Efficient global sensitivity analysis for flow-induced vibration of a nuclear reactor assembly using Kriging surrogates

Highlights: • Surrogate modeling is used to support a global sensitivity analysis (GSA) for a nuclear reactor assembly to demonstrate both the pertinence of such methods to this application as well as the significant physical insights provided by GSA. • Insight gained from the accuracy of the GSA results may be used to compare with goodness-of-fit metrics which are traditionally used to support the verification of the surrogate model. • The coupled use of surrogate modeling and GSA reduces the number of full-order simulations required, substantially reducing total computational cost. • Numerical experiments based upon an upper internals assembly of a pressurized water reactor subjected to multiple types of loadings are used to evaluate the relationship between sensitivities computed from a full-order model versus those computed from a surrogate model. • For large sample sizes, negligible variation in the resultant sensitivities is shown with respect to the particular method by which a computational design of experiment is constructed to train the Kriging surrogates which lends credence to the stability of the results. - Abstract: In this work, surrogate modeling is used to support a global sensitivity analysis (GSA) for a nuclear reactor assembly as a proof-of-concept to demonstrate both the pertinence of such methods to this application as well as the significant physical insights provided by GSA. In addition to the knowledge gained relating to the system sensitivity, insight gained from the accuracy of the GSA results may be used to compare with goodness-of-fit metrics which are traditionally used to support the verification of the surrogate model. The coupled use of surrogate modeling and GSA reduces the number of full-order (i.e., standard computationally expensive finite element analysis) simulations required, substantially reducing total computational cost. This work focuses on the use of Kriging surrogates in particular, and examines the robustness of these techniques to evaluate sensitivity by considering a variety of design of experiment strategies used to create the surrogate models. Numerical experiments based upon an inverted top-hat upper internals assembly of a pressurized water reactor subjected to base motion and fluctuating lift and drag cross-flow loadings are used to evaluate the relationship between sensitivities computed from a full-order model versus those computed from a surrogate model, highlighting the effectiveness of utilizing GSA and surrogate modeling. For large sample sizes, negligible variation in the resultant sensitivities is shown with respect to the particular method by which a computational design of experiment is constructed to train the Kriging surrogates which lends credence to the stability and veracity of the results. Additionally, for the example presented herein the historical significance of the downcomer forcing function characterization is substantiated in the sense that loads from the downcomer which act indirectly on the upper internals are shown to dominate the response relative to direct-applied cross-flow loads.