Influences of friction, geometric nonlinearities, and fixture compliance on experimentally observed toughnesses from three and four-point bend end-notched flexure tests

Three and four-point bend end-notched flexure tests, comprising a number of different test geometries, are performed on two different graphite/epoxy composites, and the toughnesses are obtained by a compliance calibration method of data reduction. The coefficient of friction along the crack plane and the flexural modulus of each material are then determined experimentally and used with nonlinear finite-element analyses to simulate these same test configurations. By using the mean experimentally observed critical load, these simulations are used to obtain the materials' toughnesses by three different methods. The first uses a previously developed 'direct energy balance approach,' which is assumed to produce the 'true' toughness. The second is by a simulated compliance calibration procedure, which is used to obtain the perceived toughness for an infinitely stiff fixture. In the third approach, experimentally determined fixture compliances, as a function of the test geometry, specimen, and crack length, are used along with the simulated compliance calibration procedure to obtain a perceived toughness that accurately accounts for the effects of friction, geometric nonlinearities, and fixture, compliance. The perceived toughnesses as obtained by these simulations are shown to accurately recreate the perceived values of toughness that are obtained from the physical tests. Moreover, the finite-element simulations indicate that the true toughness values are essentially constant for a given material, and that the three-point bend end-notched flexure test will provide perceived toughnesses that correspond quite closely to the true value, whereas the four-point bend test often will not. For these reasons, the three-point bend configuration was found to be the more preferable of the two tests.