Risk reduction in composites processing using prototype data, process simulation, and Bayesian statistics

Abstract

There is often significant risk and uncertainty associated with the development of manufacturing processes for large integrated composites structures. A good understanding of the outcome of the process is required so that process tooling and other process parameters can be designed appropriately and costly redesign and rework avoided. This paper presents an approach to risk reduction in composites processing using prior knowledge, prototype data, and model results. A Bayesian methodology for combining this information into a probability density function of the outcome of the process that explicitly accounts for reliability of the data sources is developed. The use and effectiveness of the approach is demonstrated with a case study.