Comparison of Tensile Creep Measurements on Buttonhead and Pin-Loaded Specimens of Some Silicon Nitride Ceramics
Tensile-creep measurements by two experimental techniques on two kinds of silicon nitride ceramics were compared statistically. Some of the measurements were obtained from published investigations conducted elsewhere. One technique uses buttonhead specimens, and the other technique uses pin-loaded specimens. The techniques differ in the manner of loading the specimen, of heating the specimen, and of measuring creep strain. The same model of an exponential function of applied stress and temperature was selected for representeing secondary-creep lifetimes to provide a uniform basis for comparison. Generally, the pin-loaded technique yielded somewhat smaller variability of measurements than did the buttonhead technique. Statistical F-tests indicated that predicted creep rates and lifetimes of the buttonhead specimens are distinguishable at a high level of confidence from those of the pin-loaded specimens for each material tested. Although pin-loaded specimens are considerably smaller than buttonhead specimens, no consistent trend of creep behavior can be established with respect to specimen size. Possible systematic errors in loading alignment and strain measurement of the two kinds of specimens remain unresolved, but the buttonhead specimen definitely suffers from a greater uncertainty in temperature than does the pin-loaded specimen.