Checklist for the design of Integral Bridges

using Precast, Prestressed Beams

BD 57 states that bridges with an overall length of less than 60m and with a skew not
exceeding 30° should be designed as integral, unless there is a good reason not to in
particular situations. This checklist has been prepared to assist with the design of such
bridges within Faber Maunsell and to avoid “reinventing the wheel”.

Preliminary Sizing
1. Select a beam size.
2. Calculate moments for self-weight of beam, assuming simply supported..
3. Calculate moments for self-weight of insitu slab, assuming simply supported.
4. Calculate moments due to surfacing, assuming simply supported.
5. Calculate moments due to live load, allowing for continuity at abutments and intermediate
piers.
6. Calculate prestress required.
7. Adjust beam size if required and re-calculate.

Detailed Design
8. The structure should be modelled using LUSAS. The deck should be modelled as a
grillage or other appropriate representation, e.g. plate and beam. The abutments and
intermediate piers should be modelled appropriately to reflect their structural behaviour,
e.g. an abutment wall can be modelled as a plate.
Piles should be modelled as beam members fully fixed at a distance below ground level
of:
4 x diameter for firm over-consolidated clays
5 x diameter for dense granular soils
6 x diameter for loose granular soils; and
8 x diameter for soft normally consolidated clays.
This distance ensures that the flexibility of the piles is correctly modelled, but will not
give the correct forces in the pile below ground level. (See section 27).
For piles embedded in rock, it is recommended that the point of fixity be taken as one
diameter below rock-head level.
When the abutment is founded on a spread-footing, the bottom of the wall should be
modelled as fixed in the horizontal direction and with appropriate vertical and rotational
spring stiffnesses based on the stiffness of the underlying ground and the width of the
footing. When the abutment is founded on rock, the foundation should be modelled to
rotate freely and the footing detailed to be as narrow as possible and allow for rocking
without spalling of the lower corners. Alternatively, the complete footing can be
modelled with appropriate vertical springs to represent the soil. Footings should be kept
as narrow as possible, with toes about one quarter to a third of the total width. See also
section 26.

Mahgoub: Concrete Integral 1 of 5 Rev B

9. Calculate the composite section properties of the beam and slab. For hogging, only the
transformed area of the reinforcement in the slab should be considered.
The strands in the bottom of the beam will mainly be determined by the worst sagging
(at midspan) and the strands in the top will be mainly determined by the worst hogging
(near the support, but the moment actually at the support will be carried by the insitu
concrete as an rc section since this will be at the end of the beam). Curtailment may be
determined by the maximum hogging or sagging conditions, as well as the transfer
condition.
10. Calculate initial and final stresses due to self-weight of beam, allowing for primary
creep (due to the cross-section becoming composite) and secondary creep (due to
continuity at abutments and intermediate piers). This should be based on a realistic
assumption of the age of the prestressed beams when the structure is made continuous.
In the absence of more specific data, an age of 28 days may be assumed. The creep
factor, , should be calculated assuming that the loading is applied at 3 days, but that
the relevant creep starts at the time the section becomes composite (i.e. for BS5400:
Part 4, Appendix C, km is based on 3 days and kj taken as k j ()  k j ( 28 days) ). The
forces induced by the restraint of the free creep movement should be multiplied by a
reduction factor,  cc  1  e  , to allow for the moderating effect of the restraint forces
on the rate of creep. The actual value of cc is affected by concrete aging and the
relative values of  for the precast beam and insitu slab, but it is recommended that
these effects be neglected as any adjustments are likely to be less than the normally
expected variations in creep values. Typical values for (t=∞) range from 1.65 (fcu =
50MPa) to 1.40 (fcu = 70MPa).
11. .Secondary creep effects should be calculated from the LUSAS model by applying axial
and flexural strains corresponding to the creep strains which would occur if the structure
were statically determinate. These can be input using equivalent axial (T) and
differential temperatures (T/h), if the program does not allow the direct input of strains,
i.e.
N ΔT M
T  and 
EA h EI
It is important to check that the data is correctly specified for the program being
used. In LUSAS the differential temperature should be specified as a temperature
gradient, e.g. °C/m. If in doubt, it is recommended that a single built-in beam, with a
depth not equal to unity, is modelled and loaded, and the output checked against a hand
calculation.
12. Calculate initial and final stresses due to prestress, allowing for short and long term
losses, primary creep and secondary creep due to continuity at abutments and
intermediate piers. Use the same creep factor as calculated in section 9. Some
designers prefer to consider the combined effects of beam self-weight and prestress as
these two loads are applied at the same time.
13. Calculate stresses due to differential shrinkage (primary & secondary), allowing for the
1  e 
relieving effect of creep, i.e. using a reduction factor of . This should be based

