Development of a demountable precast concrete frame building.
Thesis DisciplineCivil Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Present generation reinforced concrete (RC) and precast concrete buildings are mostly monolithic. As a result, these buildings have to be demolished when they are either obsolete in function or irreparably damaged in an earthquake shaking. Demolition leads to irrecoverable wastage of non-renewable building materials, which is against the philosophy of a sustainable building. The demolition process is environmentally unfriendly, requires careful planning to avoid any danger to nearby structures, and consumes a large amount of energy. Because of its monolithic nature, upgrading a concrete building to accommodate for any future changes is not easy. At the same time, a monolithic concrete building has a very limited structural flexibility (i.e. allowance to remove/replace existing building components or to add new building components). Because of this, a damaged monolithic concrete building which is in a repairable state requires considerable time and money to restore its full functionality. This induces substantial seismic losses contributed by direct repair cost, and more significantly by the downtime (i.e. occupancy interruption). In this research, a conceptual layout of a next generation seismically robust precast concrete frame building system which uses “dry” and removable steel connections is developed. The proposed demountable building system is structurally flexible as it allows repair/replacement of existing building components or addition of new building components.
Experimental investigation is carried out in this research to identify the most suitable “dry” connection configuration to join the precast concrete beams and columns in the proposed demountable building system. Structural performance of the tested frame sub-assemblies with three different connection configurations are evaluated by comparing the test results with analytically predicted responses and conclusions are drawn with respect to their structural performance. To investigate whether emulation of structural behaviour of a wet jointed/monolithic concrete frame system can be achieved with the use of precast concrete beams and columns connected by using the proposed “dry” connections, experimentally obtained hysteretic plots of the tested frame sub-assemblies are compared with the hysteresis behaviour of a “wet jointed” or “ductile connectors” precast concrete frame sub-assembly available in the literature. Also, the feasibility of demounting a damaged beam and replacing with a new beam of the same (or higher) capacity and achieving the same (or better) performance compared to the original sub-assembly is assessed through this test programme. Qualitative analyses are performed on the proposed connections, and based on the comparison the connection which is easy to erect and dismantle is identified.
Numerical macro models are developed to simulate the hysteretic behaviour of the tested frame sub-assemblies with the proposed “dry” connections. Numerically simulated results with different hysteresis rules and different input backbone curves are compared, and issues with the macro models in capturing strength degradation are highlighted. The accuracy of the simulated cyclic behaviour of the sub-assemblies is assessed by comparing with the experimental test results and conclusions are drawn with regard to the reliability of the macro models. To eliminate any possible damage to the ground storey columns and to avoid the need to replace the damaged columns, feasibility of pin base (instead of fixed base) columns for unbraced lateral load moment resisting frames is investigated. Different lateral load resisting frame options that render the proposed demountable building system a “low downtime” building system are then ranked in an increasing order of “structural flexibility”.
A simple hand calculation method to estimate lateral stiffness and fundamental period of a frame building (braced or unbraced) with either fixed or pin bases is theoretically derived in this thesis. To verify the reliability of the developed equations, the estimated lateral stiffness and fundamental periods are compared with the corresponding values obtained from pushover and Eigenvalue analyses (and also Rayleigh method) for a wide range of low to medium rise buildings with varying cross-sectional and geometrical configurations. Suitability of the proposed equation for frames with linear variation of section properties along the building height and simple/pin beam-column connections is also scrutinized. Reliability of the developed equation to estimate the fundamental period is also verified using the experimental test and numerical analysis results available in the literature. Since the developed equations are found to closely predict the lateral stiffness and fundamental period of a frame building, they are recommended to be used in everyday design practice.