Altair SimSolid

2020 Verification Manual

Intellectual Property Rights Notice:
Copyrights, Trademarks, Trade Secrets, Patents & Third Party Software Licenses

Altair SimSolid™
Altair Engineering Canada LTD Copyright ©2015-2018. All Rights Reserved.
Altair Engineering Copyright ©2019-2020 All Rights Reserved.

UPDATED: 9-Jul-20
Table of Contents

INTRODUCTION 6

Document Purpose 6

Test Documentation Format 6

SECTION 1: STATIC ANALYSIS TESTS 7

Test VS01: Straight Cantilever Beam 7

Test VS02: Twisted Cantilever Beam 9

Test VS03: Curved Beam 11

Test VS04: Simply Supported Thin Plate 13

Test VS05: Clamped Thin Plate 15

Test VS06: Tapered Solid Bar 17

Test VS07: Cantilever Plate 19

Test VS08: Plate with a Hole 20

Test VS09: Scordelis-Lo Roof 22

Test VS10: Stepped shaft with fillet 23

Test VS11: Raasch Challenge 24

Test VS12: Notched Plate 25

Test VS13: Simply supported beam with mid-span load 26

Test VS14: Skew Plate Under Pressure 27

Test VS15: Stress Concentration of Filleted Bar 30

SECTION 2: MODAL ANALYSIS TESTS 31

Test VM01: Free Vibration of a Wedge 31

Test VM02: In-plane Vibration of a Cantilever Plate 32

Test VM03: Cylindrical Shell 33

Test VM04: Free Thin Square Plate 34

Test VM05: Thick Skewed Plate 36

Test VM06: Thick Square Plate 37

Test VM07: In-plane Vibration of a Pin-ended Cross 39

Test VM08: Cantilevered Beam 41

Test VM09: Circular Ring -- In-plane and Out-of-plane Vibration 42

Test VM10: Triangular Wing -- Fixed Vibration 43

SECTION 3: THERMAL ANALYSIS TESTS 44

Test VT01: Cooling Fin 44

Test VT02: Insulated Wall 45

Test VT03: Cylinder with Prescribed Heat Flux 46

Test VT04: Heat Generating Plate 48

Test VT05: Heat Generating Wire 50

Test VT06: Table Frame 52

SECTION 4: THERMAL LOADING TESTS 53

Test VTL01: Thermally Loaded Support Structure 53

Test VTL02: Bimetallic Cantilever Beam under Thermal Loading 55

Test VTL03: Coupled Thermal-Structural Analysis - Constrained Straight Bar Under Uniform Temperature 57

Test VTL04: Coupled Thermal-Structural Analysis - Constrained Straight Bar Under Linear Temperature Field 59

SECTION 5: NON-LINEAR TESTS 62

Test VNL01: Coupled Analysis - Bimetallic Beam Under Thermal Load 62

Test VNL02: Z-Shaped Cantilever 65

Test VNL03: Lateral Buckling of a Right Angle Frame 67

Test VNL04: Reactions at the ends of statically undefined axially loaded plastic bar 70

Test VNL05: Residual deformations in an axially loaded plastic bar 73

Test VNL06: Pure bending of a plastic beam 76

Test VNL07: Separating beams 81

Page 4
Test VNL08: Pinched Hemispherical Shell 86

Test VNL09: Rigid Punch Plasticity 91

SECTION 6: TRANSIENT DYNAMIC TESTS 95

Test VD01: Simply supported thin square plate. Harmonic forced vibration response 95

Test VD02: Simply supported thin square plate. Periodic forced vibration response 98

Test VD03: Simply supported thin square plate. Transient forced vibration response 100

Test VD04: Simply supported thin square plate. Frequency response 102

Test VD05: Simply supported thin square plate. Random forced vibration response 105

Page 5
Introduction
SimSolid is a new generation of structural analysis software. SimSolid’s unique approach works directly
on fully featured CAD assemblies and complex part geometry models and does not use a mesh.

The SimSolid computational engine is based on significant extensions to the theory of external
approximations. External approximations are a generalization of Finite Element Method (FEM) in terms
that:
• absolutely arbitrary geometrical shapes can be used as “finite elements”
• basis functions which approximate field of interest in the part volume can be of arbitrary class
and are independent of the volume shape

SimSolid does not use the point-wise degrees of freedom (DOF) inherent in traditional FEA. SimSolid
DOF are functionals with geometrical support in form of volumes, areas, line clouds, and point clouds.
This provides the ability to handle geometrical imperfections, as well as assembly contact imperfections
like gaps, penetrations and ragged contact areas.

SimSolid controls solution accuracy using multi-pass adaptive analysis. Adaptivity can be defined on a
global or part local basis and adaptivity is always active. The SimSolid methodology is fast and efficient.
It provides superior performance metrics for computational time and memory footprint that allow very
large and/or complex assemblies to be solved quickly on desktop class PC’s.

Document Purpose
The primary purpose of this document is to demonstrate the theoretical robustness of the SimSolid
software product across a range of typical solution domains. Comparisons are made against well-known
reference solutions. Please note that while many reference solutions are based on closed form solutions
with simplifying assumptions, SimSolid solves all domains with full 3D models. This sometimes leads to
ambiguity about boundary condition handling.

A secondary purpose is to illustrate the format of typical SimSolid product inputs and outputs. However,
this document is not intended to be used as a training manual.

It is assumed the reader has a basic understanding of analysis software systems including how to define
loads, constraints, materials and specify units.

Tests are provided in six sections that cover Statics, Modal, Thermal, Thermal Loading, Non-linear and
Transient Dynamics solutions.

Test Documentation Format

In this document, tests are labeled using the format V<x><nn> where <x> is S, M, T, TL, NL or D and
<nn> is the test number. Each test is documented with a:
• Problem description - provides geometry, loading, constraints, material properties and units used
(SI or IPS)
• Reference Solution - reference results are provided along with their relevant source
• SimSolid Results - SimSolid results with a comparison to the reference solution
• Supporting Figures & Results - where appropriate figures, tables, graphs, and SimSolid results
plots are provided to illustrate the test
• Unless noted, all problems are run using SimSolid standard solution settings

Page 6
Section 1: Static Analysis Tests
Test VS01: Straight Cantilever Beam
Problem Definition

A straight cantilever beam (Fig. vs01, a) is subjected to four different loads at the free end (Fig. vs01, b):

1. extension
2. in-plane shear
3. out-of-plane shear
4. twisting

Units are IPS.

Loads are uniformly distributed along face of the beam (Fig. 1-1-2a, b, c) or along edges (Fig. 1-1-2d).

Materials properties: modulus of elasticity E = 1e+7 psi, Poisson’s ratio ν=0.3

Find tip displacements in direction of loads.

Reference
“A proposed standard set of problems to Test Finite Element Accuracy”, by R.H. MacNeal and R.L.
Harder, Finite Elements in Analysis and Design I, 1985, pp. 3-20.

Fig. vs01, a

Fig. vs01, b

Page 7
Results

Tip displacement in direction of load

Theory [Note 1] SimSolid % Difference

Axial Loading +X direction 3.0000E-05 3.0016E-05 0.05%
In-plane Shear +Y direction 1.0810E-01 1.0813E-01 0.03%
Out-of-plane Shear +Z direction 4.3210E-01 4.3205E-01 -0.01%
Twisting [Note 2] +Z direction 3.4060E-03 3.4364E-03 0.89%

Note 1 – Beam theory

Note 2 – Twist load was 1 in-lbf X-moment applied as remote load at end face of the beam.

Page 8
Test VS02: Twisted Cantilever Beam
Problem Definition
A cantilever beam (Fig. vs02), twisted by 90 degrees, is subjected to two different unit loads at the free
end (Fig. vs02):

1) in-plane shear
2) out-of-plane shear

Units are IPS.

Loads are uniformly distributed on the face.

Material properties: modulus of elasticity E=29e+6 psi, Poisson’s ratio ν=0.22

Find tip displacements in direction of loads.

Reference
“A proposed standard set of problems to Test Finite Element Accuracy”, by R.H. MacNeal and R.L.
Harder, Finite Elements in Analysis and Design I, 1985, pp. 3-20.

Results

Theory [Note 1] SimSolid % Difference

In-plane Shear +Y direction 5.4240E-03 5.4618E-03 0.70%
Out-of-plane Shear +X direction 1.7540E-03 1.6290E-03 -7.13%

Note 1 – Beam theory

Page 9
Fig. vs02, a

Fig. vs02, b

Page 10
Test VS03: Curved Beam
Problem Definition
A curved beam, spanning a 90 degree arc is fixed at one end and free at the other (Fig. vs03, a). The
beam is subjected to two different loads at the free end (Fig. vs03, b):

1) in-plane load
2) out-of-plane load

Units are IPS.

Material properties: Modulus of elasticity E=1e+7 psi, Poisson’s ratio ν=0.25

Find tip displacements in direction of loads.

