CHAPTER 4

FLEXURAL DESIGN
(PART 3)
F. SECTIONS HAVING EXCESS CAPACITY
• In practical situations, very seldom will the concrete section
chosen have exactly the required values of Sl and S2, as
found by the methods of Case-A through D, nor will the
concrete centroid be exactly at the theoretically ideal level.
• Rounding upward of concrete dimensions, provision of broad
flanges for functional reasons, or the use of standardized
cross-sectional shapes will normally result in a member
whose section properties will exceed the minimum
requirements.
• In such a case, the stresses in the concrete as the member
passes from the unloaded stage to the full service load stage
will stay within the allowable limits, but the limit stresses will
not be obtained exactly.
• An infinite number of combinations of prestress force and
eccentricity will satisfy requirements.

• Usually the design requiring the lowest value of prestress

force, and the largest practical eccentricity, will be the most
economical.

• A graphical solution indicating all acceptable combinations

of prestress force and eccentricity is helpful in making the
best choice for a given concrete cross section.
• There are four limit stresses to be satisfied: two at the initial
unloaded stage and two at the full service load stage.
• The requirements may be restated as follows for a beam
with variable eccentricity, for example:

(a)

(b)

(c)

(d)
• These equations may be rearranged to give the inverse of
the initial prestress force as a linear function of eccentricity.
For the initial stage, from Eqs. (a) and (b), respectively:

(4.18)

(4.19)
• and for the service load stage, from Eqs. (c) and (d),
respectively:

(4.20)

(4.21)
• These functional relationships are plotted in Fig. 4.13 for a typical case.

• Equation (4.18) establishes a lower bound on l/Pi (i.e., an upper bound

on Pi) such that the tensile stress limit fti is not exceeded in the initial
stage. Any value of l/Pi above the line established by Eq. (4.18) is
acceptable, as indicated by the shading.

• Similarly, Eq. (4.19) establishes another lower bound on l/Pi such that
the compressive stress limit fci is not violated.

• Upper limits on l/Pi are established by Eq. (4.20), based on the service
load tensile stress limit fts, and by Eq. (4.21), based on the service load
compressive stress limit fcs.

• The zone indicating acceptable combinations of Pi and e that meet all

four requirements is shown shaded on the graph.
FIGURE 4.13 Variation of prestress force with eccentricity for beam section with
excess capacity.
• The maximum eccentricity that can be used without
violating any of the four stress limits is found at the
intersection of Eqs. (4.18) and (4.20) in Fig. 4.13. This also
corresponds to the minimum acceptable value of Pi (i.e.,
maximum l/Pi), and probably represents the most
desirable solution.
• In many practical cases, a maximum value emax would be
established by physical limitations, based on the available
distance c2 reduced by the necessary concrete cover for
the tendons, measured from the steel centroid. Given that
condition, in the example of Fig. 4.13, the best value of l/Pi
is that at the intersection of emax with Eq. (4.20).
• The actual plotting of a diagram such as Fig. 4.13 is
unnecessary in many practical cases if its general features
are recognized. For example, for the widely used single-T
and double-T beam sections, the eccentricity will often be
controlled by the maximum depth available from the
concrete centroid to the bottom face, less an allowance for
concrete cover for the tendons.

• For that maximum eccentricity, as seen from Fig. 4.13, the

minimum prestress force Pi will be governed by Eq. (4.20),
based on the allowable service load tensile stress fts. The
required prestress force can be calculated directly from Eq.
(4.20), setting e = emax.
• The concrete stresses in the initial stage and service
load stage should then be checked, and if stress limits
are not satisfied, the design can be modified after
plotting all four equations, as in Fig. 4.13, to determine
the best combination of Pi and e. However, this will
seldom prove necessary.
• The equations just developed, which pertain to beams
with eccentricity varying along the span, are easily
modified to apply to beams with constant eccentricity
by setting M0 = 0, its value at the controlling support
section, or to beams with deflected tendons by
substituting M'0 for M0.
EXAMPLE: Design of Beam Using Standard Cross Section having
Capacity above that Required
A post-tensioned prestressed concrete roof deck is to be
designed using 8-ft width single T beams spanning 70 ft
between simple supports. Service loads include live load of
88 psf, superimposed dead load of 10 psf, and estimated
self-weight of 75 psf. Normal weight concrete is used with
specified compressive strength fc' = 5,000 psi and strength
at time of prestressing f'ci = 3,500 psi. The prestressing
force will be provided using 1/2-in. diameter Grade 270
low-relaxation strands with fpu = 270 ksi and fpy = 243 ksi.
The effectiveness ratio can be assumed to be R = 0.85.
Select an appropriate standard single-T section beam, and
determine the required magnitude and eccentricity of
prestress force based on ACI Code stress limitations.
G. BEAMS OF LIMITED DEPTH

