LRFD Temperature Multiplier

[mcox]

We are working on a three span deck truss design which has hollow stem piers. These piers were designed using the LRFD design specs. After the design was complete we decided to cross check it with the Group V loading of LFD and found that the hollow stem pier would require approx. 25% more reinforcing steel with the LFD design. It appears that the 0.5 multiplier for temperature forces in the LRFD strength checks is the reason why there is such a difference between the two codes. Could you provide any insight into the use of the 0.5 multiplier for LRFD as compared with the Group V LFD loadings. At present, there is quite a disconnect between the two situations. The disconnect disappears if the 1.2 factor for displacements is used for the LRFD situation. Did the AASHTO Committee make a conscious decision to reduce force effects due to temperature changes from the LFD specifications?

 

[wiktor]

The expansion and contraction of bridge spans, as looked upon based on the extensive database of existing bridges, yields some surprising conclusions:
Most of the existing expansion bearings installed are not moving or their movement is much smaller expected. Yet the bridge is expanding and contracting and usually no overstress could be noted. In most cases, the displacement is accommodated by elastic and non-elastic response of the supports. In other words the horizontal stiffness of the support at the level of the bearings is lesser than the horizontal stiffness of the bearing itself. In this case the pier and the bearing will accommodate the displacement of the support, proportionally to their stiffness. This phenomenon could be easily observed on older bridges equipped with steel rockers or sectional guided rollers. In most cases the bearings are not moving, or their movement is impaired.

The displacement of the span takes place, as predicted, and a factor of 1.2 would provide for adequate provision for rare occurrences of more extreme conditions.

The situation is completely different with the forces generated in the structure by these displacements. Our typical bridge is modeled as a frame, with the fixed points at the level of footings. The footings are supported by soils (or piles), which normally have quite low Young modulus. But our model considers these points to be indefinitely rigid, so our moments and shears are quite large. In reality the footings will rotate, forces are not so large, so the factor of 0.5 is quite reasonable.

There will be some special cases when the designer should use different factors – for example when the model will include soil springs acting on the piles or footing or when the bridge is founded on solid rock - in such case a factor of 1.2 for temperature and shrinkage, as for displacements should be justified.