WHAT IS A PRESTRESSED CONCRETE?

Prestressed concrete is a concrete in which internal stresses of such magnitude

and distribution have been introduced such that the stresses resulting from the given
applied loading are counteracted to a desired degree.
METHODS OF PRESTRESSING:
1. Pre-tensioning: The tendons are arranged to pass through the mould or
moulds for a number of similar members arranged end to end, and are
tensioned between external end anchorages, by which the tension is
maintained while the concrete is placed. When the concrete has hardened
sufficiently, the ends of the tendons are slowly released from the anchorages,
thus causing compression to the concrete.
Pre-tensioning is more suitable for mass production of standard members
in a factory.
2. Post-tensioning: The concrete member is cast incorporating ducts for
tendons. When the concrete has hardened sufficiently, the tendons
tensioned by jacking against one or both ends of the member, and
anchored by means of end anchorages which bear against the member at
ends or are embedded in it.

the
are
are
the

Post-tensioning is generally used on site for members cast in their final

place.
HISTORICAL DEVELOPMENT OF PRESTRESSING
The basic principle of prestressing was applied to construction perhaps
centuries ago, when ropes or metal bands were wound around wooden staves to
form barrels. When the bands were tightened, they were under tensile prestress
which in turn created compressive prestress between the staves and thus
enabled them to resist hoop tension produce by the internal liquid pressure.
1886 - Engr. P. H. Jackson of San Francisco, California obtained patents for
tightening steel tie rods in artificial stones and concrete arches to serve as floor
slabs.
1888 - C. E. W. Doehring of Germany independently secured a patent for
concrete reinforced with metal that had tensile applied to it before the slab was
loaded. These applications were based on the concept that concrete, though
strong in compression, was quite weak in tension, and prestressing the steel
against the concrete would put the concrete under compressive stressed which
could be utilized to counter balance any tensile stress produced by dead load or
live load.

G.P.Ancog

1908 - C. R. Steiner of U. S. suggested the possibility of retightening the

reinforcing rods after some shrinkage and creep of concrete had taken place, in
order to recover some of the losses.
1925 - R. E. Dill of Nebraska tried high strength steel bars coated to
prevent bond with concrete. After the concrete had cured, the steel rods were
tensioned and anchored to the concrete by means of nuts.
Modern development of prestressed concrete was credited to E. Freyssinet
of France, who in 1923 started using high-tensile wires for prestressing. Such
wires possessed ultimate strength as high as 250,000 psi.
Practical application of pre-tensioning method was first made by E. Hoyer
of Germany.
Wide application of prestressed concrete was not possible until reliable
and economical methods of tensioning and end anchorages were devised.
1939 - Freyssinet developed conical wedges for end anchorages and
designed double-acting jacks which tensioned the wires and then thrust the
male cones to the female cones for anchoring them. The system is called the
Freyssinet system.
1940 - Prof. G. Magnel of Belgium developed the Magnel system, wherein
two wires were stretched at a time and anchored with a simple metal at each
end. About this time prestressed concrete begin to acquire wide acceptance in
concrete construction.
ADVATAGES OF PRESTRESSING
1. High strength concrete with high concrete with high strength steel can be
used to yield economical proportions. The entire concrete section is fully
utilized to effectively resist the applied moment, whereas only the portion of
the section above the neutral axis is fully effective in ordinary reinforced
concrete.
2. The use of curve tendon profiles enables part of the shear force to be carried
by tendons and the precompression in the concrete tends to reduce diagonal
tension.
3. It can connect relatively longer spans and can accommodate larger load
capacity than ordinary reinforced concrete member.
4. Prestressed structures are more slender and hence more adaptable to artistic
treatment. They yield more clearance where needed.
5. Prestressed structures do not crack under working loads, and whatever cracks
may be developed under overloads will closed up as soon as the load is
removed.
6. Under dead load, the deflection is reduced owing to the cambering effect of
prestress. Under live load, the deflection is smaller becauset of the
effectiveness of the entire uncracked concrete section.
2

G.P.Ancog

7. Prestressed elements are more adaptable to precasting because of the lighter

weight.
PRESTRESSING TERMS
1. Transfer: Transfer is the operation in pretensioning in which the force in the
tendons is transferred to the concrete by bond stress.
2. Anchorages: Anchorages are mechanical devices use to transmit the
prestress to the concrete at each end of the member. Anchorages usually
bear against the member or are embedded in it.
3. Concordant cable: Concordant cable or tendon is a cable in a beam that
produces a line of pressure coincident with the tendon itself. That is the
moment due to loading is annulled due to the cable tension. A concordant
cable induces no support reactions and no secondary moments.
4. Bonded tendons: Bonded tendons are tendons which are bonded to the
concrete either directly or through grouting.
5. Unbonded tendons: Unbonded tendons are tendons which are free to
move relative to the surrounding concrete.
6. Linear prestressing: This a method of prestressing where the structure is
linearly prestressed. The prestressing tendons in a linearly prestressed
structure are necessarily straight; they can be either bent or curved but they
do not go round and round in circles.
7. Circular prestressing:
This a term applied to prestressed circular
structures where restressing wires are wound around in circles.
8. Partial prestressing:
A term given to structural members where
prestressing is done partially. In practice there is partial prestressing if there
are some tensile stress produced in the member under working load
condition.
9. Full prestressing: This is a term which denote that full prestress is applied
to the member so that no tensile stress are produced under working load
condition.
PRESTRESS LOSSES
Initial prestress in steel minus the losses is known as the effective
prestress or the design prestress. The total amount of losses to be assumed in
design will depend on the basis on which the initial prestress is measured. First,
there is the temporary jacking stress to which tendon may be subject for the
purpose of minimizing creep in steel or for which balancing frictional losses.
Then there is a slight release from that maximum stress back to the normal
jacking stress.
As soon as the prestress is transferred to the concrete, anchorage loss will
take place. The jacking stress minus the anchorage loss will be the stress at
anchorage after release and is frequently called the initial prestress.
For post-tensioning, losses due to elastic shortening will gradually take
place. For pre-tensioning, the entire amount of loss due to elastic shortening will
occur at the transfer of prestress. Other losses will take place and these will
include creep and shrinkage of concrete, creep in steel and for post-tensioned
3

G.P.Ancog

members, friction losses will also be considered. For average steel and concrete
properties, cured under average air condition, tabulated percentages may be
taken as representative of the average losses.
Losses after transfer

Average Percentage
Posttensioning
1%

Pretensioning
of
3%

Elastic shortening and bending

concrete
Creep of concrete
Shrinkage of concrete
Creep in steel ( stress relaxation )
Total

6
7
2
18%

5
6
3
15%

Anchorage take-up and friction losses will apply for post-tension members only.

STRESSES AT VARIOUS PRESTRESSING STAGES

N
Qi

Qi
sf

c
si

CB

CT

As
Beam Section

+
Due to Prestress

Due to DL

ST RE SS

Due to LL

_ +
Resultant Stress

VARI AT I O N

Initial Condition Before Transfer:

Qi
si

= initial prestressing force

= initial elongation of steel
4

G.P.Ancog

Q iL
fL
= si
AsEs
Es
si

Condition After Transfer:

Q
si

= effective prestressing force right after transfer

= sf + c
QL
=
AsEs

sf

fsL
Es

=
fcL
Ec

Let fc = stress in concrete at the level of steel after transfer:

Q
(Qe)e
+
Ac
I
fA
f A e2
= s s + s s
Ac
I
si = sf + c
fc =

fcL
fL
fsiL
= s + E
Es
Es
c
fsi = fs + nfc

2
fsi = fs + nfsAs( 1 + e )
Ac
I
Q
2
fsi =
+ nQ( 1 + e )
As
Ac
I
Q
Q
e2
fsi =
+ n A (1+ 2 )
As
r
c

1ST PRESTRESSING CONCEPT: To transform concrete into an elastic material.

G.P.Ancog

load w

My/I

Qey/I

CB

CT

As
_ Mc/I

+ Q/A
+Qec/I
Beam Section Direct
Due to Prestress
Load Effect
Eccentricity
Due to Prestress

+
Resultant Stress

Due to External
Moment M
Due to Load

S T R E S S

V A R I A T I O N

Q
(Qe)y
+
+
A
I

Resultant Stress : f =

My
I

2ND PRESTRESSING CONCEPT: Prestressing for combination of high-strength steel

with concrete.
load w

Q
+

CB

CT

As
Beam Section

+
Due to Direct
Prestress

ST R E SS

External Moment = Internal Moment

M = C or Q
= M/C
= -e

Due to Couple
C-Q
V A R I A T I O N

Resultant Stress : f =

_ +
Resultant Stress

Q
Q
+
y
A
I

3RD PRESTRESSING CONCEPT: Prestressing to achieve load balancing.

6

G.P.Ancog

load w

load w

h
+

CB

CT

As
_

+
Due to Direct
Prestress

Beam Section

S T R E S S

Due to
Net Load w
V A R I A T I O N

_
+
Resultant Stress

Equivalent uniform load from tendon:

8Qh
w =
L2
Net load
: w = w - w
Net moment: M = (1/8)wL2

Resultant Stress : f =

Q
M
+
y
A
I

THE FOUR BASIC REQUIREMENTS IN WORKING STRESS DESIGN OF

PRESTRESSED CONCRETE
1. For loading condition, prestress plus beam weight, the tensile stress at top
fiber must be less than or equal to the allowable tensile at transfer.
2. For loading condition, prestress plus beam weight, the compressive stress at
the bottom fiber must be less than or equal to the allowable compressive
stress at transfer.
3. Under the action of prestress, superimposed load including impact, the
compressive stress at the top fiber must be less than or equal to the
allowable final compressive stress.
4. Under the action of prestress, superimposed load including impact, the
tensile stress at the bottom fiber must be less than or equal to the allowable
final tensile stress.
_

yB

yT

As
Beam Section

+
+
Due to Prestress

_
Due to
Beam
Weight

+
Initial Resultant
Stress

_
Due to
Superimpsed
Load

+
Final
Resultant Stress

Requirements 1 and 2
Requirements 3 and 4

G.P.Ancog

MgYt
Q _ QeYt
> ft allowable at transfer
+
I
I
A
MgYt
Q
eY
> ft allowable at transfer
Top fiber stress
ft =
( 1 - 2t )+ I
A
r
QeYb _ MgYb
Q
< fc allowable at transfer
Requirement # 2: fc =
+
I
I
A
Q
eY
_ MgYb <
fc allowable at transfer
Bottom fiber stress
fc =
( 1+ 2b )
I
A
r
MY
MY
Q _ QeYt
Requirement # 3: fc = (
+ g t + s t < fc allowable at final
)
I
I
I
A
MY
Q _ QeYt
< fc allowable at final Top fiber stress
fc = (
) + It t
I
A
QeY
Q
_ MgYb _ MsYb > f allowable at final
Requirement # 4: ft = (
+ I b )
t
I
I
A
QeYb
MY
Q
> ft allowable at final Bottom fiber stress
ft = (
+
_ t b
)
I
I
A

Requirement # 1: ft =

I
r2 =
A
Mt = Mg + Ms = total moment due to dead load of beam plus superimposed load

= prestressing effectiveness or prestressing efficiency, = 80% for pre-tensioned

members, = 84% for post-tensioned members.

ALLOWABLE STRESSES FOR CONCRETE:

Allowable Stress at
Transfer, MPa

Code
AASHTO

Pre-tension

Allowable Stress at Final

Stage, MPa

Compressiv
e

Tensile

Compressiv
e

0.60 fci

Unbonded
reinforcements

0.40 fc

0.55 fci

ci
= .25
Bonded
reinforcements

f '

AASHTO

Posttension

= .62

ACI

0.60 fci

0.40 fc
Bonded =.50

f ci '

PCI

Posttension

0.60 fc

.25

f ci '

fc '

At precompressed tensile

f '

f '

ci
supports = .25
At ends of simple

supports = .50

Unbonded = 0.00
Severe exposure = .25

fc '

At any point bet.

Pre-tension,
Posttension

Tensile

0.45 fc

f ci '

c
zone = .50
Analysis base on
transformed cracked

section = 1.00
0.45 fc

fc '

0.00

AASHTO Cracking Stresses (Modulus of Rupture)

8

G.P.Ancog

For normal-weight concrete

For sand-lightweight concrete
For all other lightweight concrete

0.62

fc '

0.52

fc '

0.46

fc '

AASHTO Anchorage-bearing stresses

Post-tensioned anchorage at service load (but not to exceed 0.9fci)

20.60 MPa

ALLOWABLE STRESS FOR STEEL:

Code
AASHTO
ACI
PCI

Temporary Jacking
Force
0.94fpy 0.80fpu
0.94fpy 0.80fpu
0.80fpu

Immediately After
Transfer
0.82fpy 0.74fpu
0.82fpy 0.74fpu
0.70fpu

After All
Losses
0.70fpu
0.70fpu
0.60fpu

Post-tensioning tendons at anchorage, immediately after tendon anchorage 0.70f pu.

fpy ; 0.85fpu
For low relaxation prestressing steel, fpy = 0.90fpu.

Term Definitions
fpy = specified yield strength of prestressing tendons
fy = specified yield strength on non-prestressed reinforcements
fpu = specified tensile strength of prestressing tendons
fc = specified compressive strength of concrete
fci = compressive strength of concrete at time of initial prestress

Stres-Relieved Wire for Prestressed Concrete (ASTM A 421)

Min. tensile strength (psi)
Type BA
Type WA
0.192
250,000
0.196
240,000
250,000
0.250
240,000
240,000
0.276
235,000
235,000
Source: Post-Tensioning Institute
Nominal Dia. (in)

Min. stress at 1% extension (psi)

Type BA
Type WA
212,500
204,000
212,500
204,000
204,000
199,750
199,750

Stress-Relieved Seven-Wire Standard Strand for Prestressed Concrete (ASTM

A 416)
Nominal
Diameter
(in)

Breaking Strength
Min. lb)

Nominal Area
(sq. in)

Nominal Weight
(lb/1000 ft)

Min. load at
1% Extension (lb)

(0.250)
5/16 (0.313)
3/8 (0.375)

9,000
14,500
20,000

GRADE 250
0.036
0.058
0.080
9

122
197
272

7,650
12,300
17,000
G.P.Ancog

7/16 (0.438)
(0.500)
3/5 (0.600)

27,000
36,000
54,000

(0.375)
(0.438)
(0.500)
3/5 (0.600)

23,000
31,000
41,300
58,600

3/8

7/16

0.105
0.144
0.216
GRADE 270
0.085
0.115
0.153
0.217

367
490
737

23,000
30,600
45,900

290
390
520
740

19,550
26,350
35,100
49,800

Stress-Relieved Seven-Wire Compacted Strand for Prestressed Concrete

(ASTM A 779)
Nominal Diameter
(in)

0.60
0.70

Nominal Breaking
Strength
(ib)
47,000
67,440
85,430

Nominal Steel Area

(sq. in)

Nominal Weight
(lb/1000ft)

0.174
0.256
0.346

600
873
1176

Reinforcement Grades and Strengths

1982 Standard Type
Grade 40
Grade 60
Grade 40
Grade 60
Grade 60
Deformed Wire Reinforced
Deformed Wire Fabric
Smooth Wire Reinforced

Minimum Yield Strength, fy (psi)

Billet Steel (A615)
40,000
60,000
Axle Steel (A617)
40,000
60,000
Low Alloy Steel (A706)
60,000
75,000
70,000
70,000

Ultimate Strength, fu (psi)

70,000
90,000
70,000
90,000
80,000
85,000
80,000
80,000

DESIGN OF PRESTRESSED CONCRETE:

PRELIMINARY DESIGN. Internal Couple C-Q Method
1. In practice the depth h of the section is either given, known or assumed, as is the
total moment Mt of the section.
2. Under working load, the lever arm a for the internal couple could vary between 30%
to 80% of the overall height h and averages about 0.56h.

10

G.P.Ancog

fc

0.5h

h
As

0.5fc

a ~ 0.65h
~
Q

Resisting Moment and

Stress Distribution

Beam Section

Required effective prestress Q:

Q=

Area of steel:

Mt
Mt
=
a
0.65h

As =

Q
Mt
=
fs
0.65hfs

Total prestress Q = Asfs is also the force Q on the section. This force will produce an
average unit stress on the concrete fc ave.
C
Q
Asfs
=
= 0.50fc (preliminary design)
=
Ac
Ac
Ac
Asfs
Required area of concrete section: Ac =
0.50fc
fc ave =

Check results in the preliminary design and revised as needed.

ELASTIC DESIGN (No tension in concrete) ; small ratios of MG / Mt.

11

G.P.Ancog

Qh
Ac cb

kb

h
d

cgc

kt

cgs

Section
Properties

Q
Ac

Qo
Ac

ct

C
a = e+kt

cb

a = e - kb
Qo h
Ac ct
Just after transfer C at
bottom kern point
ST RESS

C
Qo

At final stage C
at top kern point

D I ST RI BU T I O N

Steps:
1. For the section obtained from preliminary design, the values of M G, kt, kb and Ac are
computed.
2. From the preliminary design section, locate c.g.s ( C at the bottom kern point)
M
e - kb = G
Qo
Stresses at top and bottom fibers will be:
Qo h
Qo h
ft = 0; fb =
A =
Ac ct ; c
fb ct
3. With the location of the c.g.s., compute the effective Q ( C at the top kern point)
f
Mt
Q=
and then Q o = Q( so )
;
fs
e+kt
Top and bottom fiber stresses will be:
Q h
Q h
fb = 0; ft =
Ac =
;
Ac cb
ft cb
4. Compute the required Ac by:
Qo h
Q h
Ac =
Ac =
;
fb ct
ft cb
5. Revised the preliminary section to meet the above requirements for Q and Ac.
1. Repeat steps 2 to 5 if necessary.

ELASTIC DESIGN (No tension in concrete), large ratios for MG / Mt.

12

G.P.Ancog

cgc

kt

ct

kb

h
d

ft

As

cgs

a = e+kt

e- a
C
Qo

cb
a=

Section
Properties

MG
Qo

fb
Just after transfer,
C above bottom
kern point
ST RESS

At final stage of loading

C at top kern point

D I ST RI BU T I O N

When the ratio of MG / Mt is large, the value of e - k b computed from e - kb = MG/Qo

may place c.g.s. outside of the practical limit (i.e. outside the beam section). Then it is
necessary to place the c.g.s. only as low as practicable and design accordingly.
Just after transfer:
(Q oe - MG )cb
e - MG/Q o
Qo
Qo
)
fb =
(
1+
+
=
I
kt
Ac
Ac
Required area:
Ac = Q o (1 + e - MG/Q o )
kt
fb

Steps:
1. From the preliminary section, compute the theoritical location of c.g.s. by
a = e - kb = MG/Qo
If it is feasible to locate c.g.s. as indicated by this equation, follow the first
procedure. If
not, locate c.g.s. at the practical lower limit and proceed as
follows.
2. Compute Q by:
Q=

f
Mt
and then Q o = Q( so )
;
ft
e+kt

3. Compute the required area by the equations:

Q h
Ac =
ft cb
Ac = Q o (1 + e - MG/Q o )
kt
fb
4. Use the greater of the two A cs and the new value of Q and revise the preliminary
section.
1. Repeat steps 1 through 4 if necessary.
13

G.P.Ancog

ELASTIC DESIGN, allowing and considering tension; small ratios of MG/Mt.

ft
ft

kt

cgc

ct

C
a = e+kt

kb
e1
e2

As cgs

cb

C
Qo

fb

fb

Stress Distribution
at Transfer

Section
Properties

ST RESS

Stress Distribution
at Working Load

D I ST RI BU T I O N

If tensile stress ft is permitted in the top fibers, the center of compression C can be
located below the bottom kern by an amount:
f I
f Ak
e1 = t
= t b
Q oct
Qo
For a given moment MG, the c.g.s. can be further located below C by the amount:
MG
e2 = Q
o
.
. . e1 + e2 =

MG + ftAkb
Qo

Lever arm acting up to the top kern point : a = e + kt

Mt - fbAkt
Prestress required
: Q=
a
Q oh ftcb
Q oh
Bottom fiber stress at transfer
: fb = A c + c
then Ac =
t
fbct - ftcb
c t
fbct
Qh
Qh
Top fiber stress under working load: ft =
+
then
A
=
c
c
Accb
b
ftcb- fbct
Steps:
1. From the properties of the preliminary section, obtain e1 + e2 and locate c.g.s.
2. Compute the net moment to be carried by the prestress: Mt - fbAkt and obtain:
Q = (Mt - fbAkt)/a and Qo = Q(fso/fs)
3. Compute the required area by the equations:
Q oh
Ac =
fbct - ftcb
Ac =

Qh
ftcb- fbct

4. Use the greater of the two Acs and revise the preliminary section.
1. Repeat steps 1 through 4 if necessary.
14

G.P.Ancog

For large ratios of MG/Mt, C will be within the kern at transfer, and the allowing of
tension on top fiber will have no effect on the design. The c.g.s. has to be located
within practical limits. Otherwise, the design is made as for the first case. The area of
concrete required is given by:
Ac

Qh
f t cb

Ac

Qo
fb

e ( M G Qo

kt

ELASTIC DESIGN, Composite Sections

ft

ft

cb

ct
kt

cgc

cgc

kb

cb

ct

in-place portion

cgs

cgs

fb

fb

a) Precast Portion, Stress

Distribution at Transfer

b) Composite Section, Stress Distribution

Under Working Load

The procedure of design follows closely the basic approach previously adopted
for noncomposite sections. It is essentially a trial-and-error process, simplified by a
systematic and fast converging procedure and assisted by the use of some simple
relations and formulas. One additional concept introduced for composite action is the
reduction of moments on the composite section to equivalent moments on the precast
portion. This is accomplished by the ratio of the section moduli of the two sections.
Steps:
1. Locate the c.g.s. For a given trial section, the c.g.s. must be so located that the
precast portion will not be overstressed and yet will possess the optimum capacity
in resisting the applied external moments. The c.g.s. must be situated as low as
possible but not lower than given by the following value of eccentricity: e = e 1 + e2
+ kb , where:
f I
e1= t
ctQ o
MG
e2 =
Qo
15

G.P.Ancog

Where: ft = allowable tension stress on top fiber of precast portion at transfer

I = moment of inertia of precast portion
ct = distance to top fiber from c.g.c. of precast portion
2. Compute the equivalent moment of the precast portion. For any moment M C acting
on the composite section, it will produce stresses on the precast portion as follows:
MCct
ft =
I
MCcb
fb =
I
where I = I of composite section, c t and cb = distance to extreme fibers of the
precast portion measured from c.g.c. of the composite section.
Let
and

I/ct
Ict
I/cb
mb =
Icb
mt =

Then we have

mtMC ct
mtMC
=
I
Ackb
mbMC cb
mbMC
fb =
=
I
Ackt
ft =

where Ac = area of the precast section

kb = bottom kern distance of precast section
kt = top kern distance of precast section.
The above indicates that MC can be modified by the coefficients m t and mb so
that it can be reduced to equivalent moments for computation based on the precast
portion properties.
3. Compute the amount of prestress required for the moments as follows. If M P = the
total moment acting on the precast portion, and f b = allowable tensile stress at the
bottom fiber, we have:
Q
MP
mM
e
(
-1+
+ b C = fb
)
Ac
Ackt
Ackt
kt
MP + mbMC - fbktAc
Q=
e + kt
MP + mbMC
if fb = 0
or
Q=
e + kt
from which compute the required
initial prestress Qo. Revised the location of c.g.s. by this new Qo if necessary.
4. In order to limit the bottom fiber stress to the allowable value at transfer, we have:
Qo
( Q oe - MG )
( Q oe - MG ) from which
Ac = 1 Q o +
fb =
+
Ac
kt
f
Ackt
b
In order to limit the top fibers of the precast portion to within allowable
compressive stress ft under working load, we have:
Q
MP + mtMC - Qe
ft =
+
Ac
Ackb
MP + mtMC - Qe
Ac = 1 Q +
kb
ft
16

G.P.Ancog

The greater of the two formulas will control the Ac required for the precast portion.
The top fiber stress of the cast- in-place top flange can be computed by the formula
f = Mc/I, using the applicable values.
Alternative Flexural Design
A - Beam with variable Eccentricities:

tension

Qi
1
Ac

St

fcsi
f1r

ct

fti
t1

Mg

compression

h
cb

fcent

Variable eccentricity profile

1

Let:

Q3e
Ac

Section

At support

t2 Mg

f2r
fts

fc1

Sb

At midspan

Concrete stress distribution at various loading stages

Qi

Qi + Mg

Qe + Mg

Qe + Mg + Mdl + Mll

fci = permissible concrete compressive stress at initial stage

fti = permissible concrete tensile stress at initial stage
fcs = permissible concrete compressive stress at service load stage
fts = permissible concrete tensile stress at service load stage
Required section modulus:

St
Sb

1 M g M dl M ll
f cs f ti

1 M g M dl M ll

f ci fts
I St ct Sb cb
ct
Sb

h St Sb

where:
St = section modulus at top
Sb = section modulus at bottom
ct = top fiber distance
cb = bottom fiber distance
= prestressing efficiency

17

G.P.Ancog

The required initial prestressing force:

Qi = Ac x [fcent] ; where: Ac = concrete area, fcent = axial compressive stress due to prestress.

ct
f ci fti fti
h
1 c
b ( f cs f ts )
h

f cent
f cent

unsymmetrical section

fts

symmetrical section

That prestressing initial prestressing force applied with eccentricity e, must produce a bending moment Qie, at the top
fiber:

Mg
Qi e
f cent f ti
St
St
Mg
Qi e
f cent fci
Sb
Sb
The required eccentricity:

e em f cent fti

St M g

Qi Qi

e em f cent fci

Sb M g

Qi Qi

em emax
B - Beams with constant eccentricity:

tension

compression

ft
ct

e = em

h
cb

2
1

Qi Stresses

Qe Stresses

fcent

Section

Constant eccentricity profile

18

fb

Concrete stress distribution at various loading stages

G.P.Ancog

Required section modulus:

St
Sb

M g M dl M ll
f cs f ti
M g M dl M ll

f ci f ts

Required constant eccentricity:

Qi e
f cent f ti
St
e f cent fti

St
Qi

e ( f cent fci )

Sb
Qi

C - Symmetrical Cross Section

Often for practical reasons, a symmetrical cross section is chosen, even though the limit stresses indicate otherwise. In
cuch case, St = Sb = S, and the stress change at the top and bottom fiber will be identical as the transverse load is
applied. As a result, only three of the four limit stresses will be achieved, in general. In such circumstanaces, there is
no unique combination of Qi and e that will serve, but rather a range of values. The best choice will normally be that
which minimizes the prestress force required.
If the bottom fiber stress controls: Sb > St:

Sb

1 M g M dl M ll

S t ; S Sb

f ci fts

If the top fiber stress controls: St > Sb:

St

1 M g M dl M ll
f cs fti

Sb ; S St

The concrete centroidal stress is computede based on service stress load condition 4:

f cent

1 cb
( f cs f ts )
h

fts

The initial prestress Qi can be found by:

e em f cent fti

St M g

Qi Qi
19

G.P.Ancog

SHEAR IN PRESTRESSED CONCRETE:

fc

Vs

C
Q

max

Vc

C
O

Portion of Beam

fc

fc

Sc

State of stress of Element A

St

Mohrs Circle of Stress for Element A

Conventional Method:
1. From the total external shear V across the section, deduct the shear V s carried by
the tendon to obtain the shear Vc carried by the concrete: Vc = V - Vs
2. Compute the distribution of Vc across the concrete section:
VQ
= c v
Ib
3. Compute the fiber stress distribution for that section due to external moment M,
the prestress Q and its eccentricity by the formula:
Q + Qec+ Mc
fc =
A
I
I
4. the maximum principal tensile stress in St corresponding to the above and fc is
computed by:
f 2
fc
St =
+ ( c )
2
2
Limiting values for principal tensile stress in prestressed concrete:
St > 0.013fc to 0.033fc

for beams w/o web reinforcements

St > 0.040fc to 0.100fc

for beams with web reinforcements

In general, shear design under working alone is not sufficient.

