| About this Abstract |
| Meeting |
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Symposium
|
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Presentation Title |
Transfer Learning for Multiscale Analysis: Modeling Delamination of Carbon-Reinforced Composite Material |
| Author(s) |
Zhengtao Yao, Philippe Hawi, Venkat Aitharaju, Jay Mahishi, Roger Ghanem |
| On-Site Speaker (Planned) |
Zhengtao Yao |
| Abstract Scope |
A novel PCE-NN-Transfer-Learning framework was developed to infer physical behavior in the presence of multiscale interactions. We specifically address inferences relevant to delamination and associated constitutive laws, and judiciously introduce cross-scale information for a carbon-reinforced composites material. Representative volume methods (RVE) with periodic boundary conditions and random material and geometric properties augment the micro scale system. Specifically, Beta probability distributions are attached to Young’s modulus, Poisson’s ratio and strength of tows and resins, cohesive strength of the interface, the shape and volume fraction of the ellipse tows. The microscale stress-material-properties-strain dataset is obtained with quadrature failure observed and analyzed. Polynomial chaos expansions (PCE) with principal components analysis for smoothing and dimension reduction are used to obtain the stiffness matrix and strength of the whole RVE trained on this dataset, providing a transition between microscale to mesoscale descriptions. Each mesoscale stiffness and strength realization is then provided as input to a user-defined material subroutine in LS-DYNA to perform bending simulations, thus yielding the macroscale bending force-displacement dataset used for training and validation. First, a neural network is trained on the micro to macro scale dataset: micro-input stress/strain series and macro-output stiffness of the bending force-displacement; it is then retrained on its unfrozen layers on the meso-input, PCE predicted stiffness matrix and strength, to the same macro-output. Finally, a PCE-transfer-learning cross-scale framework for multi-layer-stacked carbon-fibered-reinforced system, from micro(rve) scale to meso(stiffness and strength) to macro(bending stiffness) is constructed to analyze complex composites with high accuracy and low computational efforts. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |