Keywords: RC frame, performance, structural components, non-structural components, inter-storey drift, fuzzy-random analysis.

It has become a clear direction in future seismic engineering practice to design building structures meeting desired performance targets. The performance of buildings is affected by structural as well as non-structural components, and involves numerous uncertainties and vagueness. From the engineering point of view, damage to typical non-structural components incurs at a much lower level of deformation than that could induce structural damage. Furthermore, significant monetary losses may take place due to the interruption of function of the building as a result of substantial non-structural damage. Therefore, information about the performance of non-structural component is an integral part of the performance-based engineering.

In general, damage to structural components may be regarded as strain-related while damage to non-structural components is more drift-related. Strain-related criteria may as well be converted to drift criteria for framed structures. Therefore, in the present study, the performance levels for both structural and non-structural components are quantified in terms of limit inter-storey drift ratios.

Following a brief description of fuzzy set theory, a fuzzy-random model for the performance reliability analysis of RC framed structures considering both structural and non-structural damages is introduced in this paper. The limit state for each performance level is defined as an interval of inter-storey drift ratios concerning respectively the non-structural and structural damage with a membership function, while the relative importance of the two aspects is reflected through the use of an appropriate cost function. The fuzziness comes from the difference between non-structural and structural limit drift ratios while the randomness is associated with the probabilistic properties of these two limit drift ratios. Two membership functions are proposed to represent the occurrence and the exceeding of a performance level, respectively. Thus, the different non-structural and structural limit criteria can be considered in a comprehensive and computable manner for reliability analysis of the building performance. By assigning appropriate value to the -coefficient in the cost function, the relative importance of the non-structural and structural damage at a target performance state can be conveniently adjusted. An algorithm which can produce a boundedness, monotonicity and convexity preserving second-degree Bernstein polynomial is used to evaluate the membership functions.

At the present stage, the non-structural damage is represented by infill masonry walls, and an extensive survey of experimental data is conducted to establish the probabilistic description of their performance limits. On the structural side, the results from a previous study on RC column drifts will be considered as the tentative probabilistic drift model for structural components. Assuming a set of drift demand data, an example application of the proposed model is performed. Comparing with the classical reliability model based on single-threshold performance definition, the proposed model provides a good basis for incorporating different aspects into the performance assessment of a building system.

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)