Reference
“A proposed standard set of problems to Test Finite Element Accuracy”, by R.H. MacNeal and R.L.
Harder, Finite Elements in Analysis and Design I, 1985, pp. 3-20.

Results

Tip Displacement (Center face) in Direction of Load [in]

Theory [Note 1] SimSolid % Difference
In-plane Load +Y direction 8.7340E-02 8.8661E-02 1.51%
Out-of-plane Load +Z direction 5.0220E-01 5.1546E-01 2.64%

Note 1 – Curved beam theory

Page 11
Fig. vs03, a

Fig. vs03, b

Page 12
Test VS04: Simply Supported Thin Plate
Problem Definition
A square plate simply supported on four edges (Fig. vs04, a) is subjected to the following loads (Fig.
vs04, b):

1) uniform pressure
2) concentrated load

Units are IPS.

Material properties: modulus of elasticity E=1.7472e7 psi, Poisson’s ratio ν=0.3

Because solid model is used, boundary conditions on supported edges are applied along middle lines of
these edges (Fig. vs04, c).

Find displacements at center of the plate.

Reference
“A proposed standard set of problems to Test Finite Element Accuracy”, by R.H. MacNeal and R.L.
Harder, Finite Elements in Analysis and Design I, 1985, pp. 3-20.

Results

Displacement at Center [in]

Theory [Note 1] SimSolid % Difference
Uniform load 4.0620E-06 4.0733E-06 0.28%
Point load 1.1600E-05 1.1633E-05 0.28%
Note 1 – Kirchhoff’s Plate Bending Model

Page 13
.
Fig. vs04, a

Fig. vs04, b

Fig. vs04, c

Page 14
Test VS05: Clamped Thin Plate
Problem Definition
A square plate clamped on four edges (Fig. vs05, a) is subjected to the following loads (Fig. vs05, b):

1) uniform pressure
2) concentrated load

Units are IPS.

Material properties: Modulus of elasticity E=1.7472e7 psi, Poisson’s ratio ν=0.3

Find displacement at center of the plate.

Reference
“A proposed standard set of problems to Test Finite Element Accuracy”, by R.H. MacNeal and R.L.
Harder, Finite Elements in Analysis and Design I, 1985, pp. 3-20.

Results

Displacement at Center [in]

Theory [Note 1] SimSolid % Difference
Uniform load 1.2600E-06 1.2681E-06 0.64%
Point load 5.6000E-06 5.6297E-06 0.53%

Note 1 – Kirchhoff’s Plate Bending Model

Page 15
Fig. vs05, a

Fig. vs05, b

Page 16
Test VS06: Tapered Solid Bar
Problem Definition
A tapered bar of square cross section (Fig. vs06, a), is subjected to axial load applied to the end of the
bar (Fig. vs06, b). Determine the maximum axial deflection in the bar and axial stress at mid-length.

Units are IPS.

Material properties: Modulus of Elasticity E=10.4e6 psi, Poisson’s ratio ν=0.3

Load is uniformly distributed at free end of the bar. The other end is fixed.

Reference
Harris, C.O., “Introduction to Stress Analysis”, The MacMillan Co., New York, 1959, pp. 237, Problem 4.

Results

Theory SimSolid % Difference

Maximum Axial Deflection [in] 4.8080E-03 4.8324E-03 0.51%
Axial Stress (1) at Midpoint [psi] 4.4440E+03 4.4200E+03 -0.54%

(1) Stress-Z along length of bar. Value take off midpoint of XY graph.

Page 17
Fig. vs06, a

Fig. vs06, b

Page 18
Test VS07: Cantilever Plate
Problem Definition
Find bending stress at fixed end for a cantilevered plate subjected to a vertical shear load of 120 lbs.
(Fig. vs07).

Units are IPS.

Material properties: modulus of elasticity E=10.7 e6psi, Poisson’s ratio ν=0.3

Load is expected to be uniformly distributed along free end of the plate.

Reference
Ferdinand L., Singer, STRENGTH OF MATERIALS, (Harper & Row, 1962), Art. 52, p. 133.

Results

Theory SimSolid % Difference

Maximum Bending Stress [psi] 6.0000E+04 6.1128E+04 1.88%

NOTES: End constraint BC is sliding with fix line spot at middle of end face.

Fig. vs07
Page 19
Test VS08: Plate with a Hole
Problem Definition
Plate with a hole. Due to symmetry, only 1/4 section was analyzed as shown.

Units are SI.

Fig. vs08

L= 100 mm, H=50 mm, R=20 mm

Elastic Modulus=2.1e+11 Pa, Poisson's Ratio=0.3
Thickness=1 mm, F=100 N/mm^2

Reference Solution

Page 20
Results
Sliding constraints applied to left, back and bottom faces.
Load of 5000N applied to right face (100N/mm^2 x 50mm x 1mm)

Theory SimSolid % Difference

Maximum Stress [MPa] 3.7270E+02 3.7775E+02 1.35%

Page 21
Test VS09: Scordelis-Lo Roof
Problem Definition
Curved cylindrical dome. Supported on curved edges (Ux, Uz = 0), unconstrained on straight edges.
Due to symmetry, only 1/4 section was analyzed as shown.

L = 50 m
R = 25 m
t = 0.25m

Elastic Modulus = 4.32 e8 N/m^2

Poisson's Ratio = 0.0

Load = -90 N/m^2 in -Z direction

Reference Solution
The theoretical solution value of the Z displacement at the middle of free edge is given as 0.3086
though many finite element codes converge to the smaller value of 0.3024.

Results
Quarter section modeled. The Z-load determined as follows. Area of top face is
2*Pi*(R+.125)*(40/360)*25 = 438.514 m^2. To get -90 N/m^2, need a total load of 438.514*90 =
39466.25704 N.

Theory SimSolid % Difference

Maximum Z Displacement [m] 3.0860E-01 3.1098E-01 0.77%

Page 22
Test VS10: Stepped shaft with fillet
Problem Definition
Find stress concentration in filleted notch.
Reference Solution 𝜎𝑚𝑎𝑥 = 𝐾𝑡 𝜎𝑛𝑜𝑚 ≅ 2.6 𝑀𝑃𝑎

Model parameters:
• Units are SI
• Fix larger shaft face
• Apply 1 MPa extensional pressure to smaller shaft face
• D = 120 mm, d = 60 mm, r = 3 mm
• D/d = 2
• r/d = 0.05

• 𝐾𝑡 ≅ 2.6

Reference

Results

SZ Stress [MPa] 2.6000E+00 2.6060E+00 0.23%

5 adaptive passes with Adapt to features ON

Page 23
Test VS11: Raasch Challenge
Problem Definition
For thick doubly curved beam, find tip displacement at free end. Reference Solution is 4.9352 in

Units are IPS.

Material properties: modulus of elasticity E=3300 psi, Poisson’s ratio ν=0.3

Load is uniform 1.0 lbf out of plane.

•

Reference
Testing of Shell Elements using Challenging Benchmark Problems, F.T. Wong

Results
REF Solution SimSolid % Difference
Tip displacement 4.9355E+00 4.9307E+00 -0.10%

Page 24
Test VS12: Notched Plate
Problem Definition
Find Sx stress in notch and along top face. Reference Solution is 731 psi in notch, 240 psi along top
face.

Units are IPS.

Material properties: modulus of elasticity E=30e6 psi, Poisson’s ratio ν=0.3, plate thickness=0.5 inch

Load: moment of 500 lbf-in applied to right face.

Constraints are: Fixed left face

Reference
LM-ST-1 Learning Module Static Structural Analysis, Dr. Jack Xin, Dr. Liang-Wu Cai,
www.femlearning.org, 2010

Results

REF Solution SimSolid % Difference

SXX Stress in Notch [psi] 7.3100E+02 7.4939E+02 2.52%
SXX Stress along top face [psi] 2.4000E+02 2.4001E+02 0.00%

Page 25
Test VS13: Simply supported beam with mid-span load
Problem Definition
Find max displacement and stress for a pinned beam with load at the midpoint as shown.

Units are SI.

Results

REF Solution SimSolid % Difference

Max Displacement [mm] 1.0714E+00 1.0764E+00 0.46%
Max Stress [MPa] 1.5000E+02 1.5011E+02 0.07%

Page 26
Test VS14: Skew Plate Under Pressure

PROBLEM DEFINITION
A skew plate (Fig. VS14 A) is subjected to uniform pressure 0.7 KPa. Plate is simply supported at its
perimeter: Uz = 0 along edges AB, BC, CD, and AD. Ux = Uy= 0 at point A and Uy= 0 at point B to prevent
rigid body motion.

Material: elasticity modulus = 2.1e+11 Pa, Poisson's ratio = 0.3.

Find maximum principle stress at the plate center at point E.

Fig. VS14 A

REFERENCE
Test LE6 from NAFEMS Publication TNSB, Rev. 3, “The Standard NAFEMS Benchmarks,” October
1990.

