FLUID TRANSPORT MEDIA

Abstract

Improved apparatus and methods for filtering and applying coating fluids onto substrates. The apparatus and methods are useful for casting embossed sheeting, and in fluid application dies. Improved fluid filtration methods, apparatus, elements and media are also disclosed. An apparatus and method for collection of mist generated in high speed liquid film splitting processes are also disclosed.

Claims

1. A slot fluid distribution apparatus comprising: bounding first and second surfaces; bounding edge walls spanning between the first and second surfaces; an open space volume enclosed by the walls and surfaces; a slit orifice access passageway to the volume having a slit orifice length on a portion of the edge walls; at least one fluid first access passageway to the volume; wherein the passageways and the open space volume, are fluidically connected, and wherein the volume has a first flow resistances in a first direction and a second flow resistances in a second direction which are not equal.

2-24. (canceled)

25. An apparatus of claim 1 wherein at least one of the slot surfaces has a flow modifying surface structure.

26. An apparatus of claim 1 wherein the open space volume is characterized by a gap between the first and second surfaces which varies.

27. An apparatus of claim 1 further comprising flow resistance modifying periodic grid elements within the open space volume, wherein the elements are chosen from the group consisting of mesh assemblies, screen assemblies, fiber assemblies, and wire assemblies.

28. An apparatus of claim 1 wherein the first and second flow resistances are adapted to produce a substantially uniform fluid flow distribution.

29. An apparatus of claim 1 wherein the first direction is along the direction of the slit orifice access passageway and the second direction is toward the slit orifice, and wherein the average fluid first flow resistance is less than the average fluid second flow resistance.

30. An apparatus of claim 1 further comprising a first direction path length equal to the slot orifice length, a second direction path length to equal the distance between the first access passageway and the slit orifice access passageway, and wherein a dimensionless slot viscous number, defined as the first direction path length divided by the second direction path length times the square root of the average of the first flow resistances in the first direction divided by the square root of the average of the second flow resistances in the second direction, is less than one.

31. An apparatus of claim 1 wherein the edge bounding walls encompass a rectangular area having a width equal to the length of the slit orifice access passageway and a depth measured perpendicular to the slit orifice access passageway, wherein the first access passageway is placed at a distance equal to the depth and opposite one end of the slit orifice, and wherein a dimensionless rectangular slot flow number, defined as the width divided by the depth times the square root of the ratio the fluid average first flow resistance in the width direction divided by the fluid average second flow resistance in the depth direction, is less than 0.7.

32. An apparatus of claim 1 wherein the slot edge bounding walls encompass a right triangular area with a hypotenuse side and two right angle sides; wherein the slit orifice access passageway extends along a first right angle side having a first length, and the second right angle side has a second length; the first access passageway is positioned at the end of the second right angle side at the intersection with the hypotenuse side; and wherein a dimensionless triangular slot flow number, defined as the first length divided by the second length times the square root of the ratio the fluid average first flow resistance in the first length direction divided by the fluid average second flow resistance in the second length direction, is less than 0.2.

33. An apparatus of claim 1 further comprising: a distribution die fixtured with first and second external entrances, and an internal cavity connecting the first external entrance to the first access passageway and separating the first external entrance from the slot; wherein the distance through the slot from the first access passageway to the slit orifice access passageway is the slot depth; wherein the cavity spans the some portion of the length of the slit orifice; and wherein an internal flow path connects the entrances, the passageways, and the open space volume.

34. An apparatus of claim 33 further comprising a slot wherein a dimensionless slot flow resistance number, defined as the slit orifice length divided by the slot depth times the square root of the ratio of the fluid first average flow resistance in the slit orifice length direction divided by the fluid second average flow resistance in the depth direction, is less than one.

35. An apparatus of claim 1, wherein the fluid contains target contaminant particles and the slot has a probability of capturing the target contaminant particles at a capture distance from the slit orifice; wherein the target contaminant particle has a hydraulic diameter; and wherein a dimensionless blockage viscous number, defined the particle hydraulic diameter divided by the capture distance times the square root of the ratio of the fluid average first flow resistance in the slit orifice length direction divided by the fluid average second flow resistance in the particle distance direction, is less than 0.5.

36. A method of dispensing flowable material in a die comprising: providing a source of flowable material; providing a die block including a first flowable material terminal point, a second flowable material terminal point, a flow path connecting the terminal points, and at least one distribution slot, wherein the slot includes bounding first and second solid walls, bounding edge walls spanning between the first and second walls, an open space volume enclosed by the walls, at least one first access port to the open space volume, a slit orifice access port having a slit orifice length on the edge walls, flowable material first flow resistances in the slot in a first direction, flowable material second flow resistances in the slot in a second direction, and the average first resistance does not equal the average second resistance; fluidically connecting the access ports, the terminal points, the slit orifice and the open space volume; and translationing the flowable through the slit orifice access port.

37. The method of claim 36 further comprising a first flow resistances first path direction along the direction of the slit orifice, a first path length equal to the slit orifice length, a second flow resistances second direction toward the slit orifice, a second path length equal the distance from the first access port to the slit orifice, and wherein a dimensionless slot viscous number, defined as the first path length divided by the second path length times the square root of the ratio the flowable material average first flow resistance to the flowable material average second resistance, is less than one.

38. The method of claim 36 and further comprising the steps of dispensing and coating the flowable material onto a substrate.

