Date of Award


Publication Type

Doctoral Thesis

Degree Name



Mechanical, Automotive, and Materials Engineering

First Advisor

Northwood, Derek O.,


Engineering, Materials Science.




Experimental creep measurements on an AISI type 310 stainless steel and high purity polycrystalline magnesium have been conducted within the test temperature and applied stress ranges of 650$\sp\circ$C to 800$\sp\circ$C and 110 MPa to 320 MPa for the 310 stainless steel, and 150$\sp\circ$C to 250$\sp\circ$C and 20 MPa to 50 MPa for pure magnesium. The results show that: (i) the creep strain rate ratio, F, between the initial creep rate upon loading and steady-state creep rate decreases from a high value to about unity when the test temperature and/or applied stress are increased; (ii) the apparent activation energy for creep, Q$\rm\sb c,$ upon loading is lower and is closer to the activation energy for pipe diffusion, Q$\rm\sb p,$ than the value at steady-state which is of the order of the activation energy for volume diffusion, Q$\rm\sb v;$ and (iii) the initial creep rate is less strongly dependent upon the applied stress than the steady-state creep rate. The experimental results are discussed in terms of dislocation network models for creep and the internal and effective stresses. The analysis indicates that the high value for F at low temperatures and/or low stresses is in agreement with the predictions of exhaustion theories where multiplication is not taking place and easily-moved dislocations are eventually exhausted. The low value measured for the apparent activation energy for creep upon loading (creep strain ${<1}0\sp{-4})$ can be attributed to the contribution of pipe diffusion due to the large effective stress present at the onset of creep. The stress dependence of the initial and steady-state creep rates is a consequence of a change in both internal and effective stress during creep deformation, and the stress dependence of the initial strain rate offers an indirect measurement for investigating the stress dependence of the dislocation velocity in creep tests. Examination of the experimental results lends support to the theoretical models based on dislocation link length distribution (dislocation network) models for recovery creep. This analysis gives the stress dependence of the steady-state creep rate, $\dot\varepsilon\rm\sb s,$ the dislocation annihilation rate, $\rm\dot\rho\sb a,$ the average effective dislocation velocity, v, the recovery rate, R, and the strain-hardening coefficient, H, respectively, during high temperature recovery creep of crystalline materials. Micromechanical modelling was carried out for three aspects of creep, namely (i) the grain size effect in the creep and superplastic deformation of polycrystalline materials; (ii) simulation of the experimental creep results for a 310 stainless steel using the Ostrom-Lagneborg creep model; and (iii) a dislocation link length statistics model for the plastic deformation of crystalline materials. It is shown that: (i) during superplastic deformation, a solid polycrystalline material can be visualized as a two "phase" mixture, one flowing according to Newtonian viscous flow and the other deforming by power law creep; (ii) the simulation results for the 310 stainless steel obtained using the Ostrom-Lagneborg creep model are in good agreement with the experimental results and other independent studies on dislocation network models for creep; and (iii) the dislocation link length statistics model produces reasonable values for the strain-hardening coefficient and the recovery rate during a constant stress creep test, and for a strain rate change during a constant strain rate tensile test. (Abstract shortened by UMI.)Dept. of Mechanical, Automotive, and Materials Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1994 .S545. Source: Dissertation Abstracts International, Volume: 56-01, Section: B, page: 0461. Adviser: Derek O. Northwood. Thesis (Ph.D.)--University of Windsor (Canada), 1994.