A number of elastic-plastic fracture mechanics parameters have been suggested for the description of fatigue crack growth, of which three are considered in this paper: the J-integral, crack-tip opening displacement, and plastic zone size. A critical comparison of the effectiveness of each parameter may be devised from a study of Mode I fatigue crack growth rates under various states of biaxial stress. Data for one heat of AISI 316 austenitic stainless steel at 20°C and 550°C showed that, if crack closure is taken into account, the cyclic plastic zone size is the most suitable correlating parameter for stress levels up to the yield value. A second heat of 316 stainless steel showed a preference for crack opening displacement control of crack growth. The analysis is based on the Dugdale model for crack-tip plasticity.

A metallographic examination of elevated temperature multiaxial fatigue specimens of 316 stainless steel was conducted to identify the type of damage and other microstructural changes that occurred during testing. The work hardening which takes place at the beginning of the test is associated with the development of either a cell or a maze dislocation structure. Two main factors contribute to the flow stress, that is, dislocation and carbide precipitation hardening. Dislocation hardening prevails, but precipitation hardening, which increases with testing time, becomes significant for long hold-time and low-strain range tests. Fatigue cracks are predominantly transgranular. However, intergranular cracks have been observed in hold-time specimens and are related to grain-boundary sliding and grain-boundary precipitation. The fatigue fracture is a three stage process, the initial stage (Stage I) is crystallographic, extending for one to three grains. The intermediate stage (Stage II) is dominated by striations, with the exception of some featureless facets in biaxial specimens. The final stage (Stage III) is a dimple type fracture (tension fracture), beginning at shorter crack lengths in the hold-time and high-strain range specimens. Striation spacing calculations give approximate life estimates in low-cycle fatigue.

Nonproportional biaxial fatigue tests were performed using cruciform-shaped specimens of AISI 316 stainless steel at room temperature, in order to study the effect of nonproportional straining on crack growth rate. While the fluctuating stress applied perpendicular to the crack was common to all the tests, including the uniaxial test, the stress parallel to the crack was applied for only half the former period and only on either the raising or the downward part of the cycle. The crack growth rate of the specimen subjected to uniaxial loading was more than five times faster than the growth rate of the cracks in the biaxial specimen, which is contrary to the reduced endurance observed for nonproportional loading of 1% chromium-molybdenumvanadium steel in tension-torsion. Transmission electron microscopy observations of regions next to the fracture surface showed a very different dislocation substructure relating to uniaxial and biaxial loading. The degree of strain hardening in both conditions was estimated by employing dislocation models suitable for each case. If a “slipping off” mechanism for crack propagation is adopted, then the difference in crack growth rate experienced by the biaxial specimens can be explained by considering the magnitude of the strength of the slipband in each case.