39. The method of claim 36 further comprising the steps of dispensing and forming a film with the flowable material.

40. The method of claim 36 wherein the die block further comprises a cavity extending along the slit orifice direction; wherein the cavity provides a flow path between the die block first terminalpoint and the slot first access port; wherein the slot first access port is an elongated orifice; wherein the slot first direction is along the direction of the slit orifice and the second direction is toward the slit orifice; wherein a slot first path length is equal to the slot slit orifice length and a slot second path length is equal to the shortest distance from the first access port to the slot slit orifice; wherein the first and second slot flow resistances that may be equal or unequal; and wherein a dimensionless slot viscous number, defined as the first path length divided by the second path length times the square root of the ratio the average first flow resistance to the average second resistance, is less than one.

41. The method of claim 40 further comprising the step of producing a substantially uniform flow distribution through the slot slit orifice.

42. The method of claim 36 wherein the flowable material contains target contaminate particles, and further comprising the step of trapping some portion of the target contaminate particles from the flow of flowable material in the slot.

43. The method of claim 42 furthering comprising the step of trapping at least one target contaminate particle; wherein the flowable material first flow resistances, first direction is in the direction along the slit orifice length and the flowable material second flow resistances, second direction is toward the slit orifice; and further comprising the step of providing a dimensionless slot blockage number less than 0.5, wherein the dimensionless slot blockage viscous number is defined as the ratio of the target particle hydraulic diameter to the distance of the trapped target particle from the slit orifice times the square root of flowable material flow average first resistance in the first direction divided by the square root of the flowable material flow average second flow resistance.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING

[0337] In this disclosure, multiple devices are illustrated along with graphical and tabular results of the investigations of their performance.

[0338] FIG. 1a is a schematic cross-sectional view of the internal flow passages of a existing flow distribution device using a slot filled with commercial porous media.

[0339] FIG. 1b is photograph of a cross section of commercially available sintered porous media.

[0340] FIG. 1c is a grey scale contour graph of the local flow rate at points over the porous sheet area spanning the width length along the cavity and the depth distance from the cavity to the fluid exit face and where the sheet is a known porous media sheet.

[0341] FIG. 1d is a graph of the fluid flow at the exit line of the sheet when a known porous media sheet is employed.

[0342] FIG. 2 is a perspective schematic view of an improved metering sheet.

[0343] FIG. 3 is a drawing of the top, side and front views of a unit cell illustrating one flow passageway structure of an improved metering sheet.

[0344] FIGS. 4a, b and c are schematics of improved metering sheets illustrating geometric arrangements and types of the internal flow passageways structures.

[0345] FIG. 5 is an isometric schematic view of the internal feed cavity and metering sheet of a fluid distribution device of this teaching.

[0346] FIGS. 6a and b are schematic drawings illustrating possible flow passageway arrangements in an improved metering sheet.

[0347] FIGS. 7a and b are schematics with top and side views of improved metering sheets constructed with a regular array of columns.

[0348] FIGS. 8a, b and c illustrate still other flow passageway layouts for the improved metering sheet.

[0349] FIG. 9a is a schematic illustration of an internal flow passageway layout geometry designate by the term “rectangular”.

[0350] FIG. 9b is a schematic illustration of an internal flow passageway layout further improved with cross channels which are parallel to the cavity orientation.

[0351] FIG. 9c is a schematic illustration of an internal flow passageway layout further improved with cross channels which are parallel to the cavity orientation and illustrating a sheet structure where two differing base flow passageway cell sizes are employed within a sheet.

[0352] FIGS. 9d, e, and f are schematic illustrations of examples of internal flow passageway layouts employing two or more base cell geometries simultaneously.

[0353] FIG. 9g is a schematic illustration of an internal flow passageway layout geometry designate by the term “hexagonal”.

[0354] FIG. 10a is a schematic illustration of an improved metering sheet useful for coating of stripes of fluid.

[0355] FIGS. 10b and 10c are schematic illustrations of improved metering sheets useful for distributing flow from a small inlet region to a large outlet region.

[0356] FIGS. 10d and 10e are schematic illustrations of improved metering sheets useful for distributing flow from a point inlet region to a large outlet region.

[0357] FIGS. 11a and b are schematic illustrations with top and side views of improved metering sheets using a regular structure of cylindrical columns.

[0358] FIG. 11c is a schematic with top and side views of an improved teetering sheet using a regular array of spherical columns.

[0359] FIG. 11d is contains schematic side views illustrations of improved metering sheets using a regular structure of columns of additional profiles.

[0360] FIG. 12 is a grey scale contour graph of the local flow rate at points over the sheet area spanning the length along the cavity and the distance from the cavity to the fluid exit face. In this case, the sheet is a known commercial, porous media sheet where eighty percent of the flow pores of the sheet bordering the distribution cavity are blocked.

[0361] FIG. 13 is a grey scale contour graph of the local flow rate at points over the sheet area spanning the length along the cavity and the distance from the cavity to the fluid exit face. Here the flow is through an improved metering sheet where eighty percent of the flow pores of the sheet bordering the distribution cavity are blocked.

[0362] FIG. 14 is a graph of the fluid flow from a sheet along an exit line comparing the outflow from a commercial porous sheet with that of an improved metering sheet when there is identical blockage at the sheet entrance.

[0363] FIG. 15 is a schematic illustration of an improved metering sheet for distribution of flow from a circular sheet face.

[0364] FIG. 16 is a schematic illustration of a 3-dimensional cell structure of an improved metering sheet.

[0365] FIG. 17 is a schematic illustration of an improved metering sheet with a 2-D grid of square passageways.

