Structures and parts are normally designed so that their in-service deformation is solely elastic and at such a level as to not induce fatigue failure. However, for advanced applications such as gas turbines, electric power plants and nuclear reactors, it becomes impossible to avoid repeatedly stressing parts beyond their elastic limit during critical phases of operation such as start up and shut down. As a result fatigue lives for materials are reduced from many millions of cycles in the elastic range to thousands in the plastic range.
Low cycle fatigue tests which should simulate in-service plastic stressing are not standardized. The purpose of the project is to provide a clear understanding of the effects of test parameters in order that the standardization process may proceed upon a sound basis.
Four materials, 2 strain softening, 1 strain hardening and 1 strain stable were purchased and, following homogeneity testing, were distributed to the participants. Each of these made measurements in accordance with a well defined set of guidelines using their normal design of specimen and types of extensometers plus a range of failure criteria. The results of the intercomparison show that the differences between laboratories are frequently greater than a factor of 10 whilst the variations within any given laboratory are significantly less. Although the scatter depends of a variety of parameters it appears that the use of ridged specimens leads to much reduced lifetime. Excluding the results of such specimens, the scatter remains unacceptable for reasons which were not resolved by the intercomparison. However, subsequent research suggests that even small amounts of specimen misalignment could account for it. Experiments are therefore being undertaken to confirm this.
Funding SchemeCSC - Cost-sharing contracts
TW11 0LW Teddington