Label-Free Physics-Informed Strength Reduction and a Neural Operator for the Reliability of Spatially Variable Slopes

License: See LICENSE file · Updated 2026-10-06
My rating

This research develops a label-free physics-informed neural network for non-associated Mohr-Coulomb plasticity and the reliability of spatially variable c-phi slopes. Training uses a virtual-work loss defined on a finite-element mesh, while a GPU solves many random fields simultaneously. With the solver budget fixed using homogeneous slopes, the method is evaluated against the Abaqus factor of safety.

Tags: