Comparative evaluation of sigmoidal models for particle size distribution of typical tropical urban soils

Abstract

Particle size distribution (PSD) is a fundamental soil property that strongly influences hydraulic, mechanical, and deformation behavior in geotechnical engineering. Accurate mathematical representation of PSD curves is therefore meaningful for reliable soil characterization and parameter estimation. This study evaluates the fitting performance of three sigmoidal models—the Van Genuchten (1980) model, the Fredlund and Xing (1994) model, and a logistic model—for representing PSD curves of three typical soil types (clay, sandy clay, and clayey sand) in Go Vap Ward, Ho Chi Minh City, Vietnam. Laboratory particle size analyses were conducted using sieve and hydrometer tests in accordance with Vietnamese geotechnical standards. Model performance was assessed using the coefficient of determination (R²), root mean square error (RMSE) and sum of squared errors (SSE). The results show that all three sigmoidal models provide excellent agreement with experimental PSD data, with R² values ranging from 0.9616 to 0.9992. However, model performance varies depending on soil texture classes. The logistic model achieved the best fitting performance for clay soil, whereas the Fredlund and Xing model produced the lowest RMSE and SSE values for sandy clay and clayey sand. These findings highlight the importance of texture-dependent model selection and provide practical guidance for PSD modelling of tropical urban soils in geotechnical engineering applications.