Experimental optimization of process parameters to obtain class A surface finish in resin transfer molding process

Resin transfer molding (RTM) has great potential to become an efficient and economical process for fabricating large and complicated automotive composite structures. Low capital investment, excellent mechanical properties, closed mold processing, low pressure equipment, short cycle times and process versatility in component integration and assembly, make RTM very attractive for high volume automotive applications. However, a high class surface finish, required for exterior car body panels, is difficult to achieve with RTM. There are several material and process parameters that severely affect the surface quality of RTM molded components. One of the major contributing factors is the shrinkage related to the curing of thermoset resins. A volumetric shrinkage of 7-10% is common for polyester resins, which breaks the material's contact with the tool, resulting in poor surface finish. Low profile additives (LPA) are added to the polyester resin for shrinkage compensation; however the effects of LPA content on thermal, rheological and morphological properties of polyester resins, as well as surface finish aspects of resulting composites, are not well understood.In this research, the influence of material and processing parameters on the surface finish is characterized through experimental design techniques. Analytical, experimental and numerical methods are employed to optimize processing conditions for class A finish quality and shorter cycle times. Specifically, the effects of LPA content on cure kinetics, cure shrinkage and morphological changes are investigated through differential scanning calorimetry (DSC), rheological and optical microscopic techniques. Models are developed to predict cure shrinkage, LPA expansion, cure kinetics and viscosity variations of the resin. These models are then incorporated in commercial software to predict resin flow, degree-of-cure evolution and pressure variations during cure. Analytical and numerical simulation results are then validated with the experimental data. Processing windows defined for different process parameters based on analytical and numerical analyses are further used in the test matrices designed through the Taguchi method. Analysis of variance (ANOVA) and multiple regression analyses are carried out on the surface roughness of test samples molded under varying conditions. The relative influence of the most significant parameters is characterized and empirical models are developed to predict surface roughness. The effect of post-cure shrinkage on the surface roughness of the test samples is also investigated. Levels of the most significant process parameters are then set for a high class surface finish, minimum cost and short cycle times.The results show that the LPA content had no significant effect on the cure kinetics; however cure shrinkage decreased non-linearly with increasing LPA content. LPA content at 10% was found to be the minimum amount for shrinkage compensation. LPA content (≥10%) resulted in pressure increase and morphological changes during RTM manufacturing. A cure gradient was observed for low pressure injections which had a significant effect on the resin pressure and roughness. LPA was found to be the most influential parameter affecting surface finish. A minimum of 10% LPA was required for class A surface finish. Higher injection pressures and filler content improved surface quality, whereas styrene content, cure rate and temperature gradient had no effect on the surface roughness in the range tested. A direct relationship was observed between LPA content, final cure shrinkage, resin pressure and surface finish