Pbwd Optimization of Low- to Medium-Rise Steel Buildings Using Pushover Analysis

Abstract

Performance-Based Wind Design (PBWD) is a significant improvement over traditional Load and Resistance Factor Design (LRFD). Yet, integration of life-cycle cost optimization procedures within PBWD, so far, has been limited to high-rise buildings, for which the dynamic response amplification is significant. In this context, this paper introduces a novel PBWD approach to minimize expected life-cycle costs of medium-rise steel buildings. The proposed methodology combines wind vulnerability analysis with static-equivalent non-linear pushover, as low and mediumrise buildings do not experience dynamic amplifications. The inelastic analysis allows for determining structural and non-structural limit state probabilities corresponding to two damage levels, which are converted to expected economic losses using inventory information. Moreover, an active learning Kriging metamodel is employed to reduce the massive computational burden of PBWD optimization, which is tailored for discrete design variables (commercial member sections). Application case studies involve optimization of moment-resisting frames of two distinct steel buildings under different wind hazards. The results demonstrate effectiveness of the proposed approach, which can efficiently find optimal solutions with significantly smaller frame construction costs and reduced total expected costs, compared to traditional LRFD solutions.

Keywords: performance-based design, life-cycle costs, probabilistic assessment, non-linear pushover, adaptive kriging, steel buildings