Required stirrups spacing under working load:
20

G.P.Ancog

s=

Where:

Avfsv
Stbw

Av = area of web reinforcement at spacing s

fsv = allowable stress of web reinforcement
St = principal tensile stress
bw = width of beam web

ULTIMATE SHEAR DESIGN

At loads near failure, a prestressed beam is usually extensively cracked and
behaves much like an ordinary reinforced concrete beam. Accordingly, many of the
procedures and equations developed for non-prestressed beams can be applied to the
design of web reinforcement for prestressed beams also.
Shear design id based on the relation:
Vu < Vn
Where:
Vu = total shear force applied to the section at factored loads
Vn = the sum of the contributions of the concrete and web
reinforcement.
Vn = Vc + Vs and strength reduction factor = .85
In computing the factored load shear Vu , the first critical section is assumed to
be at a distance h/2 from the face of the support, and sections located a distance less
than h/2 are designed for the shear computed at h/2.
The shear force Vc resisted by the concrete after cracking has occurred is taken
equal to the shear that caused the first diagonal crack. Two types of diagonal cracks
have been observed in tests of prestressed concrete beams:
1. Flexural-shear cracks, occurring at nominal shear V ci, start as nearly vertical
flexural cracks at the tension face of the beam, then spread diagonally
upward (under the influence of diagonal tension) toward the compression
face. These are common in beams with low value of prestress force.
2. Web-shear cracks, occurring at nominal shear Vcw, start in the web due to
high diagonal tension, then spread diagonally both upward and downward.
These are often found in beams with thin webs with high prestress force.
Vci = 0.60 fc bwd + Vo + Vcr
Where:
Vo = shear caused by beam self weight ( without load factor)
Vcr = additional shear due to superimposed dead load and live loads
Vcr = V d+l Mcr
M d+l
I
Mcr = c ( 6 fc + f2pe - f2o )
c2
Where:
Vd+l/Md+l , the ratio of superimposed dead load and live load shear to
moment remain constant as the load increases to cracking load.
c2 = distance from concrete centroid to tension face
f2pe = compressive stress at tension face resulting from effective prestress
force alone
f2o = bottom-fiber stress due to beam self-weight
21

G.P.Ancog

Vci = 0.6 fc bwd + Vo +

V d+l M
cr < 1.70 fc bwd
M d+l

the shear force causing web-shear cracking can be found from an exact principal
stress calculation, in which the principal tensile stress is set equal to the direct tensile
capacity of concrete (conservatively taken equal to 4vf c according to ACI code.
Alternatively, the ACI code permits the use of the approximate expression:
Vcw= (3.5 fc + 0.3fpc)bwd + Vp
in which fpc is the compressive stress in the concrete, after losses, at the centroid of
the concrete section (or at the junction of the web and the flange when the centroid
lies in the flange) and Vp is the vertical component of the effective prestress force.
After Vci and Vcw have been calculated, then V c , the shear resistance provided by
the concrete, is taken equal to the smaller of the two values.
Calculating Mcr, Vci, and Vcw for a prestress beam is tedious matter because many
of the parameters vary along the member axis. For hand calculations, the required
quantities may be found at discrete intervals along the span, such as L/2,L/3,L/6 and
at h/2 from the support face, and stirrups spaced accordingly, or computer
spreadsheets may be used.
To shorten calculation required, the ACI code includes, as a conservative
alternative to the above procedure, an equation for finding the concrete shear
resistance Vc directly:
< 2 fc bwd
V d
Vc= (0.6 fc + 700 u )bwd
Mu
>5 fb d
c

in which Mu is the bending moment occurring simultaneously with shear force V u , but
Vud/Mu is not to be taken greater than 1.0
Contribution of shear reinforcement to the shear strength of the prestressed
beam:
Vs = Avfyd
s
The total nominal shear strength Vn is found by summing the contributions of the
steel and concrete, as indicated by:
Vu = Vn = (Vs + Vc)
Afd
= ( v y
+ Vc )
s
Normally, in practical design, the engineer will select a trial stirrup size, for which the
required stirrup spacing is found.
Avfyd
> 0.75h or 24 in.
s=
Vu - Vc

At least a certain minimum area of shear reinforcement is to be provided in all

prestressed concrete members, where the total factored shear force is greater than
(Vc) provided by concrete. Exceptions are made for slabs, footings, concrete-joist
floor construction, and certain very shallow beams. The minimum area of shear
reinforcement to be provided in all other cases is to be taken equal to the smaller of:
22

G.P.Ancog

bws
fy
Apfpus d
Av =
80fyd b
w

Av = 50

in which Ap is the cross-sectional area of the prestressing steel, f pu is the ultimate

tensile strength of the prestressing steel, and all other terms are as defined before.

ULTIMATE FLEXURAL STRENGTH OF PRESTRESSED CONCRETE

Preliminary Design.
The amount of mathematics involved in the design of prestressed concrete
section is less in ultimate design than in the elastic design, since the ultimate flexural
strength of sections can be expressed by simple semi-empirical formulas. For
preliminary design, it can be assumed that the ultimate resisting moment of bonded
prestressed sections is given by the ultimate strength of steel acting with a lever arm.
This lever arm varies with the shape of section and generally ranges between 0.60h
and 0.90h, with a common value of 0.80h. Hence the area of steel required is
approximated by:
As =

Mu
MT(LF)
=
(.8hfps)
0.8hfps

where:
LF = factor of safety or load factor
Assuming that concrete on the compressive side is stressed to 0.85f c, then the
required ultimate concrete area under compression is :
Ac =

MT(LF)
=
0.8hfps( 0.85fc )

Mu
.8hfps(.85fc )

which is supplied by the compression flange ( occasionally with the help of part of the
web). The web area and the concrete area on the tension side are designed to provide
the shear resistance and the encasement of steel, respectively. In addition, concrete
on the precompressed tension side has to stand the prestress at transfer.
Final Design.

23

G.P.Ancog

fc

cu=0.0034

0.85fc

a =1c

a/2

jd = d - a/2
As

s =fsu/Es

Section

T = Asfsu

fsu

Conditions at Ultimate Strength

Equivalent Stress Block

Steel stress:
1d - a
fps =
> fpy
cE
s
a
a = 1c
Ultimate resisting moment: ACI
Mu = Asfps(d - a/2 ) = 0.85fcab( d - a/2 )

In more convenient form:

Mu = Asfps d ( 1 - 0.59)

in terms of steel area

Mu = bd2 fc ( 1 - 0.59 ) in terms of concrete section Recttangular

f
As
sections and for
= ps ; =
T and I-sections
fc
bd

with N.A. falls

within the flange.

Mu = Asfps( d - a/2 )
a=

Asfps
fpsd
=
0.85fcb
0.85fc

Mu = ( As - Asf )fps( d - a/2 ) + Asffps( d - t/2 )

( As - Asf )fps
f
a=
Asf = 0.85 c ( b - b )t
0.85fcb
fps

T and I-sections
with N.A. falls
within the web.

p fpu
] members with bonded tendons
1 fc
fc
fps = fpe + 10,000 + 100 members with unbonded tendons with L/d< 35
fc
fps = fpe + 10,000 + 300 members with unbonded tendons with L/d > 35

fps = fpu [ 1 -

fps > [ fpy , fpe +30,000 psi ] min. for all cases

p = .55 for fpy/fpu not less than 0.80, typical high strength bars
p = .40 for fpy/fpu not less than 0.85, typical ordinary strand
p = .28 for fpy/fpu not less than 0.90, typical low-relaxation strand
PCI Code:
24

G.P.Ancog

1. Rectangular sections or flanged sections in which the neutral axis lies within the
flange.
Mu = Asfps d ( 1 - 0.60)
2. Flange sections in which N.A. falls outside the flange.
Mu = ( As-Asf )fpsd( 1 - 0.60)+ 0.85f c( b-b )t( d-.50t )
0.85fc( b-b )t
f
A -A
= ps ; = s sf ; Asf =
fps
fc
bd
3. Where information for the determination of fps is not available and provided that fpe
is not less than 0.50fpu the following values shall be used:
Bonded members
: fps = fpu( 1 - 0.50fpu/fc )
Unbonded members : fps = fpe + 15,000 psi or fps = fpe + 103 MPa
In ACI Code, .59 is used instead of 0.60.
Limitations on steel percentages:
0.85a < 0.36
1
d
fps
=
> 0.30, and when is greater than 0.30:
fc
1. Rectangular or flanged sections in which N.A. lies within the flange
Mu = 0.25fcbd2
2. Flanged sections in which N.A. falls outside the flange
Mu = 0.25fcbd2 + 0.8fc( b - b )t( d - 0.5t )
PCI
Ec = 1.80x106 + 500fc
Ec = 12.4x10

+ 500fc

psi
MPa

Load Factors:
U = 1.8( D + L )
U = 1.2D + 2.4L
U = 1.4( D + L + W )
Capacity Reduction Factors:
= 0.90 for flexure
= 0.85 for shear

Ec = 33w1.5

ACI
fc '

Ec = 0.043w1.5

fc '

psi
MPa

U = 1.4D + 1.7L
U = .75( 1.4D + 1.7L +
1.7W )
= 0.90 for flexure
= 0.85 for shear

ULTIMATE VS. ELASTIC DESIGN

At the present time, both the elastic and the ultimate designs are used for
prestressed concrete, the majority of designers still following the elastic theory. It is
difficult to state exact preference from or the other. Each has its advantages and short
comings. But, whichever method is used in design, the other one must often be
applied for checking. For example, when the elastic theory is used in design, it is the
practice to check for the ultimate strength of the section in order to find out whether
25

G.P.Ancog

the section is overstressed under certain conditions of loading and whether the
deflections are excessive. Overstressing is objectionable because it may result in
undesirable cracks and creep and fatigue effects. When the design is of conventional
types and proportions, such checking becomes unnecessary, because it is then
generally known that designing by one method will yields safe results when checked
by the other. This is, in fact, the reason why such checking is not required of reinforced
concrete structures designed by the usual codes. When we delve into new types and
proportions, it is possible that elastic design alone will not yield a sufficiently safe
structure under overloads, while the ultimate design by itself might give no guarantee
against excessive overstress under working conditions. It is therefore deemed
desirable to apply both the elastic and the ultimate methods, especially for structures
of unusual proportions.
An understanding of both theories of design is also essential informing judgment
when designing structures. Sometimes, design based on one method will yield
different proportions from those based on the other.
DESIGN EXAMPES
Investigation Problem :
1. A prestressed concrete rectangular beam 20x30 has a simple span of 24ft and is
loaded by a uniform load of 3k/ft including its own weight. The prestressing tendon
is located as shown and produces an effective prestress of 360k. Compute fiber
stresses in concrete at the midspan.
w = 3k/ft

30

20

Prestress Q = 360k

24

Beam Section
at Midspan

Beam Elevation

Solution 1: Transforming concrete into an elastic material.

Given:
Effective prestress, Q
360k
Tendon eccentricity at midspan, e = 15- 6 in
9=
Moment,
M = 1/8wL 2 = 1/8(3)(24)2 216 k-ft
=
26

G.P.Ancog

Section properties ( no hole deduction)

A=
20x30=
I =1/12bh3 =1/12x20x303=
Extreme fiber
y=30/2=

600 in2
45,000i
n4
15in

Stresses:
Q
Qey
My
+
+
A
I
I
360,000 + 360,000x6x15 + 216x1,000x12x15
=
600
45,000
45,000
= 600 + 720 + 864
= 600 - 720 + 864 = 744 psi for top fiber
= 600+ 720 - 864 = 456 osi for bottom fiber

f=

Stress diagram:
-720psi

600psi
Q
A

864psi

720psi
Qey
I

-864psi
My
I

744psi

456psi
Q + Qey + My
A
I
I

Solution 2: Apply the second concept of prestressing using the pinciple of internal
couple
method.
Take one-half of the beam as free body, thus exposing the internal couple. The
external moment at the section is:

27

G.P.Ancog

w = 3 k/ft
744psi

e
9

C
Q
456psi
Stress Distribution at
Midspan

Half Elevation of Beam

M=

wL2
3x242
= 216 k-ft
=
8
8

The internal couple is furnished by the forces C = Q = 360k which must act at a lever
arm of:
M =C = Q
216x12
= M/Q =
= 7.2 in
360
Since Q acts at 9in from the bottom, C must be acting at 16.2 in from it. Thus the
center of the compressive force C is located.
= - e = 7.2 - 6 = 1.2 in.
Stress:
Q
Q y
+
A
I
360,000 + 360,000x1.2x15
=
600
45,000
+
= 600 144
= 744 psi for top fiber
= 456 psi for bottom fiber

f=

Solution 3: Applying the concept of load balancing.

28

G.P.Ancog

w = 3k/ft

15

20

w = 2.5 k/ft

30

15

h= 6in
9

9
Section at
Midspan

24ft
Beam Elevation Showing Superimposed Load
and Equivalent Tendon Upward Force

The upward uniform force w from the tendon on the concrete is obtained from:
1
Qh = wL2
8
8Qh
w = 2
L
= 8x360x(6/12)
242
= 2.5 k/ft
Hence the net downward load on the beam is w = 3.0 - 2.5 = 0.5 k/ft, and the
resulting moment due to this net load is:
2
2
M = wL = 0.5x24 = 36 k-ft
8
8
Stresses:
Q
My
+
A
I
360,000 + (36x1,000x12)15
=
600
45,000
= 600+ 144
= 744 psi for top fiber
= 456 psi for bottom fiber

f=

Stress Diagram:
600psi

30

20

144psi

600psi
Section at
MIdspan

-144psi

Due to axial
prestress Q

744psi

Due to moment
of net load w

456psi
Resultant
stress

Design Problem:
29

G.P.Ancog

2. Design a prestressed concrete I-beam with the following data:

MT = 320 k-ft
MG = 40 k-ft
ft
= -.30 ksi
fb = -0.24 ksi

ft
fb
fso
fs

= 1.60 ksi
= 1.80 ksi
= 150 ksi
= 125 ksi

Solution:

Preliminary Design: h = k

MT

; ( k = 1.5 - 2.0) , for k = 2 , h = 35.77, try h= 36

Compute the required prestress: Q = MT/(.65h)

Q = 320x12/(.65x36) = 164.10 k
Qo = Q(fso/fs) = 164.1(150/125) = 196.92 k
Area of tendon:

As = Q/fs = MT/(.65hfs) = 164.10/125 = 1.3128 in2

Required concrete area

Ac = Asfs/(.5fc) = Q/(.5fc) = 164/(.5x1.60) = 205 in2

Trial section:
16
Section Properties

cgc

kb

18

36

kt

cgs

A1 = 4x16 = 64
A2 = 4x16 = 64
A3 = 4x28 = 112
Ac
= 240 in2

18

4
A3

16

16

A1

I1 = (1/12)bt3 + A1x162 = 16x43/12+64x162 = 16469.33

I2 = (1/12)bt3 + A2x162 = 16x43/12+64x162 = 16469.33
I3 = (1/12)b(h-2t)3
= 4x283/12
= 7317.33
I
= 40255.99 in4
2
kt = kb = r /c = ( I/A)/c =(40255.99/240)/18 = 9.32

A2

16
Trial Section

Ratio of MG/MT = 40/320 = 0.125, small ratio

Locate c.g.s:
e1+e2 =

MG + ftAkb
= 40x12+0.3x240x9.32 = 5.84 in
Qo
196.92

Hence c.g.s. can be located 5.84 below the bottom kern or 2.84 above the bottom
fiber.
Compute the net moment to be carried by the prestress:
30

G.P.Ancog

MT - fbAkt = 320x12 - .24x240x9.32 = 3303.17 k-in

The lever arm a = kt + e = 9.32 + ( 9.32+5.84 ) = 24.48 in. and the required
prestress:
M - f Ak
320x12 - .24x240x9.32
Q = T b t =
= 134.93 k
a
24.28
Q o = Q(fso/fs) = 134.93(150/125) = 161.92 k

To limit the bottom fiber stress to 1.80 ksi, we need an Ac of:

Q oh
161.92x36
Ac =
=
= 215.89 in2
fbct - ftcb
1.8x18 - 0.3x18
To keep the top fibers to 1.60 ksi, we need an Ac of:
Qh
134.93x36
Ac =
=
= 198.43 in2
ftcb- fbct
1.6x18 - 0.24x18
Ac = 215.89 in2 controls and indicates that our trial section of A c = 240 in2 > 215.89
in2 is satisfactory. Our trial section may be reduced and a new section tried over again.
3. Design a prestress concrete T- beam with the following data:
MT = 320 k-ft
ft = 1.60 ksi
MG = 210 k-ft
fb = 1.80 ksi
ft
= -.30 ksi
fso = 150 ksi
fb = -0.24 ksi
fs = 125 ksi
Solution:
MT
Preliminary Design: h = k
; ( k = 1.5 - 2.0) , for k = 2 , h = 35.77, try h= 36
Compute the required prestress: Q = MT/(.65h)
Q = 320x12/(.65x36) = 164.10 k
Qo = Q(fso/fs) = 164.1(150/125) = 196.92 k
Area of tendon:
As = Q/fs = MT/(.65hfs) = 164.10/125 = 1.3128 in2
Required concrete area
Ac = Asfs/(.5fc) = Q/(.5fc) = 164/(.5x1.60) = 205 in2
Trial section:
18
Section Properties
A1

d1

kt
cgc
kb
cb

A2

36

d2

y2= 20

ct

yc

y1= 2

A1 = 4x18 = 72
y1 = 2.0 A1y1 = 144.00
A2 = 4x32 = 128
y2 =20.0 A2y2 = 2560.00
Ac
= 200 in2
Aiyi =2704.00
yc = ct = (Aiyi )/Ac = 13.52 in, cb = 22.48 in.
d1 = 11.52in, d2 = 6.48in
I1 = (1/12)bt3 + A1xd12 = 18x43/12+72x11.522 = 9651.15
I2 = (1/12)bt3 + A2xd22 = 4x323/12+128x6.482 = 16297.43
I
= 25948.58 in4
2
kt = r /cb = ( I/A)/cb =(25948.58/20)/22.48 = 5.77
kb= r2/ct = ( I/A)/ct =(25948.58/20)/13.52 = 9.59

cgs
4
Trial Section

Ratio of MG/MT = 210/320 = 0.65625, large ratio

31

G.P.Ancog

Locate c.g.s:
MG
210x12 = 12.79 in.
=
Qo
196.92
e = 12.79 + kb = 12.79 + 9.59 = 22.38 in
e - kb =

The eccentricity e = 22.38 in. is almost equal to c b which is not feasible. We fix the
location of c.g.s. 3 in. above bottom fiber; then no tension will be in top fiber. The new
value of e:
e = cb - 3 = 19.48 in.
and the lever arm a:
a = e + kt = 19.48+ 5.77 = 25.25 in.
Compute the amount of prestress required:
M - f Ak
320x12 - .24x200x5.77
Q = T b t =
= 141.11 k
a
25.25
Q o = Q(fso/fs) = 141.11(150/125) = 169.33 k
To keep the bottom fiber stress within limits, we compute the required concrete area
from:
Qo
e - (MG/Q o)
(
Ac =
1+
)
fb
kt
169.33 (
1 + 19.48 - 210x12/169.33 ) = 169.03 in2
=
1.80
5.77
To keep the top fiber stress within limits, we compute the required concrete area from:
Qh
141.11x36
Ac =
= 155.24 in2
=
ftcb- fbct
1.6x22.48 - 0.24x13.52
Ac = 169.03 in2 controls and the area furnished from our trial section is 200 in 2 which
may be reduced if desired.
Load Balancing Problem:
4. A double cantilever beam is to be designed so that its prestress will exactly balance
the total uniform load of 23.5 kN/m on the beam. Design the beam using the least
amount of prestress assuming that the c.g.s. must have a concrete protection of
75mm. If a concentrated load P = 65 kN is applied at the midspan, compute the
maximum top and bottom fiber stresses.

32

G.P.Ancog

P = 65 kN
w = 23.5 kN/m

750

300

6m

6m

15m

Beam
Section

Solution:
In order to balance the load in the cantilever, the c.g.s at tip must coincide with
the c.g.c. with a horizontal tangent.
To use the least amount of prestress, the eccentricity over the support should be
a maximum.
h = 750/2-75 = 300 mm or 0.30 m
The prestress required:
wL2
Qh =
2
wL2
23.5x62
=
Q=
= 1410 kN
2h
2x.30
In order to balance the load at midspan, using the same prestress Q = 1410 kN, the
sag of the parabola h1 must be:
wL2
Qh1 =
8
wL2
23.5x152
=
= 0.468 m
h1 =
8Q
8x.1410
Under the combined action of the uniform load and prestress, the beam will have no
deflection anywhere and the uniform compressive stress:
Q
1410x103
= 6.267 MPa
fc =
=
Ac
(300x750)
Due to P = 65 kN:
M=

PL = 65x15 = 243.75 kN-m

4
4

The extreme fiber stresses:

33

G.P.Ancog

Q
+ 6M
Ac
bh2
1410x103 + 6x243.75x106
=
(300x750)
300x7502
= 6.267 + 8.667
= 14.934 MPa for top fiber
= -2.40 MPa for bottom fiber

f=

75
h1 = 468
h=300
6m

h=300
15m

207

6m

Cable Profile

Shear Design Problem (Working load condition)

5. A prestressed concrete beam has a rectangular section shown and is subjected to a
shear of 150k under working loads. The effective prestress in the tendons totals
300k and is inclined at an angle of arc sin = 1/6. The fiber stress distribution
under working load is 505 psi throughout. Half-inch U-stirrups are to be used ( A v =
0.40 in2), fsv = 20,000 psi.

10
30

Vc=100k

505

60

250

300k
Vs=50k
Portion of Beam

Section

Fiber stress
Shear stress
Under Working Load Condition

Solution:
Under working load, the shear carried by the tendon is
Vs = 300xsin = 300(1/6) = 50k
Vc = V - Vs = 150 - 50- = 100k
Consider the state of stress at centroidal axis:

34

G.P.Ancog

3V
2A
3(100x1000)
=
= 250 psi
2(10x60)
f/2 = 505/2 = 252.5 psi

Computing the principal tensile stress:

St = ( f/2 )2 - f/2
= (250)2 + (252.5)2 - 252.5
= 102.82 psi

Required stirrup spacing:

Avfsv
s=
Stbw
0.40x20,000
=
102.82x10
= 7.78 in c.c.

Shear Design Problem (Ultimate load condition)

6. The unsymmetrical beam shown carries an effective prestress force of 228k and
supports a superimposed dead load of 345 plf and service live load of 900 plf, in
addition to its own weight of 225 plf, on a 50 ft simple span. At the maximummoment section, the effective depth to the main steel is 24.5 in. ( eccentricity 11.4
in. ). The wires are deflected upward 15 ft from the support, and eccentricity is
reduced linearly to zero at the support. If concrete having strength fc = 5,000 psi
and stirrups with fy = 40,000 psi are used, and if prestressed wires have strength
fpu = 275 ksi, what is the required stirrup spacing at a point 10 ft from the support?

