An Experimental Study of Saturated and Unsaturated Permeabilities in Resin Transfer Molding Based on Unidirectional Flow Measurements

Abstract

Resin transfer molding (RTM) is a manufacturing process of composite parts that consists of injecting a fluid resin through a fibrous reinforcement. Permeability of the fiber bed is a key parameter that governs the flow, together with the viscosity of the resin. In this paper, the saturated and unsaturated permeabilities of a glass-woven fabric are studied for different injection pressures and porosities of the fibrous reinforcement. Unidirectional injection experiments are carried out to evaluate the unsaturated permeability of the reinforcement during the injection and the saturated permeability that governs the permanent flow established after filling of the cavity. The unsaturated permeability is evaluated by a linear regression in function of the position of the resin front during transient flow experiments performed at constant injection pressure. For each position of the resin front in the test mold, a transient permeability can be calculated by application of Darcy’s law. This transient permeability depends on the injection pressure, position of the resin front in the mold, and porosity of the reinforcement. After a certain flow length called the convergence length, the transient permeability converges to a value that is independent of the flow rate and of the injection pressure. This limit value is called here the unsaturated permeability. The convergence length is related to a maximum flow rate or average particle velocity below which Darcy’s law holds. The saturated permeability is measured after all air bubbles have been evacuated from the cavity. As a result of this experimental investigation, three parameters can be identified that are characteristic of the pore structure of the fibrous reinforcement: (1) a maximum average resin velocity can be identified below which the transient permeability remains constant; (2) the saturated and unsaturated permeabilities differ by a value that remains constant within the injection pressure range used in this study and that is independent of the fiber volume content; (3) the new notion of convergence length proposed here, namely the flow length necessary to ensure convergence of the transient permeability, depends on the two-scale pore structure of the fibrous reinforcement. This study corroborates the findings of several other investigators, who have shown that the saturated permeability is always higher than the unsaturated permeability. The difference between these two permeability values is also related to the pore structure of the fiber bed. It is hoped that this work will contribute to define a much-needed standard on permeability measurements.