Peak resistance prediction of a jack-up platform spudcan foundation in sand overlying clay based on Bayesian algorithm
In the process of positioning a jack-up platform in sand overlying clay,the spudcan foundation is susceptible to punch-through the underlying soft soil layer,resulting in puncture damage.Therefore,accurate predictions of peak resistance as well as corresponding depth of spudcan foundation in sand overlying clay are essential for operation safety of the platform.Based on the existing data of centrifuge tests,the prior probability and likelihood probability models have been constructed for both the potential peak resistance and corresponding depth of a jack-up platform spudcan foundation in sand overlying clay by introducing uncertainty factors which conforms to the Beta-PERT distribution.Besides,the Bayesian algorithm was used to calculate and update the posterior probability by using the monitoring data during the penetrating process of spudcan,so as to effectively improve the prediction accuracy of the peak resistance of spudcan foundation in sand overlying clay.On this basis,the Bayesian probabilistic prediction method presented in this paper for predicting the peak resistance of spudcan foundation in sand overlying clay was validated by recalculating 25 groups of centrifuge tests.The results show that,compared with the load-spread method and punching shear method as recommended by the current guidelines of ISO and SNAME,as well as three newly developed methods,the present Bayesian probabilistic approach could effectively improve prediction accuracy of the peak resistance of spudcan foundation in sand overlying clay.The research findings provide methodological support for the safety assessment of positioning a jack-up platform spudcan foundation in sand overlying clay.
万方数据
维普
