Does liquefaction protect overlying structures from ground shaking?

Abstract

When recordings have been obtained at sites that have been liquefied, it is clear that high-frequency components of motion are absorbed as the soil softens after the triggering of pore-pressure increase. However, it is also clear that significant acceleration peaks can be transmitted before the triggering of liquefaction effects. Also, as is now well established, liquefied soil retains some shear strength, and long-period motions continue to propagate through the liquefied layer. In this research, instrumented sites where liquefaction has been observed are examined, and comparisons between the response that would have been obtained in the absence of liquefaction and the response from actual recorded motions are carried out.

Actual records obtained from down-hole arrays at the Wildlife Site, California, in the 1987-Superstition Hills earthquake, Port Island, Kobe, in the 1995-Hyogoken-Nambu earthquake and Treasure Island-Yerba Buena Island, in the 1987-Loma Prieta earthquake, allow the effect of liquefaction of intermediate soil layers on the ground surface response to be observed. The recorded surface motions in these earthquakes are compared with simulated motion obtained from a one-dimensional equivalent linear analysis, carried out with the program ProShake version 1.1, assuming no softening of the soil occurs.

The responses obtained for those three particular cases, in terms of absolute acceleration and relative displacement spectra, are not homogeneous, and indicate that the base isolation effect in overlying structures depends mainly on factors such as their fundamental period and site soil conditions.