Quantitative microbolometer-based thermoelastic stress analysis

The use of microbolometers for Thermoelastic Stress Analysis (TSA) has the potential of substantially reducing deployment costs, as well as providing more compact and lightweight solutions in comparison to more costly, bulky and less rugged photon detector based infrared cameras. However, microbolometer performance in terms of response time and sensitivity is less than what can be achieved with photon detectors. As the use of microbolometers for TSA is becoming more widespread, it is timely to devise a means of assessing microbolometer performance and identifying the limitations of using such a system for TSA. A simulation approach is devised that enables the effect of the microbolometer response time on transient signals (such as that from TSA) to be investigated. As the TSA involves measuring a small transient temperature change, usually induced by sinusoidal cyclic loading, the simulation approach also includes the effect of a noisy non-sinusoidal signal. To identify any deleterious effects cause by a cyclic load, the transient temperature change is also generated using an optical chopper and a black body source. The variables studied are the Signal-to-Noise Ratio (SNR), the impact of different waveforms, variation in the signal mean, as well as in-built system features such as the noise reduction. A new experimentally validated calibration approach is presented that accounts for the microbolometer response time and rolling averages introduced by inbuilt noise reduction features. It is shown the calibration approach is independent of specimen material and camera frame rates, and it is demonstrated on a CFRP test specimen containing a hidden crack.