Evolution of a modern mechanical testing and design standard for high temperature materials

Abstract

The basis for high-temperature testing and analysis was established a century ago using test equipment that was available at the time. This fostered a tradition in research and engineering applications that has not changed fundamentally. The limitations of this approach are described, with emphasis on problems associated with predicting the remaining life of operating components. A new testing method and analysis philosophy is proposed based on measurement of current performance, with end of life unambiguously identified with limiting acceptable values of these current performance metrics. Instead of a single test to measure creep strength and fracture resistance, the new approach uses separate tests. For creep strength, the stress vs. creep rate curve determined at nearly constant state and covering five decades in creep rate in a one-day test is obtained from a high-precision stress relaxation test (SRT). The results for many alloys and materials often show a sigmoidal curve. Conventional long-time creep tests of steels have shown a similar effect. However, rather than recognising this as a fundamental rate-dependent phenomenon, it has been typically explained in terms of microstructural coarsening. For fracture resistance, a constant displacement rate test across a notch is used at the temperature of greatest sensitivity to high-temperature fracture. This latter test has been very successful in detecting embrittlement in superalloys. For more ductile alloys, a non-destructive measure of ductility based on the strain rate sensitivity determined from the SRT data is shown to have great potential. This introduces the concept of intrinsic ductility that can be determined without ambiguity over a wide range of stress and temperature for both new and serviced components.

Keywords:

• High-temperature alloys
• Creep rate
• Stress relaxation
• Life prediction
• Strain rate sensitivity
• Embrittlement
• Ductility
• Design for creep

Acknowledgements

Many people and companies have supported this work. However, I wish particularly to thank David Stiles, then of ARCO Alaska, who provided the initial support when this was just an idea, Philip Dowson of Elliott Corporation who provided extended financial support and eventually transferred the technology to his Company, and Tom Gibbons for his continued interest and encouragement. In addition, colleagues at GE provided stimulation and guidance in the development years. Edward Hart prompted my interest and commitment to the SRT test and to rigorous thinking in general. Robert Goldhoff and Louis Coffin were prime mentors in high temperature deformation and fracture, and Roger Bricknell worked creatively with me on environmental embrittlement. Experimental details for SRT and CDR were developed by Donald Van Steele for metals, ceramics and polymers. The original software for SRT analysis was written by Kyle Amberge for the Mac., and recently upgraded and modernised by Paul Woodford for the PC.