Finite element modelling of creep and meutral axis migration in ceramic four point bend tests

Abstract

The modelling of creep behaviour of ceramics in high temperature applications (typically turbine blades or piston crowns) and the role of creep in densification processes is important in the further development of ceramics for extreme operating conditions. Four point flexural bend tests are generally used to investigate the high temperature creep behaviour of ceramics. Flexural creep tests offer several advantages over uniaxial tests, including lower cost of preparation, avoidance of sample misalignment problems and reduced sensitivity of strain to temperature fluctuation. The disadvantage is that the stress fields are statically indeterminate and thus results become difficult to interpret. The general assumption in continuum mechanics is the nonvariance of the neutral axis location, ceramics however exhibit different creep behaviours under tensile and compressive stresses which lead to the migration of the neutral axis. Thus the relaxed outer fibre stationary stresses alter and failure times become significantly longer than expected by an equivalent uniaxial test. Thus the results of flexural creep testing may be of Anegligible quantitative value@[1] in determining rupture life at high temperature. The use of reference or skeletal stresses [2][3] may however overcome many of the current inadequacies of flexure testing. To illustrate this, a nonlinear finite element analysis (FEA) of a four point bend test is used to compare with experimental and analytical results on Synroc C [4] a multiphase ceramic designed to immobilise the radioactive elements in high-level nuclear waste HLW. The reference stresses are then calculated and compared to results derived from the FEA analysis enabling the validity of the reference stress to be examined as it applies to Synroc.