Ralph F. Krause

Creep Behavior of Improved High Temperature Silicon Nitride

New data are presented on the tensile creep behavior of silicon nitride sintered with Lu2O3. The data are compared with two earlier sets of data collected on the same material. The older sets gave results that are difficult to explain theoretically: a high value for the stress exponent, n=5.33, and no cavitation. The new set of data also gave no cavitation, but gave a stress exponent, n=1.81, that can be rationalized theoretically in terms of solution-precipitation creep of the silicon nitride grains. An analysis of variance showed that one of the earlier sets of data was statistically consistent with the newer set, whereas the other set of data was not. Combining the two sets of data that agreed statistically yields a consistent picture of creep with a low value of the stress exponent and no cavitation. The stress exponent for the combined set of data is n=1.87±0.48 (95 % confidence limits). The tensile creep mechanism of the silicon nitride containing Lu2O3, solution-precipitation, differs from those of other silicon nitrides, for which tensile creep has been attributed to cavitation. Enhancement of the creep resistance of the silicon nitride sintered with Lu2O3 may be a consequence of the fact that Lu2O3 produces a more deformation resistant amorphous phase at the two grain junctions, than do Y2O3 or Yb2O3. In parallel, reducing the amount of secondary phase below a critical limit, or increasing the viscosity of the two grain boundaries relative to three-grain junctions reduces the ability of the material to cavitate during creep, and forces the creep mechanism to change from cavitation to solution-precipitation.

Influence of Microstructure on Creep Rupture

In this paper, the effect of microstructure on both the creep and creep rupture behavior of two commercial grades of vitreous bonded aluminum oxide was investigated. Deformation and fracture occurred within the ductile, intergranular phase of the material. The creep rate was relatively insensitive to the amount of intergranular phase, but was sensitive to structural details of that phase. The creep rate could be reduced by increasing both the degree of crystallization of the intergranular phase and the viscosity of residual glass within that phase. The time-to-rupture and the strain-at-rupture increased as the amount of intergranular phase within the material increased. In this regard, an increase in the amount of intergranular phase permitted greater accommodation of strain, and hence, blunting of cavities that nucleated during the creep process. The data fit a modified Monkman-Grant curve in which the Monkman-Grant coefficient was sensitive to both stress and the amount of intergranular phase. The Monkman-Grant coefficient was not sensitive, however, to the degree of crystallization of the intergranular phase.