MODELING AND RESULTS

Two spot-points were created at the plate corners A and B in the mid-plane in order to be able to apply
constraints which eliminate rigid in-plane translation-rotations of the plate (Fig. VS14 B).

Fig. VS14 B

Page 27
To simulate simple supports, four spot-lines were created at the plate mid-plane around the plate
perimeter (Fig.VS14 C).

Fig. VS14 C

Vertical constraint Uz = 0 was applied to the spot-lines.

Final loading is shown in Fig. VS14 D

Fig. VS14 D

Page 28
Maximum principle stress distribution at the plate bottom is shown in Fig. VS14 E. Reference principle
stress in the middle of the plate is 0.802 MPa (analytical solution for thin plate formulation). SimSolid
solution in full 3D formulation gives 0.822 MPa, the difference is 2%

Fig. VS14 E

Page 29
Test VS15: Stress Concentration of Filleted Bar

PROBLEM DEFINITION
Two configurations of a filleted bar are loaded as shown.

Material: elasticity modulus = 2.1e+11 Pa, Poisson's ratio = 0.3.

Axial load: -1 MPa pressure
Sliding boundary condition at bar bottom, left and back faces

DIMENSION CONFIGURATION 1 CONFIGURATION 2

D 33 mm 45 mm
d 30 mm 30 mm
r 1.5 mm 6 mm
t 2 mm 2 mm
D/d 1.10 1.50
r/d 0.05 0.20
Kt ~2 ~1.75

Find maximum normal stress.

REFERENCE
Shigley’s Mechanical Engineering Design, Appendix A, Figure A-15-5, McGraw Hill, 2016

Reference Kt values are approximately 2 for configuration 1 and 1.8 for configuration 2. Stress
concentration given by:
𝜎 = 𝐾𝑡 ∗ 𝜎𝑜

RESULTS

Reference SimSolid % Difference

Max Prin. Stress - Configuration 1 2.00 2.04 2.00%
Max Prin. Stress - Configuration 2 1.80 1.83 1.67%

3 adaptive passes with Adapt to features and thin solids ON

Page 30
Section 2: Modal Analysis Tests
Test VM01: Free Vibration of a Wedge
Problem Definition
Determine the first fundamental frequency and vibration mode of a wedge which is constrained at the
wide end (Fig. vm01).

Units are IPS.

Material properties: modulus of elasticity E=3.e+7 psi., Poisson’s ratio ν=0,

mass density  = 7.28e-4lb - s2 /in4”

Fig. vm01

Reference
Timoshenko, S., and Young, D.N., “Vibration Problems in Engineering”, 3rd Edition, D. Van Nostrand
Co., Inc., New York, 1955, pg. 392, Art. 62.

Results

Theory SimSolid % Difference

Fundamental Frequency [Hz] 259.16 258.74 -0.16%

Page 31
Test VM02: In-plane Vibration of a Cantilever Plate
Problem Definition
Find the fundamental frequency for in-plane vibration of a cantilevered plate (Fig. vm02).

To exclude out-of-plane vibration of the plate its surface is constrained in z-direction (Uz=0)

Units are IPS.

Material properties: Modulus of elasticity E=3e7 psi, Poisson’s ratio ν=0.3,

mass density  = 7.28e-4lb - s2 /in4.

Fig. vm02

Reference
Roark, R.J. and Young, W.C., Formula for Stress and Strain, (New York:McGraw-Hill Book 6, 1982) pp.
576-578.

Results

Theory SimSolid % Difference

Fundamental Frequency [Hz] 101.33 100.29 -1.03%

Page 32
Test VM03: Cylindrical Shell
Problem Definition
Find fundamental frequency of a cantilever cylindrical shell (Fig. vm03).

Units are IPS.

Material properties: Modulus of elasticity E=3e7 psi, Poisson’s ratio ν=0.3,

mass density  = 7.28e-4lb - s2 /in4.

Fig. vm03

Reference
Roark, R.J. and Young, W.C., Formula for Stress and Strain, NY: McGraw-Hill Co. 1982, p.576.

Results

Theory SimSolid % Difference

Fundamental Frequency [Hz] 62.05 61.58 -0.76%

Page 33
Test VM04: Free Thin Square Plate
Problem Definition
Determine the first fourteen modal frequencies for a free thin square plate (Fig. vm04).

Units are SI.

Material properties: Modulus of elasticity E=2e11, Poisson's ratio ν=0.3, mass density  = 8.e3 kg/m3.

Fig. vm04

Reference
[1] The Standard NAFEMS Benchmarks TNSB, Rev. 3, 5 October, 1990.

[2] Iguchi S. Biegeschwingungen and Klangfiguren der Vierseitg freien rechteckigen Platte.
Ing. - Archiv, Bd 21, p. 5-6, 1953.

Page 34
Results
Note, that modes 4-5 and 6-7 are not available in NAFEMS solution.

REF Solution SimSolid % Difference

Flexible Mode 1 [Hz] 1.622 1.618 -0.23%
Flexible Mode 2 [Hz] 2.360 2.358 -0.08%
Flexible Mode 3 [Hz] 2.922 2.920 -0.08%
Flexible Mode 4 [Hz] 4.190 4.179 -0.26%
Flexible Mode 5 [Hz] 4.190 4.180 -0.25%
Flexible Mode 6 [Hz] 7.356 7.347 -0.12%
Flexible Mode 7 [Hz] 7.356 7.347 -0.12%
Flexible Mode 8 [Hz] 7.668 7.638 -0.39%
Flexible Mode 9 [Hz] NA 8.315
Flexible Mode 10 [Hz] NA 9.266
Flexible Mode 11 [Hz] NA 12.633
Flexible Mode 12 [Hz] NA 12.633
Flexible Mode 13 [Hz] NA 14.065
Flexible Mode 14 [Hz] NA 14.701

Page 35
Test VM05: Thick Skewed Plate
Problem Definition
Find fundamental frequency of a thick clamped skewed plate (Fig. vm05).

Units are SI.

Material properties: Modulus of elasticity E=2e11 N/m2 ,Poisson's ratio ν=0.3,

mass density  = 8.e-3 kg/m3.

Fig. vm05

Reference
The Standard NAFEMS Benchmarks TNSB, Rev. 3, 5 October, 1990.

Results

Theory SimSolid % Difference

Fundamental Frequency [Hz] 133.95 135.47 1.13%

Page 36
Test VM06: Thick Square Plate
PROBLEM DEFINITION
Determine the first 12 modes and frequencies for simply supported thick square plate (Fig. vm06).
Material properties: Modulus of elasticity E=2.e+11 Pa, Poisson's ratio ν=0.3, Mass density 8.e+3 kg/m . 3

Fig. vnl06,a

REFERENCE
The Standard NAFEMS Benchmarks TNSB, Rev. 3, 5 October, 1990, Test No. FV 52

MODELING AND RESULTS

The plate has three rigid body modes with zero frequencies. Vertical constraints are applied not at the
middle plane of the plate, but to the lower edges. This does not allow for clear separation of out-of-plane
and in-plane vibration modes. Also, in full 3D formulation constrained edges introduce stress singularity
into the solution.

ADDITIONAL COMMENTS
1. The reference solution has a very rough FEA mesh with just a few elements. Mesh lines
orientation and incompleteness of basis functions of iso-parametric elements results in artificial
“anisotropy” of the approximation
2. Vertical constraints are applied not at the middle plane of the plate, but to the lower edges. This
does not allow for clear separation of out-of-plane and in-plane vibration modes. This separation
in NAFEMS results is just because a very crude approximation is used. Also, in full 3D
formulation the constrained edges is the source of a strong singularity in the solution
3. The High Order Element (HOE) solution has only one layer of elements across thickness with
quadratic basic functions. Quadratic functions do not capture bending (you need 3rd order
functions). The Low Order Element (LOE) solution has three layers of bricks across the
thickness, but the basis functions are linear incomplete polynomials which do not capture
bending as well.
4. An additional OptiStruct model was run with a high density (800,000 element) mesh. Twenty
linear HEX elements were used through the thickness (Fig vm06b). This is felt to be a better
comparison.

Page 37
 Fig vm06 b – OptiStruct mesh

RESULTS

OptiStruct NAFEMS
Mode SimSolid % Difference % Difference
Solution Solution
Mode 4 42.45 42.47 -0.04% 45.90 -7.51%
Mode 5, 6 101.00 100.74 0.26% 109.44 -7.71%
Mode 7 148.42 147.89 0.36% 167.89 -11.60%
Mode 8 185.57 183.70 1.02% 174.00 6.65%
Mode 9 186.77 185.37 0.76% not given --
Mode 10 193.46 193.59 -0.07% 193.59 -0.07%

Page 38
Test VM07: In-plane Vibration of a Pin-ended Cross
Problem Definition
Find the first eight modal frequencies for in-plane vibration of a cross with a pin joints at ends (Fig.
vm07).

Pins are simulated via zero displacements (Fig. vm07) applied to a straight line. Front surface of the
cross is fixed from out-of plane displacements.