35

G.P.Ancog

18

Ac = 245 in2
As = 1.75 in2
Ic = 24,200 in4
ct = 13.1 in, cb = 15.9 in.
d1 = 11.52in, d2 = 6.48in
r2 = Ic/Ac = 99 in2

243.5

ct=13.1
e = 11.4

cgc

cb=15.9

19

29

Section Properties

12
Section

e10
10

15

e=11.4

50

15

Cable Profile

Solution:
At a distance from the support, the tendon eccentricity is:
10
e10 = 11.4 x
= 7.6 in.
15
The effective depth d at 10 ft from the support, d = 13.1+7.6 = 20.7 in.
According to the ACI Code, the larger value of d = .80h = .8x29 = 23.2 in will be
used.
Calculate Vci at 10 ft from support:
The bottom fiber stress due to effective prestress acting alone is:
c2 = cb = 15.9 in
Q
e10c2
288,000
7.6x15.9 = 2,610.34 psi
f2pe =
)
Ac ( 1 + r2 )= 245 ( 1 +
99

36

G.P.Ancog

The moment and shear at the section due to beam self weight alone are,
respectively:
wx
Mo,10 = o ( L - x ) = 0.255x10 ( 50 - 10 ) = 51 k-ft
2
2
Vo,10 = wo( L/2 - x ) = 0.255( 50/2 - 10 ) = 3.825 k
and the bottom fiber stress due to this load is:
Mo c2
(51,000x12)15.9 = 402.10 psi tension
f2o =
=
I
24,200
Compute the cracking moment, Mcr:
I
Mcr = c (6 fc + f2pe - f2o ) = 24,200 ( 6 5,000 + 2,600 - 402.10 )
c2
15.9
= 3990966.69 in-lbs / 12 = 332580.56 ft-lbs

The ratio of the superimposed load shear to moment at the section is:
Vd+l
w( L/2 - x )
L - 2x
=
=
Md+l
wx/2( L - x )
x( L - x )
50
2x10
=
= 0.075
10(50 - 10)
Then Vci can be computed from:
Vd+l
Vci = 0.6 fc bwd + Vo + M
< 1.7 fc bwd
d+l
= [ 0.6 5000 (5x23.2) + 3825 + 0.075(332580.56 ] / 1000
= 30.32 k
1.7 fc bwd = 1.7 5000 (5x23.2)/1000 = 13.94 k < 30.32 k, Vci ok

Calculate Vcw for:

= tan-1(11.4/(15x12)) = 3.6238o
Vp = Qsin = 288xsin 3.6238o = 18.20 k
Concrete compressive stress at the centroid:
fpc = Q/Ac = 288x1000/245 = 1175.51 psi
Thus Vcw is:
Vcw = (3.5 fc + 0.3fpc )bwd + Vp
= [( 3.5 5,000 + 0.3x1175.51 )(5x23.2) + 18200 ] / 1000
= 87.82 k

The shear force Vc is:

Vc = [ Vci , Vcw]min = [ 30.32 , 87.82 ]min = 30.32 k
The factored shear force Vu at the section:
Vu = 1.4wd( L/2 - x ) + 1.7wl( L/2 - x )
= 1.4(.255+.345)(50/2 -10 ) + 1.7( .900)(50/2 - 10) = 35.55 k
Using #3 U-stirrups, Av = 2x.11=.22 in2, and the required spacing s:
37

G.P.Ancog

s=

Avfvyd
= .85x.22x40,000x23.2
= 17.75 in
Vu - Vc
35.55x103 - .85x30.32x103

Check for maximum spacing code requirements:

b s
A f s d
Av = 50 w
Av = p pu
fvy
80 fvy d bw
5( s )
1.75 275 s 23.2
0.22 = 50
0.22 =
40,000
80 40 23.2 5
s = 35.2 in

s = 15.75 in

s = 15.75 in governs.

Comparison of stirrups spacing based on conservative approach.

L - 2x
Vu
w ( L/2 - x )
= 30/400 =0.075
=
= u
x(
L - x)
Mu
wux/2( L - x )
< 2.0 fc bwd
Vud
b
d
Vc = (0.6 fc + 700 x
) w
Mu
> 5.0 fc bwd
= [ 0.6 5000 + 700 x 0.075( 23.2/12) ] (5x23.2) / 1000
= 16.69 k
2.0 fc bwd = 2.0 5000 (5x23.2)/1000 = 16.40 k < 16.69 ok
5.0 fc bwd = 5.0 5000 (5x23.2)/1000 = 41.01 k > 16.69 ok

Required spacing s:
Avfvyd
= .85x.22x40,000x23.2
s=
= 8.12 in
Vu - Vc
35.55x103 - .85x16.69x103
The simpler but conservative approach yielded more stirrups!

Ultimate Design Problem

7. Design a prestress unsymmetrical I-beam to carry a total dead load and live load
moment of 320 k-ft having an equivalent ultimate moment of 640 k-ft. Use Grade
250 with dia. strand, fps = 220 psi, carrying an effective prestress of f pe = 143
ksi. Concrete strength is fc = 4,000 psi. and consider a bonded beam.
Solution:
Preliminary design
Trial h = k M T = 2.0v (320) = 35.77 in say 36 in.
Required area of tendon:
38

G.P.Ancog

As =

Mu
640x12
=
= 1.346 in2
(.80hfps)
0.9(.80x36 )220

The concrete compression area:

Mu
640x12
Ac =
=
= 87.146 in2
(.80hx.85fc)
0.9(.80x36 )(.85x4.0)
Trial Section for Final Design:

18

d= 32

27

36

12
Section

b = 18 in
b = 5 in
t = 4 in
d = 32 in
As = 1.346 in2
fc = 4.0 ksi
fps = 220 ksi
1 = 0.85
= As/(bd) = 0.0023368
= fps/fc = 0.12852
c = 1.18d/1= 5.709 in > t = 4 in
N.A. is within the web
Asf = (.85fc(b-b)t)/ fps = 0.8036 in2
As - Asf = 0.5424 in2
New value of :
= (As - Asf)/(bd) = 0.0033875

Area provided by compression side:

Ac = 18x4 + 5x1.709 = 80.545 in2
Moment capacity of the section:
Mu =( As-Asf )fpsd( 1 - 0.59)+ 0.85f c( b-b )t( d-.50t )
= 0.90[0.5424x220x32( 1 - .59x.0033875) +.85x4.0( 18-5)(4)(32 - 4/2)]
= 8521.61 k-in / 12 = 710.13 k-ft > 640 k-ft , ok

More exact solution:

From:

C=T
.85fcab + Asffps = ( As - Asf )fps
a=

fps[( As - Asf ) - Asf ]

.85fcb

Let ( As - Asf ) = As1 then

f [A - A ]
220( As1 - .8036) = 21.568A - 17.3325
a = ps s1 sf =
s1
.85fcb
0.85x4.0x5

From:

39

G.P.Ancog

Mu =( As1fps( d - a/2) + Asffps( d - t/2 )]

640x12 = 0.9[ As1x220( 32 - (21.568As1 - 17.3325)/2) + 0.8036x220( 32 - 4/2)]
8533.33 = 7040 As1- 2372.48 As12 + 1906.575As1 + 5303.76
3229.57 = -2372.48 As12 + 8946.575As1
As12 - 3.77098As1 + 1.36126 = 0
3.77098+ (-3.77098)2 - 4(1.36126)
As1 =
2
= (3.77098 + 2.9623) / 2
= 0.40433 in2

The required area of prestress As:

As = As1 + Asf
= 0.40433 + 0.8036
= 1.20793 in2
Which is very near to the preliminary value of A s = 1.346 in2 The trial section will be
adopted with the amount of As = 1.207932.
The web is of course to be designed by shear.
ANCHORAGE ZONE DESIGN
Bursting zones

Spalling zones

b
Idealization of the anchorage zone for
analysis

Tensile stresses at the spalling and

bursting zones

Rational design for the reinforcement for end zones must recognize that
horizontal cracking is likely. If adequate reinforcement is provided, so that the cracks
are restricted to a few inches in length and to 0.01 in. or less in width, these cracks
will not be detrimental to the performance of the beam either at service load or at the
overload stage. It should be noted that end-zone stresses in prestressed concrete
beams do not increase in proportion to loads. The failure stress fps in the tendon at
beam failure is attained only at the maximum moment section.
For pretensioned members , a very simple equation has been proposed for
end-zone reinforcement:
Qh
At = 0.021 i
fslt
where

At = total cross-sectional area of stirrups necessary, in2

Qi = initial prestress force, lb.
40

G.P.Ancog

h = total member depth, in.

fs = allowable stress in stirrups, psi
lt = transfer length, in.
An allowable stress of 20,000 psi has been found in tests to produce acceptably small
crack widths.. The required reinforcement having total area At should be distributed
over a length equal to h/5 measured from the end face of the beam, and for most
efficient crack control the first stirrup should be placed as close to the end face as
practical.
For post-tensioned members, end-zone reinforcement is often designed on
the basis of an equilibrium analysis of the cracked anchorage zone. As shown in the
figure, the end region of a post-tensioned beam with an initial prestress of Q i applied
at an eccentricity e. At some distance l from the end, the compressive stress
distribution is linear and the forces and stresses acting on the free body are in
equilibrium.

g e

Qi

Qi
e
( 1- 6
)
bh
h

Qi
e
( 1+ 6 )
bh
h

End of beam showing free body location

C
V
2

Qi
0

3
l

Qi
e
( 1+ 6 )
bh
h

Forces on free body

Any longitudinal section through the anchorage zone at a distance y from the
bottom fiber is subjected to a bending moment, which can be determined from the
forces that are acting on the ends of the block.
The bending moment may be determined for any longitudinal section as follows.
For the end block shown, two cases can be considered:
1. For y < g:

41

G.P.Ancog

y 3
y2
h
- (3+
)
)
(
h
2e h)

M = Q ie 2(
M = y
(h
Q ie

2. For y > g:

y
M
= 2(
h
Q ie

y
h

h
( 3+ 2e
)

in dimensionless form and considers

clockwise moment as positive.

) - ( 3+ 2e )( h ) +

h y
h in dimensionless form and considers
(
1)
e h
2e clockwise moment as positive.

In practical cases, moments can be calculated at increments of height, starting

from at the bottom of the beam, and plotted as a function of distance from the
bottom. From this analysis, the total required area of steel reinforcement can be found:
Mmax
At =
fs( h - x )
where:

fs = the allowable stress in stirrups ( usually taken as 20,000 psi)

x = the distance in inches from the end face to the centroid of the steel
that are within h/2 from end
All other terms are as already defined previously.
End Anchorage Problem

240k

12

e=12 5
3.5

60k
3

12

60k

3
9

h=36

8. The figure shows the end portion of a post-tensioned beam. The beam has an
interior I-shaped section and solid rectangular end blocks. The arrangement of the
post-tensioning steel is as shown in the figure. There are six cables anchored at the
ends, exerting 60 k each on the beam. It is required to design the vertical
reinforcement for the end block.

End Block Anchorage Zone

Solution:
Moment at various horizontal planes can be calculated as follows: Q i = 360k, e = 12.5
in., h = 36 in.
For 0 < y < 3 in. :
42

G.P.Ancog

M = 360x12.5 2(

y
36

36
) - (3+ 2x12.5)(36 )

from which:
M = - 0.1929y3 + 15.417y2
When 3 < y < 9 :
M = - 0.1929y3 + 15.417y2 - 240y + 720
When 9 < y < 12 :
M = - 0.1929y3 + 15.417y2 -300y +1260
ad when y > 12 :
M = - 0.1929y3 + 15.417y2 - 360y + 1980

The table below shows the moment at various longitudinal sections:

y,
in.
1.0

M,
k-in
+15.2

2.0

+60.1

3.0

+133.5

4.0

-56.8

5.0

-118.7

6.0

-206.7

7.0

-270.7

8.0

-312.1

9.0

-333.1

10.
0
11.
0
12.
0
13.
0
14.
0

-391.2
-431.3
-453.3
-518.4
-567.7

y,
in.
19.
0
20.
0
21.
0
220
.
23.
0
24.
0
25.
0
26.
0
27.
0
28.
0
29.
0
30.
0
31.
0
32.
0

M,
k-in
-617.7
-596.5
-567.7
-532.3
-491.6
-446.6
-398.6
-348.7
-298.1
-247.9
-199.2
-153.3
-111.3
-74.3
43

G.P.Ancog

15.
0
16.
0
17.
0
18.
0

-602.3
-623.5
-632.2
-630.0

33.
0
34.
0
35.
0
36.
0

-43.5
-20.1
-5.2
0.0

We can assume that x = 8 in., approximately h/4, and calculate the area of end-zone
reinforcement:
Mmax
At =
fs( h - x )
632.2x1000
=
20,000( 36 - 8 )
= 1.1289 in2
Using #4 closed stirrups, Av = 2x0.20 = 0.40 in2 , total number needed:
n = At/Av = 1.1289/0.40= 2.82, use 3 closed stirrups.

Details of end zone reinforcement is shown below:

12

e=12 5
3.5

60k
3

12

60k

3
9

h=36

x=8

[email protected] = 33

240k
2

l=h=36

Anchorage Zone

Example Design of Prestressed Composite Section:

The top flange of a composite section is given as a slab 100 mm thick and 1.50
m wide cast in place. Design a precast section with a total depth of 900 mm (including
the slab thickness) to carry the following moments: M T = 435 kN-m, MG = 55 kN-m, MP
= 135 kN-m, MC = 300 kN-m. Allowable stresses are: f t = -11.00 MPa, fb = -12.40 MPa,
44

G.P.Ancog

ft = +2.00 MPa, fb = + 1.10 MPa. Initial prestress fso = 1030 MPa, effective prestress
fse = 860 MPa.
Solution:
To assume the section, make a preliminary design, assuming a lever arm of
0.65h for the prestressing force in resisting the total moment.
Q

MT
435 x10 3

743.589 kN
0.65h 0.65(900)

Compute Qo:

f so
1030
743.589
890.577 kN
f se
860
For the inverted T-section, the concrete area required can be approximate by:
Qo Q

Ac 1.5

Q
890.577 x10 3
1 .5
107731.08 mm 2
fb
12.40

From this preliminary section, sketch a trial section and proceed to final design.

in-place portion

1500
100

170.4
cgc
900

484.6

100
cgc

629.6
cgs 315.4
100
355

Beam Trial Section

Error: Reference source not found

For the precast portion, the section properties:

100 x 355 = 35500 x 50 = 1775000
700 x 100 = 70000 x 450 = 31500000
Ac = 105500
33275000 / 105500 = 315.40 mm = cb
35500(1002/12 + 265.42) = 2.530x109
45

G.P.Ancog

70000(7002/12 + 134.62) = 4.126x109

I
= 6.656x109 / 105500 = 63090.047 = r2
kt = r2/cb = 63090.047 / 315.40 = 200.031 mm
kb = r2/ct = 63090.047 / 484.60 = 130.189 mm
For the composite section, the section properties:
100 x 1500 = 150000 x 50.0 = 7500000
105500 x 584.6 = 61675300
Ac = 255500
69175300 / 255500 = 270.74 mm 100 =
170.74 = ct
150000(1002/12 + 220.742) =
7.4339x109
105500(0.00
+ 313.862) = 10.3926x109
I of precast
= _6.6560x109
I
= 24.4825x109
9
I ct
6.656 x10 484.6
mt

0.09578
I ' c t ' 24.4825 x10 9 170.74

mb

I cb
6.656 x10 9 315.4

0.5424
I ' c b ' 24.4825 x10 9 629.26

Step 1: Location of cgs.

e1

ft ' I
2.0(6.656 x10 9 )

30.845 mm
c t Qo 484.6(890.577 x10 3 )

MG
55 x10 6
e2

61.757 mm
Qo
890.577 x10 3
e e1 e 2 k b 30.845 61.757 130.189 222.791 mm
The cgs can be located at:
cgs c b e 315.40 222.791 92.609 mm from bottom fiber
Step 2: Compute mt, mb.
mt

I ct
6.656 x10 9 484.6

0.09578
I ' c t ' 24.4825 x10 9 170.74

mb

I cb
6.656 x10 9 315.4

0.5424
I ' c b ' 24.4825 x10 9 629.26

Step 3: Compute the required Q.

46

G.P.Ancog

M P mb M C f b ' k t Ac
e kt

135 x10 6 0.5424(300 x10 6 ) 1.10(200.031)(105500)

x10 3 649.224kN
222.791 200.031
f
1030
Qo Q so 649.224
777.56 kN
f se
860

For Qo = 777.56 kN instead of 890.577 kN, revise e1 and e2 as follows:

e1 e1 previous
e 2 e 2 previous

Qo previous
Qo new
Qo previous
Qo new

30.845

890.577
35.328 mm
777.56

61.757

890.577
70.733 mm
777.56

e e1 e 2 k b 35.328 70.733 130.189 236.25 mm

Which indicates that cgs can be located at:
cgs c b e 315.4 236.25 79.15 mm above bottom fiber
Revise Q as:
Q

M P mb M C f b ' k t Ac
e kt

135 x10 6 0.5424(300 x10 6 ) 1.10(200.031)(105500)

x10 3 629.196kN
236.25 200.031
f
1030
Qo Q so 629.196
753.572 kN
f se
860

Step 4: To keep bottom fiber within allowable stress fb.

Ac

1
fb

Qo

Qo e M G
kt

753.572 x10 3 (236.25) 55 x10 6

1
3
753.572 x10

12.40
200.031

110373.67 mm 2

To keep top fiber within allowable ft.

47

G.P.Ancog

Ac

1
ft

M P mt M C Qe

kb

135 x10 6 0.09578(300 x10 6 ) 629.196 x10 3 (236.25)

1
629.196 x10 3

11.00
130.189

67734.28 mm 2
Ac = 110373.67 mm2 controls and compare to Ac of trial section = 105500.00 mm2, a
little difference but considered acceptable and therefore there is no need to revise the
trial section.

Shapes of Concrete Sections

The simplest form is the rectangular shape possessed by all solid slabs and use
of some short span beam. As far as formwork is concerned, the rectangular beam is
the most economical. But the kern distances are small, and the available lever arm for
the steel is limited. Concrete near the centroidal axis and on the tension side is not
effective in resisting moment, especially at the ultimate stage of loading.
Other shapes are frequently used for prestressed concrete:
1.
2.
3.
4.
5.

The
The
The
The
The

symmetrical I-section
unsymmetrical I-section
T-section
inverted T-section
box section

The suitability of these section will depend on the particular requirements. The
I-section is has its concrete concentrated near the extreme fibers where it can most
effectively furnish the compressive force, both at transfer of prestress and under
working and ultimate loads. The more the concrete is concentrated near the extreme
fibers, the greater the will be the kern distances and the greater will be the lever arm
furnished for the internal resisting couple. However, this principle of concentrating the
concrete in the extreme fibers cannot be carried too far, because the width and
thickness of the flanges are governed by practical considerations, and for web must
have a minimum thickness to carry the shear, to avoid buckling and to permit proper
placement of concrete.
If the MG/MT ratio is sufficiently large, there is the danger of over-stressing the
flanges at transfer, and the concrete in the bottom flange can be accordingly
diminished. This will result in an unsymmetrical I-section, which when carried to the
fullest extent becomes a T-section. A T-section is similar to that for a reinforced beams,
is often most economical, since the concrete is concentrated at the top flange where it
48

G.P.Ancog

is most effective in supplying the compressive force. It may not be economically used,
however, where the MG/MT ratio is small, because the center of pressure at transfer
may lie below the kern point. Then tensile stresses may result in the top flange and
high compressive stress in the bottom section.
The unsymmetrical I-section with a bigger bottom flange, like a rail section is not
an economical one in carrying ultimate moment, since there is relatively little concrete
on the compression flange. However there is a great deal of material to resist initial
prestress. It can be economically used for certain composite sections, where the
tension flange is precast and the compression flange is poured in place. This section
requires very little girder moment to bring the center of pressure within the kern and
hence is suitable when the M G/MT ratio is small. When carried to the extreme, this
section becomes an inverted T-beam.
The box section has the same properties of I-section in resisting moments. IN
fact, their section properties are identical.

In summary:
1.
2.
3.
4.
5.
6.

For economy in steel and concrete, it is best to put the concrete near the
extreme fibers of the compression flange.
When MG/MT ratio is small, more concrete near the tension flange may be
necessary.
When MG/MT ratio is large, there is little danger of overstressing at transfer,
and concrete in the tension flange is required only to house the tendons
properly.
If formwork is used only once, it may constitute a major cost of the beam,
so that irregular shapes for the purpose of saving concrete or steel may
not be in the interest of overall economy.
When the forms can be reused repeatedly, more complicated shapes may
be justified.
For plants producing precast elements, it is often economical to construct
forms that can be easily modified to suit different spans and depths

Rectangular SectionSymmetrical I- Section

T and Inverted T- Sections

Unsymmetrical I- Sections

49

Box Sections

G.P.Ancog

Error: Reference source not foundArrangement

of Steel
The arrangement of steel is governed by a basic principle: in order to obtain the
maximum lever arm for the internal resisting moment, it must be placed as near the
tensile edge as possible. If the cgs is very near the tensile edge, and if there is no
sufficient girder moment to bring the center of pressure near or within the kern, the
tension flange may be overcompressed at transfer while the compression flange may
be under high tensile stress.
Tension

MG
T

lower kern

T
High
compression

Section

Stress Distribution

Error: Reference source not found

lower kern

MG
T1

C
MG + MS
T

C1

T1
Section

Partial Prestress
T1 for MG Only

Total Prestress
T for MG + MS

Pretensioning in two stages to keep C within kern.

A heavy moment is desirable at transfer so that the steel can be placed as near the edge as possible. However, no
economy is achieved by adding unnecessary dead load weight to the structure in order to enable a bigger lever arm for the steel,
because whatever additional moment capacity was thus obtained would be used in carrying the additional dead load, although
some additional reserve capacity is obtained at ultimate range. Lad that can be eventually carried by the beam can be more
economically put on the structure before transfer rather than after, because moments produced by such loads will permit the
placement of steel nearer the tensile edge.

Another method sometimes used in order to permit placement of steel near the
edge is to prestress the structure in two or more stages; this is known as retensioning.
At the first stage, when the moment on the beam is small, only a portion of the
prestress will be applied; the total prestress will be applied only when additional dead
load is placed on the beam producing heavier moment on the section. thus the center
of pressure can be kept within the kern at all times, and the excessive tension in the
compression flange, as well as high compression in the tension flange can be avoided.
50

G.P.Ancog

top kern

C
Small
arm

cgs

top kern
Big
arm

T
cgs

Steel in Tension Flange Only

Steel in Both Flanges

Prestressing steel in both flanges reduces lever arm for resisting moment.
Error: Reference source not found

For certain sections, the tendons are placed in the compression flange as well as
in the tension flange. Generally speaking, this is not an economical arrangement,
because it will move the cgs nearer to the cgc and thereby decrease the resisting lever
arm. At the ultimate range, tendons in the compressive flange will neutralize some of
its compressive capacity, whereas only those in the tension flange are effective in
resisting moment. However, under certain circumstances it may be necessary to put
tendons in both flanges in spite of the resulting disadvantages. These conditions are:
1.

When the member is to be subject to loads producing both +M and M in

the section.
When the member might be subject to unexpected moments of opposite
sign, during handling process.
When the MG/MT ratio is small and the tendons cannot be suitably group
near the kern point. Then the tendons will be placed in both the tension
and the compression flanges with the resulting cgs lying near the kern.

2.
3.

Design Example, Pre-tensioned Cored Section

A pre-tensioned cored section is to be used for roof construction carrying an
added dead load of 10 psf and a design live load of 30 psf on a simple span of 70 ft. It
is made of lightweight concrete at 110 pcf, precast in factory with a transfer strength
of fci = 4000 psi and a minimum 28-day strength of f c = 5000 psi, Ec = 2.5x106 psi.
Design the tendons, using 7/16 in. 7-wire strands with As = 0.1089 in2 per strand,
fs=fsu = 250,000 psi, fi = 175,000 psi, ft = 165,000 psi, and fe = 145,000 psi, Es =
27x106 psi. Check the loss of prestress, both immediately at transfer and eventually
say at the end of 3 years. Follow PCI Building Code requirements when applicable.
8.00
1.25
2 fillet

20
2.5

9.60
cgc

2.5
10.40

51
1.0

Dynacore Section

G.P.Ancog

Error: Reference source not found

Compute for bending moments:

wG L2 358 x70 2

219,000 ft lb
8
8
w L2 (8 x10) x 70 2
MS S

49,000 ft lb
8
8
w L L2 (8 x30) x 70 2
ML

147,000 ft lb
8
8
M D M G M S 219,000 49,000 268,000 ft lb
MG

M T M D M L 268,000 147,000 415,000 ft lb

Elastic Design Method:
Compute top and bottom kerns.
r2

I c 27100

57.5 in 2
Ac
470

r 2 57.5
kt

5.53 in
c b 10.4
kb

r 2 57.5

5.99 in
ct
9.6

M G 219,000

0.5277 l arg e ratio

M T 415,000
Since the girder load is relatively heavy, the cgs can be polaced as low as possible.
Assuming a clear concrete protection of 1.5 in for fire resistance and a protection of
cgs of 2.4 in, the total available from cgs to top kern:
h = cb cover + kt = 10.4 2.4 + 5.53 = 13.53 in.
This will give a resisting moment MR = Feh up to zero tension in bottom fiber.
6 f c ' 6 5000 423 psi
Allowing a maximum tension of
PCI Code, we have the
resisting moment from zero tension to 423 psi tension:
52

G.P.Ancog

fI 423x 27100 1

x 91,600 ft lb
c
10.4
12
This value in moment of 91,600 ft-lb may be found to be too high since its constitutes
22% of the total moment of 415,000 ft-lb. If fully utilized in design, it will appreciably
reduce the required amount of prestressing steel and hence the ultimate moment.
Furthermore, we will be using a great deal of tension in the bottom flange concrete,
which once cracked, may not be able to resist that tension. Hence it would be well to
use perhaps only half of that value in our first attempt to determine the required
prestress; thus:
M

h
1
91,600
Fe M T M 415,000
12
2
2
Fe 318,000 lbs
As

Fe 318,000

2.2 in 2
f e 145,000

7
No. of 16 in strands required = 2.20/A o = 2.20/0.1089 = 20.2 pcs. Using 20 strands,
we can check the stresses at transfer and under total load as follows.