• In previous cases, the moment due to beam's

own weight was compensated by increasing the
tendon eccentricity by the amount Mo/Pi, below
the eccentricity that would produce the limiting
stresses, where Pi to be acting alone.

• The entire stress range fti - fcs at the top surface,

and the full stress range fci - fts at the bottom
surface, reduced only by the effect of loss of
prestress force, were available to resist the
superimposed load moment Md + Ml.
• Full compensation for member self-weight may
not be possible if the beam depth is limited for
architectural reasons, or clearance
requirements, or if there is a high ratio of self-
weight to superimposed loads, as would be the
case for long spans.
• The problem will become evident when the
required tendon eccentricity is calculated from
Eq. (4.6). The calculated eccentricity may
provide insufficient concrete cover for the
tendon, and in extreme cases, may indicate that
the tendon be placed outside of the concrete
section, obviously not a practical arrangement.
• It is possible to calculate the "critical span" for a
member of given cross section, up to which the self-
weight can be compensated by tendon eccentricity, and
beyond which it cannot.
• If the calculations indicate that for the concrete section
choosen, the desired tendon eccentricity cannot be
achieved, the section proportions should be modified if
possible, providing the same section moduli with a
section that is deeper and narrower and that can
accommodate the desired tendons.
• If it is not possible to make the section deeper, then the
section moduli must be increased for the member of
restricted depth, since the stress ranges available to
accommodate the superimposed loads are reduced.
• The equations for required section moduli are
found by :

(4.22)

(4.23)

where μ is a number, less than unity,

defining the fractional part of the
stresses due to the self-weight to be
accounted.
• Comparing Eqs. (4.1) and (4.2) with Eqs. (4.22) and
(4.23), it is seen that they differ only by the inclusion
of the factor μR in the last two equations.
•
• Since μ is not known at the outset of a design, it
must be assumed; a good initial guess is zero.

• The concrete centroidal stress and the initial

prestress force can be found by using the equations
of Section 4.2A without change:
• The maximum eccentricity emax is determined on the basis of the
minimum distance from the steel centroid to the bottom face of
the beam, based on requirements of tendon spacing and
concrete cover.

(4.24)
• The eccentricity is now set equal to the available emax and Eq.
(4.24) is solved for μ, the only unknown.
• If this value differs appreciably from that assumed in calculating
the section moduli, a revised value of μ is adopted and the
calculations repeated.
• It is clear from Eqs. (4.22) and (4.23) that the calculations are not
sensitive to changes in μ, and as a result, the iterative design
process converges to a satisfactory degree very quickly, often
in only one cycle.
 4.3 SHAPE SELECTION AND
FLEXURAL EFFICIENCY
• One of the unique features of prestressed concrete design is the
freedom to select cross-sectional proportions and dimensions
to suit the special requirements of the job at hand.
• The steel structural designer is limited to choosing from a
number of readily available, usually symmetrical, cross-
sectional shapes. In timber design, rectangular sections are
used almost without exception.
• In the case of prestressed concrete, not only can the member
depth be changed, but the web thickness modified and the
flange widths and thicknesses varied independently to produce
a beam with nearly ideal proportions for a given case.
• Particularly for post-tensioned beams of medium or long span,
the careful design of the cross section is an important part of
the total design process.
• For short span beams, in which the dead load of the
beam is likely to be only a small fraction of the total
load to be carried, rectangular members such as in Fig.
4.16a may provide the most economical solution,
because forming costs are minimized.

