CA1258643A - Cylindrical fibrous structures and method of manufacture - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE
Cylindrical fibrous structures comprising a fibrous mass of nonwoven, synthetic, polymeric microfibers wherein the microfibers are substantially free of fiber-to-fiber bonding and secured to each other by mechanical entanglement or intertwining, the fiber structure having a substantially constant voids volume over at least a substantial portion of the structure and, prefer-ably, a graded fiber diameter structure, both as measured In the radial direction. The structures are particularly useful as depth filters.

Description

j L~

This Inven~ion relates to cyllndrlcal flbrous struc-tures, more par~lcularly to cyllndrlcal Fibrous structures com-~rlslng nonwoven, synthetlc, po I ymcrlc mlcroflbers p~rtlcularly useful as depth fllters ~or a varlety o~ fluld clarlflcatlon 5 appllcatlons.

Thls applIcatlon Is a dlvlslonal applIcat!on of copend-lng applicatlon No. 471,536 flled January 4, 1985.

Nonwoven structures formed from a varlety of materlals, Includlng natural and synthetlc flbers In both staple and contln-uous -Form, have long been known and used in depth fllter opera-tlons. Such depth fllters generaliy have a range of pore dlame-ters. If the fllter medlum Is thln, the larger partlcles In the fluld belng flltered wlll pass through those areas havlng the larger pores. If the effluent passlng through the fllter medium Is then passed through a second equal layer, some of the larger partlcles remalnlng In the fluld wlll be removed as they encounter more flnely pored areas. Slmlla~ly, use of a thlrd equal fllter layer wlll remove addltlonal large partlcles, fur-ther Increaslng the flltratlon efflclency. Use of a thlc~ layer of f 1 Iter medlum wlll have the same effect as uslng multlple lay-ers of equal total thlckness. The Increased efflclency 50 obtalned is one of the motlvations for uslng depth filtratlon.
To be useful for a glven appllcatlon, a depth ;3 Lilter m~s~ provide the requisite level of effici-ency, that is, an acceptable level of removal of particles of a specified size present in a fluid ! being filtered. ~nother important measure of the 5performance of a filter is the time to clogging in a given type of service, that is, the time at which the pressl-re across the filter has either reached a level at which an undesirable or unacceptable power input is required to maintain adequate flow, or the poten-lOtial Eor filter collapse with the accompanying lossof integrity and effluent contamination is too high.
To extend filter life, it has long been the practice to design depth filters such that their density is lower in the upstream portions, thus pro-]5viding relatively larger pores upstream and smallerpores downstream. By virtue of the graded density, the contaminated fluid passes through progressively smaller pores, and particulate material being fil-tered from the incident fluid penetrates to varying 20depths according to its size, thereby allowing the - filter element to accomodate more solids (a higher dirt capacity) without affecting flow and consequent-ly providing a longer eEfective life for the depth filter. Stated otherwise, in theory, the larger 25upstream pores remove larger particles which would otherwise clog the downstream, finer pores and filter life is thereby extended.
The density oE the Eilter meclium is, however, in itself, an important determinant of the medium's 30behavior in service. The op-timum density of a filter medium is determined by two factors:
(1) In order to have a high dirt capacity, the percent voids volume in the depth filter should be as high as possible. The reasons for this may be seen 35by comparing a gravel screen made using woven wire 3L25~ 3 with a metal plate e~ual in size to the screen but containing a single hole. The metal plate will be clogged by a single oversized particle, while the screen, re~uixing a large number of particles to 5become clogged, will remain in service longer.

(2) In a fibrous depth filter, there is an upper limit beyond which Eurther increasing the per-cent voids volume becomes undesirable. As the voids volume is increased, the fibrous depth filter is more 10 readily compressed by the pressure drop generated by the fluid passing through it; this is particularly troublesome when the fluid is viscous where, if the percent voids volume is too high, the filter medium will collapse at a very low differential pressure.
]5 As it collapses, the pores become smaller and the differential pressure increases, causing still more compression. The resulting rapid increase in pres-sure drop then tends to reduce life rather than - as might otherwise be expected with a high voids volume 20 filter - extending it. Use of a very low density - (high voids volume) can also ma)ce the filter very soft and thereby easily damaged in normal handling.
Thus, there is a practical upper limit to voids volume, the value of which depends on the clean dif-25 ferential pressure at which the filter is to be used.For any given type of service there will be an opti-mum percent voids volume at which filter life will be at a maximum.
~s noted above, attempts have previously been 30 made to provide depth ilters from fibrous materials and to extend their effective life by providing a graduated porosity, accomplished by a density profile with the density increasing in the direction of flow of the fluid being filtered. These attempts have met 35 with some success but the filter structures have _ --4--5 ~3 6 L~

substantial limitations. These include relatively short life due to the limited range through which pore diameters can be changed, and reduction in pore diameters ~ue to compression when used with viscous 5 fluids or at very high liquid flow ra-tes.
The present invention, then, is directed to cylindrical fibrous structures, particularly useful as depth filters, and a method of manufacturing them which substantially overcomes the shortcomings of the 10 cylindrical fibrous depth filters which have hereto-fore been used. ~s will become apparent from the following description of the invention, the cylin-drical fibrous structures in accordance with this invention typically have, relative to fibrous cylin-]5 drical depth filters of the type previously avail-able, extended filter life, i.e., higher dirt capac-ity at e~ual efficiency, or better efficiency at equal lifel or both better efficiency and higher dirt capacity. They also have the ability to remove much 20 finer particulate contaminants -than have heretofore been capable of being removed by previously available commercial fibrous cylindrical depth filters.
In the method in accordance with this invention, synthetic ~olymeric material is fiberized by extru-25 sion into a high velocity gas stream and collected asa mass of mechanically entangled or intertwined fi-bers in the form of a hollow or annular cylindrical structure, as will be described in more detail below.
In the course of investigating this process, and the 30 products thereof, several surprising observations were made:
(a) Increasing the percent voids volume (by decreasing the density) generally yielded only a small increase in filter life.
(b) When the project was initiated, it was ~25~3~43 assume~ th~t iE two filters differing in fiber dia-meter but otherwise equal were compared, the filter with the finer ~iber would be more compressible.
Contrary to this assumption, it was found that the 5 filter with finer fibers had a resistance to com-pression that was substantially equal to that with coarser fibers, providing that the densities of the two filters, i.e., the percent voids volume, were the same. As a consequence of this discovery, cylin~
lO drical filter structures using fibers as fine as about 1.5 micrometers were prepared and found to have satisfactory resistance to compression. Cylindrical filter structures, or cylindrical filter elements as they are sometimes referrcd to herein, made with ]5 fibers in the range of from about 1.5 to about 2.5 micrometers and with annular thicknesses of the fib-rous mass of about 0.6 inch (1.5 cm), had extraor-dinarily high efficiencies, for example, in excess of 99.9999 percent for removal of bacteria organisms as 20 small as 0.3 microme-ter in diameter.
(c) secause of -the desirable characteristics obtainable with depth filters prepared from these very fine fibers, cylindrical filter structures with fine fibers in their downsteam portions and coarser 25 fibers upstream but with a constant voids volume throughout were prepared. These filter elements combined extraordinary efficiencies, e.g., 9~.999 percent, removal for bacteria organisms as small as 0.3 micrometer in diameter with relatively high dirt 30 capacities comparable to dirt capacities of much coarser conventional cylindrical depth filter ele-ments.
The cylindrical fibrous structures in accordance with this invention comprise a fibrous mass of non- -35 woven, synthelic, polymeric microfibers, the fibrous -6- ~ 5 ~6L~-~

mass having a substantialiy constant voids volume over at least a substantial portion of the fibrous mass, preferably at least the major portion, as meas-ured in the radial direction. The microfibers are 5substantially free of fiber-to-fiber bonding and are secured to each other by mechanical entanglement or intertwining. Filter structures in accordance with the subject invention are preferably supported by the incorporation of a hollow, open, relatively rigid, lOcentral suppor-t member or core, with the fibrous mass of microfibers on the exterior of the support member.
~lso, for most applications, it is preferred that the fibrous mass have a substantially constant -voids volume and a graded fiber diameter structure over at ]5least a portion thereof as measured in the radial direction, obtained by progressively varying the fiber diameter as the cylindrical fibrous structure is built up while simultaneously holding the voids volume cons-tant.
The method of manufacturing the cylindrical fibrous structures in accordance with the subject invention comprises the steps of:
(a) extruding synthetic, polymeric Inaterial from a fiberizing die and attenuating the e~truded poly-25meric material to form synthetic, polymeric micro-fibers by -the application of one or more gas streams directed toward a rotating mandrel and a forming roll in operative relationship with tlle mandrel;
(b) cooling the synthetic, polymeric microEibers 30prior to their collection on the mandrel to a temper-ature below that at which they bond or fuse to each other to substantially eliminate fiber-to-fiber bond-ing; and (c) collecting the cooled microfibers on the 35mandrel as a nonwoven, fibrous mass while applying a 64:~

