Vibrothermography nondestructive evaluation (NDE) is in the early stages of research and development, and there exists uncertainty in the fundamental mechanisms and processes by which heat generation occurs. Holland et al. [30] have developed a set of tools which simulate and predict the outcome of a vibrothermography inspection by breaking the inspection into three distinct processes: vibrational excitation, heat generation, and thermal imaging. The stage of vibrothermography which is not well understood is the process by which vibrations are converted to heat at the crack surface. It has been shown that crack closure and closure state impact the resulting heat generation [39; 55; 56]. Despite this, research into the link between partial crack closure and vibrothermography is limited [40; 55; 56].

This work seeks to rectify this gap in knowledge by modeling the behavior of a partially closed crack in response to static external loading and a dynamic vibration. The residual strains left by the plastic wake during fatigue crack growth manifest themselves as contact stresses acting at the crack surface interface. In response to an applied load below the crack opening stress, the crack closure state will evolve, but the crack will remain partially closed. The crack closure model developed in this work is based in linear elastic fracture mechanics (LEFM) and describes the behavior of a partially closed crack in response to a tensile external load and non-uniform closure stress distribution. The model builds on work by Fleck [18] to describe the effective length, crack opening displacement, and crack tip stress field for a partially closed crack. These quantities are solved for by first establishing an equilibrium condition which governs the effective or apparent length of the partially closed crack. The equilibrium condition states that, under any external or crack surface loading, the effective crack tip will be located where the effective stress intensity factor is zero. In LEFM, this is equivalent to saying that the effective crack tip is located where the stress singularity vanishes. If the closure stresses are unknown, the model provides an algorithm with which to solve for the distribution, given measurements of the effective crack length as a function of external load.

Within literature, a number of heating mechanisms have been proposed as being dominant in vibrothermography. These include strain hysteresis, adhesion hysteresis, plastic flow, thermoelasticity, and sliding friction. Based on experimental observation and theory, this work eliminates strain hysteresis, thermoelasticity, and plastic flow as plausible heating mechanisms. This leaves friction and adhesion hysteresis as the only plausible mechanisms. Frictional heating is based on the classical Coulomb friction model, while adhesion hysteresis heating comes from irreversibility in surface adhesion. Adhesion hysteresis only satisfies the experimental observation that heating vanishes for high compressive loading if surface roughness and the instability of surface adhesion is considered. By understanding the fundamental behavior of a partially closed crack in response to non-uniform loading, and the link between crack surface motion and heat generation, we are one step closer to a fully predictive vibrothermography heat generation model. Future work is needed to extend the crack closure model to a two-dimensional semi-elliptical surface crack and better understand the distinction between frictional and adhesion heating.