TITLE NO.

67-37

Ultimate Strength of Prestressed

Concrete Piles and Columns

By ARTHUR R. ANDERSON and SAAD E. MOUSTAFA

This paper deals with the ultimate strength of crete columns, subjected to combinations of axial
prestressed concrete prismatic sections subjected force and bending. A strength reduction factor q,
to axial loads and bending. Although the tech- of 0.7 has already been included in the diagrams
niques developed are applicable to prestressed
concrete columns in general, the potential for for axial loads in the range from 100 percent down
practical application is mainly related to the de- to about 10 percent of the ultimate load, and be-
sign of prestressed concrete piles, which func- low this level, the q, factor increases linearly to a
tion structurally as a special case of reinforced maximum of 0.9 for the case of pure flexure. The
concrete columns. The full potential of prestressed sudden break in the curves results from this arbi-
concrete piles has not been recognized by struc-
tural designers because readily applicable criteria trary variation in q,. Use of a varying value of q, is
based on ultimate loads and moments for design
have not been available. TABLE I - PILE SECTION PROPERTIES
This paper contains interaction diagrams for the
ultimate capacity of a series of standard pre-
PILE D d A I r ~ESTII!ESS w LmalC, FT
StCTION in in in 2 in 4 in N PSl l~t 2-PT 3-PT
stressed concrete piles produced in the Pacific I) (2) (3) (4) (4)
Northwest. These piles are furnished in concrete

[Q] 830 2.9 5

strengths of 6000, 7000, and 8000 psi, all stressed 10 100 1240 110 75 100
to an effective prestress level of 1200 psi. 12 144 1728 3.5 7 1200 1!'>8 80 110
Applications of the interaction diagrams to de-
sign are included, and a comparison of the ulti- 14 196 3200 4.0 9 1130 215 85 , 20
mate strength of prestressed concrete piles pre- 1:--D---j

'b
dicted by the authors' method is made with 14 162 2105 3.6 8 1220 178 80 105
previously published work.
The construction of the interaction diagrams is 16~ 225 4080 4.2 II 1210 247 85 115
explained, and the computer program used for 18 268 5750 4.6 13 II 90 295 90 120
the construction is included in the paper.
20 331 8770 5.1 16 1200 363 95

~D~
125
Keywords: axial loads; bending; bridge foun-
dations; columns (supports); compressive strength; 24 477 18180 6.1 2.2. 1150 525 105 135
computer programs; concrete piles; deflec-

a
tion; eccentricity; foundations; interaction dia-
grams; moments; prestressed concrete; slenderness
ratio; stresses; structural design; ultimate strength 24 IS 300 15700 7.2 15 1230 330 120 150
method; wharves.

• THE ULTIMATE CAPACITIES of the prestressed ~D~

concrete piles shown on load-moment interaction

a
36 26 487 60000 II. I 24 122.0 533 140 190
diagrams are based on an effective concrete pre-
stress of 1200 psi and for concrete cylinder com- 48 38 675 158 200 15.3 34 1230 740 165 220
pression strengths of 6000, 7000, and 8000 psi. Sec- 54 44 770 233400 17.4 38 1210 845 lBO 240
tion properties for the piles are given in Table 1.
Notes:
DESCRIPTION OF THE INTERACTION 1. Number of tendons, ¥.! in. diameter 270 ksi stress-
relieved seven wire strands.
DIAGRAMS 2. Effective precompression in the concrete, after
losses in steel tension due to shrinkage and creep.
These diagrams are essentially three-dimen- 3. Weight of pile based on 158 lb per cu ft concrete.
sional charts in which the parameters are: (1) ul- 4. Max pile lengths for zero flexural tension in con-
crete with optimum location of support points.
timate applied axial force; (2) ultimate applied 5. Concrete cover over reinforcement, 2 in. clear.
bending moment; and (3) slenderness ratio h' jr. 6. Spiral reinforcement, pitch 2 in. at ends, 4 in. over
remaining length. No. 4 wire gage for solid pile sec-
The piles are assumed to act as prestressed con- tions, No. 2 wire gage for hollow sections.

620 ACI JOURNAL I AUGUST 1970

an attempt to accommodate the ACI philosophy
that assumes a strength reduction factor cp of 0.9 ACI member Saad E. Moustafa is senior research and plan-
ning engineer, Concrete Technology Corporation, Tacoma,
for flexural members and 0.7 for columns. Wash. He received a BS from Cairo University, United Arab
Republic (UAR), in 1961, a MS from Cornell University in
1966, and a PhD degree from the University of California in
DEVELOPMENT OF THE ULTIMATE 1968. From 1961-1964, he worked as a professional engineer
STRENGTH THEORY FOR PRESTRESSED in Cairo (UAR), where he participated in the design of several
CONCRETE PILES reinforced concrete high rise buildings. Dr. Moustafa also
taught at Cario University.
It is assumed that the pile section is symmetri- ACI past president Arthur R. Anderson is cofounder, Con-
cal about one axis, and that the strain distribution crete Technology Corporation and principal, ABAM Engineers,
Inc. He received his BS in civil engineering from the University
across the section is linear at ultimate load. Be- of Washington in 1934, his MS from the Massachusetts Institute
cause the relationship between compression stress of Technology in 1935, and his DSc from MIT in 1938. He is
the author of numerous technical papers published in the
and strain in concrete is nonlinear, a stress dia- United States and abroad, dealing with scientific work in civil
gram as proposed by Hognestad (see Fig. 1) ,1 was engineering end the technology of concrete. Dr. Anderson is a
member of several national and international professional en-
used as a general basis for ultimate strength gineering societies. In 1968 he was the recipient of the
analysis. Alfred Lindau Medal and back in 1962 of the Construction
Practice Award (both from the ACI). From 1946 to 1951 he
The stress in a prestressed concrete pile at ulti- was a consulting engineer in Stamford, Conn., and from 1941
mate load results from three causes: (1) prestress, to 1946 the head of the Technical Department, Cramp Ship-
yard, US Navy Bureau of Ships, Philadelphia, Penn. Currently,
(2) applied axial force, and (3) flexure due to Dr. Anderson is a member of ACI Committees 114, Research
eccentricity of axial force or to applied bending and Development, 117, Tolerances, 315, Detailing Reinforced
Concrete, and 443, Concrete Bridge Design.
moment. When the pile stands as a laterally un-
supported column, an incremental increase in
bending is produced by the magnification of ec-
centricity due to lateral deflection of the pile.
Fig. 2 illustrates the combinations of strain
across the pile section and the stress distribution
at ultimate load resulting from the causes men-
tioned above. Fig. 2 (a) illustrates the strain and
stress conditions in the concrete, while Fig. 2 (b)
deals with the strains and stresses in the pre-
stressing tendons.
The application of Eq. (1) and (2) and the re-
lationship between stress and strain illustrated in
Fig. 2 (a) and 2 (b) are in generalized form re-
lationships. The stress-strain curve for concrete in
Fig. 1 has been reduced to an equivalent rectan-
gular stress block for bending (see Fig. 3), as
recommended by ACI 318-63 and is the basis on
which the interaction diagrams in this paper were
constructed.
-
e.u.
S TIL\IN