Units are SI.

Material properties: Modulus of elasticity E=2e11 N/m2 ,Poisson’s ratio ν=0.3,

mass density  = 8000 kg/m3.

Fig. vm07

Reference
NAFEMS, SBNFA (November 1987), Test 1.

Page 39
Results

REF Solution SimSolid % Difference

Mode 1 [Hz] 11.336 11.316 -0.18%
Mode 2 [Hz] 17.709 17.733 0.14%
Mode 3 [Hz] 17.709 17.790 0.46%
Mode 4 [Hz] 17.709 17.790 0.46%
Mode 5 [Hz] 45.345 45.126 -0.48%
Mode 6 [Hz] 57.390 57.202 -0.33%
Mode 7 [Hz] 57.390 57.204 -0.32%
Mode 8 [Hz] 57.390 57.303 -0.15%

Page 40
Test VM08: Cantilevered Beam
Problem Definition
Normal mode dynamic analysis of a cantilevered beam (shown below). Left end fully constrained, back
face has slider constraint to remove 3rd dimension.

Units are SI.

Material properties: Modulus of elasticity E=2e11 N/m2 , Poisson’s ratio ν=0.3,

mass density  = 8000 kg/m3.

Reference
NAFEMS Finite Element Methods & Standards, Abbassian, F., Dawswell, D. J., and Knowles,
N. C. Selected Benchmarks for Natural Frequency Analysis Test No. 71. Glasgow: NAFEMS,
Nov., 1987.

Results

REF Solution SimSolid % Difference

Mode 1 [Hz] 1.001 1.010 0.85%
Mode 2 [Hz] 6.327 6.321 -0.09%
Mode 3 [Hz] 17.720 17.679 -0.23%
Mode 4 [Hz] 34.720 34.586 -0.39%
Mode 5 [Hz] 57.390 57.052 -0.59%
Mode 6 [Hz] 85.730 85.002 -0.85%

Page 41
Test VM09: Circular Ring -- In-plane and Out-of-plane Vibration
Problem Definition
Normal mode dynamic analysis of an unconstrained circular ring (shown below).

Units are SI.

Material properties: Modulus of elasticity E=2e11 N/m2 , Poisson’s ratio ν=0.3,

mass density  = 8000 kg/m3.

Reference
NAFEMS Finite Element Methods & Standards, Abbassian, F., Dawswell, D. J., and Knowles,
N. C. Selected Benchmarks for Natural Frequency Analysis, Test No. 6. Glasgow: NAFEMS,
Nov, 1987.

Results

REF % NAFEMS %
Mode SimSolid
Solution Difference Solution Difference
7, 8 out-of-plane 51.486 51.850 -0.70% 52.290 -1.54%
9, 10 in-plane 53.140 53.380 -0.45% 53.970 -1.54%
11, 12 out-of-plane 147.770 148.800 -0.69% 149.700 -1.29%
13, 14 in-plane 149.430 151.000 -1.04% 152.400 -1.95%
15 out-of-plane 280.770 287.000 -2.17% 288.300 -2.61%
16 in-plane 280.770 289.500 -3.02% 288.300 -2.61%

Page 42
Test VM10: Triangular Wing -- Fixed Vibration
Problem Definition
Normal mode dynamic analysis of a fixed right triangular plate, with 6-inch length and 0.034-inch
thickness (shown below).

Units are IPS.

Material properties: Modulus of elasticity E=6.5e6 lb/in2 , Poisson’s ratio ν=0.3541,

Weight density= 0.06414 lb/in3, mass density  = 1.66128e-4 lb-s2/in4

Reference
“ASME Pressure Vessel and Piping 1972 Computer Programs Verification.", I.S. Tuba and
W.B. Wright, ASME publication, I-24, Problem 2

Results

REF Solution SimSolid % Difference

Mode 1 [Hz] 55.900 55.692 -0.37%
Mode 2 [Hz] 210.900 209.934 -0.46%
Mode 3 [Hz] 293.500 290.228 -1.11%

Page 43
Section 3: Thermal Analysis Tests

Test VT01: Cooling Fin

Problem Definition
A cooling fin 1x1x8 inch is surrounded by fluid with one end maintained at temperature
T=100 F, and the other end insulated (Fig. vt01). Find the temperature at the insulated end.

Ambient temperature Ta = 0. Material properties: thermal conductivity k = 25 BTU/(hr-ft-F), thermal

convection coefficient h = 1BTU/(hr-ft^2-F).

Prescribed
temperature

Convection

Insulated end

Fig. vt01

Reference
Kreith, F. Principles of Heat Transfer. 2nd ed. P.A.: International Textbook Co., 1959, pg. 48

Results

REF Solution SimSolid % Difference

Mean Temp at Insulated End [deg-F] 68.592 68.580 -0.02%
Thermal Flux Through Heated End [BTU/(hr*in^2] 17.500 17.153 -1.98%

NOTE – Reference solution is 1D model

Page 44
Test VT02: Insulated Wall
Problem Definition
A wall consists of two layers (firebrick and insulating brick) of thickness 9 inch and 5 inch
correspondingly. The temperature at firebrick surface is 3000 F and convection coefficient is
12 BTU/(hr-ft^2-F). The temperature at outer surface of insulating brick is 80 F and convection coefficient
is 2 BTU/(hr-ft^2-F). Find temperatures at firebrick and insulating brick surfaces.

Material properties: firebrick thermal conductivity is 0.8 BTU/(hr-ft-F), insulating brick thermal conductivity
is 0.1 BTU/(hr-ft-F).

Reference
Kreith, F. Principles of Heat Transfer. 2nd ed. P.A.: International Textbook Co., 1959, pg. 32

Solution Results
The wall is simulated by two solid blocks: 9x24x24 inch (firebrick), and 5x24x24 inch (insulating brick)
(Fig. vt02). Thermal resistance in contact between the blocks is neglected. Side surfaces of the blocks
are insulated in order to allow heat transfer only in the direction perpendicular to the wall.

Convection

Insulated

Convection
Insulated

Fig. vt02 boundary conditions Fig. vt03 Temperature distribution

REF Solution SimSolid % Difference

Temp at Firebrick [deg-F] 2957.0 2957.0 0.00%
Temp at Insulating Brick [deg-F] 336.0 336.3 0.09%
Heat flux Through Wall [BTU/(hr*ft^2)] 513.0 515.0 0.39%

NOTE – Reference solution is 1D model

Page 45
Test VT03: Cylinder with Prescribed Heat Flux
Problem Definition
A cylinder is loaded with a prescribed heat flux along a strip on side surface (see Fig. vt03a). The
bottom is maintained at zero temperature and the top and the rest of side surface are insulated. Find
temperature at boundary between strip with applied flux and the rest of side surface of the cylinder (point
A at Fig. vt03a).

Cylinder height is 50 mm, diameter 200 mm, the width of the heat flux strip 20 mm.
Material properties: thermal conductivity is 52 W/(m-K).

Heat flux value 5.e+5 W/(m^2).

Insulated

Heat flux
A

Insulated

Zero temperature

Fig. vt03a

Reference
This problem is a standard NAFEMS benchmark: NAFEMS, BMTTA (S), No. 15 (i).

Page 46
Results

Fig. vt03b Temperature distribution

REF Solution SimSolid % Difference

Temperature at Point A [deg-C] 213.6 212.5 -0.51%

Page 47
Test VT04: Heat Generating Plate
Problem Definition
A fluid is heated by a steel plate (Fig. vt04). Heat is uniformly generated in the plate volume. Find
temperature in the plate mid-plane and heat flow rate to the fluid.

A piece of plate is simulated by a 1x1x0.5 inch solid. Side surfaces of the solid are insulated in order to
allow heat transfer only in the direction perpendicular to the plate surface.

Plate material thermal conductivity is 25 BTU/(hr-ft-F), surface convection coefficient is 13.969738

BTU/(hr-ft^2-F), fluid temperature is 150 F, volume heat generation rate is 1.e+5 BTU/(hr-ft^3).

Reference
Kreith, F. Principles of Heat Transfer. 2nd ed. P.A.: International Textbook Co., 1959, pg. 42, ex. 2-9

Results
REF Solution SimSolid % Difference
Temperature at Mid-plane [deg-F] 299.1 300.0 0.30%
Heat Flow Rate to Fluid [BTU/hr] 2083.3 2083.3 0.00%

Page 48
 Insulated

Convection
Insulated

Fig. vt04 Temperature distribution in the plate

Page 49
Test VT05: Heat Generating Wire
Problem Definition
A bare steel wire is generating uniform heat at rate 111311.7 BTU/(hr-ft^3) (Fig. vt05). The wire is
surrounded by air at temperature 70 F. Surface convection coefficient between the wire and the air is 5
BTU/(hr-ft^2-F). Find temperature at the wire centerline and at the surface, and heat dissipation rate rate
to the air.