[0366] FIG. 18 is a schematic illustration of an improved metering sheet with a 3-D grid of passageways.

[0367] FIG. 19 is an isometric illustration of an improved metering sheet with a 3-D grid of passageways.

[0368] FIG. 20 is a schematic illustration of a node point within porous media.

[0369] FIG. 21 is a graph of filter experiments illustrating the change in flow as addition particles are captured by the filter.

[0370] FIG. 22 is a graph of filter experiment results plotting the particles trapped or escaped as a function of the pore capture probability.

[0371] FIG. 23 is a graph of filter experiment results plotting the particles trapped or escaped as a function of the pore capture probability.

[0372] FIG. 24 is a graph of filter experiment results plotting the particles captured as a function of the capture probability.

[0373] FIG. 25 is a graph plotting a preferred pore capture probability distribution.

[0374] FIG. 26 is a graph plotting a preferred pore capture probability distribution.

[0375] FIG. 27 is a schematic illustration of a section of filtering grid which includes auxiliary channels.

[0376] FIGS. 28a, 28b and 28c are schematic illustrations of filtering grids which include auxiliary channels.

[0377] FIG. 29 is a graph of filter experiment results plotting particles trapped versus auxiliary channel resistance.

[0378] FIGS. 30, 31 and 32 are schematic illustrations of filtering grids which include auxiliary channels.

[0379] FIG. 33 is a schematic illustration of a section of a conventional porous, sintered, metal sheet modified with auxiliary micro-channels.

[0380] FIG. 34 is a schematic illustration of a section of a conventional porous, sintered, metal sheet modified with auxiliary micro-channels spanning 3-dimensions.

[0381] FIGS. 35a, 35b and 35c are schematic illustrations of filtering grids including both inlet and outlet auxiliary channels.

[0382] FIG. 36 is a schematic cross-sectional illustration of a conventional Z-fold filter element.

[0383] FIG. 37 is a schematic cross-sectional illustration of a known filter element.

[0384] FIG. 38 is a schematic cross-sectional illustration of a block of filter media containing a preferred array of inlet and outlet auxiliary channels.

[0385] FIG. 39 is a schematic cross-sectional illustration of a block of filter media with randomly distributed filtering pores and containing a preferred array of inlet and outlet auxiliary channels.

[0386] FIGS. 40 and 41 are schematic illustrations of a block of filter media containing a preferred arrays of inlet and outlet auxiliary channels.

[0387] FIGS. 42 and 43 are schematic cross-sectional illustrations of example arrays of inlet and outlet auxiliary channels within filter media.

[0388] FIGS. 44 and 45 contain tabulated results of filtration experiments comparing filter media of this invention with known Z-fold elements.

[0389] FIGS. 46 is a perspective schematic view of the internal flow passages of a known flow distribution device using a slot.

[0390] FIG. 47 is a perspective schematic view of the internal flow passages of a known flow distribution device using multiple bores and orifices.

[0391] FIG. 48 is a graph of the output of a known multi-orifice die with three plugged bores.

[0392] FIG. 49a is a perspective schematic view of a section of one embodiment of the inventive die using one row of bores.

[0393] FIG. 49b is a perspective schematic view of a section of one embodiment of the inventive die using two rows of bores.

[0394] FIG. 49c is a perspective schematic view of a section of still another embodiment with a die using two rows of bores.

[0395] FIG. 49d is an isometric schematic view of the flow passageways of still another embodiment with a die concept using two rows of bores.

[0396] FIGS. 50a, b, c and d are schematics of die cavity and drilled hole layouts.

[0397] FIG. 51a is a graph of the output of a multi-orifice, drilled hole die illustrating the utility of auxiliary channels.

[0398] FIG. 51b is a graph of the output of a multi-orifice, drilled hole die with three plugged bores with and without the use of the channels of the invention.

[0399] FIG. 52 is a graph of the output of a multi-orifice die with 45% of the bores plugged illustrating how channels improved filtration and uniform flow distribution.

[0400] FIG. 53 is schematic cross sectional view of the internal flow passages of an embodiment of the invention.

[0401] FIG. 54 is schematic cross sectional view of the internal flow passages of another embodiment of the invention.

[0402] FIG. 55 is a partial perspective schematic view of another embodiment of the invention.

[0403] FIG. 56 is a graph of the lengthwise flow distribution of a known multi-orifice flow distribution device as a function of the dimensionless viscous number.

[0404] FIG. 57 is a graph illustrating the utility of the invention and demonstrating improving the flow distribution from a multi-orifice flow device by selective blocking of bore entrances.

[0405] FIG. 58 is a graph of the uniformity index for a device of the invention as a function of an auxiliary channel position and dimensionless viscous number when the viscous number is Nvs=0.0001.

[0406] FIG. 59 is a graph of the uniformity index for a device of the invention as a function of the channel position and dimensionless viscous number when significant bore clogging occurs and when the viscous number is Nvs=0.0001.

[0407] FIG. 60 is a graph of the uniformity index for a device of the invention as a function of the channel position and dimensionless viscous number when significant bore clogging occurs and when the viscous number is Nvs=1.

[0408] FIG. 61 is a graph of the uniformity index for a device of the invention as a function of an auxiliary channel position and the channel dimensionless viscous number when the viscous number is Nvs=1.

[0409] FIG. 62 is a graph of the local flow from the bore orifices when the intersecting bore geometry of FIG. 50a is employed and there is blockage of two bores.