Fig. I -Stress-strain diagram for concrete

It should be noted that the prestressing steel
(seven wire strands), tensioned to an effective
stress of 154 ksi, relaxes tension when the pile
sustains an applied axial compression force. How- where
ever, when the pile is subjected to bending, some
of the tendons ultimately may be subjected to P, ultimate axial load
an increase in tension, depending on the ratio of fc concrete stress at any point across the
bending to axial force. section
For a given cross section, the values of the ul- A 8; area of tendon i
timate resisting moment and ultimate axial load fsi stress in tendon i
determine the location of the neutral axis. If one M,.* ultimate resisting moment
arbitrarily assumes a position of the neutral axis, y distance from centroid to element dAc
as shown in Fig. 2 (a), only one value of utimate Yi distance from centroid to tendon i
load and ultimate moment will satisfy the equi- M"* is the magnified moment resisted by the
librium conditions: pile section at the point of failure. The external
n moment capacity can then be determined from
Pu = ffc dAc + :SAsds£
i=l
(1) the following:
1. In the case of a pile cut off at grade in
n stiff soil, the slenderness ratio h' /r is equal to zero
M,.* = ffc Y dAc +:SA,.; Y; fsi (2)
because the surrounding soil prevents lateral
i=l

ACI JOURNAL I AUGUST 1970 621

deflection or magnification of moment due to the ultimate applied moment Mu will be less than
eccentricity. Thus, for continuously laterally Mu * due to the lateral deflection amplification
supported piles with a zero slenderness ratio, the factor. Thus, the ratio Mu */M,. is dependent on
ultimate resisting moment equals the ultimate the ratio h' /r. Furthermore, the ratio Mu */M, de-
applied moment Mu * = M". pends on the values of the ultimate load P 11 , the
2. When the pile extends above grade or is effective height of the pile h', the section stiffness
driven into very soft soils, it serves as a laterally EI, and the distribution of bending moment along
unsupported column (or partially supported) with the pile, on which the following assumption was
a slenderness ratio greater than zero. In this case, used:

I ULTI~: STRAIN, I

TO EXTlilliiALJ +
I.LTTOTALULTIMATE
~TOCTRAIN DUE~ + ~TIIAIN
I;OTAL STRAINl
(a) EfFECTIVE
PRESTRESS AXIAL LOAD
DUE l8ENDINGJ
STRAIN = J
LAT UIJ'IMATii
STIIE:SSl
J

I
HFECTIVE
PRiSTI!ESS I
H
~- r:ESTIIE55 STRAIN
CONCRE:TE)

Asi=x ',
~·~---j;;~~
I+ -+-l
---+---1
_:____+

\-+- +/
~~---

(b) ~ FFECTIVEJ
PRESTRESS
STRAIN
+ ~STAAl
AXIALN LOAD,
DUE TO+:
KNDIN6 MOMENT
J - tTOTAL
-
AT ULTISTilAIJ
M.CI'E ~ [ AT ULTIMATEJ
TOTAL STRESSl

+MAGNIFICATION

Fig. 2 - (a} Concrete strain and stresses in pile section and (b) steel strains and
stresses in pile tendons

lu•O.OO! r;. o.asf~

© I--@ u
~ u

Fig. 3 - Strain distribution and

assumed stress block in pile
1-L-
section

N.J. N.J..
-

fo RECTANGULAR STRESS BLOCk:

STRAI"-1 DISTRIBUTI0"-1

622 ACI JOURNAL I AUGUST, 1970

 The unsupported pile, with an ultimate load Pu ~P(e+AMAX)
and eccentricity e (see Fig. 4), is at the threshold
of failure due to a combination of axial load and
+
bending moment P u ( e .6.maz) • .6.max is the total \~I
r--1\.
lateral deflection caused by the applied moment
~
and the incremental deflection caused by the
magnification factor.
From the classical theory of elastic buckling: 2
M,* = Mu sec\}Puh'2 /4EI (3) ~MAX

