Professional Documents
Culture Documents
for Proportioning
Rectangular Prestressed
Concrete Columns
John R. Salmons
Professor of Civil Engineering
University of Missouri-Columbia
Columbia, Missouri
Donald G. McLaughlin
Structural Engineer
Burns and McDonnel Engineering Co.
Columbia, Missouri
(Formerly Graduate Student
University of Missouri-Columbia)
120
the column. With an additional equa-
tion these charts can be used to obtain a
modified effective stress in prestressing
Synopsis
steel after losses. The generalized in- A series of nondimensional ulti-
teraction curves are presented in the mate load- moment interaction
Appendix as well as design examples curves for short, rectangular con-
illustrating how the interaction curves crete columns with axial prestress-
and the stress modification equations ing are presented. Several design
should be used. examples are included to show the
application of the design charts.
General nondimensional equa-
Governing Equations tions are given which are based on
nonlinear approximations for both
The derivation of equations gener- the concrete and prestressing steel
ated for a general prestressed concrete stress-strain curves, linear strains
column parallels the development of across the cross section, strain
those for load-moment interaction rela- compatibility and equilibrium. The
tionships for any specific column. For steel location is generalized by re-
the general case, the equations depend placement of individual strands with
upon the assuming of a neutral axis lo- a rectangular prestressed steel
cation and satisfying the compatibility tube.
of strains. The concrete compression The series of curves presented
force and steel forces are calculated are for a steel strength of 270 ksi
from the strain distribution. These (1861.7 Nlmm2), concrete strengths
forces are nondimensionalized and, of 5, 6, 7, and 8 ksi (34.48, 41.37,
when used in the appropriate equations 48.26, and 55.16 N/mm 2 ), and
of equilibrium, produce a single point geometric parameters correspond-
on the generalized load-moment curve.` ing to the more commonly used col-
The entire interaction curve is obtained umns. In addition, full and partial
by varying the neutral axis location stressing of the prestressing steel
within an appropriate range. were considered.
The concrete compression force is
calculated from a concrete stress-strain
relationship which accounts for a por-
tion of the nonlinear region of the
curve. This solution is more accurate due to prestressing, precompression of
over the entire range of the interaction the concrete and bending of the mem-
curve than one based on the rectangular ber. The force in the steel tube is cal-
stress block approximation. The general culated from these strains using an ac-
equations are designed so that any con- tual stress-strain relationship for pre-
crete stress-strain curve can easily be stressing strand.
substituted for the one used for the so- The stress stage in the steel is
hition presented. analyzed using three specific cases:
The steel forces are calculated in Case 1 is appropriate when all the steel
terms of a steel percentage rather than strains are in the linear range of the
for a particular number of strands. For stress-strain curve, Case 3 is used when
analysis, the prestressing strands have all steel strains are in the nonlinear
been replaced by a thin rectangular range, Case 2 is a combination of Cases
tube of prestressing steel with an 1 and 3 and deals with steel strains in
equivalent area of steel. The total strain both the linear and nonlinear portions
in the tube consists of the steel strain of the stress-strain curve. Other pre-
a)
m
U
0
U
Concrete Strain
l (
f=
E f z
ForO c E^ . E p : 0.85f^ ^2 ` – I `^
`E p1 1Ep
<
For €< E, -- E cu : tc = 0.85f'
P
stressing steel stress-strain relation- 8
ships could he used in the analysis pre- K fc b t = K, + K
sented here; however, the translonna-
tion would be more involved than that R R` + R,
necessary for the replacement of the = f bts
concrete stress-strain relationship. where
P„ = ultimate axial load acting on
NONDIMENSIONAL PARAMETERS column
The governing equations are used to M„ = ultimate moment acting on
generate nondimensional load versus column
moment interaction curves for pre- = concrete strength
stressed concrete columns. The evalu- b = width of cross section
ation of the equations is accomplished t = depth of cross section
by incrementing the depth of the com- K, = portion of K due to compres-
pressed concrete area (k xt) which is sive force in concrete
measured from the extreme fiber on the R c = portion of R die to compres-
compression face. The solution of each sive force in concrete
increment of k„t produces nondimen- K 8 = portion ofK due to steel forces
sional loads (K) and corresponding R s = portion ofR due to steel forces
nondimensional moments (R) which
represent a point on an interaction
CONCRETE COMPRESSION
curve. Each of the two nondimensional
PARAMETERS
terms can be considered in two parts
which correspond to the contribution of The parameters K C and R1 are depen-
the concrete and steel respectively. dent upon the compressive force of the
These nondimensional terms K and R concrete (C c) and the moment (M e) of
are defined as: this force about the bending axis of the
122
1 t/2
t/2
k„t
N.A.
