CVX7640 – Structural Design

Session 28

Analysis of a Prestressed Concrete

Section
Aim:

To introduce the concepts used in the analysis of

a prestressed concrete section under different
loading conditions.

Objective:

To be able to analyse a prestressed concrete

member section under flexure, to satisfy the three
defined Classes at Serviceability Limit State and
evaluate the capacity at the Ultimate Limit State.

28.1 Introduction

The design of a prestressed concrete structure involves

many considerations, the most important of which is the
determination of the stress distributions in the individual
members of the structure. In most types of structures it is
usually sufficient to consider certain critical sections
where the stresses are the greatest. However, in
prestressed concrete structures, with high stresses by the
prestress force, more sections have to be considered as
critical and the stress distributions checked for different
stages of loading. The practical means of carrying this out
will be discussed in a subsequent session.

This session deals with the distribution of flexural stresses

at the Serviceability and Ultimate limit states. These two
distributions are different, but in determining those same
three basic principles as given below are employed;

(a) Strain distribution

This is assumed to be linear in elastic bending theory, and
this assumption is also found to be sufficiently true for
concrete members even up to the point of faiture. The
strain in the steel in pretensioned and bonded post-
tensioned members is assumed to be the same as that in
the concrete at the same level.

(b) Material stress-strain curves

These have been described earlier for steel and concrete.

95
CVX7640 – Structural Design

(c) Equilibrium
At any section in a prestressed concrete member there
must be equilibrium between the stress resultants in the
steel and concrete and the applied bending moment and
axial load (if any) at that section.

The basic difference between the analysis of sections at

the serviceability and ultimate limit states is that in
principle different regions of the stress-strain curves
identified under (b) are used in each case.

28. 2 Serviceability limit state

The analysis or sections in Class 1 and 2 members at the

serviceability limit state is carried out by treating the
section as linearly elastic and using ordinary bending
theory. (The analysis for Class 3 members will be
considered later). This is justified by the fact that, at the
service load, the stress- strain curve for steel is linear, and
that for concrete is approximately so. Furthermore, Class
1 and 2 prestressed concrete members remain uncracked
at service loads , justifying the use of the second moment
or area based on the gross concrete section.

In was shown earlier that, in for a prestressed concrete

member not subject to any loading, the concrete behaves
at any cross-section as if it were subjected to an axial force
P and a bending moment Pe, where e is the eccentricity of
the prestress force at that section, as shown in Fig. 28.1.
Centroid of the section Longitudinal segment

Cross section Zt
e
≡ P P
P P
Zb Pe Pe
Figure 28.1:
Eccentrically prestressed P/Ac Pe/Zt P/Ac-Pe/Zt
section with equivalent -
loading and stress distribution + + =
+
Pe/Zb P/Ac+Pe/Zb
Stress distribution at the segment

Thus the extreme values of the stress distribution due to

an eccentric prestress force can be written as;

ft = (P/Ac) - (Pe/Zt),
fb = (P/Ac) + (Pe/Zb)

where; e is taken as positive if it is below the member

centroidal axis, Ac is the cross-sectional area and where ft

96
 CVX7640 – Structural Design

and fb are the stresses and Zt and Zb the section moduli for
the top and bottom fibres of the member respectively. The
sign convention used is that compressive stresses and
sagging bending moments are positive. If now an external
sagging bending moment M, is applied to the section, an
additional distribution of stresses is introduced and the
resultant stress distribution due to prestress force and
applied bending moment may be found by superposition.

ft = (P/Ac) - (Pe/Zt) + (M/Zt) Eq. 28.1(a)

fb = (P/Ac) + (Pe/Zb) - (M/Zb) Eq. 28.1(b)

If in addition to the applied bending moment at the

section there is also an applied axial load, then the force P
in Equations 28.1(a) and (b) is the sum of the prestress
force and the applied axial load.

