Please choose your delivery country and your customer group
Optically excited lockin-thermography (OLT) becomes increasingly popular since it is robust and easy to apply. However, the full potential of this method is not used, if only one lockin-frequency is evaluated. More information can be gained by correlating two or more measurements. This allows to gather information on features like the kind of defect, thickness of the specimen, properties of boundary layers, lateral heat flow etc. Certain defects are correlated to specific patterns in a scatter plot, akin to a fingerprint. Specimens with and without a certain feature can be distinguished more easily using suitable filters in the scatter plot. This is apparently of importance for automated defect detection and classification. This paper covers feature-extraction using scatter plots. A wedge was used to measure the depth dependence of the phase angle for two lockin-frequencies. According to the thermal wave theory the scatter plot of the wedge shows a closed curve. Another specimen located next to the wedge contains flat bottom holes in different depths drilled from the rear side. The lateral heat flux near the holes results in a different shape of the correlated scatterplot curves, and therefore can be extracted from phase images in order to reveal hidden structures and defects.