Cleavage fracture and irradiation embrittlement of fusion reactor alloys: mechanisms, multiscale models, toughness measurements and implications to structural integrity assessment

We describe the highly efficient master curves-shifts (MC-AT) method to measure and apply cleavage fracture toughness, K-Jc(T), data and show that it is applicable to 9Cr martensitic steels. A reference temperature, T-0, indexes the invariant MC shape on an absolute temperature scale. Then, To shifts (AT) are used to account for various effects of size and geometry, loading rate and irradiation embrittlement (DeltaT(i)). The paper outlines a multiscale model, relating atomic to structural scale fracture processes, that underpins the MC-AT method. At the atomic scale, we propose that the intrinsic microarrest toughness, K-mu(T), of the body-centered cubic ferrite lattice dictates an invariant shape of the macroscopic K-Jc(T) curve. K-Jc(T) can be modeled in terms of the true stress-strain (sigma-epsilon) constitutive law, sigma (T, epsilon), combined with a temperature-dependent critical local stress, sigma*(T) and stressed volume, V*. The local fracture properties, sigma*(T)-V*, are governed by coarse-scale brittle trigger particles and K-mu(T). Irradiation (and high strain rate) induced increases in the yield stress, Deltasigma(y), lead to DeltaT(i) with typical DeltaT(i)/Deltasigma(y) 0.6 +/- 0.15 degreesC/MPa. However, DeltaT(i) associated with decreases in a* and V* can result from a number of potential non-hardening embrittlement (NHE) mechanisms, including a large amount of He on grain boundaries. Estimates based on available data suggest that this occurs at >500-700 appm bulk He. Hardening and NHE are synergistic, and can lead to very large DeltaT(i). NHE is signaled by large (>1 degreesC/MPa), or even negative, values of DeltaT(i)/Deltasigma(y) (for Deltasigma(y) < 0), and is often coupled with increasing amounts of intergranular fracture. The measured and effective fracture toughness pertinent to structures almost always depends on the size and geometry of the cracked body, and is typically significantly greater than K-Jc. Size and geometry effects arise from both weakest link statistics, related to the volume under high stress near a crack tip, and constraint loss associated with large amounts of deformation in small specimens and shallow surface cracks. We describe micromechanical models that can be used to adjust the toughness measured using small specimens to both the intrinsic material K-Jc and the effective toughness pertinent to a structure. Finally, using a simple example, we illustrate the profound implications of size-geometry effects on the fracture of fusion structures. This assessment is based on a metric of strength and ductility, specified as the ratios of the critical load and displacement at fracture to the corresponding yield load and displacement, P-c/P-y and Delta(c)/Delta(y), respectively. Even in cases where the material experiences very brittle elastic fracture in standard tests, or in heavy sections, with Pc/Py < 1, the extrinsic factors pertinent to fusion structures (small shallow cracks in thin sections, etc.) lead to P-c/P-y > 1 and Delta(c)/Delta(y) much greater than 1. Indeed, in some circumstances, the benefits of irradiation due to increases in P-c may more than offset the liabilities of the decreases in Delta(c). (C) 2003 Elsevier B.V. All rights reserved.