Ft = 20(0.1089)(165,000) = 359,000 lb
Fe = 20(0.1089)(145,000) = 316,000 lb
e = cb = cover = 10.4 2.4 = 8.00 in.
Stresses at transfer:

Ft Ft ec M G c

Ac
Ic
Ic
359000 359000(8)(9.6 or 10.4) 219000 x12(9.6 or 10.4)

470
27100
27100

Top fiber:
fT = 763.829 1017.387 + 930.952 = 677.394 psi compression
Bottom fiber:
fB = 763.829 + 1102.169 1008.531 = 857.467 psi compression
These stresses indicate a near rectangular stress block, and a small amount of camber
under the girders own weight, which is usually desirable.
Under total design load:
53

G.P.Ancog

Fe Fe ec M T c

Ac
Ic
Ic
316000 316000(8)(9.6 or 10.4) 415000 x12(9.6 or 10.4)

470
27100
27100

Top fiber:
fT = 672.340 895.527 + 1764.132 = 1540.845 psi compression
Bottom fiber:
fB = 672.340 + 970.154 1911.143 = -268.649 psi tension
The top fiber compression is well within the allowable 0.45fc = 2250 psi, and the
bottom fiber tension is also less than the permissible 423 psi; this is expected when
choosing 20 7/16 in. strands.
To locate the strands at the end of the section, we assume that the strands are
spaced at a minimum of 1 in c.c. Depending on the hardware available and the
practice at the plant, this spacing may vary considerably. Further we will assume that
one point harping is preferred for simplicity of production (actually it frequently
happens that 2-point or 3-poinrt harping may be used, then the profile may be easily
arranged both for camber control and for stress control along the entire length of the
span), then we can make a tentative arrangement. IN order to produce no tension in
the top, the cgs at the ends should be within the kern. In order to produce a
rectangular stress block at the ends, the cgs should be at the cgc. In order not to have
the intermediate points control the design cgs should be located within a limiting zone.
The cgs at the end should also be located so as to produce the best camber effect.
Taking all this into consideration and using 4 strands for each web, a tentative
arrangement is shown below.

cgs

3 @1 3/4
=5 1/4
4

6 5/8
2 3/8
1 1/2 clear
cgs Location

4 @7/16 =1

Strand Location
Error: Reference source not found
Strand and cgs location
54

G.P.Ancog

The stress at the end can only be critical at transfer and will be computed as
follows, assuming no external moment.
f

Ft Ft ec

Ac
Ic

359000 359000(10.4 6.63)(9.6 or 10.4)

470
27100
763.829 479.443 284.386 psi compressio n, top fiber
763.829 519.397 1283.226 psi compressio n, bottom fiber

These stresses are well within the allowable range. Hence the design is
considered satisfactory so far as the flexural elastic stresses are considered.
Ultimate Strength:
Compute ultimate moment, PCI Code:
Mu1 = 1.8( D + L) = 1.8(415,000) = 747,000 ft-lb
Mu2 = 1.2D + 2.4L = 1.2(268,000) + 2.4(147,000) = 675,000 ft-lb
Mu = [Mu1, Mu2]max = [747000, 675000]max = 747,000 ft-lb
The expected actual ultimate resisting moment strength of the section may be
estimated as follows:
T ' NAo f su 20 x 0.1089(250,000) 545,000 lb
Ac '

Mu
545,000

128 in 2
.85 f c ' .85 x5,000

For width b = 8ft or 96 in and thickness of 1.25in, flange area = 120 in 2 with only
8 in2 to be additionally furnished by web fillets. Hence the center of compression can
be assumed at t/2 = 1.25/2 = 0.63 in. the lever arm at ultimate is approximately
computed as:
z = h cover t/2 = 20 2.4 0.63 = 16.97 in.
16.97
770,000 ft lb 747,000 ft lb
12
The 20-7/16 in. strands are just about right, while 19 strands would not have been
sufficient.
M ' T ' z 545,000

Camber and Deflection

Error: Reference source not found

Camber at transfer due to prestress only.

55

G.P.Ancog

Ec = 2,500,000 psi
Ft = (20x0.1089)(165,000) = 359,000 lb
M1 = Fty1 = 359,000(6.63 2.4) = 1,520,000 in-lb
M2 = Fty2 = 359,000(10.4 6.63) = 1,355,000 in-lb

L2

8EI

(70 x12) 2
2
2

M 2 3 M 1 8 x 2.5 x10 6 x 27100 1355000 3 1520000 3.09 in upward

Deflection due to beam own weight.

wL2
219,000 ft lb
8
(70 x12) 2
L2 5

Mg
(219000 x12) 2.86 in downward
6
8 EI 6
8 x 2.5 x10 27100 6

Hence the net camber at transfer, 1-day old is: 3.09 2.86 = 0.23 in, upward.
Mg

Depending on the age of concrete at which the additional dead load is placed,
the deflection will vary. If the dead load is added early, then after one year, the
amount of deflection or camber can be approximated by using an average value of E
= 3,500,000 psi, for 3,000 psi < f c < 6,000 psi and a creep factor of 1.80 for the
effects of prestress and dead load. Using the effective prestress of f c = 145,000 psi,
the deflection can be calculated as follows:
MD = 268,000 ft-lb
Fe = nAofse = (20x0.1089)(145,000) = 316,000 lb
M1 = Fey1 = 316,000(6.63 2.4) = 1,330,000 in-lb
M2 = Fey2 = 316,000(10.4 6.63) = 1,190,000 in-lb

L2
8EI

2
5

M D creep 1.8
2
1

3
6

(70 x12) 2
2
5

1190000 1330000 368000 x12 1.8

6

3
6
8 x 2.5 x10 x 27100

1.01 in downward

which indicates that, in the course of 1 year, the beam will have a maximum
downward deflection of 1.01 in., which probably will not be objectionable on a span of
70 ft. But it does indicate that either a more careful study of the camber and deflection
history is desirable, or that some means to reduce the deflection is needed, such as
placing the cgs at ends further below the present location.
The instantaneous deflection due to live load may be estimated using a higher
value of Ec =, say 3,800,000 psi. (immediate deflection).

56

G.P.Ancog

M L 147,000 ft lb
(70 x12) 2
L2 5

ML
(147000 x12) 1.26 in downward
6
8 EI 6
8 x3.8 x10 27100 6

1.26
1
which is

of the span length, not considered excessive

70 x12 667
Design for Shear: Shear is critical only under total load, hence,
VT

wL (358 80 240) x 70

23,700 lb
2
2

Vertical component from the tendons, for Fe = 316,000 lb, is:

y1
6.63 2.4
316,000
3200 lb
35 x12
L 2
Net shear, assuming the prestress is already transmitted to the concrete, is:
V s Fe tan Fe

Vc = VT Vs = 23,000 3,200 = 20,500 lb

Compute the Q-value for shear, use the top portion of section with respect to the cgc.
Top flange: 120 x 8.97 =
Fillet:
16 x 7.68 =
Web:
104 x 4.18 =
Q
=
V Q 20500(1633)
v c
99 psi
Ib
27100(5 x 2.5)

1075
125
435
1633 in3

For average prestress:

Fe 316000

672 psi
Ac
470
Principal tension is:
2

f
672
f
672
S t v c c 99 2
14 psi

2
2
2
2
which is very low. However, the end block stresses should be considered and a certain
amount of web reinforcement used near the ends.
2

Loss of Prestress
Elastic shortening at transfer: Es = 27x106 psi, Ec = 2.5x106 psi.
57

G.P.Ancog

f ct

Ft 359,000

765 psi
Ac
470

27 x10 6
f s nf ct
765 8250 psi
2.5 x10 6
The loss at the level of tendons at midspan may be slightly higher since the fiber
stress in concrete is somewhat higher. This also indicates that the estimated loss of
10,000 psi from the initial prestress of 175,000 psi to 165,000 psi at transfer is fairly
good estimation.

Loss of prestress due to creep may be estimated at 2 times the elastic shortening,
thus:
f s 2 x8250 16,500 psi
Loss of prestress due to shrinkage may be estimated assuming a coefficient of 0.0003,
at
f s s E s 0.0003(27 x10 6 ) 8,100 psi

Total loss of prestress, excluding loss or gain due to bending of members, is,
f st 8250 16500 8100 32850 psi
which is fairly close to the assumed loss of 30,000 psi in the problem statement.
Design Example, Post-tensioned Bridge Girder
Precast girder of a highway bridge are to be post-tensioned, grouted, then lifted
to the bridge site to be connected together by concrete poured in placed, as shown in
the figure below. The two-lane bridge is to carry H20-S16-44 loading, and the girders
are spaced 6 ft on centers. Overall length of the girder is 96 ft, with 95 ft between
centers of supports. Maximum live load moment for one lane, M max = 1433 k-ft,
maximum live load shear, Vmax = 61.3 k/lane. The dead load due to 2 bituminous
paving is 150 plf and the in-place slab and diaphragm is 133 plf. The weight of the
girder is taken equal to 940 plf. Following the ASSHO specs for Highway Bridges, when
applicable, design the interior girder as follows:
(a)
Design the midspan section, indicating the required amount of
prestressing steel.
(b)
Design the end section, showing the mild steel stirrups.
(c)
Design the longitudinal layout of girder showing the profile for the cgs and
the intermediate and end diaphragms.
(d)
Investigate the factor of safety of the girder for cracking and ultimate
strengths.
(e)
Compute the deflection of the girder at transfer and under the working
load.
(f)
Detail the midspan and the end sections using the Freyssinet system.
58

G.P.Ancog

Compute the loss of prestress due to friction and the initial prestress required at
the jack. Strength of concrete is to be 4,500 psi at 28 days and 4,000 psi at transfer.
The high tensile steel used is to have a minimum ultimate tensile strength of f s=
250,000 psi and a minimum yield strength of f y = 200,000 psi at 0.2% plastic set. The
steel stress at transfer will be 165,000 psi and the effective prestress at 25,000 psi
loss is 135,000 psi. Es = 28,000,000 psi, Ec = 4,000,000 psi. Use intermediate-grade
reinforcing bars for the mild-steel reinforcement, fy = 40,000 psi.
48.0
6

5.0

24.0

18.5

Half Elevation

CL

4.5
2 bitumen surface

6.0 cc of girders

Precast diaphragm Diaphragm

Precast diaphragm Diaphragm

Midspan Section
Error: Reference source not foundSolution

a)
Maximum live load for 1 lane: Mmax = 1433 k-ft
For the interior girder spacing of 6 ft and a lane width of 10 ft:
6.0
M LL 1433
860 k ft
10
50
50
I

0.227 0.30, ok
L 125 95 125
Impact factor:
Impact moment:
M IL M LL ( I ) 860(0.227) 195 k ft
Total live load plus impact: M LL M IL 860 195 1055 k ft

M LL M max

S
10

Error: Reference source not found

59

G.P.Ancog

Compute dead load moments:

Added dead load:
Bituminous paving
In-place slab
Total =
Weight of girder =

=
150 plf
=
133 plf
283 plf
940 plf

1 2 1
wL (.283)95 2 319 k ft
8
8
Added dead load moment:
1
1
M G wG L2 (.940)95 2 1060 k ft
8
8
Girder moment:
Total moment: M T M G M D M LL M IL 2434 k ft
MD

Section properties of the trial section which resulted from preliminary design and
trials.
6
3
38
7
22

x
x
x
x
x

54
3
8
7
8

=
=
=
=
=

324
9
304
48
176
862

972
63
7600
2040
8450
19125 / 862 = 22.5 in = ct
52 22.5 = 29.5 in = cb
The moment of inertia of the concrete section about cgc.
324(62/12 +
9(32/18
+
2
304(38 /12 +
49(72/18
+
2
176(8 /12 +

x
x
x
x
x

3
=
7
=
25
=
41.67 =
48
=

19.52)
15.52)
2.52)
19.172)
25.52)

=
=
=
=
=
60

124000
2200
38600
18000
115000
297800 / 862 = 345 = r2
G.P.Ancog

kt = r2/cb = 345 / 29.5 = 11.7 in.

kb = r2/ct = 345 / 22.5 = 15.4 in.

M G 1060

0.435, relatively l arg e

M T 2434
, the cgs can be located as low as
practicable without producing any tension in the top fiber.
Assume cgs is located 4 above the bottom fiber, compute the total arm for the
internal resisting moment.
a k tt c b cov er 11.7 29.5 4 37.2 in
Total effective prestress required:

M T 2434 x12

786 k
a
37.2
25000
.1515 15%
For a loss of prestress of 165000
, the initial prestress required:
F

Fo

F
786

925 k
.85 .85

AASHTO specs for post tensioned members:

Allowable stress at transfer: f b .55 f c ' .55(4000) 2200 psi
f .40 f c ' at 28 days .40( 4500) 1800 psi
Allowable stress at final stage: t
Required concrete area, at final stage of loading:
Ac

Fh
786 x52

770 in 2
f t c b 1.8 x 29.5

Required concrete area at transfer:

Ac

Fo
fb

e M G Fo 925
25.5 1060 x12 / 925
2
1

841 in
kt
2
.
2
11
.
7

Actual gross area provided is Ag = 862 in2, which seem s to be just about
sufficient for the required area of Ac = 841 in2.
It generally takes two or three trials to arrive at this adopted section rather than
just one trial as illustrated here.
61

G.P.Ancog

To supply the effective prestress of 786 k at an allowable stress of 135 ksi, the
required steel area is:
F 786
As

5.83 in 2
f s 135
b)
Shearing stresses can be checked for two sections, one at the support and
another at 5 from the support where the web is 8 thick. At the support the web is 22
thick; shear is evidently not controlling. At 5 from the support:
VLL = 61.3 k/lane
Vlane

S
6
61.3 36.8 k / girder
10
10

VLL for girder =

Impact : V IL V LL ( I ) 36.8(0.227) 8.4 k

Bituminous paving + in-place concrete: wL / 2 .283(42.5) 12.0 k

Girder weight: wG L / 2 .940(42.5) 40.0 k
Total shear: VT = 36.8 + 8.4 + 12.0 + 40.0 = 97.2 k
Determine the shear carried by tendons assuming a parabolic profile with rise h
computed as follows:
h = e = cb cover = 29.5 4 = 25.5 in.
Provide some eccentricity at support such that h 24 in.( 2 ft)
L 8 Fh L 4 Fh 4(786)( 2)
2

65.5 k
2
L
96
L 2
At 5.5 ft from end of girder:
Vs w

Vs

x
42.5
V max
65.5 58.0
L2
48

Hence Vc of concrete:
Vc 97.2 58.0 39.2 k
Compute the value of Q:
324
9
8

x
x
x

19.5 =
15.5 =
16.52/2=
Q
=

6318
139.5
1089
7546.5 in3

Maximum shearing stress occurs at the cgc:

62

G.P.Ancog

Vc Q 39.2 x1000(7546)

124 psi
Ib
297800(8)

Compressive fiber stress at cgc:

fc

F 786 x1000

912 psi
Ac
862

The principal tensile stress is:

s t v 2 f c 2 f c 2
2

124 2 456 2 456 20 psi

Allowable value of principal tension:

s 0.03 f c ' 0.03(4500) 135 psi s t 20 psi
No stirrups are needed under working load.
Investigate the ultimate strength for shear, AASHTO specs:
V ' 1.5D 2.5( L I )
1.5(52.0) 2.5(36.8 8.4) 191 k

Shear carried by the tendons:

Vs = 58.0 k
Hence:
Vu V 'V s ' 191 58 133 k
Shear carried by concrete:
Vc 0.02 f c ' bjd .02(4500)(8)(7 8)(52) x10 3 32.76 k

Stirrups required:
Av

1 Vu V c s 1 133 32.76 x10 3 s

0.027538s in 2
2 f y ' jd
2 40000(7 8)(52)

Using in U-stirrups, Av = 0.40 in2:

63

G.P.Ancog

Av
0.40

14.525 say 14.50 in

0.027538 .027538

Similar computations can be made for other points along the girder. So far as
shear is concerned, more stirrups are required near the ends than along the middle
portion of the girder, but the reverse is true when considering the effect of combined
moment and shear. Hence judgment should be exercised in the actual spacing of the
stirrups.
Minimum web reinforcements, AASHTO specs:
Av 0.0025b' s
For b = 8 in, Av = 0.40 in2:

Av
0.40

20 in
0.0025(8) 0.0025(8)
For the end section, stirrups are required to distribute the anchorage stresses.
Since the anchorage are fairly uniformly distributed, the computed tensile stresses in
the anchorage zone will be low and analysis is not required. Normal stirrups, however,
are provided.
s

c)
A half elevation of the girder is shown in the sketch layout. The midspan section
is adopted for the entire girder, except 5 ft near the ends where a uniform web
thickness equal to the bottom flange width of 22 in. is used in order to accommodate
the end anchorages, to permit the curving up of some tendons, and to distribute the
prestress. Three intermediate diaphragms are placed along the length of the span.
Sometimes transverse prestressing is employed to bind the girders together. But for
this design, transverse dowels are provided in these diaphragms to be joined together
by in-placed concrete. The amount of steel is not excessive for these diaphragms;
some nominal reinforcements are employed as shown on the sketch.
The most common location of cgs for a simple span is a parabola with cgs near
the cgc at the ends. Such a profile will give ample moment resistance along the entire
beam. If cgs is above the cgc at the ends, the tendons will carry greater shear but lose
some of the reserve moment resistance. If cgs is below the cgc at the ends, the
tendons will carry less shear, but the positive prestressing moment at the ends will
tend to decrease the principal tension. Also note that the cgcs for the midspan and
the end sections actually differ slightly. For this design, the cgs will be placed a little
below the cgc of the end section.
d)
Computation of cracking moment.
Resisting moment up to zero stress in the bottom fiber:
64

G.P.Ancog

M r Fa 786

37.2
2434 k ft
12

Modulus of rupture taken equal to f 0.14 f c ' 0.14(4500) 630 psi

M

fI 630(297800)
1

530 k ft
cb
29.5
12 x1000

Total resisting moment at cracking:

M cr M R M 2434 530 2964 ft
Overall factor of safety against cracking:

F .S .

M cr 2964

1.22
M T 2434

Factor of safety for live load and impact:

F .S .

M cr M G M D 2964 (1060 319)

1.50
M LL M IL
860 195

This indicates that the girder will begin to crack only when the live load plus
impact is increased by as much as 50%.
The ultimate resisting moment computation:
Assuming that the ultimate strength of steel to be developed:
T f s ' As 250000(5.83) x10 3 1460 k
Average compressive stress in concrete 0.85fc = 0.85(4500) = 3.82 ksi

4.5
6
y
3 fillet
Precast Diaphragm

Precast Diaphragm

3-2

4-4

1:1
8
1-10

Section

65

G.P.Ancog

C T
.85 f c ' A T
T
1460

382 in 2
.85 f c ' 3.82
6
x
54
=
324
3
x
3
=
9
8
x
c
=
8c
Total =
333+8c = 382; c = 6.125 in.
y = 6 + c = 12.125 in. from top (N. A. location)
A

The cg of compression zone:

324
9
49

x
x
x

3
=
7
=
9.0625 =
Total =

972
63
444.0625
1479.0625 / 382 = 3.87 in. cg below top fiber.

The ultimate lever arm:

z h cov er cg 52 4 3.87 44.13 in.

The ultimate moment Mu:

44.13
5369.15 k ft
12
M u M u .90(5369.15) 4832.24 k ft
Factored moment:
M u Tz 1460

M u 1.5 D 2.5( LL IL ) 1.5(1379) 2.5(1055) 4706 k ft

Hence the ultimate strength is considered sufficient.
e)
Deflection Computation
1.5
M 1 Fe1 925
115.63 k ft
12
M 2 Fy 925(2) 1850 k ft
M D 1060 k ft
Instantaneous deflection due to prestress, upward deflection:

66

G.P.Ancog

M 1 L2 5M 2 L2

8 EI
48 EI
2
3
3
115.63 5(1850) 96 (12) x10

2.78 in
8
48 4 x10 6 (297800)

cgc
e1 =1.5
cgs

y = 2

parabolic

F = 925 k

Eccentricity of Prestress

-115.63 k
-1850 k
parabolic

Moment Due to Prestress

parabolic
+1060 k

Moment Due MG
Diagram for Deflection Computation

Error: Reference source not foundOwing

to loss of prestress of 15%:

f se
135000
p
2.78 2.345 in.
f so
160000
Downward deflection due to MG:

pe

5M G L2 5(1060) 96 2 (12) 3 x10 3

G

1.47 in.
48 EI
48 4 x10 6 ( 297800)

The immediate deflection at transfer:

i p G 2.78 1.47 1.31 in.

Added dead load deflection due to in-place concrete:
67

G.P.Ancog

MD
319
G
1.47 0.44 in.
MG
1060
Long term deflection, assume creep coefficient Cc = 2.0, and loss of prestress:

f pe ( G D )C c 2.345 (1.47 0.44) 2.00 0.87 in., upward

Instantaneous downward deflection due to design live load and impact,
assuming a parabolic moment diagram:

M L I
1055
G
1.47 1.46 in.
MG
1060

f)
The above design will be applicable to most prestressing systems now in use
today, although minor modifications may be desirable for certain cases. For purpose of
illustration, detailed arrangement of the tendons is shown for the Freyssinet system.
Using cables of 18 - 0.196 in. wires, 11 tendons are required, Ao = 0.543 in2:
Steel area as required:

Ast = 5.83 in2

Steel area provided:

Ast = 11(0.543) = 5.973 in2

10

11

6
8

24.2
cgs

27.8

8
8

4- 4

8
10 11

cgs
4.2

4 1 2 3
8 5 9 6 7

3
3
2

End Section

End Elevation

Midspan Section

Cable Location For Freyssinet System

The midspan and end sections are drawn showing arrangement of tendons to
give the required locations of cgs. Curving the tendons in both horizontal and vertical
planes is necessary to conform with the required location of the cgs. It is noted that
that some deviation from the required parabola is permissible, because it will not
affect the strength of the girder. A recommended order for tensioning the cables is
indicated as shown.
Loss for anchorage slip in Freyssinet cones may be assumed to average 0.2 in.,
which if average throughout the entire length of 96 ft, indicates a loss of prestress
equal to:
68

G.P.Ancog

0.2
Es
28 x10 6 4900 psi
L
96 x12

To estimate the frictional loss, assume a coefficient of friction = 0.35 for

Freyssinet cables in metal sheathing and a K = 0.0010 per ft for wobble effect.. the
average rate of change in direction for the parabolic cable is given by:

8h 8(2)

0.167 radian
L
96
The maximum frictional loss at the far end, if tensioned only from one end is:

F2 F1 e ( KL )
e ( KL )
KL
% loss ( KL)100%
KL .35(.167) 0.0010(96) 0.058 0.096 0.154 or 15.4%
Ave. Fa F2

The controlling point is the midspan which has a loss equal to of total 15.4% =
7.7%.

f f %loss ( f so ) 7.7%(160000) 12,300 psi.

If the tendons are overstressed by 12,300 psi at the anchorages, the anchorage
loss of 4,900 psi be balanced. Using the Freyssinet jack, there is an additional loss at
f j
the jack of about
= 8,000 psi. Therefore the maximum initial stress at the jack
should be:

f i f so f f f j 160000 12300 8000 180,300 psi

According to AASHTO specs, the temporary jacking stress:
48.0

f j 5.00.80 f s ' 0.80(18.5

250000) 200,000 psi

Girder Layout
6

24.0

The
prestress should not exceed 0.60f s = 0.60(250000)=150,000 psi,
2 effective
0

4
or 0.80f2

=
0.80(200000)
= 160,000 psi. Hence the stress in the steel is considered
y
CL
safe in all conditions.
r @ 20
8@6
8@12
symmetrical

Half Elevation

Girder Layout

4.5

1 6

6.0 cc of girders

2 bitumen surface

6
Precast girder
Continuous key
4-4

3-2

Precast diaphragm
Precast diaphragm
1:169
cgs
1 10

Midspan Section

G.P.Ancog
8

Continuous Beams
Continuity, Pros and Cons

wc

Continuous Beam

Simple Beam
w

wc

a
V=0

a
V

L/2

L/2

FBD of Half Span

V=0

C
T

FBD of Half Span

w' L2
T ' a'
8

wc ' L2
2T ' a '
8
Moment Diagram
Continuous Span

Moment Diagram
Simple Span

Load-Carrying Capacity Of
A Continuous Beam

Load-Carrying Capacity Of
A Simple Beam

Let:
T = ultimate tension developed in the tendons
a = lever arm
w = ultimate load produced by tendons
Simple beams:
70

G.P.Ancog

w' L2
T ' a'
8
8T ' a '
w' 2
L
Continuous beams:
wc ' L2
2T ' a '
8
16T ' a '
wc '
L2
The ultimate load w carried by the simple beam is controlled by the capacity of
the midspan section and cannot be increased by any change in the end eccentricities
of the cgs.
For the continuous beam, there are two resisting moments, one at midspan and
another over the support. Hence the load-carrying-capacity is definitely affected by
the position of the cgs over the intermediate support. w c = 2w, this means that twice
the load on the simple beam span can be carried by the continuous span for the same
amount of concrete and steel. This represents a very significant basic economy that
should be realized by engineers designing prestressed concrete structures. Because of
this strength inherent in continuous construction, it is possible to employ smaller
sections for the same load and span, thus reducing the dead weight of the structure
and attaining all the resulting economies.
Although it is generally conceded that continuity is economical in reinforced
concrete, it is seldom known that, from certain point of view, even greater economy
can be attained in prestressed construction.
Economical design of continuous prestressed beams can be achieved in several
ways. Owing to the variation of moment along the beam, the concrete section and the
amount of steel are often varied accordingly. The peak of the negative moments can
be reinforced with non-prestressed steel, thus reducing the amount of prestressing
steel. Advantage can be taken of the redundant reactions to obtain favorable lines of
pressure in the concrete. Design can be based on the ultimate strength of such beams,
applying the principles of limit design.
Like any other type of construction, there are advantages and also
shortcomings, which, under certain conditions, could outweigh the advantages. The
choice of a particular type of design must be made after considering all the factors
involved in the job.
Disadvantages inherent in continuous prestressed concrete beams can be
enumerated as follows:

71

G.P.Ancog

1.

