Sample Preparation of Silts for Liquefaction Testing

Abstract One of the most important aspects of cyclic testing in the laboratory is using samples that are representative of their in-situ conditions. Since undisturbed samples of cohesionless soils are typically too difficult or costly to obtain, reconstituted samples need to be prepared using a method that most closely replicates the in-situ stress, density, and fabric. Research has clearly shown the effect of sample preparation methods on the liquefaction resistance of sands, and it is believed that wet pluviation methods most closely approximate the in-situ fabric of fluvial soils. For pure silts, however, these methods are limited because only very loose samples can be made. This paper introduces a new modified moist tamping method that can be used to reconstitute denser specimens of silt. It was found that samples tamped at an initial saturation level of about 55 % gave comparable cyclic strengths to samples prepared from a slurry and to specimens trimmed from an in-situ block sample. The method can be considered a cost-effective alternative for the liquefaction testing of silts.