Fatigue failure is a concern when high‐strength, high‐toughness silicon nitride ceramics are used in mechanical components and the growth of natural flaws will determine the usable upper bound strength. In this study a fracture resistance curve (R‐curve) model is incorporated into an established method for deducing natural flaw growth rates from a combination of strength and fatigue life data for smooth specimens. Experimental data for a commercial silicon nitride, SL200, were examined. When compared with results deduced using a constant fracture toughness model, the new method gives more physically realistic growth rate results. Specifically, by incorporating the R‐curve the deduced fatigue threshold is equal to the reported intrinsic toughness for crack propagation of 2.2 MPa√m, whereas the constant fracture toughness model gives a physically unrealistic threshold value. Furthermore, much better agreement is achieved with the growth rates measured using macroscopic compact‐tension specimens. Overall, it is concluded that the R‐curve effect should not be ignored when deducing the fatigue crack growth rates of natural flaws in high‐toughness silicon nitride ceramics.

Silicon nitride exhibits fatigue based on cyclic crack propagation which is critical for components under repeated loading. Lifetime predictions for such components are usually based on power law formulations and are most sensitive to the crack growth exponent n. Various statistical procedures exist to determine the parameters from cyclic tests of un‐cracked smooth samples. In this work, an analysis is presented for silicon nitride (SL200) lifetime data with a focus on the relation between load ratio and the exponent n. It is found that n increases with the load ratio which has also been observed for macroscopic cracks in silicon nitride earlier. A high degree of uncertainty is associated with the crack growth exponent due to the sparse lifetime database. A pooling strategy is presented which increases the sample size of the underlying lifetime distribution and successfully decreases the scatter in the crack propagation curves and the uncertainty in the crack propagation exponent and, thus, in the lifetime prediction.

Fatigue behavior under cyclic torsion loading shows a complicated dependency between the number of cycles to failure N f and the height of the applied load σ appl . A recently published result is chosen for which three clearly different regions are observed in which the normally straight log(N f )‐log(σ) plot shows different slopes. Whereas the high‐stress region near the torsion strength can be interpreted as normal mode‐I failure of randomly oriented surface cracks, the steep part may be understood as the consequence of friction degradation at grinding cracks which are under pure mode‐II and mode‐III loading. At lowest loads, quasi‐static subcritical crack growth support has to be expected with its commonly higher load exponents.

The work investigates the cyclic fatigue behavior of alumina. It reveals that for cyclic torsion tests the maximum stress criterion is not valid. Cyclic torsion loading leads to a reduced lifetime compared to cyclic tensile-compression. Also an increased cyclic fatigue effect compared to four point bending tests was revealed.

Abstract Tools for hot wire rolling are exposed to high thermal, mechanical and tribological loadings during service. Ceramic materials such as silicon nitride possess good mechanical and tribological properties up to temperatures where most of today' roller materials do not exhibit acceptable reliability. In the present paper basic mechanical properties under thermal and mechanical loading as well as the damage behaviour of Si3N4 have been investigated in high cycle thermal fatigue and dynamic as well as cyclic four point bending tests. In fretting fatigue tests the behaviour of Si3N4 under near-service conditions, as they appear during application in the hot wire rolling process, has been analysed. It is shown that in the temperature range investigated pure thermally induced loading leads to the same damage mechanism as pure mechanical loading and therefore thermally induced stresses can be directly added to any mechanical loading. The isothermal four point bending tests confirm the existence of a cyclic fatigue effect. With the fretting fatigue experiments the limits of high-performance ceramics such as Si3N4 are shown. Under the complex loading, surface flaws are produced leading to a severe lifetime reduction compared to plain mechanical and thermal loading.