in which the flexural stiffness EI of the member

is assumed to have a constant value along the full MAG.
height of the member.
In the case of a prestressed concrete pile, the
column behavior is nonelastic, and an elaborate
step-by-step integration procedure would be re-
__.J/
quired, in which the actual properties of each
cross section along the pile would be considered in
determining the final deflected shape. Thus, to
calculate the ultimate resisting moment Mu * at DE!='LECTI ON MOMENT
the point of failure, one would have to take into
Fig. 4 - Deflections and bending moments in laterally
account the varying values of EI depending on the unsupported pile
conditions of stress and strain along the pile.
To reduce the labor of computing Mu*, a sim- limit rr./2, the maximum permissable slenderness
plifying but conservative assumption was adopted. ratio is reached, which occurs when the bending
The minimum value of EI prevailing at the sec- moment in the pile is zero, and the ultimate load
tion of failure was assumed as a constant EI for P, has .zero eccentricity. No physical explanation
the total effective height h'. can be given when the term \}Puh'2 /4EI is greater
Thus, both the elastic modulus of the concrete than rr./2.
and the moment of inertia of the pile section are
subject to reductions, depending on the strain APPLICATION OF THE ULTIMATE STRENGTH
level in the compressed zone of the pile section THEORY
and also depending on whether or not the pile
section is cracked. In constructing the interaction diagrams a suf-
Referring to Eq. 3, which is the secant formula ficient number of neutral axis positions were se-
for long columns, the basis for the selection of lected to obtain values of Pu and Mu* ranging from
the appropriate value of EI is derived as follows: the case of concentric axial load to the case of
pure flexure. For each neutral axis position
Curvature at failure:
chosen, values of the slenderness ratio h' /r varied
Eu/C = Mu*/EI from zero to the maximum, where the upper limit
Therefore: is reached when the value of \}Puh'2 /4EI is equal
to rr./2. From a large number of computations, the
EI = cMu*/Eu (4)
three-dimensional interaction diagrams were ob-
where c is the distance from the neutral axis to tained.
the extreme fiber under maximum compression The theory described above is applicable to
and Eu is the ultimate strain in the maximum structural members having one axis of symmetry,
compression fiber. designed to sustain combinations of axial com-
Rewriting Eq. (3), the ultimate applied moment pression and bending. The members can be com-
is: posed of more than one material, each having
arbitrary stress-strain characteristics.
Mu = Mu*/secyP,.h'2 /4EI (5)
In preparing the interaction diagrams for pre-
For a given position of the neutral axis, unique stressed concrete piles, the computational pro-
values of the ultimate load P,. and the ultimate cedure is ideally suited to computer solution. The
resisting moment Mu * will be satisfied by Eq. (1) computer program, in Fortran IV language is
and (2). The value of the ultimate applied mo- given in the Appendix.
ment then depends on the value of the term The interaction diagrams were constructed us-
\}Puh'2 /4EI in Eq. (5), whose numerical value ing a computer and a digital incremental plotter,
must be between the limits zero and rr./2. At the and are shown in Fig. 5 through 10.

ACI JOURNAL I AUGUST 1970 623

 0 10 to 30

~~o ~~o
f~ 6ooopsi f~ GOOD psi

.~~~--~-~c:J--~
( 1000 psi

...J

~ 8+---~-+--~~+-----T+--~~+------1
~1'1
u.J
t:c
~g+-~~-~\-----+-~---++---~~-----1
!:i"'
;:::)

~o
10

~o 0
0
<D

f~ sooo psi ·( aooo psi

g
Ill

0
0

"''
0
~

g
N

0
60 0 20 .co 60 100

ULTIMATE MOMENT, K~FT.

Fig. 5 - Interaction diagrams for I0 and 12 in. square piles

624 ACI JOURNAL I AUGUST, 1970

 ULTIMATE MOI'1ENT, IHI 7'5 100 125 \50
100 0 25 50
0 20 40 60 80
14

~0
0. . ~o
fc 6000 psl t~ E~OOO psi

~~~~--T0.
~0,
fc 7000 psl
If)
0..

~0
0 fc 8000 psi

0
~ +-'.,.._--+---t---'"'<'-1

so 100 0 25 50 75 150
ULTIMATE MOMENT, K-FT.

Fig. 6 -Interaction diagrams for 14 in. octagon and 14 in. square piles

ACI JOURNAL I AUGUST 1970 625

 ll)
a..
~0
0
~~~--~~~~~~