B.A.
Cc
Yc
b b
Steel Tube
t
B.A gt
t
.9t
124
1.
!1Ii
gt /
t N.A.
B.A.
gt/2
CBI
b
N
N
m
u)
Steel Strain
Fig. 5. Stress- strain curve for 270 ksi (1861.7 Nlmrre ) prestressing steel.
lar section in which all the prestressing load versus moment interaction curves
steel stresses are within the nonlinear for a specific column. Points on an in-
portion of the stress-strain curve. teraction curve are calculated by as-
suming a neutral axis location, satisfy-
E vl > E NL E 9 4 > EYL
ing the compatibility of strains, calcu-
The nondimensional farce and mo- Iating the concrete compression force
ment due to steel (K„ R 1 ) can be cal- and steel forces, and solving the equa-
culated for this case with the following tions of equilibrium. For general use,
expressions: the concrete compression force and the
steel forces were nondimensionalized.
K pp (3 Al + f94 177)
+ 2 (13) With this one deviation a general de-
f'
4
sign aid was developed for rectangular
prestressed concrete columns.
R • gpP
8f I
J PI S4
126
gion. The amount the curve translates the depth of the section. This allows a
(AK) for a change in the effective pre- smooth reliable curve between the bal-
stressing level (Af, ) can be found by ance point and the load axis. However,
modifying the basic nondimensional at higher steel percentages and lower g
equation for the nondimensional load values, the concrete stress-strain rela-
K: tionship used in this derivation also
K =K, +K& shows a slight inconsistency. This is
Since all prestressing steel strains are caused by the inability of the stress-
in the linear portion of the stress-strain strain approximation to give accurate
curve, equations for modification to results for the concrete force and its lo-
K(AK) were developed in terms of the cation as the nonlinear portion of the
change in prestressing (Afte ), curve moves oil' the cross section. At
higher steel percentages and lower g
AK = P. Or .f e I +
Pp
(15) values, the concrete compression force
and its location is a major determinant
This equation can be solved for Af1e : of the load and moment capacity of the
column. The result of this inconsistency
.fc AK is a small discontinuity in the curves
4 }ae = (16)
Pv f ET
near the load axis.
L Er.
Prestressed Steel Tube
With Eqs. (15) and (16) and the inter- Approximation
action curves, a prestressed concrete
column can be designed with any level The replacement of the prestressing
of effective prestressing provided all strand with a prestressed steel tube is
the strands are stressed to the same the primary approximation in the deri-
level. Complete derivations of all vation of the equations used to generate
equations are available in a previous the load vs. moment interaction curves.
publications The accuracy of the solutions obtained
with this approximation was checked
by calculating points on the interaction
curves for specific columns. These spe-
Effects of Derivation citie columns consisted of various steel
Approximations percentages, numbers of strand, tub val-
ues, and values of g. The values ob-
Concrete Stress-Strain
tained were then nondimensionalized
and checked with the general interaction
Relationship
curves.
A nonlinear approximation for the For the case in which all the pre-
concrete stress-strain relationship is stressing steel strains are in the linear
used in order to produce more accurate portion of the stress-strain curve, the er-
results regardless of the location of the rors were in the range of 2 percent.
neutral axis. This capability is impor- However, certain factors determine the
tant since the development of an inter- magnitude of the error. For a given
action diagram requires the depth to steel percentage, the column with the
the neutral axis to vary from zero to in- largest number of strands had the least
finity. The details of the concrete error. Increasing the g value or the tlb
stress-strain approximation are shown value increased the error produced by
in Fig. 1. the prestressed steel tube, Increasing
The concrete stress-strain approxi- the steel percentage increased the
mation is especially useful as the depth magnitude of the error.
