Frequency dependent heat generation during vibrothermographic testing of composite materials
This investigation concerns the frequency dependent heat generation behavior and the heat generation mechanisms for the thermal patterns of delamination in fiber reinforced composites during a vibrothermographic test, which uses real time thermography as a nondestructive evaluation of a structure or a component excited with mechanical vibration. A local resonance model was proposed in the past to describe the frequency dependent heat generation behavior during a vibrothermographic test, and this model was used as a basis for writing software for calculating the natural frequencies of a plate with the size of delamination. Vibrothermographic tests were performed on three glass-epoxy panels that each contained four different sized simulated delaminations. Comparison between the observed vibrothermal peak frequencies and the natural frequencies predicted by the local resonance model, and investigations of the thermoelastic emission field in the delamination region using SPATE, were made to determine the validity of the local resonance model. A significant conclusion of the results is that the local resonance is indeed the mechanics model for the frequency dependent heat generation behavior.
A careful measurement of the degree of heating of both sides of [0₅] glass-epoxy panel with delaminations on the 2-3 ply interface, and comparison between the predicted heat patterns generated from a finite difference heat transfer program and observed heat patterns, was made to identify the heat generation mechanism. The results show that the majority of heat generation during vibrothermographic testing results from higher stresses or strains due to local resonance. The heat generation was affected by the combination of the principal strains and shear strain for the lower modes of resonant vibration, and was dominated by the shear strain for the higher modes of resonant vibration.
Impact damaged graphite-epoxy panels were also inspected constituting an application of vibrothermography on real damaged components. The degree of heating of the damage were measured through a frequency range, and the damage severity was inspected by ultrasonic C-scan and edge replication. From comparison of two plots of the degree of heating versus exciting frequency, either the area under the curve or the number of vibrothermal peak frequencies, the severity of the damage can be qualitatively identified.