Keywords: seismic, precast, concrete, wall, shaking table, nonlinear, inelastic, panel, finite element.

The satisfactory performance of large precast panel structures during strong earthquakes attracted attention towards their potential use in seismic zones. Mass production of typical structural components and repetitive erection procedures enhances the reliability of their performance. However, their economic advantage must not be compromised with complicated seismic-resistant connections. Significant progress has been made in understanding the dynamic behavior of precast panelized structures and the development of a rational seismic design procedures. Much of the ongoing research is focused on the weak links in such structure namely; the horizontal and vertical joints. The challenge is to develop joints that are simple yet seismically viable. One of the most important experimental schemes to study seismic performance is the shaking table tests of large assemblages. Such tests, induce in the structure seismic forces defined by the dynamic structural characteristics, and allow for correlation between peak base acceleration and the extent of damage in the structure. Most importantly, the recorded time histories of the various response parameters are used to verify the validity of proposed analytical and computer models. Subsequently, valid analytical models can be used to investigate the effects of different design features and controlling response mechanisms.

This paper reports on the correlations studies carried out using the shaking table recorded results of a one third scale three-storey precast panel wall, connected together using a proposed seismic horizontal joints. The test models considered , are simple walls subjected to two levels of shaking; 0.180g peak and a 0.67g peak. El-Centro NS earthquake record was used in the shaking tests. The test setup allowed for the bottom horizontal joint to be the focal point for measurement and for the subsequent finite element modeling during the severe shaking.

The computer program DRAIN-2D was used in the correlation studies. Most of the damage during the severe shaking took place at the bottom horizontal joint as intended. So, the nonlinear inelastic model was comprised of primarily elastic elements for all the panels and the upper joints and an inelastic horizontal joint model for the lower joint. The joint model utilized dimensionless discrete spring to model the joint's rocking and slippage mechanisms. A number of new spring elements were developed and incorporated in DRAIN-2D. The final correlating model accounted for a number of mechanisms in the horizontal joint that had a pronounced effect on the seismic response of the structure, such as; steel softening, steel buckling , and shear compression interaction in concrete.The inelastic joint model model is a set of dimensionless axial springs arranged in parallel across the width of the joint. The axial springs model the various materials within the joint that resist the overturning moment and shear force. The force deformation characteristics of these springs are determined based on the constitutive material properties. Three different types of springs are used to develop the initial inelastic joint model. Two mechanisms of resistance need to be modeled; overturning mechanism and shearing mechanism. Based on the fact that no shearing distress was observed and the fact that the edge keys inhibited any slippage as proven by the measured slippage across the joint, the shear mechanism was idealized using linear elastic horizontal springs. The final joint model retained its basic spring-elements arrangement employed in the initial model. However, the joint rotational mechanism was refined to account for the number of mechanisms that were anticipated to improve the correlation capabilities. New spring elements were developed and included in the final joint model. The new improvements can be summarized as follows: Concrete compression softening due to continual gap opening and closing; Concrete compression softening due to inelastic deformation; Modeling of reinforcement bar at the exact location they occupy in the actual wall; Modeling of rupturing using bilinear 4-node spring elements; Modeling of buckling using bilinear 4-node spring elements. The last correlation run, included two rupturing bars and buckling bar on the right side of the joint. Bilinear spring elements were used for all three bars on the left side. The correlation was satisfactory Two levels of correlation was used: ability to predict maximum response values; this was achieved easily: ability to correlate favorably with the recorded response time history; this is much harder to achieve for the full length of the signal. The analytical studies carried out in this study, lead to the following conclusions. The nonlinear inelastic model developed in this study is robust and can predict the peak dynamic response parameters reliably. The joint model with discrete spring elements arrangement is capable of model the joint rocking fairly well. It can simulate the joint gap continual opening and closing.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £135 +P&P)