force on the exterior surface of the collected micro-fibers on the mandrel by the forming roll to form the cylinclrical structure, wherein the process variables are controlled to provide the collected fibrous mass 5 with a substantially constant voids volume over a substantial part thereof, pre-Eerably at leas-t the major portion, as measured in the radial direction.
It is preferred, especially or coarser fibers, that cooling of the microfibers be enhanced by the 10 injection oE a cooling fluid into the stream of the microfibers prior to their impingement on the mandrel or the forming roll to assist in eliminating fiber-to-fiber bonding.
~dditionally, it is preferred that the attenu-- ]5 ated microfibers impinge on the forming roll which is held at a temperature substantially belo~ the melting or softening point of the fibers to further enhance cooling prior to -the microfibers being transferred to and collected on the rotating mandrel, thereby pro-20 viding additional cooling and further reducing the likelihood of undesirable fiber-to-fiber bonding.
Preferably, the stream of microfibers is directed -toward the forming roll and mandrel in such a manner that at least the major portion contact the forming 25 roll first (where they are cooled further) and from where -they are then transferred to the rotating man-drel. Also, if the forming roll is wet, particularly when microfiber collection on the mandrel is initi-ated, more consistent start-ups are obtained due to 30 better (more uniform) transfer of the microfibers to the mandrel, i.e., the potential Eor undesirable layer-to-la~er bonding is reduced and a smoother wrapping with minimized clumping and a more regular or uniform laydown of fibers is obtained.
In addition, the apparatus is preferably de--8- 1~ 3 signed so as to allow free access of secondary air in order to assist in the rapid cooling of the hot, freshly formed fibers.
Figure 1 is a perspective view of an apparatus 5 which can be used to form the c~lindrical filter structures in accordance with this invention;
Figure 2 is a perspective view showing an an~
cil]ar~ collection means which can be used with the apparatus oE Figure l;
Figure 3 is a partially cut away perspective view of a cylindrical filter structure in accordance with this invention;
Figure ~ is a yraph of resin (or polymeric mate-rial) pressure versus fiber diameter;
]5 Figure 5 is a graph of fiberizing air pressure versus fiber diameter;
Figure 6 is a graph of forming roll air pressure versus fiber diameter;
Figure 7 is a graph of resin rate versus resin 20 pressure; and Figure 8 is a graph of particle diameter for which the removal rating equ~ls 99.9 percent versus fiber diameter.
The subject invention will be better understood 25 by reference to the drawings. Turning first to Fig-ure l, there is shown an apparatus useful for forming the cylindrical filter structures in accordance with the subject invention comprising a fiberizer or fi-berizing die lO to which molten resin is delivered by 30 a motor-driven extruder 11 and to which hot com-pressed gas, preferably air, is delivered Erom a heater 12. The fiberizer lO contains a multiplicity of individual extrusion nozzles 13 by which the mol-ten resin is converted to fibers. In the preferred 35 mode illustra-ted in Figure 1, the hot resin ~or poly-_9_ ~5~ 3 meric material) stream delivered from the extruder ]1to the fiberizer 10 issues from each nozzle under pressure (fiberiziny air pressure). The molten, thermoplastic pol~meric microfibers generally desig-5 nated 1~ are formed as the resin is extruded from thenozzles 13 and attenuated b~ the jets of hot gas referred to above which carry the microfibers upward in the direction of a cylindrical Eorming roll 15 which is in operative, rotating relationship with the 10 power driven rotating, and preferably also recipro-cating, man~rel 16. The Eorming roll 15 may be cooled, e.g., by passing unheated ambient air through i-ts internal portions. When a supported filter struc-ture is being formed, prior to initiation of col-]5 lection of the microfibers on the mandrel, one ormore open, relatively rigid central support members or filter cores 17 ~such as that shown in detail in Figure 3) are placed on the mandrel. Alternatively, as discussed below, for some low pressure applica-20 tions a central support member or core may not berequired, in which event the cylindrical filter struc-ture can be formed directly on a solid mandrel. The mandrel 16 and the filter cores 17 are designed such that the cores ro-tate with the mandrel, either by 25 means of friction between the fil-ter cores and the mandrel or by use of springs or other arrangemen-t.
The forming roll 15 is preferably mounted on bearings so that it is freely rotatable, i.e., it rotates freely when in contact with the mandrel 16 or 30 with fibrous material collected on the mandrel 16 or with filter support cores 17 (as illustrated in Fig-ure 3) which may be placed on the mandrel 16. Ad-ditionally, the forming roll 15 is preferably biased, for example, by an air cylinder 18, operating on the 35 shaft 9 which is rotatably mounted Oll bearings. The --10-- ~ LP~t3 air cylinder 1~ through the shaft 9 applies bias to the orming roll 15 in a controlled manner towards or away from the mandrel 16. Depending on the fric-tional characteristics of the air cylinder and the 5 charactex of the fibers being collected, damping of the shaft 9 may be desirable to prevent vibration of the forming roll 15.
The mandrel 16 is rotated by a motor (not shown), generally at a rate oE from about 50 to about 500 10 rpm, and, in the preferred embodiment shown in Figure 1, is reciprocated axially at a rate generally be--tween about 10 (3.0 meters) and about 300 feet (91.4 meters) per minute. The length of stroke of the reciprocating mandrel will depend on the desired ]5 length of the cylindrical filter structure or struc-tures being formed.
Especially when making relatively coarse fibers, a suspension of finely divided water droplets 19, or other cooling fluid, is preferably injected into the 20 stream of fibers 14 from one or both sides by the nozzles 20, impinging on the stream of fibers a short distance above the extrusion nozzles 13, e.g., 1 to 5 inches (2.5 to 12.7 cm) to cool the microfibers and help prevent fiber-to-fiber bonding.
In operation, the microfibers are projected upward in the direction of the forming roll 15 and the mandrel 16, generally at least in part impinging on the forming roll 15, from where they are contin-uously transferred to the filter cores 17 mounted on 30 the rotating, reciprocating mandrel 16. As the man-drel 16 rotates and reciprocates, the diameter of the cylindrical mass of flbers collected on the ilter cores 17 increases.
It is generally preferred that at least the 35 major portion of the fibrous s-tream 1~ imp;n~e on the 5f~L~

forming roll 15 rather than on the mandrel 16 as this results in a more uniform and more reproducible pro-duct in whicll the fibers exhibit little or no unde-sirable interfiber bonding, i.e., they are substan-5 tially free of fiber-to-fiber bonding.
Under some conditions, particularly when col-lecting fibers less than about 1.8 to 2 micrometers in diameter, an auxiliary collection member 22, as is shown in Figure 2, may be used to advantage. This 10 member can be a flat, stationary - relative to the forming roll - sheet or plate. Alternatively, it may have a moderate radius with the concave side facing downwards toward the fibrous stream 14. It is pre-ferably mounted such that one edge is about 0.1 inch ]5 (0.25 cm) or less from the forming roll surface. The collection member 22 is secured by brackets 23 to the frame 24 supporting the forming roll 15. The func-tion of member 22 is to collect fine fibers which would otherwise bypass the forming roll. As rapidly 20 as these fibers collect on member 22, they are trans-ferred to the forming roll 15 and thence to the ro-tating reciprocating mandrel 16.
The system used for fiberizing the resin or polymeric material can take a varie-ty of forms, many 25 of which have been set forth in the patent and jour-nal literature. See, for example, the paper titled "Superfine Thermoplastic Fibers" in the ~ugust 1956, Volume 48, Number 8, edition of Industrial and ~n-gineeriny Chemistry. The resin stream or streams can 30 be discontinuous ~i.e., delivered by individual noz-zles) or continuous (i.e., deliverecl through a slot), and the air stream or streams can be similarly con-tinuous or discontinuous. ~dditionally, combination of these design configurations can be used, e.g., 35 when preparing a filter element Erom two or more 6~1~

di~ferent polymeric materials.
Also, a number of process variables can be con-trolled to provide any desired combination of fiber diameter and voids volumes within the limits of the 5 apparatus. As will be evident from consideration of the Examples below, it is preferred that four var-iables be used in operating the apparatus of Figure 1. These are:
(1) The Rate Of Delivery Of Resin (Or Polymeric Material) To The Fiberizing Die:
This xate is adjusted by increasing or de-creasing the pressure developea by -the extruder, which in turn is accomplished b~ changing its speed.
]5 As the rate is increased, coarser fiber is generated and voids volume of the collected fiber cylinder tends to decrease.
(2) The Fiberizing Gas Flow ~ate:
This rate is adjusted by altering the pres-20 sure at which the gas, typically air, is delivered to the fiberizing die. As the flow rate of the gas stream (or streams) is increased, the fiber diameter becomes smaller and the voids volume tends to in-crease.

(3) The Forming Roll Pressure:
The forming roll pressure is varied as re-quired to maintain the voids volume constant. For example, if the fiberizing gas rate is decreased to increase the diameter of the microfibers, the forming 30 roll pressure must be decreased to maintain a con-stant voids volume.
(~) The Quantity And Type Of Fiber Cooling:
These include the quantity of secondary air, - die to collector distance (see below), the tempera-35 ture of the forming roll on which it is preferred -13- 1~5~

that the fibers impinge, the quantity and mode of delivery of liquid coolant, and the rate of rotation and reciprocation of the mandrel. When fiber dia--meter is smaller than about 3 to 6 micrometers, wa-5 ter cooling is not required, although it can be used.The effect of these various cooling means on the density of the collected fiber varies and must be determined empirically.
Other process variables which influence the 10 character oE the formed filter cylinder but which once set - in the preferred mode of opera-tion - no longer need be altered include:
(1) The fiberizing die to collec-tor distance (DCD), if too large, permits the fiber to form bun-]5 dles prior to deposition on the forming roll (a phen-omenon known as "roping"), causing the formation of a non-uniform product. If DCD is too small, the fibers may be insufficiently cooled when collected and this may result in melting or softening, which tends to 20 close off pores and obstruct free flow of fluids through the fibrous mass when used as a filtering device.
The optimum DCD depends upon the diameter and velocity of the fibers and upon the rapidity with 25 which they are cooled and is best determined by trial ar;d error.
DCD can be used as a controlling variable and, indeed, was so used during the early phases of the development of this invention but was discontinued 30 because it proved easier and adequate to vary the four variables listed above.
~ 2) The temperature of the resin or polymeric material supplied to the fiberizing die has a strong effect on product characteristics. As this tempera-35 ture is increased, fiber diame-ter decreases, but ~ 6~3 excessive temperatures cause the production of very short fibers and shot, as well as significant reduc-tion in resin molecular weight due to depolymeriza-tion. 5'he optimum temperature is best determined by 5 trial and error since it depends on a number of fac-tors, including the particular polymeric material, the nature of the structure desired and the par-ticulars of a given apparatus, for example, the ex-truder size as related to the resin flow rate.
(3) The temperature of the fiberizing air has a relatively minor effect, provided that it is held within about 50 degrees F' (2~ degrees C) of the resin temperature.
(~) The temperature of the forming roll is ]5 preferably low, for example, near ambient, to help prevent interfiber fusing of the fibers collected on it prior to transfer to the rotating reciprocating mandrel.
(S) The rate of rotation of the mandrel; higher 20 rotation rates help to prevent interfiber bonding.
(6) The rate of reciprocation of the mandrel;
higher reciprocation (or axial translation) rates help to prevent interfiber bonding.
sy the method in accordance with this invention, 25 the fiber diameter of the cylindrical fibrous struc-tures can be varied in a continuous or step-wise rnanner from one part of the cylindrical structure of the fibrous mass to another - as measured in the radial direction - by varying the resin and fiber-30 izing air flow rates while the voids volume is main-tained substantially constant by varying the forming roll bias force on the cylindrical mass of fibers as the structure is formed on the rotating mandrel. As may be seen in Figure 8, if the voids volume is con-35 stant, the pore size varies with the fiber diameter.

~15- ~2~8~

By the method in accordance with this invention, the pore diameter can be varied continuously or stepwise from one part of the filter to another in any desired manner.
When the desired outside diameter of the cylin-drical fibrous structure has been reached, the opera-tion is terminated by discontinuing the flow of resin and air onto the forming roll 15, discontinuing or reversing the bias oE the forming roll 15 and stop-10 ping the mandrel 16, following which the formed cyl-indrical fiber structure together with the core or cores is removed from the mandrel 16. The ends of the resulting cylindrical structure are then cut to length and if more than one core 17 has been used, ]5 additional cuts are made to separate each section, thereby forming individual cylindrical filter struc-tures, sometimes referred to herein as filter cylin-ders, filter elements or simply as elements. ~ cy-lindrical filter structure in accordance with this 20 invention is illustrated in Figure ~. The cylin-drical filter structure generally designated 30 is comprised of the hollow support core 17 and a fibrous mass of nonwoven, synthetic, polymeric microfibers 31.
~s noted above, for some applications it may be desirable to Eorm the cylindrical fibrous structures in accordance with this invention directly on the mandrel without -the use of an internal support or core. For most purposes, however, it is desired that 30 the structure, when used as a filter, be able to withstand, without collapse or loss of integrity, diEferential pressures of 40 psi (2.~1 ~g/cm2) or higher. The voids volumes of the unbonded fibrous mass of the filter structures in accordance with this 35 invention which yield desirable combinations of high -16- ~5~3 efficiency and long life in service are, in general, too higll to withstana pressures of this magnitude and would collapse if an internal support member were not provided. Accordingly, for most applications, it is 5 desirable to form the filter on a hollow foraminous, or open, relatively rigid central support member or core desi~ned in such a way as to provide support for the collected fiber or fibrous mass. The central support mernber or core 17 must be open or foraminous 10 in nature, as illustrated in the perspective view of a typical supported cylindrical filter structure in Figure 3, since it must provide adequate passages for flow of filtered fluid into the central portion of the core (outside/in filter configuration) or, con-]5 versely, passage of fluid to be filtered from thehollow cen-ter of the filter structure into the fib-rous mass (inside/out configuration). Typically, the core, which is relatively rigid vis-a-vis the mass of collected fibers on the exterior thereof in order to 20 provide the requisite support, will have openings 32 with spans preferably on the order of one-quarter inch (0.6 cm) or less and, generally, not more than one-half inch (1.3 cm).
The central support member or core can be made 25 by a variety of processes and from a variety of mat-erials, for example, from synthetic resin by injec-tion molding or extrusion, or from metal by conven-tional processes. While not required, the core may have a multiplicity of small protuberances on its 30 exterior to assist in securing the microfibers to the exterior of the core.
Another alternative is to build up a support core of self-bonded fibers on the mandrel by operat-ing under conditions such that fiber-to-fiber bonding 35 occurs during the first part of the formation of the -17- ~ ~5~