[0410] FIG. 62 is a graph of the local flow from the bore orifices when the intersecting bore geometry of FIG. 50b is employed and there is blockage of two bores. The influence of number of bore intersections is shown.

[0411] FIG. 64 is an isometric schematic of the basic cube element of the three dimensional flow grid of this teaching.

[0412] FIG. 65 is an isometric schematic of a volume filled with an array of basic cube elements of flow passages that may be used to model the flow within the volume.

[0413] FIG. 66a illustrates a 2D square grid of passageways that may be studied by the flow model of this teaching.

[0414] FIG. 66b illustrates grid geometries that may be modeled with the passageway flow model of this teaching.

[0415] FIG. 66c illustrates an interlaced rectangular grid geometry that may be modeled with the passageway flow model of this teaching.

[0416] FIG. 66d illustrates a triangular grid geometry that may be modeled with the passageway flow model of this teaching.

[0417] FIG. 66e illustrates a diamond grid geometry that may be modeled with the passageway flow model of this teaching.

[0418] FIG. 66f illustrates a hexagonal grid geometry that may be modeled with the passageway flow model of this teaching.

[0419] FIG. 67a is a chart illustrating the flow distribution from an improved, two layer, multi-orifice die each with one auxiliary cross channel per layer and with 45 percent of the bores blocked.

[0420] FIG. 67b is a chart illustrating the flow distribution from an improved multi-orifice die with four auxiliary cross channels per layer and with 45 percent of the bores blocked.

[0421] FIG. 68 is a schematic cross-sectional view of a known coating distribution die using a slot.

[0422] FIGS. 69a, 69b and 70a are schematic cross-sectional views of a blade coating devices of this invention.

[0423] FIGS. 70b, 70c, 70d, 70e are compound coating devices where the die block containing the distribution cavity is physically separated from the applicator means.

[0424] FIG. 71 is a schematic cross-sectional view of a known casting die station for producing embossed webs.

[0425] FIG. 72 is a schematic cross-sectional view of a known free span die coating station.

[0426] FIG. 73 is a schematic cross-sectional view of a known transfer coating station.

[0427] FIG. 74 is a schematic cross-sectional view of an improved transfer coating station.

[0428] FIG. 75 is a schematic, close-up, cross-sectional view of lips of a die of an improved coating station.

[0429] FIGS. 76 and 77 are cross-sectional views of polymeric lips of this invention.

[0430] FIGS. 78a and 78b are cross-sectional views of compound polymeric lips of this invention.

[0431] FIGS. 79a and 79b are cross-sectional views of disposable coating dies employing non-metallic polymeric lips and improved fluid distribution sheets.

[0432] FIGS. 79c, 80 and 81 are cross-sectional views of a coating dies and mounting attachments illustrating embodiments of this invention.

[0433] FIG. 82 is a schematic cross-sectional view of the out-running side of a two roll nip region where mist is generated.

[0434] FIG. 83 is a schematic view of the mist collection device of this invention in its operational position in a roll nip.

[0435] FIG. 84 is a schematic end view of a mist collection die.

[0436] FIG. 85 shows schematic top and end view of a mist collection lower plate of a mist collection die.

[0437] FIG. 86 is a graph of the uniformity for a distribution die using improved porous media.

[0438] FIG. 87 is a graph of the uniformity index for a drilled hole die with one auxiliary channel.

[0439] FIG. 88 is a graph of the uniformity index for flow from metering sheets.

[0440] FIG. 89 is a schematic edge view of an improved slot with bidirectional flow conductance.

[0441] FIG. 90a is a graph of the uniformity index for a flow from corner fed polygon slots.

[0442] FIG. 90b is a graph of the uniformity index for a flow from a rectangular slots where the feed point is within the rectangular area.

[0443] FIG. 91 is a schematic top view of an improved slot with bidirectional flow conductance.

[0444] FIG. 92 is a schematic isometric view of an improved conventional porous media with bidirectional flow conductance.

[0445] FIG. 93 is a graph of the uniformity index for flow from a cavity and slot system.

[0446] FIG. 94a is a schematic top view of an improved slot distribution system with a polygonal major surface shape.

[0447] FIG. 94b is a schematic top view of an improved rectangular slot distribution system where fluid is fed at a point within the rectangle.

[0448] FIG. 95 is a schematic top view of an improved slot distribution system with an irregular major surface shape.

[0449] FIG. 96 is a schematic top view of an improved slot distribution system where the polygonal major surface shape of FIG. 95 has been mathematical transformationally mapped onto a rectangular plane.

[0450] FIG. 97 is a graph of the uniformity index for flow from a slot with blockage.

[0451] FIG. 98 is a graph of the uniformity index for flow from slots with non-uniform gaps.

[0452] FIG. 99 is a schematic cross-sectional view a multi-orifice die with another embodiment of the invention.

[0453] FIG. 100 presents schematic views of die shims of the invention.

[0454] FIG. 101 is a schematic view of a die shim of the invention.

[0455] FIG. 102 is a schematic cross-sectional view a multi-orifice die of the invention.

[0456] FIG. 103 is a graph of the uniformity index for flow from a porous sheet with blockage.

[0457] While the above identified figures set forth several preferred embodiments of the invention, other embodiments are also contemplated, as noted in the discussion. In all cases, this disclosure presents the invention by way of representation and not limitation. It should be understood that numerous other modification and embodiments can be devised by those with ordinary skills in the art which fall within the scope and spirit of the principles of the invention.

