Influence of mechanical and probabilistic models on the reliability estimates of fibre-reinforced cross-ply laminates

• Reliability estimates of FRP laminates are not robust.
• Three factors may influence greatly the estimates.
• Definition of ‘failure due to matrix cracking’.
• Laminate stiffness degradation in the mechanical model.
• Choice of probability distribution function for the ply properties.

Numerous methodologies for estimating the reliability of fibre-reinforced plastics have been published in the past few decades. Several of these methodologies use different mechanical and probabilistic models, each one based on a number of assumptions and approximations. The objective of this study is to assess common assumptions and approximations made on mechanical and probabilistic models used in reliability analyses of fibre-reinforced plastic laminates. The assessment consists of two parts: a theoretical overview of the models and their justification, and an investigation of the quantitative influence of the models on the reliability estimates of a group of fibre-reinforced cross-ply laminates. The reliability estimates are calculated through Monte Carlo simulations using different mechanical and probabilistic models. This study concludes that both mechanical and probabilistic models can significantly influence the reliability estimations. For the mechanical models, the factor with the greatest influence is the definition and modelling of matrix cracking. While, for the probabilistic models, the choice of probability distribution for modelling ply property variability has the greatest influence.