Comparison of various 9–12%Cr steels under fatigue and creep-fatigue loadings at high temperature (original) (raw)

The present article compares the cyclic behaviour of various 9-12%Cr steels, both commercial grades and optimized materials (in terms of creep strength). These materials were subjected to high temperature fatigue and creep-fatigue loadings. TEM examinations of the microstructure after cyclic loadings were also carried out. It appears that all the tempered ferritic-martensitic steels suffer from a cyclic softening effect linked to the coarsening of the subgrains and laths and to the decrease of the dislocation density. These changes of the microstructure lead to a drastic loss in creep strength for all the materials under study. However, due to a better precipitation state, several materials optimized for their creep strength still present a good creep resistance after cyclic softening. These results are discussed and compared to the literature in terms of the physical mechanisms responsible for cyclic and creep deformation at the microstructural scale. (B. Fournier). related dislocation glide and climb. Indeed such modifications of the microstrcuture are not observed after pure ageing [9,10]. Different obstacles hinder the dislocation motion such as LABs or high-angle boundaries (HABs), the other dislocations, large precipitates such as the M 23 C 6 (and Laves phases after long term creep tests), mainly at HABs , small precipitates such as the MX (NbC, NbN, VC, VN) [12] and the solid solution elements (Mo, W) . These kinds of microstructural features are present within the standard commercial steels, Grade 91 (P91) and Grade 92 (P92). Their creep resistance is close although the P92 steel is a bit more resistant. The large precipitates (diameter of about 100 nm) are usually located at the LABs and HABs and are supposed to pin the boundaries. The smaller precipitates (diameter of a few 10 nm) are usually homogeneously distributed, in the subgrain interiors as well as along the boundaries. As for the solid solution, they could reduce the mobility of the subgrain dislocations and boundaries. During the last ten years, additional material improvements have led to more resistant ferritic-martensitic steels. The reduction of the carbon concentration leads to the dispersion of fine MX precipitates along boundaries whereas the increase in the boron concentration leads to a reduced growth of the M 23 C 6 which thus pin more efficiently the boundaries during creep . This way, the minimum creep strain rate can be reduced by one order of magnitude leading to a lifetime longer by one order of magnitude. This is expected since the same Monkmann-Grant curve seems to be valid for almost all ferritic-martensitic steels and for a large temperature range . This can be explained by the simulation of the necking process combined with creep-induced material softening . The addition of nitrogen and tungsten leads to the formation of 0921-5093/$ -see front matter