Accurate laboratory measurements of crack initiation and growth are of vital importance for characterising material behaviour for use in the residual life assessment of structural components. The Potential Drop (PD) technique is one of the most common methods of performing these measurements, but such measurements are also sensitive to large inelastic strains which are often erroneously interpreted as crack growth. Despite the maturity of the PD technique, the extent of these errors is not fully understood and the most appropriate method of suppressing them is unknown.
In this thesis typical errors in the measurement of crack extension due to large inelastic strains have been quantified experimentally. These errors depend on the PD configuration and in some cases the configurations recommended in the standards are susceptible to particularly large errors. Optimum configurations for common fracture specimens have been identified but despite these mitigating measures, when testing ductile materials, the errors due to strain remain large compared to other sources of error common to the PD technique.
A sequentially coupled structural-electrical FE modelling approach has been developed which is capable of predicting the influence of strain on PD. This provides a powerful tool for decoupling the effects of strain from crack extension. It has been used in conjunction with experimental measurements, performed using a novel low frequency ACPD system (which behaves in a quasi-DC manner), to develop procedures for accurately measuring crack initiation and growth during fracture toughness and creep crack growth testing. It is demonstrated that some of the common methods of interpreting PD measurements during these tests are not fit for purpose. The proposed method of interpreting creep crack growth data has been used to re-validate creep crack initiation prediction models provided in the R5 assessment procedure.