descriptionPublicationkeyboard_double_arrow_right Article , Preprint , Other literature type 13 Jan 2020Embargo end date: 01 Jan 2021 English Publisher:Springer Science and Business Media LLCJournal:Archives of Computational Methods in Engineering, volume 27, pages 1,337-1,362 (issn: 1134-3060, eissn: 1886-1784,

Authors: Pedro Díez; Marc Rocas; Marc Rocas; Sergio Zlotnik; Alberto García-González; Xabier Larráyoz;

doi: 10.1007/s11831-019-09397-x , 10.1016/j.finel.2021.103556 , 10.48550/arxiv.2102.07673

arXiv: http://arxiv.org/abs/2102.07673

handle: 2117/340337 , 2117/356919

Nonintrusive Stochastic Finite Elements for Crashworthiness with VPS/Pamcrash

- Summary
- Subjects
- Metrics

Abstract

Uncertainty Quantification (UQ) is a key discipline for computational modeling of complex systems, enhancing reliability of engineering simulations. In crashworthiness, having an accurate assessment of the behavior of the model uncertainty allows reducing the number of prototypes and associated costs. Carrying out UQ in this framework is especially challenging because it requires highly expensive simulations. In this context, surrogate models (metamodels) allow drastically reducing the computational cost of Monte Carlo process. Different techniques to describe the metamodel are considered, Ordinary Kriging, Polynomial Response Surfaces and a novel strategy (based on Proper Generalized Decomposition) denoted by Separated Response Surface (SRS). A large number of uncertain input parameters may jeopardize the efficiency of the metamodels. Thus, previous to define a metamodel, kernel Principal Component Analysis (kPCA) is found to be effective to simplify the model outcome description. A benchmark crash test is used to show the efficiency of combining metamodels with kPCA.

Related Organizations

Universitat Politècnica de Catalunya
Spain
Institució dels Centres de Recerca de Catalunya
Spain
Universitat Polite`cnica de Catalunya
Spain
International Center for Numerical Methods in Engineering
Spain
Seat (Spain)
Spain

View all View all

Keywords

FOS: Computer and information sciences, Computer Science - Machine Learning, Surrogate modeling, :45 Integral equations::45H05 Miscellaneous special kernels [Classificació AMS], :65 Numerical analysis::65H Nonlinear algebraic or transcendental equations [Classificació AMS], Classificació AMS::45 Integral equations::45H05 Miscellaneous special kernels, Crashworthiness, Kernel principal component analysis (kPCA), Numerical analysis--Simulation methods, Classificació AMS::65 Numerical analysis::65K Mathematical programming, 510, :Matemàtiques i estadística::Anàlisi numèrica [Àrees temàtiques de la UPC], Machine Learning (cs.LG), Methodology (stat.ME), FOS: Mathematics, Àrees temàtiques de la UPC::Matemàtiques i estadística::Anàlisi matemàtica::Equacions funcionals, Mathematics - Numerical Analysis, :65 Numerical analysis::65K Mathematical programming, optimization and variational techniques [Classificació AMS], Integral equations, Àrees temàtiques de la UPC::Matemàtiques i estadística::Anàlisi numèrica::Mètodes numèrics, Uncertainty quantification, Statistics - Methodology, Separated response surface (SRS), Anàlisi numèrica, :Matemàtiques i estadística::Anàlisi numèrica::Mètodes numèrics [Àrees temàtiques de la UPC], Àrees temàtiques de la UPC::Matemàtiques i estadística::Anàlisi numèrica, Classificació AMS::65 Numerical analysis::65K Mathematical programming, optimization and variational techniques, Numerical Analysis (math.NA), Equacions integrals, 620, Kriging, Classificació AMS::65 Numerical analysis::65H Nonlinear algebraic or transcendental equations, :Matemàtiques i estadística::Anàlisi matemàtica::Equacions funcionals [Àrees temàtiques de la UPC], optimization and variational techniques, Numerical analysis

Impact byBIP!

	selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).	15
	popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.	Top 10%
	influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).	Average
	impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.	Top 10%