Precast Concrete Terraces Under Static Incremental Loading; A laboratory investigation

The paper describes an experimental investigation of a set of L-shaped precast concrete terrace units subjected to static incremental loading to assess their structural performance and estimate their stiffness. A series of loading-unloading tests were carried out on uncracked (as delivered from the factory) and cracked (after the first complete loading-unloading cycle) units. The variation of parameters, such as displacements and strains, with the applied load was recorded and presented graphically. The reduction in stiffness of the units due to cracks was then estimated from these graphs.
The predominant mode of failure was found to be due to gradual propagation towards the top, of cracks starting from the soffit (tension zone) and around the symmetry line (where maximum bending occurred). The strain distribution with depth for the vertical “beam” part of the terrace unit was predominantly linear, with tension at the bottom and compression at the top. However, a large portion of the horizontal part of the unit (slab) followed closely the behaviour of the beam, to give tension rather than compression at the top. This could have important implications for the design of the units. Their deformed shape was considerably more complex than that assumed in the initial design, with downward displacements combined with rotation about a longitudinal horizontal axis and warping at the free slab corners.
A series of finite element models were developed, depicting closely the true behaviour of the units, to assist the study of the structural aspects that otherwise would not be easy to identify (such as, the formation of a 'bowl' at the central region of the units). It was concluded that the present methods and procedures of evaluating and designing precast concrete terraces were not sufficiently comprehensive. Further tests are required, combined with more rigorous analytical work and the establishment of benchmarks, in order to substantially minimise the uncertainties surrounding their performance in service.