Comparison of risk-based robustness indices in progressive collapse analysis of building structures

A current challenge in Structural Engineering is the identification of a robustness index for the design of structures subject to low-probability high-consequence load events. Robustness indices proposed in the literature have not been subject to comprehensive scrutiny. Few comparisons presented so far have been limited to simple academic examples involving idealized systems, which cannot suggest robustness values for practical design. In this paper a comprehensive evaluation of two risk-based robustness indices is presented, addressing progressive collapse of regular frame structures subject to damage by abnormal loads. Herein frames of different aspect ratios are considered, subject to local damage of different size, in hazard-independent and hazard-dependent (blast loading) scenarios. System damage states include plastic bending failure of beams, as well as local and global column crushing. Moreover, robustness indices are compared for structures conventionally designed, for structures strengthened following the codified Alternate Path Method (APM), and for structures resulting from riskbased optimization, for different hazard probability levels. Results suggest that the robustness index of Baker, Schubert and Faber performs better when comparing conventional to APM-strengthened codified designs. The Praxedes, Yuan and He robustness index is less sensitive but appears to be more appropriate to describe structures resulting from risk-based optimization.