cl~~----~~~~----~~--~~----+-----4
~~. 0
Tc 7ooo psl

<(
0
_J

x~~~-----~~--~t-~~----~~~~tt-----~
.,.+
<(
LLJ

!<o
~~-l--~~~~~~----~~----t+---iir----~
~
:;::,

~~~5:==~:t~~~~§§~~~~liO~~.J,0 '~~~~~0~~~8[0~~~~§§~~~~~l_~2~~0
ULTIMATE MOMENT, K-FT.

Fig. 7 - Interaction diagrams for 161f2 and 18 in. octagon piles

626 ACI JOURNAL I AUGUST, 1970

 300 31>0 7
ULTIMATE MOMENT, K-Ft

Fig. 8 - Interaction diagrams for 20 in. solid and 24 in. hollow octagon piles

ACI JOURNAL I AUGUST 1970

827
 0 fc. 1000 psi

0
2+----ir---~--~-+--~~--~~--~

.~~~~Loo~~z~oo~~~3jo§o§§§§4~oo~~~5olo-----6oo o·oMr=~~zoto~~4joEo~§§6~oo§§~~s~oo~~~~o~olo__J_jzoo
ULTIMATE MOMENT, K-f"T.

Fig. 9 -Interaction diagrams for 24 in. solid octagon and 36 in. outside diameter, 26 in. inside diameter hollow
cylinder piles

ACI JOURNAL/ AUGUST, 1970

 &~~~~~-t~---r--
0
N

r:J==e~$~. O·t:3~~~~
3~0
700 1750 2100 0 450 900 1350 1800
ULTIMATE MOMENT, K-FT.
Fig. 10 -Interaction diagrams for 48 in. outside diameter, 38 in. inside diameter and 54 in. outside diameter, 44
in. inside diameter hollow cylinder piles
ACI JOURNAL I AUGUST 1970
 DISCUSSION OF THE COMPUTER PROGRAM pendent on the end conditions. End conditions in-
The basic purpose of the program is to perform clude the depth below grade to the assumed point
numerically the integration of Eq. (1) and (2), to of fixity and the condition at the pile head. The
obtain values of P" and Mu * for each assumed relationship between unsupported length land ef-
position of the neutral axis, and then to solve Eq. fective height h' is given for three cases in Fig. 11,
(5) to obtain values of M,. for each assumed value where the pile head is assumed to be laterally
of the slenderness ratio h' jr. supported. If the pile head is free to move lateral-
As mentioned above, the properties for con- ly, the member shown in Fig. 11 (a) becomes un-
crete (stress-strain curves) were based on the stable; in Fig. 11 (b) h' becomes twice l, and in
provisions of ACI 318-63, Section 1503 (g). A Fig. 11 (c) equal to l.
rectangular stress block was used, and at ultimate For a given effective height h', radius of gyra-
resistance, a concrete stress intensity of 0.85 fc' tion r, applied ultimate load Pu, and applied ulti-
was assumed uniformly distributed over the rec- mate moment Mu, a prestressed concrete pile may
tangular block shown in Fig. 4, as recommended be selected from the charts. The tentative pile sec-
by ACI 318-63. The maximum strain at the tion and concrete strength can be determined,
extreme compression fiber at ultimate resistance once the load factors for design have been de-
was assumed equal to 0.003. cided.
The strength reduction factor cp recommended The design of prestressed concrete piles serving
by ACI 318-63 was also included in the computa- as unsupported columns should include the ef-
tions. A factor cp = 0. 7 was used for all axial loads fect of long-term behavior of the concrete under
P,. in exces~ of 0.1 fc' A 0 • For Pu values of zero sustained load. Of special interest is the influence
(pure flexure), a value cp = 0.9 was used, and cp of creep-deflection on the stability of long, slender
was assumed to decrease linearly to 0.7 when P,. concrete columns.
reached 0.1 fc' Ac. ACI 318-63 contains a provision under Section
The stress-strain curves for steel were based on 916 (d) which takes into account the long-term
typical diagrams for 1h in. diameter, 270 ksi lateral deflection of reinforced concrete columns
stress-relieved seven-wire prestressing strands. due to creep. ACI 318-63 provides that the analysis
for creep-deflection be based on a reduced elastic
APPLICATION OF THE INTERACTION modulus for concrete, where the value of Ec for
DIAGRAMS TO STRUCTURAL ENGINEERING the long-term sustained load deflection shall not
be more than one-third the value of:
In the case of foundation piles in stiff soils, cut
off at grade and connected to footings, the slen- Ec = 33 (w)l-5 (fc')0.5
derness ratio h' jr is zero. The selection of an ef-
fective height h' and an appropriate load factor is
a matter of judgment and experience. The soil EXAMPLES OF PRESTRESSED CONCRETE PILE
conditions and the ratio of dead to live load will DESIGN USING THE INTERACTION
obviously influence the load factor selected. DIAGRAMS
When the piles are driven into soft soils or ex- Building foundation piles
tend above grade, the effective length h' is de- The piles under the building are cut off at the foot-
ing line. The concrete footings are cast-in-place around
the pile heads. The full length of the pile is surrounded
by stiff soil. Select a pile for an ultimate load of 360
kips, assuming an eccentricity of 0.1D. Assume h'/r =0.
By inspection of the interaction diagrams, one finds
two economical pile sections meeting the criteria above:
1. 10 in. square pile, fc' = 8000 psi
2. 12 in. square pile, fc' = 6000 psi
Deep foundation piles, installed by heavy equip-
ment
The piles are cut off at grade and encased in massive
concrete footings. An ultimate load of 900 kips per pile
is required, with an assumed eccentricity of 0.05D and
assumed slenderness ratio, h'/r of zero.
By inspection of the interaction diagrams, one finds
(a) (b) (c) three choices or prestressed concrete piles available:

h=i h·0.71 h· 0.51. 1. 161h in. octagon pile, fc' = sooo psi