DETAILED DESCRIPTION

[0458] The following description of the invention is provided as a teaching of the invention in its best, currently known embodiments. To this end, those ordinarily skilled in the relevant art will recognize and appreciate that many changes can be made to the various aspects of the invention described herein, while still obtaining the beneficial results of the present invention. They will also recognize that the invention covers such a broad range of uses that the best embodiments for specific application may not be the optimum for others. It will also be apparent that some of the desired benefits of the present invention can be obtained by selecting some of the features of the present invention without utilizing other features. Accordingly, those who work in the art will recognize that many modifications and adaptations to the present invention are possible and can even be desirable in certain circumstances and are a part of the present invention. Thus, the following description is provided as illustrative of the principles of the present invention and not in limitation thereof.

[0459] As used herein, the singular forms “a,” “an” and “the” include plural referents unless the context clearly dictates otherwise. Thus, for example, reference to a “metering sheet” includes embodiments having two or more such sheets unless the context clearly indicates otherwise.

[0460] Ranges can be expressed herein as from “about” one particular value, and/or to “about” another particular value. When such a range is expressed, another embodiment includes from the one particular value and/or to the other particular value. Similarly, when values are expressed as approximations, by use of the antecedent “about,” it will be understood that the particular value forms another embodiment. It will be further understood that the endpoints of each of the ranges are significant both in relation to the other endpoint, and independently of the other endpoint.

[0461] As used herein, the term or phrase “fluid communication” is intended to include aspects where a fluid may be caused to flow from point or object to another.

[0462] As used herein, “flow resistance” of a passageway relates the flow through the passageway to the pressure drop through the passageway. The flow rate multiplied by the flow resistance equals the pressure drop from one end to the other end of the passageway.

[0463] References are made to size scales. As used herein, the term or phrase “nanometer scale size” or “on the scale of nanometers” is intended to include all sizes ranging from 1 to less than 1000 nanometers. The term or phrase “micron scale size” or “on the scale of microns” is intended to include all sizes ranging from 1 to less than 1000 microns. The term or phrase “millimeter scale size” or “on the scale of millimeters” is intended to include all sizes ranging from 1 to less than 10 millimeters.

[0464] As used herein, the prefix “micro-” is intended to refer to sizes of about 1 micron to about 1000 microns. “Micro-channels” refers to channels with hydraulic diameters ranging from about 1 to 1000 microns.

[0465] A primary teaching of this invention is an apparatus for distributing fluid along a line comprising slot like volume that encompasses a fluid conveying means. Such means conveys fluid from an inlet edge, inlet point or inlet points to an outlet edge. The volume has first and second major surfaces in close proximity where flow passes between the surfaces, and it has various shapes. In some examples the apparatus is a sheet where flow passes edgewise through it. Within portions of the volume, the ratio of the flow resistance of the fluid in a first direction to the flow resistance perpendicular to the first direction when measured at the same flow rate does not equal one. These distributing means and articles include but are not limited to enclosed slots, porous media, drilled passageways, machined pathways, micro-channeled sheets, and multiple layers of such. The multiple layers maybe interconnected.

[0466] It has been found that when fluid is distributed from a point through the volume to a discharge line, generally uniform outflow may be achieved when special ranges of parameters are used. A slot like volume is characterized by two major surfaces. These are generally near parallel or near concentric surfaces in close proximity to each other. The distance between them is the thickness or gap. It may be constant, or varying, or slowly varying, or on the average constant, or periodically varying, or varying about a constant value, or varying about an average value or systemically varying. The major surfaces each have an area generally with the same area shapes with perimeters that fit within a rectangle. The rectangle has width and depth lengths.

[0467] The major surfaces may be flat planes or surfaces with elevation changes about some reference plane in Cartesian coordinates, or a diameter in cylindrical coordinates, or a diameter in spherical coordinates. This sheet like volume may be folded, twisted, bent, or deformed in many different ways. This slot like volume may also be thought of as a membrane that has and internal

[0468] This sheet or membrane like functional volume may be formed of a void volume enclosed by bounding solid surfaces or fluid restraining planes. A slot is an example. Additionally, this volume may be filled or partially filled with material or porous material. The volume between the major surfaces again is bounded by an edge or multiple edge boundaries. The sheet like volume is used to direct fluid from an inlet to an outlet edge.

[0469] We teach improvements to existing technology to achieve a prescribed flow or a uniform flow from a sheet outlet edge region. FIG. 95 shows a top view generalized sheet like volume shape 9500. The top major surface area 9505 is flat to simplify the illustration, but this is not a requirement. Dashed lines 9510 and 9530 define an outlet edge flow region where fluid flows outward and normal from the volume edge.

[0470] The area 9505, and the defining perimeter edge of the volume as projected in the top view, and the outlet edge defining lines 9510 and 9530 may generally be mapped onto a straight sided polygon as shown in FIG. 96. Here the defining perimeter is now the rectangle edge of area 9605, and the discharge edge is confined between line 9610 and 9630. Any point on the surface will have a corresponding point on the rectangle. Flow from the outlet edge is shown by arrows 9640.

[0471] When the major surface of the volume can be mathematically transformed into a polygon, the findings and teachings for flow distribution in and from the polygon concerning the effect of flow resistances on flow distribution may be mathematically interpreted to obtain analogous effects in the original irregular shaped volume. If a prescribed flow an outlet edge for an irregular shape is desired, and if the shape may be transformed into a polygon such as the rectangle or triangle in the following descriptions, then the teachings for obtaining a flow distribution in the transformed space may be used to determine the local variability of flow resistances in the original irregular shape necessary to obtain an equivalent distribution.

