LAUSR

Predicting the swelling pressure (Ps) in subgrades constitutes a new approach in pavement structural design. It is shown, however, that current conventional testing programs could lead to erroneous results. Several approaches to estimate Ps have so far been proposed; yet, no universally accepted method has been suggested. In this study, 40 subgrade soil samples taken from a route in Tebessa city (Algeria) were subjected to different laboratory tests. Values of free swelling pressure were measured by using an odometer. A statistical model capable of indirectly estimating Ps is provided. The model is based on a correlation of Ps with dry density ($$\rho _{\mathrm{d}})$$ρd), water content (w) and plastic limit, and has been checked by variance inflation factors method to examine multicollinearity. The proposed equation demonstrates a high correlation coefficient ($$R^{2})$$R2) of 0.887 when compared to the direct laboratory results. Excellent estimation capability was achieved using this method in comparison with previously developed models. The equation, proposed in this paper, provides another evaluation method for the design engineers to estimate swelling pressure in practice.