Seismic Assessment of Pre-1936 Dual RC Wall-Riveted Steel Frame Buildings

Abstract

In the doctoral research project reported herein, the relative earthquake hazard vulnerability of dual Reinforced Concrete (RC) wall-riveted steel frame buildings constructed in New Zealand prior to 1936 was assessed. Based on a preliminary assessment of the strength and ductility capacity of the lateral force resisting components of an existing building, it was identified that the seismic behaviour of these buildings is dictated by the performance of the structural walls. Therefore, an experimental and numerical investigation was undertaken on the behaviour of wall components (wall-piers and spandrels) having reinforcement detailing similar to those found in an existing pre- 1936 building. The material and geometrical characteristics of the experimentally investigated wall components were replicated from those of the case study building. The lateral force resisting capacity of these components was found to be limited by their flexural strength. From the experimental investigation, it was found that the lateral force resisting capacity and the stiffness of some of these components degrade rapidly, due to the low quantity of reinforcing bars provided within the components. However, the wall components having spliced longitudinal reinforcement performed better than was expected, with strength and ductility capacity that were comparable with those components having continuous longitudinal reinforcement. The results of the experimental investigation were compared with provisions contained in assessment documents that are currently in use in New Zealand and the USA for the seismic assessment of existing RC walls. The provisions for the required splice length of plain round bars that are typically found in pre-1936 RC buildings were found to be excessively conservative, and thus underestimate the strength of walls having spliced longitudinal reinforcement. Additionally, the provisions contained in these documents were found to overestimate the effective stiffness of pre-1936 walls that carry no or a low magnitude of axial loads. Consequently, strength and effective stiffness values were proposed herein based on the results of the experimental investigation. Numerical models of the experimentally investigated wall components were developed using a finite element modelling package and were verified against data obtained from the experimental program. The numerical models resulted in prediction of the initial stiffness, the lateral force resisting capacity and the stiffness and strength degradation characteristics of the experimentally investigated components to an acceptable accuracy. The research project culminated by providing guidance for practicing structural engineers seeking to undertake seismic assessment of pre-1936 dual RC wall- steel frame buildings using the Nonlinear Dynamic Procedure (NDP). The procedure demonstrated a seismic response that could not be identified when using simplified analysis procedures. The detailed assessment also highlighted the adverse effects that the characteristics and location of the walls can have on the seismic performance of these buildings through introduction of significant torsional and vertical irregularities.