[0472] An analogy is that if the polygon were a very elastic rubber sheet, it may be stretched and manipulated into the irregular shape of interest, then properties of the polygon may be transformed to obtain the equivalent flow results in the irregular shape.

[0473] Various implementations of the invention and systems that incorporate it are illustrated.

[0474] A. Existing Dies

[0475] FIG. 46 is a perspective schematic view of internal flow passages of a known flow distribution device using a slot. The flow passages consist of an internal cavity 501 into which a flow is introduced as illustrated by arrow 503. The fluid flows down the length of the cavity. Along its length flow enters a slot 502 which connects to the exterior of the die. The fluid exits as illustrated by the arrows 504 from the slot orifice 505. In coating and other flow distribution devices the cavity is also referred to as the manifold or distribution manifold.

[0476] FIG. 47 is a perspective schematic view of internal flow passages of known flow distribution devices using multiple bores and orifices. The passages consist of a cavity 510, and bores 512, and orifices 514. The orifices are located on a discharge face which is not shown. A flow is introduced as illustrated by arrow 516. It is distributed down the length of the cavity 510, and enters the bores 512. The bores connect the cavity with the exterior of the die where the flow exits through the orifices 514.

[0477] B. Existing Modeling

[0478] Flow models may be used to accurately describe the distribution of flow rates from the slot orifice of a slot die. We have found that in many eases Stokes flow and lubrication flow models work well. A model with a three dimensional grid of flow passages is briefly summarized in the following description.

[0479] In the past, a simple approach to modeling the flow in a porous media and media flow passageways was used. It considered the media as a continuum where flows versus pressure drop are related by a Darcy constant. Commercially important three dimensional porous structures have been modeled by the Darcy approach where flows in individual passageways are ignored and the multitude of channels in a region are averaged together.

[0480] C. New 3D Flow Grid Modeling

[0481] Three-dimensional modeling of flow through interconnected channels, tubes, bores and passageways has been developed. It has been inventively used to study, to optimize, and to define new and improve devices and methods for fluid distribution. Here the flows in each passageway are individually calculated and variations in flow between them are studied. In contrast to the Darcy approach, the mathematically more intensive task of calculating flows for every individual flow passageway is accomplished.

[0482] A general three dimensional model has been developed which may be used to describe the flow in any system of flow passageways or designed porous media. It is illustrated in FIG. 64. It is constructed of arrays of channels, assembled in grids of passageways. The grids are arranged so that they outline the edges of what is called in topology a convex quadrilateral. The grids extend in x, y and z directions. The base element of this model is a single array of passageways represented by the lines 750 connecting eight node points represented by the small circles 752, and where the passageways form the outlining edges of one convex quadrilateral. For simplicity, this quadrilateral may be thought of as a cube without limiting the general usefulness of the model. Twenty-four additional passageways 754 extend from the eight corners or node points. These additional passageways provide flow paths for fluid entering and exiting the base cube. With the cube aligned with orthogonal x, y and z axis, these additional passageways extend in x, y and z axis directions from each node.

[0483] The base cube element consists then of twelve passageways 750, and twenty-four additional connection passages 754. Each has an arbitrary resistance to fluid flow through it. When the flow is pressure driven through the grid of flow paths, at least one pair of the additional passageways serve as inflow and outflow paths to and from the cube.

[0484] Thirty-six equations must be generated to solve for the flow in each flow path. Eight mass balances may be written, one for each node where passages intersect. Multiple closed loops of passageways may be defined. Because the pressure drop around any closed loop of paths must sum to zero, flow loop equations may be written for the cube. Five independent equations may be written. Additionally, a pressure balance equation for any pair of inlet and outlet passageways and an arbitrary flow path between them is used when the pressure differential between the inlet and outlet is known. Still further for any passageway that is not used, the flow rate is set to zero. In this manner 36 equations may be written, and the flow in every passageway may be found by solving the equations.

[0485] For Newtonian fluids the equation set will comprise a linear system of equations which may be solved by various methods. When the fluid flow properties depend upon the rate of flow, the equation set will be non-linear.

[0486] An alternative method of solving for the flows in the 3D grid is as follows. If pressure boundary conditions are known, and the flows at any node must sum to zero. The pressures at the nodes may be calculated noting that the flow from one node to the next equals the path flow conductance times the pressure change. A system of equations is developed for solving for the pressures at each node. Once the pressures are known, the path flow rates are calculated using the flow conduction relationship.

[0487] By using an assemblage of the base cubes, an assembly of fluid flow passageways may be used to span a volume, a 3-dimensional space.

[0488] One can think of a general of assemblages of cubes where each edge of the cube is a flow duct. Blocks of cubes may be stacked together to form a composite spanning the volume to be modeled. An example geometry is shown in FIG. 65 where each bold line represents a flow channel. Here an assemblage of cubes 760 fills a three dimensional volume. Fluid is flowed into the assemblage through an inlet tube indicated by arrow 766. Fluid exits through a tube indicated by arrow 768.

[0489] D. 2D Flow Model

[0490] A simplified form of the cube model may be used to model flow channels and passageways existing in a plane of two dimensions. FIG. 66a illustrates a simple two dimensional model of flow paths. Here an assemblage of passageways 772 fills a two dimensional (2D) plane. Fluid flows into the assemblage through an inlet tubes indicated by arrows 776. Fluid exits through a tubes indicated by arrows 774. The intersections of passageways are nodal points. The passageways are arranged in square grids connecting the nodes.