to the neutral axis becomes larger than For the case in which all the pre-
CONCLUDING REMARKS
The purpose of this study was to pro- the concrete and steel strengths and the
vide the engineer with a versatile de- ultimate load and moment acting on the
sign aid for rectangular prestressed column. From these parameters various
concrete columns without accounting column sizes, steel percentages and
for length effects. This was accom- effective prestressing levels can be
plished by the development of a gen- evaluated until a combination is found
eral formulation for a prestressed col- which satisfies the design require-
umn analysis and of nondimensional ments. This method offers efficiency
load versus moment interaction curves. and versatility in either preliminary or
The input information necessary in final design of rectangular prestressed
order to use these design aids is only concrete columns.
REFERENCES
1. Everard, N, J., and Cohen, E., Ultimate ter 1I, pp. 389-399.
Strength Design of Reinforced Concrete PCI Design Handbook -Precast Pre-
Columns, Interim Report of ACI Com- stressed Concrete, 2nd Edition, Pre-
mittee 340, ACI Publications SP-7. stressed Concrete Institute, Chicago, Il-
2. Notes on ACI 318-71 Building Code Re- linois, 1978, Chapter 8, p. 17.
quirements With Design Applications, McLaughlin, D. G., "Prestressed Con-
Portland Cement Association, 1972, crete Column Design," MS Thesis, De-
Chapter 10, p. 26. partment of Civil Engineering, Univer-
3. Nilson. A. H., Design of Prestressed Con- sity of Missouri-Columbia, December
crete, John Wiley and Sons, 1978, Chap- 1980.
128
APPENDIX A - DEVELOPMENT OF
DESIGN CHARTS
The following charts were formulated then be used as illustrated in the design
using the numbered equations in the examples.
previous presentation. Two variables f, The design curves for prestressed
and g are considered in various combi- concrete columns are similar to those
nations for rectangular columns with for regular reinforced columns in shape.
equal prestressing steel in each face. The prestressed concrete column
The steel percentage pp , is based on curves exhibit a less defined balanced
the gross cross section of the column. point but do have a concrete compres-
The following data were used in the sion control region and a tension con-
development of the design charts: trol region for each steel percentage
= OA03 in./in. (0.003 cm/cm) (see Fig. B2, Appendix B).
E, = 57.5 v' .) (psi) (ksi) An interesting feature of the charts
fp„ = 270 ksi (1861.7 Nlmm2 ) presented here is that the curves for
fde = 154.9 ksi (1068.0 N/mm2) prestressing steel percentages cross
E= 27,500 ksi (189.61 kN/mm2). each other. This occurs because in the
The charts were developed without compression control region, the higher
accounting for column length effects. the prestressing steel percentage, the
Also, the workmanship factor, 0, has not lower the column capacity. Hence, in
been applied to the curves. Therefore, the compression control region it is de-
the first step in design would be to ad- sirable to use less prestressing steel.
just the loads for slenderness effects by However, in the tension control region
using a moment magnifier approach the higher the percentage of prestress-
and applying the workmanship factor, ing steel, the higher the moment car-
. The appropriate design charts can rying capacity of the column.
0.5C 0.50
U
o2
04c 0.4C
NN
03C 0.3C
0 2( /2 0.2C //
0,! C 0.10
0.00 0.02 004 006 0.06 0.10 0.12 ^i4 0.00 0.02 0.04 0.06 0.06 0.10 0.12 D.ia
3'
P;Q fSe =154.9 ksi 9t pp fSe =154.9 ksi
00
^oa 9= 0.8 t Qc g= 0.9 t
0.70
P, oop S p^
O ^ 0 ^
0.60
00/
U 050
0.40
0.30
/// 020
_ 7 _ 7/__
7/ _
0. 0
W Ke/t=%e^f^bt2
Ke/t=pie/f'bt2
r^ -aPT yin CHART NJO. 6
o_sc 0.90
N
f^ = 6.Oksi ^• f^ = 6.0 ks ^
0.8c
fp =270.0 ksi L. 0.80
fpu =270.0 ksi b .
