Fitting soil-water characteristic curve by using unimodal and bimodal soil physical properties

Abstract

The soil-water characteristic curve (SWCC) is the relationship between the volumetric water content, ɵ and the matric suction, S (= Ua - UW). The SWCC is requested by many researchers in order to determine the behaviour of soil for further analysis in engineering purposes, for example, shear strength. In the past decade, few fitting methods have been developed to describe the SWCC for a particular soil, for example, Fredlund and Xing (1994). These fitting functions fit the experimental SWCC raw data from laboratory test, to form a curve for further prediction and analyse which costly and time consuming. Thus, Zapata (1994) introduced correlation formulae to D60 from soil physical property and SWCC fitting parameters in order to achieve time and cost saving instead of doing grain size distribution analysis, sieve analysis and hydrometer test. Two set of data were selected, clayed sand and silt. By plotting the grain size distribution (GSD) curves from laboratory and fitted GSD curves were calculated and plotted. From the fitted GSD, found the bmodal fitting equation has better fitting results (R2 = 0.97 (clayed sand) and R2 = 0.98 (silt)) and the soil physical properties, D60, was determined for further analysis in Zapata (1999) equations as fitting parameters (af, nf, and mf) in Fredlund and Xing (1994) model. The results were compared to the Fredlund and Xing (1994) model without employing Zapata (1999) equations and found that the behaviour of SWCC with Zapata (1999) equations can produce similar smooth curves.