Open Access
Subscription or Fee Access
Transverse Failure Behavior of Continuous Fiber Reinforced Composites
Abstract
Composite material reveals exceptional high stiffness, strength and stiffness-toweight ratio values which properties can advantageously be used in many demanding engineering applications. However, due to the processing variations and the resulting microstructures (including defects) the macroscopic material properties reveal a high data scatter. A stochastic methodology was described by a research group also with our contribution to predict the properties of UD composites. The present paper is a simplification and includes further development of this methodology for uniaxial thermoset and thermoplastic fiber reinforced composites. Stochastic microcells for UD laminates were generated using a software tool developed in above project. The matrix material properties correspond to a specific epoxy resin typically used for high performance composites and to a polycarbonate recently used in thermoplastic UD tapes. The basic material data were determined in terms of elastic modulus, tensile strength, failure strain and fracture toughness values. The microcell model consists of the typical failure modes of a 2D cell, matrix failure (elastic-plastic) of the matrices and cohesive strength for the fiber matrix interface. To simplify the continuous distribution functions, discrete values over a wide range were selected for all of the parameters used in the models and simulations with these discrete values were carried out. The cells were exposed to tensile and compressive monotonic loading transverse to the fiber direction. The macroscopic transverse strength (Yt, Yc and S) values over a wide material parameter range were predicted for conventional composite failure models. The predicted failure values were compared with corresponding experimental values which were measured using UD laminates in transverse to fiber direction. The sufficient amount of the specimen tests enabled a statistical analysis of the experimental data.
DOI
10.12783/asc35/34855
10.12783/asc35/34855