2.

3.

4.

5.

6.

7.

8.

Frictional loss in continuous tendons. This can be serious if

there are many reversed curves, if the curves possess large
deflection angles, or if the tendons are excessively long. Such loss
can be minimized by using relatively straight cables in undulating or
haunched beams. The usual method of overtensioning, of stressing
from both ends, can also be used to reduce frictional losses.
Shortening of long continuous beams under prestress. This
may produce excessive lateral force and moments in the
supportingcolumns, if they are rigidly connected to the beams
during prestressing. Provisions are usually made to permit
movement at the beam bearings or rocking of the columns.
Secondary stresses. Secondary stresses due to prestressing,
creep and shrinkage effects, temperature changes, and settlements
of supports could be serious for continuous structures unless they
are controlled or allowed for in the design. One interesting point in
continuous prestressed structures is that these secondary stresses
can often be utilized to good advantage so that they will add to the
economy of the structure.
Concurrence of maximum moment and shear over supports.
It is believed that the concurrence of maximum moment and shear
at the same section may decrease the ultimate capacity of the
beam. This happens over the support of most continuous beams.
Hence care must be taken to reinforce such points properly for both
shear and moment if high ultimate strength is desired. The elastic
strength, however, is not affected by such concurrence.
Reversal of moments. If live loads are much heavier than dead
load, and if partial loadings on the spans are considered, continuous
beams can be subjected to serious reversal of moments. This can be
sometimes be overcome by proper design, such extensive used of
nonprestressed steel in combination with prestressed concrete.
Moment peaks. Peaks of maximum negative moments may
sometimes control the number of tendons required for the entire
length of the beam. These peaks, however, can be strengthened by
employing deeper sections or by adding prestressed and prestressed
reinforcements over the portions where they are needed.
Difficulty in achieving continuity for precast elements. It is
easy and natural to obtain continuity for cast-in-place construction,
but continuity of precast elements cannot always be achieved
without special effort. On account of difficulties in handling precast
continuous beams, they are often precast as simple elements, to be
made continuous after they are erected in place.
Difficulty in designing. It is more difficult to design continuous
rather than simple structures. But with the development of simpler
methods, the design of continuous prestressed concrete beams can
be made into a more or less routine procedure applying basic
principles for continuous structures familiar to most engineers

72

G.P.Ancog

Tendon Layouts for Continuous Beams

Continuous beams may be divided into two classes: fully continuous beams and partilly continuous beams. For full
continuity, all the tendons are prestressed in place and are generally continuous from one end to the other., although some can be
anchored at intermediate points if found desirable. The concrete can be either poured in place or made of blocks assembled on
falsework The tendons may be encased in the concrete during pouring, threaded through preformed holes, or placed outside the
webs. They may be either bonded or unbonded, depending on the requirements of the structures.

Precast elements can also be made fully continuous by coupling the tendons
together with a high tensile rod and then stressing one or both ends of the tendons.
Fully continuous beams tendon layouts:

Curve tendons with straight beam, often used for slabs or short span beams, where simple formworkis more important than saving of steel and concrete. The main object

Haunched or curve beams, often used for longer spans or heavier loads, this will not only save concrete and steel but permit the use of straight tendons, likewise position

Compromise of the above two arrangements, using curve beams and slightly curve tendons. This would permit optimum depth of beam as well as ideal position of steel

Overlapping tendons. Cables protruding at intermediate points offer a possibility of varying prestressing force along the beam. The arrangement has no reversed curves i

73

G.P.Ancog

Partially continuous beam tendon layouts:

Continuous prestressed cables placed in conduits or grooves left in the structure. After erection, concrete is poured between the beams over the supports

Poured-in-place concrete Tendons stressed after erection

Tendons stressed before erection

Continuous tendons placed over the supports only. This saves steel but requires more anchorages than first layout. More of the anchorages are placed at

Poured-in-place concrete Cap cables placed and stressed after erection

Tendons stressed before erection

Supplying continuity using cap cables. These tendons are usually made of wires or small strands, can be conveniently stressed from the soffit of the bea

Continuous elements over the supports transversely prestressed. Tensile elements are are inserted over the supports and attach them to the precast beam

74

G.P.Ancog

Temporary anchorage

Beam with tendons prestrressed

Tend

coupler

Poured-in-place concre

Coupler over supports. This is specially applicable to high-tensile bars, but also to other forms of tendons to obtain continuity. This permits the stressing

75

G.P.Ancog

In-place concrete

nonprestressed steel

Moment in concrete due to prestressing in a simple beam:

Nonprestressed steel over the supports. Continuity can be achieved by employing nonprestressed over the supports. This is specially true for composite

Beam Elevation
e

T
C
M = Te = Ce

R=0
FBD of Section

Beam Elevation

For simple beam, no matter how the beam is prestressed, only the internal stresses
will be affected by prestressing. The external reactions, being determined by statics,
will depend on the dead load and live load (including the weight of the beam). Without
load on the beam, no matter how the beam is prestress internally, the external
reactions will be zero, hence the external moment will be zero. With no external
moment on the beam, the internal resisting moment must be zero, hence the C-line
beam under prestress
not held bywith
supports.
(line of pressure Bending
in the ofconcrete)
must if
coincide
the T-line in the steel (which is the
cgs line). The C-line in the concrete being known, the moment in the concrete at any
section can be determined by:

M Te Ce
Moment of concrete due to prestressing in a continuous beam:
Reactions exerted to hold the beam in place.
M

Moment diagram due to reactions.

C-line

cgs line
or T-line

a=
76

M
T

Deviation of C-line from cgs line due to moment produce by secondary reactions.

G.P.Ancog

When the continuous beam is prestressed, it bends and deflects. The bending of
the beam can be such that the beam will tend to deflect itself away from the suuports.
If the beam is refrained from deflecting at these supports, reactions must be exerted
on the beam to hold the it there. Thus reactions are induced when a continuous beam
is prestressed. These induced reactions produce moments in the beam. To resist these
moments, the C-line must be at a distance a from the T-line, such that the internal
resisting moment equals the external moment M caused by the induced reactions.
a

M
T

In simple beam, the C-line coincides with the T-line. The stress distribution in the
concrete at any section is given by the location of the T-line. No secondary moment is
created. In continuous beams, the C-line usually deviates from the T-line. The stress
distribution at any section is given by the C-line which does not coincide with the tline. The difference between the two beams lies in the presence of external reactions
and moments in the continuous beam, produced as a result of prestressing. Since the
external moment is soley produced of external reactions, and since the reactions are
only applied at the supports, the variation of moment between the supports is a linear
one.. If T remains constant between supports, then the deviation a, being directly
proportional to M, also has to vary linearly.
77

G.P.Ancog

The difference between the simple and a continuous beam under prestress can
be represented by the existence of secondary moments. Once these moments over
the supports are determined, they can be interpolated for any point along the beam.
Moment in the concrete given by the eccentricity of the prestressed is
designated as the primary moment, such as would exist if the beam were simple. On
account of the primary moment acting on a continuous beam, the secondary moments
caused by the induced reactions can be computed. The resulting moment due to
prestress, then is the algebraic sum of the primary moment and secondary moment.
Secondary moment + Primary moment = Resulting moment
Assumptions for the Method of Analysis and Design Adopted:
1.
2.
3.
4.
5.

The eccentricities of the prestressing cables are small compared to the

length of the members.
Frictional loss of prestress is negligible (where frictional loss is appreciable,
it should be taken into account).
The same tendons run through the entire length of the member, (varying
steel areas can be included with some modifications).
The axial component ofC-line
the prestress is constant for the member and is
equal to the prestressing force F.
The primary moment M1 at any section in the concrete is given by:
e1

M 1 Fe1

e2 =

M2
F

cgs-line

(a) Beam Elevation

Procedure of Analysis:
1.
2.
3.

M1 = Fe1

First treat the entire beam as if it had no supports. Plot the moment
diagram for the concrete produced by the eccentricity of the prestress.
Compute the(b) Primary
loading
on Diagram
the beam
ccorresponding to that moment
Moment
Due To Prestress
diagram. This is the loading produced by the steel on the concrete.
With this loading acting on the continuous beam as it is actually
supported, compute the resulting moment by moment distribution or other
similar method.
(c) Shear Diagram For (b)

(d) Load Diagram For (c)

M2

78

(e) Resultiing Moment Diagram Due To Prestress, From Loading In (c)

G.P.Ancog

Since the C-line deviates linearly from the cgs-line, it will have the same intrinsic
shape as the cgs-line, and can be plotted. It is usually not necessary to compute the
secondary moment, which is represented by the deviation from the C-line and the cgsline. If desired, it can be computed by the simple relation:
Secondary moment = M2 M1

M 2 M1
F

Hints for plotting the loading diagrams directly from the moment diagrams:
1.

At the end of the tendons, the force F from the tendons on the concrete
can be resolved into 3 components:
A.
An axial force, Fcos1 = F (since cos 1 =1), acting at the end of the
anchorage. This is usually has no effect on the bending moment but
may produce moment in a rigid frame, owing to the axial shortening
effect.
B.
A transverse force Fsin1 = F1 = Ftan1, applied at the support and
balanced by the vertical reaction from the support directly beneath.
This again produces no moment in a continuous beam, unless it is
applied away from the support. Its effect in a rigid frame is small.
79

G.P.Ancog

F5
Forces From Tendons On Concrete

C.

2.

F4

A moment, Fcos1e = Fe, acting at the end of the beam. This will
produce moment along the entireFlength
of the continuous beam
3
and it must be included when following the moment distribution
procedure.
F2

Along the span of the member, where the cgs or the cgc-line of the
member curves, transverse loads are applied to the concrete. Two
common cases can be considered:
A.
When the moment diagram takes the shape of a parabolic or a
circular curve (note: owing to the assumption of a flat curvature,
parabolic and circular curves are considered to have the same effect
in producing transverse loads), a uniformly distributed load is
applied to the concrete along the length of the curve. The total force
on each curve is given by the change in slope between the two end
tangents; thus the total force at 2 is given by:

W F sin 2 F 2

B.

3.

For practical purposes, the load W can be considered as uniformly

distributed along the length of the curve.
When the moment diagram changes direction sharply, the force can
be considered as concentrated at one point; the moment at f4, for
example is:

F sin 4 F 4
Over the interior supports, where the moment diagram changes direction,
a load is applied directly over these supports. Again two cases can be
considered:
A.
If the moment diagram curves gradually over the support, again a
uniformly distributed load is applied as shown for 3. This will affect
moments in the beam, and the load must be considered in
performing the moment distribution.
B.
If the moment diagram is bent abruptly over the supports, a
concentrated load is applied thereon. Such a concentrated load is
directly reacted by the support underneath and produces no
moments on the beam. It can be neglected in performing the
moment distribution.

Obtaining Loading Diagrams Due To Prestress:

80

G.P.Ancog

F = 250k
0.4

cgc-line
0.2

0.8

cgs-line

0.9

0.08

30

0.176 rad

D
20

250(-.20) = -50.00 k
250(-.80) =-200k
250(+.40) = +100k
250(-.90) = -225k

25

25

a) Beam Elevation
+100k
0.00

-50k
-225k

-200k

b) Primary Moment Diagram Due To Prestress

Mb = Ma + Av
Av = aL = Mb - Ma
a = [ -200 (-50)]/30 = -5

+15k

-5k
a = [ +100 (-200)]/20 = +15
[0 100]/25
= -4
[-225 0]2/25 = -18
-22
+[0 100]/50 = - 2

+20k

-24k

c) Shear Diagram From b)

-24
[0 (-225)]2/25 = 18
-[0 100]/50
= 2
20

+20k

Vb = Va + Aw
P1 = [-5 0]
= -5
P2 = [15 (-5)] = 20
P3 = [-24 (15)] = -39
w = [20-(-24)]/50 = 0.88 k/ft
P4 = [0-(20)]
= -20

Example 1:-5k

-20k
-39k
FEM:
Concentrated load:
d) Loading Diagram From c)
A continuous
prestressed
concrete beam withFEM:
the load:
bonded tendons is shown in
MAB = (Pab2)/L2
Uniform
MAB = 20(30)(202)/502
= +96 eccentricity at A, is bent
MAB
= - MBA = wL2/12
the figure. The
cgs has an
sharply
at D and B, and has a
MBA = (Pa2b)/L2
MAB = -MBA = 0.88(502)/12
parabolic curve
the span
MBA = for
20(302)(20)/502
= -144 BC. Locate the line of pressure
= 183 (the C-line) in the concrete

due to prestress alone, not considering the dead load of the beam. Consider a
rel. k = 1
rel. k = 1
prestress of 250 k.
FEM
Dist
CO
Dist
Bal

-50

+96
-96

-50

-144

+183

-73
-29
-246

+92
-29
+246

-183
+183

0.00

e) Moment Distribution For Loading In d)

0.98

cgc-line
0.2
C-line

0.45

81

0.61

d) C-line Due To Prestress From a) and e)

G.P.Ancog

The line of pressure for the entire beam can be computed by plotting its moment
diagram and dividing the ordinates by the value of the prestress.

82

G.P.Ancog

50
0.20'
250
246
eB
0.98
250
eA

If desired the secondary moment over the center support can be computed as:
M sec ondary M resul tan t M primary
M sec 246 100 146 k ft
Example 2:
For the prestressed beam in Ex. 1, a uniform load of 1.2 k/ft is applied to the entire
length of the two spans (including weight of the beam itself). Locate the line of
pressure in the concrete due to the combined action of the prestress and external
load. Compute the stress in concrete section at B, I = 39,700 in4, Ac = 288 in2.
w = 1.2 k/ft

12
6
cgc-line
24

cgs-line

50

Concrete Section
FEM

-250
Dist +250
CO
____
0.00

50

+250

-250

+125
+375

-125
+375

-250
+250
____
0.00
+211

+211

M-diag.
Due to DL

-375
C-line

C-line.
Due to MDL

0.74

0.54
1.50

C-line

Resulting C-line.
due to combined MP+MDL

0.09

-.52

0.13

By dividing the moments by the prestress F, the shifting of the C-line can be
obtained due to MDL.
83

G.P.Ancog

Computing the stresses:

At B: MB = FeB = 250(-.52) = -130 k
Top fiber:

F Mc
250 130 x12(18)

A
I
288
39700
0.867 0.707 0.160 ksi

fT

Bottom fiber:

F Mc
250 130 x12(18)

A
I
288
39700
0.867 0.707 1.574 ksi

fB

Note:
The prestress moments and the dead load moments can be combined together
and then make the moment distribution. The results would obviously be the same.
Linear Transformation and Concordancy of Cables
The design of continuous prestress beams is more complicated problem. In
analysis, the concrete section, the steel, and the location of the steel are already
known or assumed. It is only necessary to compute the stresses for the given loading
conditions. This is not true in design, which is essentially a trial-and-error process in an
effort to reach the best proportions. The designer must be well acquainted with the
method of analysis before he can perform efficiently in design. IN order to design well,
we must be conversant with some of the mechanics of continuous prestressed beams.
Linear Transformation
The designer should be able to perform linear transformation with ease and skill
and to obtain either concordant or nonconcordant cables to satisfy the most desirable
conditions.
When the position of cgs-line or of a C-line is moved over the interior supports of
a continuous beam without changing the intrinsic shape (i.e. the curvature and bends)
of the line within each individual span, the line is said to be linearly transformed.
Explanation:
1.
2.
3.

First, the position of the line is moved only over the interior supports
whenever desired, but not at the ends of beam.
Strictly speaking, a line can still be termed linearly transformed if it is
moved at the ends.
For purpose of design, linear transformation without involving movement
st the ends is much more useful.
84

G.P.Ancog

4.

BY linear transformation, the intrinsic shape of the line within each span
remains unchanged; only the amount of bending of the line over the
interior supports changed.
The C-line resulting from perstressing a continuous beam is linearly
transformed line from the cgs line.

5.

2nd cgs line

0.6

0.3
0.4
0.5

F = 200 k

0.8
0.2

1.0

1.0

0.2

1st cgs line

25
50

50

50

a) Linear Transformation of cgs Line

160k

M for 2nd cgs line

80k

140k
20k

200

-40k

200

-60k
-160k

M for 1st cgs line

b) Primary Moment due to Prestress

14.4 k
80 k-ft

0.64 k/ft

c) Loading diagram For Both Moment Diagrams

Theorem:
In a continuous beam, any cgs line can be linearly transformed without changing
the position of the resulting C-line.
This means that the linear transformation of cgs line does not affect the stresses
in the concrete, since the C-line remains unchanged. Thus the two cgs line in the
above figure will produce the same stresses in the concrete, despite their apparently
divergent locations.
Any bending of cgs over the support will produce transverse forces acting on the
beam which are directly counteracted by reactions from the supports. Hence the
loading will not affect the moment along the beam. Since the moment is not affected,
the C-line is not affected. Thus, linear transformation involving bending of the cgs line
over the interior supports will not change the location of the C-line. On the other hand,
85

G.P.Ancog

any movement of the cgs line at the ends of the beam changes the magnitude of the
applied end moment which do affect the moments on all spans. Hence, linear
transformation cannot involve the movement of the cgs line over the ends of the
beam or over the exterior support of a cantilever, but it can involve movement of the
cgs line over the interior supports.
The above theorem, permitting the linear transformation of the cgs line without
changing the C-line, offers many possible adjustments in the location of the cgs line
which cannot be easily accomplished without that knowledge.
Concordant Cable In Continuous Beam
Concordant cable in continuous beam is a cgs line which produces a C-line
coincident with the cgs line.
In other words, a concordant cable produces no secondary moment. Thus every
cable in a statically determinate structure is concordant, because no external reaction
is induced, and there is no secondary moment in the structure. For a continuous beam,
on the other hand, external reactions will usually be induced by prestressing. These
reactions will produce secondary moments in the beam, and the C-line will shift away
from the cgs line. When this happens, the cable is termed nonconcordant. When by
chance or by purpose, no reactions are induced in a continuous beam by prestressing,
then there will be no secondary moments and the cable is a concordant one.
When a concordant cable is prestressed, it will tend to produce no deflection of
the beam over the supports, and hence no reactions will be induced (not considering
the weight of the beam).
Besides the fact that concordant cable line is easier for analysis, there is seldom
a necessity for using a concordant one.
While no significant reason can be given for preferring a concordant cable, there
is even less justification for locating a nonconcordant cable for the sake of
nonconcordancy. The real choice of a good cgs location depends on the production of a
desirable C-line and the satisfaction of other practical requirements, but not on the
concordancy or nonconcordancy of the cable.. A concordant cable, being easier to
compute, is slightly preferred, other things being equal.

86

G.P.Ancog

Concordant cable, C-line

coincides with cgs line

R = 0 due to
prestress

R=0

R=0

a) Continuous Beam With Concordant Cable

cgs line, a non=
coincides cable

R 0 due to
prestress

Secondary moment

C-line due to prestress

R0

R0

b) Continuous Beam With Nonconcordant Cable

A convenient procedure in design is to obtain a concordant cable that gives good

position of the cgs in resisting the external moment. If that location falls outside the
beam, it can be linearly transformed to give a practical location without changing its
C-line. According to this procedure, the finding of locations for concordant cables
become a useful means to an end.
Theorem for obtaining a concordant cable.
.
Every real moment diagram for a continuous beam on nonsetting supports,
produced by any combination of external loadings, whether transverse loads or
moments, plotted on a scale, is one location for a concordant cable in that beam.
Example 3:
For a continuous beam loaded as shown, obtain some desirable locations for
concordant cables to support that loading.
Solution:
Note: Every moment diagram plotted to any scale is a concordant cable.

87

G.P.Ancog

Example Beam:
48 k

1.60 k/ft

25

50

50

a) Continuous Beam With Any Loading

475 k-ft
-291

-363

19

b) Moment Diagram For a)

9.50

7.26

5.82

c) One Concordant Cable From b)

4.75

3.63

2.91

d) Another Concordant Cable From b)

e) Continuous Beam With Uniform Loads And End Moments

f) Concordant Cable From Loading In e)

1.
2.

If we plot the continuous beam moment diagram for a given loading, we

obtain b)
Two concordant cable locations are shown in c) and d), both are
proportional to the moment diagram in b) and hence both are concordant.
88

G.P.Ancog

3.
4.
5.

6.

f) gives another location of a concordant cable, which is proportional to the

moment diagram for loading in e).
Many similar concordant cables can be found by drawing all kinds of
moment diagrams.
The most desirable concordant cable will be governed by practical
requirements of the particular problem as well as by the ability of the
cable to resist the applied loads. For example, the location in c) gives
larger resisting arms for the steel but may overstress the concrete if the
weight of the beam is light, in which case d) may be a better location.
f) does not suit this particular loading as well but gives a symmetrical
layout and may carry other loadings, such as beams own weight, more
efficiently.

Corollaries

of

the

General

Theorem

1.

The reverse of the theorem is true: The eccentricity of any concordant

cable measured from the cgc is a moment diagram for some system of
loading on the continuous beam plotted to some scale.

2.

Any C-line is a concordant cable, since it is obtained by computing the

moments due to a system of loads on a continuous beam.

3.

Superposing two or more concordant cables will result in another

concordant cable. Superposing a nonconcordant cable and a concordant
cable will result in a noncocordant cable.

4.

When a sudden change in direction is desired, a concentrated load is

applied. When a gradual change is desired, a uniform load is applied. One
moment diagram can thus be modified into another by the addition of
loads. Hence one concordant cable can be easily modified into another.

5.

In order to obtain a concordant cable from another by linear

transformation involving moving of eccentricities over the ends of a
continuous beam, the following procedure can be used:
a.
Apply an end moment on the continuous beam; compute the
moment diagram due to that moment.
b.
When one end is moved by a given moment, the entire cable must
be transformed linearly in proportion to that moment diagram.
c.
If the movement of eccentricities at both ends is desired, apply end
moments at both ends proportional to the respective amount of
movements, and shift the entire cable in proportion to the moment
diagram so obtained. This will yield another concordant cable.

Much ingenuity can be exercised in the location of concordant cables, but it

should be left to the skill of the designer after he understands the basic theorem
and its main corollaries. When applied to rigid frames, the effect of sidesway and
89

G.P.Ancog

rib shortening should be additionally considered. For varying prestress along the
beam, the moment diagram should be divided by the corresponding prestress at
each point in order to obtain the location of a concordant cable. Or the tendons
may be treated separately. If each individual tendons or group of tendons forms
a concordant cable, then, when acting together, they also form a concordant
cable.

Example 4:
Obtain a new concordant cable, with its intrinsic shape the same as that of
Example 3-c), but with the right end of the cable 4 above the cgc.
19

9.50

7.26

5.82

Ex. 3-c)
0.25

0.375
1.00

0.095

19

a) Moment Diagram for Continuous Beam Due to Unit Moment at Right End
8.50

5.76
4

6.20
19

b) Concordant Cable From Ex. 3-c) With 4 Displacement at Right End

Solution:
1.
2.

Apply a unit moment at the right end of the beam; by moment distribution
method, plot the moment diagram as in a)
The concordant cable can now be linearly transformed in proportion to the
moment diagram in a), giving a new concordant cable as in b).

Note:
The same moment diagram in a) can be used to shift the end eccentricity any
other amount, not only for 4 as illustrated. Also, owing to the symmetry of the beam,
a) can be similarly used for moving the end eccentricity at the left. A combination of
90

G.P.Ancog

two moment diagrams due to a moment at each end will permit the simultaneous
shifting of both end eccentricities.
An infinite number of concordant cables can be obtained by rotating one
concordant cable about the points of inflection, because such rotation simply
represents the addition of one concordant cable to another, and should result in a
concordant one. The points of infliction in these moment diagram are sometimes
called nodal points.

Cable Location
Cable location means the location of the centroid of the tendons, that is the cgs
line. After the cgs line is determined, the location of the individual position of the
various tendons is an easier problem.
Designing a continuous prestressed-concrete beam, like that of any other
continuous structure, is essentially a procedure of trial and error. Knowledge regarding
the analysis of such structures, together with a systematic approach to the solution,
will aid greatly in arriving at desired results.
Recommended Steps for Designing a Continuous Prestressed Beam
1.
2.

3.

Assume section of members for dead load computation.

Compute maximum and minimum moments at critical points for various
combinations of dead, live, and other external loads. Compute the amount
of prestress required for these moments and the corresponding depth of
concrete. Modify section of members and repeat steps 1 and 2 if
Mmax
necessary.
Plot the top and bottom kern lines for the members. From the bottom kern
MG
line, plot:

a min

M min
F

Mmin

also

aG

MG
Fo

Moment Diagram for One Span of a

Beamsmallest
= theContinuous
algebraic

where Mmin
moment. The distance amin and aG
should be plotted upward for M and downward for +M.
Top Fiber
Top Kern

aG
amax
aG
amin

91
Bottom
Fiber
Limiting Zone for C-line Due to Prestress

Bottom Kern
G.P.Ancog

For the top kern, plot

a max

M max
F

also

aG

MG
Fo

again, upward for M and downward for +M.

The shaded area, between the limits of these four lines, obtained by
amin, amax and aG, represents the zone in which the line of pressure must lie
if no tension is permitted. When the zone is too wide, an excess of
prestress or of concrete section or of girder load is generally indicated. If
the limiting line from one kern crosses the a limiting line from another
kern, an inadequacy is evident. An ideal layout is obtained when there
exist a narrow limiting zone within the beam where the centroid of the
cables can be conveniently located.
4.

5.

