Phase-Field Fracture in Multiphase Materials: An IFENN Approach
Dulal, Prakash
Dulal, Prakash
Citations
Altmetric:
License
License
Abstract
The finite element method (FEM) has been widely used in engineering for decades, but simulating fracture remains costly due to the need for extremely fine meshes near cracks. This cost increases further in multiphase materials where varying properties influence crack behavior. In this poster presentation, we present the Integrated Finite Element Neural Network (IFENN), a hybrid method combining FEM with physics-informed convolutional neural networks (PI-CNNs) for efficient crack prediction in materials with inclusions. FEM handles the mechanical problem, while the PI-CNN solves the phase-field fracture problem in a staggered manner. The network is trained using a physics-based loss function that accounts for spatially varying material toughness, requiring only minimal data from a simple single-inclusion problem. Once trained, the model generalizes to problems with multiple inclusions under different loads, boundary conditions, and shapes, delivering accurate crack paths and forces at significantly reduced computational cost.
Description
Date
1/1/2026
Student Status
First Year (Graduating in 2029)
Journal Title
Journal ISSN
Volume Title
Type of presentation
Poster
Collections
Research Projects
Organizational Units
Journal Issue
Citation
DOI
Advisor(s)
Department
Program/Major
Civil and Enviromental Engineering
College/School
College of Engineering and Mathematical Sciences
Organization
Research Category
Engineering