Achieving buildability and robustness in precast concrete structures

Abstract

This research develops guidance on the rational assessment of structural robustness in precast concrete framing systems. The procedure evaluates first the behaviour of the local connections, followed by the global performance. The joint assessment is applied to two precast concrete framed building systems developed by Laing O'Rourke, namely, the M-Frame and the D-Frame. The robustness performance of the whole structure was done for the D-Frame system only.

The M-Frame incorporates a beam-to-beam moment resisting connection which is designed to be stronger than the connected members. The connection consists of top and bottom connector plates, which are welded to vertical dowel bars which are embedded in the connected beams. Satisfactory behaviour of the connection was established through laboratory testing. Procedures are developed for the design and micro-modelling of the connection. The thesis defines this type of connection as "strong" since the flexural strength of the connection exceeds that of the connected beams. This characteristic makes it equivalent to a monolithic beam.

The beams of the D-Frame are connected at a half-lapped joint, which was initially designed to meet the tie-based rules for robustness. The connection is defined as "weak" since following column removal, deformations are controlled by joint opening, with the connected precast members remaining relatively straight. Flexural continuity was initially provided by top and bottom plates, but preliminary analysis showed the joint to lack ductility. An alternative connection was developed in which continuity of flexural reinforcement across the joint is achieved using bespoke brackets. Laboratory testing showed that the extension of the connected bars can be controlled by debonding the bars to either side of the bespoke brackets. Localised failure by the bar bending was identified. A component-based model is developed to capture the response of the joint.

Performance-based analysis of the whole D-Frame system is carried out. Sudden column removal is chosen as a threat-independent damaging event. The pseudo-static resistance is used as a measure of structural robustness. This was done at different structural idealisation levels to identify the contribution of individual members. The evaluation of different half-lapped joint connections was done for an individual damaged beam. The bracketed connector exhibited the highest performance. Spatial effects were then considered by modelling the entire floor system in 3D. It is shown that the proposed D-Frame system can arrest the progression of the damage under sudden column loss. Further optimisation of the system and connections is recommended.