on a realistic assumption of the age of the prestressed beams when the insitu slab is
cast. In the absence of more specific data, an age of 28 days should be assumed. 

Mahgoub: Concrete Integral 2 of 5 Rev B

 should be calculated assuming that the age at loading is when the section becomes
composite.
14. Decide whether beams will be made continuous before the insitu slab is cast. If this is
done there will be a benefit initially, but this will mostly dissipate with the effects of creep.
There may also be a Health & Safety benefit as the beams will be stabilised by the
permanent works earlier in the construction process.
15. Calculate the initial and final stresses due to self-weight of insitu slab, allowing for the
construction sequence and for primary creep and secondary creep due to continuity at
abutments and intermediate piers.  should be calculated assuming that the age at
loading is the age when the section becomes composite. In the absence of more
specific data, an age of 28 days should be assumed, (i.e. for BS5400: Part 4, Appendix
C, km is based on 28 days and kj taken as k j () =1.0). Typical values for (t=∞) range
from 1.70 (fcu = 50MPa) to 1.40 (fcu = 70MPa).
16. Creep will also take place between the section being made composite and the
application of live load after the bridge has come into service. This can be significant
and may be considered by the designer. However, a check should be made that no
heavy construction traffic will use the bridge in the meantime. Typically, for a bridge
which is made composite 28 days after casting the beams, the percentage of the total
load induced by creep that will take place in the following month is:
25% due to the self-weight of the beam and prestress, and
55% due to the self-weight of the slab.
17. Calculate the stresses due to self-weight of surfacing and other superimposed dead
loads, allowing for continuity at abutments and intermediate piers. It can be assumed
that these stresses will not be affected significantly by creep providing the structure is
not changed after this loading is applied.
.
18. The earth pressure distribution behind the abutments and attached wing walls should be
calculated using k* and k0 in accordance with BA 42/96 Amdt no.1, May 2003. These
pressures should be applied to the abutment members of the LUSAS model above
ground level. The forces and moments in the structure should then be calculated. Two
loadcases need to be considered – where the earth pressure is disadvantageous, k*
should be used; where the earth pressure is advantageous, it should be based on k0 or
0.5 x k*, whichever is the smaller. No live load surcharge need be considered, except as
indicated in the following sections. Earth pressure is a permanent load. It can be
assumed that the stresses due to earth pressure will not be affected significantly by
creep providing the structure is not changed after this loading is applied.
Wing walls attached to abutments should be kept as small as possible, particularly their
projection perpendicular to the deck centre-line, to minimise the amount of soil and
structure that has to move with the abutment during thermal expansion of the deck.
Where the earth pressure on the abutment is based on k*, a pragmatic approach is to
base the corresponding pressure on the wing wall on an effective coefficient, k’, given
by:
k’ = k*sin2 + k0cos2
where  is the angle between the wing wall and the longitudinal axis of the bridge
19. Calculate the effects of differential settlement. This is a long-term effect and should be
calculated using a long-term stiffness for the structure or the 28-day stiffness and a

