Civil Engineering Association

Objectives
• To understand the temperature-induced cyclic soil–structure interaction mechanisms governing the performance of integral bridges at the serviceability limit state. Particular attention is given to the build-up of lateral backfill soil pressure on the bridge abutments.
• To identify possible inadequacies in the current design advice note BA 42/96.
• To improve the knowledge base from which future designs will benefit.
Project
The integral bridge abutment wall investigated in this project is of the stiff wall type with a pinned base. Repeating temperature changes of the bridge create longitudinal movements in the bridge deck and rotation of the rigid abutment wall about its base. The magnitude of the wall rotations is governed by the bridge length, the type of deck construction and changes in bridge temperature (i.e. daily and seasonal temperature fluctuations). The escalation of lateral soil pressure behind the bridge abutments is then related to the wall rotations. The research project includes:
• an experimental programme for soil elements (Leighton Buzzard sand) tested under both cyclic stress and cyclic strain imposition, for a range of soil parameters;
• the formulation of a new constitutive soil model, for the prediction of cyclic stress–strain behaviour in particular;
• the results from a series of seven 1 in 12 scale model retaining wall experiments (to simulate typical seasonal wall movements in concrete bridges having lengths of 60, 120 and 160 m);
• a series of numerical simulations (50 cases in total) for a 7 m high abutment wall; and
• a comparative study relating bridge behaviour during seasonal temperature cycles only to bridge behaviour for the combined effect of daily and seasonal temperature changes.
The soil model, which is of the incremental type, has been developed specifically for the prediction of soil behaviour under plane strain loading. It has been validated against a comprehensive range of soil test data obtained under both monotonic and cyclic loading. The model was then incorporated in a semi-empirical retaining wall analysis for the modelling of a 7 m high stiff abutment wall. Bridge lengths of 40–160 m were considered together with a range of typical UK daily and seasonal temperature fluctuations. Computer predictions using the retaining wall analysis have been compared with experimental results from the model retaining wall.