| Popis: |
Lives predicted by two crack growth rate models are compared with the lives measured in field and laboratory spectrum tests on rail steel. The two models are shown to be equivalent, except that one model reflects average rates and the other, the fastest, measured rates from constant amplitude tests. Comparison with the field test suggests that the stress intensity factor model used to represent a detail fracture growing in a rail head requires some minor modifications. The laboratory tests, performed with load spectra in real sequence order, are shown to be reasonable simulations of the field test. Both the field and laboratory spectrum test crack growth lives are shorter than the lives predicted from the average rate model. This result is attributed to acceleration (load interaction) in the tests. |
