Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage
We propose a computational procedure to derive a data-driven surrogate constitutive model capturing the elastic–plastic response and progressive damage of a heterogeneous solid. This is demonstrated by analysing the response to deformation of a volume element of a non-linear, n-phase random composit...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2024-02-01
|
| Series: | Materials & Design |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S0264127524001102 |
| _version_ | 1797301616606445568 |
|---|---|
| author | Weijian Ge Vito L. Tagarielli |
| author_facet | Weijian Ge Vito L. Tagarielli |
| author_sort | Weijian Ge |
| collection | DOAJ |
| description | We propose a computational procedure to derive a data-driven surrogate constitutive model capturing the elastic–plastic response and progressive damage of a heterogeneous solid. This is demonstrated by analysing the response to deformation of a volume element of a non-linear, n-phase random composite, used in this study as a model material. Finite Element simulations are conducted, imposing pseudo-random, multiaxial, non-proportional histories of macroscopic strain to such volume element. The corresponding predicted histories of macroscopic stresses and other variables are recorded, to form part of a training dataset for the surrogate model. Essential additional training data is obtained by recording the changes in the homogenised stiffness matrix of the volume element during the deformation, by performing a series of linear perturbation analyses. Supervised machine learning is applied to the data, proposing suitable sets of inputs and outputs and implementing a phenomenological constitutive model based on simple neural networks. This results in a data-driven model of high accuracy. |
| first_indexed | 2024-03-07T23:25:08Z |
| format | Article |
| id | doaj.art-c79d13768a95410880e13d10d8890a5e |
| institution | Directory Open Access Journal |
| issn | 0264-1275 |
| language | English |
| last_indexed | 2024-03-07T23:25:08Z |
| publishDate | 2024-02-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Materials & Design |
| spelling | doaj.art-c79d13768a95410880e13d10d8890a5e2024-02-21T05:24:24ZengElsevierMaterials & Design0264-12752024-02-01238112738Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damageWeijian Ge0Vito L. Tagarielli1Department of Aeronautics, Imperial College London, SW72AZ London, UKCorresponding author.; Department of Aeronautics, Imperial College London, SW72AZ London, UKWe propose a computational procedure to derive a data-driven surrogate constitutive model capturing the elastic–plastic response and progressive damage of a heterogeneous solid. This is demonstrated by analysing the response to deformation of a volume element of a non-linear, n-phase random composite, used in this study as a model material. Finite Element simulations are conducted, imposing pseudo-random, multiaxial, non-proportional histories of macroscopic strain to such volume element. The corresponding predicted histories of macroscopic stresses and other variables are recorded, to form part of a training dataset for the surrogate model. Essential additional training data is obtained by recording the changes in the homogenised stiffness matrix of the volume element during the deformation, by performing a series of linear perturbation analyses. Supervised machine learning is applied to the data, proposing suitable sets of inputs and outputs and implementing a phenomenological constitutive model based on simple neural networks. This results in a data-driven model of high accuracy.http://www.sciencedirect.com/science/article/pii/S0264127524001102PlasticityDamageMachine LearningData-drivenConstitutive Model |
| spellingShingle | Weijian Ge Vito L. Tagarielli Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage Materials & Design Plasticity Damage Machine Learning Data-driven Constitutive Model |
| title | Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage |
| title_full | Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage |
| title_fullStr | Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage |
| title_full_unstemmed | Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage |
| title_short | Data-driven homogenisation of the response of heterogeneous ductile solids with isotropic damage |
| title_sort | data driven homogenisation of the response of heterogeneous ductile solids with isotropic damage |
| topic | Plasticity Damage Machine Learning Data-driven Constitutive Model |
| url | http://www.sciencedirect.com/science/article/pii/S0264127524001102 |
| work_keys_str_mv | AT weijiange datadrivenhomogenisationoftheresponseofheterogeneousductilesolidswithisotropicdamage AT vitoltagarielli datadrivenhomogenisationoftheresponseofheterogeneousductilesolidswithisotropicdamage |