Select a trial cable location within the above zone. Note that, if the cable
follows the shape of some moment diagram, it will be a concordant cable.If
this location is a concordant cable, it is a satisfactory solution. If it is a
nonconcordant cable, the C-line can be determined by moment
distribution method. If the C-line still lies within the limiting zone, then the
two locations are possible: either the trial location giving a nonconcordant
cable, or a new location following the C-line. If this C-line lies outside the
zones, a new cable location can be tried. An attempt should be made to
get a concordant cable within the zone. It is generally best to try a
concordant cable, because they coincide with their C-lines and give a nore
direct solution.
The concordant cable within the limiting zone obtained in Step 4 is a good
location for resisting external moment, but it may or may not be a good
practical location

Example 5:
92

G.P.Ancog

A pedestrian bridge of prestressed-concrete slab, has a three-span symmetrical

continuous layout as shown. The bridge is 9-4 wide with a uniform thickness of 13
(neglecting curb). The total effective prestressing force is 1,230,000 lbs after
deducting a loss of 15%. Design live load is 50 psf. Choose a suitable location for the
cable, allowing no tension in the concrete, fc = 5,000 psi.
Solution:
Following the procedures described above and considering 1-ft unit width of slab:
Step 1:
The section is already chosen, and the dead load is 162 plf or 162 plf for a
1-ft width.
Step2:

The amount of prestress is already chosen; it is:

F

1230000
132 k ,
9.33

Fo

F
132

156 k
1 0.15

Step 3:
Kern lines for a rectangular section are located h/3. the maximum and
minimum moment diagrams together with the girder moment diagram are shown for
one-half of the structure. These diagrams are divided by the respective prestress, F for
those with live loads, and Fo for dead load only. The a values thus obtained are plotted
from the kern lines as shown, giving the limits for the zone within which the C-line due
to prestressing must lie.
Step 4:
Using the moment diagrams as guides, select a trial cgs location within
the zone as shown. For purposes of illustration, assume the cgs line to posses the
following characteristics.
1.
2.
3.
4.

Passing through the cgc (mid-depth of slab) at end supports.

One sharp bend for each side span.
One sharp bend over each intermediate support.
A parabolic curve for the center span.

For the cgs location, the corresponding loading on the concrete is shown in d),
moment distribution for which gives a moment diagram as in e). Dividing the moment
diagram by the prestress yields a C-line as shown in f) which is very close to the trial
location and is still within the limiting zone. Hence this C-line is a location for a
satisfactory concordant cable.
Step 5:
A more practical location for the cgs is shown in g), affording a better
location for the steel. This is obtained by linearly transforming the concordant cable
93

G.P.Ancog

into a nonconcordant one. This nonconcordant cable will yield the same C-line as the
concordant one and hence will serve the same purpose as far as the stresses are
concerned.

40

60

a) Bridge Elevation

40

40

30

CL

9-4

a) Half Bridge Elevation

22.2 k-ft
14.2 k-ft

13

LL on AB, CD

41.5 k-ft

LL on BC

29.3 k-ft

b) Bridge Section
DL only

DL only

CL

b) Maximum & Minimum Moment Curves

LL on AB, BC
43.6
46.7
53.9
58.4
DL only

LL on AB, CD

LL on BC

kt
cgc
4
kb

4.5
cgs

LL on AB, BC

DL only

15

CL

c) Limiting Zone & Trial cgs Location

5.5 k
0.208 k/ft

DF 0.69
FEM+32.2
d) Loading
Dist-32.2
CO
Dist
----------Total0.00

0.31

-19.3+62.4

& Moment Distribution for Prestress in c)

-16.1
-18.6
- 8.4
-54.0l 54.0
54.0
31.3

e) Moment Diagram from d)

39.6
4.90

2.85

3.60

f) C-line from e)
94

3.09

G.P.Ancog

4.25
4.25

g) Nonconcordant Cable

Analysis of Continuous Prestressed Concrete Beam

L1

L2

L3

1. Area-Moment Method of Analysis

a
)

L1
A

Pe
EI
B

95

G.P.Ancog

L1

L2

MSBA

b
)

L3

MSCB

MSBC
B

MSCD
C

Procedure:
1. Considere the spans cut at interior supports and each span is considered as a
simple beam on which moment, Pe, (primary moment) are imposed due to
prestress..
2. The slope of the end span at the first interior support due to prestressing may
then be calculated as:

AB

Pe(L1 ) x
EI
L1

3. Apply a moment on the end span at B (M SBA, the fixed-end secondary moment
at B) which will rotate the beam back to zero slope. For two span symmetrical
structure, this moment would be the secondary moment due to prestressing.
For three or more spans or unsymmetrical structures, it is necessary to
distribute the fixed-end secondary moments. Application of MSBA will result
in a triangular moment diagram over the end span and the rotation of the
support due to this moment diagram will be:

AB

M SBA L1
3EI

Equating rotations due to prestressing and due to secondary moment MSBA can be
obtained:

Pe(L1 ) x M SBA L1

EI
L1
3EI

M SBA

5.
6.
7.

3
L1

Pe(L x
1

In similar manner, the rotations due to prestress moments are calculated

for the other end span and for the interior spans.
For the interior spans, moments are imposed at each end of the span to
rotate the beam back to zero slope. It is necessary then to solve 2
simultaneous equations to determine values of MSBC and MSCB.
After obtaining MSBC, MSCB and MSCD (the latter in a manner similar to M SBA),
the moment distribution technique may be used to calculate the final
secondary moments at the two interior supports.
96

G.P.Ancog

L
2

8.
9.

8P(eb ec)
L2

After determining the final secondary moments at supports, the secondary

moment effect in any interior span point is readily obtained by Plinear
interpolation between the secondary moments at supports.
The final moments due to prestressing at any point are then calculated as
the algebraic sum of the primary moments, Pe, and the secondary
moment MS.

2. Equivalent Load Method

The effect of the prestressing force on a beam can be
determined by
considering the prestressing force to be replaced by equivalent external loads. After
obtaining the equivalent loads, fixed end moments can be determined and the final
fixed end moments can be determined by applying moment distribution technique or
any other method of structural analysis.
Simple Span:

The uniform load w:

97

G.P.Ancog

wP

d
ds

1 dy

dx
R
d2y
dx 2
d2
P 2
dx
w

P
R

3
2

dy

dx
2

ds
d

3
2

The eccentricity can be expressed as:

4(eb e c ) 2
x ec 2P(1-1)eb
2
1 (1-)L2
L
The moment due to prestressing force P at a distance x from the origin:
e( x )

4( e b e c ) 2

x ec
2
L

M Pe( x) P

1)eb
Equivalent Load 2P(1for Continuous
Beam:
(1-)(1--1)L

Forces exerted on concrete in typical end and interior spans when

eccentricity varies parabolically.
End Span:
2P(1-)eb
(L)2

eb

eb

MBA

1eb

Peb
Ptan

B
L

(1 - 1L

1L

End Span

End Span

98

G.P.Ancog

4P(1 2)eb
(1/2 2)L2

Fixed End Moment:

M BA

Peb
2
2
2 1 3 3 1 1 2 1 1 1 1 1
4

4P(1 2)eb
2L2

Eccentricity:
x = 0 to x = L:

( )eb 1
(L)

x2

2( 1 )eb
x e b
(L)

x = L to x = (1 1)L
e

(1 1 )eb
2 (1 1 )eb
2 (1 1 )eb
2
x

1eb
(1 )(1 1 ) L
(1 )(1 1 )
(1 )(1 1 ) L2

x = (1 1)L to x = L

(1 1 )eb 2 2(1 1 )eb

(1 1 )eb
x
x
eb
2
1 (1 ) L
1 (1 )
1 (1 ) L

Interior Span:
e

eb

eb
2eb
A

B
2L

( - 2)L

( - 2)L

2L

Interior Span

99

G.P.Ancog

Fixed End Moments:

M AB M BA

2 Peb
- 1)(1 1)eb
(1 2 )(1 4(222) +(12
- )4L2
3

Eccentricity:
x -

x = - L/2 to x = - (1/2 2)L

2(1 2 )eb 2 2(1 2 )eb

(1 2 )eb
x
x
eb
2
2L
2 2
2L

x = -(1/2 2)L to (1/2 2)L

24(1 - 1)eb

2(1 2 )eb 2 (1 - )4L3

e
x 2 eb
(1 2 2 ) L2
x = (1/2 2)L to x =x2 L/2
+

2(1 2 )eb 2 2(1 2 )eb

(1 2 )eb
x
x
eb
2
2L
2 2
2L

Forces exerted on concrete in typical end and interior spans when

eccentricity varies as a parabola and a fourth-degree curve.
-12(1 -e1)eb
(1 - )4L4

eb

eb

1eb

(1 )L

2P(1 )eb
L2

End Span

MBA
P

Peb

End Span

100

B
G.P.Ancog

Fixed End Moment:

M BA

Peb
5 2 1 (3 2 12 10) 4 (2 3)
20

Eccentricity:
x = 0 to x = L

( 1 ) e b 2 2( 1 ) e b
x
x eb
(L)
(L) 2

x = L to x = L
e

(1 1 )eb 4 4 (1 1 )eb 3 2(2 2 2 1)(1 1 )eb 2 4 (2 1)(1 1 )eb

x
x
x
x
(1 ) 4 L4
(1 ) 4 L3
(1 ) 4 L2
(1 ) 4 L

2 ( 2 4 2)(1 1 )eb

1 eb
(1 ) 4
e

eb

eb
2eb
A

B
L/2

L/2

Interior Span

MAB
MBA
P

- 192(1 2)ebx2 + 16(1 2)eb

L4
L2

A
Interior Span

101

G.P.Ancog

Fixed End Moments:

M AB M BA

8 Peb
(1 2 )
15

Eccentricity:

16(1 2 )eb 4 8(1 2 )eb 2

x
x 2 eb
L4
L2

After the prestressing moments have been determined at the supports,

prestressing moments between supports may be obtained by superimposing the
simple beam moment diagram due to prestressing loads on the base line provided by
connecting the total prestressing moments at supports. However, it is generally
simpler to obtain interior prestressing moments using the tendon profile. Using this
approach, the interior moments may be expressed as:
M Pe M AB ( M BA M AB )

x
x

P e A (e B e A )
L
L

where:
MAB and MBA are the total prestressing moments at two adjacent supports
P is the tendon force
eA, eB & e are the eccentricities at A, B & x, respectively
Eccentricities are taken as negative when it falls below the cgc, and positive above
cgc.

(eB eA)x
L
eA

(eB eA)

eB

eA

(MBA MAB)x
L
MAB

MAB

(MBA - MAB)

MBA

3. Load Balancing Method of Analysis

The basic concept of load balancing also represents the influence of the tendons
by equivalent loads. The tendon is selected to directly counteract the imposed loading
at the indicated eccentricity, e. Since the moment induced by the tendon and the load
offset each other, the net stress in the beam will be the axial compression due to
prestressing, f = P/A. If it is desired to design the beam for zero stress at the bottom
fiber at center span (or for any value of stress less than the modulus of rupture), it is
only necessary to reduce the amount of prestressing provided. The net stress on the
section may be calculated from:
102

G.P.Ancog

P M net c

A
I

where Mnet is the net (unbalanced) bending moment on the section.

For continuous beam designs, the tendon geometry would be assumed to have a
parabolic shape, as shown.
Ptan

Ptan

Load w

wp

cgc

cgs
e
L

Simple Beam

L2

L1

L3

Continuous Beam

While load balancing is an extremely powerful concept for visualizing the effect
of prestressing on the structure, there are some qualifications that should be kept in
mind when using the technique in final detailed designs.
1.

2.
3.

It is not practical to install tendons with a break incurvature over

supports. The stiffness of the tendons requires a reverse curvature in
the tendon profile with a point of contraflexure some distance from
the supports. This reverse curvature modifies the loads imposed by
prestressing from those assumed using load balancing.
The load balancing technique should be adjusted to counter friction
losses and the resulting variation of force in the tendon along its
length.
As in equivalent load method, load balancing does not account
directly for moment induced by tendons anchored eccentrically at
ends free to rotate.

Prelimary Sizing of Prestressed Members

There are no set span-depth limits for prestressed members, but some values in
the table below are provided as a guide to the preliminary sizing of members.
Typical span-depth ratios
Continuous span
Roof
Floor
50
45
45-48
40-45

Element Type
1-way solid slab
2-way solid slabs, supported on columns only
103

Simple span
Roof
Floor
45
40
G.P.Ancog

2-way waffle slabs (36 pans)

Beams
One-way joists

40
35
42

35
30
38

35
30
38

30
26
35

These ratios maybe increased if calculations verify that deflection, camber, and
vibration frequency and amplitude are not objectionable.

Design Example 1: 2-Span Cast-In-Place T-Beam

Design a 2-span cast-in-place T-beam of normal weight concrete. Use f c = 5000
psi and fpu = 240000 psi (1/4 dia. wire, post-tensioned). Beams are 26.9 ft on centers.
2

20x24 cols., 10 long

3
b = 182
6
30

20
14
57-00

57-00

Section properties:
Ag
yt
yb
Ig
Zt
Zb

=
=
=
=
=
=

1512 in2
8.00 in
28.00 in
133000 in4
16630 in3
4752 in3

Dead load = 2455 lb/ft (including full slab)

Live load = 50 psf reduced to 30 psf on columns and beams
Moments
From the above information, the dead load moments were calculated (by
moment distribution) to be 385.5 k-ft at the exterior columns and 804.3 k-ft at the
interior column and +402.1 k-ft at midspan. The corresponding live load moments (not
considering alternate panel loading) are 126.7, -264.4 and +132.2 k-ft, respectively.
Moment at
Exterior columns
Interior columns
-385.5
Mid-span

DL
-385.5
-804.3
+402.1

LL
-126.7
-264.4
+132.2

Total
-512.2
-804.3
-1068.7
+534.3
-264.4

-126.7

interior col

exterior col

+132.2

+402.1
57-0

104

G.P.Ancog

Post-tensioning
The effects of post-tensioning are treated separately from the effects of loads
and are dependent on the magnitude and position of the tendon force and on frame
action. Assume two-end stressing. A tendon profile is selected:

28
3
L/12
= 4.75

(11/12)(57)=52.25

Secondary Moments
Secondary moments induced by post-tensioning are accounted for by
considering the effects of restrained rotations. For simplicity, a post-tensioning force of
1
2
1 kip is used.

b
a

52.25

4.75

Mp = Pe; P = 1 kip
a = (28 3) x 1 = 25 k-in
b = (8 4) x 1 = 4 k-in

Using conjugate beam method:

Rotation @ 1:
2
1
3
52.25
2
EI 1 x 25 x52.25 x
4.75 x 4.75 x 4 x (4.75)
471.3
8
2
3
3
57
Rotation @ 2:
105

G.P.Ancog

 2 52.25 2
2
5

x
x 25 x 4.75 x 4 x 52.25 ( 4.75) 386.9
2
3
8

EI 2

Fixed end moments:

2 2 1 555.7

19.50 k in
28.5
FEM @ 1: = L 2

1 2 2 302.5

10.61 k in
28.5
FEM @ 2: = L 2
Moments (per kip of force)
By moment distribution, moments at exterior columns are +11.31 k-in, at
interior column, + 14.71 k-in, and at mid-span, + 13.01 k-in. These secondary
moments can be considered to be:
M S P(e)
So the resulting e values, in this example, are 11.31, at columns 1 & 3, 14.71 at
column 2, and 13.01 at mid-span.
Required Post-Tensioning Force:
Allow a tensile stress of: f =

6 f c ' 6 5000 424 psi

Mid-span M due to allowable tensile stress:

M f '

I
4752
f ' Z b 424
167.9 k ft
cb
12(1000)

Support M due to allowable tensile stress:

M f '

I
16630
f ' Z t 424
587.6 k ft
ct
12(1000)

Final Required Prestress @ Exterior Support:

106

G.P.Ancog

M T M sup port (512.2 587.6)12

40.6 k
Zt
16630
0 11.31
e e
1512
Ag

Final Required Prestress @ Interior Support:

M T M sup port (1068.7 587.6)12

194.3 k
Zt
16630
4 14.71
e e
1512
Ag

Final Required Prestress @ Mid-span:

M T M midspan (534.3 167.9)12

290.5 k
Zb
4752
25 13.01
e e
1512
Ag

The final prestressing required is the maximum computed:

P Pexterior , Pint erior , Pmidspan

max

40.6,194.3,290.5 max 290.5

Mid-span prestressing force governs.

Approximate Prestress Loss Values
Post-tensioning
Tendon Material
Stress relieved 270K strand, stress relieved 240K
wire
Bar

Prestress Loss, psi

Slabs
Beams & Joists
30,000

35,000

20,000

25,000

This Table of approximate pretsress losses were developed to provide common Post-Tensioning Industry
basis for determining tendon requirements on projects in which the magnitude of prestress losses is not
specified by the designer. These loss values were based on the use of normal weight concrete and on
average values of concrete strength, prestress lebvel, and exposure conditions. Actual values of losses
may vary significantly above or below the Table values in cases where the concrete is stressed at low
strengths, where the concrete is highly prestressed, or in a very dry or very weight exposure conditions.
The table values do not include losses due to friction.

Friction Losses
1

= 0.20, K = 0.001
4

25

107
28.5

28.5

G.P.Ancog

Stressing simultaneously at supports 1 & 3:

Po Px e ( KL )
For circular curve:

2e

L 2

tan 1

2(25 / 12)
.145 radians
28.5

1 tan 1

2(29 / 12)
.170 radians
28.5

2 tan 1
For mid-span:

1 = 0.20(.145) = 0.0290
KL = 0.001(28.5) = 0.0285
0.0575
Po Px e ( KL 1 ) Px e ( 0.0575) 1.059 Px

0.8 f pu
Po / Px

0.8 f pu
1.059

0.755 f pu 0.7 f pu

, so use 0.7fpu.

Initial steel stress = 0.7fpu = 0.7(240)

= 168.00 ksi
Losses taken
= 25.00 ksi
Final steel stress
= 143.00 ksi
For interior support:
= 0.20(.145 + 2(0.168))
KL = 0.001(57)

= 0.0962
= 0.0570
0.1532
108

G.P.Ancog

Po Px e ( KL 1 ) Px e ( 0.1532) 1.166 Px

0.8 f pu
Po / Px

0.8 f pu
1.166

0.686 f pu 0.686(240) 164.7 ksi

Initial steel stress = 0.6867fpu = 0.686(240)

= 164.70 ksi
Losses taken
= 25.00 ksi
Final steel stress
= 139.70 ksi

Determine Number of Wires:

Use dia. wires, Ao = 0.0491 in2, mid-span control the number of wires
requirements.
N

P
290.5

41.37,
f e ( Ao ) 143(0.0491)
use 42 dia. wires.

Check total prestress at interior support:

P Nf e Ao 42(139.7)(0.0491) 288.1 k 194.3 k

, ok!

Ultimate Moment Analysis:

Ultimate load wu: 1.4DL + 1.7LL = 1.4(2455) + 1.7(807) = 4808 plf
Total service load: w = DL + LL = 2455 + 807 = 3262 plf
At supports 1 & 3:

Mu

wu
4808
MT
(512.2) 755 k ft
w
3262

At support 2:

Mu

wu
4808
MT
(1068.7) 1575 k ft
w
3262
109

G.P.Ancog

At mid-span:

Mu

wu
4808
MT
(534.3) 788 k ft
w
3262

Moment provided:(Bonded reinforcement)

At supports 1 & 3:

As NAo 42(0.0491)

0.00526
bd
bd
14(28)

f pu

1
240

f s f pu 1 (
) 240 1 .5(0.00526)
209.7 ksi
2
fc '
5

f s
209.7

(0.00526)
0.221 0.3 max .
fc '
5

M u As f s d 1 .5 [.9(42 x0.0491)( 209.7)( 28)(1 .5 x0.221)] / 12 789.7 k ft

Mu provided = 789.7 k-ft > Mu required = 755 k-ft, ok

At mid-span:

As NAo 42(0.0491)

0.000343
bd
bd
182(33)

f pu

1
240

f s f pu 1 (
) 240 1 .5(0.000343)
238.0 ksi
2
fc '
5

f s
238.0

(0.000343)
0.016 0.3 max .
fc '
5

M u As f s d 1 .5 [.9(42 x 0.0491)( 238.0)(33)(1 .5 x 0.016)] / 12 1203 k ft

Mu provided = 1203 k-ft > Mu required = 788.0 k-ft, ok

At support 2:

As NAo 42(0.0491)

0.00460
bd
bd
14(32)

f pu

1
240

f s f pu 1 (
) 240 1 .5(0.00460)
213.5 ksi
2
fc '
5

f s
213.5

(0.00460)
0.196 0.3 max .
fc '
5

M u As f s d 1 .5 [.9(42 x0.0491)( 213.5)(32)(1 .5 x0.196)] / 12 934.0 k ft

Mu provided = 934.0 k-ft < Mu required = 1575 k-ft, so add Grade 60 rebars.
110

G.P.Ancog

Calculate moment at face of column:

L M u @ 2 M umidspan
(1575 755)

4.808(28.5)
137.0 14.4 151.4 k
2
L
57
2
Approx. moment reduction: M = 3 Vu(a); a = (col. size) = (20) = 10 in.
2
10
M (151.4) 84.0 k ft
3
12
Moment at face of column: Mu = 1575 84 = 1491 k-ft
Vu wu

Required additional rebars:

Asb

M u M uprovided

f y d
fy
fc '

(1491 934)(12)
3.87 in 2
0.9(60)(32)

3.87 60
Asb f y

.103

bd f c ' 14(32) 5

0.196 0.103 0.299 0.30 max, ok!

Check stresses at service load:
At exterior supports:
fse = 143.00 ksi
P = NAofse = 42(0.0491)(143) = 295 k

Fiber Stress:
Top fiber stress:
ft

P P(e e) M T

A
Z
Z

P P(e e) M T

A
Zt
Zt

295 295(0 11.31) 512(12)

1512
16630
16630
0.195 0.201 0.369 0.027 ksi, compressio n

Bottom fiber stress:

111

G.P.Ancog

fb

P P (e e) M T

A
Zb
Zb

295 295(0 11.31) 512(12)

1512
4752
4752
0.195 0.702 1.293 0.786 ksi, compressio n

At mid-span:

Fiber stress:

P P(e e) M T

A
Z
Z

Top fiber stress:

ft

P P ( e e ) M T

A
Zt
Zt

295 295( 25.00 13.01) 534(12)

1512
16630
16630
0.195 0.213 0.385 0.367 ksi, compressio n

Bottom fiber stress:

fb

P P (e e) M T

A
Zb
Zb

295 295(25.00 13.01) 534(12)

1512
4752
4752
0.195 0.744 1.348 0.409 ksi, tension

409 psi f ' 6 f ' c 6 5000 424 psi , ok!

At interior support
fse = 139.7.00 ksi
P = NAofse = 42(0.0491)(139.7) = 288 k

Fiber Stress:

P P(e e) M T

A
Z
Z

Top fiber stress:

112

G.P.Ancog

ft

P P (e e) M T

A
Zt
Zt

288 288(4 14.71) 1069(12)

1512
16630
16630
0.190 0.324 0.771 0.257 ksi, tension

Bottom fiber stress:

fb

P P(e e) M T

A
Zb
Zb

288 288(4 14.71) 1069(12)

1512
4752
4752
0.190 1.134 2.699 1.755 ksi, compressio n
1755 psi f c 0.45 f c ' .45(5000) 2250 psi , ok

These stresses computed are all within allowable limits. Repeat stress check for
initial prestress force of 0.80fpu, allowable stress for compression fc = 0.6fc, and
f '
tension f = 3 c .
To complete the design, check mid-span deflection and design the shear
reinforcements.

Design Example: Flat Plate

Design a two-way post-tensioned flat plate with 22-00 square bays as shown.

10

cols. 16 x 16

22-00 transverse
design strip
10
3 bays @ 22-00

Loading:
Superimposed dead load (partition)
Live load

20 psf
40 psf
113

G.P.Ancog

Materials:
Concrete (normal weight)
Compressive strength, fc

150 pcf
4000 psi

Prestressing steel:
Use dia. strand unbonded, fpu

270 ksi

Non-prestressed reinf., ASTM A615-60, fy

60 ksi

Select slab thickness:

Use span-depth ratio of 40

L 22(12)

6.6 in
40
40
use 6.5" slab.
t

Loading for 22 ft strip:

6 .5
22
(150)
1.79 klf
12
1000
Slab dead load:
22
20
0.44 klf
1000
Superimposed dead load (partition):
40

Live load:

22
0.88 klf
1000

Select tendon profile:

1.25

3
3

3.25 3.25

1.25
0.0

-2.0

+2.0

-1.0

+2.0

-2.0

0.0

Tendon Profile

For min. cover and dia. tendons:

Min. distance face of slab to cgs of tendons = 0.75 + 0.5/2 = 1.00, use 1.25

114

G.P.Ancog

With equal length spans, simple beam moments are identical for all spans,
therefore set total drape (net vertical profile distance) of the tendons the same for
each span as shown in the tendon profile.
Since the tendon is anchored at slab mid-depth, the end span sets the max. midspan drape as the sum of the negative eccentricity and the average positive
eccentricity:
Mid-span drape = 2 + (2 + 0)/2 = 3
Mid-span eccentricity of 1 is then calculated for the interior span to give equal
total drape in all spans.
For unequal span frames, the tendon eccentricity is usually set at the maximum
allowed over each support to get maximum ultimate strength capacity.
Select balancing prestress force:
Balance total dead load: 1.79 + 0.44 = 2.23 klf

Pe'

wb L2
, where P prestress force after all losses , e' total drape
8
wb L8
simple span oment
8

wb L2 2.23(22) 2 (12)
P

540 k for 22' design strip

8e'
8(3)
At this stage, the balanced load portion of the design is complete. A prestress
force of 540 k following the profile shown in the figure for tendon profile, will produce a
uniform upward load of 2.223 klf, which is equal and opposite to the uniform total
dead load. IN this idealized state, bending moment stresses are zero, and the slab is in
direct compression caused by the prestress.
fc = P/A at balanced condition
Similarly, bending stresses at time of prestressing are calculated by summation
of initial compressive stress (Po/A) and bending stress due to the distributed
overbalanced moment (Mob) caused by the difference between balanced uniform load
and the uniform load at time of prestressing (usually slab dead load).

fc

Po M ub

A
Z

Idealized stress at balanced condition

115

G.P.Ancog

P = 540 k
fc

A = 6.5(22x12) = 1716 in2

P 540 x1000

315 psi , compressio n

A
1716

Idealized stress at service loading

Perform moment distribution for the unbalanced uniform live load of 0.88 klf and
results are plotted as follows:
20.9

-29.9

20.9

15.9

-29.9
-42.0

-37.1

-37.1 -42.0

Unbalanced Moments

Check stresses at maximum moment:

fc

P M ub 540(1000) 42.0(12)(1000)

A
Z
1716
22 x12(6.5) 2 / 6

315 271 44 psi, compressio n

therefore, the slab has no tensile tress under full service load.