• For medium and long spans, the more efficient flanged

shapes of Figs. 4.16b through 4.16e are preferred.

• The steel centroid may be kept lower, for such sections,

without exceeding stress limits in the unloaded stage.

• The internal lever arm between tension and

compression resultant forces at the service load and
ultimate load stages is maximized.
FIGURE 4.16 Types of beam cross sections, (a) Rectangular
beam, (b) Symmetrical I-beam, (c) T-beam. (d) Inverted
T-beam. (e) Hollow-box beam .
• In choosing a section, often the most expedient procedure is to
start with a trial section, the properties of which match very
closely the desired values of S1 and S2.
• The tabulated section constants of Appendix A will prove
helpful in this connection.
• The trial section is then modified as required.
• With only two conditions to be satisfied (provision of the
required S1 and S2) and six independent section dimensions to
be chosen in general (Fig. 4.16), it is clear that there are many
possible solutions. The designer must choose the best.
• It is not always necessary or desirable to satisfy requirements
for S1 and S2 exactly.
• In some cases a broad top flange is desired to provide a useful
surface, as for bridge decks or building floors.
• Considering the flexural strength of a member, it is often desirable to
provide a generous top flange so that, should the beam be
overloaded, the concrete stresses will remain low, and a ductile failure
insured though yielding of the tensile steel.
• In such cases, the compressive stress at the top face of the beam at
full service load will be well below the allowed value.
• For long-span girders with a high ratio of self-weight to superimposed
load, there is little danger of overstressing the bottom flange in
compression at transfer, and the bottom flange may, in some cases, be
eliminated altogether, resulting in a T- section.
• In practice, the beam depth may be selected first, on the basis of a
desirable span-depth ratio, or from requirements of headroom and
clearance.
• The ratio of span to total depth for prestressed construction is
considerably higher than is usual for reinforced concrete.
Table 4.1 Typical Span-Depth Ratios for Prestressed Members

I-beams and single-T beams 24-36

Double-T beams 30-40

Bridge girders 25-30

One-way solid slabs 35-50

One-way hollow core slabs 40-50

Two-way solid flat plates 40-50

• Once the depth of a beam is chosen, it is desirable to have a
measure of the relative flexural efficiency of the cross sections
being compared.
• In general terms, the ratio of section modulus to concrete area,
S/AC will serve as such a measure.
• A beam characterized by a high ratio of S/AC would represent a
more efficient use of the material than one with a low ratio.
• For the general case of nonsymmetrical cross sections it is
equivalent to maximize the ratios Sl /Ac and S2/Ac since Sl = I/cl
and S2 = I/c2, these ratios can be rewritten by eqs (a) & (b)
respectively as follows:

(a)

(b)
• For given values of cl and c2, the most efficient
cross section is the one with the largest value
of radius of gyration, that is, the one in which
the concrete area is concentrated as nearly as
practical toward the extreme top and bottom
surfaces.

• The terms at the right of Eqs. (a) and (b) will be

recognized as the lower kern (k2) and upper
kern (k1) dimensions of the section,
respectively.
• It is convenient to express the distances represented by Eqs. (a)
and (b) in non dimensional form, in terms of the distances cl
and c2 to the top and bottom faces of the member. With kl =
r2/c2 and k2 = r2/c1 these ratios are

• Consequently, the single expression

•The efficiency factor Q can be used as a

convenient basis for comparing the flexural
efficiency of competing cross sections of a
given depth.
• The efficiency factor Q can also be expressed in geometric
terms. Noting that h = c1 + c2

indicating that the Q factor is nothing more or

less than the ratio of the total kern depth to the
total section depth.

• Clearly, those I- and T-shaped cross sections

with relatively thin webs and flanges will display
higher Q factors than will cross sections with
thicker parts.
• Practical considerations place an upper limit on the
degree of slenderness that can be achieved.
• The overall proportions of a beam must be chosen
considering the possibility of lateral buckling of the
loaded member, if it is not supported against lateral
movement by connected construction.
• This is particularly a factor during the handling of
precast members. Thin compression flanges always
present the danger of local buckling when loaded.
• Thin overhanging flanges are also vulnerable to
breakage during the handling and erection of precast
elements.
• The minimum thickness of the web is often
determined by minimum clearances needed for
the prestressing tendons and the auxiliary
reinforcement, and the requirements of
concrete cover for the outermost steel.

