Preliminary study for flaw detection in nodular cast iron by cyclic loading and thermography

The objective of this paper is to deduce thermographic parameters from infrared (IR) picture series which indicate the presence of bulk defects in nodular cast iron. For this task, specimens manufactured with and without artificial defects are sinusoidally loaded in the range of the fatigue limit for a short period. During this time, the IR picture series were recorded with a high-speed IR camera. After a picture alignment, the time-dependency of the average gray values in picture subsets was analyzed by discrete Fourier transform (DFT) to get the magnitude and phase shift of the fundamental frequency of the local IR radiation power. The results of the experiments and FE simulations indicate consistently that the volume defects can be detected on the basis of the average and standard deviation of the phase shift distributions on the surface of the specimens.

► A new NDT method of volume flaw detection is presented using cyclic loading, thermography, DIC and DFT.
► Experimental data are compared with FE simulations.
► The flaws are detectable through the phase shift between cyclic loading and first harmonics of IR intensity.
► About 175 cycles beneath the fatigue limit are enough to detect the flaws.