Example 28.1
A simply supported pretensioned concrete beam has
dimensions as shown in Fig. 28.2 and spans 15 m. It has
an initial prestress force or 1100 kN applied to it and it
carries a uniformly distributed imposed load of 12 kN/m.
Determine the extreme fibre stresses at midspan; (i) under
the self weight or the beam, if the short-term losses are
10% and the eccentricity is 325 mm below the beam
centroid; (ii) under the service load, when the prestress
force has been reduced by a further 10%.
200

150

Cross section at mid span

for simply supported
350

prestressed beam in
Example 28.1
325
200

400

For the beam cross-section:

Ac = 2.13 x 105 mm2
Zb = Zt = 35.12 x 106 mm3
w = 5.1 kNm
Mi = 5.1 x 152/8 = 143.4 kNm
Ms = 17.1 x 152/18= 480.9 kNm
Pi = 0.9 x 1100 = 990 kN
Pi = 0.8 x 1100 = 880 kN

990 *103 990 *103 * 325 143.4 *106

(i) ft   
2.13 *103 35.12 *106 35.12 *106

97
CVX7640 – Structural Design

ft = 4.65 - 9.16 + 4.08 = - 0.43 N/mm2.

fb = 4.65 + 9.16 - 4.08 = 9.73 N/mm2.

880 *103 880 *103 * 325 480.9 *106

(ii) ft   
2.13 *103 35.12 *106 35.12 *106
ft = 4.13 - 8.14 + 13.69
= -1.42 N/mm2

-0.43 9.68

At Transfer At Service

9.73 -1.42
Stress distribution at the mid span of the Beam

The stress distributions shown in the Example 28.1 are

typical of those in a prestressed concrete member under
maximum and minimum loads, and illustrates an
important difference from reinforced concrete is that, with
prestressed concrete the minimum load condition is
always an important one. These four stress conditions
lead to a method of design for prestressed concrete
sections which will be discussed in further detail later.

So far the prestress force in a prestressed concrete

member has been considered to be provided by one layer
of tendons, so that the resultant prestress force coincides
with the physical location of the layer of tendons at each
section. However, there is usually more than one layer of
tendons in prestressed concrete members. In this case the
resultant prestress force coincides with the location of the
resultant of all the individual prestressing tendons, even
if it is not physically possible to locate a tendon at this
position.

For post-tensioned members where the duct diameter is

not negligible in comparison with the section dimensions,
due allowance for the duct must be made when
determining the member section properties. For
pretensioned members the transformed cross-section
should, strictly speaking, be used. However, in practice,
the section properties are generally determined on the
basis of the gross cross-section.

28.3 Additional steel stress due to bending

In the case of un-grouted post-tensioned members there is

no bond between the prestressing steel and the
surrounding concrete, but with pretensioned and grouted

98
 CVX7640 – Structural Design

post-tensioned members, bond is present and bending of

the member induces stress in the steel, as in a reinforced
concrete member. It is the bond between the steel and
concrete which makes the ultimate load behaviour of
pretensioned and grouted post-tensioned members very
similar to that of reinforced concrete members, and
different from that of un-grouted post-tensioned
members. The bond enables composite behaviour
between the steel and concrete to take place, and the extra
stresses induced in the steel at the serviceability limit state
may be determined by using the transformed cross-
section properties.

Example 28.2

The beam in Example 28.1 is pretensioned with tendons

having a total cross- sectional area of 845 mm2. Determine
the stress in the tendons under the service load.

All of the loads acting on the beam are resisted by the

transformed concrete section, as shown in Fig. 28 4. The
transformed area of the prestressing steel is mAps, where
m is the modular ratio Es/Ee. For typical values of Es and
Transformed Cross Ec of 195kN/mm2 and 28kN/mm2 respectively, m is
section at mid span for approximately 7.0.
simply supported
prestressed beam
It can be shown that y’ = 384 mm and that the second
moment of area of the transformed section is 1.38x1010
mm4. The eccentricity about the centroid of the
transformed section is 316 mm.
y’

The steel stress induced by the service load is given by.

y
f ps  mM s  M i 
I
= 7.0 x (480.9- 143.4) x 106 x 316/(1.38 x 1010)
mAps = 54 N/mm2.

The effective steel prestress after all losses have occurred

is given by;
fpe = 880 x 103/845 = 1041 N/mm2

So, the total steel stress fpb is now 1041+54 = 1095 N/mm2.
The extra stress induced by bending in this as well as
most other cases, is thus small, and could usually be
ignored.

28.4 Post-cracking behaviour

If the service load on the beam is increased, then the

tensile stress at the soffit of the beam will increase

99
CVX7640 – Structural Design

proportionately, until the modulus of rupture is reached.

