Permeability of Glass Fabric Reinforced Vinyl Ester Composite by VARIM Process

Abstract

The vacuum-assisted resin infusion molding (VARIM) method has been increasingly used in manufacturing large composite structures, such as MW-scale wind turbine blades, due to its simplicity and cost-effectiveness. Understanding permeability of the composite is essential in controlling resin infusion for efficient manufacturing process development. Deformation of the vacuum bagging in the VARIM results in ever-changing permeability during infusion. The objectives of this research are to: (1) identify unique features of the apparent permeability during VARIM of glass-fabric laminate composite, (2) investigate the effects of process variables and manufacturing variations on the apparent permeability, and (3) determine the effective permeability value for use in VARIM simulations. In the experimental program, a series of glass-fabric panels were placed in a variety of laminate lay-ups and infused with vinyl ester resin and a selected low-viscosity fluid (corn oil). The process was monitored with embedded thermocouples, a video camera, a digital viscometer and an LVDT to monitor flow front motion and dimensional stability. Two methods were used to evaluate the VARIM composite apparent permeability. In the first method, permeability and part quality of the composite panels were determined by vinyl ester resin flow-front tracking and burn-out tests. In the second method, the apparent permeability of corn-oil infused composite panels was determined as a function of mass flow-rate and panel thickness. Resin fluid motion and other manufacturing parameters were found to govern the apparent permeability in the first method. With an improved lay-up design the second method was used and the resulting composite dimensional stability was found to be more significant in pre-infusion compaction tests than that during infusion. Distinct features were obtained in both effective and apparent permeabilities. Infusion times in the experiment are compared favorably with the results obtained from FLUENT 16.0’s volume of fluid method, using inputs from the second experimental method.