Softening behaviour and correction of the fracture energy

The area under the load-displacement softening curve gives the total external work on the test specimen and not the fracture energy. The fracture energy follows from half this area that is equal to the critical strain energy release rate at the first crack increment. For wood this is correctly applied for mode II. For mode I however, as for other materials, the total area is wrongly regarded, a factor 2 is too high. In some applications, based on crack increment cycles, the error is even a multiple of this factor 2. On the other hand, the measurements at softening may show an apparent decrease of the specific fracture energy that can be explained by a small crack joining mechanism when the ultimate state of the ligament of the test specimen is reached. Post fracture behaviour is thus not comparable with the behaviour of macro crack initiation. It is further shown, by the kinetics of the process, that the irreversible work of an ultimate loading cycle is proportional to the activation energy of the fracture process and not to the driving force of the process. This explains why the crack velocity decreases with the increase of this irreversible work and increases with the stress intensity increase. The fracture energy is a function of the Griffith strength and is thus related to the effective width of the test specimen and not to the ligament length. This also has to be corrected. Based on the derivation of the softening curve, the reported fracture toughness of 720 kN m(-1.5) of double-edge notched tests is corrected to 330 kN m(-1.5) and the value of 467 kN m(-1.5) based on the fracture energy, of the compact tension tests, is also corrected to the right value of 330 kN m(-1.5). A revision of published mode I data, based on the fracture energy obtained by the area of the softening curve, is thus necessary. (C) 2007 Elsevier Ltd. All rights reserved.