If this is, say, 3.6 N/mm2 for the concrete in the beam in
Example 28.1, then the bending moment Mcr which will
cause this stress to be reached is given by
880 *103 880 *103 * 325 M cr *106
f b  3.6   
2.13 *105 35.12 *106 35.12 *106
Mcr = 557.5 kNm.

If the service load bending moment is increased beyond

this value, the concrete in the tensile zone must be
assumed to have cracked and the stresses in the section
must be found using a cracked transformed section, the
contribution of all concrete below the neutral axis being
neglected. It is assumed that there is still bond between
the steel and concrete, even though the concrete
surrounding the steel is cracked. This is the same
assumption that is made in reinforced concrete section
analysis.
εc ft

Fig. 28.2:
x

Stress-strain distributions Strain diagram Stress diagram

for cracked section
fps
εpe εp

The strain and stress diagrams for a cracked section are

shown in Fig. 28.2 and the procedure for a cracked-section
analysis is as follows:

(a) Choose a strain in the concrete extreme fibres, εc ;

(b) Choose a neutral axis depth, x ;
(c) Determine the concrete and steel stresses from the
relevant stress-strain curves, neglecting the concrete
in tension below the neutral axis;
(d) Check whether total compression equals total tension
within the section (for no applied axial load). If it
does, determine the moment of resistance of the
section. If not, go back to steps (a) and (b) and repeat
steps (c) and (d);
(e) Repeat steps (a)-(d) until the moment of resistance
equals the applied bending moment.

Note: that when determining the force in the prestressing

steel, the total strain εps comprises two components, that
due to the bending action of the beam εs and that due to
the effective prestrain in the tendon, εpe

A cracked-section analysis is extremely laborious, since

there are two unknowns, εc and x, and is best carried out

100
 CVX7640 – Structural Design

by programming. For a given value of εc, x may be found

from consideration of internal equilibrium. The resulting
moment of resistance is then compared with the applied
bending moment. The procedure is exactly the same as
that which will be used in the next section to analyse a
member at the ultimate load. There, the value of εc is
fixed, leaving the only unknown as x, which simplifies the
calculations considerably.

Cracked-section analyses are useful in determining the

steel stresses in Class 3 members, and these are used as a
means of checking that the crack widths are not excessive.

Example 28.3

For the beam in Example 28.1 use a cracked-section

analysis for an applied bending moment of 557.5 kN m to
determine the stress in the steel.
880 * 10 3
 pe   0.00534
845 * 195 * 10 3
The stress and strain distributions which give equal
tension and compression and also balance the applied
bending moment are found using the trial and error
approach outlined above, and are shown in Figure below.
0.000476 13.33
931.4 kN
Stress-strain distributions
423

599
Strain diagram Stress diagram

931.4 kN
0.00534 0.000312

From the figure, the moment of resistance is given by

Mt = 931.4 X 0.599 = 557.9 kNm
fps = (0.000312 + 0.00534) x 195 x 103 = 1102 N/mm2.

It has been assumed in the foregoing that the concrete

stress-strain curve is linear. Although this is a reasonable
approximation for the initial region of the stress-strain
curve, for the analysis of a cracked section at much higher
loads the full non-linear curves for both steel and concrete
must be considered.

28.5 Ultimate load behaviour

As the service load is increased still further, both the

concrete and steel stresses will increase, following the
respective stress-strain curves, which are the same form
as given in Fig 26.3 and 26.4 which are the design stress-
strain curves, with γm taken as 1.0. The design, or

101
CVX7640 – Structural Design

allowable, ultimate load on the beam, which incorporates

partial factors of safety, will be considered later.

As the applied load is increased, the strain and stress

distributions across the section change, the strain
distribution remaining linear as described in Section 28.1
The stress in the extreme fibres of the beam section
follows the stress-strain curve in Fig.26.3 (with γm = 1.0)
and will eventually reach the limiting value of 0.67fcu
although the extreme-fibre concrete strain continues to
increase, until it reaches its maximum value, εcu = 0.0035.
This strain has been found to be the average maximum
that concrete of all grades can withstand before crushing
of the material. At all times the total compression in the
concrete and the tension in the steel are equal (for no
applied axial load) and the moment of resistance is
always given by C*z or T*z, where C & T are the concrete
compression force equal to tension in steel respectively
with ‘z’ being the lever-arm between the two forces.

