Nonwoven metallic filter media are widely used in the manufacture of fine chemicals requiring extraordinary purity. Because their separation mechanism is complicated, their design and selection still rely on empirical knowledge. To optimize their design, the estimation method for the separation particle size is updated.

Nonwoven filter media having a depth filtration mechanism can be regarded as multiple layers with a unit thickness that is thin enough to perform surface filtration. Since each layer represents an individual filter, the number of layers of the unit thickness media, n, is equivalent to the number of filtration trials. Thus, the partial separation efficiency, Δη, at a particle diameter, D_p, is as follows.

where, p is the probability of particle passage, f(x) is the probability density function for the pore size distribution. This distribution can be described as the size variation of the spaces bounded by randomly oriented line segments.

As the number of [...] increases, the separation size decreases and the pressure drop increases.

Consequently, by using this estimation method based on the probability model, optimized nonwoven filter media with well-balanced performance between the separation size and pressure drop can be designed on a rational basis