Numerical methods in engineering

Predicting how cracks form, grow, and ultimately cause structures to fail is one of the central problems in computational mechanics, with direct consequences for the safety of bridges, aircraft, pressure vessels, and any engineered component subject to stress. Classical finite element methods struggle with crack propagation because a moving crack discontinuity constantly violates the assumptions baked into a fixed mesh, which has driven the development of alternatives such as the Extended Finite Element Method, meshless and radial-basis-function approaches, peridynamics, and phase-field models — each representing a different mathematical strategy for handling sharp or diffuse discontinuities without remeshing. A persistent open question is how to make these methods computationally tractable at engineering scales while remaining faithful to the physical fracture processes occurring at much finer length scales, particularly in brittle materials where crack paths are highly sensitive to microstructure and loading history. Current research is pushing toward robust coupling of these frameworks, better handling of three-dimensional crack branching and coalescence, and validation against controlled experiments to establish which modeling choices actually matter in practice.

Works

71,186

Total citations

1,178,014

Keywords

FractureMeshless MethodsExtended Finite Element MethodPeridynamicsPhase-Field ModelingRadial Basis Functions