
6 pages
Theory of Structures SMTS 2 S.I Units by B C Punmia, Ashok Kumar Jain and Arun Kumar Jain
No ratings yet
Theory of Structures SMTS 2 S.I Units by B C Punmia, Ashok Kumar Jain and Arun Kumar Jain
498 pages
Influence Lines for Structures
60% (5)
Influence Lines for Structures
51 pages
Prestressed Concrete
67% (3)
Prestressed Concrete
5 pages
Dokumen - Tips - Steel Structures 5th Edition Solutions Manual
No ratings yet
Dokumen - Tips - Steel Structures 5th Edition Solutions Manual
131 pages
Module 1.2 - Flexure in Beams
No ratings yet
Module 1.2 - Flexure in Beams
8 pages
Desiga of Steel Structures
100% (10)
Desiga of Steel Structures
206 pages
02 Structural Analysis 2 (Advanced)
100% (7)
02 Structural Analysis 2 (Advanced)
274 pages
507 39 Solutions-Instructor-manual Ch7 DRCS
100% (1)
507 39 Solutions-Instructor-manual Ch7 DRCS
13 pages
Udoeyo Structural Analysis
100% (4)
Udoeyo Structural Analysis
391 pages
Continous Beam - Flexibility Method
No ratings yet
Continous Beam - Flexibility Method
11 pages
The Behaviour and Design of Steel Structures To BS5950
100% (5)
The Behaviour and Design of Steel Structures To BS5950
237 pages
AE December 2019 16 Civ A1
No ratings yet
AE December 2019 16 Civ A1
7 pages
Reinforced Concrete Design, 3rd Ed, Leet
92% (13)
Reinforced Concrete Design, 3rd Ed, Leet
562 pages
Composite Beam Design Guide
100% (1)
Composite Beam Design Guide
4 pages
Flexibility Method in Structural Analysis
No ratings yet
Flexibility Method in Structural Analysis
56 pages
Torsion and Column
No ratings yet
Torsion and Column
31 pages
Design of Concrete Structures
100% (2)
Design of Concrete Structures
783 pages
Structural Design for Engineers
100% (1)
Structural Design for Engineers
220 pages
Theory of Indeterminate Structure by Syed Ali Rizwan PDF
100% (2)
Theory of Indeterminate Structure by Syed Ali Rizwan PDF
470 pages
Advanced Theory of Structures (Frame Analysis)
100% (4)
Advanced Theory of Structures (Frame Analysis)
216 pages
Learning Module: College of Engineering and Information Technology
No ratings yet
Learning Module: College of Engineering and Information Technology
11 pages
Limit State Design of Structural Steelwork PDF
100% (1)
Limit State Design of Structural Steelwork PDF
152 pages
Tailieuxd.com Resource Links
75% (4)
Tailieuxd.com Resource Links
136 pages
Analysis of Statically Indeterminate Structures by The Force Method
No ratings yet
Analysis of Statically Indeterminate Structures by The Force Method
57 pages
Design of Tension Members
No ratings yet
Design of Tension Members
14 pages
Structural Analysis
100% (5)
Structural Analysis
866 pages
Steel & Timber Structural Design Guide
No ratings yet
Steel & Timber Structural Design Guide
37 pages
Structural Steel Design Guide
100% (4)
Structural Steel Design Guide
358 pages
Chapter One: Introduction, Material and Design Concepts: Y.Boopathi-Civil /lecturer
No ratings yet
Chapter One: Introduction, Material and Design Concepts: Y.Boopathi-Civil /lecturer
70 pages
Lec 1 A
No ratings yet
Lec 1 A
55 pages
Steel Structures Design Course
No ratings yet
Steel Structures Design Course
79 pages
4 Fracture Mechanics - Intro-Rev
No ratings yet
4 Fracture Mechanics - Intro-Rev
19 pages
Earthquake Impact on Gravity Dams
No ratings yet
Earthquake Impact on Gravity Dams
99 pages
Carbon Fiber PDF
No ratings yet
Carbon Fiber PDF
6 pages
Zinc-Coated Parallel and Helical Steel Wire Structural Strand
No ratings yet
Zinc-Coated Parallel and Helical Steel Wire Structural Strand
6 pages
MT&QC Lab Manual
100% (1)
MT&QC Lab Manual
58 pages
Full
No ratings yet
Full
6 pages
Developments On Seismic Retrofit of RC Shear Walls With FRP
No ratings yet
Developments On Seismic Retrofit of RC Shear Walls With FRP
20 pages
Effect of Solid-Solution Treatment On High-Temperature Properties and Creep Fracture Mechanism of LPBF IN625
No ratings yet
Effect of Solid-Solution Treatment On High-Temperature Properties and Creep Fracture Mechanism of LPBF IN625
23 pages
Thamrin - 2020 - IOP - Conf. - Ser. - Mater. - Sci. - Eng. - 713 - 012029 PDF
No ratings yet
Thamrin - 2020 - IOP - Conf. - Ser. - Mater. - Sci. - Eng. - 713 - 012029 PDF
8 pages
Concrete Design Prokon
100% (3)
Concrete Design Prokon
206 pages
Bridge Failure Analysis Report
No ratings yet
Bridge Failure Analysis Report
11 pages
Torsion Test in Applied Mechanics Lab
No ratings yet
Torsion Test in Applied Mechanics Lab
19 pages
Elastic-Plastic Fracture Mechanics For Initiation and Propagation of Notch Fatigue Cracks
No ratings yet
Elastic-Plastic Fracture Mechanics For Initiation and Propagation of Notch Fatigue Cracks
16 pages
Flooring: Technical Manual
No ratings yet
Flooring: Technical Manual
35 pages
Composite Materials for M.Tech Students
No ratings yet
Composite Materials for M.Tech Students
2 pages
Bond F.C., 1961. Crushing and Grinding Calculations
90% (10)
Bond F.C., 1961. Crushing and Grinding Calculations
17 pages
An Study On Steel Fiber Reinforced Concrete
100% (1)
An Study On Steel Fiber Reinforced Concrete
26 pages
2024 Subjects Description
No ratings yet
2024 Subjects Description
9 pages
Tme 313 - Semester Major Take Home Exercises - 2023 2024
No ratings yet
Tme 313 - Semester Major Take Home Exercises - 2023 2024
11 pages
Failure Analysis For Endineering Materials
No ratings yet
Failure Analysis For Endineering Materials
18 pages
GELITA UpToDate38210x279,4 LY02
No ratings yet
GELITA UpToDate38210x279,4 LY02
16 pages
Fracture Mechanics Reasoning QA
No ratings yet
Fracture Mechanics Reasoning QA
4 pages
Pole Vault and The Pole
No ratings yet
Pole Vault and The Pole
10 pages
ASTM E-Books List by Title & Publisher
100% (1)
ASTM E-Books List by Title & Publisher
61 pages
Principles of Reinforced Concrete - Module 4
No ratings yet
Principles of Reinforced Concrete - Module 4
12 pages
NALCO Corrosionmanual PDF
100% (2)
NALCO Corrosionmanual PDF
126 pages
Principal Plane
No ratings yet
Principal Plane
7 pages
Failure Mechanism-Based Reliability Assessment Framework: A Case Study in An Automotive Component
No ratings yet
Failure Mechanism-Based Reliability Assessment Framework: A Case Study in An Automotive Component
9 pages
4461r 91
No ratings yet
4461r 91
146 pages
Tensile Strength of Soil
No ratings yet
Tensile Strength of Soil
17 pages