INTRODUCING A NEWLY DEVELOPED FABRIC FOR AIR FILTRATION

: Woven and nonwoven fabrics present fi ltration effi ciency higher than other air fi ltration media. Fabrics are selected according to air ﬂ ow conditions and particle characteristics. The majority of air fi ltration media are nonwoven fabrics because of their cost, but they need high fi ltration area for high effi ciency. Modifi ed construction of woven fabric introduces high performance in air fi ltration and decreases fi lter size, which tends to have better competition abilities. The designed fabrics have considerable thickness and suitable pore characteristics by applying roving instead of weft yarns. Four factors (roving count and their turns per inch, picks per inch and fabric designs) were varied in order to study the effect of these factors on their performance in fi ltration. Optimum operating conditions for a determined range of air permeability and pore size were obtained.


Introduction
According to Tharewal et al. [1], in textile industry, the generation of fl y and dust affects both laborers' health and product quality. In hospitals, the level of air cleaning and fi ltration effi ciency are very important factors to be controlled. According to Das et al. [2], the various problems of contaminated air cause several types of air fi lters to be developed. The main objective of air fi ltration is to remove solid particles from air stream [3][4][5]. Filters can be used to recover solid particles [1]. Das et al. [2] and Kothari et al. [6] stated that fabric fi lters, especially woven and nonwoven, are considered the best type of air fi ltration media. These porous or semi-porous structures with considerable thickness allow air stream to pass through various points of fi ltration. According to Irwin [5] and Bird et al. [7], particles bigger than the pore size of fi lter media are stopped by the membrane. This fi ltration behavior is called as surface fi ltration. Many layers of dust accumulate on the fi lter surface, which increase fi ltration effi ciency. This fi ltration behavior is called as cake fi ltration. According to Lawrence et al. [3] and Bird et al. [7], the fabric itself provides the support and true fi ltering usually occurs through the retained dust cake; this is true for woven fi lters. Depth fi ltration occurs whenever the pore size of fi ltration media is bigger than the particle size. Moreover, some particles block some pores that allow fi ltration effi ciency to increase. This fi ltration behavior occurs whenever fi ltration media are thick layers of fi bers that are randomly or less orderly arranged like nonwoven fi lter media. On the other hand, successive blockage during fi ltration decreases the lifetime of fi lter. Nonwoven fi lter media would not be completely cleaned because of the impeded solid particles [7,8].
Air permeability, porosity and pore characteristic affect the performance of air fi lters [9][10][11]. Textile materials contain three kinds of pores such as closed pores, blind pores and through pores [12]. Jena and Gupta [13] stated that the largest through pore diameter determines the effi ciency of the fi ltration medium. Maini et al. [14] found that the existence of these pores is a result of the voids within the fi ber itself and the voids between the fi bers. According to Guangbiao and Fumei [15], inter-yarn porosity has a very strong infl uence on fabric permeability. Fabric permeability was considered as two parts of inter-yarn and inter-fi ber interstices.
According to Tugrul [16], these properties depend mainly on material geometry and construction. Fabric construction, mass per unit area, thickness and packing density are factors controlling air permeability and pore characteristics of woven fi lters [1,16]. Modifi ed construction of woven fabric, through replacing the ordinary weft yarns by roving, will increase fabric volume through increasing its thickness. This allows the air stream to path through more and more fi ltration points. In addition, fi ltration effi ciency is expected to be high with the minimum fi ltration area.
with maximum TPI can sustain stresses during insertion as wefts. Roving should be rewind on cone package to be suitable for loom creel. On the rapier loom, the rapier head has the ability to grip and genteelly insert roving across the shed. The fabric constructions consist of constant warp of 20/2 Ne, 13 TPI and 22 ends per inch. Warp and weft have the same material, which consists of 50% cotton or 50% cotton waste. The weft yarns were obtained by varying both roving count (0.8, 1, 1.2 Ne) and TPI (1.4, 1.8, 2.2). Three levels of picks per inch (PPI; 16, 20, 24) and three fabric designs (plain, twill, sateen) were applied for obtaining the modifi ed fabrics. Regression analysis was applied on the obtained results to attain the best operating conditions. Some fabrics of the suitable fi ltration properties were selected for the design of air fi lters. Selected fabrics for fi lter design had the same fabric construction, except weave design, which was as follows: roving count 1 Ne, 1.4 TPI and 16 PPI. The designed fi lters were subjected to air fi ltration tests in order to study and evaluate their performance.

Fabric testing
Fabric mass was evaluated according to ASTM D 3776-96 (reapproved 2002). The air permeability test was carried out according to ISO 9237. The maximum pore size was measured according to British standards (BS 3321). Fabric thickness was evaluated according to ASTM standard D 46 (1975). Packing density was calculated based on the results of fabric mess and thickness tests through equation P d = 1000 * M/T, where P d is the fabric packing density (gm/cm 3 ), M the fabric mass per square meter (gm/m 2 ) and T the fabric thickness (mm).

