Civil Engineering Reference
In-Depth Information
To put this possible misalignment into context, consider a 3 m deep bridge deck
being built on a vertical radius of 10,000 m, in segments 20 m long. The height of the
chord of the 20 m long segment of the circle is only 5 mm. If the soffi t form were 1 mm
out of true, giving a 6 mm height to the chord, the bridge would be built to a radius
of 8,333 m. As the piers would defi ne the correct vertical radius, the deck would be
subjected to a sagging moment as it was launched. On the assumption that the top and
bottom fi bre distances were respectively 1 m and 2 m, in the short term, the stresses
on the top and bottom fi bres would be respectively +0.7 MPa and -1.4 MPa (using
E
/
R
=
/
y
). In the longer term these stresses would reduce to less than a third of these
initial values due to the effect of relaxation, and would become insignifi cant (
3.9.2
).
However, during launching this very small geometric error signifi cantly increases the
risk of the section cracking. This example underlines the importance of designing a
section that is reinforced to behave in a ductile manner as explained below.
σ
15.8.4 Launching stresses
The choice of the limiting bending stresses to be observed during launching is
fundamental to both the satisfactory behaviour of the bridge in the construction phase,
and to its economy.
Some authorities insist on there being no bending tensile stresses in the deck during
launching under the combination of self weight, temperature gradient and differential
settlement. This leads to an excessive amount of fi rst stage prestress, which is both
uneconomic and gives a false sense of security. If the deck is designed to be fully
prestressed during launching, there is considered to be little justifi cation for providing
any more than nominal longitudinal passive reinforcement. This leaves the deck brittle,
and highly vulnerable to severe cracking in the event of unexpected bending moments.
Such cracking may or may not require repair, but inevitably halts construction while
its causes are investigated.
It is more rational to reduce the central prestress and to increase the passive
reinforcement to render the deck ductile. For a bridge which is designed for a conser-
vative client and is designated as Class 1 or 2 in service, the launching prestress may be
designed to deliver zero tensile stresses under self weight plus predictable settlement
effects. The effects of temperature gradients and ±3 mm of misalignment as described
above should then be checked at Class 2 (tensile stresses not exceeding -2.5 MPa).
The top and bottom slabs should be reinforced longitudinally with a minimum of
0.6 per cent of the slab area, providing for ductile behaviour in direct tension in the
event of cracking (
3.7.2
). Such a design will not crack if the construction is carried
out to a high standard, but if tolerances are exceeded, cracking will be controlled
and within specifi cation. However, it would be better to design the deck as partially
prestressed during launching, with crack widths limited to 0.25 mm. This will reduce
the central prestress and increase the longitudinal reinforcement, rendering the bridge
more ductile. Once the second stage prestress is installed, there is no reason such fi ne
bending cracks should not close, when the deck reverts to Class 1 or 2.
If the bridge deck is designed to be partially prestressed in service, then partial
prestressing should also be adopted during launching. It must be checked that there
is suffi cient reinforcement in the top and bottom fl anges to limit crack widths to the
specifi ed value and to provide for ductile behaviour at all points along the deck.