.1 f , - 7000 psi
2.. 18 in. octag<m p~ e, fc • -:=, 6000 psi
r\~, \\ - \~\\~,"~' Q' '"~ ~~"~"\Q\\\ ()\\ ,,,,t\\-., ne\qn\ 3 2.0 in. octagon plle, c -
. ~C\ )QURtt~ll ~UGUSl, \910
.\ a\ \1\\~t~\\~ Ut\\U~~ot\ed ?\\e
Marine wharf facility Pile columns for long-span bridge
The wharf is to be constructed with a laterally sup- The bridge is to be supported on piles with an et·
ported reinforced concrete deck, vertically supported fective height h' of 72 ft. The piles are required to
on piles with a maximum free standing length of 60 ft carry a dead load of 1200 kips and a live load of 80
above the ground. The deck is very stiff when com- kips. Load factors for design are 1.5 for dead load and
pared to the piles. Each pile must carry 50 kips dead 2.5 for live load. Assume an initial eccentricity of 0.1D
load and 130 kips live load. Based on the soil data, the and make provisions for lateral creep deflection to
piles are assumed fixed 10 ft below the ground line. The determine effective eccentricity.
pile heads are assumed fixed into the deck. Load fac- For trial design, the initial ultimate load is:
tors for design are 1.5 for dead load and 1.8 for live
load.
Pu = 1.5DL + 2.5LL
The prestressed concrete pile section is determined (1.5) (1,200) + 2.5 (80)
by trial solution. 1st trial: 16% in. octagon pile. The 2000 kips
effective height is:
The ratio of dead to total load is 1200/1280 = 0.94.
h' = 0.5 (60 + 10) (12) = 420 in. Because of the very high ratio of permanent to total
The radius of gyration r = 4.2 in. with h'/r 100. design load, assume the eetfective is 50 percent greater
Assume an initial eccentricity = O.lD = 1.65 in. The than the einitial due to lateral creep deflection.
From inspection of the interaction diagrams let us try
required ultimate load is:
a 54 in. outside diameter, 44 in. inside diameter hollow
Pu = (1.5) (50) + 1.8 (130) = 309 kips cylinder pile, 7000 psi concrete. The slenderness ratio
h'/r for this pile is (72) (12) /17.4 = 49.8, which ap-
The required ultimate moment is: pears reasonable for the first trial. Then the trial ulti-
mate moment is:
Mu = 309 X 1. 65
12
= 42.5 ft-kips
Mu = (1.5) CPu) (0.1) (D)
The lateral deflection from dead load creep will ( 1.5) (2000) ( 5.4) /12
cause an increase in eccentricity. For the first trial, an
estimate of the magnification of eccentricity due to 1350 ft-kips
creep deflection caused by sustained moment is made. We next investigate the creep deflection due to a
The creep magnification can increase substantially sustained dead load, Pd of 1200 kips using one-third the
the effective eccentricity depending on the intensity of elastic modulus for 7000 psi concrete ( 1825 ksi),
sustained moment on the pile. For a slender pile with a
h' = 864 in., and the moment of inertia I = 233,400
ratio of dead load to total load of 50/180 = 0.278 let us in.4 Using the secant formula:
assume a slenderness ratio of 100, and a creep magni-
fication factor of 1.25, or an effective e = (1.25) (O.lD). eetfective = einitial sec yPdh'2!4EI
Then the trial magnified ultimate moment is (1.25)
(42.5) = 53.1 ft-kips. From the interaction diagrams, The value under the radical term is:
one finds that a 16% in. octagon pile of 7000 psi con- v'C1200) (864)2/(4) (1825) (233,400) = 0.725 radians
crete with a slenderness ratio of 100 is capable of (or 41.5 deg)
sustaining an ultimate load of 309 kips with an ulti-
mate moment of 53.1 ft-kips. The value of the secant is 1.34 and the eetfective = 1.34
Having tentatively selected this pile, the creep-de- The design requirements are then compared with
e;nitial.

flection under the designed sustained load moment is the capability of the 54 in. outside diameter, 44 in. in-
next calculated. The effective elastic modulus under side diameter hollow cylinder pile with 7000 psi con-
sustained load is assumed for 7000 psi concrete at one- crete:
third the ACI 318-63 value, namely, 1825 ksi. From
Table 1, the moment of inertia of a 16% in. octagon is
Required Available
4080 in.4
From the secant formula, the creep magnification is Pu 2000 kips Pu 2000 kips
calculated, and: Mu 1205 ft-kips Mu = 1320 ft-kips

Overhead bridge craneway

eeffective = einitial sec v'Pdh'2/4EI An overhead bridge craneway is supported on free-
standing piles, carrying a dead load of 20 kips and a
where Pd is the sustained dead load (50 kips) and the live load of 100 kips, with an eccentricity of O.lD. The
value under the radical term is: effective height of the pile is 24 ft.
For this structure the load factors will be assumed
v'C50) (420)2/(4) (1824) (4080) = 0.552 radians 1.25 for dead load and 2.0 for live load. For the fi-rst
(or 31.6 deg)
trial, with the ratio of dead load to total load amount-
The value of secant 31.6 deg is 1.18 which is less than ing to 20/120, or 0.167, let us assume the ratio, effective
1.25 (our assumed value above for creep magnification) to initial eccentricity is 1.2. The ultimate load is
and Mu required = (1.18) (42.5) = 50.2 ft-kips. 1.25 (20) + (2) (100) or 225 kips. From inspection of
The design -requirements are compared with the the interaction diagrams, a 12 in. square pile, with
capability of the 16% in. octagon 7000 psi concrete pile. h'/r = (24) (12)/3.5 = 82 and fc' = 7000 psi, appears
satisfactory for a trial design, with Pu = 225 kips and:
