Efficient intensity measures and classification algorithms for collapse prediction of tall buildings

Nenad Bijelic, Ting Lin, & Gregory Deierlein

In Preparation 2018, SCEC Contribution #8040

The emergence and maturity of physics-based simulations, such as the ones from Southern California Earthquake Center CyberShake project, offer a unique opportunity to re-examine the relationships between engineering demand parameters and input ground motions. In contrast to approaches based on recorded seismograms, using extensive sets of site-specific unscaled CyberShake motions requires no assumptions about how properties of motions change with scaling (which in turn implicitly requires assumptions on intensity measures (IMs) that control the response). In this study, we focus on collapse response with the following objectives: (1) examine seismogram features that drive the collapse response of tall buildings; (2) contrast utility of different IMs for collapse prediction; and (3) develop efficient and reliable collapse classification algorithms. To that end, we apply machine learning techniques to results of nonlinear time history analyses of an archetype 20-story tall building performed using ~2,000,000 ground motions generated for CyberShake sites in the Los Angeles basin. Performed feature selection (based on regularized logistic regression) generally confirms existing understanding of collapse predictors as gained from scaled recorded motions but also reveals the benefit of some novel IMs, in particular the RSx spectral features. In addition, the statistical interrogations of the large collection of hazard-consistent simulations also demonstrate the utility of different IMs for collapse predictions in a way that is not possible with recorded motions. Finally, a small subset of robust IMs is identified and used in development of an efficient collapse classification algorithm which is tested on benchmark results throughout the Los Angeles basin.

Bijelic, N., Lin, T., & Deierlein, G. (2018). Efficient intensity measures and classification algorithms for collapse prediction of tall buildings. Earthquake Engineering and Structural Dynamics, (in preparation).

Related Projects & Working Groups
Ground Motion Simulation Validation (GMSV) Technical Activity Group (TAG), Earthquake Engineering Implementation Interface (EEII)