~ibrous structure, e.g., by minimizing the type and quantity of fiber cooling, following which the method in accordance with this invention is carried out under conditions such that fiber-to-fiber bonding is 5 substantially eliminated. The resulting structure h~s the internal support necessary to prevent col-lapse of the element under conventional operating pressures and has the added benefit that the portion of the structure which is self-bonded (the central 10 support member) has some filtering capability.
For very low pressure service, for example, in the rarlge of about 5 to about 25 psi (0.35 to 1.76 kg/cm2), the cylindrical depth filters in accordance with this invention can be made directly on a smooth ]5 mandrel and used without a core. It is, of course, also possible to make a coreless cylindrical filter structure and subsequently incorporate a core or central support member therein.
The preferred fibrous structures prepared by the 20 method in accordance with the subject invention are comprised of a fibrous mass of nonwoven, synthetic, polymeric microfibers which are substantially free of fiber-to-fiber bonding, secured to each other by mechanical entanglement or intertwining, and wherein 25 the fibrous mass has a substantially constant voids volume, typically in the range of from about 60 to about 95 percent, more preferably from about 64 to about 93 percent and even more preferably from about 75 to about 85 percent. When polypropylene is used 30 as the resin, the most preferred voids volume is about 82 percent. Typically, the annular thickness of the cylindrical fibrous structures in accordance with this invention, particularly when used as depth filters, is in the range of from 0.4 to 1 inches (1.0 - 35 to 2.5 cm), preferably in the range of 0.5 to 0.8 .25~

inches (1.3 to 2.0 cm), and more preEerably in the r~nge of 0.6 to 0.7 inches (1.5 to 1.8 cm). ~s will become more evident from the following Examples, the combination of these characteristics in the cylin-5 drical filter structures in accordance with thisinvention result in high filter efficiency and en-hanced dirt capacity or life.
Polyrneric materials particularly well suited for use in accordance with this invention are thermo-10 plastics such as the polyolefins, particularly poly-propylene and polymethylpentene, polyamides, par-ticularly nylon 6, nylon 610, nylon 10, nylon ll, nylon 12, and polyesters, particularl~ polybutylene tereph-thalate and polyethylene terephthalate. Other ]5 suitable, but less preferxed, polymers are addition polymers such as polyvinyl fluoride, polyvinylidene fluoride and their copolymers, and polycarbonates.
The method in accordance with this invention can also be applied to solutions of resins in appropriate 20 solvents, in which case temperatures can vary down to ambient or lower. In this mode the solvent must be at least largely evaporated before the fibers are collected to avoid fiber-to-fiber bonding.
Thermoset resins in partially polymeri~ed form 25 can be fiberized but are not a preferred starting material as operation with them is more comple~.
The fiber diameters can be varied from about 1.5 micrometers or less up to about 20 micrometers or more. However, when the product is made in the pre-30 ferred voids volume range of 75 to ~5 percent, fiberdiameters above about 20 micrometers make elements so coarse as to have little use for filtration applica-tions.
Fiber aspect ratios are large. e.g., l,000 or 35 hlgher. Indeed, it is very difficult even by micro--19- ~5~6~

:`
- scopic examination, to determine length to diameter ratios as fiber ends are difficult to ind.
: Various additives, such as activated carbon, ion exchange resins, and the like, can be incorporated 5into the cylindrical fibxous structures in accordance with this invention by, for example, feeding them ; into the stream o fibers prior to laydown. ~lso, the cylindrical fibrous structures in accordance with this invention can be formed in any desired length.
10 The cylindrical fibrous structures can be further processed byr for example, the application of an external support and the incorporation oE end caps shaped so as to fit within the particular filter ~' assembly in which the resulting filter element is to ]5 be used.
The term "substantially free of fiber-to-fiber bonding" r as used hereinr refers to the characteris-tics of the microfibers ma~ing up the fibrous mass portion of the cylindrical fibrous structures in ~ 20 accordance with this invention. The microfibers are !' mechanically entangled or intertwined. It is this mechanical entanglement which provides the structural integrity oE the fibrous mass portion of the struc-ture. Whrn examined under a microscope at lOx to 25 lOOx the fibrous portion of the filter structure may display random fiber-to-fiber bonding but such bond-ing is in an amount that would not be significantly detrimental to filter Eunc-tion nor contribute in any material way to the structural integrity of the fil-30 ter. Additionally, it is possible, by the use oftweezersr to separate out fibers which have clean, smooth proEiles, free of protuberances and oE unsep-arable clumps of fibers of the type which typically appear on fibers in structures containing substantial 35 fiber-to-fiber bonding.

-20~ ~2 5~L~3 The term "substantially constant voids volume", as u~l herein, means the average voids volume o the fibrous mass portion of the cylindrical filter struc-ture varies by no more than about 1 to 2 percent.
5 Voids volume determinations or, alternatively, den-sities, were carried out by use of a series of 5 U-sh~ped gauges. Diameters oE the gauges were selected such that the difference between each successive gauge represented one-fifth of the total volume of 10 the collected fiber on the cylindrical filter struc-ture as it was being formed. With the polymeric material or resin delivered at a constant rate, the time required to reach the diameter of each gauge was recorded. This procedure was repeated as ten suc-]5 cessive filter cylinders were prepared under the same conditions and the times averaged. The percent voids - volume determined by this procedure was found accur-ate to within about 2 percent, and in the case of finer fibers, to within about 1 percent.
As may be seen in Figure 8 and in Table II be-low, filter elements each having a constant voids volume of 82 percent and a constant fiber diameter throughout (but varying from element to element rom 1.9 up to 12.6 mlcrometers) provided removal ratings 25 varying from less than 1 micrometer, e.g., 0.5 micro-meters or even less r up to over 40 micrometers.
One configuration which is useful because it provides prefiltration for a very wide range of final filters is made using a program for forming roll 30 pressure, resin rate, fiberizing air rate, and cool-ing water Elow, which produces a constant density element with fibers varying in diameter from about 1.9 micrometer at the id (downstream) to about 12.6 micrometers at the od (upstream). The manner in 35 which the fiber diameters are profiled can be varied ii, ; ~

. .

-21~ 3 widel~, for e~ample, for some applications a higher proportion o~ Eine ~ibers could be used while for others more coarse fibers might be preEerred. For general prefilter service~ filter elements have beer-5 made in which the diameters of the fibers ~orm ageometric progression. In such a construction, if the element is divided into N cylindrical portions, each containing the same weight (and volume) of fi-bers, then the fiber diameter of each portion is 10 larger than that of the adjacent downstream portion by the factor F, where F = (12.6) N-l 1 . 9 ]5 For example, if the number of sections were 20 (N =
20), F would be 1.105.
Another configuration, which is desirable be-cause it combines absolute filtration with prefiltra-tion~ is one in which the downstream portion of the 20 filter is made using constant fiber diameter, while the upstream portion is profiled from the fiber dia-meter of the downstream portion up to a larger dia-meter. The constant fiber diameter downstream por-tion of the filter element may comprise ~rom about 20 25 to about 80 percen-t of the total volume of the fib-rous filter mass and, correspondingly, the upstream profiled portion of -the filter element may constitute from about 80 to about 20 volume percent. One pre-ferred configuration is a filter element in which 30 about the first 50 percent by volume of the element has a constant fiber diameter of 1.9 microme-ters (downstream portion) and the upstream portion has a graded fiber diameter structure, i.e., it is pro-filed, with the fiber diameter ranging from 1.9 mi-35 crometers up to 8 to 12 misrometers. Another con--22- ~ S~

: figuration is as that immediately above but the con-stant Eiher diameter portion has fibers with diameters of 8 micrometers an~l in the graded fiber diameter portion they range f~om 8 up to 12 to 16 mic~ometers.
Still another desirable configuration is one in which the downstream portion of the filter is made using a constant voids volume, with a constant fiber diameter, and the upstream portion has both a pro-filed Eiber diameter and a profiled voids volume.
Filter elements have been made using fibers as small as 1.6 micrometers in diameter. Still finer fibers could be used but are not preferred because production rates become progressively lower and col-lection oE the fibers becomes more diEficult, with a ; ]5 larger proportion not collected on any o~ the working parts of the apparatus. Other elernents have been made with fibers as coarse as 16 micrometers but such elements have removal ratings so large as to have limited practical application or, if made quite dense, 20 relatively low dirt capacity. Filter elements made with their downstream portions composed of fibers of 13 micrometers or smaller can, however, for some applications, beneEit from upstream layers profiled up to 16 micrometers or even higher.
Filter elements in accordance with this inven-tion have been made using polyprop~lene resin with voids volumes varying from 64 to 93 percent. Voids volumes above about 85 to 88 percent are not pre-ferred because they are deformed by relatively low 30 differential pressure, for example, as low as 5 to 10 p~si (0.35 to 0.7 kg/cm2), with consequent change of pore diameter. Voids volumes below about 75 percent are not generally preferred for use in the operating range of most filters, which is up to about 40 to 60 35 psi (2.81 to 4.22 kg/cm2) differential pressure, -23~ 3 because filter life decreases as the voids volume is decreased. One exception is the use of lower voids volumes w}len making filters with fiber diameters at the low end of the practical fiber diameter range, 5 for example, 1~6 to 2.0 micrometers. Such filters remove finer particles than would be obtained using higher voids volume and are useful for that reason.
For filter operations with filter elements made from fibers having diameters from above about 1.6 to 2.0 10 micrometers and opera~ing at differential pressures of up to a~ou~ 40 to 60 psi (2.81 to 4.22 Xg/cm2), the preferred voids volume range is 78 to 85 percent.
For applications in which differential pressures exceed 60 psi (~.22 kg/cm2), and up to several hun-]5 dred psi (14 ~g/cm2 or higher), lower voids volumesdown to 60 percent or even less may be needed to prevent collapse under pressure. Somewhat higher voids volumes can be used with filter elements pre-pared from resin materials of relatively higher mod-20 ulus, such as nylon 6, which have better resistanceto deformation compaxed with polypropylene.

EEficiency, Removal Rating ~nd Dirt Capacity (Life):

These characteristics were determined for 2.5 od x 1.1 id x 10 inch (6.35 x 2.8 x 25.4 cm) long ele-ments using a modified version of the F2 test devel-oped in the 1970s at Oklahoma State University. In this test a suspension of an artificial contaminant 30 in an appropriate test fluid is passed through the test filter while continuously sampling the fluid upstream and downstream of the filtex under test.
The samples are analyzed by automatic particle count-ers for their contents of five or more different 35 preselected particle diameters and the ratio of the -2~

upstream count to downstream count is automatically recorded. T~lis ratio, known in the industry as the beta (~) ratio, provides the xemoval efficiency at each oE the preselected particle diameters.
The beta ratio for each oE the five or more dia-meters tested is plotted as the ordinate against particle diameter as the abscissa, usually on a graph in which the ordinate is a logarithmic scale and the abscissa is a log2 scale. A smooth curve is then 10 drawn between the points. The beta ratio for any diameter within the range tested can then be read from this curve. ~fficiency at a particular particle diameter is calculated from the beta ratio by the formula:
]5 Efficiency, percent = 100 (1 -l/beta).

