Centrifuge Testing of Foundation Behavior Using Full Jackup Rig Models

ABSTRACT The response of independent leg jack-up rigs to environmental loading is highly sensitive to foundation behavior. Because of this sensitivity and the many uncertainties involved there is an urgent need for experimental investigation of jack-up foundation behavior. The Drum Centrifuge offers an economical means for obtaining such data through properly scaled model results. An experimental program is described in which tests were conducted on both a single spud model and a full 3-leg model on sand. A description of this program and some preliminary results of these studies are presented herein. The results are examined in the context of work hardening plasticity theory and are shown to be consistent with this idealization. The theoretical formulation herein incorporates well known aspects of foundation behavior, specifically classical bearing capacity theory. Although results examined to date are quite limited, they indicate that considerable soil fixity can exist under working loads even on loose sands. Considerable work remains to validate and generalize these results. For example, a broader range of loading and soil characteristics needs to be considered. The effects of cyclic loading for undrained or partially drained conditions and the effects of sand density variations also need to be explored. INTRODUCTION The water depth limits of independent leg jack-up rigs are profoundly affected by their foundation behavior, primarily the rotational restraint provided by the soil on the jack-up' s foundation elements (spud cans). For economic reasons operators typically desire to employ jack-up rigs at their maximum safe water depth limits. That is, they desire to use the least expensive (usually the smallest) rig that will safely do the job. There is therefore a keen interest in developing reliable models of spUd can-soil interaction behavior. Unfortunately, developing the requisite accuracy with the desired confidence level has proved rather difficult. A few field measurement programs (2, 3, 5) have been carried out in which the foundation rotational stiffness has been inferred from structural response. These programs have been invaluable in establishing a baseline for model calibration and additional measurements are clearly needed. On the other hand, such measurements are limited as to the range of loads, water depths and foundation conditions that can be observed. For example, it is unlikely that the design wave load would occur during a measurement program. This is important because the foundation rotational restraint is greater under the lighter loads, and decreases under greater loads thus exhibiting nonlinear behavior. Hence, extension (and generalization) of field measurements to include a wider range of loads, water depths and foundation conditions will require additional analytical and experimental studies. The program described herein falls in the latter category. It involves the use of a special geotechnical centrifuge to carry out properly scaled model tests of both individual spud cans under combined vertical, horizontal and moment loading and three leg jack-up models subjected to lateral loading. Prior to describing the centrifuge tests and results, a description of the structural/ foundation behavior of jack-ups is presented below followed by a brief discussion of the principles of centrifuge modeling.