Modeling II

Fibrous filters are generally characterized by their collection efficiency and pressure drop. In this work we report on our CFD simulations aimed at predicting the permeability and nanoparticle collection efficiency of filter media made up of short fibers. The simulations are performed in a 3-D space for fibrous structures constructed according to given fiber orientation distribution functions. It is shown that shortening the fiber length causes significant increase in the Solid Volume Fraction (SVF) of the media and therefore, affects its permeability and collection efficiency.

There are numerous correlations for predicting the pressure drop of filters with unimodal fiber diameter.  There is, however, no available expression for such filters made of bimodal fiber distribution. Air flow through a series of 2-D models of fibrous filters having bimodal fiber diameter distribution was numerically studied. An expression for an equivalent unimodal diameter was found which can be used with the existing unimodal analytical or empirical expressions to obtain the pressure drop of filters having bimodal fiber diameter distribution.

In this work, 3-D structures resembling nanofiber filter media are simulated and challenged with nanoparticle aerosols at reduced operating pressures. For the range of fiber diameters considered in this paper, the free molecular flow regime is dominant. Nanoparticle capture efficiency of nanofiber webs, due to Brownian diffusion and interception, is calculated for particle diameters ranging from 50 nm to 500 nm. Our simulations show that by decreasing the fiber diameter, the minimum collection efficiency of filtration media having identical pressure drops increases. This effect is accompanied by a decrease in the particle diameter associated with these minimum efficiencies – the most penetrating particle diameter.

Nonwoven may be digitized by several methods. Two are the input as images of a real media and the generation from average properties such as porosity, fiber diameters and fiber orientation. Real media may be understood better, and generated media may provide insights into possible improvements. In either case, the agreement between reality and the computer representation of the nonwoven must be established. Here we discuss the ideas and present some examples of the computer simulation of mercury intrusion porosimetry by the pore morphology method. This method allows establishing relations between production parameters like grammage and fiber diameters and porosimetry measurements that would require a lot more work to establish in real experiments.

Three most important functions of non-woven diapers and hygiene absorbent products are to absorb, immobilize and contain body fluids. But the absorption capacity is not the only measure of the product performance. Today’s customers expect a dry feel, leak-proof, small sized aesthetic and comfortable product. Satisfying all these requirements is a really challenging problem that requires synthesis of the latest knowledge in fields of science and engineering. Minimizing product cost while maintaining or improving the overall performance of the product is the most important task during the product development and design cycle. Typical non-woven absorbent products contain several layers of absorptive materials of different absorbency and mechanical strength. Fluids quickly spread in the matrix system by the action of capillary forces and then are gradually immobilized on the surface of the highly absorbent polymers. To account for the complex physics with different size and time scales, we have developed a CFD wicking model that has been continuously enhanced for several years. So far the model has not been able to account for an external deformation of the non-woven porous media and its effect on the matrix properties and fluid distribution. Typical example includes a situation when a baby sits on a diaper and leakage can occur. Therefore, a fluid-structure interaction model that can address this issue has been developed. New features of the model will be demonstrated on several realistic examples.