W. Brostow
Departments of Materials Science and Physics, University of North Texas, Denton, TX 76203-5308, USA
Josef Kubát
Department of Poymeric Materials, Chalmers University of Technology, S-412 96 Gothenburg, Sweden
Michael J. Kubát
Department of Polymer Technology, Royal Institute of Technology, S-100 44 Stockholm, Sweden
ABSTRACT
Experimental evidence on stress relaxation is analyzed first for a wide variety of classes of materials: metals and their alloys, synthetic and natural polymers, glasses and frozen non-polymeric organic liquids. Common features of curves ð(t) of relaxation of stress ð as a function of time t are discussed, and the importance of the internal stress ði(for time = infinity) noted. Theoretical approaches are then reviewed, with particular attention to the cooperative model and its modifications; that model corresponds well to the experimental results. Some simulation results obtained by the method of molecular dynamics are reported for ideal metal lattices, metal lattices with defects, and for polymeric systems. In agreement with both experiments and the cooperative theory, the simulated ð(log10 t) curves exhibit three regions: initial, nearly horizontal, starting at ð0; central descending approximately linearly; and final, corresponding to ði. In agreement with the theory, the slope of the simulated central part is proportional to the initial effective stress ð0* = ð0 - ði. The time range taken by the central part is strongly dependent on the defect concentration: the lower the defect concentration, the shorter the range. Imposition in the beginning of a high strain destroys largely the resistance of a material to deformation, resulting in low values of the internal stress ði. On the joint basis of experimental, theoretical, and experimental results, we explain the mechanism of stress relaxation in terms of deformations occurring in the immediate environment of the defects. Simulations show several common features in the behavior of metals and polymers. Apart from the defect concentration, the amount of free volume vf is also important.
*Mech. Comp. Mater. 1995, 31, 591.