Integration of Statistical and Physical Models of Short Fatigue Crack Growth.
Abstract
A computer program has been written that generates Monte Carlo simulations of small fatigue crack growth. The simulations allow predictions of the statistics of growth given postulated or empirical laws defining the effects of local stochastic microstructure. Variations in crack shape and irregularity of the crack front, correlations between surface and subsurface growth rates, and the probability and effects of part or all of the crack front being temporarily arrested at grain boundaries can all be calculated. The simulation of embedded and surface cracks of irregular shape is made feasible by the development of approximate, fast algorithms for estimating variations in the Mode I stress intensity factor, k sub i around an irregular plane crack. The simulations provide a convenient way of generating large, detailed, synthetic data bases for small crack growth under diverse conditions. These data bases can then be used in lieu of experimental data to test simpler, but more efficient and robust probabilistic models of crack growth such as the one developed in the first year of this program. They also provide the theoretical framework required for understanding the role of fluctuations in the shape of small cracks in determining their rate of growth. By comparing the predicted statistics of the shape of small cracks with experimental data such as from striations, postulated laws of the dependence of growth rate on stochastic microstructure can be tested.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1987
- Accession Number
- ADA179725
Entities
People
- B. N. Cox
- W. L. Morris