Development and use of an improved filament-winding process model

Composite materials are being used to solve an ever-increasing variety of engineering problems because they provide greatly improved mechanical properties and the flexibility to tailor them. A process that is often used to fabricate cylindrical composite structures and other surfaces of revolution is filament-winding. The objective of this research is to develop a model of the filament-winding process. At a time when concurrent engineering philosophy is being accepted and implemented, design should incorporate as much of the manufacturing process as possible. This is critical in a process like filament-winding where the residual stresses associated with fabrication can be as significant as the stresses encountered during service. Thus the optimum design is dependent upon the physical and chemical phenomena that occur throughout manufacture as well as in the service life. Previous filament-winding process models relied on a linear elastic assumption to describe the material behavior during cure. However, it is known that many of today's polymeric resins exhibit a time dependent response even during the cure cycle. A major contribution of this work is the incorporation of this viscoelastic response into the process model. Although these materials and mechanics issues are a major focus, they are incorporated into a comprehensive multidisciplinary process model that includes heat transfer, chemorheology, and compaction. It predicts, as a function of position and time, the temperature, degree of cure, viscosity, and deformation of a thick composite cylinder of arbitrary lay-up as well as the resulting residual stress state. The model is validated through an experimental program consisting of a materials characterization study of an AT-400/Fiberite 934 graphite epoxy composite and the measurement of processing induced strains. Good correlation is achieved between model predictions and experimental results. The various geometric and material input parameters are also investigated to determine their impact and relative importance on the predicted residual stresses. Recommendations for future work such as cure cycle optimization and extension to other material systems and geometries are made.