,^-ePv
U 0.5{ 0.50
QJ a7
If
Y 0.4C 0.4C
0.3c 0.3C
0.2C 0 2C
0-IC OHC
0.00 0.00
0.02 0,04 0.06 008 0.10 012 0.14 0.02 0.04 0.06 0.06 0.10 0.12 0.14
Ke/t=P"e/f^bt2 Ke/t=PUe^f^bt2
CHART NO. 7 CHART NO. 8
0.90
= 6.0ksi •
b u
fPv = 270.0 ksi fPu =270.0 ksi b I "
080 c^#
se =154.9 k i _ fse=154.9 ks.l
°•^a #
g= 0.8 9 = 0.9
e 0.7( e
o ^^^ PU 0040) Pu
qo^9 pad a9
0.61
0.S(
II
0.41
0.3'
02
_LL
0.1
0.00 _
0.02 0.04
7
0.06 0.06 0.10 012 0.14 0.00 0.02 0.04 0,06 0.06 OJU 0.12 0i4
W Ke/t=Pue^ffbt2
w Ke/t=Pue/f"bt2
r'I-1LQT Kin 0
0.91
0 9(
f^ = 7.0 ksi f'= 7.0ksi
fp U =270.0 ksi b fpu =270.0 ksi
0.81 Os(
fse=154.9 ksi 9t
^ao fse^ I54.9 ksi 9t t 5'
_____
Os
= 0.s ° 9 = 0.7 t
07' e 0.71
Pu
Q o 0o a ^
0.61
0.51
0.5(
N
a?
0.4(
0.4c
4
0.3(
03C
0 2(
02(
O,IC 0.€(
0.00 0.02 0.04 0.06 O.OB 010 0.12 0.14 0.0p 002 0.04 006 0.08 0.10 012 0.14
g= 0.8 # 9 = 0.9 t
e 070
a e
Pu
Q o ^^
0.60
-U 0.50
0.40
030
0.20
_ 7/
Ii
0.10
0.00
_ _
0.00 002 0.06 0.08 0.10 0.12 0.14 0.02 0.04 0.06 0.08 0.40 0.12 0J4
004
W
Ke/t=PUe^ffbt2 Ke/t = PLi e/ fbt2
CHART NO. 13 CHART NO. 14
0.0 090
Cn
fC = 8.0 k s i ^ • • f^ = 8.Oksi
* -, -®per
--
l =270.0 ksi . f
pu =270.0 ksi
0.5C 080
f
se =154.9 ksit gt
' •a fse=154,9 k
° 4O
O g= 0.6 t Of 9 = 0.7 t
0 7( 0 7C Q e
pu
0.6( 0.60
0.5C 050
dJ
0.4( 040
0.3( 0.30
0.21 0.20
01( 01C
0.00 0.02 0.06 0.06 0.10 0.12 0.14 0'00 0.02 0.06 0.08 0.10 0.12 0.14
004 0.04
Ke/t=pie/rbt2 Ke/t=Pue^f"bt2
CHART M( I 1-APT MO I^
no 0.9 0.9{
0 fc = 8. f^ = 8.0 ks;
c
ksi fpu
0.81
fSe =154.9 ksi 9t %,a fSel 54.9 k si gt
9p - 0.8 t 0.9 'f
0.7 e 0.71
03 Qa a^
o °Q,
m
Q
qa, °• pu
0.6 0.61
m
N
r
U
0.51 0.5(
d
N
II
0.41 0.41
0.3( 0.3C
0.21 02
0.0c o.oc
V-tie U.U4 U.Ub U.Ut$ U.IU 0.12 0.14 U.uc c1.U4 u.Ub U.Ua U.w 0.12 0.i4
W
Ke/t=Pe^ffbt2 Ke/t=Fbe^ffbt2
i N (D
N
2'I I 1 2" , 2"
16"
138
r'I-IGQT NIfl r,
0.91
fc = 6.Oksi
b ^,^^aPU
fC=270.0 ksi
:gel
'°o fse=1`^4-9 ksi gt
0
9 = 0.7 t
0.71
0
' q e 0
o s P
K= 0.6 3 qo °o> ^'
00
a 6, C
00^' O
U,
U,
K _ 0.512
a
0.51 E
0
o V
O
^^ I 0
041 II Balance Point
O U)
10 10
:.3C O I I O 6
O
II II O
0 2( 0
Tv'7__
0.1C
0.00
/ ^.vc u,U^+ vun U.Ua 0.10 0.12 0.14
Ke/t=PUe/f^bt2
t
=0.7(0.63)6(16)j6 P„ = 425 kips
= 677 kips > 550 kips M„ = 700 in.-kips fP=
f, = 5ksi = 0.7
M= O(K )S, b(t)^
First calculate K, K(elt), g and pn.