As an example, if beta = 1000, efficiency = 99.9 percent.
~nless otherwise stated, the removal ratings cited in the examples presented below are the parti-cle diameters at which beta equals 1,000 and the efficiency is 99.9 percent.
E-fficiencies in the range of from 1 to about 20 25 to 25 micrometers were determined using as the test contaminant a suspension of AC fine test dust, a natural silicious dust supplied by the AC Spark Plug Company. Prior to use, a suspension of the dust in water was mixed until the dispersion was stable.
30 Test flow rate was 10 liters/minute of the aqueous suspension. This same procedure was applied to fil-ters having efficiencies oE less than 1 micrometer by determining eEficiencies at usually 1, 1.2, 1.5, 2, 2.5 and 3 micrometers and extrapolating the data to 35 under 1 micrometer.

-25~ 86~3 ~ fficiencies above about 20 micrometers were determined using Potter's Industries Incorporated fl3000 spherical glass beads suspended in MIL-H-5606 hydraulic fluid. These glass beads have a dis-tribu-5 tion of si~es ranging from less than 15 micrometersup to 50 ~o 55 micrometers and higher. The viscosity of this fluid is approximately 12 centipoise at the test temperature of 100 degrees F (37.8 degrees C).
Test flow rate was 20 liters per minute. The higher 10 viscosity and flow rate serve to keep beads up to about 100 micrometers in diameter in suspension.
Filters in the 20 to 25 micrometer range were often tested by both methods. The resulting effici-ency and dirt capacity data were usually comparable.
]5 In both the a~ueous and oil based tests, pres-sure drop across the test filters was measured as the test suspension Elowed through the filter and was recorded as a function of time. The quantity of contaminant incident on the filter required to devel-20 op a differential pressure of 60 psi (4.2 kg~cm2) is recorded as the dirt capacity or "life" of the test element.
It is characteristic of depth filters, particu-larly in -the coarser grades, that efficiency tends to 25 be reduced at large differential pressure. Since filters are rarely exposed to differential pressures as high as 60 psi (4.2 kg/cm2), efficiency data are reported as an average of about the initial two-thirds of the total life of the filter.
As noted above, data reported as less than 1 micrometer are obtained by extrapolation. In order to provide assurance that the extrapolated data were reasonably near to correct, or at least conservative, a number of the filter elements with high efficien-35 cies at under 1 micrometer were further tested by , -26- 1~5~

passing suspensions of bacteria of known dimensions thrQug}l them. The upstream and downstream bacteria concentrations were used to calculate eEficiency. In all cases the efficiencies so deterrnined either con-; 5 firmed the extrapolated ~2 test data or indicated a still higher efEiciency.
: The test in the finer ranges using the ~C dust described above showed significant and reproducible beta ratios as high as 100,000 to 1,000,000 and 10 there~ore permitted measurement of efficiencies of up ; to and over 99.999 percent while the smaller number of glass beads permi-tted computation of efficiencies up to about 99.99 percent at up to about 40 micro-meters and to successively lower efficiencies at ]5 larger diameters.

Filtration Testing Using Bacteria:

Filtration of suspensions of bacteria of known j20 size is a very useEul high sensitivity method for determining Eilter efficiency. This test method is :particularly appropriate for application to filters made using finer fibers and moderate to high density because bacteria removal is one of the important 25 prospective applications for the finer grades of filters in accordance with this invention.
Bacteria removal tests were run in the following manner:
(a) A suspension in water of a pure strain of a 30 bacterium of known dimensions was prepared at a con-centration of about 101 to 5 x 1012 organisms per liter.
(b) The filter element was placed in an appro-priate housing and 1 liter of the bacteria suspension 35 passed through the element a-t a rate of 0.5 to 1 ' ~
' ~ - ;:. .

-~7- ~ 6~

litex per minute.
(c) ~liquots of the effluent from the filter were collected and diluted with sterile water to 10, 100, 1,000, etcetera, fold. Each such diluted ali-5 quot was then cultured in a Petri dish in an appro-priate growth medium. ~ach bacterium present de-veloped, within 24 to 48 hours, into a colony of bacteria large enough to be seen at low magnification using a microscope. The number of colonies in some 10 dilutions was so great that the colonies could not be counted, while in others there were too few to be statistically significant. Ilowever, there was always t at least one dilution providing a useful count, from which the total number of bacteria in the effluent . ]5 could be calculated. Knowing the influent count and the effluent count, efficiency can be calculated.
The bacteria used in developing this invention included Pseudomonas diminuta (Ps.d) and Serratia marcescens (Serr. m.), the dimensions of which are 20 respectively 0.3 micrometer diameter x 0.6-0.8 mi-; crometer long and 0.5 micrometer diameter x 0~8 micrometer long.
The in-vention will be better understood by re-ference to -the following Examples, which are offered 25 by way of illustration.

- , . , -~B~ 3 ~x~mple 1: Preparation Of ~ Cylindrical Filter Struct~re Of Uniform ~iber Size And Uniform Voids VQlume (Ungraded):
_ _ _ The apparatus described above was used to pre-pare a supported cylindrical filter structure with a us~ble central section 36 inches long (91.4 cm long~.
The fiberi~ing die length was 6-1/4 inches (15.9 cm), the str~ke of the reciprocating mandrel was 43-3/4 10 inches (111.1 cm), the mandrel rotation rate was 150 rpm, the axial translation rate was 500 inches (1270 cm) per minute, and the die to collector distance (DCD) was 12-1/4 inches (31.1 cm). The mandrel was fitted with three hollow foraminous (latticed) filter cores, each 1.1 inches (2~.8 cm) inside diameter ~id), 1.3 inches (3.3 cm) outside diameter (OD) by 9.B
inches (24.9 cm) long, of the type illustrated in Figure 3. Polypropylene resin having a melt flow index of 30 to 35, was heated to 720 degrees F ~382 degrees C) and the extruder rpm adjusted so as to give a total resin flow rate of 1.83 grams per second through the spaced nozzles, each having a gas stream surrounding the resin extrusion capillary, at a resin pressure of 625 psi (43.9 kilograms per cm2). (Poly-propylene resin having a melt flow index of 30 to 35was the resin used in all the Examples herein ~nless otherwise noted.) ~ fiberizing air pressure of 4 psi (0.28 kilograms per cm2) was used. The average fiber diameter produced under these conditions had pre-viously been determined to be 12.5 micrometers. Themicrofibers having 12.5 micrometer diameters so pro-duced were directed onto the air cooled Eorming roll which was biased towards the mandrel by an air cylin-der pressurized to 8 psi (0.56 ky/cm2) and were thence transferred to the filter cores on the rotating/

_~9_ ~ ~ 5~

reciprocating mandrel. Fiberizing ~nd collection were continued until the od of the fibrous cylindri-cal filter structure (sometimes referred to herein as a '`filter cylindern~ a "filter element" or simply as an "element~J reached 2.5 inches (6.35 cm)~
The average density of the fibrous porti~n of the cylindrical filter structure of this Example was such as to yield a voids volume of about 81 percent (voids volume, in percent, equals 100(1-D/d), where D
equals the apparent density and d equals the density of the resin, which is 0.9 grams per cubic centimeter for the polypropylene used).
The central section of the cylindrical filter structure was cut into three sections to make three filter cylinders, each 9 ~ inches (24.9 cm) long and each having a corresponding 9.8 inches ~24~9 cm) long filter core on the interior thereof. The respective voids volumes of the three filter cylinders were equal within the measurement error; each had a voids volume o~ 81.2 percent.
The three filter cylinders denoted A through C
below were assembled into housings which provided appropriate end sealing means and were tested using the F2 test method described above, yielding the results listed below:

Filter Life or Removal Rating, Dirt Capacity (micrometersj (qrams) -30- 1~ 5~

Microscopic examination of the elements was performed. With the exception of a limited number of small l~calized areas in which some undesirable fiber softening had occurred, the individual fibers could be pulled out of the mass using tweezers with no evidence of aahesion to neighboring fibers, i.e., the pr~files ~f the fibers were smooth with no protuber-ances inaicating fiber-to-fiber bonding.
' It should be noted that in this Example, as in the following Examples where fibers of 2.5 micrometers or larger were formed, water spray was used to pro-vide enhanced cooling of the fibers, thereby assist-ing in minimizing undesirable fiber-to-fiber bonding.
The water spray was appl-ied in the general manner illustrated in Figure 1 at application rates suffi-ciently high, in connect~on with other cooling tech-niques as descxibed, to provide structures substanti ally free of fiber-to-fiber bonding, e.~., ;n the range of about 80 to 140 cubic centimeters per min-ute.

Example 2: Distribution Of Vo;dsVolume Within Cylindrical Filter Structures Made Using Constant Forminq Roll Pressure:
Using the apparatus described above and the general procedure described in Example 1, a series of elements, each having uniform fiber size~ were pre-pared directly on a 1.3 inch ~3.3 cm) od solid collec-tion mandrel. That is, the elements of this Examplediffered from those prepared in Example 1 in that they did not contain a central support member or core. The elements, denoted in Table I below as filters D through H, each had constant fiber dia-meters but from filter to filter the fiber diameter L 9~ 3~
~31-varied from 12.S micrometers down to 2 5 micrometers, as set out in Table I. A series of five ~U-shapedN
gauges were prepared with the d.iameter of the first gauge such that t}-e volume of fiber collected re-presented one-fifth of the total volume of fiber collected between 1.3 inches (3.~ cm), the id of the formed filter cylinder, and the 2.5 inch ~6.35 cm) od of the finished filter cylinder or element. Similar-ly, the difEerence in diameter between the second and first gauges - the second having a larger diameter than the first - represented one-fifth of the volume of fiber collected between 1.3 inches (3.3 cm) and 2.5 inches (6.35 cm). In li~e manner, the difference in diameter between the.second and third gauges re-15 presented one-fifth of the volume of fiber collected between 1.3 inches (3 3 ~m) and 2.5 inches ~6.35 cm), etcetera, up through the fifth gauge. Resin flow rate was held constant and, as the filter cylinder diameter increased or built up during formation of 20 each of the elements D through H, the time required to reach the diameter of each of the five gauges was recorded. These times were then used to determine the percent voids volume of each of the five sections of each of the elements, with the results shown in 25 Table I bel~w:

. . _ ,' ' ' ~.

- 3 2~ , t~

h o o r~) t`J 0 .
~J O r~
O ~ (~ CO Cl:l ~0 O
~ ~ C~

CO CO
. ~ . . .
C r~ O _1 O ~ ~
_~ . ., Vv O t` In O
C ~ ~

~ h V C
aJ c ~ ~r ~ ~r ~ ~ CO ~
a) P~ J

a U~
E i~
~ al u~ O co u- u ,1 V
a h t) _~
h E

a ~

3 ~ 5~t~jL~ ~

~ he measurement of voids volume within the fibrous mass (or, in effect, the average density, since the density of the fibers is a constant, i e., 0.9 qrams per cc) by the method described above is not precise. It i5 believed that the voids volume in the ~ilter elements is uniform or nearly so through-out the thickness of the filter cylinder ard that any errors caused by using the averages of these voids volumes is small. Thus, in those Examples reported which are made using a range of fiber diameters on a single element by adjusting the orming roll pres-sure, the voids volume is believed to be constant throughout the fibrous mass within about l to 2 per-cent.

Examples 3 through 12 Preparation Of A
Constant Voids Volume Cylindrical Filter Structure Profiled To Provide A Wide Range f Pore Diameters By Varying_Fiber_Diameter:
Examples 3 through 9 below demonstrate the preparation of constant or near constant voids volume filters with removal ratings varying from less than 1 micrometer up to 40 micrometers. Examples lO, 11 and 12 below show how the data generated in Examples 3 through 9 can be used to prepare graded fiber dia-meter, constant voids volume filter elements.