Wire diameter is 0.75 inch. Wire material thermal conductivity is 13 BTU/(hr-ft-F).

Reference
W.M.Rohsenow, H.Y.Choi, Heat, Mass and Momentum Transfer, 2nd ed. Prentice-Hall, Inc.,Englewood
Cliffs, NJ, 1963, pg. 106, ex. 6.5.

Results
The wire is simulated by a 5 inch long cylinder. End surfaces of the cylinder are insulated in order to
allow heat transfer only in the direction perpendicular to the cylinder centerline. Heat dissipation rate was
calculated through heat flow rate per wire 1 ft length.

Insulated

Convection

Fig. vt05 Temperature distribution in the wire

Page 50
 REF Solution SimSolid % Difference
Temperature at Centerline [deg-F] 419.9 418.6 -0.31%
Temperature at Surface [deg-F] 417.9 416.5 -0.34%
Heat Dissipation Rate [BTU/hr] 341.5 341.2 -0.09%

Page 51
Test VT06: Table Frame
Problem Definition
Find temperature at points A, B, C and D of the assembly of 8 tubes.

Units are SI.

Material properties: AISI 304 steel used.

Dimensions: Hollow tubular frame, L=2.25m, H=1.5m, D=1.5m

•

Reference
LM-TH-1 Learning Module 1, Thermal Analysis, Dr. Jack Xin, Dr. Liang-Wu Cai, www.femlearning.org,
2010.

Results

Location REF Solution SimSolid % Difference

A 110.2 112.0 1.63%
B 182.0 180.8 -0.66%
C 107.6 109.1 1.39%
D 143 143 0.00%

Page 52
Section 4: Thermal Loading Tests

Test VTL01: Thermally Loaded Support Structure

Problem Definition
A support structure of two copper bars, steel bar, and a rigid beam is subjected to a load and a
temperature rise after assembly (see Fig. vtl01a). Find stresses in the bars.

Steel bar material properties: modulus of elasticity 30.e+6 psi, coefficient of thermal expansion 70.e-7
in/in-F, copper bar correspondingly 16.e+6 psi and 92.e-7 in/in-F. Total load applied to the rigid beam is
4000 lbf, temperature change 10 F. Bars have cross-sectional area 0.1 inch^2 and length 20 inch.

Structure components are simulated by solids. Dimensions of the support bars are 20x0.5x0.2 inch, rigid
beam – 12x1x1 inch. The rods are equally spaced at a distance of 5 inch. Rigid beam is simulated as an
absolutely rigid body.

Reference
S.Timoshenko, Strength of Material, Part 1, Elementary Theory and Problems, 3rd Edition, D. Van
Nostrand Co., Inc., New York, NY, 1955, pg. 30, problem 9

Results

Distribution of max principal stress in the structure is shown in Fig. vtl01b.

REF Solution SimSolid % Difference

Stress in steel [psi] 19695.0 19697.0 0.01%
Stress in copper [psi] 10152 10151 -0.01%

Page 53
 Immovable

Copper
Copper

Steel

Rigid beam Fig. vtl01a

Fig. vtl01b

Page 54
Test VTL02: Bimetallic Cantilever Beam under Thermal Loading
Problem Definition
A cantilever beam is built from two equal thickness layers of different metals (Fig. vtl02a). The beam is
fixed at centerline at one end (translations and rotation around the centerline are set to zero) and is
subjected to uniform temperature change of 100 F. Determine the displacements at the centerline at free
end and outer fiber bending stress.

Both layers of materials have the same dimensions 10x1x0.05 inch, the same modulus of elasticity
3.e+7 psi, and zero Poisson’s ratio. Coefficient of thermal expansion of top layer is 2.e-5 in/in-F, bottom
layer – 1.e-5 in/in-F.

Fig. vtl02a

Reference
R.J. Roark, W.C. Young, Formulas for Stress and Strain, 5th Edition, McGraw-Hill Book Co.,Inc., New
York, NY, 1975, pp. 113-114

Results
Layers of the beam are simulated by solids glued together. In order to reproduce as close as possible
boundary conditions of beam theory used in reference, the edge at centerline at “fixed” end of the beam
is constrained in lateral directions. To eliminate rotations around the edge and sliding along the edge a
point at the corner of upper fiber is fixed in two directions (see Fig. vtl02b).

Because the beam is simulated using full 3D solid model, the constraints applied at the beam “fixed” end
introduce local disturbances which quickly dissipate along the beam axis (Fig. vtl02d).

Page 55
Fig. vtl02b Constraints at ‘fixed” end

Fig. vtl02c Beam deformation

Fig. vtl02d Distribution of bending stress X at “fixed” end

REF Solution SimSolid % Difference

Free End, Middle Fiber, Disp X [in] 1.5000E-02 1.4940E-02 -0.40%
Free End, Middle Fiber, Disp Y [in] 7.5000E-01 7.4940E-01 -0.08%
Top Fiber, Stress [psi] 7.5000E+03 7.5680E+03 0.91%

Page 56
Test VTL03: Coupled Thermal-Structural Analysis - Constrained Straight
Bar Under Uniform Temperature
Problem Definition
A straight 1x1x10 mm bar (Fig. vtl03a) is constrained at ends and is thermally loaded. Temperature
change of 100 C is applied to one end of the bar, the rest of the bar surface is insulated. Material
properties: modulus of elasticity 2.1e+11 pa, and Poisson’s ratio 0.3, coefficient of thermal expansion
1.2e-5 m/m-C, thermal conductivity 43 W/m-K.

Fig. vtl03a

Fig. vtl03b

Page 57
Reference
W.C. Young, R.G. Budynas, Roark’s Formulas for Stress and Strain, Seventh Edition, McGraw-Hill Book
Co.,Inc., 2001, p. 759

Results
Bar is simulated using 3D solid model. At first, thermal analysis is performed. Due to applied thermal
conditions bar reaches constant temperature of 100 C everywhere.

At second, the result of thermal analysis is used as input for structural analysis. In order to make use of
reference data obtained for 1D model of the bar the following constraints are applied: sliding supports
are applied to end surfaces of the bar and two vertices at one end of the bar are constrained in order to
eliminate rigid body motion of the bar (Fig. vtl03b).

REF Solution SimSolid % Difference

Stress along bar (compression) [pa] 2.520E+08 2.516E+08 -0.17%

Page 58
Test VTL04: Coupled Thermal-Structural Analysis - Constrained Straight
Bar Under Linear Temperature Field
Problem Definition
A straight 1x1x10 mm bar (Fig. vtl04a) is constrained at ends and is thermally loaded. A uniform
temperature of 100 C is applied to the top of the bar, and a uniform temperature of -100 C is applied to
the bottom of the bar. The rest of the bar surfaces are insulated. Material properties: modulus of
elasticity 2.1e+11 pa, and Poisson’s ratio 0.3, coefficient of thermal expansion 1.2e-5 m/m-C, thermal
conductivity 43 W/m-K.

Temperature 100 C on top

Sliding
Insulated Sliding

Temperature -100 C at bottom

Fig. vtl04a

Vertex is constrained in
vertical direction
Vertex is constrained in both lateral
directions

Page 59
Fig. vtl04b

Reference
W.C. Young, R.G. Budynas, Roark’s Formulas for Stress and Strain, Seventh Edition, McGraw-Hill Book
Co.,Inc., 2001, p. 759

Solution Results
Bar is simulated using 3D solid model. Temperature distribution obtained in thermal analysis is shown in
Fig. vtl04c. Temperature changes linearly across the bar from 100 C at the top to -100 C at the bottom.

Fig. vtl04c Temperature distribution

Obtained temperature field was used as input for Structural analysis. In order to make use of reference
data obtained for 1D model of the bar the following constraints are applied: sliding supports are applied
to end surfaces of the bar and two vertices at one end of the bar are constrained in order to eliminate
rigid body motion of the bar (Fig. vtl04b).

Bar deformation and stress X distribution is shown in Fig. vtl04d.

Page 60
Fig. vtl04d Bar deformation and stress X distribution

REF Solution SimSolid % Difference

Stress at the top/bottom of the bar
2.520E+08 2.516E+08 -0.17%
(compression/tension) [pa]

REF Solution is 1D model

Page 61
Section 5: Non-linear Tests

Test VNL01: Coupled Analysis - Bimetallic Beam Under Thermal Load

Problem Definition
A bimetallic beam consists of two layers of materials with different coefficients of thermal expansion. The
beam is loaded with uniform temperature applied to its top and bottom surfaces (Fig. vnl01, a). Sides
and ends of the beam are insulated. One end of the beam is supported by a hinge, the other end is
supported by a sliding hinge. Find temperature at the material interface and maximum deflection of the
beam.

The layers of materials are equal: 0.05x0.5x10 inch. Material properties:

Material 1: elasticity modulus 1.e+7 psi, Poisson's ratio 0, coefficient of thermal expansion 14.5e-6 1/F,
thermal conductivity 5 BTU/(hr-in-F).
Material 2: elasticity modulus 1.e+7 psi, Poisson's ratio 0, coefficient of thermal expansion 2.5e-6 1/F,
thermal conductivity 5 BTU/(hr-in-F).

