Sir,
The claim has been made in the past, and more recently (for ice) by Reference Mellor and TestaMellor and Testa (1969[a]), that the stress dependence of the secondary creep rate (that is, the steady-state creep rate) changes at low stresses. We have pointed out before (Reference WeertmanWeertman, 1967) that experimental creep rates cited in favor of such a claim usually can be dismissed because they clearly do not correspond to true secondary creep. Our argument is the following: In order to be certain that, at a given stress, it is true secondary creep which is being measured rather than transient creep, it is necessary to obtain the creep rate over a total creep strain of at least the order of 0.1 (10%). Thus the smallest steady-state creep rate that can be measured reliably in a year-long laboratory test is about 10−8/s. In Reference Mellor and TestaMellor and Testa’s (1969[a]) work the tests used to show a different stress dependence involved creep rates in the range of 10−10/s to 10−11/s. These creep rates were described as secondary creep rates. (The 0° C tests (Reference Mellor and TestaMellor and Testa, 1969[b]) led to very much faster creep rates. The authors pointed out, however, that the 0° C tests had the complicating factor of grain growth. These particular tests need not be taken as proof of a change of stress dependence of creep rate.)
It should be emphasized that Mellor and Testa are well aware of the difficulty of obtaining a true secondary creep rate at low stresses. They point out that their conclusion about the stress dependence at low stresses is not conclusive because of this difficulty. The purpose of our letter is to point out that although Mellor and Testa used creep tests that ran for almost one year, the time duration of their tests is, nevertheless, many orders of magnitude too short to establish a true secondary creep rate at low stresses. Therefore, no conclusion at all can be made of the stress dependence of the secondary creep rate at low stresses.
An example can be cited to prove that the stress dependence of the creep rate can be quite different at very small creep strains from what it is at large creep strains. In a microcreep experiment (which repeated Chalmers’ original experiment) on tin crystals Reference HarrisHarris and others (1966) find (as did Chalmers) that the creep rate is proportional to stress. The test temperature was room temperature and the stresses ranged from about 1 to 20 bar. The total creep strain displacements were so small that an optical interferometer technique was required to measure them. Yet in this same stress range creep experiments on tin single crystals which were carried out at about 200° C and which extended to large strains led to creep rates proportional to the stress raised to about the 5th power (Reference Weertman and BreenWeertman and Breen, 1956).
9 May 1969
