Weakest-link failure prediction for ceramic components is usually based on strength data obtained from 3- and 4-point bend tests. However, in the route from test specimen to actual component several pitfalls are encountered. Firstly, the bend tests themselves do have their problems. Secondly, there is the problem of the interpretation of the uniaxial strength data in terms of a failure criterium for multiaxial stress states, for which at this juncture no universally applicable model is available. In this contribution the difficulties encountered will be summarized. Besides alternatives will be mentioned which are based on microstructural insight and anisotropic deformation models.
|Journal||KGK, Tijdschrift voor Klei, Glas en Keramiek|
|Publication status||Published - 1991|