Centrifuge Modeling of the Effect of Preshaking on the Liquefaction Resistance of Silty Sand Deposits

Field observations suggest that preshaken natural sands in some seismic regions have high liquefaction resistance as a result of geologic aging and/or preshaking. This paper focuses on the young silty sand deposits located in the Imperial Valley of California. Recent deposition and intense seismic activity in the Valley suggest that preshaking is the main cause of their increased liquefaction resistance. The first part of the paper examines the liquefiable layer at the Wildlife site, which may have been deposited by flooding approximately between 1905-1907. The site was instrumented with accelerometers and piezometers in 2005, providing data over the last 10 years. The following conclusions are reached from this and from the catalog information on earthquakes before 2005: (1) Since 1907, the Wildlife layer has been subjected to approximately 60-70 earthquakes having a(max) >= 0.1 g at the site, which caused pore pressure buildup in the layer; (2) most of these earthquakes generated excess pore pressures but generally did not liquefy the soil (Events A); and (3) approximately 10 or 20% of all earthquakes were capable of liquefying the layer immediately after deposition (Events B). This information was used to plan a centrifuge experiment that crudely simulated the history of the Wildlife site. In this test, 66 base shakings were applied to the base of a 6-m prototype homogeneous deposit of loose saturated silty sand, with a ratio of one Event B for every 10 Events A. Events B liquefied the deposit at the beginning but not at the end of the experiment. Events A liquefied the deposit at very shallow depths at the beginning but stopped liquefying it very soon into the experiment. Finally, an Event B caused the next Event A to generate more excess pore pressures, with this effect being canceled rapidly by a couple of subsequent Events A. The lack of liquefaction by Events B after heavy preshaking in the experiment is consistent with theWildlife layer response to the 2010, M-w = 7.2, El Mayor-Cucupah earthquake, an Event B that generated only a 19% pore pressure ratio at the site. (C) 2016 American Society of Civil Engineers.