On crack propagation shapes in elastic bodies (Q1881498)

The paper deals with evaluation of the energy functional and its derivatives for a mixed boundary value problem of linear elasticity in a domain with a singularly perturbed boundary. An elastic body \(\Omega\) contains a crack \(\Gamma^\delta\) that includes a straight (unperturbed) section and a small extension determined by a function \(x_2= \psi(x_1)\) over a small interval \((x_1\in(0,\delta))\). The authors claim that the derivative, with respect to \(\delta\), of the energy functional ``does not depend on the function \(\psi\) provided that function \(\psi\) is chosen from a quite narrow class, i.e., if this function is smooth enough''. Furthermore, the article states that ``the Griffith criterion does not distinguish any possible crack propagation shapes determined by the functions \(\psi\in W^{2,0}_p(0,1)\)''. One would expect to see an analysis of singular fields in the vicinity of singularly perturbed boundary \(\Gamma^\delta\), but such analysis is absent in the present article.