By the time the limiting concrete strain has been reached,

the total strain in the prestressing steel εpb can either be (a)
greater than ε2 in which case the steel will have yielded
before the concrete finally crushes - a ductile failure (such
a beam is termed under-reinforced); or (b) less than ε2, in
which case the steel will not have yielded before the
concrete finally crushes - a brittle failure (such a beam is
termed over-reinforced). If the steel strain equals ε2, then
the section is said to be balanced.

This situation is analogous to that in reinforced concrete

members at the ultimate limit state, the only difference
being that the initial strain εpe in the steel must be
considered. As with reinforced concrete members it is the
ductile failure which is desirable, since it is gradual and
gives ample warning. The load deflection curves for
typical under- and over-reinforced members will be
discussed later.

Example 28.4

Determine the ultimate applied load that the beam in

Example 28.1 can support if fcu = 40 N/mm2 and fpu = 1860
N/mm2.

The procedure is similar to that used in the cracked-

section analysis of Example 28.3 except that now the
extreme-fibre concrete strain is fixed at εcu = 0.0035. The
stress and strain distributions determined from material
properties are shown in the figure below.

102
 CVX7640 – Structural Design

0.0035 0.67fcu
0.566x C
x 0.00152
Stress-strain distributions 0.434x
at ultimate load z
Strain diagram Stress diagram

T
εpe εp fpb

d  x  
 pb  0.000534   0.0035
 x 
Where; ‘d’ is the effective depth or the tendon.

With the value of εc is fixed at εcu = 0.0035, ‘x’ is

determined by considering internal equilibrium. With
irregular sections it may be easier to determine ‘x’ by trial
and error rather than to find it directly. Note that the
concrete stress block is made up or two portions,
corresponding to the two portions or the concrete stress-
strain curve, one rectangular and one parabolic. The
dimensions of the stress block are dependent on the
characteristic strength or the concrete, and are shown in
Figure below for fcu = 40 N/mm2.
0.0035 26.8
1571.7 kN
171
Stress-strain distributions
at ultimate load for this 625
Strain diagram Stress diagram
example
1571.7 kN
0.00534 0.01083 1860

The ultimate moment of resistance of the section is then

given by;
Mu= 1571.7 x 0.625 = 983.3 kN m.

This corresponds to a uniformly distributed load of, 34.9

kN/m. The steel strain for this neutral axis depth is
greater than the yield strain, so this is an example of a
ductile, under-reinforced section.

It should be noted here that in the above example, the

load of 34.9 kN/m is that which would cause actual
failure of the beam. What is usually required is the
maximum safe, or design, load that can be supported. The
full analysis of the behaviour of a prestressed concrete
beam from transfer of prestress force to ultimate load has
been shown in Examples 28.1 - 28.4 to illustrate the basic
approaches to the analysis. For uncracked section,
analysis by elastic bending theory can be used. Once the
concrete has cracked, this theory is still applicable, except
that the cracked-section properties must be considered.
Once the concrete stress in the extreme fibres approach
the non-linear portion of the concrete stress-strain curve,

103
CVX7640 – Structural Design

then elastic bending theory is no longer valid, and the

analysis must be carried out using the three basic
principles stated in Section 28.1 Throughout, however, the
equilibrium conditions between internal stress resultants
and applied loads have been shown in order to
emphasize the same basic behaviour of the beam at all
levels of load.

28.6 Design ultimate strength

The uniform load of 34.9 kN/m determined in Example

28.4 is that which would cause physical collapse of the
beam. What is more usually required is the safe ultimate
load that the beam can support, that is the load which
gives an adequate factor of safety against failure of the
materials. This is found by introducing the partial factors
of safety for the steel and concrete material properties
described in a previous session. The typical stress-strain
curves for concrete and steel are now modified to those
shown in Fig. 28.3 with γm for concrete and steel taken as
1.5 and 1.15 respectively.
Stress
Figure 28.3: 0.45fcu
Design Stress-Strain curves for Concrete
(γm = 1.5) and prestressing steel (γm =
1.15) & Stress/Strain distribution for For Concrete
example 28.5

Stress

0.87fpu

0.70fpu Strain

For Steel 0.0035 0.45fcu

0.645x C
x 0.00124
0.354x
z
ε1 Strain diagram Stress diagram
ε2 Strain
T
εpe εp fpb

The stress-strain curve for concrete leads to the strain and

stress distributions within the section shown in Fig.28.3.
Again, the dimensions of the stress block are dependent
on the characteristic strength of the concrete, and are
shown in Fig. 28.3 for fcu = 40 N/mm2.

