Predicting the Mechanical Properties of Biopolymer Gels Using Neural Networks Trained on Discrete Fiber Network Data

Abstract

Biopolymer gels, such as those made out of fibrin or collagen, are widely used in tissue engineering applications and biomedical research. Moreover, fibrin naturally assembles into gels in vivo during wound healing and thrombus formation. Macroscale biopolymer gel mechanics are dictated by the microscale fiber network. Hence, accurate description of biopolymer gels can be achieved using representative volume elements (RVE) that explicitly model the discrete fiber networks of the microscale. These RVE models, however, cannot be efficiently used to model the macroscale due to the challenges and computational demands of multiscale coupling. Here, we propose the use of an artificial, fully connected neural network (FCNN) to efficiently capture the behavior of the RVE models. The FCNN was trained on 1100 fiber networks subjected to 121 biaxial deformations. The stress data from the RVE, together with the total energy and the condition of incompressibility of the surrounding matrix, were used to determine the derivatives of an unknown strain energy function with respect to the deformation invariants. During training, the loss function was modified to ensure convexity of the strain energy function and symmetry of its Hessian. A general FCNN model was coded into a user material subroutine (UMAT) in the software Abaqus. In this work, the FCNN trained on the discrete fiber network data was used in finite element simulations of fibrin gels using our UMAT. We anticipate that this work will enable further integration of machine learning tools with computational mechanics. It will also improve computational modeling of biological materials characterized by a multiscale structure.

arXiv comment20 pages, 9 figures, for associated files please see https://bitbucket.org/buganzalab/nn_rve/src/master/

References in corpus (11)

• Adam: A Method for Stochastic Optimization
• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Integrating Machine Learning and Multiscale Modeling: Perspectives, Challenges, and Opportunities in the Biological, Biomedical, and Behavioral Sciences
• Stress controls the mechanics of collagen networks
• Geometric deep learning for computational mechanics Part I: Anisotropic Hyperelasticity
• Learning Constitutive Relations using Symmetric Positive Definite Neural Networks
• A partitioned model order reduction approach to rationalise computational expenses in multiscale fracture mechanics
• Machine learning materials physics: Integrable deep neural networks enable scale bridging by learning free energy functions
• Bayesian Neural Networks
• Backpropagation generalized for output derivatives

Cited by in corpus (7)

• Deep Learning in Deterministic Computational Mechanics
• Data-driven Tissue Mechanics with Polyconvex Neural Ordinary Differential Equations
• Data-driven anisotropic finite viscoelasticity using neural ordinary differential equations
• Automated Data-Driven Discovery of Material Models Based on Symbolic Regression: A Case Study on Human Brain Cortex
• Towards understanding structure-function relationships in random fiber networks
• TopoGEN: topology-driven microstructure generation for in silico modeling of fiber network mechanics
• Learning Mechanically Driven Emergent Behavior with Message Passing Neural Networks