Application of the singular value decomposition to lock-in thermography for flaw characterization

In this work, a methodology for fast and accurate estimation of the geometrical parameters of buried flat flaws by means of lock-in thermography is proposed. With this goal, first, a modulation frequency selection procedure is developed which allows to select the optimal experimental frequencies for flaw characterization. Second, the application of the singular value decomposition as a processing tool to the lock-in thermographic problem is studied. As a result, its reduction to the first singular value has been demonstrated, which enables an effective space-frequency uncoupling. In consequence, the shape and depth of the studied defect can be accurately determined from space and frequency domains respectively. Overall, the proposed method provides a complete experimental and theoretical methodology for an accurate and very fast flaw geometrical characterization. • Lock-in infrared thermography as non-destructive technique for a fast and accurate flaw characterization. • Development of a modulation frequency selection procedure for optimal defect characterization. • Application of the singular value decomposition to lock-in thermography for flaw geometrical parametric retrieval.