Semiempirical approach to predict shrinkage and warpage of fiber-reinforced polymers using measured material properties in finite element model

Abstract

The automotive industry has great interest in designing and producing lightweight high-performance components using fiber-reinforced polymers (FRPs), primarily due to their high specific strengths. Injection molding of FRP is one of the preferred processes to meet low-cost, high-volume objectives. It is imperative to account for shrinkage and warpage while designing the tools for injection molding. However, predicting shrinkage and warpage of injection-molded FRP parts remains a challenge. This is because both the structural and thermal properties depend on the condition of the fibers in the resin, that is, variation in the orientation, length, and concentration throughout the part. Additional challenges come from the fact that the material properties of polymers are a function of temperature, which varies as the parts cool. In this study, we are presenting a finite element-based semiempirical approach to address both these challenges and predict warpage due to cooling for a fiber-reinforced resin component in solid phase. The approach is demonstrated to predict warpage of an injection-molded flat plaque made of glass fiber-reinforced polypropylene, cooled from 160°C to room temperature of 23°C. First, the fiber orientation in the plaque is estimated. Next the material properties for the combined material, that is, glass and resin, are measured as a function of temperature. Then the combined material properties and calculated fiber orientations are used to estimate the ‘in-mold’ condition resin properties using reverse engineering. Finally, the warpage of the plaque is predicted using the estimated resin properties and fiber orientations. Warpage predictions using this method compare well with the measured experimental results. Our study demonstrates that valid predictions for shrinkage and warpage of injection-molded fiber-reinforced thermoplastic parts in solid phase can be made if accurate material properties are used.