Harper–Dorn creep: Controversy, requirements, and origin

This paper reviews the current status of research on Harper–Dorn creep, Newtonian creep behavior that was first reported by Harper and Dorn in their studies on the creep of Al at very low stresses and temperature near the melting point, and that, unlike Nabarro–Herring creep, is apparently independent of grain size, d. An examination of the results reported for Al under the conditions of small strains (∼0.01) and those of recent experiments performed at strains (>0.1) reveals that Harper–Dorn creep does not always occur in large-grained Al at very low stresses; that Harper–Dorn creep is Newtonian in nature only at very small strains (<0.01), and that its occurrence in large-grained Al (d > 400 μm) requires two conditions: the dislocation density in the annealed samples is low (103–104 cm−2) and the purity level of the material is high (99.999%). It is shown that the aforementioned requirements for Harper–Dorn creep can provide a possible explanation for the conflicting evidence regarding the occurrence of Harper–Dorn creep in Al when creep tested at very low stresses. In addition, preliminary results on Harper–Dorn creep in Pb that were obtained in long-term tests lend support to the second requirement related to the purity level of the material.