Example 28.5
Determine the ultimate moment of resistance of the beam
in Example 28 using the stress-strain curves in Fig. 28.3.

104
 CVX7640 – Structural Design

As in Example 28.4, ‘x’ may be found by equating the

total compression and tension forces. For this example,
the stress and strain distributions which satisfy this
requirement are shown in the figure below.
0.0035 17.9
1367.2kN
190
Stress-strain distributions
at ultimate load for this 603
Strain diagram Stress diagram
example
1367.2 kN
0.00534 0.00939 1618

It can be seen that the steel stress at failure of the beam is

equal to the yield stress and that therefore the mode of
failure is ductile.
The ultimate moment of resistance is given by;
Mu = 1367.2 x 0.603 = 824.4 kNm.

For an applied bending moment of this amount, the total

ultimate uniformly distributed load is 29.3 kN/m. The
allowable characteristic imposed load ‘w’ is now given
such that;
1.4 x 5.1 + 1.6 w = 29.3 then w = 13.9 kN/m

The characteristic service load is thus 1.6*w = 19.0 kN/m.

In order to illustrate the different approaches using the

serviceability and ultimate limit states, consider the beam
designed as a Class 1 member with zero tension under the
service load. For the same prestress force and eccentricity
used in these examples, the allowable service load can be
shown to be 15.3 kN/m For the ultimate limit state
approach, the service load is 19.0 kN/m and so this bears
out what was stated for Class 1, and for many Class 2,
members, design is usually based on the serviceability
limit state.

28.7 Simplified Concrete Stress Block

In order to simplify the calculations involved using the

concrete stress block shown in Fig.28.3 a simplified
rectangular stress block is given in BS8110, shown in Fig.
28.4. This stress block gives the same total concrete force
in compression as that in Fig. 28.3 and enables ultimate
strength calculations to be performed quickly by hand.
0.0035 0.45fcu
0.9x C
Figure 28.4: x
Simplified stress block as given
z
in BS 8110 Strain diagram Stress diagram

T
εpe εp fpb

105
CVX7640 – Structural Design

Example 28.6

Determine the design ultimate moment of resistance of

the beam in Example 28.1 using the BS8110 simplified
stress block.

By equating tension and compression within the section,

the neutral axis depth is found to be 211 mm, and the
steel stress, from Fig. 28.3 is equal to the yield stress. The
ultimate moment of resistance is then given by

Mu = 0.87 x 1860 x 845(700 - 0.9 x 211/2) x 10-6

= 827.3 kNm.

The agreement between this value and that in Example

28.5 is very close.

28.8 Code Formula and Design Tables

As an alternative to the two methods described

previously, which both use the stated basic principles, the
following formula for rectangular sections is given in
BS8110.
M u  f pb Aps d  d n 
Where fpb is the tensile stress in the prestressing steel at
failure, and dn is the depth to the centroid of the concrete
stress block, taken as 0.45x, provided that, in a flanged
beam, the flange thickness is not less than 0.9x. Values of
fpb and x for pretensioned and bonded post-tensioned
members are given in Table 28.1.

Table 28. 1 Values of fpb and x

f pu Aps fpb/0.87fpu x/d

fpe/fpu = fpe/fpu =
f cu bd 0.6 0.5 0.4 0.6 0.5 0 .4
0.05 1.0 1.0 1.0 0.11 0.11 0.11
0.10 1.0 1.0 1.0 0.22 0.22 0.22
0.15 0.99 0.97 0.95 0.32 0.32 0.31
0.20 0.92 0.9 0.88 0.4 0.39 0.38
0.25 0.88 0.86 0.84 0.48 0.47 0.46
0.30 0.85 0.83 0.8 0.55 0.54 0.52
0.35 0.83 0.8 0.76 0.63 0.6 0.58
0.40 0.81 0.77 0.72 0.7 0.67 0.62
0.45 0.79 0.74 0.68 0.77 0.72 0.66
0.50 0.77 0.71 0.64 0.83 0.77 0.69

The table may be used for rectangular beams and for T

beams where the neutral axis lies within the compression
flange. For T -beams where the neutral axis lies below the

106
 CVX7640 – Structural Design

flange, the more rigorous methods described previously

should be used.

