This paper describes a metamodelling approach to investigate the behaviour of a plate anchor in spatially variable soil. The approach explores the effect of variability in undrained shear strength on the monotonic holding capacity. The problem is firstly analysed through a selected number of two-dimensional finite element (FE) analyses. As this is computationally expensive, relatively few analyses are performed, with the metamodel developed to map the response over a wider range of parameters. This uses mathematical operators calibrated to emulate the FE models, which can be built on a relatively small data set, with the overall objective to retain the accuracy of the original FE model at negligible computational cost. In future, this approach may support design by considering uncertainties that are common for any offshore foundation problem, providing a tool that can be easily coupled with traditional probabilistic design approaches.
Mentani Alessio, Govoni Laura, Gaudin Christophe, Watson Phil, Tian Yinghui (2023). Metamodeling to emulate plate anchor response in spatially variable soil. London : Society for Underwater Technology.
Metamodeling to emulate plate anchor response in spatially variable soil
Mentani Alessio
Primo
Writing – Review & Editing
;Govoni LauraSecondo
Funding Acquisition
;
2023
Abstract
This paper describes a metamodelling approach to investigate the behaviour of a plate anchor in spatially variable soil. The approach explores the effect of variability in undrained shear strength on the monotonic holding capacity. The problem is firstly analysed through a selected number of two-dimensional finite element (FE) analyses. As this is computationally expensive, relatively few analyses are performed, with the metamodel developed to map the response over a wider range of parameters. This uses mathematical operators calibrated to emulate the FE models, which can be built on a relatively small data set, with the overall objective to retain the accuracy of the original FE model at negligible computational cost. In future, this approach may support design by considering uncertainties that are common for any offshore foundation problem, providing a tool that can be easily coupled with traditional probabilistic design approaches.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.