Required Available Mu = (225) (1/12) (1.2) = 22.5 ft-kips
Pu 309 kips Pu = 309 kips The magnification of eccentricity due to creep under
Mu = 50.2 ft-kips Mu = 56 ft-kips dead load, Pd = 20 kips, with h' = 288 in.,

ACI JOURNAL I AUGUST 1970 631

E =
lfa Ec = 1,825 ksi, I = 1,728 in.4 is next investi- Two curves from Reference 3 for 14 in. octagon
gated: piles of 7600 and 8400 psi concrete are reproduced
eeffective = einitia! sec y'Pdh'2/4EI in Fig. 12, which are based on a .p factor of 1.0.
The ultimate loads for three test piles with their
The value under the radical term is:
concrete compression strengths are plotted at
y(20) (288)/(4) (1825) (1728) = 0.363 radians Points 1, 2, and 3. For comparison with the
(or 20.8 deg) authors' interaction diagram, a curve for the 8000
The value of the secant is 1.07, and eeffective 1.07 = psi concrete 14 in. octagon pile is drawn, based on
The design requirements are then compared with
e;nitial. a .p factor of 0.7. It can be seen that good agree-
the capacity of the 12-in. square pile, with fc' 7000 = ment between Hromadik's tests and the authors'
psi. curve is reached when the latter is multiplied by
Required Available
1/.p.
It is evident that Hromadik's method would be
Pu = 225 kips Pu = 250 kips less conservative than the authors' generalized in-
Mu = 24.1 ft-kips Mu = 35 ft-kips teraction diagrams, which are based on the actual
moment-curvature relationship at the ultimate
COMPARISON OF THE INTE.RACTION load condition rather than a determination based
DIACRAMS WITH RESULTS OF PREVIOUS on the assumption that the full section moment of
WORK
inertia and the tangent modulus apply for the
The most important previous work on ultimate flexural stiffness at the critical load.
strength of slender prestressed concrete piles was In engineering practice, the authors feel that
published by Hromadik. 3 He load-tested three pre- long, slender prestressed concrete piles rarely, if
stressed piles and developed a method of analysis ever, can be free of eccentricity, and some bending
for elastic buckling of slender piles under con- must be anticipated regardless of painstaking ef-
centric axial load, when no lateral deflection of forts to achieve true concentricity of loads.
the pile was assumed until the load reached
Euler's critical value:
TABLE 2- COMPARISON OF ZIA AND MOREA-
Per = :rtz EI/h'2 DITH'S RESULTS WITH AUTHORS' RESULTS
Zia and Moreadith Authors
The method did, however, take into account the
Pu 170 kips 135 kps
nonlinear compression stress-strain relationship of .p 1.0 0.7
concrete as published by Hognestad. 1 Puf.P 170 kips 193 kips

1400 ...----.--~---,---,.---,1-~r---r---, More recently, work published by Zia and

---~-- I8400 Ipsi, Moreadith4 included a numerical example of the
', .rf~ = lf·I.O ultimate strength of a 10 in. square prestressed
1200
~----
I '>, {

..(>1·'FROM 1-!ROMAD!.W concrete column, made from 6500 psi concrete,

with an eccentricity of 0.3D and a slenderness
f, =7600 psi. ll\ ' ratio of 43.
1000 f-cj)• 1.0------- ~,,.---\1---+---+--t---1----l
The results of Zia and Moreadith's method are
~ \\ given in Table 2 and compared directly with the
:i2 \
~800 \\ results obtained from the authors' interaction dia-
0 ~ ,,
gram.
~ r---~ \' In its tentative recommendations for prestressed
1.11 600 ' ~ 8080 psi. (2) - concrete piles, the PCI committee on prestressed
~- .f,, .. 8000 ps1.• """ 1\. '~~~·
• ~' 350 psi.. ((I)
3)
concrete columns" included the following provi-
~ EiE:WERALIZE:D INTkRACTION I A. l 8520 psL sions for the design of short and intermediate
::J 400 1-(e•O, ~-0.7)-;--------- ' \ ~,'-----01 length columns:

200+--~--+--+-~--~--~r--~~ "'~~~~.. The maximum allowable axial load P on short

columns (slenderness ratios from zero to 30), and

~
with closely spaced spirals is:

0~-~-~--+--+--+--4-~f--~
P = 0.34fc'Au + f.As'- 0.40 F. (6)
0 25 50 75 100 125 150 175 200 (0.44 + 0.000045 fee)
SLENOEii!NE!:»S RATIO, hjr The maximum allowable concentric axial load
on short columns reinforced with separate lateral
Fig. 12- Ultimate concentric load capacity of 14 in.
ties or light spiral reinforcement shall be 85 per-
octagon piles: comparison of Hrornadik's method with
authors' generalized interaction diagram cent of that given by Eq. (6), or:

632 ACI JOURNAL I AUGUST, 1970

 TABLE 3 -ALLOWABLE LOADS IN KIPS FOR 14 IN. OCTAGON PILES OF
8000 PSI CONCRETE
e=O e = 0.1D e = 0.2D
h'/r
PCI Authors PCI Authors PCI Authors
0 348 364 250 281 195 205
30 348 354 250 267 195 192
50 303 345 217 248 170 173
75 242 315 174 206 135 134
100 181 257 130 143 101 96

P = 0.85 [0.34 f c' Au + f .As' - 0.40 Fe

(7)
linear stress-strain relationship of concrete, and
(0.44 + 0.000045 fcc)] the elastic stability of the pile as a slender column
is recognized.
Short columns subjected to combined axial load The mathematical solution is given in Fortran
and bending, where the eccentricity e is less than IV language for computer analysis, enabling the
e,the maximum allowable load N shall be given reader to construct interaction diagrams for pre-
by: stressed concrete columns stressed to various pre-
stress levels and with various concrete
strengths (see Tables A1 and A2 in the Appen-
(8) dix) . The program can be modified to accommo-
date analyses of arbitrary prestressed concrete
