Cost-Benefit Importance Vectors for Performance-Based Structural Engineering

This paper offers importance vectors to identify the structural parameters that are most cost efficient to change to improve the reliability of an underperforming design. The methodology comprises the synthesis of reliability sensitivity measures and construction cost data to obtain practically useful parameter rankings. The results expose the change in the reliability per dollar spent on different parameters. A novel distinction is made between design parameters and tolerance parameters in order to include the possibility of improving the reliability by allocating resources to reduce variability in certain geometry and material parameters. The developments are intended as a tool for the emerging performance-based engineering approach. Sophisticated structural models are employed in conjunction with reliability analysis. In this paper, the costs are estimated based on RSMeans construction cost data, while the probabilistic models for material and geometry parameters are based on the JCSS model code. A detailed example of a six-story, threebay reinforced concrete frame structure is presented to demonstrate the computation and interpretation of the cost-benefit importance vectors.