Composites manufacturing involves many sources of uncertainty associated with material properties variation and boundary conditions variability. The aim of this paper is to summarise the state of the art on experimental and stochastic simulation methodologies and results focusing on statistical characterisation and the influence of inputs variability on the main steps of composites manufacturing, including process-induced defects, as well as to highlight the interdependencies between the process parameters. Uncertainty introduced by experimental methods and modelling practices is also included.


The findings presented in this study highlight the importance of variability in composites manufacturing and thus the need for future development and incorporation of stochastic simulation schemes into the existing commercial simulation tools. This implies that stochastic simulation should play a major role in process design; adopting stochastic simulation tools will have tremendous benefits in terms of costs.

Benchmark guidelines should be developed regarding characterisation techniques and modelling practices in all manufacturing steps, to minimise property measurement and model uncertainties. 


T.S. Mesogitis, A.A. Skordos, School of Applied Sciences, Manufacturing and Materials Department, Composites Centre, Cranfield University, UK, and A.C. Long, Division of Materials, Mechanics & Structures, University of Nottingham, UK.

Further information

This paper was published in Composites Part A: Applied Science and Manufacturing, Volume 57, February 2014, pages 67-75 and is available on ScienceDirect.com.