A revision of the double-torsion technique for brittle materials

Abstract

Serious doubts about the constant-K characteristics of double-torsion specimens have been expressed in recent years. However, compliance calibration shows that the energy release rate and hence the stress intensity factor are independent of the crack length in this technique. Relatively large scattering in the results obtained by the load-relaxation method makes the data unreliable. The load-relaxation experiments should be carried out in combination with other methods. A theoretical calculation of the load-deformation curve under a constant displacement rate gives a rough estimation of the stress corrosion index. The subcritical crack growth data can be given, in principle, by only one experimental run of the constant displacement rate method of double-torsion testing.