Particle collection of medium performance air filters consisting of binary fibers under dust loaded conditions

doi:10.1016/S1383-5866(99)00086-6

Separation and Purification Technology

Volume 19, Issues 1–2, 1 June 2000, Pages 145-152

https://doi.org/10.1016/S1383-5866(99)00086-6 Get rights and content

Abstract

Evolutions of collection efficiency and pressure drop of an air filter with dust load vary with the filter structure, such as mean fiber size, fiber size distribution, packing density, as well as the captured particle size and filtration velocity. In the present work, time change in collection performance of medium performance air filters composing of binary fibers is studied in order to clarify the effect of fiber size distribution on the collection performance. As a result, it was found that, during the period of depth filtration, the time change in collection performance of binary filters can be predicted by assigning unique fractional collection efficiency raising factors, overall collection efficiency raising factors and pressure drop raising factors of individual fibers, suggesting that there are no interactions between fibers with different sizes. This study confirms that mixing fibers with various diameters is one of the effective means to adjust the time change in collection efficiency as well as initial collection performance.

Introduction

Medium performance air filters are usually used in air conditioners. They are also installed upstream of High Efficiency Particulate Air (HEPA) filter in air-intake ducts of clean room to reduce particle load to HEPA filter.

There are many experimental studies dealing with collection performance of air filters under dust loaded conditions. Kimura and Iinoya [2] and Yoshioka et al. [6] obtained empirical correlation among the filter performance, the deposited mass and the total dust load to air filters. However, the relationships between filter structure, (i.e. mean fiber size, fiber size distribution, packing density) and collection performance of filter with dust load have not been well established yet. The present work focused on obtaining the relationships between the fiber size and the collection performance with dust load and optimizing the fiber size distribution to lengthen the life of air filters. As a fundamental study for the influence of fiber size distribution on the collection performance, model filters consisting of two distinct fiber sizes were studied in the present work.

Section snippets

Initial performance of binary filter

Suppose we have filters consisting of uniformly mixed fibers with two distinct diameters, d_f1, d_f2 (d_f1<d_f2). Assuming that there are no interactions between fibers with different sizes, the initial fractional collection efficiency E₀(d_p) without dust load is given by Eq. (1). $E_{0} d_{p} =1− exp − 4αL π 1−α z_{1} η_{1_{0}} d_{p} d_{f1} + z_{2} η_{2_{0}} d_{p} d_{f2}$ where, η_1₀(d_p), η_2₀(d_p) are the initial single fiber fractional collection efficiencies, L is the thickness of filter, α is the packing density, and, z₁, z₂ are the mass fractions of

Experimental

Fig. 1 shows experimental set up. Two fibers were uniformly mixed at various mass fractions to make a fibrous bed with thickness of 1 cm and diameter of 5 cm. Although the mass fractions of individual fibers were varied, the packing density of fibrous bed was adjusted at α=0.03. Polydisperse Kanto-loam particles (JIS Z 8901 No. 11) were generated by a fluidized bed dust generator (KANOMAX Model 3211). The generated aerosol was then passed through an impactor with a 50% cut-off size of 3 μm. The

Initial fractional collection efficiency E₀(d_p)

Fig. 3 compares the experimental initial fractional collection efficiencies with those predicted by Eq. (1) for binary filters at various mass fractions of coarse fibers, z₂ In the prediction of initial single fiber fractional collection efficiencies in Eq. (1), they were given as a sum of inertial fractional efficiency η₁(d_p) by Takahashi [5](Eq. (13)) and diffusion–interception efficiency η_DR(d_p) by Stechkina et al. [4](Eq. (14)). $η_{1} d_{p} = exp −0.462α^{−0.5} Stk$ $η_{DR} d_{p} =2.7 Pe^{−2/3} + 1 2κ 2 1+R ln 1+R − 1+R + 1 1+R$

Conclusions

The present work studied the influence of fiber size distribution on the collection performance by using binary filters. The following conclusions are obtained.

