Experimental Study of Shaft Resistance of Model Piles in Fluidized and Nonfluidized Fine Sand

Abstract

Torpedo piles installed by dynamic penetration have been used as anchors in the Brazilian offshore oil production infrastructure practice for two decades. Dynamic penetration aided by fluidization of the soil during pile penetration is now being contemplated as a method of installation that would allow deeper penetration. The two key design questions in connection with torpedo piles are how far they penetrate and what their pullout capacity is. In a companion paper, the authors addressed the first question, whereas in the present one the second question is attended through laboratory tests using model piles, essentially pipes simulating torpedo piles without wings. The model piles were installed in two different ways: by fluidization, which enabled the piles to sink by their own weight, and by monotonic jacking. Pullout tests were then performed on the model piles in both fluidized and nonfluidized sandy soils prepared at two initial relative densities. Results from the laboratory tests indicate that shaft uplift capacity of fluidized piles is essentially independent of the sand initial relative density. The measured values of the coefficient of lateral earth pressure (Ks) derived from the fluidized model tests are lower than those reported for other methods of pile installation, in some cases being lower than K0. Finally, the shaft resistance of fluidized piles increases after installation as the soil reconsolidates and particles rearrange.