= 0.7(0.063)6(16)(16)2
K = P°
= 1083 in. -kips > 875 in. -kips
The 16 x 16-in. column with four 425
= 0.62
3/s-in. diameter strands is adequate. 0.7(5)14(14)
K = 0.62 0 ° oy s R^
AKi V 6C
JK
05C
u
`S 34C
0.3C
II I
0.2C
0. IC
Ke/t=%e^ffbt2
140
8-1/2" Strands
4-
Lr+ ^NI}
f ee = e M. 2250
^t = K t bt 20.7(5)(15)20)2
5(0.04) Of,'
27500 )j = 0.100
0.0031 I 1 + 0.0031(
L ( 57.5 V1 5000 . gt = (20 in.) — 2(2 in.)
g = 16120 = 0.8
= 63.2 ksi Enter Chart No. 3 with K — 0.125 and
The required prestressing level is: R = 0.100; the required prestressing
steel percentage is 0.0035. The re-
fe= 154.9-63.2=91.7ksi quired area of prestressing (A p1 ) is cal-
culated as follows:
The 14 x 14-in, column with four
A, = pbt = 0.0035 (16)20 = 1.12 in.R
1/z-in, diameter strand will work with f8e
This relates to eight ½-in, diameter
91,7 ksi.
strands with a total prestressing steel
DESIGN EXAMPLE 3
Design a rectangular column using
the following data:
Assume strong axis bending. METRIC (SI) UNIT
P„ = 140 kips M 2250 in.-kips
EQUIVALENTS
1 in. = 25.4 mm
= 5 ksi f = 270 ksi 1 in. 2 — 645.6 mm2
1 ft = 0.3048m
fire = 154.9ksi bit — 0.8 ¢=0.7 1 f2= 0.0929 m2
1 psf = 47.99 NIm2
Trial Design 1 1 psf = 47.88 Pa
Initiall y assume that b = 16 in. and t 1 psi = 0.006895 MPa
= 20 in. and calculate the following pa- 1 psi — 0.006895 N/mm2
rameters: 1 pcf — 16.02 kg/m2
1 kip = 4.448 kN
P. 1 kip/ft = 14.594 kN/m
K =
f^ bt 1 in.-kip= 113.0N•m
— 140 1 ft-kip = 1.356 kN•m
= 0.125
0.7(5)(16)20
l e ^O9 >
0.6C
Dp
05C
a
Y 04C
o.9C
0.2C // -i
K = 0.130
010 K= 0.125 cv
o o
r
o to
0.00
Y L ii
002 004 006 006 cc 0.10 Cr 0.12 0.14
Ke/t=Pie/f'bt2
Fig. B5. Design chart for Example 3.
area of 1.224 in.' The results of trial de- 0.125, R = 0.100 to a steel percentage of
sign 1 are shown in Fig. B4. 0.0038 (see Fig. B5).
The coordinates corresponding to this
Analysis of Trial Design 1 end point are as follows:
The interaction curves can be used to K = 0.13 and K(elt) = 0.012
find the load and moment capacity of The actual design load capacity of the
the column by using the following pa- column is given by the following equa-
rameters: tions:
K = 0.125 R=0.100 g=0.8 Pa = 0.7 (0.13) (5) (16)20
A n , __ 1.224 = 145.6 kips > 140 kips
= 0.0038 0.7 (0.102) (5) (16) (20)2
P° = bt 16(20)
= 2285 in.-kips > 2250 in.-kips
A line is drawn from the origin on The 16 x 20-in, column with eight
Chart No. 3 through the point K = b-in. diameter strands is adequate.
142
APPENDIX C — NOTATION