JOURNAL OF MATERIALS SCIENCE 17 (1982) 1066-1078 Effects of anelastic deformation on high-temperature stress relaxation of polycrystalline MgO and AI203 YASURO IKUMA, RONALDS. GORDON Department of Materials Science and Engineering, University of Utah, Salt Lake City, Utah 84112, USA High-temperature (1160 to 1450~ C) deformation of dense polycrystalline (10 to 90#m) AI2 03 and MgO doped with Fe (up to 2.65 cation %) was studied by stress relaxation, dead-load creep and creep recovery. In some cases, all three deformation tests were conducted on a single specimen. A comparison of strain rate-stress data calculated from both stress relaxation and dead-load creep experiments revealed discrepancies in both the magnitude of the strain rates and the dependence between the strain rate and stress. These differences were attributed to the existence of anelastic deformation effects. The correction of stress relaxation data in the low stress regime for linear anelasticity led to strain rate-stress data in reasonably close agreement with results obtained from dead-load creep tests conducted in the viscous creep regime. Creep recovery experiments indicated that anelastic deformation in these ceramic materials was relatively insensitive to changes in temperatureand grain size over the range of variables studied. 1. Introduction An analysis for the calculation of plastic strain rates and stresses from stress relaxation measure- ments taken under conditions of four-point bending has been proposed by Shetty and Gordon [1]. In this analysis, the plastic deflection rate, ~, of the specimen is calculated from = _c _@ dr' where C is the total elastic compliance of the system, including the specimen and the machine components such as the load cell and dP/dt is the rate of load relaxation. Shetty and Gordon assumed that stress relaxation is due entirely to the conversion of elastic strain to plastic strain. In the present paper, the validity of this assumption will be analysed by comparing strain rates calcu- lated from stress relaxation and dead-load creep tests. It will be shown that anelastic effects have a significant effect on the interpretation of stress relaxation data. The effect of anelasticity on stress relaxation in metallic systems has been 1066 recognized by previous investigators [2-4] ; however, the direct comparison of strain rates calculated from stress relaxation and dead-load creep tests is rare in deformation tests at high temperatures on ceramic materials [5 ]. 2. Experimental procedure Specimens tested in this study included dense polycrystalline MgO doped with 2.65 cation% Fe and polycrystalline A12Oa doped with 1 to 2 cation%Fe, with grain sizes, Gs, between 10 and 90/am. These materials and dopants were selected because extensive dead-load creep data were available [6, 7] in these systems for compari- son with stress relaxation data. The methods used in the preparation of these specimens have been described elsewhere [6, 7]. Stress relaxation tests were performed in four-point bending using an apparatus described by Shetty and Gordon [1 ]. Creep recovery experiments were performed using a four-point dead-load creep apparatus, with a remote unloading system which is described elsewhere [8]. 0022-2461/82/041066-13504.06/0 9 1982 Chapman and Hall Ltd.