Stochastic modeling of fatigue crack dynamics for risk analysis and remaining life prediction

This paper presents a stochastic dynamic model of fatigue crack propagation in metallic materials which are commonly encountered in mechanical structures and machine components of complex systems. The (non-stationary) statistics of the crack growth process are obtained without solving stochastic differential equations in the Wiener integral or Ito integral setting. The crack propagation model thus allows real-time execution of decision algorithms for risk assessment and life prediction on inexpensive platforms (such as a Pentium processor). The model predictions are in close agreement with experimental data of fatigue crack statistics for 2024-T3 and 7075-T6 aluminum alloys.