Temperature at top and bottom surfaces is 400 F.

Fig. vnl01, a

Reference
B.A.Boley, J.H.Weiner, Theory of Thermal Stress, R.E.Krieger Publishing Co, Malabar, FL, 1985, pg.

Page 62
429

Results
Layers of different materials are simulated by solids. Hinged supports are simulated by applying
immovable edge constraint at contact surface at one end of the bar (Fig. vnl01, b) and sliding edge
support at the other end (Fig. vnl01, c).

Fig. vnl01, b Fig. vnl01, c

Fig. vnl01, d Deformed shape of the beam

Page 63
Fig. vnl01, e Sliding of the simply supported end of the beam due to non-linear deformation

REF Solution SimSolid % Difference

Temperature at Contact [deg-F] 400.00 400.00 0.00%
Max Deflection [in] 0.90 0.90 -0.45%

REF Solution is 1D model

Page 64
Test VNL02: Z-Shaped Cantilever
PROBLEM DEFINITION
A Z-shaped cantilever is under action of a transverse load of max 4000 N (Fig. VNL02 A).

Fig. VNL02 A

The load is conservative, its direction does not change while the cantilever deforms. Material properties:
elasticity modulus 2.e+11 pa, Poisson's ratio 0.3.

REFERENCE
A.A.Becker, Background to Finite Element Analysis of Geometric Non-linearity Benchmarks, Published
by NAFEMS, Ref:-R0065, pg. 61 Page 66

Page 65
RESULTS

Table below contains values of vertical displacement Uz for different loads in mm.

Load [N] SOL Reference SimSolid % Difference

300 113.0 113.3 0.24%
500 124.1 122.1 -1.58%
1000 131.0 133.3 1.78%
2500 139.5 143.4 2.82%
4000 144.0 144.8 0.53%

SOL Reference is thin shell model

Von Mises stress distribution at load 4000 N is shown in Fig. VNL02 B

Fig. VNL02 B

Page 66
Test VNL03: Lateral Buckling of a Right Angle Frame
Problem Definition
Perform analysis of buckling and post-buckling of a thin-walled frame-like structure in the shape of right
angle (Fig. vnl03, a). One end of the structure is clamped, and a conservative load in X-directions is
applied to the other end.

Material properties: elasticity modulus 7.124e+10 pa, Poisson's ratio 0.31.

255 mm

30 mm

Clamped end

Plate thickness 0.6 mm

255 mm

30 mm

X
Loaded end

Fig. vnl03, a Geometry and loading

Reference
R.Kouhia, On Kinematical Relations of Spatial Framed Structures, J. Computers & Structures, Vol. 40,
No.5, 1991, pp. 1185-1191

Page 67
Results
Structure is simulated as a solid. Reference critical load found from 1D beam bending model of the
structure is P = 0.820035 N. Negligible perturbation load of Pz = P*1.e-4 = 8.2e-5 N is applied in lateral
Z-direction at the loaded end of the structure in order to cause a very small initial deflection of the
structure from the flat. The main load was gradually increased from 0 to 1.6*P.

Graph in Fig. vnl03, b shows maximum deflections vs. applied load. The load-lateral deflection curve is
in a good agreement with the results reported in the reference. Transversal post-buckling behavior at
first is associated with a very small transversal stiffness, but then the load carrying capacity increases
(load-displacement curve flattens). In-plane behavior is almost monotonic, no capacity increase occurs
in the range of load change.

3D solution predicts the critical load approximately 10% lower than 1D beam model solution. Fig. vnl03,
c and vnl03, d show Von Mises stress distribution which is substantially different from the one assumed
in the beam bending model.

250

200

150
Deflection

100

50

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Load, N

Max in-plane deflection X, mm Max transversal deflection Z, mm

Max transversal deflection Z, mm [SIMSOLID] Max in-plane deflection X, mm [SIMSOLID]

Fig. vnl03, b Load-deflection graphs

Page 68
Fig. vnl03, c Deformed shape and Von Mises stress plot at load 1.1 N (front view to the left and top
view to the right)

Fig. vnl03, d Deformed shape and Von Mises stress plot at load 1.4 N

Page 69
Test VNL04: Reactions at the ends of statically undefined axially loaded
plastic bar
PROBLEM DEFINITION
Find reactions at the fixed ends and maximum displacement of a bar axially loaded beyond plasticity
(Fig. vnl04, a).

N F
R

A
L

Fig. vnl04, a

Bar dimensions are 10x10x200 mm. Distance between loaded point and left end A=50 mm. Strain-stress
curve of the bar material is defined by the power law:
𝜎 = 𝐾𝜀 𝑛 ,
where K is strength coefficient, value n must be in the range [0,1].
At n=0 material is perfectly plastic and at n=1 material is elastic.

Reactions study is performed for the material with the following properties: K=530 MPa, n=0.26,
Poisson's ratio 0. Corresponded strain-stress curve is shown in Fig. vnl04, b

Fig. vnl04, b
The study was performed for the following load F values: 30000 N, 47000 N, 55000 N, and 60000 N.
These loads cover the full range of elastic-plastic response of the bar.

Page 70
REFERENCE SOLUTION

One-dimensional analytical reference solution can be found as follows.

Let us designate ε1 is tensile strain at the left span of the bar, ε2 is compressive strain at the right span
of the bar, N is reaction force at left end of the bar, R=F-N is reaction force at the right end of the bar.
The length of the bar does not change under the load, therefore
𝐴 𝐿−𝐴
∫ 𝜀1𝑑𝑥 − ∫ 𝜀2𝑑𝑥 = 0
0 0

or
𝐴 𝐿−𝐴
𝑛 𝑛
∫ √𝑁/(𝐾 ∗ 𝐴)𝑑𝑥 − ∫ √(𝐹 − 𝑁)/(𝐾 ∗ 𝐴)𝑑𝑥 = 0
0 0

Here A is the bar cross section area.

From this equation one can find the reaction at the left end of the bar

𝑁 = 𝐹 ⁄(1 + (𝑎/𝑏)𝑛 )
and
𝑅 = 𝐹−𝑁
at the right end.

MODELING AND RESULTS

Bar was modeled as a 3D solid with immovable ends. Axial force F could not be applied precisely at the
solid bar axis, so four line spots were created at the bar sides and total load F was uniformly distributed
over the spots (Fig. vnl04, c).

Fig. vnl04, c

Reaction force results are summarized in the following table:

SOL Reference, SimSolid,

Force F [N] % Difference
Reaction [N] Reaction [N]
30000 17128 18260 6.61%
47000 26834 27290 1.70%
55000 31401 31900 1.59%
60000 34256 34740 1.41%

Page 71
Typical Von Mises stress distributions are shown in Fig. vnl04, d, e. As one can see the distribution has
high gradients at load application lines. Yet the reactions values well correlate to the 1D solution
because the reactions are applied far from the active force.

Fig. vnl04, d. Von Mises stress distribution at load F=30000 N

Fig. vnl04, e. Von Mises stress distribution at load F=60000 N

Page 72
Test VNL05: Residual deformations in an axially loaded plastic bar
PROBLEM DEFINITION
Find residual deformations in a bar axially loaded beyond plasticity (Fig. vnl05, a).

Fig. vnl05, a

Bar dimensions are 10x10x200 mm. Strain-stress curve of the bar material is defined by the power law:
𝜎 = 𝐾𝜀 𝑛 ,
where K is strength coefficient, value n must be in the range [0,1].
At n=0 material is perfectly plastic and at n=1 material is elastic. Left end of the bar is clamped, right end
is loaded with force F
Residual deformation study is performed for the material with the following properties: K=530 MPa,
n=0.26, Poisson's ratio 0. Corresponded strain-stress curve is shown in Fig. vnl05, b. Note, that elasticity
modulus defined by the first two points of the strain-stress curve is E=2.67324e+10 Pa.

Fig. vnl05, b
The study was performed for the following load F values: 20000 N, 25000 N, and 30000 N.

Page 73
REFERENCE SOLUTION

One-dimensional analytical reference solution can be found as follows.

Let us designate ε is total strain in the bar, εe is elastic component of the total strain, εr is residual strain.
At strain ε and stress σ the residual strain is

εr = ε - εe = (σ/K)1/n - σ/E = (F/(K*A))1/n - F/(E*A),

where A is cross-section area of the bar.

Then residual displacement at the right end of the bar

𝐿 𝑛
𝐹
𝑈𝑟 = ∫ ( √ − 𝐹/(𝐸 ∗ 𝐴))𝑑𝑥
0 𝐾∗𝐴

MODELING AND RESULTS

Bar was modeled as a 3D solid. Left end of the solid was fixed, right end loaded with axial force (Fig.
vnl04, c).