Mahgoub: Concrete Integral 3 of 5 Rev B

 1  e 
reduction factor of . The same value of  as used in section 15 should be

adopted.
20. Calculate the stresses due to live load (HA and HB) and the accidental wheel load,
allowing for continuity at abutments and intermediate piers. (Remember the fL factors at
SLS.)
21. Overall thermal strains are given in BA 42/96 and are used in calculating the earth
pressures to be considered (see section 0). The free-standing structure, i.e. without
earth pressure loading should be analysed for the restraint of these strains, assuming
that they can be either positive or negative. The positive temperature case (i.e.
expansion) should be combined with earth pressure loading due to k* and live load on
the structure, see also section 26 for consideration of the abutment walls. The negative
temperature case (i.e. contraction) should, conservatively, be combined with zero earth
pressure and no live load.
22. Calculate the stresses due to differential temperature, allowing for primary effects and
secondary effects due to continuity at abutments and intermediate piers.
23. Overall thermal effects (section 21) and differential temperature effects (section 22)
should be combined in accordance with BD 37/01 clause 5.4.5.2, using the appropriate
partial factors for load combination 3.
24. Longitudinal forces in accordance with BD 37/01 should be applied to the deck. No live
load horizontal forces need be considered on the soil behind the abutments. The
longitudinal forces shall be combined with an earth pressure based on 0.5 x k* on the
abutment resisting the longitudinal force, and an earth pressure on the other wall (acting
in the direction of the longitudinal load) based on ka and a live load surcharge. The
loads shall be combined using the partial load factors for load combination 4. If the
difference between the total soil forces on the abutments is greater than the longitudinal
force on the deck, no further consideration is required.
25. The deck should be designed at SLS and checked at ULS.
It is often not possible to comply with Class 1 serviceability requirements of BD 24 in
hogging regions. At integral abutments and over continuous supports, it is acceptable in
the UK (see BA 42/96) to design prestressed pre-tensioned beams as reinforced
concrete providing due allowance is made for compressive stresses due to prestess.
Note that in the Republic of Ireland the beams should be designed as Class 2 in these
regions.
26. The abutment walls and intermediate columns should be designed to resist the forces
obtained from the LUSAS analysis.
For loadcases where the abutment wall rotates into the fill and where the base slab has
a significant length of heel, consideration should be given to base fixity when the bottom
of the abutment has been modelled as pinned. Conservatively, the frame analysis may
be re-run with the base fully fixed for these loadcases. For forward rotation of the
abutment wall away from the fill, the length of the toe is unlikely to be long enough for
base fixity to need consideration.
27. Design of Piles
The method of modelling the piles described in section 8, will give the correct moments
and forces in the piles above ground level. Below ground level, the forces in piles
founded in soils must allow for soil/pile interaction. A simple way of doing this is to use

Mahgoub: Concrete Integral 4 of 5 Rev B

 the coefficients derived by Reese and Matlock (see Tomlinson “Pile Design and
Construction Practice”). Where the shear force and bending moment in the pile at
ground level act in the same sense (i.e. the shear force will induce moments acting in
the same direction as the applied moment), the maximum bending moment in the pile
can be taken as:
for M ≥ VT, 0.9M + 0.68VT ≥ M
for VT ≥ M, 0.82M + 0.76VT.

See Tomlinson for the definition of T  5 EI pile / n , where n = the coefficient of horizontal
subgrade modulus variation.
28. Design of Spread Footings.
Since the top of the abutments are propped by the deck, stability checks are confined to
checking that the footing has sufficient sliding resistance, e.g. by taking moments about
the centre-line of the deck. Earth pressures should be based on ka and should include
live load surcharge. A load factor against sliding of 1.5 is required.
The distribution of bearing pressure under the footing depends on many factors and is
difficult to determine. Therefore, the following pragmatic approach is recommended.
The mean bearing pressure under the footing should be calculated including both the
vertical force in the abutment wall and the self-weight of the abutment wall, the footing
and the fill immediately above the footing. It should not exceed 0.5 x the maximum
allowable peak bearing pressure.
The reinforcement in the footing should be designed assuming that the soil pressure
under it has a triangular distribution, with the peak pressure assumed to occur at either
the heel or toe of the footing.
29. Design of Wing Walls
Independent wing walls should be designed for stability under active pressures (ka) and
live load surcharge. The reinforcement should be designed for the greater of:
at-rest pressures (k0) and live load surcharge, or
pressures due to compaction, as proposed by Ingold “A Reconsideration of
Retaining Wall Design”, The Structural Engineer, Vol 60B, No. 4, Dec 1982.
The reinforcement in wing walls cantilevering from the abutment should be designed for
earth pressures corresponding to the greatest of:
k’ (see section 18);
at rest pressures (k0) and live load surcharge; or
pressures due to compaction, as proposed by Ingold (ibid.)

Mahgoub: Concrete Integral 5 of 5 Rev B