The effect of solvent damage and fibre waviness on the strength of cross-plied carbon fibre reinforced poly(aryl sulphone)

The effect of damage on carbon-fibre reinforced poly(aryl suiphone) by a solvent of the type used by BAe Military Aircraft Division as a paint stripper has been investigated together with the effect of fibre waviness upon the mechanical strength of carbon-fibre reinforced poly(aryl suiphone). Mathematical models for solvent effects were derived and comparison was made with the practical results obtained both from tension loading and three-point loading in bending. Good agreement between the theoretical and test results has been observed. It was found that solvent damage is minimal on the tensile behaviour of (90,0)s laminates but caused a reduction in the strain at failure in (0,90)s laminates. Theoretical models were developed for fibre waviness, where the waviness is in the plane of the ply. The theoretical results are compared to the results of tensile testing. It is reported that, contrary to the work of others, fibre waviness can have a marked affect on laminate behaviour in tension and, for a wave ratio of 0.1, can cause the load at failure in cross-plied laminates to be reduced to only 24% of the the load at failure in an equivalent laminate free of fibre waviness. Specimens were put through heating and cooling cycles in order to reduce residual stresses. Estimates of the value of these stresses were based upon the curvature of asymmetric laminates. The results compared favourably with published values for similar materials. Various 'protective' coats were investigated to reduce the amount of solvent damage. One protective coat was found to give reduced damage with increasing coats. Test results were compared to those obtained from (i) The Tsai-Hill failure criterion, (ii) The Han-Smith maximum shear stress criterion. The test results were found to be in good agreement with both criteria when failure did not occur at positions containing fibre waviness. When failure occurs at fibre waviness it is demonstrated that if the exact point of failure within fibre waviness is known then corrections can be made to the test results. When the corrections were made there was good agreement between test results and both criteria