[0491] FIG. 66b illustrates how more complex nodal connectivity may be simulated by setting the flow resistances of a repeating pattern of passageways to infinity in the square grid in FIG. 66a. FIGS. 66c through 66f illustrate other equivalent 2-dimensional path geometries that may be studied by manipulating the resistances of the model in FIG. 66a. All of these variations may be created by setting selected passageway flow resistances to zero and to infinity.

[0492] Using a large mesh of the squares or the triangles, nearly any assembly of interconnecting 2D flow paths may be approximated. In a like manner using a cubic grid assembled of multiple cube elements of FIG. 64, assemblies of interconnecting 3D flow paths may be approximated.

[0493] E. Improved Multiple Orifice Dies

[0494] The minimization of the cost of coating may be accomplished through the use of the improved multi-orifice dies. These dies have a generally lower fabrication cost. It is a teaching to fabricate these dies from polymeric materials.

[0495] In the following description of multi-orifice dies, the term “bore” refers to a passageway which conveys fluid between a supply cavity and an exit orifice on or in the proximity of a die face. The bore is a member of a multiplicity of bores arranged to connect a line or lines of multiple exit orifices. Drilled holes are an example. It is within the scope of this invention to have a slot orifice extend along the length of the die on the die face. The slot intercepts the bores before they reach the die face. Such a slot may perform the function of merging the separate streams of fluid exiting from the bores into a ribbon of fluid at or near the die face.

[0496] The terms “auxiliary channel in the die width” and “cross channel” refer to flow a passageway which is generally perpendicular to the bores. Alternatively, the passageway may be at an angle to the bores. The passageway intercepts at least two bores.

[0497] The term “uniformity index” is a dimensionless number that refers to the flow distribution from the bores and is abbreviated UI. It is defined here as the difference between the maximum and minimum bore flow rates at the exit face divided by the average bore flow rate. A perfect uniform flow distribution has a uniformity index equal to zero.

[0498] 1. Modeling Multi-Orifice Dies

[0499] A model of the flows in multi-orifice fluid distribution devices was created. For devices with flow passages like those illustrated in FIG. 47, it is found that the flow distribution from the die orifices is primarily a function of the ratio of the viscous flow resistance down the cavity 510 to the total viscous flow resistance through the bores 512. This is referred to as the dimensionless viscous number, Nvm.

[0500] The models are valid when the assumption of Stokes flow is accurate. That is when conditions are such that inertia and gravity have insignificant influences, and when bore and channel entrance losses are small.

[0501] 2. The Clogging Problem

[0502] When the viscous number approaches zero a perfect, uniform flow distribution is achieved, but this does not prevent clogging. When large particles contaminate the fluid, they may clog the entrances to bores. An example of the problem presented by clogging three bores of a FIG. 47 type die at dimensionless positions 0.05, 0.47, and 0.79 is illustrated in FIG. 48. Here the ratio of the local flow rate from a bore to the average flow rate from the bores is plotted versus the dimensionless position along the die discharge face. The result of clogging is a disrupted flow distribution from the respective orifices. For each clogged bore there is a corresponding lack of coating exiting at the die face. Flow enters the cavity at position zero. If the die is being used to coat a substrate, defective product will result and force shutdown of production.

[0503] a. Auxiliary Cross Channels

[0504] Through extensive investigation, it has been discovered that the clogging problem of multi-orifice dies may be substantially diminished. An improved multiple orifice die is obtained by adding at least one cross die auxiliary channel. More than one is most preferred. The channels are illustrated in FIG. 49a. The multi-orifice die body 540 contains an internal cavity 542 which extends down the length of the body. Multiple bores 544 extend from the cavity to an exterior face of the die 545. Fluid is introduced to the cavity by a means not shown and then flows to the bores 544. The fluid flows through the bores and exits from orifices 546 on the die face 545.

[0505] Auxiliary channels 547 and 548 are positioned to intercept the bores 544 and provide for the transfer of fluid from one bore to at least one other bore. It is preferred that at least one of the channels extends down the widthwise length of the die and engages with substantially all of the bores 544.

[0506] b. Layered Bores:

[0507] Through extensive investigation, it has additionally been found that the clogging problem of multi-orifice dies may be made insignificant using additional rows or layers of bores stacked upon each other. This improved multiple orifice die is illustrated in FIG. 49b. The die body 640 contains an internal cavity 642 which extends down the length of the body. Multiple bores extend from the cavity to an exterior face of the die 645. These bores are arranged in a first row of bores 643 arrange along a dashed line 652, and a second row 644 arrange along a dashed line 652. These lines need not be straight.

[0508] Fluid is introduced to the cavity 642 by a means not shown and then flows to the bores 643 and 644. The fluid flows through the bores and exits from orifices 646 on the die face 645.

[0509] Auxiliary cross channels 647 and 648 are positioned to intercept the bores 643 and 644 and provide for the transfer of fluid from one bore to at least one other bore hole. Although two channels are illustrated, improvement is achieved using only one. More than one channel is preferred. Improve performance is achieved when the channel provides for the transfer of fluid from one bore, a first bore, to at least a second bore or a number of additional bores. The first bore is a member of a first line of bores such as line 652 and the second bore may be a member of the same line of bores or a member of another line of bores. It is preferred that the channel connect both adjacent bores in the same line and adjacent bores in another line.

[0510] It is preferred that at least one of the channels extends down the length of the die and engages with substantially all of the bores 643 and 644 of lines 651 and 652.

