Reduced order model based metamodeling for multiparametric local heat treatment simulations
Abstract
The objective of the present work is to combine reduced order model (ROM) methods with parameter-
to-solution metamodeling techniques to predict instantly the mechanical state at the end of Post Welding
Heat Treatments on industrial steel pipelines. The ROM method used is based on a proper orthogonal
decomposition (POD) and Grassmann interpolation strategy with parametric finite elements simulations.
Additionally to the accuracy estimation of the resulting surrogate model, global variance and sensitivity
analysis are carried out to determine operation parameters influence qualitative and quantitative.
Fichier principal
csma_Reduced_order_model_based_metamodeling_for_multiparametric_local_heat_treatment_simulations.pdf (685.19 Ko)
Télécharger le fichier
Origin | Files produced by the author(s) |
---|