Stress state factor evaluation based on a fractographic analysis for use in the crack growth FASTRAN retardation model of the AFGROW computing code

The FASTRAN crack growth retardation model uses stress state descriptors to consider the local stress state ahead of the crack tip. Using the FASTRAN model implemented in the AFGROW computing code, the descriptor is input into the retardation model as a constant tensile (compressive) constraint factor along with the corresponding crack growth rate. per presents an evaluation based on a fractographic analysis of fracture surfaces of middle tension M(T) specimens from Al-alloy, which were used for the standard crack growth rate measurement under constant amplitude loading. There are typically three zones on the fracture surface: the flat zone that is perpendicular to the loading force and represents the tension mode of specimen cracking, the slant zone that is approximately 45° deviated from the loading force direction and a transient zone between them. It is shown that a significant correlation between crack length, local crack growth rate and the sizes of the zones exist. lengths relevant to the zones and corresponding crack growth rates were processed by statistical procedures. A continuous uniform distribution is a very good statistical model of the logarithm of the crack growth rates on both sides of the transient zone. In the retardation model, this distribution can be accounted for by using the means of the probability functions. scale structural model was tested for fatigue crack growth under a random loading process representing real operational conditions. Using the typical values of constraint factors for a “pure” plain stress and plain strain states and the crack growth rates evaluated by the fractographic analysis as input parameters, the crack growth curves calculated by the AFGROW code using the FASTRAN retardation model were compared with the crack growth data measured in the experiment. GROW output crack growth curves fit to the experimental measured ones without any additional tuning of the retardation model parameters.