sections having one axis of symmetry, and rein-
where
forced with stressed or unstressed reinforcement,

P = 0.17fc'Au ( 1 - 95 i:~ ): ,. 0.08fc' Au (9)

all materials having arbitrary stress-strain charac-
teristics.
The results obtained from the interaction dia-
grams are in reasonable agreement with results
e- = o.23D (1 + 2oo j:~ ) L o.5D oo> of previously published work.

For intermediate length columns (slenderness REFERENCES

ratios between 30 and 100), the load and moment 1. Hognestad, E., "A Study of Combined Bending
computed from the analysis shall be reduced by and Axial Load in Reinforced Concrete Members," Bul-
the factor R, where: letin No. 399, Engineering Experiment Station, Univer-
sity of Illinois, Urbana, Nov. 1951, 128 pp.
R = 1.22 - 0.007 (h' /r) (11) 2. Timoshenko, S. P., and Gere, J. M., Theory of
Elastic Stability, McGraw-Hill Book Company, New
but not greater than 1.0. York, 2nd Edition, 1961, 541 pp.
To compare the results of the PCI and the 3. Hromadik, J. J., "Column Strength of Long Piles,"
authors' method, the allowable capacity of a 14 in. ACI JouRNAL, Proceedings V. 59, No. 6, June 1962, pp.
octagon, 8000 psi prestressed concrete pile is given 757-778.
4. Zia, P., and Moreadith, F. L., "Ultimate Load
in Table 3. To make this comparison, the load fac- Capacity of Prestressed Concrete Columns," ACI JoUR-
tor applicable to the authors' ultimate strength in- NAL, Proceedings V. 63, No. 7, July 1966, pp. 767-788.
teraction diagram values has been derived. The 5. PCI Committee on Prestressed Concrete Columns,
PCI method is based on a load factor of 2.5, with "Tentative Recommendations for the Design of Pre-
a 15 percent strength reduction factor included in stressed Concrete Columns," Journal, Prestressed Con-
crete Institute, V. 13, No. 5, Oct. 1968, pp. 12-21.
Eq. (7). On the other hand, the authors' curves
include a </> factor of 0.7. Thus to equate the two
CONVERSION FACTORS
methods the load factor applicable to the authors'
curves is (2.5) (0.7) /0.85, or 2.06. Multiply By To obtain
in. 25.4 mm
CONCLUSIONS ft 30.48 em
sq in. 6.4516 cm2
The interaction diagrams are a convenient de- cu ft 0.0283168 ms
sign aid for the engineer. The graphical presenta- lb (avdp) 0.45359237 kg
tion of ultimate load and moment capacity of kips 453.59237 kg
psi 0.070307 kgf/cm2
prestressed concrete piles as a function of concrete ksi 70.307 kgf/cm2
strength and slenderness ratio enable the designer ft-kips 138.255 kg-m
to rapidly identify an economical pile section.
Note: The expression for Ec (elastic modulus of con-
The generalized interaction diagram is based on crete) measured as a function of density and compres-
a rigorous mathematical concept in which the non- sion strength of the concrete cannot be directly trans-
lated from British to SI Units.
AC/ JOURNAL I AUGUST 1970
633
 APPENDIX
NOTATION

Ac net area of concrete section Eu ultimate strain in the maximum compression

Ag gross area of column fiber
As area of longitudinal prestressing steel ¢ capacity reduction factor
c distance from neutral axis 'to extreme fiber in fee effective prestress in the concrete
flexure fc' compressive cylinder strength of concrete
D diameter of width or pile section fs allowable stress in longitudinal nonprestressed
E modulus of elasticity reinforcement at 40 percent of the yield
e eccentricity of resultant load on a pile strength but not to exceed 30,000 psi
e maximum permissable eccentricity of P fs' ultimate strength of prestressed reinforcement
fc compression stress in the concrete Fe effective axial prestress force
p allowable axial load on a short column h' effective height of column
p maximum allowable axial load acting in com- I moment of inertia
bination with the maximum allowable moment Mu ultimate applied moment
Mu* ultimate resisting moment at point of failure
M
Pu ultimate applied axial load to pile or column M maximum allowable bending moment acting in
R a capacity reduction factor for intermediate combination with the maximum allowable axial
length columns load = Pe
r radius of gyration of gross concrete section N allowable eccentric load normal to the cross
w weight of concrete in lb per cu ft sections of the column acting with eccentricity
y distance from section centroid to flexure fiber e
in question
lateral deflection of pile or column This paper was received by the Institute Nov. 3, 1969.

TABLE AI -INPUT DATA FOR COMPUTER SOLUTION

Card Subject
1 Number of interaction diagrams needed Nl
The following set of five cards is repeated Nl times one set after the other. Each
set contains information for each section and material properties:
2 1-36 6A6 Headline # 1 to appear on the top of the interaction
diagram (usually to give size of pile)
3 1-36 6A6 Headline #2 to appear on the top of the interaction
diagram usually to indicate concrete strength
4 1-10 FlO.O Concrete strength in ksi positive for compression (4
ksi or greater)
5 1-10 F10.0 Outside diameter of pile, in.
5 11-20 FlO.O Inside diameter of pile, in .. , if hollow
5 21-30 F10.0 Radius of circle of prestressing steel, in.
5 31-40 F10.0 Number of 270 kip prestressing strands
5 41-50 F10.0 Cross-sectional area of each strand, sq in.
5 51-60 F10.0 "X" cross section identification number where:
X
X
=
= - 1. for square section