Filter testing
Three properties are measured in order to determine the performances of tested fi lter according to International Standard ISO 5001/2000 and Egyptian Standard ES 918/2007 [17,18]. It is also recommended that the most important properties affecting fi ltration performance are resistance to air fl ow, average fi ltration effi ciency and fi lter dust capacity. The previous properties were measured during the air fi ltration test, which are classifi ed into three tests.

Resistance to air fl ow test:
The test is carried out by the device shown in fi gure 2.1. Without dust loading and eject the absolute fi lter which shown in fi gure 2.2.a and housing 2.2.b [19]. High-effi ciency particulate air absolute fi lter introduces a minimum effi ciency of 99.97% when tested at particles' diameter of 0.3 mm, which is recommended to use in air fi ltration tests to evaluate and test the performance and fi ltration effi ciency of a new fi lter before operation. The properties of the used absolute fi lter are shown in fi gure 2.2.c. Resistance to the fl ow test is carried out through operating the system with and without the tested fi lter and then, the pressure drops Dp 1 and Dp 2 are determined in both cases. Pressure drop of the fi lter can be calculated by the following equation: Dp = Dp 2 -Dp 1 . The calculation is repeated with several values of air fl ow rate, and pressure drop of the fi lter is determined for each fl ow rate.

Filtration effi ciency test:
The system is operated for 10 minutes before dust loading to adjust the air fl ow rate, and then, the dust injector is operated. The fi ltration effi ciency test can be performed as follows: masses of tested and absolute fi lters are measured before and after the test to determine the dust mass on the tested fi lter (M t ) and absolute fi lter (M a ). Filtration effi ciency can be determined by the following equation: .

Dust capacity of a fi lter:
It is defi ned as the ratio of dust mass accumulated on the tested fi lter (M t ) to the total dust mass injected during the fi ltration test. Dust concentration should be within the range (1 to 3 gm/m 3 ) of air. Dust particles are prepared according to ISO 12103-A4 for the coarse dust test. The dust preparation device mainly consists of fi ve metallic meshes. Table 2.1 shows pore diameter of each mesh, the range of particles' size on its surface and the percentage of sharing in total dust mix.  table 3.1, which demonstrate that the proposed regression represents signifi cantly the obtained data since R 2 is higher than 0.91 and the F-signifi cance tends to zero. Also the roving counts, PPI and their interactions are the most important factors affecting all the tested properties. The regressions coeffi cients in case of air permeability and maximum pore size are the same, with the same sign and rank. The effects of different factors on some properties are discussed according to the experimental data as follows.

Fabric air permeability
Figure 3.2 shows that air permeability increases as roving count increases for minimum and medium PPI. Whenever roving count becomes fi ner than 1 Ne, permeability decreases especially for plain weave of 24 PPI. Practically, plain fabric is not woven easily for maximum PPI or maximum cover. So, coarse roving forces weaving loom to decrease the actual PPI in the produced fabric, while fi ne roving allows the actual PPI to increase and tend to be as the same as the designed 24PPI. However, for coarse counts, the actual PPI is lesser than the designed one. So, air permeability decreases as roving count becomes fi ner than 1 Ne for maximum PPI. It is normal relationship between roving count and air permeability at 16 PPI. As shown in fi gure 3.3, air permeability increases as roving TPI increases up to 1.8, and then, there is a little reduction in air permeability. The maximum vales of air permeability can be obtained at the medium level of roving TPI. It can be found that the newly developed fabric depends on roving in its structure. Air passes through voids between roving within fabric structure, and also, air can path through pores between     fibers within roving structure. Roving diameter decreases as its TPI increases. So, voids or spaces between roving within the weave structure have to be enlarged. However, TPI higher than 1.8 decreases the voids between fibers within the roving structure and decreases the resultant fabric permeability. Fabric permeability can be controlled by controlling twist level for the same roving count.

Fabric pore size
As stated the effect of studied factors on pore size is typically as that on air permeability. It can be observed from figure 3.4 that pore size increases as roving TPI increases up to 1.8 TPI and then pore size decreases, which is explained previously at air permeability discussion. The figure shows that minimum values of pore size can be obtained at plain weave. Figure 3.5 shows that the rate of reduction in pore size due to increasing roving TPI, more than 1.8, in twill weave is lower than that of plain weave. This rate of reduction is the least for sateen weave, which is considered to introduce soft fabric and is easily woven with high PPI, as shown in figure 3.6. Increasing roving count results in increasing pore size in the fabric produced of 16 and 20 PPI. There is a small reduction in the pore size for the fabric of 24 PPI, which is discussed previously at air permeability discussion. Since air permeability and pore size are geometrical properties of a material, so yarn or roving structures, count, PPI, TPI and fabric design affect these properties.