Step 1:
A series of supported cylindrical filter struc-tures or elements (Examples 3-9), each with constant or near constant voids volume of 82 + 1 percent and a uniform fiber diameter, was prepared. While the fiber diameter within an individual element was con--3~

stant over the serle~ oE seven element~ formed, the fiber diameters ran~ed from 1.9 to 12.6 micrometers as set out in Table II below. As also set out in Table II below, the filter elements provided removal ratings in the range o~ from less than 1 up to 40 micrometers.
These ~lements were prepared using the general procedure and apparatus of Example 1 but the resin pressure, fiberizing air pressure and the forming roll air pressure were varied in order to obtain fiber diameters spanning the range from 1.9 to 12.6 micrometers as noted in Table II. The average voids volume of each test element was held as closely as possible to 82 percent, the average deviation being less than 0.4 percent. Conditions were controlled to substantially eliminate fiber-to-fiber bonding in the formed elements by the methods described above; most importantly, the fibers were collected on the forming roll rather than the mandrel, and water spray was used when iber diame~er was 2.5 micrometers or greater. Each element was ~2 tested and the removal rating (diameter of particles in the incident fluid at which the removal efficiency equalled 99.9 per-cent) and the dirt capacit~ (or life) were deter-mined. The preparation conditions and the test re-sults obtained are shown in Table II below. Fi~ures

4, 5, 6, 7 and 8 graphically show the relationship of the important parameters o~ Table II.

~35--U) ' V ~ ~ CO ~ o ~ o a c~ ~ o ~1 ~ ~D 0 0 o ~

~ ~ I U~
:~ ~ O ~J
o a) ~ ~ ~ r` o o o ~; E E v ~ ~--I 0 o I U~
O -v o Va~ ~ ~ o ~0 a E E E
Cll .
C C
v~ a, ~ O ~ O ~ ~ ~ ~ d' V C ~ ~
E O (L~ 4 al ~ 1 co o ~ o ,- ~ ~ ~, 1 ~
E _ _ _ ._ _ _ _ E ~ Q~ ~ ~ ~ ~ o~
w a Q~ , . u~ ~ o u~ In ~n In u~
.rl V ~ D~ ~ O O ~ N N N N

O O r:E3 ~n ~ 0 ~n J- C~ X _ _ ~ ~ ~ o o 4-~
. . o o o o ~o ~ In .~
~N
O E ~ ~ ~ ~ 'n o ~D
aJ ~ ~n ~ ~ ~r o r~ m D,-- ~ ~ ~ _l ~ o o O ~ u~ o In O O U~ O a:
~ ~ ~ u~
a) R. QJ

E~
r~ E ~ ~ n ~ 0 O~
WZ

_t~e_~:

Figures 4 through 7 can be used used to prepare an operating plan which will make ilter ~lements wit~ any combin~tion of fiber diameters between 1 9 and 12.6 micrometers.
In general, it i~ preferred to construct ele-ments in which the liquid beinq filtered will flow from the outside of the elements toward the inside and then exit through the filter core (an outside/in configuration). However, in some circumstances, for example, when it is desired to retain the collected solids within the ~ilter cartridge, the direction can be reversed (inside/out configuration). In either case, it is generally advantageous to have the pores profiled from large at the upstream side to small at the downstream side by providing fibers of aecreasing diameter in a graded or profiled manner in the direc-tion of fluid flow, i.e., in the radial direction, while maintaining th~ voids volume substantially constant. ~
The configuration o~ the profile can vary wide-ly. In some applications ~t ma~ be desirable to have the upstream portion of the filter graded with the downstream portion of uniform pore size. Alterna-tively, especially if intended for use as a prefil-ter, the entire thickness of the fibrous portion of the filter can be varied in an appropriate profile with the largest pores upstream to the smallest pores downstream~ Example 10 illustrates a filter of the latter type in which progressively larger fiber dia-meters are used as the filter is built up. In Ex-ample 10, the fiber diameters are varied as a ~eo metric progression. Varying the fiber diameters as a geometric progression i5 believed to provide a 37 ~ 5 ~ L~

Eilter;ng element well adapted to a wide variety of non-speci~ic applications. For any specific applica-tions, other schemes can be used, for example, linear, square root, or logarithmic, etcetera. Alternatively,

5 the fi~er diameters can be graded in a continuous manner without discrete steps in the radial direction, a form of gradation referred to herein as "continuous-ly profiled".

Example 10: Element Made With Constant Voids Volume And Varyinq Fiber Diameter Throuqhout Using the general procedure and apparatus of Example 1, the data o~ Figures ~ through 7 (generated in Examples 3 through 9) were used as follows:
(a) the total volume of the fibrous portion of the filter element to be formed (2.5 inches (6.~5 cm) od x l.30 inches (3.30 cm) ia x 9 . 8 inches ~24.9 cm) long) was divided into 15 equal incremental volumes;
(b) the fiber ~iameter range from l.9 to 12~6 micrometers was then divide~ into 14 steps of in-creasing fiber diameter, each fiber diameter being 14.447 percent larger than the preceding one, the first being 1.9 micrometers and the last 12.6 micro-meters (this forming a geometric progression of fi~er diameters as set out in Table III below);
(c) the operating conditions required to ob-tain the 15 flber diameters set out in Table III
below were then read from Figures 4, 5 and 6.
In this Example, the filter is designed to have an equal incremental volume of the fibrous mass at each of the selected fiber diameters~ Because the voids volume is constant within experimental error at - 82 percent and the density is therefore correspondingly constant, it was required that an equal weight of the -3~-- ~S~

microf;bers be deposited in each of the 15 incrernental volumes. Since the resin x~t~ i~ a function of the fiber di~meter, Figure 7 was used to calculate the time required to deposit an equal weight in each of the increment~l volumes. The result was the operating program set out in Table III below.

N ~ ~D ~1 o ~ Cl) r` ~ ~ ~ ~ ~ ~ a~
0 3 ~ u~l ~ ~1 ~ a~ ~D ~ ~`~ ~ ~ CO 1~ U~ ur~ ~r U1 U ........... o ~ ~ ~
~ ~ ~ ~ ,~ ~ o o o o O O
CJ` ~ ~
C h X
E u~
~ h`~
O ~ o a~ u~ O r~ ~ _ 0 ~ ~cr ~ O
C4 ~s a u ~ ~ ~ ~ ~ ~ ~ ~ .t ~ ~

a)-- __~___________~
h ~ ~ ~r N ~ O ~D t~ ol ~ ~ ~ ~ I` (~) u^l CU~ ...............
~ ~ oooooC~
N ~J _ _ _ _ _ _, _ _ _ _ _ _ _ _ _ ._1 h X
h fl, _ ~1) It~ 1~ Lr) . . .
Ul o ~ o o q~ r ~ ~ ~ o a~
~: a~ u~ ~O ~O ~n E ~ CD _ a~
O
h ~ ~ ~ `~ ~ ~ ~r ~ ~r Yr ~r ~r ~r ~r ~ ~r ~ ~ ~ _ _ _ ~
c ~n--W . o o o u) u~
m ~ O _ _.
J-E ~ ~ ~1~ ~1 ~ t~ ~`
~ ~n a) c~ ~
E C
,~ o E~ V
aJ

h h a aJ ~
O o~D O ~D
E h ........ ~ ......
I 1 _I

' 3 ~ 5 t~ L~ ~ ~

The 9.9 inch (24.9 cm) long filter ele~ent, prepared a5 described above, on a l.l inch (2~79 cm1 ia x 1.3 inch (3.30 cm~ od x 9.~ lnch (24.9 cm~ long core, had an od of 2-l/2 inches 56.35 cm~, i.e., the 5 fibrous mass had an inside diameter of l.~ inches (3.30 cm) and an outside diameter of 2-l/2 inches (6.35 cm). The filter element exhibited the fol-lowing properties:
Clean pressure drop was l.8 psi (0.13 kg/cm2) at lO liters o water per rninut~. The life or dirt capacity was 83 grams to 60 psi (4.22 kg/cm21 dif-ferential pressure filtration efficiency was in excess of 90 percent at l.0 micrometer, 99 percent at 3.7 micrometers, 99.9 percent at 5 micrometers and 99.99 percent at 5.6 micrometers; bacteria removal efficiency tested using 0.~ micrometer diameter Pseu-domonas diminuta (Ps. d~ organism wa~ 99.997 percent.
Because of its very high dirt capacity (long life) and removal capability over the full range of particle diameters f~om O.l to 40 micrometers, this type oE filter i5 particu ~ rly well suited as a pre-fllter. For example, it could b~ used to precede an absolute rated final filter when used in critical applications, ~uch as ~terilization of parenterals or for providing water for use in the manufacture o microelectronic devices. Because of it~ wide range capability, it would also serve well as a prefilter for a coarser after filter, ~or example~ one rated at 5 or lO micrometers. It can al50 be used as the only filter in the system for many other applications.

..... _ _ .
-~5 Ex~mple 11: Filter Element With Constant ~iber Di~meter Of 1.~ Micrometers For The Inner 50 Percent Of The Fibrous Portion ~ The Element And Varying Fiber Diameter For The Outer 50 Percent, With Constant Voids Volume Thro qhout:

The filter element of this Example was prepared - in the same general manner a~ that of Example 10 and, as with the filter element of Example 10, had a voids volume of 82 percent but differed from Example 10 in that the initial 50 percent by weight of the fibrous portion of the element was made up of fibers having a constant diameter of 1.9 micrometers with the halance varied, again as a geometric progression~ from 1.9 to 12.6 micrometers. This was accomplished by the oper-ating program set out in Table IV below.

. _ . _ _ . .

~, ~ ~ _ _ _ ~ _ _ _ ~ ~ ~ _ _ _ _ ~
O ::1 E u~ ~ n ~
a~ \ _ ~ _ _ _ _ _ _ _ _ _ _ _ _ _ ~, _, n O -~ U~ o ~ u~ o ~ ~ o ~ ~ O ~ ~ I~
r~ ~ Q ~n ~r ~ r~ ~ ~ ~ ~ _l ~ ~ ~

o -- _ _ _ _ _ _ _ _ _ ,_ _ _ _ _ _ ~ ~ ~ r~l ~ o ~D c~ ~ ~ 0 ~ n E u~ n r~
...............
~ ~ _ _ _ _ _ _ _ _ _ _ ~_ _ _ _ _ In n n o ~ o o er un ~ ~ r ~ o a~ n ~ ~ Q u) ~ \D ~D In ~r ~ ~ ~ ~ _~ ~

H L u7 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ r~ _1 u) ~ u) a~ ~ o o o u~ m un In In un u- u~ m u n In r~ a) h u~ O _l O s~ N ~ 1 N N N N 1~ N t`J N

V
t~ ~ ~ N O ~ I` ~ ~ o a:~ )~ un _~ ~ a~ o ~ ~ ~ ~7 ~r m ~ ~ o ~ ~ ~ ~1~ ~ ~ r~
Q) tJ ~
E c -_~ o ~ u~
v C

~ a v E
~ O ~ ~ un ~ ~ r r~ r ~ o ~D
f7 ~ O ,~ P U7 ~ 1~ ~ C~
~ n :~ ~ ,~

The resultinq element had the same idt od and length as that of Example 10. It exhibited the fol-lowing properties:
Clean pressure ~rop: 4.3 psi (0.30 kg/cm2~ at a S test flow rate of 1~ liters of water per minute Life or di~t capacity: 36 grams to 60 psi (4.22 kg/cm2) dif~erential pressure;
Filtration efficiency: in excess of 99 percent at 0.7 micrometer (as estimated by extrapolation), measured as 99.9 percent at 1.4 micrometers, as 99.99 percent at 2.2 micrometers, and as 99.999 percent at 3 micrometers.
Because of its very high efficiency at 2.2 micrometers, this filter element can, for nearly all purposes, be rated as 2.2 micrometers absolute. It also provides very useful levels of removal for par-ticles as fine as 0.7 micrometer. These high effi-ciencies, coupled with the very hiyh dirt capacity of 36 grams under the F2 test as described above, pro-vides a highly usef~l~filter with a long servicelie. ~
It should also be noted that while the efficl-ency o~ this filter element is f;ner than that of Example 4, its service life (dirt capacity) is over 4 tirnes higher and, indeed r is equal to the service life of a uniform pore filter with a removal rating of about 20 micrometers at 99.9 percent efficiency.
In order to further characterize this ultra-fine, long lived filter element~ an element made in a similar manner was tested by passing through it a suspension of Pseudomonas diminuta bacteria. This organism is cylindrical in shape, with a diameter of 0.3 micrometer. ~fficiency of removal was 99.997 percent.
Elements made in th~ manner as described above -44~ 5 3 are well ~u~ted for th~ ~iltr~t~on o a var~ty o product~ from which yea~t and bacteria are to be removed, yielding not only a liqu~d e~fluent free of or greatly reduced in its content of yeast and bac teria, but also one with high clarity.

