Freestanding block overturning fragilities: Numerical simulation and experimental validation

The overturning fragilities of symmetric and asymmetric freestanding blocks, ranging in height from 0.54 to 3.6m and with height-to-width ratios ranging from 2.1 to 6.6, are determined numerically. A probabilistic formulation regularizes the overturning responses when exposed to earthquake-like random-vibration waveforms. The peak amplitude of the forcing excitation (peak ground acceleration or PGA) is parameterized as a function of the block size, block shape, overturning probability, and either the PGA normalized peak ground velocity (PGV/PGA), spectral acceleration at 1 s (S-a(1)/PGA), or spectral acceleration at 2s (S-a(2)/PGA). These later intensity measures are correlated with the duration of the predominant acceleration pulse. The overturning fragilities are compared with shake table experiments using blocks ranging in height from similar to 0.2 to 1.2 m and with height-to-width ratios ranging from similar to 2 to 10. Excitations utilized in the shake table experiments include recordings of the 1979 Imperial Valley, 1985 Michoacan, 1999 Duzce, 1999 Chi-Chi, and 2002 Denali Earthquakes along with synthetic waveforms. The overturning fragilities accurately represent the overturning responses of blocks with simple basal contact conditions. Objects with multiple rocking points, such as precariously balanced rocks, are more fragile than predicted. Nondestructive tilting tests are used to account for blocks with complex basal contact conditions, demonstrating that these blocks overturn similarly to more slender blocks with simple contact conditions. Copyright (C) 2008 John Wiley & Sons, Ltd.