Hiroshi Hohjo

Correlation between Microstructure and Fatigue Behavior of Silicon Nitride Ceramics.

Microstructures of seven kinds of silicon nitride ceramics were quantified by electron cyclotron resonance plasma etching and sequential image analysis. And the correlations between microstructure and fracture toughness, intrinsic bending strength or cyclic fatigue crack growth exponent were estimated by using multiple regression analysis.The characterized microstructure parameters, composed of the mode and distribution of grain size, the fraction of needle-like grains and the amount of additives, were correlated to the experimental mechanical properties with high correlation coefficients. For improving the fracture toughness, intrinsic bending strength and cyclic fatigue crack growth exponent of silicon nitride ceramics at the same time, it may be effective to increase the needle-like grains, according to the weight factors of microstructure constituents.Fatigue crack growth passages in the materials with highest and lowest fatigue crack growth exponents were analyzed. As a result, in the material with a high exponent, crack arrest and bridging were more often caused by the intergranular fracture of needle-like grains. A larger amount of additives and a broader distribution of grain size make crack arrest and bridging tend to occur more often.

Fatigue Lifetime Prediction Method for Ceramics with Defects.

A new method of predicting the fatigue lifetime of ceramics with various sizes of defects has been developed. In this method, first, specimens each of which has an artificial crack of constant size are used to estimate fatigue crack propagation parameters. In the vicinity of the tip of an artificial crack induced by Knoop indentation, a tensile stress field is generated. Then, the stress intensity factor acting as the driving force of fatigue crack growth was considered the sum of two terms associated with the applied stress and residual stress field. By numerical computation inserting the stress intensity factor into the power law of crack propagation, crack propagation parameters are estimated. Subsequently, using the obtained parameters, the fatigue lifetimes of the smooth specimens with fracture originating from a small defect, are predicted.This method was applied to three types of silicon nitride ceramics. As a result, the cyclic fatigue lifetimes of precracked specimens, which were set free from residual stress field by surface removal, were precisely predicted using the obtained parameters. Thus, the values of these parameters were proved to be adequate. Furthermore, the predicted fatigue lifetimes of the smooth specimens were in good agreement with the experimental results, considering the scatter of static strength.

Influences of Defects upon the Fatigue Strength of Si3N4 Ceramics.

Microstructure of engineering ceramics, especially the distribution of defects and the density, sometimes varies with the lot of manufacture. For this reason, it is important to investigate the influences of the microstructure upon the mechanical properties.In this study, the rotating bending fatigue tests were carried out on two lots of Si3N4 ceramics, and the influences of the defects and density upon the fatigue strength were investigated. Bulk densities of lots A and B were 3.02g/cm3 and 3.16g/cm3, respectively. Average static bending strength of lot B was about 20 percent higher than that of lot A due to the differences in defects distribution and fracture toughness. However, the two lots scarcely differed in rotating bending fatigue strength. This is considered to be because the fatigue strength largely depends on the size and kind of defects.Fatigue strength was predicted on the basis of a fatigue datum in the case when the fracture origin was the largest pore defect. Crack propagation rate was assumed to be in proportion to the power of the stress intensity factor. Consequently, when the fracture origin was a pore defect with inclusion or a micropore cluster, the fatigue strength was higher than the predicted value, in most cases. Therefore, the fatigue life prediction based on pore defects leads to reliable estimation.