The liquefaction resistance of partially saturated soil was experimentally investigated for one clean sand and one silty sand collected from a site in Christchurch, in an area severely affected by liquefaction in the 2010–2011 Canterbury earthquakes. A series of cyclic undrained tests were performed on fully and partially saturated sand and silty sand specimens, in conjunction with evaluation of saturation conditions in situ based on comprehensive field measurements of P-wave velocity (Vp) in Christchurch deposits. The Skempton’s B-value and P-wave velocity were comparatively used as measures for partial saturation in the laboratory. B-value - Vp relationships from the test results indicate that Vp steadily increases with the B-value until a threshold B-value is reached beyond which Vp remains unchanged at values indicating full saturation, i.e. Vp ;≥ ;1600 ;m/s. In general, the liquefaction resistance of tested sand and silty sand increases with a decrease in the B-value or Vp, i.e. with a reduction in the degree of saturation. Furthermore, test results suggest existence of threshold B-values and Vp for tested soils beyond which no significant increase in the liquefaction resistance was observed. This threshold B-values and Vp were found to be dependent on soil type and applied confining stress. The effects of partial saturation on liquefaction strength are different for the sand and silty sand when using Vp as a measure for the degree of saturation. While a gradual rate of increase in liquefaction strength with decreasing Vp is observed for the tested sand, the liquefaction strength of silty sand shows similar gradual increase with a decrease in Vp up to about 800 ;m/s, which is then followed by an abrupt increase in the liquefaction strength for Vp ;< ;800 ;m/s. Generally good agreement between liquefaction strength of tested soils and published data was observed, with a clear distinctive feature in the behaviour of the silty sand as compared to clean sands.
Read full abstract