1.
In early stage of filtration, constituent fibers of binary filter independently contribute to the collection performance and therefore we may predict the initial collection performance without considering interaction between two distinct fibers.
2.
The time change in fractional collection performance of binary filters during depth filtration

Acknowledgements

This work was supported financially by a grant from Hosokawa Powder Technology Foundation.

References (6)

C.N. Davis
Proc. Inst. Mech. Eng. Lond.
(1952)
N. Kimura et al.
Kagaku Kogaku
(1964)
N. Kimura et al.
Kagaku Kogaku
(1969)

There are more references available in the full text version of this article.

Cited by (28)

Impact of fiber diameter polydispersity on the permeability of fibrous media
2022, Chemical Engineering Science
Citation Excerpt :
The details of this second set of numerical experiments can also be found in the Supplementary Material provided. Finally, despite the limited experimental data available in the literature for which sufficient details on the fiber size distributions used were provided, a comparison was made between the extended Davies’ correlation proposed from the simulation campaign and the experimental data reported in Payen et al. (2012) and Sakano et al. (2000) for mono- and bimodal fiber distributions. In this section, permeability results from the two sets of computations described in Section 3 are presented.
The classical empirical correlation proposed by Davies has long been recognized, particularly in air filtration, for its ability to predict the permeability of fibrous media. Several models have been proposed in the literature to extend its use for polydisperse fiber media and are reviewed in this paper. To assess their validity, the fluid permeability of more than 500 numerically generated fiber mats with a large variety of polymodal fiber distributions, porosities and mat thicknesses was predicted using the lattice Boltzmann method. A modified Davies’ correlation combining the volume-to-surface-area equivalent diameter, the coefficient of variation of the fiber diameter distribution and an empirical constant was shown to collapse all data onto a single master curve. The quality of the new correlation was further assessed on a new set of simulations carried out using a completely different numerical procedure and on a set of experimental data from the literature.
Transport through random bimodal and multimodal fiber structures
2020, Chemical Engineering Science
Citation Excerpt :
The authors report earlier X-ray scattering results indicating a bimodal distribution in such gels, with a radius of 1.5 nm for 87% of the fibers and 4.5 nm for 13% (i.e., β = 0.333 and γ = 6.69 in our notation). Sakano et al. (2000) measured the pressure drop and particle collection efficiency of PET (polyethylene terephthalate) air filters of fiber diameters 12 μm and 39 μm, for three different values of relative fiber density and ε = 0.97. Brown and Thorpe (2001) measured the pressure drop and capture efficiency for submicron aerosol flows through filters made of glass fibers of diameters 0.76 μm and 3.0 μm.
Random walk simulation results are reported for the transport and structural properties of random bimodal fibrous media for a range of values of the porosity, relative fiber size, and relative fiber density. Analytical expressions are derived for the mean intercept length of random bimodal and multimodal fiber structures, and are validated through simulations. A fiber dispersity factor emerges from the derivations, accounting for the deviation of the mean intercept length of such structures from that of unimodal fiber beds of the same porosity. Tortuosity factors in all diffusion regimes for in-plane and transverse flow through bimodal fibrous media of moderate or high porosity are practically independent of the relative fiber size and density and equal to those reported earlier for beds of unimodal fibers. The same holds for the formation factor of such structures, accounting for the dimensionless effective bulk diffusivity, thermal and electrical conductivity, magnetic permeability, dielectric constant and refractive index. The dimensionless viscous permeability of bimodal and multimodal fiber structures of moderate or high porosity may be obtained from that of unimodal fiber structures through simple multiplication by the square of the dispersity factor. This result is in line with a multitude of experimental data of the literature.
Filtration of Solid Aerosols
2017, Aerosol Filtration
During filtration, the collection of particles by the filter results in an increase in the pressure drop across the filter. For solid aerosols, this evolution, based on filtration time or the mass of the collected particles, can typically be divided into three phases. The duration of each phase is based on the filter media, the aerosol and operating conditions. Each phase is separated by more or less clearly marked transition zones.
Permeability prediction of an analytical pore-scale model for layered and isotropic fibrous porous media
