Creep-rupture ductility of engineering materials

This Special Issue contains papers from the Low Rupture Ductility of Materials Workshop held at the University of Loughborough on 2/3 June-2016, which was organised by the High Temperature Committee of the Institute of Mining, Minerals and Materials (IOM³) in collaboration with the ESIS High Temperature Materials Testing Committee (HTMTC). The main focus of the presentations was on metallurgical considerations, modelling and testing, with aims of:

•	considering what is the most appropriate measure of creep-rupture ductility, and whether it should be regarded as a material property, and
•	identifying what is an acceptable level of creep-rupture ductility, in particular for the creep strength enhanced ferritic steels.

The workshop had been dedicated to the memory of Doug Thornton who had sadly passed away in March 2015. Starting his working life as an English Electric apprentice in the 1960s, he had risen to become Chief Metallurgist of GEC Turbine Generators (ultimately to become Alstom). Over the years, Doug was at least in part responsible for the industrial implementation of many engineering metallurgical advances, perhaps most notably the practical adoption of advanced martensitic 9/10%Cr steels in the form of large forgings and castings for super-critical turbo-machinery components. Outside Alstom, he was best known as the co-founder of ECCC (the European Creep Collaborative Committee), which under his leadership during the 1990s and early 2000s was responsible for many important European industry coordinated recommendations concerning long-time creep properties for inclusion in European Product and Design Standards.

Selected papers cover the Creep ductility of 1CrMoV rotor steel, by Stuart Holdsworth, and Creep ductility considerations for high energy components manufactured from creep strength enhanced steels, by Jonathan Parker. A paper by Peter Skelton concerns the Deformation, diffusion, and ductility during creep in terms of continuous void nucleation and creep-fatigue damage. Finally, two papers by David Woodford deal with: Intrinsic ductility for structural materials as a function of stress and temperature, and a Comparison of creep strength and intrinsic ductility for serviced and reheat treated T91 steel based on stress relaxation testing. Other Ductility Workshop contributions are planned for inclusion in future issues of Materials at High Temperatures.

Traditional measures of creep-rupture ductility are the final elongation (A_R) and/or reduction of area (Z_R) determined from uniaxial specimens at final fracture, as used by Holdsworth and Parker. However, it is acknowledged that rupture elongation in particular, but also reduction of area at rupture are influenced by specimen geometry, e.g. the dimensions of gauge section diameter and parallel length, and their relative proportions. An alternative to these quantities is the true creep strain at rupture determined from Z_R (see Holdsworth), but this parameter can also be sensitive to specimen geometry. As can be seen from Figure 1, a significant contribution to A_R (and Z_R) is the tertiary creep strain, and it is the sensitivity of this component to specimen geometry, material response, the extent of necking and the associated influence of multi-axiality which means that the traditional quantities cannot really be regarded as unique material properties. Nevertheless, A_R (and Z_R) are parameters which have been routinely collected (mostly with standard specimen geometries) for at least the last 60 years, and are thereby an invaluable commodity (in particular if used in a qualitative way which acknowledges the limitations associated with these parameters). These are the ductility parameters which are commonly found in most material creep property databases.

Figure 1. Variation of creep strain with time to rupture.

A general rule-of-thumb is that low creep ductility A_R is less than 5%. This originated from the consideration of a large body of evidence for 1CrMoV steel. However, there is no shortage of evidence to show that such a threshold ductility level is very material dependent, e.g. [1,2].

Many steels exhibit a loss in creep ductility with time at temperature, under stress, which can ultimately recover after very long times. Perhaps one of the most widely studied steels in this respect is 1CrMoV steel, and the Special Issue paper of Holdsworth reviews the evidence for this alloy, with a particular focus on the influence of material pedigree. Some practical implications of low creep ductility in this class of steel are also considered. While 1CrMoV was one of the most investigated steels in the 1970s and 1980s, more recently by far the main interest is in the advanced martensitic creep strength enhanced steels. The Parker Special Issue paper considers the evidence for P91 and P92 creep strength enhanced steels.

The sensitivity of the tertiary strain component of A_R (and Z_R) to specimen geometry has at least in part been responsible for the consideration of other creep ductility quantities, such as the Monkman–Grant ductility [3] (A_MG, as considered by Skelton) and the intrinsic ductility parameter proposed by Woodford. The A_MG parameter has been adopted by a number of research engineers, but is not widely accepted, not least because it can ignore a significant primary creep strain component, which may be regarded as unjustified. Primary creep strains can be as high as 0.5% in some alloys. Nevertheless, A_MG is importantly not dominated by a specimen dependent tertiary creep strain component, and the ratio λ = A_R/A_MG is regarded by some as the correct measure of creep ductility since it provides an accurate means of representing the capacity of a material for stress redistribution [4].

The Skelton Special Issue paper deduces a Monkman–Grant ductility strain rate relationship which is used to calculate the cyclic creep damage in tests on austenitic steels at temperatures in the range 550–650 °C. The approach is responsible for lower bound ductility estimates compared with other calculation methods, and thereby can provide an alternative pessimistic solution for the determination of cyclic creep damage and ductility exhaustion.

For many years, Woodford has promoted a concept of characterising the creep strength properties of a range of engineering alloys to long times using the results from short duration stress relaxation tests. His two papers in this Special Issue extend the application of this approach to the determination of so-called intrinsic ductility for 1CrMoV and T91 steels. Woodford’s intrinsic ductility directly relates to Hart’s strain rate sensitivity parameter [5], and thereby opens the debate as to whether ductility reflects an end of life or deformation condition.

Stuart Holdsworth
Empa: Swiss Federal Laboratories for Materials Science & Technology
[email protected]