What can be added to the fracture mechanics of metal fatigue that has not already been said since the 1900s? From the view point of the material and structure engineer, there are many aspects of failure by fatigue that are in need of attention, particularly when the size and time of the working components are changed by orders of magnitude from those considered by st traditional means. The 21 century marks an era of technology transition where structures are made larger and devices are made smaller, rendering the method of destructive testing unpractical. While health monitoring entered the field of science and engineering, the practitioners are discovering that the correlation between the signal and the location of interest depends on a priori knowledge of where failure may initiate. This information is not easy to find because the integrity of the physical system will change with time. Required is software that can self-adjust in time according to the monitored data. In this connection, effective application of health monitoring can use a predictive model of fatigue crack growth. Earlier fatigue crack growth models assumed functional dependence on the maximum stress and the size of the pre-existing crack or defect. Various possibilities were examined in the hope that the data could be grouped such that linear interpolation would apply.