Example 12: Filter Element With Constant Fiber Diameter Of ~.9 Micrometers For The Inner 59 Percent Of The Fibrou~ Mas-~, And Varying Fiber Diameters For The outer 41 Percent~ With Constant Voids Volume Of 82 Percent Throuqhout:

The filter element of this Example was prepared in the general manner of Example 10. However, the initial 59 percent by weight of the fibrou~ portion of the supported filter element of thi~ Example had fibers with diameters of 2.9 micrometers with the balance varîed as a geometric progr~ssion from 2.9 to 12.6 micrometer~. This wa~ accomplished by the oper-ating program set out in ~able V below~

~, 5 ~ ~ j L~ ~3 _~ ~ ~ o a~ ~ t--~ c~ ~ ~ ~ ~D
O ~I E ~ C~
............
.Ioooooo ~ aJ tl _ _ ~_ _ _ _ _ _ _ _ _ _ C ~X
E
~ 1 O ~ ~n o ~ D ~ ~ o a~
1:4 ~ Q r~

~__ ______~______ ~ ~ o ~ c~ a~
t7~ ~ E ~ cs7 ~ _ ~D ~ ~ c~ r~ ~D In r~
C ~n t) ............
~ ~oooooo N a) t~ _ _ _ _ _ _ _ _ _ _ _ _ -,1 u~ In Ll~
~ 1 . . .
.,1 _1 U~ O ~ q- ~ o a~ r~ u~ ' ' O
:~ )~ t~
~ _________.__~_ ~ C u~--m u7 ~ -~S a) )~ v~ ~3 N N ~ N N N (`i ~1 N N N
E~p:; ~ O
a~ ~ . .
.~ ~ I
0 ~ ~ O cr~ ~ ~D ~ ~ ~
~r ~n ~ r a~ co cr~ o ~ ~ r1 E ~ ~t ~ ~1 ~ ~ ,J ~ ~ N t`J N
V~
C
._1 o E~ ~ t) 1~ a U~

~.

a S~ V
Q) a) E
O . . . ~ a~ o o ~D
~ n :E

~46~

The ilter element o~ thi~ Example was prepared by the g~neral proceaure described ~n Example 11 above and had the same id, od and length as the ele-ment of E~ample 11. It exhibited the following pro-perties:
Clean pressure drop: 1,5 psi (0.11 kg/cm~) at atest flow .rate o~ 10 liters per minute of water;
Life or dirt capacity: 53 grams;
Filtration efficiency: 90 percent at 1.1 micro-meters, 99.9 percent at 4.6 micrometers and 9g.99percent at 5.8 micrometers.
Filters of the type of this Example have ap-plications in.fielcls such as filtration of magnetic particle suspensions used for video recording tape manufacture and for processing photographic film emulsions.

Example 13: Filter With Fiber Diameters Varyin~ From 8.5 to 12 S Micrometers-Using a procedure ~i~ilar to that of Example 10 but starting with 8 5 micrometer aiameter fibers, a filter element was prepared which had the following properties:
Life or dirt capacity: ~15 grams up to a pres-sure drop o~ 0.6 p~i (0.042 kg/cm2) with a removal rating of 24 micrometers. The 115 gram life to 0.6 psi ~0.042 kg/cm2) i~ very much higher wh~n compared with the best commercially available ilters o equal removal rating which use a varying density ~tructure, as opposed to a sub~tantially uniform void~ volume and corresponding substantially uniform density with a varying or graded fiber diameter structure.

f 5~
.
Examples 14 throuqh 17:

~ number of filters have been described in the literature which ;eek to obtain increased life or dirt capacity using fibers of uniform diameter by varying ~he pore diameters ~rom larger upstream to smaller downstream by decreasing the voids volume, i;e., increasiny the density, of the filter medium in a progressive manner. The characteristics to be expected of such a filter can be projected using the test results obtained in the following Examples 14 through 17 The filter elements of this group of Examples were made in the same general manner as Example 1.
Each was prepared with uniform fiber diameter of 3.2 micrometers and with uniform voids volume in each element. However, the voids volume varied from one element to the next~ as noted in Table VI below.

.

L~ ~3 ~ .

.

~ C O ~ ~ ~ ~ CO
O -- ~ a.
E ~ t~ v ~1 ~ ~ ~
a) ~ ~ o E E

Vl a~ ~ ~ ~ ~D ~
h r` ~D ~ o t~

a.
a ~) O O ~ ~ c~
:>
__ ~ ~ _ _ _ O~ ~ ~D O
r~ ~ r~ 1--a) ~
O
, .,, U~
~t E ~v~
> ~ ~
O o ~u,o u~ ~n o a~
~ ~:
_.
a) ~
_ ~ _ _ O
C u~ ~ c~ ~ n~
N ~1) ~ ~`J ~`J ~'J ~ ,, _ _ _ _ O O O O

r~a ~ _ _ _ a L~ ~ u~
:~ ~C ~ '7 ~ t'7 U~ _ ~
C~> O O O
~IJ 1`~ 07 O O o O

a~

E
~ E
X ~
Z

3~3~

It may be deduced from the abo~e aata that lf one were to make a compo~ite filter, wlth voids vol-ume grad~ated from 79 to 88 percent, its life would be not better than 10.3 grams under the F2 test.
Furtherl its removal rating wo~ld be somewhere be-tween 1.7 and 3. a micrometers. These characteris-tics, when compared with the data of Examples 10-12, show that an element made with a constant voids vol- -ume, but with varying fiber diameter, would have at least four times the life or dirt capacity at equal efficiency.
The conclusion that far better life can be obtainea using constant voids volume with varying fiber diameter, as opposed to constant fiber diameter and varying voids volume ~or aensity), is also sup-ported by similar data (not presented herein) at fiber diameters other than 3.2 micrometers.

Examples 18 throuqh 21:

The same conclu~ions~regarding the infèriority of the approach of varying the void~ volume or den-sity to obtain a graaed pore structure vis-a-vis varying the fiber diameter while maintaining a sub-stantially uniform voids ~olume is also supported bythe subject Examples 18-21. The filter elements of Examples 18-20 were prepared on 1.3 inch (3.3 cm) od coresO In each of Examples 18~ 19 and 20, the fiber deposited between 1.3 inches (3.3 cm) diameter and 2.1 lnches (5.3 cm) diameter was identical, IOe., the fiber diameter used was 3;2 micrometers. The ~oids volume of this portion of all three Examples was 83 percent.
For the filter element of Example 18, the de-35 position of fibers was terminated at the 2.1 inches ~50--(5.3 cm) diameter~
The fllter element of Example 19 wa~ graded or profiled between 2.1 inches (5.3 cm) diameter and 2.5 inches (6.35 cm) diameter by varying the diameter of the fiber from 3.2 micrometer to 12.5 micrometer diameter in the manner of Examples 11 and 12 while maintaining the voids volume con~tant at 83 percent.
The filter element of Example 20 was profiled between 2.1 inches (5.3 cm~ diameter and 2.5 inches (6.35 cm) diameter by increasing the voids volume from B3 to over 90 percent while maintaining the fiber diameter constant at 3.2 micrometers.
As indicated in Table VII below, Example 21 was prepared in the same manner as Ex~mple 20 except that 15 the fiber diameter ~as 3.6 micrometers throughout.
The characteristics of the ~our elements of Examples 18-21 are set out in Table VII below~

;~0 ! :~

_. ~

-51- 1.25~
. .

U~
~ a~
~E
~ ~ O ~O C~ ~D ~
0-~ ~ . . . .
E~ 1~ Ur ~ (.~1 ~\1 ~r Q~ n', E

a) ~(:~:) cl:~ ~1 .
~ ~ ~r ~ ~

l ~ ~3 v ~ a~ E ~ o ~)a ~ o Q~
~ ~ o ~ a) a~ a~ ~ v a) ~ ~ v o IIn O aJ h U U~ V E E E ~ J v F ~ tr~ a) ~
n ro r~ E ~ _ ~J E ~ ~ h C h ~ ::~ O .C C E ~ ~ ~ O .C C E ~; h a) aJ ,~ o _ _ v 3 --~ ~ o ~ ~ ~
Q) ~ a u~~ v ~ E _l ~ ~I v ~ ~ o o ra~ ~ I ~ ~ O ~ c -~ ~ v V ~ ~ a a) ~ > ~ C ~ ~ a C ~
t.) O ut (~7 ~J C L: ~ ~ O E C_ c~ C ~1 tU E ~ ~ S ta E
n5 v ~ h l ~1 -1 a h ~-~ v O O ~ u~ u~ ~ ~ ~ O U~ -~ Ql ~ ~ ~ O
C C) ~O E E (V ~u h ~ ~ ~ O C a) Vl h ~ v ~ ~ ~ _ ~ O ~ ~ -I O O ~ ~ U ~ O O ~ ~ ~ C
H C ~ ~ a) 0 -1 ~ ~ ,C C O O 1~ 0 ~1 O ~ h a) ta - o~_~
~> ~ E V a E~ ~ ~ ~ :~: 3 ~ ~ ~ P~ ~ P~ ~ U :~:

m ~ ~ ~ ~ ~ ~ ~ h ~: -- a) a)-- a) a) -- a) a) ~
OP ~ ~ ~P ~ ~d~ ~ ~ dP ~ 1 o I ~ aJ Q) ~a) a) ~ a) ~ aJ a) _1 ~ ~ a:l ~ E cO E`E ~ E~ E a:~ E3 E
E--_ O r~l _ O ~_ O 1~ ~ O
Q~ ~ a ~ u ~ h--~ h -~ ~ ~ ~ h-~
E u n E ~ E~~ n E~ u U O ~1 ~ v h ~ç h ,~ 4,i ?,1 ,~
n3 v Ql )~ o u~ ~, o u~ :E 1~ o u~ h o Ul 5 h a~ u E ~ ~ a~ ~ ~ a~ u~
v V ~a .~ ~ .a ,, ~ ~,q ,. ~ ~-1 ~ r C ~ Q~ 0 -1 C O ~ ~ o -~ ~: O ~-~ C: O -_~
E ~ cl =) > r~ 1~ ~ ~ ~ :>
,_ E _ u~ u~ _ U ~ . .
1~ _ . U~ .
a) ~ In _ '~_ ~O
U) _ _ _ u~ E E ~:
c) _ u~ u~ n OQO~-t ~I~I ~ ~
~I) _ 5 ~ a~
E ~ ~ ~ o Z

In Examples 19 and 20, ~oiaS volume~ above 90 percent were not used because they were deemed too soft and compress;ble to be practical.
From the data set out in Table VII and by interpolating between Examples 20 and 21, it may be seen that this type of element, if made with a fi~er dia~eter such as to yield a re~oval rating o~ 2.8 micrometers, i.e.~ equal to that o~ Example 19~ would have a dirt c~pacity of only about 12 grams or about one-quarter that of Example 19.

