I have always, and I believe correctly, considered cohesize/bridging models linear elasticity. This is because whether we apply them to holes or cracks, we are solving the equations of ELASTICITY, with nonlinear boundary conditions along the part(s) of the crack along which a traction-separation law is applied. So the material model is totally elastic. All we did was to augment the elasticity theory with an ad hoc physical condition (inspired by experiments such as Dugdale).
Recent comments