Search

Efficient models are needed in eddy current evaluation to enable fast and accurate predictions also in terms of probability of detection (POD). Here, calculations of the transfer impedance in a transmitter–receiver sensor are considered using the finite element method. It is shown that numerical integration around the flaw gives accurate values even if the solution without the flaw is calculated analytically which also is decreasing the computational time. A closed crack model is validated and used in a model based POD study showing limitations with transmitter–receiver sensors. It is finally suggested how to link the modelled POD to design parameters of components. [ABSTRACT FROM AUTHOR]

Abstract: The eddy current inspection of small fatigue cracks in Ti–6AL–4V is evaluated in both a finite element model and experiments. The crack was created in a fatigue process and an eddy current measurement was carried out as the resulting crack was subjected to different levels of static load. The signal showed a strong dependency of the time between the creation of the fatigue crack and the eddy current measurement. This dependency is proposed to be related to oxides forming on the crack faces. The oxide is favourable for the detection of fatigue cracks. The narrow width of the fatigue crack is important to consider in eddy current inspection and as static loads are applied across the crack faces, electrical connections arise within the crack, which has a strong influence on the eddy current signal. Four different models of the contact behaviour were implemented within the finite element model. It is shown that the electrical connections that arise within small fatigue cracks, as well as the influence from the narrow opening as tensile loads are applied, can be predicted by a finite element model of the eddy current method. [Copyright &y& Elsevier]