Automatic Algorithm for Quantitative Pulsed Phase Thermography Calculations

Pulsed Phase Thermography (PPT) has been successfully applied for defect detection purposes on a variety of materials. A great deal of work has been done to evaluate the potential of PPT for quantitative applications, using for instance statistical methods, Neural Networks or wavelets (Ibarra-Castanedo et al., 2004). However, calibration requirements and lengthy computation subroutines, preclude their use on most NDT applications. A new inversion technique, based on phase delay data, has been recently proposed by the authors (Ibarra-Castanedo and Maldague, 2004a). Quantification is carried out by correlating the defect depth with its corresponding blind frequency, f b , i.e. the frequency at which the defect becomes visible on the frequency spectra. Estimation of f b is performed however using phase contrast calculations; as a result a non-defective zone on the sample surface is needed. As will be stressed, provided that thermal data is correctly sampled, phase profiles present a characteristic pattern that can be used on automatic f b retrieval for a particular defect depth without defining a sound area. Introduction: Pulsed Phase Thermography (PPT) combines interesting features from two older thermographic techniques, i.e. PPT is as rapid and easy to deploy as Pulse Thermograhy, and, after processing the thermal data, it provides phase delay images as Lock-In Thermography. It is well known that phase is less affected than thermal data by problems such as non-uniform heating, surface emissivity variations and non-planar surfaces. Accordingly, PPT is a safe and easy to deploy NDT technique, giving the possibility to rapidly inspect large and complex surfaces. Figure 1 summarizes the different steps involved on PPT. The specimen surface is first stimulated with a thermal pulse, e.g. using photographic flashes n, varying from a few seconds to several milliseconds depending on the thermal properties of the material being inspected. Once the pulse reaches the specimen o, the thermal front propagates through the material while the cool down process begins at the surface. The principle of defect detection is based on the fact that, at the surface, defective zones will be at higher or lower temperatures with respect to non-defective zones, depending on the thermal properties of both, the material and the defect. The temperature evolution on the surface is monitored using an infrared camera p. A thermal map of the surface, or thermogram, is recorded at regular time intervals. A 3D matrix is then formed, in reference to Figure 1, x and y …