Researchers have developed a novel mesh-free method for phase-field modeling of brittle fracture, utilizing a single neural network to represent displacement and phase fields. This approach trains the network by directly minimizing the incremental energy, incorporating multiresolution feature encoding via C1 quadratic B-spline grids. The method demonstrates critical pairings that allow for accurate crack propagation prediction and classification of crack states, outperforming existing deep Ritz baselines on benchmark datasets. AI
IMPACT Introduces a novel neural network approach for simulating complex physical phenomena like brittle fracture.
RANK_REASON Research paper published on arXiv detailing a new computational method. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →