A novel time domain analysis approach for thermal non-destructive testing and evaluation of solids

Abstract

In recent decades, InfraRed Thermography (IRT) for Thermal Non-Destructive Testing & Evaluation (TNDT&E) has encountered wide spread applications for the characterization of various solid materials due to its non-contact, whole-field, quick and non-invasive inspection capabilities. The principle of this technique is based on the mapping of thermal profile over the test object in order to reveal its surface and sub-surface anomalies. Due to potential capabilities of IRT, it finds numerous applications in various fields such as electrical, aeronautical, civil, mechanical, automotive, and bio-medical engineering etc. It can be implemented either in passive or active mode for NDT&E applications. In passive thermography, natural thermal profile of the test object is obtained in order to detect its surface and sub-surface anomalies in the absence of any external heat stimulus. Inadequate thermal contrast from deeper sub-surface features, and inability to provide quantitative analysis, limit the applicability of this approach. In contrast, active thermography uses a pre-defined controlled heat stimulus to launch thermal waves into the test object. In order to extract deeper sub-surface information, various signal, image and data processing schemes are further employed onto the recorded temporal thermal response of the object. Among the various active infrared thermographic methods, Pulse Thermography (PT), Lock-in Thermography (LT), Pulsed Phase Thermography (PPT) and Frequency Modulated Thermal Wave Imaging (FMTWI) are predominately in use. FMTWI overcomes the requirement of high peak power heat sources of pulsed based techniques (PT & PPT) and repetitive experimentation of LT. It probes thermal waves into the test object within a suitable band of frequencies in a limited time span decided by thermal properties of the specimen and its physical dimensions. Frequency domain phase based data analysis scheme redistributes the imposed energy into the individual frequency components leading to a limited test resolution and sensitivity for detecting the sub-surface defects with a chosen frequency component. In an attempt to overcome the limitations of frequency domain phase based approach, this work introduces a novel matched filter based time domain (phase and correlation coefficient) analysis scheme. This analysis makes use of the advantages of concentrated energy in time achieved through matched filtering approach adopted onto the obtained thermal response to non-stationary aperiodic excitation. This is achieved by analyzing phase information in the time domain instead of frequency domain to characterize the specimen without disintegrating the energies of associated frequencies contributed in defect detection. The defect detection performance of FMTWI is further enhanced by reshaping the spectra of frequency modulated signal with Gaussian envelope function. Detection enhancement facilitated by incorporating Gaussian window function has been studied and compared with the conventional FMTWI. In order to increase the probing bandwidth and energy for better depth resolution, the continuous chirp has been digitized and used for stimulation. Analytical basis for temperature evolution with this Digitized Frequency Modulated Thermal Wave Imaging (DFMTWI) has been developed and its defect detection capabilities have been studied by using novel time domain phase analysis scheme and compared with conventional frequency domain approach. Further, applicability of novel coded excitation schemes have been introduced to infrared thermography. Investigations have been carried out for finding out the capabilities of these methods. Features and potential abilities of pulse compression method have been illustrated.