%0 Journal Article %1 fe7e5a97e46c47e29cb9fea6f05df623 %A Gerritzen, Johannes %A Hornig, Andreas %A Winkler, Peter %A Gude, Maik %D 2024 %I Elsevier Science B.V. %J Computational Materials Science %K area_architectures topic_engineering Constitutive FIS_scads Fiber Machine Neural Parameter identification learning, modeling, networks, plastics, reinforced %R 10.1016/j.commatsci.2024.113274 %T A methodology for direct parameter identification for experimental results using machine learning — Real world application to the highly non-linear deformation behavior of FRP %V 244 (2024) %X In this work, we demonstrate how Machine Learning (ML) techniques can be employed to externalize the knowledge and time intensive process of material parameter identification. This is done on the example of a recent data driven material model for the non-linear shear behavior of glass fiber reinforced polypropylene (GF/PP) (Gerritzen, 2022). A convolutional neural network (CNN) based model architecture is trained to predict material modeling parameters based on the input of stress–strain-curves. The optimal model architecture and training setup are determined by hyperparameter optimization (HPO). Solely virtual data, generated using the target material model, is used throughout the training and HPO. The final CNN is capable of calculating model parameter combinations from experimental stress–strain-curves which lead to excellent agreement between experimental and associated model curve.

@article{fe7e5a97e46c47e29cb9fea6f05df623, abstract = {In this work, we demonstrate how Machine Learning (ML) techniques can be employed to externalize the knowledge and time intensive process of material parameter identification. This is done on the example of a recent data driven material model for the non-linear shear behavior of glass fiber reinforced polypropylene (GF/PP) (Gerritzen, 2022). A convolutional neural network (CNN) based model architecture is trained to predict material modeling parameters based on the input of stress–strain-curves. The optimal model architecture and training setup are determined by hyperparameter optimization (HPO). Solely virtual data, generated using the target material model, is used throughout the training and HPO. The final CNN is capable of calculating model parameter combinations from experimental stress–strain-curves which lead to excellent agreement between experimental and associated model curve.}, added-at = {2024-11-28T16:27:18.000+0100}, author = {Gerritzen, Johannes and Hornig, Andreas and Winkler, Peter and Gude, Maik}, biburl = {https://puma.scadsai.uni-leipzig.de/bibtex/2add81a52dde9fdbe201a60f11f51da1c/scadsfct}, doi = {10.1016/j.commatsci.2024.113274}, interhash = {e45da739ed33fc18e4c9e48dd9c9e9bc}, intrahash = {add81a52dde9fdbe201a60f11f51da1c}, issn = {0927-0256}, journal = {Computational Materials Science}, keywords = {area_architectures topic_engineering Constitutive FIS_scads Fiber Machine Neural Parameter identification learning, modeling, networks, plastics, reinforced}, language = {English}, month = sep, note = {Publisher Copyright: {\textcopyright} 2024 The Author(s)}, publisher = {Elsevier Science B.V.}, timestamp = {2024-12-06T14:36:51.000+0100}, title = {A methodology for direct parameter identification for experimental results using machine learning — Real world application to the highly non-linear deformation behavior of FRP}, volume = {244 (2024)}, year = 2024 }