Idealized stress at initial loading:

Assume initial tendon stress = 0.7fpu, and effective tendon stress = 0.6fpu.
.7
.7
(315) 367 psi
(2.23) 2.60 klf upward
Then initial Po/A= .6
, and initial wb = .6
Uniform slab dead load
Net loading = 2.60 1.79

= 1.79 klf downward

= 0.81 klf upward (overbalanced)

Since the overbalanced initial loading (0.81 klf) is less than the unbalanced service
Po
P
loading (0.88 klf), and A is greater than A , it is apparent that the slab has no tensile
stresses at initial loading.
Comparison of idealized design with more rigorous analysis.
Since tendons cannot be placed with sharp peaks of tendon profile, the actual
tendon profile produces load reversals as it changes from concave upward to concave
downward, and resultant moments change accordingly. Using equivalent loading and
tendon infliction points at 1/15 of span lengths and from structural analysis (moment
distribution, computer program, etc.), the following results are:
116

G.P.Ancog

Ext. Col.
Mid-span
-27.4
-18.8
Check stresses at max. moment:

fc

Int. Col.
-51.5

Mid-span
-9.9

P M ub
51.5

315
(271) 17 psi , tension
A
Z
42.0

Check Ultimate Flexure

From moment distribution (moments in k-ft)
Moment
Ext. col
Mid-span
MDL
-61.2
+53.7
MLL
-28.8
+19.7
Mu actual
-134.3
+108.7
Mu
-133.5
+226.5
capacity

Int. col
-100.7
-38.4
-206.3
-231.2

Mid-span
+42.7
+17.4
+89.4
+180.6

Calculation of ultimate moment capacity:

Assume effective prestress of prestressing steel, fse = 0.6fs = 0.6(270) = 162 ksi, then
A ps

P
540

3.33 in 2
f se 162

At interior column:
d = 6.5 1.25 = 5.25 in

A ps

3.33
0.0024
bd (22 x12)(5.25)
f '
4
f s f se 10 c 162 10
189 ksi
100
100(0.0024)
, (specs)
f s
189

0.0024
0.113 0.30, ok
fc '
4

A ps f s
.85 f c ' b

3.33(189)
0.70 in
.85(4)( 22 x12)

M u A ps f s d .9[3.33(189)(5.25 .7 / 2)] / 12 231.2 k ft

2

Check punching shear stress:

117

G.P.Ancog

For 22 x 22 panel:
VDL = 1.4(2.23 klf)(22) = 68.7 k
VLL = 1.7(0.88 klf)(22) = 32.9 k
Vu
= 101.6 k
For a 16 x 16 cols., bo = 4(b + d) = 4(16 + 5.25) = 85 in.
v 4.2 f c ' 4.2 4000 265.63 psi
Concrete shear capacity: uc
Actual shear stress:
V
101.6 x1000
vu u
267.85 265.63
bo d .85(85)(5.25)
taken ok.
Although the value obtained is somewhat higher, empirical equations developed
from the following tests indicate adequate punching shear capacity.
Lin-Scordelis-May Tests:
Vu (0.175 0.0000242 f c '0.00002 P / S )bdf c '
P = 540 k, S = 22 x 12, b = 4(16) = 64 in., d = 5.25 in.
Vu [0.175 2.42 x10 5 (4000) 2.0 x10 5 (540 x10 3 /( 22 x12))]( 64)(5.25)( 4.0) 161.0 k
> 101.6 k
Grow-Vanderbilt Tests:
Vu = (360 +0.30fc)bd = [360 + 0.3(315)](64)(5.25)/1000 = 153.00 k > 101.6 k
Minimum bonded reinforcements:
From ACI Code, when tension exists in the precompressed tensile zone at a load
of DL + 1.2LL, a minimum amount of bonded reinforcement is required. At interior
column, the moment at DL + 1.2LL = 100.7 + 1.2(38.4) = 147.0 k-ft:
P M s ( M DL 1.2 M LL )

A
Z
Z
P = 540 k, A = 1716 in2, Z = 1859 in3 , Ms = 87.6 k-ft ( from moment
distribution)
fc

Top fiber:

540 x10 3 87.6 x12 147(12)

0.069 ksi
1716
1859
1859

fc

540 x10 3 87.6 x12 147(12)

-0.069
0.699 ksi
1716
1859
1859
0.58

Bottom fiber:

fc

6.5

118
+0.699

G.P.Ancog

Tension in concrete:
1
1
f ' cb (. 069)(0.58)( 22 x12) 5.28 k
2
2
As f s Tc

Tc

As

Tc
5.28

0.18 in 2
0.5 f y 0.5(60)

Provide: 2 - #3 top bars at int. columns.

Tension Members
Prestressed tension members combine the strength of a high-tensile steel with
the rigidity of concrete and provide a unique resistance to tension consistent with
small deformations that cannot be obtained by either steel or concrete acting alone.
The rigidity of prestressed concrete serves well, especially for long tension members
such as tie rods for arches or staybacks for wharves and retaining walls. When
prestressed, concrete is given strength to resist any local bending and at the same
time steel is stiffened and protected.
The basic behavior of prestressed tension members:
1.

2.

3.

The member can be considered as essentially made of concrete which is put

under uniform compression so that it can carry tension produced by external
loads. If the concrete has not cracked, it is able to carry a total tensile force
equal to the total effective precompression plus the tensile capacity of the
concrete itself.
The member can be considered as essentially made of high-tensile steel
which is pre-elongated to reduce its deflection under load. From this view
point, the ultimate strength is often limited by excessive elongation of the
steel which usually takes place at the cracking of the concrete.
The member can be considered as a combined steel and concrete member
whose strains and stresses before cracking can be evaluated, assuming
elastic behavior and taking into account the effect of plastic flow.
A

(n-1)As

P
Section at A-A

119
Prestressed Concrete Tension Member.

Transformed
Section

G.P.Ancog

Initial prestress:

f co

Fo
Ac

Effective prestress:
fc

F
Ac

Transformed section:
At nAs Ac unbonded section
At Ag (n 1) As

grouted section

Stress produce by P:
fc

P
At for concrete

f s nf c

nP
At for steel

Es
E c consider the effect of creep & shrinkage
Resultant stress (due to effective prestress plus external load):
n

fc

F
P

Ac At for concrete

fs fc

nP
At for steel

To find load P to produce resultant stress to zero:

A
F
P

0 ; P F t F (1 n )
Ac At
Ac
120

G.P.Ancog

Instantaneous unit strain:

Fo
EAc

Under the action of external load P:

P
EAt

Comparison of strains in a prestressed concrete and in ordinary steel member:

Given:
fs = 20,000 psi, Es = 30x106 psi

fs
20000

0.00067
E s 30 x10 6
fc from 1,000 psi to 0, Ec = 4x106 psi

Given:

fc
1000

0.00025
E c 4 x10 6
fse = 125,000 psi, Es = 30x106 psi

Given:

f se 125000

0.00417
E s 30 x10 6
, which 6 times that of structural steel and 16
times that of prestress concrete.

Example 1: Tension Member

A straight concrete member 150 ft long is prestressed with a high tensile steel
strand through the centroid of the section. The strand is anchored to the concrete with
end anchorages but separated from it by bond-breaking agents along the length. A c =
80 in2, As = 0.80 in2, fc = 4,000 psi, fs = 250,000 psi, fo = 150,000 psi, fse = 127,500
psi, Ec = 4x106 psi, Es = 30x106 psi.
a)
Compute the allowable external load on the member, allowing no tension in
the concrete.
b)
Compute the shortening of concrete due to prestress, assuming a creep
coefficient of 1.5.
c)
Compute the lengthening of the member due to the external load obtained
in a), neglecting creep.
d)
If the member were designed of structural steel with an allowable stress of
20,000 psi, compute the lengthening under the load.
e)
Compute the lengthening if the strand is used alone by itself with an
allowable stress of 127,500 psi.
121

G.P.Ancog

Solution:
a)

Allowable external load, allowing no tension:

P F (1 n ) f se As (1 n ) 127500(.80)(1

b)

30 x10 6 0.80
) 110,000 lbs
4 x10 6 80

Under initial prestress, compute the shortening:

Fo L 150000(0.80)(150 x12)

0.675 in.
E c Ac
4 x10 6 (80)

If effective prestress is considered:

f se
127500
0.675
0.573 in.
fo
150000

If creep is considered:

c C c 0.573(1.5) 0.860 in.

c)

30 x10 6
At Ac nAs 80
(0.80) 86 in 2
6
4 x10
Under the external load of 110,000 lbs:
:

d)

closely

check

the

shortening

of

concrete in b).
For a structural steel stressed to 20,500 psi, the elongation is:

s
e)

PL
110000(150 x12)

0.575 in.
E c At
4 x10 6 (86)
,

f s L 20000(150 x12)

1.20 in.
Es
30 x10 6

For high tensile steel stressed to 127,500 psi:

f se L 127500(150 x12)

7.65 in.
Es
30 x10 6

Generally speaking, pretsressed-concrete tension members have a very low

reserve strength above the point of zero stress. In order to get a sufficient factor of
safety, it may be necessary to design the member so that, under working loads, there
will be some residual compression in the concrete.

122

G.P.Ancog

Example 2:
For the tension member in Example 1, what working load can it carry using a
factor of safety of 2.0 against the cracking of concrete, assuming the direct tensile
strength of concrete to be 0.08f c = 320 psi? Compute the residual compression in
concrete under the working load.
Solution:
For fc = 320 psi: F = fseAs = 127500(0.80) = 102000 lbs, Ac = 80 in2, At = 86 in2:
F
P
102000 P

320;

Ac At
80
86
P 137,000 lbs , which is the cracking load .

For a factor safety of 2.0:

Pworking

Pcracking
FS

137000
68,500 lbs
2 .0

The residual compression computed:

fc

F
P 102000 68500

1275 795 480 psi

Ac At
80
86

Circular Prestressing

Circular prestressing is employed to denote the prestressing of circular

structures such as pipes and tanks where the prestressing wires are wound in circles.
In most prestressed circular structures, prestress is applied both circumferentially and
q
longitudinally,
the circumferential prestressing being circular and the longitudinal
prestress actually linear.
cgs line

Fo

C-line = cgc line

C = Fo

C = Fo Fo

T=q

Due to prestress, Fo

Due

Due to circular prestress, the stress in concrete is always axial an equal to:

f co

Fo
Ac , at transfer
123

G.P.Ancog

fc

F
Ac , at final stage of loading after all losses.

With application of internal liquid pressure:

fc

qR
; At Ac (n 1) As
At

The resultant prestress in the concrete under the effective prestress F and the internal
pressure q:
fc

F qR

Ac
At

In order to be exact, the value of n has to be chosen correctly, considering the level of
stress and the effect of creep.
The criteria for designing prestressed tanks vary. The usual practice is to provide
a slight residual compression in the concrete under the working pressure. This is
accomplished as follows:
1.

Assume that hoop tension produced by internal pressure is entirely carried by

the effective prestress in steel; then F As f s qR and the total required steel
is:
As

qR
f s , the total initial prestress is then, Fo As f so

The concrete area to resist the initial prestress Fo:

Fo
f c , from the value of the required A , the thickness t of tank can be
c
obtained.
Ac

2.

Corresponding to the adopted value of A c, the stress in the concrete and steel
under internal pressure q can be obtained as:
F qR

Ac At , the stress in concrete

f s nf c
, the stress in steel

fc

fs f

Since the serviceability of the tank is impaired as soon as the concrete begins to
crack, it is of utmost importance that adequate factor of safety be provided against
124

G.P.Ancog

cracking. If a factor of safety m against cracking is required, the design procedure

adopted will be as follows:
1.

Assuming ft = tensile strength in concrete at cracking (which averages 0.08f c

but may be zero if the concrete has previously cracked or if precast blocks are
used):
ft

2.

F mqR

Ac
At

At the same time in order to limit the maximum compression in concrete to f c,

then:

Fo
f c , and
f A f
mqR
f t se s c
f o As
( f o As f c ) nAs

Ac

3.

Determine the required steel area:

As

mqR
[ f se ( f t / f c ) f o ](1 nf c / f o )

4.

After As is obtained, Fo and Ac can be computed by:

5.

Fo
A f
s o
fc
fc
Stresses can be evaluated using the following equations:
Ac

F qR

Ac At , the stress in concrete

f s nf c
, the stress in steel

fc

fs f

Prestress losses for tanks:

Average loss of prestress
=
25,000 psi
Usual allowance of loss
=
35,000 psi, considered conservative
At extreme adverse condition =
40,000 psi
Creep and shrinkage allowance:
elastic and creep strain in concrete =
0.0005
shrinkage
=
0.0005
Total loss
=
0.0010
which amounts to about 28,000 psi, taking Es = 28x106 psi.
125

G.P.Ancog

Example 3: Tank Prestressing

Determine the area of steel wire required per foot of height of a prestressed
concrete water tank 60 ft inside diameter to resist 20 ft of water pressure. Compute
thickness of concrete required, fc = 3,000 psi, fc = 750 psi, n = 10, fo = 150,000 psi, fse
= 120,000 psi. Neglect mortar coating in the calculations. Design both steel and
concrete on the following two bases:
1.
2.

Assuming all hoop tension carried by effective prestress.

for a load factor of 1.25, producing zero stress in the concrete.

Solution:
Pressure of 20 ft of water:
q h 62.4(20) 1248 psf

a) Assuming all hoop tension carried by the effective prestress:

Required steel area:
As

qR 1248(30)

0.312 in 2
f se
120000

Required concrete area:

F0 (0.312 x150000)

62.5 in 2
fc
750
For a height of 12 , the thickness required is:
Ac

Ac 62.5

5.2 in., adopt 5.5 in.

b
12
and under the action of the internal pressure:

fc

F qR

; Ac 5.5(12) 66 in 2 ; At Ac nAs 66 10(0.312) 69.12 in 2

Ac At

fc

0.312 x120000 1248(30)

26 psi
66
69.12

b) For a load factor of 1.25, producing zero stress:

Required steel area:

126

G.P.Ancog

As

mqR
[ f se ( f t / f c ) f o ](1 nf c / f o )
1.25(1248)(30)
0.372 in 2
[120000 (0 /( 750))(150000)](1 10( 750) / 150000)

Required concrete area:

Fo
0.372(150000)

74.4 in 2
fc
750
Thickness required:
Ac

Ac 74.4

6.2 in.
b
12
, adopt t = 6.5 in.

Resultant stress in concrete under full water pressure: Ac = 6.5(12) = 78 in2

At = Ac + nAs = 78 + 10(0.372) = 81.72 in2
fc

F qR
.372(120000) 1248(30)

115 psi
Ac At
78
81.72

Note: By designing by the second method, gives heavier sections for both concrete and steel.
The design can be economized if some tension in the concrete is allowed at 25% overload.

Vertical Prestressing in Tanks

load carried by
vertical element
load carried by
ring tension

HR

0.0

Vertical Section
of Tank Wall
and Deflection
under Load

0.1
0.2

K=1.6
K=0.8
K=0.4

0.3

0.7

K=12.0

0.6

2R

0.5

Moment in
Vertical Element

Moment and deflection in vertical element of tank wall

K=6.0

K=3.0

Height, H

0.4

Load Distribution
between Horizontal
t
& Vertical Elements

0.8
0.9
1.0
.0

.1 .2

.3 .4

.5 .6 .7

Coefficient, C
Fixed Base, Free Top

.8

.9 1.0

127

K=

H2
2Rt

Ring Tension T = CHR

G.P.Ancog

0.0
0.1
0.2

2R

0.3

0.8
0.9

K=12.0

0.7

K=6.0

K=0.8

0.6

K=3.0

0.5

K=1.6

K=0.4

Height, H

0.4

1.0
.0

.1 .2

.3 .4

.5 .6 .7

.8

.9 1.0

Coefficient, C
Hinged Base, Free Top

K=

H2
2Rt

Ring Tension T = CHR

Horizontal elements of the tank wall are subjected to hoop tension and the
vertical elements are subjected to bending. The variation of bending in the vertical
elements will depend on several factors.
1.
2.
3.
4.
5.

The condition of support at the bottom of the wall, whether fixed, hinged, free
to slide, or restrained by friction.
The condition of support at the top of the wall, whether fully or partially
restrained or free to move.
The variation of concrete thickness along the height of the wall.
The variation of pressure along the depth, whether triangular or trapezoidal.
The ratio of the height of tank to its diameter.

Vertical prestressing should be designed to stand the stresses produced by various

possible combinations of the following forces:
1.
2.
3.

The vertical weight of the roof and the walls themselves.

The vertical moments produced by internal liquid pressure.
The vertical moments produced by the applied circumferential prestress.
128

G.P.Ancog

In addition to the above, stresses may be produced as a result of differential

temperature between the inner and outer faces of the wall, and by shrinkage of
concrete walls unless they are entirely free to slide on the foundation. These factors
cannot be easily evaluated and hence are often neglected or provided for indirectly in
an overall factor of safety..
It must be noted that the maximum stresses in the concrete usually exist when
the tank is empty, because then the circumferential prestress would have its full
effect. When the tank is filled, the liquid pressure tends to counterbalance the effect of
circumferential prestress and the vertical moments are smaller. Since it is more
convenient to use the same amount of vertical prestress throughout the entire height
of the wall, the amount will be controlled by the point of maximum moment. By
properly locating the vertical tendons to resist such moment, a most economical
design can be obtained. However, efforts are seldom made to do so, and the amount
of prestress as well as the location of tendons is generally determined empirically
rather than by any logical method of design.
Example 4: Design of Tank Wall
A 1-ft vertical element of a water tank is shown in the figure below. It carries
1500 lbs of weight from the roof. At a point 20 ft below the top, the vertical moments
are: for initial circumferential prestress M = 3200 ft-lb (tension on the inside fibers)
which reduces to 2500 ft-lb eventually. For full fluid pressure, M = 2400 ft-lb (tension
on the outside face). The vertical prestressing wire is located 2.75 in from the inside
face and exerts an initial prestress of 11,000 lb/ft, which reduces to 8,000 lb/ft
eventually. Compute the stresses in the extreme vertical fibers of the concrete under
the initial and final conditions, considering both an empty and a full tank.
1500 lbs

8
20

inside of tank
A

12

cgs
2.75

Vertical Element of Wall

Section A-A

Solution:
129

G.P.Ancog

Computation for stresses in concrete:

Stages of loading condition

Initial

Fiber

Inside

A. Weight of roof: f = P/A = 1500/(8x12)

-16
-21
B. Weight of wall: f = h/A = 150(20)/(8x12)
C. Axial component of vertical prestress: f =
P/A
-115
(11,000)/(8x12) =
(8000)/(8x12) =
D. Eccentricity of vertical prestress: f = 6M/
(bd2) = 6Pe/(bd2)
6(11000x1.25)/(12x82)
-107
6(8000x1.25)/(12x82)
E. Vertical moments due to circumferential
prestress
+300
f = 6M/(bd2) = 6(3200x12)/(12x82)
6(2500x12)/(12x82)
+41
Total for tank empty
F. Vertical moment due to liquid pressure
f = 6M/(bd2) = 6(2400x12)/(12x82)
Total for tank full

-225
-184

Final

Outsid
e

Insid
e

Outsid
e

-16
-21

-16
-21

-16
-21

-83

-83

-78

+78

+23
4
+36

-234
-276

-225
198

+225
-51

-115

+107

-300
-345
+225
-120

Slight tension of 36 to 41 psi exists on the inside vertical fibers when the tank is
empty. Otherwise, compressive stresses are obtained throughout.
Dome Ring Prestressing
Dome prestressing wires
and prestress temporary
erection bars

Dome shell

Vertical prestressing wires

Circumferential prestressing wires

Pneumatic mortar

Reservoir wall

130
Typical Section of Dome Ring for Tanks

G.P.Ancog

The dome roof itself is made of concrete or pneumatic mortar with thickness
varying from 2 to 6. For domes of large diameter, variable thickness may be
employed and thickness greater than 6 are used for the lower portion. Before
concreting the dome, some erection bars are prestressed around the base of the
dome. After the hardening of the shell concrete, wires are prestressed around it.
During this operation, the dome shell rises from its forms as it is compressed, thus
simplifying the careful procedure for decentering required for non-prestressed domes.
Methods and formulas, though available for the analysis of dome stresses under
uniform loads, are applicable only to points on the domes removed from the
discontinuous edge. The computation of stresses in the edge ring becomes very
complicated problem if the edge ring is prestressed. However, for purposes of design,
a conventional method is available. It consists of prestressing the ring to induce
sufficient compressive stresses to counteractTotal
the
tensile
dome
load W stresses set up in the ring
under the maximum live and dead loads. With this prestress, it is usually possible to
raise the dome from its falsework, since only the dead load is actually acting on the
dome.

Spherical dome

For a spherical dome:

Design for Prestress in Edge Ring of Dome

Vertical reaction per unit length along the edge member:

W
2R sin
The horizontal reaction per unit length:
V

131

G.P.Ancog

H V cot

W cot
2R sin

Required prestressing force F to resist the horizontal reaction:

F HR sin

W
cot
2

The effective prestressing F having been determined, the cross-sectional area of

the ring concrete can be obtained by:

Fo
fc
R = radius of dome
W = total vertical load on the dome
Fo = initial prestressing force, fc = allowable compressive stress in concrete.
Ac

It is desirable to keep fc at a relatively low value, say 0.2fc and not greater than
800 psi. This is necessary in order to minimize excessive strain in the edge ring which
might in turn produce high stresses in the shell. It must be further observed that this
procedure of design is satisfactory only when there is no possibility of heavy
overloads, because the prestressed edge ring does not posses a high factor of safety
against overloads, although the factor of safety is sufficient for ordinary roof loading.
Example 5: Prestress Dome Ring
A spherical dome as shown the figure, carries a total live and dead load of 900k.
Design the prestress in the edge ring and the cross-sectional area of the concrete
required for the edge ring. Loss of prestress = 20%, fc = 600 psi.
Total W = 900k
spherical dome
120 ft

45o

Required prestress:
132

G.P.Ancog

W
900
cot
cot 45 o 143 k
2
2

Initial prestressing force:

Fo

F
143

179 k
(1 .2) .80

Required area of concrete:

Ac

Fo 143x1000

298 in 2
fc
600

Prestressing Compression Members and Piles

A

P
A

Column under load P

Column prestressed with force F

Moment at A-A due to deflection

No moment at B-B due to deflection

Column Action Due To Prestress

Column Action Due To Load

A prestressed concrete member

A
F

F Concrete as freebody
A
A
F

F
A
Balancing Action of Concrete and Steel133

Steel as freebody
G.P.Ancog

Consider an ordinary column under an external load P, when the column

deflects, additional moment in section A-A is created by deflection, because the
external load now acts with a different eccentricity on that section.
Consider a column internally prestressed but not externally loaded, so long as
the steel and concrete deflect together, there is no change in the eccentricity of the
prestress on the concrete, no matter how the member is deflected. Hence there is no
change in moment due to any deflection of the member and no column action.
When an external load is applied to a prestressed-concrete column, any
deflection of the column will change the moment, and the column action will result.
Considering the concrete alone, it is a column under direct compression, and any
slight bending of the column will result in an eccentricity in a section such as in A-A.
But considering the steel as a freebody, there will exist an equal eccentricity with an
equal but opposite force, producing a tendency to straighten itself out. The tendency
to straighten is exactly equal and opposite to the tendency to buckle and hence the
resulting effect is zero. This is not true when the member is externally prestressed, say
against the abutments, because there will be no balancing effect from the prestressing
element, and column action will result.
C

C
Before Bending

Section C-C

C
After Bending

Section C-C

Steel and Concrete in Contact After Bending

E

Steel and Concrete in Contact at Several Points

If steel and concrete are not in direct contact along the entire length, the
problem will be different. The concrete under compression will have a tendency to
deflect laterally. That deflection will not at first bring the steel to deflect together with
it; hence the eccentricity of prestress on the concrete is actually changed, thus
resulting a column action. After a certain amount of deflection, the steel is brought
134

G.P.Ancog

into contact with the concrete and the two will begin to deflect together. Hence the
column action is limited to the differential deflection of the two materials.
If the steel is in contact with the concrete at several points, say at E and F, but
not along the entire length, the column action is limited to the length between the
points of contact. If such length is short, column action will not be serious.
Curve Members

Eccentric Prestress
And Column Action
Bent Members Under
Concentric Prestress

If a curve or bent member is subjected to internal prestress, and if the prestress

is concentric at all sections (the cgs line coinciding with the cgc line), then the
concrete is behaving like an arch subject to axial force with the corresponding
exception that the applied force from the steel will move with the deflection of the
concrete and will always remain concentric. Hence there is no tendency to buckle as in
ordinary arch under external loads, whose line of pressure is determined by the loads
and may not shift together with the deflection of the arch.
If the prestress is eccentric, the compression in the concrete is still equal and
opposite to the tension in the steel. Any deflection of the member will still displace
both of them together, and there will be no column action due to prestress. The effect
of an eccentric prestress on the concrete, however, will produce deflection of the
member should be used in computing column effects due to external loads.
As far as column action is concerned, it is immaterial whether there is any
frictional loss along the length of the prestressing tendon, because the tension in the
stell is always balanced by the compression in the concrete at any section, whatever
frictional loss may occur. Hence, whether there is frictional loss or not, there will be no
column action due to prestress.
Compression Members
A prestressed-concrete compression member is one that carries external
compressive load. A member that is simply compressed by its prestress is not a
compression member. A prestressed member is not under column action due to its
own prestress, but it is subject to column action under the external compressive load
just like a column of any other material.
135

G.P.Ancog

Column under prestress F with an eccentricity e, loaded with concentric load

P plus external moment M.
Short column condition:
fc

F Fec P Mc

A
Ic
At
It

Effect of axial prestressing on the ultimate strength of columns.

