Thermal equivalence networks for analysis of transient thermography

In recent years, advances in high-speed infrared camera technology have opened up many new fields of application. One area which benefited greatly is the field of thermal testing. Increased sensitivity and faster sampling rates have opened a bridge to thermal testing techniques that could previously only be performed via electrical contacts or thermocouples. An example is the method network identification by deconvolution, which is investigated here. This work begins with an extensive collection of the mathematical theory, established practices as well as modern research results. Detailed descriptions of all procedures implemented in this work are given. To evaluate the established algorithms systematically, test series on theoretical thermal transients are conducted. Special measures of accuracy are developed to rate the performance of each method. Building on these results, a new method for network identification is presented, which is called optimization-based network identification. The method achieves a greatly enhanced accuracy and noise resistance. Tests on thermal transient measurement data confirm these findings by improving established analysis techniques such as the transient dual interface method. The combination of network identification and infrared thermography resulted in the invention of a new non-destructive testing technique, which was patented as a result. The work concludes with three application examples using this thermographic network identification.