Sample disturbance of cemented collapsible soils

A comprehensive study of sample disturbance effects on the response of collapsible soils was performed. Various factors potentially affecting degree of disturbance were studied. It was found that the effect of tube length was statistically insignificant for lengths up to 305 mm for 71 mm diameter specimens, and likewise for sampler tube area ratio up to 56%. No significant difference was found between driven and pushed sampler tubes, or between specimens trimmed into the test ring compared to those extruded directly into the ring. The response-to-wetting parameters for block specimens were not significantly different from those of hammered tubes for well-cemented soils with low gravel content. However, for very lightly cemented soils and high-gravel-content soils a moderately small, but statistically significant, difference in response was observed. The block samples exhibited lower initial dry unit weights and more collapse strain upon wetting. The difference was ascribed to more densification occurring during tube sampling combined with selective sampling, which unavoidably occurs when specimens are trimmed out of very lightly cemented or very gravelly block samples.