Vibrothermography is a nondestructive testing technique that uses vibration-induced heating to locate cracks inside parts. Industrial application of vibrothermography has typically been limited due to a lack of complete understanding of the factors influencing vibrothermographic crack heating. In other words, in otherwise identical circumstances, some cracks heat differently. This thesis presents modified versions of three journal papers related to developing a predictive forward model of the vibrothermography process. The first paper discusses an empirical model for vibrothermographic crack heating, including factors for vibrational amplitude, crack mobility and crack closure. The paper provides details of the model and the data collection process. This paper also demonstrates a method for measuring crack closure with vibrothermography. An accompanying appendix also provides complete experimental results. The second paper describes an unexpected increase in crack heating after introducing a lubricating oil into a crack. It was also discovered that the amount of increase in heating decreases as excitation amplitudes are increased, to the point that, at high excitation amplitudes, overall crack heating may decrease. A proposed hypothesis is presented. A third paper discusses a tool developed to help manage and interpret the large quantities of data collected during this work. Databrowse is an extensible web-based platform for viewing, querying, and transforming laboratory data. The details of the system and our experiences are provided.