Fig. vnl05, c

Results on residual deformations are summarized in the following table:

SOL Reference, SimSolid, Residual

Force F [N] % Difference
Residual Disp [mm] Disp [mm]
20000 3.22 3.43 6.78%
25000 9.16 9.172 0.13%
30000 19.077 19.14 0.33%

Page 74
Typical Von Mises stress distributions are shown in Fig. vnl05, d.

Fig. vnl05, d. Von Mises stress distribution at load F=30000 N

Page 75
Test VNL06: Pure bending of a plastic beam
PROBLEM DEFINITION
Find elastic core and maximum displacements in a cantilever beam loaded with a moment (Fig. vnl06,
a).

h M

Fig. vnl06, a

Beam dimensions are b x h x L, where b=10 mm, h=40 mm, and L=200 mm. Material of the beam is rigid
plastic with the strain-stress curve shown in Fig. vnl06, b:

Fig. vnl06, b
As one can see from the curve, yield stress σy = 2.1188e+8 Pa, yield strain εy = 1.73425e-3.
Poisson's ratio 0.

The study was performed for the following moment M values: 776.893 N*m and 847.518 N*m.

Page 76
REFERENCE SOLUTION

Plastic beam bending theory assumes that there are two zones in the beam material: plastic zone at the
outer layers of the beam and elastic core at the beam centerline. Corresponding stress distribution for
material without hardening is shown in Fig. vnl06, c.

H h

Fig. vnl06, c

The relation between applied moment M and the size H of the elastic core in the beam

𝑀 = σy*b*(h2/4 - H2/3),
here b is beam thickness.
Maximum deflection
Umax = εy*L2/(2*H).

REFERENCE
Mase, George E., “Theory and Problems of Continuum Mechanics”, McGraw-Hill Company, New York,
1970.

MODELING AND RESULTS

The theory of bending of plastic beams is based on hypothesis of flat cross-sections which remain flat
during deformations. In order to emulate this basic assumption as close as possible the problem was
modeled as an assembly of two solids. One solid represented the beam itself with material properties
defined by the curve shown in Fig. vnl06, b. Another small solid was attached to the beam end and was
set as an absolutely rigid body (Fig. vnl06, d). The rigid solid was loaded with the moment M, it served
as a moment transfer element.

Fig. vnl06, d

Page 77
Figures below show the results of the simulation which are summarized in the following tables

SOL Reference, Elastic SimSolid, Approximate

Moment M [N*m] % Difference
Core Size [mm] elastic core size [mm]

776.893 10.00 12.00 20.00%

847.518 5.00 6.00 20.00%

SOL Reference, Max SimSolid, Max

Moment M [N*m] % Difference
Displacement [mm] Displacement [mm]
776.893 3.468 3.351 -3.37%
847.518 6.937 6.918 -0.27%

It should be mentioned that SimSolid 3D solution does not have sharp boundary between elastic and
plastic zones as it is in simplified beam bending formulations. This makes difficult direct comparison.
Nevertheless, the elastic zone approximate sizes are in good correlation.

Fig. vnl06, e. Stress X at M=776.893 N*m

Page 78
 Fig. vnl06, j. Stress X at M=847.518 N*m

Fig. vnl06, k. Von Mises stress at M=776.893 N*m

Fig. vnl06, l. Von Mises stress at M=847.518 N*m

Page 79
Fig. vnl06, m. Residual Von Mises stress after unloading at M=776.893 N*m

Fig. vnl06, n. Residual Von Mises stress after unloading at M=847.518 N*m

Page 80
Test VNL07: Separating beams
PROBLEM DEFINITION
Two similar beams with dimensions 10x10x100 mm are in contact (Fig. vnl07, a). Find total or partial
beam separation/slippage depending on the applied loads. Friction coefficient is f=0.15. Material
properties: elasticity modulus 2.1e+11 Pa, Poisson's ratio 0.3.

P Fy

Fx

Fig. vnl07, a
Five load cases are considered. In all load cases the lower beam bottom is fixed, the upper beam is
loaded on its top with total vertical load of 2 N uniformly distributed over the surface. The right end of the
upper beam is loaded either with vertical load Fy or with horizontal load Fx uniformly distributed over the
surface.

Load case Fx, [N] Fy, [N]

1 0 1
2 0 0.9
3 0.3 0
4 0.31 0
5 0.29 0

Page 81
MODELING

Beams were simulated as two solids shown in Fig. vnl07, b. Contact condition at the connection was set to
"Separating" with friction coefficient 0.15.

Fig. vnl07, b

REFERENCE SOLUTIONS AND RESULTS

Load case 1.
P Fy

A
Fig. vnl07, c

Let us consider the equilibrium of the upper beam. If Fy=0, then force N is equilibrated by the reaction
force in contact. As Fy grows, the beam deforms, and the contact separation starts at its right end and
expands to the left. Ultimately, the contact fully separates and contact area degenerates into a line (point
A in Fig. vnl07, c). The value of force Fy which results in full separation can be found from moments
equilibrium equation

𝐿
∑ 𝑀(𝐴) = 𝑃 ∗ ( ) − 𝐹𝑦 ∗ 𝐿 = 0
2

This ultimate value is Fy =1 N.

SimSolid result for this value of Fy is shown in Fig. vnl07, d. As one can see full separation did happen
and the contact occurs only along the single edge which causes stress concentration at the beam
corner.

Page 82
 Fig. vnl07, d

Load case 2.

In this load case force Fy=0.9 N is not sufficient to cause full separation of the contact (Fig. vnl07, e).

Page 83
 Fig. vnl07, e

Load case 3.

Fx

Friction forces

Fig. vnl07, g

In this load case normal separation does occur because there is no detaching force - Fy=0. Shear force
Fx tends to cause beam slippage which is resisted by friction forces distributed over the contact area. In
equilibrium the force projections onto horizontal axis:
F - Fx = 0,
where F is total friction force.
Maximum total friction force is
F = f*P.

Page 84
Total slippage (or total tangential separation) starts when Fx >= f*P. When Fx < f*P only partial slippage
is possible. Maximum total friction force is Fmax = 0.15*2 = 0.3 N.

The result for Fx = 0.3 N is shown in Fig. vnl07, k. As one can see full slippage develops in contact area,
but left lower edge of the upper beam is still almost coincides with the edge of the lower beam.

Fig. vnl07, k

Load case 4.
In this load case the shifting force Fx is increased to 0.31 N. The force exceeds the maximum friction
force and the upper beam not only deforms but also starts moving as a rigid body.

Load case 5.
In this load case the shifting force Fx is decreased to 0.29 N in order to not to exceed the maximum
friction force. As one can see only partial slippage occurs in this case. There is a clear "sticking" contact
area at the left end of the beams.

Page 85
Test VNL08: Pinched Hemispherical Shell
PROBLEM DEFINITION
A hemispherical shell is loaded with inward and outward concentrated forces at point A and point B
respectively. The hemisphere has 18˚ hole at the top and the quadrant of the hemisphere is modeled
utilizing symmetric boundary conditions (Fig. vnl08, a). Correspondingly, forces P shown in the Fig.
vnl08,a are acting on the quadrant. Displacement at points A and B are to be determined for force values
P = 40, 60, and 100 lbf.

Fig. vnl08,a

Material properties: elasticity modulus 6.825e+7 psi, Poisson's ratio 0.3.

REFERENCE
Test 3DNLG-9 from NAFEMS Publication R0024 “A Review of Benchmark Problems for Geometric Non-
linear Behaviour of 3-D Beams and Shells (SUMMARY).”

MODELING AND RESULTS

Symmetry conditions were simulated via sliding boundary conditions applied at faces coinciding with
symmetry planes (Fig. vnl08,b). Concentrated forces were applied in points of outer face of the sphere
(Fig. vnl08,c,d). In order to eliminate rigid body motion along Z-axis a point on the top of the sphere was
constrained in Z-direction (Fig. vnl08,e).

Page 86
In Fig.vnl08,m,n true deformations of the shell at load P = 100 lbf are shown. Results of comparison are
summarized in the table below.

40 -3.280 -3.100 -5.49%

60 -4.360 -4.169 -4.38%
100 -5.950 -5.642 -5.18%

REF Solution, Uy at SimSolid, Displacement

Load [lbf] % Difference
point B [in] Ux at point B [in]
40 -2.330 -2.219 -4.76%
60 -2.830 -2.727 -3.64%
100 -3.430 -3.306 -3.62%
REF Solution is thin shell model

Fig. vnl08,b

Page 87
Fig. vnl08,c,d

Fig. vnl08,e

Page 88
Fig. vnl08, m

Page 89
Fig. vnl08, n

Page 90
Test VNL09: Rigid Punch Plasticity

PROBLEM DEFINITION
A 2D plane strain punch is pressed on larger block of finite dimensions. Punch is assumed rigid and
contact condition between punch and the block are assumed frictionless sliding (line BC). Vertical
downward displacement D is applied to the punch (Fig. vnl09,a). Symmetry condition is assumed along
line ABL, and sliding condition is applied along line LK.

