Abstract
Biaxial stress rapture tests under constant internal gas pressure have been carried out on 20% Cr ; 25% Ni : Nb stainless steel thin walled tubing at 650°C, 750°C and 850°C. The initial tests examined the effect of grain size, carbide distribution, heat treatment temperature and cold work, content on the strength and rupture strain of the austenitic stainless-steel. For creep strength, it was found that a heat treatment temperature of 1100°C was essential to solution the Nb & C in austenite for subsequent matrix strengthening by fine precipitation of NbC at typical service temperatures Optimum creep strength was obtained by a combination of high solution treatment temperature (1100-1150°C) with 5 per cent cold work and an ageing treatment at 800°C for dispersion strengthening by precipitate dislocation interactions Treatments that improved the creep strength adversely affected the ductility as measured by the rupture strain. Creep strength was reduced and ductility restored by using heat treatment temperatures of ≤ 1000°C and time combinations that overaged and coarsened the NbC precipitates" In this condition coarse grained structures 70-175 micron size had similar ductility and only marginally better strength than structures of 12-15 micron grain size. Creep and rupture life expressions were formulated from analyses of the stress rupture data. The activation energies were similar to that for self diffusion in austenite and the stress exponents, n, were typical of those obtained for other austenitic steels. The stability of structures with thermal and mechanical treatments that gave the best creep strength was examined by a series of ageing treatments at a typical service temperature of 800°C for times up to 10,000 hrs. The strength was reduced to a constant value and the ductility increased after 1000 hrs ageing by coarsening of the original fine NbC precipitates and recrystallisation of the cold worked matrix. A constant relationship between creep rate and rupture life - etr = 10(%) was found for this material in the weak overaged condition. The constant was independent of pretreatment, stress and test temperature. Additional stress rupture tests were made on material from eleven different billets of 20% Cr : 25% Ni : Nb stainless-steel and a similar constant relationship was obtained. The explanation for this phenomenon was found by analysing theoretically the strain time relationship of biaxial and uniaxial stress rupture tests and comparing them against experimental creep curves. The constant of 10 represents the strain the material would have attained under constant stress in the same time as that to reach the end of secondary creep under constant pressure. The biaxial creep and stress data were converted to effective stress and effective minimum creep rate using the Levy-Mises criterion for deformation so that comparison could be made with uniaxial data. Uniaxial tests were carried out and compared with the converted biaxial data and good agreement was observed. There was no effect of test temperature on the rupture strain whereas at creep rates of < 0.005 %/h a pronounced drop in ductility occurred that was independent of heat treatment and test temperature. In order to establish the meaning or usefulness of rupture strain as a ductility parameter, ruptured specimens were examined metallographically to assess the effect of creep rate and strain on the density of cavitational damage. Wedge cracks were present in the surface regions at creep rates of 0.15-0.005 %/h. However, at creep rates < 0.005 %/h. where the low rupture strains occurred, spheroidal cavities were distributed uniformly throughout the matrix. Strains of ~5 per cent were necessary to initiate the wedge cracks and failure occurred by the growth and propagation of surface initiated cracks. Spheroidal cavities were present at < 1 per cent strain and failure occurred at low strains by interlinkage of cavities.