Some inherently unreliable practices in present day fracture mechanics

A number of current practices in fracture mechanics which use quantities near a crack tip to make conclusions about response at the crack tip itself are examined. Specifically these include: stress and displacement matching to estimate stress intensity factors, monitoring local stress and strain values to predict fracture, and both crack opening angle and crack opening displacement as fracture criteria. By means of a pair of counter applications, all of these procedures are demonstrated to have the potential of leading to completely incorrect conclusions. An understanding of what causes this inadequate performance then indicates that such procedures may be unreliable in general and prompts suggestions as to alternatives.