Extension of the Griffith's fracture criteria to saturated clays

Inglis 1] has solved the problem of distribution of stress in an elastic plate around an elliptical hole. His works clarify the role of cracks in the failure of an elastic material. However, his solution cannot be applied to saturated clay because he considers only total stresses, while, in saturated clay, the criterion of rupture should be expressed in terms of effective and not total stresses. The solution of Atkinson and Craster 2] using Biot's poroelasticity theory, shows that there is no high pore pressure in the vicinity of the crack tips for saturated clay. The major difference between this approach and the Biot's theory of is that, in saturated clay, strain is a function of the variation of the effective stress 3], while, in poroelastic media, strain is only a function of the variation of the total stress 4, Equation 2.2]. Also in their solution there is continuity between the pore fluid and the inner fluid in the crack. Their solution is valid for poroelastic media involving a movement of the pore fluid. In our solution there is no movement of the pore fluid (Undrained condition). In this paper we have solved the same problem as Inglis 1], but for the particular case of saturated clay obeying elastic law. By solving this problem we obtained the expressions for pore pressure, effective stress, total stress and displacements. The results show that not only the total stress but also the pore pressure and the effective stress are also high in the vicinity of the crack tips. A new failure criterion, based on Griffith's strain energy principle 5] and maximum tensile stress 6], valid for saturated clay is developed in this paper. Copyright © 2003 John Wiley & Sons, Ltd.