Playground
Compare the surrogates with FEM on held-out test cases, or run them live in your browser on a case you set up.
A benchmark for scientific machine learning
Fast neural surrogates for phase-field fracture
Compare operator-learning architectures on two high-fidelity fracture datasets, then see how they predict crack initiation and propagation in a single forward pass.
Fracture-Bench brings neural surrogate models to phase-field fracture simulation. It compares operator-learning and generative architectures on shared data splits and evaluation metrics across two high-fidelity datasets. The playground shows their predictions next to the finite-element results, so the differences between architectures can be explored directly, and it runs the surrogates live in your browser on a case you set up yourself. You can find the project resources below:
Compare the surrogates with FEM on held-out test cases, or run them live in your browser on a case you set up.
Training code for FNO, DeepONet and Latent DeepONet, and the simulation code behind both datasets.
5,000 high-fidelity phase-field simulations in two datasets, openly available in the JHU Data Archive.
Fracture-Bench: a controlled benchmark suite for machine-learning surrogates in phase-field fracture mechanics.
Weights of FNO, DeepONet and Latent DeepONet for both datasets, ready to load in JAX.
The datasets, the surrogates and the main results of the benchmark on one page.
Department of Civil and Systems Engineering and Data Science and AI Institute, Johns Hopkins University, Baltimore, MD 21218, USA