Ideal Forming Analysis for Random Fiber Preforms

Abstract

In resin transfer molding, the manufacture of the fiber preform controls many aspects of part quality. These include defects such as wrinkling and tearing, as well as spatial variations in fiber volume fraction and permeability. We develop a mathematical model and numerical method for analyzing preforming of random fiber mats. The model uses an ideal forming theory, which maps a fiber sheet to the mold surface by minimizing the integral of a formability function over the mold surface. The scalar formability function depends on the local deformation, and exhibits large values under conditions that promote either tearing or wrinkling of the mat. The model is implemented as a finite element simulation for arbitrarily shaped three-dimensional preforms. Results include the shape of the initial fiber sheet, and values of the formability function and the principal stretch ratios over the mold surface. This information is used to predict the presence of defects in the preform. Example calculations are shown for an axisymmetric hat shape and for a box with a flange. The calculation requires a modest amount of input data and, rather than predict the exact result of the forming operation, it shows the best result that is possible. Thus, it is a useful tool in the early stages of part and mold design.