Cross Correlation of Eddy Current Images for Detection of Fatigue Crack Propagation
There is a need within the NDE industry to evaluate the growth of damage such as fatigue cracking and plastic deformation over the life of a structure. To accomplish this aim, data collected over a period of time must be compared spatially while quantitative measures must be developed to determine the extent of crack propagation in a damaged region. To facilitate automated inspection and characterization of defects in metals, the application of cross-correlation techniques and frequency mixing to sequential eddy current images of fatigued welds and fatigue cracks in compact tension and reduced section specimens has been investigated. Results indicate that cross-correlation techniques can be used to align scan images to provide a history of damage development and that frequency mixing can enhance the signal-to-noise ratios of fatigue cracks in welded steel. Localized damage was detected by segmenting the images into smaller cross-correlation regions for improved resolution. The technique may be used to examine magnetic eddy current data from ferrous welds and seams and to identify cracks or regions of cold working in transportation and energy-generation systems.