A comprehensive study on soil parameters influencing collapsibility of loess through data analysis

Abstract

Loessial soils may settle and collapse under significant loading or water inundation. Therefore, their mechanical behavior and collapse potential must be assessed before undertaking any construction work. However, conducting a thorough empirical assessment of soil collapsibility would be both costly and time-consuming. Therefore, in this study, a dataset, including the results of 756 oedometer test records on loessial soils, as well as the physical properties of the samples, has been gathered from 27 published papers. Subsequently, a grey relational analysis (GRA) was used to identify the most influential factors among the thirteen most significant factors. The results showed that the most influential factors on the collapsibility of loess are the plasticity index, dry unit weight, initial void ratio, and degree of saturation. Finally, using the multivariate non-linear regression (MNLR) technique, a prediction equation has been derived to predict the collapse potential of loess with an acceptable level of accuracy.