Example 28.7

Determine the ultimate moment of resistance of the beam

in Example 28.1 using Code Formula and Table 28.1.

On the initial assumption that 0.9x  200, then

f pu Aps 1860 845
 . = 0.14
f cu bd 40 400 x700
f pe 880 * 10 3
  0.56
f pu 845 * 1860

From Table 28.1;

fpb=0.97*0.87*1860 = 1570 N/mm2
x = 0.30*700 = 210 mm (i.e. 0.9x =189 mm)

Therefore, according to Code Formula;

Mu = 1570 x 845(700 - 0.45 x 210) x 10-6 = 803.3kNm

28.10 Untensioned Reinforcement

It is usually found that the ultimate strength of Class 1

members is satisfactory, but with some Class 2, and with
most Class 3, members it is often found that the ultimate
moment of resistance based on the prestressing steel
alone is insufficient.

In such cases, either the section size should be increased,

or untensioned reinforcement should be introduced. In
order to find the moment of resistance of a section with
both tensioned and untensioned steel, the basic principles
stated in Section 28.1 are still used, except that the
relevant stress-strain curve for the untensioned
reinforcement must also now be considered. Such a curve,
using the characteristic yield strength of the
reinforcement, fy, is given in BS8110 and is shown in Fig.
28.5

Figure 28.5:
Design stress-strain curve for
reinforcement according to
the code BS 8110

107
CVX7640 – Structural Design

Example 28.8
Determine the design ultimate moment of resistance of
the beam section in Example 28.1, if 4T10 bars are added,
at the same level as the prestressing steel. Assume that fy
= 460 N/mm2.

The same basic stress block shown in Fig. 28.4 will be

used here, except that the extra stress in the untensioned
reinforcement must now be added. In this case the depth
of the neutral axis is best found using a trial-and-error
approach since it should not be assumed that both types
of steel have yielded. The results of this analysis are
shown in Table 28.2.

Table 28. 2 Neutral axis depth for beam in Example 28.8

X (mm) εpb εst fpb (N/mm2) fst (N/mm2) T(kN) C(kN)
200 0.0141 0.00875 1618 400 1493 1296
220 0.013 0.00764 1602 400 1480 1426
225 0.0127 0.00739 1590 400 1469 1458

The strain in the prestressing steel is given by εpb = εpc + εp

The strain in the untensioned reinforcement εs is equal to
εp since both types of steel are at the same level.

From Table 28.2 it can be seen that, with sufficient

accuracy, x may be taken as 225 mm. The depth of the
stress block is then 203 mm, but it is sufficiently accurate
to consider only the flange as in compression. The
ultimate moment of resistance is then given by;
Mu = 0.45 x 40 x 400 x 200(700 -200/2) x 10-6 = 864.0 kNm

The ultimate moment of resistance has thus been

increased from the value of 827.3 kN m found in Example
28.6.

The full analysis of sections with untensioned

reinforcement has been shown in the above example in
order to illustrate the basic principles. However, in most
practical cases, it is sufficiently accurate to replace the
cross-sectional area of this reinforcement by an equivalent
area Asfy/fpu and then to analyse the section using any of
the methods described previously.

The presence of steel, either tensioned or untensioned, in

the compression zone can be treated in a similar manner
to the strain compatibility method described above for
steel in the tension zone.

Untensioned reinforcement really becomes useful only

after the concrete in a section has cracked, and

108
 CVX7640 – Structural Design

particularly at the ultimate limit state. At service loads,

the stress in this steel is small and may even be
compressive, depending on the location within the
section. The addition of untensioned reinforcement is
very useful in limiting cracking and providing sufficient
ultimate strength capacity soon after transfer, when the
concrete is still immature. It is good practice to provide
untensioned reinforcement in any region of a member
where tension is likely to occur. It is of particular use at
the supports of members with straight tendons, where the
allowable tensile stresses, particularly at transfer, may be
exceeded. This reinforcement will also resist any cracking
that may occur due to accidental mishandling of
structural members. This may be caused, for instance, by
a simply supported beam being lifted at its centre.