Under the action of an external compressive load, the column will shorten and
the prestress in the steel will be decreased.
uc = 0.003 for concrete
fe = effective prestress
fs = remaining prestress
Es = modulus of elasticity of steel
f s f e uc E s
Numerical values:
uc = 0.003, fe = 120,000 psi, Es = 30x106 psi
Remaining prestress, fs = 120000 0.003(30x106) = 30,000 psi
Remark:
The major part of prestress may be lost at the ultimate compressive strength of
the concrete. This means that the ultimate load carrying capacity of the column is not
much decreased by prestressing. If the column fails on the tensile side as the result of
bending or buckling, the steel on that side can be stressed to near its ultimate
strength.
The ultimate strength of concentrically prestressed slender columns under axial
loads has been investigated both theoretically and experimentally at various
universities. The general conclusion is that the axial prestressing of a slender column
has no effect on the superimposed axial load which will cause that column to buckle. If
the prestressing exceeds the difference between the buckling stress and the ultimate
strength of the concrete, the column will fail in compression before it will buckle. When
the superimposed load is not axial, prestressing could increase both the cracking and
the ultimate strength.
The buckling of the compressive flange of prestressed beam is subject to the
same reasoning. There is no danger of flange buckling produced by internal prestress
in a beam. For external loads, the tendency to buckle in the flange is governed by the
usual theory of elasticity, so long as there is no cracks in the concrete. After cracking
or near ultimate load, little is known about the buckling of the compressive flange in
prestressed beams.
Example 1
A concrete column 16 by 16 in cross section and 18 ft high is pre-tensioned
with 8 3/8 in. wires, which are end anchored to the concrete. The effective prestress
136

G.P.Ancog

16

2 to center of steel

2 to center of steel

16

Column Section
is 100,000

psi in the steel. For a concentric compressive load of 80 k and a horizontal

load of 8 k at the midheight of the column, compute the maximum and minimum
stresses in the column, assuming it to be hinged at the ends. Investigate the
secondary moments in the column due to deflection. Discuss the safety of the column
under such loads and also during handling. Assume n = 7, f c = 4,000 psi, fs =
200,000 psi, Ec = 4x106 psi.

80
8 kk

Column

Solution:
Stress in concrete due to prestress:

F (8 x0.11)( 100,000)

344 psi
Ac
(16 x16 8 x 0.11)
Stress due to axial load P = 80 k, disregarding deflection of column:

P
P
80,000
80,000

306 psi
At Ag (n 1) As 16 x16 (7 1)8 x0.11
261
The maximum bending moment occurs at midheight:
M

PL 8(18)

36 k ft
4
4

It of the transformed section:

It

bh 3
16 4
2(n 1) As i x 2
2(7 1)(3 x0.11)(6) 2 5460 142 5602 in 4
12
12

Extreme fiber stresses:

Mc (36 x1000)(12)(8)

616 psi
It
5602
137

G.P.Ancog

Maximum and minimum stresses:

f max 344 306 616 1266 psi compressio n
f min 344 306 616 34 psi compressio n
Maximum deflection of the column due to the horizontal load:
PL3
(8 x1000)(18) 3 (12 3 )

0.075 in.
48 E c I t
48(4 x10 6 )(5602)
Increase in moment due to axial load:
M P (80 x1000)(0.075) 6000 in lb (0.50 k ft)

Very small to materially affect the column stresses, and may be neglected. The
maximum compressive stress of 1266 psi would appear to be high for a reinforcedconcrete column but is not excessive for a prestressed member which more a beam
than a column in this example.
The safety of the column can be determined only if we know the ultimate
strength of the column under such combined axial and transverse loads and also if we
know the possibilities of overloading, that is, to what extent the axial or the horizontal
loads may be increased, and whether eccentricity of the applied axial load may be
possible.
For the purpose of investigation, let us assume that both the horizontal and the
axial load are increased by 50% while, in addition, there will be an eccentricity of 2 in.
for the axial load. The stresses can be computed:
Stress due to axial load:
1.5(-306) = - 459 psi
Moment due to eccentricity: M = 1.5(80)(2) = 240 k-in (20 k-ft)
20
616( ) 342 psi
36
Stress due to eccentricity:
Stress due to horizontal load:
1.5( 616) 924 psi
Resulting stresses:
f max 344 459 342 924 2069 psi
f min 344 459 342 924 463 psi
Remark:
The maximum compressive stress of 2069 psi is only about 0.52f c, while the
tensile stress is below the modulus of rupture of about 0.12f c, hence the column
138

G.P.Ancog

would not have cracked, and the midheight deflection can still be computed by the
elastic theory not to be more than 0.2 in., which is not a significant value. Thus it can
be concluded that the column is safe.
Investigating the handling stresses:
Assume the column is picked up at the midheight:
Resulting moment:
2
wc L2 Ag c L
[16 x16(150) / 144](9) 2
M

10.8 k ft
2
2
2

Maximum tensile stress:

Mc (10.8 x1000)(12)(8)

185 psi 344 psi

It
5602

Since the maximum tensile stress during handling, + 185 psi is much less than
the precompression stress of 344 psi, the column is safe during handling.
Column Under Eccentric Load

Precast bearing walls and columns can be prestressed to improve their elastic
behavior and handling characteristics, and to increase their resistance to lateral forces
both in the elastic and the ultimate ranges. They cannot be designed following rules of
thumb applied to reinforced concrete walls and columns. But they can be designed on
P and properties of materials. The behavior and
the basis of principles of mechanics
e
strength of prestressed columns under eccentric loading can be predicted with fair
precision. The degree of accuracy will depend on the choice of values of modulus of
elasticity, the modulus of rupture, and the compressive strength of the concrete.
Before cracking, the stresses and deflections can be calculated assuming the column
to behave elastically. The stress at any section is the sum of the stresses due to
prestress, direct axial load, moment due to the eccentricity, and the moment due to

the deflection.

P
Column under
eccentric load

139

G.P.Ancog

Column stresses:
f max
min

F
P Pec Pc

At At
It
It

where:
F = effective total pretress including all losses except elastic shortening of
concrete due to superimposed load.
P = superimposed load.
e = eccentricity of load from the centroid of section.
c = distance to the extreme fiber from the centroid of the section
At = area of transformed section
It = moment of inertia of the transformed section
= deflection of column at the section.
Critical stress occur at the midheight of the column, where the deflection of the
column is given by the well-known secant formula:

PL2
e sec
1

4Ec I t

By the elastic theory, cracking can be assumed to occur when the fiber stress
reaches the modulus of rupture. Beyond cracking, the elastic theory is no longer
accurate. An estimation of the ultimate load can be made by the elastic theory,
assuming it to be the load at which the extreme fiber stress reaches the compressive
strength of concrete. Since such an approximate analysis can be way off (by perhaps
some 10% even for ordinary cases), it is desirable to apply plastic analysis, taking into
140

G.P.Ancog

1.5

account the cracking of concrete under tension, the plasticity of concrete under
As: 6 - 3/8 strands
compression, and the plasticity of steel. Furthermore, while the elastic analysis
mentioned above will generally err on the conservative side, it is conceivable that,
Section A-A
under unusual conditions, erratic conclusions could be reached unless plastic analysis
is applied.
Elevation

Example 2: Column with Eccentric Load

A pre-tensioned concrete pin-ended column has elevation and section as shown.
The effective prestress in the six 3/8 in. 7-wire strands (as = 0.08 in 2 each) is 150,000
psi or 12,000 lbs per strand. Es = 30x106 psi. concrete has cylinder strength of 5,700
psi, modulus of rupture of 600 psi, and E c = 4x106 psi. It is loaded by load P with an
eccentricity of 1.5 in. along the weak direction. Compute the cracking and the ultimate
L
value of P using the elastic theory, assuming
noncracked section.

Solution:
a) Compute properties of the section:
n

E s 30 x10 6

7.5
Ec
4 x10 6

At Ag (n 1) As (8 x12) (7.5 1)( 6 x 0.08) 96 3 99 in 2

I t I g 2(n 1) Ax ( x 2 )

12 x83
2(7.5 1)(3 x.08)( 2.5 2 ) 512 19 531 in 2
12

b) Compute deflection of the column at midheight by the secant formula for several
values of P, say P = 50, 100, 120, 140, and 160 k. Numerical computation is given
for P = 120 k.

141

G.P.Ancog

PL2
120 x10 3 (249) 2

e sec
1 1.5 sec

1
6

4
E
I
4
(
4
x
10
)(
531
)
c t

1.5(sec 0.935 1) 1.5(sec 53.7 1) 1.5(1.688 1)

1.03 in.
c) Compute stresses in concrete at midspan:
F
P Pec Pe

At At
It
It
min
F = Asfse = (6x.08)(150,000)x10-3 = 72 k, assuming it is not reduced by the
presence of P, and using P = 120 k, D = 1.03 in:
f max

72 x10 3 120 x10 3 120 x10 3 (1.5)( 4) 120 x10 3 (1.03)( 4)

99
99
531
531
728 1210 1355 930
4223 psi

f max
min

f max

f min 347 psi

d) Plot elastic stresses and deflections for various loads P, assuming noncracked
section. From the graph, determine cracking load and ultimate load.
Load P
kips

Due to P - Pe

200
Resulting tensile stress
Due to P

Due to P + Pe

160
140

Cracking load = 130 k

fc = 5,700 psi
-7000 -6000 -5000 -4000 -3000 -2000 -1000

Maximum compressive stress, psi

120
100

Modulus of rupture = 600 psi psi

Resulting compressive stress

Stress due to F = 728 psi

Ultimate load = 144 k

Due to P

180

1000

2000

80

Deflection at midheight
of column

60
40
20

3000

Max. tensile stress, psi

Deflection, inches

b) Deflection by Elastic Theory

a) Stresses by Elastic Theory

142

G.P.Ancog

Ultimate Strength of Prestressed Column Under Combined Axial Plus

Bending

Ultimate strength under combined axial load plus bending moment

M
t
2

t
2

Steel
Column

s1
c s2

Strain at

Section
T2

T1

CL of column
Stress block
in concrete

fc

Applying equilibrium equation:

Fv 0 : P C T1 T2
M 0 : (T1 T2 )

yc

yt
Cyc
2

By assuming a location for the neutral axis at ultimate load, setting c as the
ultimate strain in concrete and fc as the ultimate stress of concrete, and by assigning
ultimate stress distribution curves for concrete, it is possible to compute the
combination of P and M that results in this ultimate failure.
For slender columns, the value of M just computed should include the effect of
deflection, which can be computed by numerical procedure provided the load-moment
curvature relationship of the column section is known.
Example 3: Ultimate column load
For the column section shown in example 2, if the ultimate neutral axis were
located at 3 in. from one edge, compute the combined P and M producing that failure.
143

G.P.Ancog

Solution:
Assume c = 0.003, by proportion, changes strains for the steel are:

P
M
t
2
5
2.5

t
2

Steel
Column

3
2.5

s1
c s2

Strain at

Section
T2
CL of column

T1

Stress block
in concrete

.85fc
yc
6

Es = 30x10 psi, ec = 0.003:

C
a

3.5
3.5
c
(0.003) 0.0021
5
5
1.5
1.5

c
(0.003) 0.0009
5
5

s1
s2

As = 3(0.08) = 0.24 in2

T1 ( f se s1 E s ) As (150000 (0.0009(30 x10 6 ))( 0.24) 42.5 k

T2 ( f se s 2 E s ) As (150000 (0.0021)(30 x10 6 ))( 0.24) 20.8 k
Using trapezoidal stress distribution with average ultimate stress at 0.85fc and a = 3:

144

G.P.Ancog

C .85 f c ' b(

ac
35
) .85(5700)(12)(
) x10 3 232 k
2
2

y c 1.96 in
P C T1 T2 232 42.5 20.8 169 k

t
Cyc (42.5 20.8)( 2.5) 232(1.96) 508.97 k in
2
M 508.97
e

3.0116 in
P
169
M (T1 T2 )

which indicate that a load of 169 k with a total eccentricity of 3.0116 in. (including
column deflection if any) will produce failure in the column, when the ultimate neutral
axis is assumed located at 3 in. from the edge.
Cross section

can

Various combination of P and M can be plotted and a column interaction diagram

for the column with given amount of prestress.

Elevation
be
drawn

se s

Mu
Pu

P
M

Strain
*tdistribution
M

se
uc

0.85fc

0.85fc

Ts
s

Ts

Ts
1

o2
3

cu

se + s

Cross Section
b

Elevation

Strain Variation

Jensen
Idealization

Ts
Hognestad
Idealization

Stress Block Idealization

Conditions at ultimate in a prestressed concrete column.

Systematic determination of the interaction diagram using various strain

distribution.

145

G.P.Ancog

A systematic way to determine points of the interaction diagram is illustrated in

the above picture. One may start with a uniform strain distribution over all the section,
at the concrete strain o and maximum concrete stress, and determine the maximum
axial force that can be supported by the section. This is given by position 1 of the
various strain-distribution planes. The failure criterion that requires that the middepth
strain be equal to o is satisfied by rotating the strain-distribution plane about the
middepth section. Any strain-distribution contained between planes 1 and 2 satisfies
the failure criterion.

Pu

Mu

Typical ultimate column interaction diagram.

The top branch of the interaction diagram shown above between points 1 and 2 was
obtained from the strain-distribution planes bound by planes 1 and 2. The top branch
is characterized by failures due to large axial loads and small bending moments. The
rest of the interaction diagram, between points 2 and 3, may be obtained by satisfying
the criterion that failure of the section occurs when the crushing strain uc is attained
at the extreme fiber, between limits set by planes 2 and 3. The later corresponds to
failure of the section due to bending moment alone, such as in the conventional case
of a beam section, and cannot be predicted a priori. It can be obtained by considering
the column as pure beam action, no axial load.
Example 4:
146

G.P.Ancog

A prestressed-conrete column with a 16 x 10 rectangular cross section is

reinforced with 8 in. strands, four in each wide face. The strength of the concrete
is fc = 6000 psi. the strain at which maximum stress occurs, o, and the crushing strain
uc are assumed equal to 0.002 and 0.003, respectively. The stress-strain diagram for
concrete is described by k1, k2, and k3. For this problem assume k1 = 0.75, k2 = 0.42,
and k3 = 0.85. For the steel, se = 0.005, fsu = 208 ksi.
Solution:
Two levels of initial prestrain in the prestressing steel, ese are considered. In the first
case, full prestressing to se = 0.005 is applied while the second case the section has
no prestressing at all, se = 0.

Plane 1:
10

16

Plan

8 strand
As = As = 0.575
se = 0.005
fc = 6 ksi

Stresses in steel:
fs = fs = s Es = 0.003(30x103) = 90 ksi
Forces:
C = k3fcbt = 5.1(16x10) = 816 k
Ts = Ts = Asfs = 0.575(90) = 51.8 k

Strain

Strains in steel:
es = es = eo = 0.002
s = s = se = s = 0.005 0.002 = 0.003

o = 0.002
Pu

Vertical equilibrium:
Pu = C Ts Ts = 712 k
Mu = 0

k3fc = .85(6)
= 5.1 ksi

Stresses
Ts C

Ts

147

G.P.Ancog

Plane 2:

10

16

8 strand
As = As = 0.575
se = 0.005
fc = 6 ksi

Plan

Strain

s oc = 0.003
c = 10
Pu

Stresses

k1k3fc = (.75)(.85)(6)
= 3.8 ksi
Ts

Strains in steel:
es = uc(d/c) = 0.003(8/10) = 0.0024
s = se s = 0.005 0.0024 = 0.0026
es = euc(d/c) = 0.003(2/10) = 0.0006
s = se - s = 0.005 0.0006 = 0.0044
Stresses in steel:
fs = s Es = 0.0026(30x103) = 78 ksi
fs = sEs = 0.0044(30x103) = 132 ksi
Forces:
C = k1k3fcbt = 3.8(16x10) = 608 k
Ts = As fs = 0.575(132) = 76 k
Ts = Asfs = 0.575(78) = 45 k
Vertical equilibrium:
Pu = C Ts Ts = 487 k
Mu = C(t/2 k2c) + Ts(t/2 d) Ts(t/2 d)
Mu = 608(5 - .42x10) + 76(5 2) 45(5 2)
Mu = 580 k-in

C Ts

10

16

Plan

8 strand
As = As = 0.575
se = 0.005
fc = 6 ksi

Plane 3:
2

Strain

oc = 0.003

Mu

Stresses
C
Ts

k1k3fc = (.75)(.85)(6)
= 3.8 ksi

Strains in steel:
es = uc[(c-d)/c] = 0.003[(c 2)/c
s = se s = 0.005 0.003[(c 2)/c]
es = euc[(d c)/c] = 0.003[(8 c)/c]
s = se + s = 0.005 0.003[(8 c/c]
Stresses in steel:
fs = s Es = (30x103){0.005 0.003[(c 2)/c]}
fs = 150 90[(c 2)/c]
fs = sEs = (30x103){.005 + 0.003[(8 c)/c]} = fsu
Forces:
C = k1k3fcbc = 3.8(16c)
Ts = As fs = 0.575fs
Ts = Asfs = 0.575fs
Determination of c:
Fv = 0: = C Ts Ts = 487 k
0 = 3.8(16c) 0.575(150 90[(c 2)/c] - 0.575fs
Assume c = 3 in., then fs = fsu = 208 ksi and substituting:

Ts

3.8(16x3) = 0.575(150 90[(3 2)/3] + 0.575(208)

182 < 69 + 119 = 188, therefore c = 3 in., try c = 3(185/182) = 3.05 in.
then fsu = 208 ksi,
3.8(16x3.05) = .575(150 90[(3.05 2)/3.05] + 0.575(208)
186 68 + 119, c = 3.05 in.
Pu = 0
Mu = k1k3fcbc(h/2 k2c) Asfs(h/2 d) + Asfs(h/2
148 d)
Mu = .75x.85x6x16x3.05(5 - .42x3.05) - .575(150 90[(3.05 2)/3.05](5 2) + 0.575(208)(5 2)
Mu = 844 k-in

G.P.Ancog

10

16

8 strand
As = As = 0.575
se = 0.005
fc = 6 ksi

900

se = 0

800
700

se = 0.005

Axial load P, k

600

tensile strain when se = 0.005

500 tensile strain when zero
zero
se = 0
400
300
200
100
0

200

400

600

800

1000 2000

Bending moment M, k-in

Effect of initial prestrain on the interaction digram of a prestressed-concrete column.

The interaction diagram when the initial strain se = 0 is shown. The section
behaves like a conventional like a conventional reinforced concrete except that it is
reinforced with high-strength steel which lacks a definite yield point. This account for a
flat curvature of the diagram, as shown above, and for the absence of a sharp
balanced point, a definite characteristic of the interaction diagram of conventionally
reinforced concrete sections. The initial prestrain in the steel reduces the capacity of
the section to resist external load. This, of course, is due to the section of the internal
compression force imposed by the prestressing steel on the concrete.
Zero Tensile Strain column Interaction diagram
The stress block in the concrete is not fully developed when the maximum strain
in the section is smaller than uc. From the Hognestad Stress-strain diagram:

149

G.P.Ancog

C (0.85 f c ' bt )
Mc

4
o

1 4
3 o

1
1 4
(0.85 f c ' bt 2 4 1
6
o
2 o

where all symbols have been defined as before with the exception of 4 which is the
strain in the concrete at the near end fiber. The preceding expressions are valid for the
range 0 4 o when the strain in the concrete at the far end is zero.

P
M

Elevation

Strain
distribution

uc
4

Systematic determination of the zero tensile-strain interaction diagram using various strain distribution.

10

Plan

16

Plane 4:

8
strand
As = As
= 0.575
se = 0.005
fc = 6 ksi
2

Strain

s 4 = 0.0015
c = 10
e

Stresses

Ts

C Ts

Pu

Strains in steel:
es = 4(d/c) = 0.0015(8/10) = 0.0012
s = se s = 0.005 0.0012 = 0.0038
es = e4(d/c) = 0.0015(2/10) = 0.0003
s = se - s = 0.005 0.0003 = 0.0047
Stresses in steel:
fs = s Es = 0.0038(30x103) = 114 ksi
fs = sEs = 0.0047(30x103) = 141 ksi
Forces:
C = .85fcbt(e4/eo)(1 1/3(e4/eo)
= .85(6)(16)(10)(.0015/.002)(1 1/3(.0015/.002)
= 459 k
Ts = As fs = 0.575(114) = 65.6 k
Ts = Asfs = 0.575(141) = 81.1 k
Vertical equilibrium:
P = C Ts Ts = 312 k
M = (1/6)Ct(1 (4/o))/(1 1/3(4/o)) + Ts(h/2 d)
Ts(h/2 d)
M = 459(10/6)(1 (.0015/.002))/(1 1/3(.0015/.002)
+ 65.6(5 2) 81.1(5 2)
= 638 + 243 197 = 684 k-in
150

G.P.Ancog

Prestressed Concrete Piles

Since piles are subjected to tensile stresses during transportation, driving and under
certain service conditions, the desirability of prestressing is evident.
Design of prestressed piles
Experience seems to indicate that a prestress of about 700 psi in the piles will
ensure safety during handling and driving under normal conditions. While the amount
of prestress required will vary with the size and shape of the pile, the hammer blow,
and the cushioning effects, as well as the soil conditions, it is obviously impractical to
vary the prestress in each pile. Of course, higher or lower values than 700 psi may be
desirable for special cases.
The bearing capacity of concrete piles is seldom if ever considered by their
strength under direct compression, but it is convenient to express the bearing capacity
in terms of the compressive strength or stresses. Strictly speaking, if the bearing
capacity were limited by the compressive stress, there would be no need for
prestressing. Therefore current formulas are empirical in nature.
The design load on such piles is often based on ultimate strength using arbitrary
factor of safety of about 4. Such a higher factor of safety is hardly necessary so far as
the service load is concerned, but it is believed that for piles so designed, the
compressive stress during driving will seldom be critical, and it should be possible to
attain the desired bearing value without damaging the pile.
If the cylinder strength of concrete is fc, the ultimate strength of the concrete in
a pile can be safely assumed 0.85f c. At ultimate load, the amount of prestress
remaining in the tendons is approximately 60% of the effective prestress. Thus if a
6,000 psi concrete is prestressed to an effective prestress of 700 psi, the ultimate
strength can be computed by the formula:
N = (0.85fc 0.60fc)Ac = (0.85fc 0.60(700))Ac
where Ac = cross-sectional of the concrete pile. Using a factor of safety of 4.0, the
design load N = N.
151

G.P.Ancog

Joint standard set up by AASHTO and the Prestressed Concrete Institute state
that the maximum compressive stress fpc on prstressed concrete piles (in addition to
the effective prestress) for fc = 5,000 psi shall not exceed the following:
fpc = 1000 psi
fpc = 1160 16L/D

for 0 < L/D < 10

for 10 < L/D < 25

where
L = effective length of pile, taken as the actual length of pile when both
ends are hinged
D = diameter or width of pile
For 5,000 psi < fc < 9000 psi, these stresses may be increased indirect
proportion to the design concrete strength. The buckling load of a pre-tensioned
concrete pile can be computed using the Eulers column formula:
N cr
where
E
I
L
k

2 EI
(kL) 2

Ncr = critical buckling load

= modulus of elasticity of concrete
= moment of inertia of the concrete pile section
= length of pile
= effective column length factor

The value of E should be chosen to fit the duration of loading that is, a higher
value should be used for dynamic load and a reduced value for sustained load. Since
possibility of an increase in actual load is remote, a factor of safety of 2.0 is
considered sufficient. Thus the allowable load is often set at:
N

N cr
2 EI

2
2(kL) 2

If no tensile stress is allowed, a high factor of safety is obtained for

concentrically prestressed members subjected to bending. It is therefore often
permissible to allow tensile stresses in the concrete under design moments.
The joint AASHTO-PCI Standards allow 250 psi tension for normal design loads,
while values up to 600 psi have been permitted for earthquake and other infrequent
loads. Thus if fe is the effective prestress in the pile concrete, the allowable moment
under earthquake loading is:
M ( fe ft )

I
c

where
I = moment of inertia of concrete section
c = distance from neutral axis to extreme fiber
152

G.P.Ancog

The ultimate moment is computed by:

M u CAs f s ' d
where
C = a coefficient depending on shape of pile section, etc. varying between
(0.32 0.38)
As = total area of prestressing steel
fs = ultimate strength of steel
d = diameter or size of pile
The design moment, based on the ultimate load, should have a factor of safety of 2.0,
while a factor of safety of 1.5 to 1.7 will be sufficient for earthquake and wind loads.
According to the elastic theory, the existence of direct external loads delays
cracking of the concrete pile and thereby increase the moment-carrying capacity. ON
the other hand, the ultimate moment capacity is reduced by the presence of direct
external loads. Hence, when the design is for combined moment and direct loads, the
moment capacity of the pile should be checked by both the elastic and the ultimateload theories, as described in columns under eccentric loads.
REFERENCES
1. Narbey Khachaturian, and German Gurfunkel, Prestressed Concrete, McGraw-Hill,
1969.
2. T. Y. Lin, Design of Prestressed Concrete Structures, John Wiley & Sons, 2e, 1963.
3. Phil M. Ferguson, Reinforced Concrete Fundamentals, Joh Wiley & Sons, 3e, 1973.
4. Arthur H. Nelson, Design of Concrete Structures, McGraw-Hill, 12e, 1997.
5. Design Manual, Portland Cement Association.

153

G.P.Ancog