A Model to Predict Strength of Cement-Stabilized Clay with High Initial Water Contents

Estimating the strength of cement-stabilized clay (CSC) is crucial for efficient and reliable design of ground improvement. However, current strength models of CSC are inadequate for soft clay soils at high water contents. This paper proposes a new strength model of CSC based on a set of unconfined compressive strength (UCS) data. The experimental data indicate that the water–cement ratio alone cannot account for the strength variation effectively, and the initial water content of CSC should also be given more consideration. A power function of liquidity index was incorporated into Abrams’ equation to consider both cementation strength and clay fabric. The proposed model was shown to outperform existing models. For other clays in the literature, the power exponents of the proposed model vary over a narrow range, but the parameter q1(28d) varies significantly from 8,054 to 25,194 kPa. The values of q1(28d) were found to correlate well with clay activity and clay content. Hence, a multiple regression equation was established for q1(28d) using clay activity and clay content. Using the multiregression equation for q1(28d) and the mean values of power exponents, the proposed model can provide mean UCS estimates without the need for calibration tests. In addition, the mean UCS can be used to determine the characteristic UCS for preliminary design purposes, as specified in Eurocodes.