(A) study on analysis of flow field and prediction of fiber orientation during flow molding of short fiber reinforced composites

Abstract

Fiber orientations caused by the flow during injection molding of short fiber reinforced polymer composites were investigated by numerical simulation. The simulations were conducted both in thickness plane and in three-dimensional domain. For the analysis in the thickness plane, the Lagrangian scheme was employed for the finite element analysis. Flow fields were solved by using a penalty method with Uzawa``s scheme and orientation fields were solved by using the equation of change for second order orientation tensor. A generalized Newtonian fluid whose rheological behavior is independent or dependent of fiber orientation was assumed. Automatic mesh generation using an elliptic grid generator was developed for quadrilateral elements. Mold filling and orientation analyses were performed for cavities of rectangular cross-section. To determine the orientation state in other cross-sectional geometries, numerical analyses were also performed for two orientation in the thickness plane, the orientation field can be classified into three regions in the flow direction and three layers in the thickness direction. It was observed that critical values, upper limits of the degree of orientation, are present. Effects of initial orientation at the inlet on the orientation field were examined. Coupling effects between flow field and orientation field were shown. For the analysis of three-dimensional domain, generalized Hele-Shaw model was adopted and solved by a FEM/FDM hybrid scheme with control volume approach. Orientation fields were also solved by using the equation of change for second order orientation tensor. Forth order Runge-Kutta method was applied to solve five expanded equations. To determine the orientation field, the analyses were performed for the tensile bar specimen and typical shape having obstacles. An experimental method to determine the interaction coefficient was developed. Using image processing techniques, interaction coefficient was determined succes...