High-temperature creep experiments often reveal a transition at very low stresses to a region where the stress exponent is reduced to a value lying typically in the range of ϳ1 to 2. This region is generally associated with the occurrence of a new creep mechanism, such as grain-boundary sliding, diffusion creep, and/or Harper-Dorn creep. Several recent reports have suggested that diffusion creep and Harper-Dorn creep may not be viable creep mechanisms. This article examines these two processes and demonstrates that there is good evidence supporting the occurrence of both creep mechanisms under at least some experimental conditions.
I. HISTORICAL BACKGROUNDexperiments were undertaken to determine the precise functional relationship between the steady-state creep rate and CREEP deformation refers to the unrecoverable plastic external experimental parameters such as stress and temperastrain occurring in a material when it is subjected to a conture. This latter approach had two very significant advanstant applied stress (or a constant applied load) over an tages over the earlier attempts to develop constitutive extended period of time. Creep processes are diffusionrelationships as in the classic Andrade t 1/3 law. First, the controlled, and they become of particular importance in approach, when combined with theory, permitted an assessmaterials experiencing extensive periods of time at elevated ment of the precise atomistic processes occurring during temperatures, where these high temperatures are generally creep deformation, and thus it led to the concept of specific above ϳ0.4 T m , where T m is the absolute melting temperature and well-defined rate-controlling creep mechanisms. Secof the material.ond, the approach provided, for the first time and over at least The small-scale industries of the 19th century tended to a reasonable range of experimental conditions, a predictive operate at relatively low temperatures so that the occurrence capability of the effect of changes in the operating stresses of any creep deformation in mechanical parts was generally and temperatures. The many publications of this era are well neither appreciated nor of significant magnitude to seriously documented in the creep literature, and they culminated impair the industrial operation. However, this situation began in an extended and comprehensive overview of the creep to change in the very early days of the 20th century when behavior of a wide range of metals and metallic alloys. [4] there was a concerted effort to increase the operating temperThis article follows on from this more recent approach, atures, and therefore the overall efficiency, of conventional and it is concerned specifically with the significance of working plants, such as steam boilers. It is not surprising, steady-state creep and the interpretation of rate-controlling therefore, that the first scientific publication dealing exclucreep mechanisms at very low stress levels. As will be sively with creep deformation should appear almost exactly demonstrated, although the ...