Examples 22 tllrouqh 24:

The filter elemerts of these Examples were made in a manner similar to that of Example 12 except that the starting fiber diameters varied from 3,2 to 4.8 micrometers insteaa of the 2.9 micrometers of Example 12. l'he resulting structures had voids volumes of 82 percent and tne characteristics set out in Ta~le VIII
below:
~ .
TABL~ VIII

Exa~ple Starting ~ife Removal 25 Number Fiber Dia-Gr~ms Rating, meter, Mi- Micrometers crometers 22 3.2 52 9 23 4.0 67 1~
24 4.8 63 14 -53- ~ S~

Examples 25 through 28: Filter Element Series Illustr~ting EfEect Of Variation O Voids Volume:

A series of 2-1/2 inch (6.35 cm) od x 1~1 inch (2.79 cm~ id x 9.8 inch (24.9 cm) long filter ele-ments were prepared at conditions such as to produce 2.2 micrometer fibers. By varying the water spray rate and the forming roll pressure, voids volume was varied from 72.1 percent to 91.8 percent.
The properties of the resulting filters are set out in Table IX below:

r ~.

~;

. . .
O

t) ._~

U~
. ~ ~P
E ~ u~ 'r o O c~ o h cn .
tr~
~J

t,31 E~
~1 ~;
E~

al w E ~ ~
In ~o o u~ E ~ `
, ~`I o r~ ,~
O O
:~ ~ ~, Q~
_ IJ
E a u~ ~ r` CD
)~ ~
2:

` -55~

.
The dlrt cap~clty oE Examples 25 through 28 w~s plotted ~s the ordinate against the 99.99 perc~nt removal rating ~ the abscissa ana a llne dxawn throuqh the four experimental points. On thi~ line it was seen that the dirt capacity of a filter made in the rnanner of Examples 25-28 with 99 99 percent efficiency at 2.2 microme~ers would have a dirt capa-city of 5.9 grams. This i~ in markea contrast to the dirt capacity of the filter element of Example 11, which had a 99.99 percent efficiency at 2.2 micro-meters but a dirt capacity of 36 grams, i.e., six times greater.

Examples 29 through 34: Filtering Element Series Illustrating The Effect Of Variation OE Voids Volume Or Density:

A ~eries of Eilter elements similar to those of Examples 25 through 28 was prepared using f ibers having 12.5 micrometer diameters and with vvid~ vol-ume~ varied from 63.6 to 89.8 percent.
The characterlstics o~ the elements are set out ` in Table X below:

. .

- 5 6- ~l~ 5~
o .,.
t, ., JJ
~o ~ a~
h o r~
a~ ~ ~ I
E
h ...~
d~
C~
_~ _ C a~ ~rco In a~ ~ ~ ~
X_ a~ ,~ U
a) u cr:
m a~ a) ~: ~ 41 v d .~
E, ~ 1 0 a~ t`3 s~
CL~ ~ E
C~
~, ~ U7 tn n ~
E _ o a ' ~ n~ ~ ~ u~
-_1 h ~ r~

~ ~ ~e; c ~ Q) a~
C4 ~
QJ U~
. . . . .
r~ r o In ~ O
O O U~
:~ ~ 5:
U~
E t~
r~ h ~ D.
U~ o o a~, a E n I~ E
X ~ _~ N
X , . ~_ _ .

_57~ ~5~6~3 When tlle above elements axe compared with ele-ments oE e~ual efficiency made by the method in accord-ance with this invention, life is seen to be much highex for the latter. To illustrate this, Exarnple 5 29 above may be compared to Example 13 when both are tested with a glass bead contaminant suspended in ~IIL-EI-5606 hydraulic fluid. They have virtuall~
iclentical removal ratings of 14 and 15 micrometers, respe~ively, at 99.9 percent efficiency. However, 10 the dirt capacity of Example 13 is ~ grams versus only 12 grams for Example 29 (when tested to 60 pSi t4.2 k~/cm2) pressure drop).

Examples 35 through 37:
]5 This group of Examples compares the collapse pressures of a filter element made with interfiber bonding induced by use of molten or softened fibers (Example 35) with filter elements made using fibers 20 substantially free of this type of interfiber bonding (Examples 36 and 37).
Example 35 is a purchased specimen of a commer-cially available ~Iytrex brand filter (available from Osmonics, Inc.) made using polypropylene fibers and 25 which is characterized by the presence of very strong interfiber bonding. On examination, the bonding was seen to be caused by the adhesion to each other of melted or softened fibers to form a coherent mass.
Example 35 had no internal support core. Examples 36 30 and 37 were prepared by the method in accordance with this invention, also with no support cores. Collapse pressures were determined by individually wrapping the outside of each of the test elements with a thin water-impervious plastic film, sealing the ends of 35 the element and then applying pressure to the exter-ior Q~ the el~ment with water in a transparent hous-ing so that the Eailure of the element could be ob-served.
The dimensions, rating, voids volume and col-5 l~pse pressure for each element are set out below:
Example 35: ~Iytrex 20 micrometer element, 2.75 inch (7 cm) od x 1-3~8 inch (3.5 cm) id x 10 inch (25.~ cm) long, average voids volume 7~.7 percent.
Collapse pressure was 80 psid (5.63 kg/cm2).
Example 36: A twenty micrometer rated element was made by the method ln accordance with this inven-tion, 2.75 inch (7 cm) od x 1-3/8 inch (3.5 cm) id x 10 inch (2S.4 cm) long with no support core. Voids volume was 75 percent. Collapse pressure was 16 psid ]5 (1.13 kg/cm2).
Example 37: The element of this Example was similar to that of Example 36 except that the voids volume was 81.5 percent. Collapse pressure was S
psid (0.35 kg/cm2).
The much lower collapse pressures of the ele-ments in accordance with this invention are due to the substantial absence of interfiber bonding. Con-versely, the Elytrex element was sufficiently strengthened by interfiber bonding that it had the 25 necessary strength to withstand up to 80 psid (5.63 kg/cm2 ) .

Examples 38 and 39:

Nylon 6 resin was fed into the same apparatus and processed in the same general manner as previous-ly described using polypropylene. The operating conditions and properties o~ the resulting elements (when tested using the F2 test~ are described below:

_59_ ~ 5~L~3 Example 38 Example~39 Resin Temperature Degrees F 693 693 (Degrees C)(417, (417) 5 Resin Pressu~e, psi 300 300 (kg/cm2) (21) (21) Eiberizing Air Pressure, psi 40 12 (kg/cm2) (2.8) (0.84) 10 Fiber Diameter, micrometers 2.3 4.0 Voids Volume, Percent 80.5 74.9 Removal Rating, micrometers 5.0 6.1 ]5 Life, grams 12.2 10.2 Example 40:

Filter elements are prepared in accordance with 20 thisinvention in a similar manner to the polypropy-lene elements previously described, but the resin used is polymethylpentene. The properties and char-acteristics of the products are very similar to those obtained using polypropylene.

Exam~le 41:
-A filter element made in the manner of Example25 was tested by passing through it 1,000 ml of a 30 suspension of Pseudomonas diminuta (Ps d), a 0.3 micrometer bacterium, in water. The effluent was analyzed for its content of this bacterium. Whereas the total number of bacteria in the influent was 2.3 x 1012, the effluent content was found to be 1.6 x 35 105, indicating a reduction by a factor of 1.4 x 107.

-~io-This corresponds to an efficiency of 99.99999 per-cent.

E.Yam~~l 2:
Two filter elements (42A and 42B) were made in a manner generally similar to Example 11 except that the inner 50 percent of this element consisted of 1.7 micrometer diameter fibers and the outer 50 percent 10 was proEiled up to 12.5 micrometer fiber diameter.
Both Eilters were tested with Ps d in the same manner as in Example 41 and, in addition, were resterilized and retested using Serratia marcescens (Serr m) as the test organism. The results are shown in Table XI
]5 below.

~2~

o~
C~
~ o~ a~
:~ ~ .
U Q) ~ ar~
C U~ s~
u . ~ ~D
'~ ~D
C~ ~ cr D~ a~

r~ ~
o o E

X X
E
~ ~ ~ a~
a~ ~ .
V ~ ~
C
C~
o o X X
X ~ U~
3 ~ ~

o o o C~
E ~ _~
X X
~,. a: 0 Q) E u~
oJ
v C~ ~ ~
:~ o o ~ X ~C

a. ~tc m E
X

: ' ' ., " , ,, E~amE~ 43:

The general procedure of Example 10 is repeated except that the variation of the resin flow rate, of 5 the fiberizing air flow rate and of the forming roll pressure is continuous as opposed to stepwise, pro-ducing a filter element with a continuously graded fiber diameter structure, i.e., a continuously pro-filed structure, with characteristics comparable to 10 those of the filter element of Example 10.

Eilter elements are prepared in accordance with ]5 this invention and in a similar manner to the poly-propylene elements previously described, but the resin used is polybutylene terephthalate (PBT). The proper-ties and characteristics of the products are similar to those obtained using polypropylene. How-20ever, because of the higher melting point of PBT andits resistance to hydrocarbons, filter elements pre-paxed from PBT will be useful at higher temperatures and in service where they will come in contact with hydrocarbons which might cause the polypropylene ~5fibers to swell.

Example ~5 Relative Compressibility Of Course And Fine Fibers: -Filter elernents were prepared with essentially uniform fiber diameter throughout in the manner of Fxample 1, differing only with respect to fiber dia-meter. Using a tool resembling a laboratory cork borer having an inside diameter of 0.58 inches (1.473 35cm), specimens were cut from the fibrous portion of .

~3L~D r eacil ~ilter element perpendicular to tlle longitudinal axis of the element, forming a generall~ cylindrical specimen ~bo~t 0.6 inches (1.52 cm) in length and about 0.58 inches (1.47 cm) in diameter.
Measured forces of 10 psi tO.7 kg/cm2), 20 psi (1 4 kg/cm2), 60 psi ~4.22 kg/cm2), 90 psi (6.33 kg/cm2) and 120 psi (8.44 kg/cm2) were individually and sequentially applied to the ends oE each of the cylinders while the thickness of the cylinders at 10 e~ch level of applied force was simultaneously mea-sured.
Three elements with Eiber diameters of 2.0, 6.8 and 12 micrometers were each individually tested in this manner. The decrease in thickness of each, when ~5 plotted against force applied, was very similar Eor all three fiber diameters~

Example 46: Collapse Of Filter Flement When Vsed With High Viscosity_Fluids At Hi~ Flow Rates~
. ~

Two filter elements 46~ and 46B were prepared using identical procedures in a manner generally similar to Example 24 except that the average voids volume was 2 percent lower (80 percent cf 82 per-25 cent) Element 46A was tested using the F2 methoddescribed above at 10 liters/minute of water. It had a removal rating of 1l.2 micrometers and a dirt capa-city of 53 grams. The clean pressure drop, prior to the test, was 0.7 psi (.05 kg/cm2) a-t am~ient temper-30 ature of 20 to 25 degrees F (6.67 to 3.89 degrees C).
Element 46B was placed in an F2 test standwhich used hydraulic fluid MIL-H-5606 at 100 degrees F (37.8 degrees C). At this temperature the viscos-ity of MIL-H-5606 is 12.7 centipoise or 12.7 times 35 that of water. No test contaminant was added to the 5~

system, instead clean fluid was flowed through the element at the flow rates set out in Table XII below.

