Estimating relative density of sand with cone penetration test

Abstract

As a result of difficulties in obtaining undisturbed samples in cohesionless soils, empirical correlations based on cone penetration test (CPT) results, often developed from calibration chamber experiments, are widely used for determining many soil parameters for geotechnical investigation. This paper describes the application of 19 reduced-scale calibration chamber CPTs for examining the empirical correlations used for predicting the relative density of loose to medium-dense sands. A reduced-scale subtraction cone with an apex angle of 60° and a net area ratio of 0·75 was used in the laboratory tests. Current methods for estimating relative density present very large differences among themselves as well as with the experiments of this study, rendering them meaningless for general application in all sands. The state parameter is suggested as a more reliable and universal alternative to relative density for estimating the degree of compaction of any sand deposit from CPT data. It is demonstrated that an empirical method provides reasonable estimates of the state parameter for the CPT experiments of this study. Compared with relative density, the state parameter method can be particularly useful for examining the liquefaction susceptibility of sands.