Reproduce “Fracture-Bench: A Controlled Benchmark Suite for Machine Learning Surrogates in Phase-Field Fracture Mechanics” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Meenu Krishnan U., Maryam Hakimzadeh, Indrashish Saha, Meer Mehran Rashid, Lori Graham-Brady, Somdatta Goswami
Fracture-Bench compares five surrogate architectures on two-dimensional, quasi-static phase-field fracture datasets, where crack tips 0.02 mm apart can yield different paths. FNO and Latent DeepONet reach test MSEs of 2 × 10⁻⁶ and 7 × 10⁻⁶ on trajectory prediction; cDDPM reaches 0.45 × 10⁻⁴ on final-state prediction, but inference takes 706 seconds per sample. These tests do not establish generalization.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.