In liquid composite molding (LCM) processes, the usual practice is to simulate the mold filling process using a flow simulation software to predict the filling time and filling pattern and subsequently optimize the process (e.g. location of resin inlets and outlets) for a good quality impregnation. A prerequisite for such a simulation is the experimentally determined or computationally estimated permeability of the fiber preform to be molded. In fact perme- ability is a determining parameter in the governing equations (via Darcy’s law) of resin flow inside a porous fibrous media. On the other hand, permeability is influenced by several pa- rameters among which the fiber volume fraction (Vf), reinforcement architecture and its in- tra-ply shearing are the most important ones. Because of these observations, many research- ers have extensively studied different methods of permeability measurement and have char- acterized the permeability of several preforms.
In 1988 Adams et al. [55] proposed a methodology to solve the governing equations of two-dimensional radial flow in an anisotropic porous media based on experimental data of elliptical flow front radii versus time. The original method was further developed by Chan and Hwang [56] and Griffin et al. [57] in 1991 and 1995, respectively. In 1999, Weitzenböck et al. [58, 59] propose a new solving technique for a more general case in which the assump- tion of symmetric permeability tensor is lifted. In this method, instead of recording the major and minor radii of the elliptical flow front, three points on the flow front along three orienta- tions are recorded with time. Hoes et al. [60] proposed a new permeability measurement set up for radial injection experiment which uses two steel plates as top and bottom mold haves and electrical sensors embedded in the top mold half to detect the flow front. Permeability of
several woven glass fabrics are tested and their statistical distributions are acquired. Endruweit et al. used radial flow injection method to experimentally study the effect of shear angle on the principal permeabilities of various glass fiber fabrics [61]. Effects of stochastic variation of fiber angles and spacing between fibers on permeability uncertainty are also sim- ulated in other works of Endruweit et al. using 2D radial flow injection method [62, 63]. Apart from radial injection technique, linear flow method is also used for permeability meas- urement [64]. Pan et al. [65] used one-dimensional permeability measurement technique to examine influence of process parameters on permeability variance of knitted and woven glass fabric preforms. It is reported that edge protection, fabric areal weight and complexity of mold shape, have significant influence on the permeability behavior.
As a result of increasing popularity of natural fiber composites, permeability of natural fiber reinforcements has also been studied in recent years. Umer et al. characterized the per- meability and compaction response of wood fiber mats [66] and flax fiber mats [67]. For the former mat type the permeability is characterized versus fiber volume fraction while for the latter one, it is characterized versus fiber length and diameter as well as fiber volume fraction. It is claimed that the permeability of natural fiber mats and fabrics is dominated by the char- acteristics of open channels. Therefore, swelling of fibers impregnated with glucose syrup, which is a polar liquid, is responsible for lower measured permeability compared to the situ- ation where non-polar motor oil is used. For flax fiber mats the permeability of medium yarn diameter (560 μm) mats is reported to be highest in comparison with both small (350 μm) and large (810 μm) yarn diameter mats. Moreover, the permeability of 50 mm chopped yarn length was found to be higher as compared to its 15 mm counterpart. Permeability of medium
yarn diameter and 50 mm yarn length mats is reported about 0.5× 10−8 m2, for 20 % fiber volume fraction which is equal to that of glass fiber mats at the same fiber volume fraction. Mekic et al. [68] evaluated the in-plane and out-of-plane permeabilities of preforms made of randomly oriented flax fibers ranging from 12 mm to 50 mm in length and 10 μm to 50 μm in diameter. It is concluded that the in-plane and out-of-plane permeabilities fall in the range of those obtained for randomly oriented fiberglass mats. So, traditional liquid molding tech- niques can be used for the fabrication of natural fiber composites. In 2010, bidirectional wo- ven jute fabrics are characterized in terms of saturated and unsaturated permeability by Fran- cucci et al. [69]. It is claimed that fluid absorption and fiber swelling are two mechanisms respectively responsible for lower unsaturated and saturated permeability values compared to the permeability of glass fiber fabrics. In unsaturated permeability measurement, hydro- philic nature of plant fibers removes fluid from the main stream and thus decreases flow velocity. On the other hand, saturation of natural fibers causes swelling, thus reducing the porosity and increasing flow resistance. The permeability-porosity relationship is also inves- tigated for bidirectional jute fabrics [69, 70] and random sisal mats [70] based on the empir- ical Carman–Kozeny model. Recently Lebrun et al. [71] studied the permeability behavior of unidirectional hemp/paper and flax/paper reinforcements, as well as their individual con- stituents of UD flax layers and absorption papers, using resin infusion molding method and epoxy resin as test liquid. Tensile properties of the resulted composites are as well investi- gated. These reinforcements have shown lower permeability levels, particularly the hybrid reinforcements, compared to the UD glass reinforcements. It was concluded that the overall
permeability must be improved. More recently, the effect of test fluid on saturated permea- bility of twill weave flax fabrics was examined by Nguyen et al. [72]. Saturated permeability is reported dependent on the test fluid because of different swelling behaviors of flax fibers to each of the test fluid. The modified Kozeny– Carman model with two model constants is also claimed to well simulate the experimental saturated permeability data. In another work [73], a model for resin flow inside flax fiber preforms which takes into account fiber’s mass sink effect and swelling during the mold filling process is proposed.
Despite considerable amount of research on permeability evaluation, an international standard for permeability measurement is still lacking. In response to this, two international permeability benchmarks were conducted [74, 75] and a guideline was published for meas- uring the unsaturated permeability of preforms using the linear injection method [76].