Thermoelastic and elastoplastic effects measured by means of a standard thermocamera

Thermography has been applied in a frequency range that is seldom analyzed in the literature: above one hundred Hertz. The high excitation frequencies and the uniform specimen loading allowed the assessment of the thermoelastic effect, even using an ordinary thermocamera, which is commonly assumed to be inadequate to evaluate thermoelasticity. The thermoelastic effect is assessed by means of a procedure based on Fourier analysis, so it is possible to filter the signal noise. The methodology allows the evaluation of both loading frequency and stress amplitude. The first one is estimated with an error depending mainly on spatial resolution of analysed images; the second one shows larger variance. An error analysis has been performed thanks to the strain gage positioned on the back of the specimens. This analysis has proved that the methodology is able to give an unbiased estimate of stress amplitude, consequently lower variance can be obtained if results coming from more than one image are averaged. This methodology should not be regarded to as an alternate method to differential thermography (specimens need to be uniformly stressed), but it can be successfully implemented whenever standard thermography has been planned for the analysis of a uniformly stressed specimen. This allows one to check correct specimen loading and constraining conditions (presence of instability, specimen faults, etc.). Moreover the authors have numerically estimated the ratio of elastoplastic to thermoelastic effect in order to define in which conditions the evaluation of thermoelastic effect is crucial for the elastoplastic measurements accuracy.