0. for octagon section
X = 1. for circular section
Maximum ultimate moment expected, ft-kips, for mo-
6 1-10 F10.0
ment scale, should be even multiple of 7
6 11-20 FlO.O Maximum ultimate axial load expected in kips for axial
load scale, should be even multiple of B

ACI JOURNAL I AUGUST, 1970

634
 >
n
-
Q
c:
:u TABLE A2- COMPUTER PROGRAM
z:

->
r-
>
c:
In
c:
en
COMM~ NAMEII61oNAMf2 I 61
COMMON XXA! l0ol501 oYYAI lOolSOI
COMMON XA!l321oYAIInl
DIMENSION XYMI 101
API = 3ol4159
AP2 = API*API
FFC
XP
XM
= FCifY,C,FCU,DD11
= XP
XM + A*FFC*Y
=
AA.A = AAA + A
All = A*Y•Y +Ali
300 CONTINUE
104 CONTINUE
00 400 N=ltMM
+ A*FFC
AML
AML
::: XMM
= ASS I AMLI
XXACJ,KJ = AML
YYAfJtKl = APL
6 00 CONTINUE
WRITEC6tl51 fXYMfNitN=lt91
WRITEf6tl61HH
FCR = EX PLI AAA
DO 10 11=1o9
DO 20 JJ=t.l30
XAIJJJ = XXA{Jt,JJI
YAIJJJ = YYAIIIoJJI
20 CONTINUE
CALL LINE IXA.,YAtl30t!tOtQJ
10 CONTINUE

-
CALL PLOTS RETURN
-f XN = N Al = FCR/FCU
READ! 5o! IN! END
DO 1000 !=loN1 Y = RRS•COSITHTA*IXN- lei I Al = 4.*Al
CD READC5•21 fNAMEIU:.J,K=l•6•
£1 = CC - ,5•001 + Yl*EU/C Al = 4. - Al
.......
Q READI5o21 IHAME21KioK=1•6l
Ell = El + .0055 Al = ABSIAll SIBFTC FCI.
FSI = FSULTIE!I I Al = SQRTIAll Fl:JNCTJON FCtCY,CtFCtJtDOll
READI5o3JFCJ
READ ( 5t4 10Dlt002 tRRS tZMMtASS, TYPE
FS = FSJ - 154. A2 = f2e- A})/7. YC • Y - DDl/2• + C
READ l 5, 5) xx;.t,xxP XM = XM + ASS*FS*Y A2=le-A2 AK = e85 + e05*1FCU + 4el
THTA = 2o*API /ZOlfo\
XPR = XPR + ASS*FSJ ETG = EEC*A2 CR = C - C*AK
MM = ZMM
RECS = 28000,/FEC IF !EXPL + fXPLI 112tllltl12 IF IYC- CR110,20,20
FCU = -FCJ AAA = AAA + ASS*IRECS - lei 112 AHST = AP2*ETG*AAA/EXPL 10 FC I = 0 •
FCJ = lOOO.•FCJ All =All+ ASS*<RECS- lei*Y*Y AHST = ABSI AHST I GO TO 30
400 COI'HINUE HH = SORTIAHSTI 20 FCI = eB5*FCU
WRITE16olll
IF 1002 + DD21 lOS,J05ol06 WRITE'I6t171HH 30 RETURN
WRITEI6ol21 INAI'fll~loK=lo61 1 OS GO TO 110
WRITE16oi21 !NA'IE21KloK•lo6l 111 WRITEI6o181 END
106 DO SOO N=l•400
EEC = 33a*fl58a**la51 XN = N
2000 CONTINUE
EEC • EEC*SCRT!FCJI/1000, CALL SAAn£XXMtXXPJ
EU • -,003 Y = .50125*002 - .0025*DD2*XN CALL PLOTI12.0,0.,-3J SIBFTC FSULTo
WRITE!6ol31 RR = .s•noz 1000 CONTINUE FUNCTION FSULTIESI
WRITEH,l91 BB = 2•*SORTIRR*RR - Y*YI CALL PLOTIO.,Q.,9991 DIMENSION FSI151•EPSI151
DO 2000 K=lol30 A = e0025*002*AA 1 FORMAT 1141 FSili=O.O
XIC. : K FFC = FCJfY,(,FCUtODll 2 FORMAT 16A61 FSI21=180.0
XC t leSl - aOl*XK XP = XP - A*FFC 3 FORMAT 1Fl0o21 FSI31=190o0
XM : XM - A*FFC*Y 4 FORMAT i6Fl0.4) FSI4l=200e0
C = XC*DDI
XP • O, AAA = AAA - A 5 FORMAT 12Fl0o?l FSIS1=210.0
XM = o. All= Ali- A•Y•Y 11 FOR"1AT 11Hltl/////////////////////t FSI61=220o0
500 CONTINUE
XPR • o, ISH I FSI71=230.0
110 CONTINUE 12 FORMAT 13~Xt6A6t//l FSI81=240.0
AAA • Oo
All = O, EXPl = XP + XPR 13 FORMAT 11Hltl///l FS191=24S.O
Ell = C>XM/EU 14 FORMAT !/t10X,F15.7tll FSI10i=2SOo0
IF !TYPE +TYPEI 10lol02ol03
101 DO 100 N=loiOO IF IEXPL + EXPLI 90J,901o902 15 FORMAT 130X,4£EI5.7t5Xlt/t30X, FSI11J=255.~

XN N = 901 GO TO 903
902 EXPL = 0.
l4CE15.7t5Xltlt~OX,fl5e71
16 FORMAT l/t32X,Fl0e2t
FS112i=260o5
FSI131=263o0
A = aOl•DDl*DDl
903 CONTINUE 129H PLASTIC BUCKLING H/R FSI 14J=264e0
Y = .505*001 - .01 *DD!*XN
FFC = FCifYtCtFCU,ODl) XPM = el*FCU*AAA 17 FORMAT l/t32XtF10.z, FSI1Si=270o0
XP = XP + A*FFC IF IEXPL - XPMI 401o402o402 129H TANGENT MODULUS CRITICAL H/RI E"PS!li=O,O
401 FJF = o7 18 FORMAT l/tlOXt62H----------- EPS!21=0,00657
XM = XM + A•FFC*Y
AAA =
AAA + A GO TO 403 1-------------,/) EPS13i=0,00700
402 FIF = e9 - e2*EXPL/XPM 19 FORMAT 110Xtl1HP-ULT. AND, EPSI4)=0e00743
Ail = A*Y*Y + All
l 00 CONTINUE 403 EXPL = FIF*EXPL 129HCORRESPO~DING M-ULT, FOR H/R• E"PSIS1=0,00786
GO TO 104 XM = FIF*X""/17• 215H= o. TO 200. I EPS16i=O,OOB35
102 XL = SQRTI2ol HH = Oe STOP EPSI71=0,00890
AA = XL*DDl/C2a*fXl+2.) I WRlTE16ol4lEXPL FND EPSI81=0,00968
DO 200 ~= 1 t40{l
RG = AI 1/AAA EPS!9)=0,01020
XN = N RG = SORTIRGJ EPSilOJ=O.OllO
Y = e50125•0Dl - .0025*00l*XN DO 600 J=l•9 SIBFTC SAAD. EPSI111=0.012S
XN = J SUBROUTINE SAAD!XXMtXXPl EPS112I=O.Ol50
YY = ABS!Yl
H = 25.•RG*IXN- lei COMMON NAME1161,NAME2161 EPSI131•0o0175
IF IVY- AAI 4lo41o42
41 BB = DOl BETA= EXPL*H*H/f4•*flll COMMON XXAI10t150),YYA!l0tl50J EPSI141=0.0200
GO TO 44 BETA = AASIAETAI COMMON XAII321,YAI1321 EP511SI=Oo0350
42 88 -=- 2.•tAA + •5*001 -YY)
BETA = SQRT!BETAI XO = 0, K=2
IF !BETA - 1oS7J 107o108o108 XAI1311 = ·o, IFIES15t6t6
44 A = a0025*D01*RR 107 F~ = 1./!COSIAETAII
FFC = FCI rv,c,Fcu,oot, XAI1321 = XXM/7. F~=-~c:;
X~M = XM/FM YAC13ll = Oe
XP = XP + A*FFC K=l
X~ = XM + A*FFC*Y
XYMfJ) = XMM YAI1321 = XXP/8. 6 IFCES-O.Q3501Bt7t7
HH = H/RG CALL AXISIXOt0etl5HULTIMATE MOMENT• 7 FSULT=270e0*(-l.OI**K
AAA = AAA + A GO TO JOQ 16H KeFT.,-21t7.,0.,
All = A*Y*Y + All 108 XMM = Oe
RFTUR~

200 CONTINUE 2XA{l'Hl,XA.f13?1t10.0J 8 DO 50 N=2.15

XYM(J) = XM~ CALL AXISCXO,O.,l5HULTJMATE AXIAL t IFI EPSI N 1-ES150,50 ,60
GO TO 104 AH = AP2*Eli/EXPL 19HLOAD KI~S,24t8et90.,
103 DO 300 N=lo400 50 CONTINUE
AH = ABS!AHI 2YAI13ll,yAfl32 I ,10.01 60 FSULT= ( -1.1 **K*IF.S IN l-1 FSIN l-FSI~-
XN = N AH = SORTIAHI XOl = 1. + XO 11 I I* I EPS IN I -ES I /I EPS IN I -EPS I N-1 I I I
Y = eS0125*001 - e002S•DDl*XN HH = AH/RG CALL SYMBOLCXOltBe5t0.14tNAME1t RFTUR~
R't = •11j*fl0l 109 CONTINUE 10et36J END
88 c. 2•*SORTIPR*RR - Y*YI APL = FXPL CALL SYMAOLCXQ1,A.r,o.14tNAME2,
~
A = .0025*001*88 APL = ABSIAPLI l0et36J
C4
Ul