Fabric packing density
Packing density is determined based on results of both fabric mass and thickness tests. As stated before, packing density α (mass/thickness), for the same fabric area. As shown in fi gure 3.7, packing density increases as roving TPI increases for the medium 20 PPI. It can be explained that increasing TPI results in decreasing roving diameter, which decreases fabric thickness, and packing density increases as the mass remains constant. At 24 PPI, whenever roving TPI increases, also diameter reduces, which decreases both fabric thickness and its mass. It can be explained that the actual PPI in the produced fabric tends to be as the same as the designed PPI for the fi ner roving with a lower diameter to give the same fabric cover. Then, fabric mass tends to increase as actual PPI in the fabric increases. So, the resultant packing density remains constant in this case. As shown in fi gure 3.8, the fi ner the count, the lower the packing density. At 1.8 roving TPI, relations between measured properties and studied factors will be normal. The rate of reduction in packing density for 24 PPI is higher than that for 16 PPI, which is discussed previously. The maximum PPI in the produced fabric needs roving with a low diameter to be woven easily. Finer roving means lower diameter, and actual PPI increases. In general, sateen weave has the least packing density than twill or plain weaves. Sateen weave is much thicker than other designs, plain and twill, according to the results of the fabric thickness test. So, sateen is more permeable than twill or plain weaves because of its low packing density. Figure 3.9 shows the designed fi lters that were prepared for the air fi ltration test. This type of fi lters, which is known as surface fi lter, needs area of fi ltration media lower than that for nonwoven pocket fi lters, which consume large area of fabric. The performance of the designed fi ltration media was evaluated as follows.

Filtration performance of the designed fi lters
3.4.1. Resistance to air fl ow Figure 3.10 illustrates the relation between air fl ow rate and pressure drop for a plain fi lter. The fi gure presents that fi lter's pressure drop increases rapidly as the air fl ow rate increases from 3.3 to 3.5 m 3 /min. Within this range, it can be noticed that pressure drop increases by 40% as the air fl ow rate increases by 6%. The pressure drop increases slightly whenever the air       Within the previous range of air fl ow rate, it can be noticed that pressure drop increases by 58% as the air fl ow rate increases by 33%. The pressure drop increases slightly whenever the air fl ow rate increases from 5.3 to 6.6 m 3 /min. It is recommended for this fi lter that the optimum fi ltration performance can be achieved whenever the air fl ow rate varies between 5.3 and 6.6 m 3 /min. Within the previous range, it can be noticed that pressure drop increases by 8% as the air fl ow rate increases by 25%. So, the air fl ow rate has a little effect on the pressure drop of a fi lter.
Each of the three previous fi gures can be divided into two zones according to the slope of the line representing the relationship between air fl ow rate and pressure drop. The fi rst zone is the high slope zone, while the second zone is the low slope zone. The slope of the line in the fi rst zone is the highest one for plain fi lter than twill fi lter, while sateen fi lter has the least slope. It is also the same for the slope of the line in the second zone. This fl ow behavior assists fi ltration performance of sateen fi lters with higher air fl ow rates and acceptable pressure drop.     efficiency for particles greater than 25 mm, which are suitable for air filtration applications in spinning mills, especially at the blow room area and vehicles air filters also. Dust capacities of the designed filters are approximately 4% for plain filter, 8% for twill filter and 7% for sateen filter. These low values of the dust capacities of the designed filters increase the estimated time for each filtration cycle before cleaning the filter.

CONCLUSIONS
Air passes through voids between fibers within the yarn structure and between yarns within the fabric structure. 1.8 TPI is the optimum weft twist for high permeability, but 1.4 TPI is the optimum weft twist whenever the minimum pore size is required. Increasing PPI by 50% results in decreasing air permeability by 60% to 65% and decreasing pore size by 40% to 50%. Fabric thickness of sateen weave is higher than other designs; therefore, low packing density and resistance to air flow are achieved in sateen weave. For weft counts 0.8 to 1 Ne, 1.4 to 1.8 TPI and 16 to 20 PPI, the fabric has a reasonable range of air permeability and pore size for some filtration applications like textile industry and some vehicles motors. Controlling both yarn and fabric structures is a very important factor to achieve the designed air permeability and pore size. Nonwoven filter media are disposal materials. In addition, they cannot be cleaned completely after filtration because of the impeded particles within their structure. Finally, the designed fabric increases the competition abilities of woven filters, range of air flow rate and filtration efficiency and decreases the filtration area without scarifying efficiency.  Plain  100  50  100  72  24  0  100  3  588   Twill  100  37  100  52  36  0  100  2  456   Sateen  100  56  100  84  35  0  100  2  500