2017, Chemical Engineering Science
In this study an existing analytical pore-scale model for flow parallel and perpendicular to 1D unidirectional fibres are used and adapted to propose permeability predictions for in-plane and through plane flow in layered 2D fibre arrangements as well as flow through 3D isotropic fibrous porous media. This is done by application of a weighted average equation in which different weights are assigned to flow parallel and perpendicular to 1D fibre arrangements, depending on the fibre orientation and average flow direction. Different arrays are considered, based on the degree of staggering of fibres with respect to the average flow direction. The effect of fibre orientation on permeability is investigated for flow through layered 2D fibre arrangements. It is illustrated how the permeability at low solid volume fractions reduces through the incorporation of the effect of developing flow. The effect of blocked pores at high solid volume fractions is also accounted for by introducing a percolation threshold solid volume fraction beyond which no more seepage takes place. A unimodal equivalent radius is furthermore introduced into the models in order to predict the permeability of bimodal fibrous media. The proposed permeability predictions are applicable over the entire range of solid volume fractions. Comparison with theoretical models from the literature as well as available experimental and numerical data proves satisfactory. The model characteristics that distinguish the proposed models from many other models in the literature are (i) being physically adaptable to extend its range of applicability, whilst at the same time (ii) balancing accuracy and simplicity (from an analytical point of view), and (iii) containing no empirical coefficients.
Study of the influence of synthetic filter media compressive behavior on its dust holding capacity
2015, Separation and Purification Technology
Citation Excerpt :
In comparison, using coarse fibers provides more stiffness and a three dimensional architecture to the media leading to an improvement of the dust holding capacity. The combination of the two types of fibers in blends shows a better nonwoven filtration compared to the equivalent unimodal diameter structures especially in the early stage of the filtration [3,4,5]. Beyond fiber characteristics, filtration behavior also depends on the fiber architecture in the filter medium and the resulting medium properties which are given by a wide variety of process configuration and parameters.
In depth filtration, a challenge is to obtain the highest dust holding capacity (DHC) reaching a critical pressure drop in order to improve the filter medium lifespan. Such property can be driven by the design of the fibrous structure. Indeed, various structural parameters and properties can influence the behavior of the fibrous filter medium and the evolution of its filtration properties during the clogging. These parameters are fiber properties, morphology and blend, medium structure such as multilayer architecture, thickness and pore distribution along with medium properties such as compressive behavior in the thickness direction and flexural rigidity. The focus of this work concerns the study of the relations between the DHC and the compressive properties of bilayer fibrous structures. A method for estimating the number of fiber-to-fiber contact points of a fibrous structure composed of three types of fibers is proposed. The results show the possibility to significantly increase the DHC performance by controlling the compressive behavior of the fibrous medium.
Pressure drop model for nanostructured deposits
2014, Separation and Purification Technology
This study presents a new pressure drop model developed for cakes composed of nanostructured particles. The cake structure is understood as a tangle of chains composed by juxtaposed primary particles with (aggregates) or without (agglomerates) a partial overlap. Since cake porosity is one of the main parameters determining aeraulic resistance, an experiment protocol based on the changes in deposit thickness as a function of the cake mass per surface area has been developed to accurately determine this parameter. To this end, the pressure drop and the porosity of the cakes created by the filtration of carbon nanoparticles aggregates and agglomerates on PTFE membrane were measured. The aggregate and agglomerate count median mobility diameters range from 91 nm to 170 nm and from 48 nm to 62 nm, respectively. The associated Peclet numbers range from 0.19 to 53 for filtration velocities of 0.01, 0.05 and 0.09 m/s. Initial experimental results indicate that the porosity of the cakes ranges from 0.94 to 0.984 in correlation with the Peclet number of the aggregates or agglomerates. The agreement between experimental results and the pressure drop model is fairly good. Of the experimental values, 95% are within plus or minus 25% of the theoretical value.

View all citing articles on Scopus

View full text