28.11 Class 3 members

The three different classes prestressed concrete members

and the analysis for the serviceability limit state outlined
previously is suitable for Class 1 and 2 members, since it
is assumed that the concrete remains uncracked. For Class
3 members, however, the concrete is assumed to have
cracked, and the aim is to limit the crack widths to
acceptable levels, depending on the degree of exposure of
the member.

A procedure is given in BS8110 whereby it is assumed

that the section remains uncracked, and that hypothetical
tensile stresses exist corresponding to the specified
maximum sizes of crack. These hypothetical stresses are
shown in Table 28.3 and may be modified by the factors
in Table 28.4 which take into account the reduced
hypothetical tensile stresses which should be used with
increased depth of the member. If additional
reinforcement is contained within the tension zone, close
to the face of the member, the hypothetical tensile stresses
may be increased by an amount which is proportional to
the cross-sectional area of the untensioned reinforcement
expressed as a percentage of the cross-sectional area of
the concrete in the tension zone. For 1% of reinforcement
the stresses in Table 28.3 may be increased by 4.0 N/mm2
for Groups A and B and by 3.0 N/mm2 for Group C. For
other percentages, the stresses may be increased in
proportion, except that the total hypothetical tensile stress
should not exceed 0.25fcu.

The procedure given in BS8110 should really be regarded

as a preliminary design method, suitable for estimating
the section size and amount of prestressing steel. With

109
CVX7640 – Structural Design

this method, the tensile area of the concrete has a

considerable effect on the steel stress, while in reality the
stress in the steel is very little affected by this area of the
concrete. Tests have shown that crack widths are more
related to the steel stress than to the hypothetical tensile
stress in the concrete.

Table 28. 3 Hypothetical tensile stresses for Class 3 members

Group Limiting Design stress (N/mm2)
crack for concrete grade
width 30 40 50 ≤
(mm)
A Pre-tensioned tendons 0.1 - 4.1 4.8
0.2 - 5 5.8
B Grouted post-tensioned 0.1 3.2 4.1 4.8
tendons 0.2 3.8 5 5.8
C Pre-tensioned tendons 0.1 - 5.3 6.3
distributed in the tensile 0.2 - 6.3 7.3
zone and positioned close
to the tension faces of the
concrete

Table 28. 4 Modification factors for Class 3 tensile stresses

Depth of Member (mm) Factor
200 ≥ 1.1
400 1.0
600 0.9
800 0.8
1000 ≤ 0.7

Alternative design methods will be outlined in lesson 30,

but in all the design methods. the serviceability limit state
of cracking must be checked. This can be carried out
either by determining the crack widths directly, or by
finding the stress in the steel adjacent to the tensile face of
the member and checking that it is below a certain limit.
One crack-width formula (Concrete Society, 1983) is;
w = (0.08 x 4cfst /Es) in mm,

where ‘c’ is the cover to the untensioned steel and ‘fst’ is

the stress in the untensioned steel, or increase of stress in
the prestressing steel at the service load, determined
using a cracked-section analysis described previously.

As an alternative to using above equation, it is

recommended (Concrete Society, 1983) that the stress, fst
to be limited to 150 N/mm2 where the whole section is in
compression under the permanent load (usually the dead
load of the structure plus some proportion of the imposed
load). This stress may be increased to 200 N/mm2 if the
full service load is of short duration. In this case any
cracks which form will be temporary and will close on

110
 CVX7640 – Structural Design

removal of the extra load. The allowable compressive

stresses under service load, and all stresses at transfer,
should be the same as for Class 2 members.

Example 28.9
For the beam in Example 28.1 determine the maximum
service load, which the beam can support if it is designed
as a Class 3 member with a maximum crack width of 0.2
mm.

Ignoring the effect of the untensioned steel, from Table

28.3 the hypothetical tensile stress is 5.0N/mm2. From
Table 28.4 the modification factor, by interpolation, is
0.825. Thus the allowable hypothetical stress is given by;
Fht = 0.825*5.0 = 4.13 N/mm2

In order to find the maximum service load, the tensile

stress at the serviceability limit state is given by;
880 x103 880 x103 x325 M s x106
   4.13
2.13 x105 35.12 x106 35.12 x106

The corresponding uniform load is 20.5 kN/m and this

should be compared with the load of 15.3kN/m when the
beam was being treated as a Class 1 member.

The analysis of Class 3 members at the ultimate limit state

is identical to that of Class 1 and 2 members.

111
CVX7640 – Structural Design

112