Victor MatrayLMPS, Faisal AmlaniLMPS, Frédéric FeyelLMPS, David NéronLMPS
{"title":"非参数几何图形的还原阶建模与图神经网络耦合混合数值方法:结构动力学问题的应用","authors":"Victor MatrayLMPS, Faisal AmlaniLMPS, Frédéric FeyelLMPS, David NéronLMPS","doi":"arxiv-2406.02615","DOIUrl":null,"url":null,"abstract":"This work introduces a new approach for accelerating the numerical analysis\nof time-domain partial differential equations (PDEs) governing complex physical\nsystems. The methodology is based on a combination of a classical reduced-order\nmodeling (ROM) framework and recently-introduced Graph Neural Networks (GNNs),\nwhere the latter is trained on highly heterogeneous databases of varying\nnumerical discretization sizes. The proposed techniques are shown to be\nparticularly suitable for non-parametric geometries, ultimately enabling the\ntreatment of a diverse range of geometries and topologies. Performance studies\nare presented in an application context related to the design of aircraft seats\nand their corresponding mechanical responses to shocks, where the main\nmotivation is to reduce the computational burden and enable the rapid design\niteration for such problems that entail non-parametric geometries. The methods\nproposed here are straightforwardly applicable to other scientific or\nengineering problems requiring a large number of finite element-based numerical\nsimulations, with the potential to significantly enhance efficiency while\nmaintaining reasonable accuracy.","PeriodicalId":501482,"journal":{"name":"arXiv - PHYS - Classical Physics","volume":"25 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-06-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A hybrid numerical methodology coupling Reduced Order Modeling and Graph Neural Networks for non-parametric geometries: applications to structural dynamics problems\",\"authors\":\"Victor MatrayLMPS, Faisal AmlaniLMPS, Frédéric FeyelLMPS, David NéronLMPS\",\"doi\":\"arxiv-2406.02615\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This work introduces a new approach for accelerating the numerical analysis\\nof time-domain partial differential equations (PDEs) governing complex physical\\nsystems. The methodology is based on a combination of a classical reduced-order\\nmodeling (ROM) framework and recently-introduced Graph Neural Networks (GNNs),\\nwhere the latter is trained on highly heterogeneous databases of varying\\nnumerical discretization sizes. The proposed techniques are shown to be\\nparticularly suitable for non-parametric geometries, ultimately enabling the\\ntreatment of a diverse range of geometries and topologies. Performance studies\\nare presented in an application context related to the design of aircraft seats\\nand their corresponding mechanical responses to shocks, where the main\\nmotivation is to reduce the computational burden and enable the rapid design\\niteration for such problems that entail non-parametric geometries. The methods\\nproposed here are straightforwardly applicable to other scientific or\\nengineering problems requiring a large number of finite element-based numerical\\nsimulations, with the potential to significantly enhance efficiency while\\nmaintaining reasonable accuracy.\",\"PeriodicalId\":501482,\"journal\":{\"name\":\"arXiv - PHYS - Classical Physics\",\"volume\":\"25 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-06-03\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Classical Physics\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2406.02615\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Classical Physics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2406.02615","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
A hybrid numerical methodology coupling Reduced Order Modeling and Graph Neural Networks for non-parametric geometries: applications to structural dynamics problems
This work introduces a new approach for accelerating the numerical analysis
of time-domain partial differential equations (PDEs) governing complex physical
systems. The methodology is based on a combination of a classical reduced-order
modeling (ROM) framework and recently-introduced Graph Neural Networks (GNNs),
where the latter is trained on highly heterogeneous databases of varying
numerical discretization sizes. The proposed techniques are shown to be
particularly suitable for non-parametric geometries, ultimately enabling the
treatment of a diverse range of geometries and topologies. Performance studies
are presented in an application context related to the design of aircraft seats
and their corresponding mechanical responses to shocks, where the main
motivation is to reduce the computational burden and enable the rapid design
iteration for such problems that entail non-parametric geometries. The methods
proposed here are straightforwardly applicable to other scientific or
engineering problems requiring a large number of finite element-based numerical
simulations, with the potential to significantly enhance efficiency while
maintaining reasonable accuracy.