• Although web thicknesses of less than 5 in. may

be satisfactory to carry shear stresses with a
reasonable amount of web reinforcement, such
thin webs are often difficult to place without
risk of voids or air pockets.

• Normally a 5-in. web width should be

considered the practical minimum.
• Web widths are usually smaller in European
practice than in the United States.
• In general, well-designed I-beams have an
efficiency factor close to 0.50.
• Q factors less than about 0.45 indicate too
heavy a section, and values larger than about
0.55 indicate an excessively slender section
of questionable practicality.
• Many standards sections, such as, single and
double-T beams, have relatively low Q
values, but are used because of their
practical advantages.
 4.4 STANDARD SECTIONS
• Certain standard cross-sectional shapes have been evolved over
the years for floor and roof panels, wall panels, beams, and
columns, as well as for short and medium span highway bridge
girders.
• Members with these standard shapes can be mass-produced by
pre-casting plants, often using long-line methods and
permanent, reusable metal forms.
• Great cost savings are possible, compared with construction that
requires special forming, either in a pre-casting plant or at the
construction site.
• Standard sections are often used, even though properties may
not be optimum for a particular set of design constraints and
even though the section efficiency could be improved by shape
modification.
 4.5 CONCRETE PROTECTION AND SPACING OF
TENDONS
• To provide tendons, ducts, and reinforcing bars with adequate
protection against corrosion and fire, it is necessary to maintain
a certain minimum thickness of concrete cover outside of the
outermost steel.
• The thickness required varies depending on the type of
construction, condition of exposure, class of member, and type
of reinforcement.
• The ACI Code imposes the minimum cover distances for
prestressed concrete shown in Table 4.2.
• When tendons are enclosed in ducts, as is usual for post-
tensioned members, cover distances apply to the duct and the
metal end fittings.
• Normally, tendons are enclosed in transverse stirrup bar
reinforcement, in which case cover distances are measured to
the outside of the stirrups, and are somewhat less than required
for primary reinforcement, which may consist of both tendons
and non-prestressed reinforcing bars.

• If the prestressed member is designed for a service load tension

in excess of limits of Table 3.1, cracking of the concrete is likely
and according to ACI Code, the minimum cover requirements of
Table 4.2 must be increased 50 percent.

• For precast members, improved dimensional control in

construction of forms and placement of steel justifies some
reduction in cover for the non-prestressed bar reinforcement,
and special provisions of the ACI Code account for this.
• Normally, it is necessary to maintain a certain minimum distance
between parallel tendons or reinforcing bars to ensure proper
placement of concrete around them.
• Air pockets below the steel should be avoided, and full contact
between the bars or tendons and concrete is desirable to
optimize bond strength.
• The ACI Code specifies that the minimum clear distance between
adjacent bars in a layer should not be less than the nominal bar
diameter db, nor 1 in.
• Where the reinforcement is placed in two or more layers, the
clear distance between layers must not be less than 1 in. and the
bars in the upper layer should be placed directly above those in
the bottom layer.
• These provisions are interpreted to apply to prestressing strands,
as well as reinforcing bars, although "bundling" is permitted
• At the ends of pre tensioned members, spacing
requirements are increased to ensure proper
transfer of prestress force by bond.
• According to ACI Code, the clear distance
between pre tensioning tendons in the end
region must not be less than 4 db for wires and
3db for strands.
• Elsewhere along the span, it is permissible to
bundle the reinforcement, that is, to place
groups of no more than four tendons or bars in
contact so that they act as a unit.
• A unit of bundled tendons or bars is treated as a
single bar or tendon of diameter corresponding to
the total equivalent area.
• When post-tensioning ducts are arranged closely
together vertically, adequate spacing must be
provided to prevent the tendons, when tensioned,
from breaking through the duct.
• Experience indicates that a clear spacing of ducts
of one and one-third times the size of the coarse
aggregate and not less than 1 in. will be
satisfactory.