An investigation of the factors determining creep strength and ductility in Grade 92 steel

Summary The ECCC creep rupture data sets, initially collated for design strength assessment, can also provide a valuable resource for wider investigations. Thus, variations in properties between different casts or batches with different chemical composition, heat treatment, manufacturing route and product form can be analysed. This can be an effective means of determining how alloy composition and heat treatment may be optimised. This paper describes an investigation into the causes of the large cast-to-cast variations in long term creep ductility observed in the ECCC Grade 92 data set. The analysis shows that a major factor, accounting for the bulk of the variability, is the normalising and tempering heat treatment. Under-normalised batches, especially with normalising times below 30-60 minutes, show inferior long term ductility. Light tempering may exacerbate these effects. Under-normalising is also well correlated with high ambient temperature tensile strength and with high creep rupture strength under short term test conditions, when virtually all test failures are highly ductile. In longer term uniaxial creep testing, when the under-normalised batches show much reduced ductility, their relative uniaxial creep rupture strength values tend to be somewhat poorer, though not exceptional. The potentially greater harmful influence of low creep ductility on performance under multiaxial loading and cyclic operation merits further study. Long term uniaxial creep rupture strength tends to be relatively poor when batches are either severely over- normalised and tempered (hence weak and ductile), or severely under-normalised and tempered (hence strong but brittle). However, there is also substantial scatter, no doubt due to compositional variations. Hence, whilst intermediate levels of manufacturing heat treatment broadly appear to be optimal, some batches with intermediate heat treatment levels do also appear amongst those with the poorest long term creep performance. The microstructural factors which may be involved, and relationships with parallel Japanese work on Grade 91, are discussed. In conclusion, there is every prospect that low ductility Grade 92 can be avoided by suitable heat treatment specifications.