Dimensionless numerical sensitivity analysis of narrow cracks by means of infrared lock-in thermography

Abstract

The non-destructive detection of defects in materials, such as cracks, delaminations and others, is a critical issue to be addressed in many technological applications like automotive, aeronautical or aerospace industries. In this frame, lock-in infrared thermography is identified as a highly suitable solution offering not only a qualitative detection of the defects but also their quantitative characterization. With the goal of maximizing its capabilities on the crack detection and characterization, in this work, a dimensionless numerical global sensitivity analysis of the lock-in infrared thermography is developed. First, a complete dimensionless reformulation of the thermographic investigation is provided, not only limited to the defect geometrical parameters, but also including the corresponding experimental parameters. As a consequence, the constraints of particular experimental setups or material properties can be removed by means of an appropriate choice of length, time and temperature scales. This leads to a set of dimensionless parameters and equations which preserve the full physical information of the experiment. The resulting model has been numerically solved and successfully validated by using experimental thermographic data over laboratory calibrated cracked material samples. Second, the developed dimensionless numerical model has been used as input for a global sensitivity analysis able to determine the correlations between the proposed dimensionless parameters and their corresponding impact on the thermographic results. Furthermore, the ranges of sensitivity and predominance of each dimensionless parameter are obtained, which provide quantitative parametric selection criteria for a maximized efficiency and accuracy on the thermographic crack detection and characterization.