T~BLE XII

Flow Of MIL-H-5606 Pressure Drop psi Liters/Minute (kg/cm2) 1 1.3 (.09) ~ 11.0 (.77) 7 25.5(1.79) 46.5(3.27~
13 cartridge failed due to core collapse ]5 at approximately 80 psi (5.63 kg~cm2) The pressure drop through the filter elements in accordance with this invention is proportional to 20 flow and to viscosity. Based on the aqueous pressure drop at lO liters/minute of 0~7 psi (.05 kg/cm2), the calculated pressure drops for MIL-H-5606 are 8 psi ~.56 kg/cm2) at 10 liters/minute (vs. ~6.5 psi (3.27 kg/cm2) measured) and about 11 psi (.77 kg/cm2) at 13 25 liters/minute (vs. approximately 80 p5i (5.62 kg/cm2) measured).
The much higher pressure drops when using the viscous fluid are due to the compression of the fil-ter medium. This example illustrates that use of 30excessively high voids volumes is not desirable for applications in which high flow of viscous fluids at high pressure drop are involved, particularly with the finer grades of filter element.
--65- ~ 5~L~ 3 -Example 47 Filter Element With The Inner T~ `hirds With Voids Volume O~ 74 Percent ~nd The O-lter One-Third Profiled In Fiber Diameter To 12.5 ~icrometers Also At 74 Percent Voids Volume ~ filter ele~ent is prepared using the general procedure of Example 11 modified as follows:
The initial 67 percent by weight of the fibrous 10 mass of the element is made up of 1.6 micrometer diameter fibers prepared using a water spray coolant and with the forming roll air pressure adjusted to obtain a voids volume of 74 percent. The outer 33 percent by weight of the fibrous mass of the element is applied also using a water spray coolant but while adjusting the resin rate, fiberizing air pressure and the forming roll pressure in a manner such as to profile the fiber diameter in a continuous manner from 1.6 to 12.5 micrometers while maintainins a uniform voids volume of 74 percent. The resulting filter element will have a ~itre reduction in excess of 107( 99.99999 percent efficiency) when tested using 0.3 micrometer diametèr Psuedomonas diminuta organisms and will have a much higher dirt capacity than a similar cylindrical ilter element in which 100 percent of the weight of the weight of the fibrous mass is made up of 1.6 micrometer fibers with a 74 percent voids volume.

- - - . . .

~ 6~ 3 E~ample ~8: Filter element With The Inner Two-Thirds With ~oids Volume Of 74 Percent And The Out~r One-Third Profiled In Both Fiber Diameter ~nd Voids Volume, ~espectively, Up To 12.5 ~~icromete~s And 85 Percent Voids Volume:

A filter element is prepared using the general procedure of Example 47 but modified as follows:
The outer 33 percent by weight of the fibrous mass of the element is applied also using a water spray coolant while adjusting the resin rate, fiber-izing air pressure, and forming roll yressure in a manner such as to profile both the fiber diameter and the voids volume simultaneously, both in continuous fashion. The Eiber diameter is profiled from 1.6 up to 12.5 micrometers at the od and the voids volume is pro~ilea from 74 percent up to 55 percent.
The resulting filter element, when tested for efficiency using the Pseudomonas diminuta bacteria, will have an efficiency essentially equal to that of Example 4? but with a some~hat higher dirt capacity.

-67- ~ ~5~3 The cylindrlcal fibrous structures in accord~
ance with tl~e subject invention Eind use in a variety o~ filtration applications. The filter elements in accordance with the subjeet invention combine extend-5 ed ~ilter life, i.e., higher dirt capacity, at e~ualefficiency, or better efficiency at equal life, or both better efficiency and higher dirt capacity than previously available commercial fibrous cylindrical depth filters. A combination oE high dirt capacity 10 (lon~ life) and removal capacilities over a wide range of particle diameters ma]ces filter elements in accordance with this invention useful as prefilters, for example, to precede an absolute rated final fil-ter when used in critical applications, SUC}I as ster-]5 ilization of parenterals or for providing water foruse in the manuEacture of microelectronic devices~
Filter elements in accordance with this inven-tion are also well suited for the filtration of a wide variety of products from which yeast and bac-20 teria are to be removed, yielding not only a liquideffluent Eree of or greatly reduced in its content of yeast and bacteria, but also one with high clarity.
~ilter elements in aecordanee with this invention may also be used where high titre reduetions, coupled 25 with high dirt capaeities, are required for removal of bacteria.
In addi-tion to their primary use as depth fil-ters with high effieieney and extended life, the eylindrical fibrous struetures in aecordanee with 30 this invention also find use as eoalescers and in insulation applications.

Claims (35)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A cylindrical fibrous structure comprising a fibrous mass of nonwoven, synthetic, polymeric microfibers, said microfibers substantially free of fiber-to-fiber bonding and secured to each other by mechanical entanglement or intertwining, and said fibrous mass having a substantially constant voids vol-ume and varying fiber diameter over at least a substantial por-tion thereof as measured in the radial direction to achieve a graded pore size over said portion.

2. The cylindrical fibrous structure of claim 1 wherein said fibrous mass has a voids volume In the range of from about 60 to about 95 percent.

3. The cylindrical fibrous structure of claim 2 wherein said fibrous mass has a voids volume In the range of from about 64 to about 93 percent and said microfibers have diameters In the range of from about 1.5 to about 20 micrometers.

4. The cylindrical fibrous structure of claim 3 wherein said fibrous mass has a voids volume In the range of from about 75 to about 85 percent and said microfibers have diameters In the range from about 1.9 to about 12.6 micrometers.

5. The cylindrical fibrous structure of claim 1 wherein said nonwoven synthetic polymeric microfibers are com-prised of a thermoplastic selected from the class consisting of polyolefins, polyamides and polyesters.

6. The cylindrical fibrous structure of claim 5 wherein said thermoplastic is polypropylene.

7. The cylindrical fibrous structure of claim 1 wherein said fibrous mass has a substantially constant voids vol-ume throughout as measured in the radial direction.

8. The cylindrical fibrous structure of claim 1 wherein said fibrous mass has a continuously graded fiber diame-ter structure as measured in the radial direction.

9. The cylindrical fibrous structure of claim 1 wherein said fibrous mass has a graded fiber diameter structure over substantially its entire structure as measured in the radial direction.

10. The cylindrical fibrous structure of claim 1 wherein said fibrous mass has in a downstream portion thereof a substantially constant voids volume and in an upstream portion thereof the same substantially constant voids volume as in said downstream portion and a graded fiber diameter structure all as measured in the radial direction.

11. The cylindrical fibrous structure of claim 1 wherein said fibrous structure has a removal rating of from 0.5 to 40 micrometers.

12. A cylindrical fibrous structure comprising a fibrous mass of nonwoven, synthetic, polymeric thermoplastic microfibers having a voids volume In the range of from about 64 to about 93 percent. said micrometers having diameters in the range of from about 1.5 to about 20 micrometers and said microfibers being substantially free of fiber-to-fiber bonding and secured to each other by mechanical entanglement or Inter-twining, said fibrous mass having a substantially constant voids volume and a graded fiber diameter structure over at least an upstream portion thereof both as measured in the radial direc-tion to achieve a graded pore size over said portion.

13. The cylindrical fibrous structure of claim 12 wherein said thermoplastic is polypropylene, said voids volume is in the range of from about 75 to about 85 percent, and said microfibers have diameters in the range of from about 1.9 about 12.5 micrometers.

14. The cylindrical fibrous structure of claim 13 wherein said cylindrical fibrous structure has an efficiency for the removal of Pseudomonas diminuta of at least about 75 percent.

15. The cylindrical fibrous structure of claim 14 wherein said cylindrical fibrous structure has an efficiency for the removal of Pseudomonas diminuta of at least about 90 percent.

16. The cylindrical fibrous structure of claim 15 wherein said cylindrical fibrous structure has an efficiency for the removal of Pseudomonas diminuta of at least about 99.9 percent.

17. The cylindrical fibrous structure of claim 12 wherein said fibrous structure has a removal rating of from 0.5 to 40 micrometers.

18. A cylindrical fibrous structure comprising an open, relatively rigid, central support member and a fibrous mass of nonwoven, synthetic, polymeric microfibers on the exterior of said support, said microfibers substantially free of fiber-to-fiber bonding and secured to each other by mechanical entangle-ment or intertwining and said fibrous mass having a substantially constant voids volume and varying fiber diameter over at least a substantial portion thereof as measured in the radial direction to achieve a graded pore size over said portion.

19. The cylindrical fibrous structure of claim 18 wherein said fibrous mass has a voids volume in the range of from about 60 to about 95 percent.

20. The cylindrical fibrous structure of claim 19 wherein said fibrous mass has a voids volume in the range of from about 64 to about 93 percent and said microfibers have diameters in the range of from about 1.5 to about 20 micrometers.

21. The cylindrical fibrous structure of claim 20 wherein said fibrous mass has a voids volume in the range of from about 75 to about 85 percent and said microfibers have diameters in the range from about 1.9 to about 12.6 micrometers.

22. The cylindrical fibrous structure of claim 18 wherein said nonwoven synthetic polymeric microfibers are com-prised of a thermoplastic selected from the class consisting of polyolefins, polyamides and polyesters.

23. The cylindrical fibrous structure of claim 22 wherein said thermoplastic is polypropylene.

24. The cylindrical fibrous structure of claim 18 wherein said fibrous mass has a substantially constant voids vol-ume throughout as measured in the radial direction.

25. The cylindrical fibrous structure of claim 18 wherein said fibrous mass has a continuously graded fiber diame-ter structure as measured in the radial direction.

26. The cylindrical fibrous structure of claim 18 wherein said fibrous mass has a graded fiber diameter structure over substantially its entire structure as measured in the radial direction.

27. The cylindrical fibrous structure of claim 18 wherein said fibrous mass has in a downstream portion thereof a substantially constant voids volume and in an upstream portion thereof the same substantially constant voids volume as in said downstream portion and a graded fiber diameter structure, all as measured in the radial direction.

28. The cylindrical fibrous structure of claim 18 wherein said central support member comprises a core of self-bonded fibers.

29. The cylindrical fibrous structure of claim 19 wherein said fibrous structure has a removal rating of from 0.
to 40 micrometers.

30. A cylindrical fibrous structure comprising an open relatively rigid, central support member and a fibrous mass of nonwoven, synthetic, polymeric, thermoplastic microfibers having a voids volume in the range of from about 64 to about 93 percent, said microfibers having diameters in the range of from about 1.5 to about 20 micrometers and said microfibers being substantially free of fiber-to-fiber bonding and secured to each other by mechanical entanglement or intertwining, said fibrous mass having a substantially constant voids volume and a graded fiber diameter structure over at least an upstream portion thereof, both as mea-sured in the radial direction.

31. The cylindrical fibrous structure of claim 30 wherein said thermoplastic is polypropylene, said voids volume is in the range of from about 75 to about 85 percent, and said microfibers have diameters In the range of from about 1.9 to about 12.6 micrometers.

32. The cylindrical fibrous structure of claim 31 wherein said cylindrical fibrous structure has an efficiency for the removal of Pseudomonas diminuta of at least about 75 percent.

33. The cylindrical fibrous structure of claim 32 wherein said cylindrical fibrous structure has an efficiency for the removal of Pseudomonas diminuta of at least about 90 percent.

34. The cylindrical fibrous structure of claim 33 wherein said cylindrical fibrous has an efficiency for the removal of Pseudomonas diminuta of at least about 99.9 percent.

35. The cylindrical fibrous structure of claim 30 wherein said fibrous structure has a removal rating of from 0.5 to 40 micrometers.