A physics-guided neural network framework for elastic plates: Comparison of governing equations-based and energy-based approaches
Wei Li,Martin Z. Bazant,Juner Zhu
DOI: https://doi.org/10.1016/j.cma.2021.113933
IF: 6.588
2021-09-01
Computer Methods in Applied Mechanics and Engineering
Abstract:<p>One of the obstacles hindering the scaling-up of the initial successes of machine learning in practical engineering applications is the dependence of the accuracy on the size and quality of the database that "drives" the algorithms. Incorporating the already-known physical laws into the training process can significantly reduce the request for data. In this study, we establish a neural network-based computational framework to characterize the finite deformation of elastic plates, which in classic theories is described by the Föppl–von Kármán (FvK) equations with a set of boundary conditions (BCs). A <a href="/topics/computer-science/neural-networks">neural network</a> is constructed by taking the spatial coordinates as the input and the displacement field as the output to approximate the exact solution of the FvK equations. The physical information (PDEs, BCs, and potential energy) is then incorporated into the loss function, and a pseudo dataset is sampled without knowing the exact solution to finally train the neural network. The accuracy of the modeling framework is carefully examined by applying it to four different loading cases: in-plane tension with non-uniformly distributed stretching forces, in-plane central-hole tension, out-of-plane deflection, and buckling under compression. Three ways of formulating the loss function are compared: (1) purely data-driven, (2) PDE-based, and (3) energy-based. Through the comparison with the finite element simulations, it is found that all the three approaches can characterize the elastic deformation of plates with a satisfactory accuracy if trained with a proper strategy in terms of sampling size and resolution. Compared with incorporating the PDEs and BCs in the loss, using the total potential energy shows certain advantage in terms of the simplicity of hyperparameter tuning and the computational efficiency.</p>
mechanics,engineering, multidisciplinary,mathematics, interdisciplinary applications