A
D

B
C

L
K

Fig. vnl09,a

Dimensions: AB = 40 mm, BC = 80 mm, BL = 160 mm, LK = 200 mm.

The elastic properties are: elasticity modulus 1.e+3 MPa, Poisson's ratio 0.3. Two plasticity models are
used:
a) elastic perfectly plastic with yield stress 1 MPa
b) Isotropic hardening with tangent modulus 0.1e+3 MPa

REFERENCE
Test PL-7 from NAFEMS Publication R0049, “Background to Material Non-Linear Benchmarks”.

MODELING AND RESULTS

Geometry of rigid punch and the base was simulated with two 3D solid plates with thickness 10 mm (Fig.
vnl09,b). Plane strain condition assumes zero strain perpendicular to the plate side. In order to meet the
condition zero normal displacements were applied to both sides of the plates (Fig. vnl09,c).

Page 91
Fig. vnl09, b,c

Prescribed displacement of the punch was applied in six steps as follows: D = 0.04, 0.12, 0.14, 0.16,
0.18. 0.24 mm.

Table below contains values of the reaction force at the punch per 1 mm of the plates thickness.

SimSolid reaction, N Reference reaction, N

D, mm Perfect plasticity Hardening Perfect plasticity Hardening
0.04 29.57 29.57 30 30
0.12 86.95 87.27 86 87
0.14 99.93 100.65 99 100
0.16 108.22 111.17 109 113
0.18 108.87 113.03 110 114
0.24 111.75 122.20 111 125

Page 92
Figures below show evolution of equivalent stress with the punch displacement increments

Perfectly plastic material Isotropic hardening material

D=0.04 mm

D = 0.12 mm

D = 0.14 mm

Page 93
 Perfectly plastic material Isotropic hardening material

D = 0.16 mm

D = 0.18 mm

D = 0.24 mm

Page 94
Section 6: Transient Dynamic Tests

Test VD01: Simply supported thin square plate. Harmonic forced vibration
response

PROBLEM DEFINITION
A simply supported thin square plate 10 x 10 x 0.05 m is subject to uniform pressure P=100 Pa which
changes in time as the following function
P=100*sin(ω*t),
here ω=2*PI*f, f is excitation frequency.
The excitation frequency changes in range from 0 to just above first resonant frequency of the plate.

16 modes are used to approximate dynamics solution, 2% modal damping is assumed in all modes.

Material properties: elasticity modulus 2.e+11 Pa, Poisson's ratio 0.3, density 8.e+3 kg/m^3.

REFERENCE
Test 13H from NAFEMS Publication R0016, “Selected Benchmarks for Forced Vibration” J.Maguire,
D.J.Dawswell, L.Gould.

MODELING AND RESULTS

Plate is simulated as a 3D solid body. Spot-lines were created at the mid-plane of the plate in order to
apply hinge supports (see fig. vd01,a).

Fig. vd01,a

Peak response is reached when excitation frequency becomes equal to first natural frequency of the
plate. It happens 19.82 seconds after the beginning of the excitation.

Page 95
Fig. vd01,b

The following table contains peak response comparison.

Deflection Y, mm Surface stress, MPa

45.34 31.84 SimSolid, solid model
45.42 35.08 Reference, thin plate model

Fig. vd01,d shows transient and steady state portion of response.

Page 96
Fig. vd01,d

Page 97
Test VD02: Simply supported thin square plate. Periodic forced vibration
response

PROBLEM DEFINITION
A simply supported thin square plate 10 x 10 x 0.05 m is subject to uniform pressure P=100 Pa which
changes in time as the following function
P=100*(sin(ω*t) - sin(3ω*t))
here ω=2*PI*f, f=1.2Hz is excitation frequency.

16 modes are used to approximate dynamics solution, 2% modal damping is assumed in all modes.

Material properties: elasticity modulus 2.e+11 Pa, Poisson's ratio 0.3, density 8.e+3 kg/m^3.

REFERENCE
Test 13P from NAFEMS Publication R0016, “Selected Benchmarks for Forced Vibration” J.Maguire,
D.J.Dawswell, L.Gould.

MODELING AND RESULTS

Plate is simulated as a 3D solid body. Spot-lines were created at the mid-plane of the plate in order to
apply hinge supports (see fig. vd02,a). In order to simulate periodic load changing according to the
formula above, two pressure loads were applied to the plate:
P=100*sin(ω*t), to the top surface of the plate, and
P=-100*sin(3ω*t) to the bottom surface of the plate

Fig.vd02,a
Table below contains typical values at steady state portion of the dynamics solution (see Fig. vd02,b).

Deflection Y, mm Surface stress,

MPa
-2.91 2.048 SimSolid, solid model
-2.863 2.018 Reference, thin plate
model

Page 98
Fig. vd02,b

Page 99
Test VD03: Simply supported thin square plate. Transient forced vibration
response

PROBLEM DEFINITION
A simply supported thin square plate 10 x 10 x 0.05 m is subject to suddenly applied uniform pressure
P=100 Pa which remains constant in time. Find peak dynamics response of the plate.
16 modes are used to approximate dynamics solution, 2% modal damping is assumed in all modes.

Material properties: elasticity modulus 2.e+11 Pa, Poisson's ratio 0.3, density 8.e+3 kg/m^3.

REFERENCE
Test 13T from NAFEMS Publication R0016, “Selected Benchmarks for Forced Vibration” J.Maguire,
D.J.Dawswell, L.Gould.

MODELING AND RESULTS

Plate is simulated as a 3D solid body. Spot-lines were created at the mid-plane of the plate in order to
apply hinge supports (see fig. vd03,a).

Fig. vd03,a

Table below contains values at peak.

Peak Deflection Y, mm Surface stress,

time, MPa
sec
0.211 -3.47 2.72 SimSolid, solid model
0.210 -3.52 2.41 Reference, thin plate
model

Figure below shows transient response evolution in the plate center.

Page 100
Page 101
Test VD04: Simply supported thin square plate. Frequency response

PROBLEM DEFINITION
A simply supported thin square plate 10 x 10 x 0.05 m is subject to uniform pressure P=100 Pa which
changes in time as the following function
P=100*sin(ω*t),
here ω=2*PI*f, f is excitation frequency.

16 modes are used to approximate dynamics solution, 2% modal damping is assumed in all modes.
Material properties: elasticity modulus 2.e+11 Pa, Poisson's ratio 0.3, density 8.e+3 kg/m^3.

This test is similar to test VD01, but it is solved in frequency domain.

REFERENCE
Test 13H from NAFEMS Publication R0016, “Selected Benchmarks for Forced Vibration” J.Maguire,
D.J.Dawswell, L.Gould.

MODELING AND RESULTS

Plate is simulated as a 3D solid body. Spot-lines were created at the mid-plane of the plate in order to
apply hinge supports (see fig. vd04,a).

Fig. vd04,a

Peak response is reached when excitation frequency becomes equal to first natural frequency of the
plate. Fig.vd04,b shows frequency sweep result for Y-displacement amplitude in picked point with
coordinates [m] X = 4.9469, Y = 0.05., Z = -4.9158

Page 102
Fig. vd04,b

Page 103
Fig. vd04,c

The following table contains peak response comparison.

Deflection Y Surface stress
amplitude, mm amplitude, MPa
SimSolid 45.68 32.52
3D Solid model
Reference 45.42 35.08
Thin plate model

Page 104
Test VD05: Simply supported thin square plate. Random forced vibration
response

PROBLEM DEFINITION
A simply supported thin square plate 10 x 10 x 0.05 m is subject to uniform pressure P=100 Pa which
forces random vibrations with uniform Power Spectral Density function ("white noise")

16 modes are used to approximate dynamics solution, 2% modal damping is assumed in all modes.
Material properties: elasticity modulus 2.e+11 Pa, Poisson's ratio 0.3, density 8.e+3 kg/m^3.

REFERENCE
Test 13H from NAFEMS Publication R0016, “Selected Benchmarks for Forced Vibration” J.Maguire,
D.J.Dawswell, L.Gould.

MODELING AND RESULTS

Plate is simulated as a 3D solid body. Spot-lines were created at the mid-plane of the plate in order to
apply hinge supports (see fig. vd05,a).

Page 105
Fig. vd05,a

Peak response is reached when excitation frequency becomes equal to first natural frequency of the
plate. Fig.vd05,b shows PSD response for Y-displacement in picked point with coordinates [m] X =
4.9627, Y = 0.05, Z = -5.0418

Fig. vd05,b

Page 106
Fig. vd05,c shows distribution of Y-displacement PSD.

Fig. vd05,c

The following table contains peak response comparison.

PSD Y- PSD surface
displacement, stress X,
mm^2/Hz MPa^2/Hz
SimSolid 2086.8 1058.4
3D Solid model
Reference 2063.2 1025.44
Thin plate model

Page 107