Predicting Peak Resistance of Spudcan Penetrating Sand Overlying Clay

Accurately predicting peak penetration resistance $q_{\text{peak}}$ during spudcan installation into sand overlying clay is crucial to an offshore mobile jack-up industry still suffering regular punch-through failures. This paper describes a series of spudcan penetration tests performed on medium-loose sand overlying clay and compares the response to existing centrifuge data from tests performed on dense sand overlying clay. Together these data demonstrate that punch-through is a potential problem for both dense and loose sand overlying clay soil stratigraphies. Using this experimental database, a failure-stress-dependent model has been modified to account for the embedment depth, and the depth of occurrence of $q_{\text{peak}}$ is shown to be a function of the sand thickness $H_s$. The model then was recalibrated, taking these findings into account, for a larger range of material properties and ratios of sand thickness to spudcan diameter ($H_s/D$). Finally, the performance of the modified and recalibrated model is verified by comparing its predictions with those calculated using current guidelines. The comparisons show that the modified model yields more accurate predictions of $q_{\text{peak}}$ over the range of $H_s/D$ ratios of practical interest, which when used in practice will potentially mitigate the risk of unexpected punch-through on sand overlying clay stratigraphies.