Determination of soil permeability coefficient following an updated grading entropy method

Abstract

This paper presents a critical review of the grading entropy approach of permeability-coefficient predictions (k P) for coarse-grained soils. The approach applies the grading entropy theory to particle-size distributions (PSDs), such that the entirety of each gradation curve can be interpreted as a single point on a grading entropy chart, plotting its normalised entropy increment (B) against relative base grading entropy (A) values. Published data sets of measured permeability-coefficient (k M) values for saturated compacted silty sand, sand and gravel materials are examined to understand the dependence of A and B on various gradation parameters and the void ratio (e). In particular, logk M negatively correlates with logB and positively correlates with logA and e (loge). As such, power functions of the form k P = C 1 A C 2 B C 3 e C 4 prove statistically superior, noting that the fitting coefficient C 1 to C 4 values are specific to the PSD range and densification (compaction) states investigated for the permeability tests. Recommendations are given for increasing the predictive power, including separate models for well-graded and poorly graded materials and the addition of a particle shape factor and specific surface parameters in the regression correlation.