[0511] i. Auxiliary Cross Channels

[0512] Another useful embodiment of our invention channel geometry is illustrated in FIG. 49c. In this embodiment, the first layer of bores 664 arranged along exit line 671 is interconnected by auxiliary channels 669 and 670. The second layer of bores 663 is arranged along line exit 672 and the bores are interconnected by auxiliary cross channels 667 and 668. There is no interconnection between the two layers except at the cavity 662.

[0513] The auxiliary channels 667 and 668 are positioned to intercept the bores 663 and provide for the transfer of fluid from one bore to at least one other bore hole. Although two channels are illustrated, improvement is achieved using only one. More than one channel is preferred. Improve performance is achieved when the channel provides for the transfer of fluid from one bore, a first bore, to at least a second bore or a number of additional bores. This first bore is a member of a first line of bores arranged along a line such as 672 and the second bore is a member of the same line of bores.

[0514] ii. Auxiliary Inter-layer Channels

[0515] Still another useful embodiment of our inventive channel geometry is illustrated in FIG. 49d. An isometric view of a portion of bores, and auxiliary channels is shown. Bores 680 and 681 are arranged along lines 684 and 685 respectively into layers one above the other. These bores terminate at discharge orifices 689 and 690. The bores 681 along line 685 are connected in the top layer by auxiliary cross channel 682. The bores 680 along line 684 are connected in the bottom layer by auxiliary channel 683. Additionally, layer interconnect auxiliary channels 686, 687 and 688 provide passageways for flow between the layers. These inter-layer auxiliary channels further improve the distribution of flow from orifices 689 and 690. It is preferred that the auxiliary layer interconnect channels simultaneously intersect both the bores and the intra-layer auxiliary cross channels. However, improvement in flow distribution may also be obtained by other locations.

[0516] Improved flow distribution is obtained when contaminants are present with only interlayer auxiliary channels present. It has been found that an auxiliary passageway, such as channel 547 or 548 in FIG. 49a or channel 647 or 648 in FIG. 49b or channels 683 and 682 in FIG. 49d are useful in eliminating flow defects. Significantly when two or more layers of bores are employed, improved flow distribution along the die face may be obtained using layer interconnect auxiliary flow channels that connect two or more lines or layers of bores. Most preferred is the use of two or more lines or layers of bores with one or more intra-layer cross flow auxiliary channels and one or more arrays of interlayer auxiliary channels.

[0517] In the case of a single line of bores, auxiliary channels are helpful. When a cavity to bore entrance is plugged by any means, the total disruption in flow from the die does not result if at least one auxiliary channel is present. Referring to FIG. 49a, the presence of auxiliary channels like 547 or 548 enable flow from all the orifices 546 in the presence of clogging. The channels produce not only flow from all orifices negating plugging problems at the cavity, but they also improved orifice flow uniformity down the line of the die discharge orifices. This new technology allows acceptable die outflow uniformity when the prior art would not. If the die is being used to coat a fluid onto a web translating past it in proximity to the face 545, a continuous distribution of coating fluid is produced.

[0518] 3. Model Parameters

[0519] Modeling has been used in designing liquid distribution devices using bores and channels that function well. Fluid dynamically, it is useful study the flow distributions in and from a die for the case of a constant cross-sectional area primary cavity distributing flow to bores. One prime parameter is the dimensionless viscous number, Nvsm. A second is the dimensionless cross channel viscous number, Nvcm. Nvsm is defined as the total viscous resistance to flow down the cavity (manifold) divided by the composite total flow resistance through the bores. For n bores of equal resistance carrying flow in parallel, the composite total flow resistance equals 1/n times the flow resistance through one bore. When the bores have un-equal resistances, the reciprocal of the total resistance equals the sum of the reciprocals of the resistances of the individual bores.

[0520] A preferred value is Nvsm less than 1.0 when no auxiliary channels are present.

[0521] Nvcm is defined as the total viscous resistance to flow down the cavity (manifold) divided by the composite total flow resistance through all the auxiliary cross channels. Preferred values of Nvcm are greater than 0.001 .

[0522] Nvim is defined as the total viscous resistance to flow down the cavity divided by the composite total flow resistance through all the inter-layer channels. Preferred values of Nvim greater than 0.001

[0523] 4. 2D Model Shows Cross Channels Improve Results

[0524] When the auxiliary cross channels are uniform in size and uniformly spaced, and when one layer of bores is present, a useful characterizing parameter is the ratio of the number of uniformly spaced bores along the die width minus one, to the number of auxiliary channels plus one. This ratio is symbolized as Kw/Kd. This is further multiplied by the square root of the bore flow resistance between the auxiliaries divided by the auxiliary channel flow resistance between the bores. A constant area, end fed cavity is assumed. Symbolically this parameter is Nspm=Kw/Kd sqrt(Rw/Rc). It has equivalent fluid dynamic relevance and meaning to the terms used in describing flow in slots and porous media.

[0525] FIG. 87 illustrates that decreasing the dimensionless number Nspm improves the uniformity independent of the value of the viscous parameter Nvsm.

[0526] FIG. 51a illustrates how dramatically the UI may be improved when channels are added. It also shows the importance of the dimensionless number Nspm which equals the number of drilled holes minus 1, divided by the number of channel plus one, times the square root of the ratio of hole conductivity between channels to channel conductivity between holes. The graph illustrates that a die design where the value of Nspm is less than 3 is preferred.

[0527] FIG. 51b illustrates that the presence of channels like 547 or 548 of FIG. 49a simultaneously eliminates clogging defects and may also improve the flow distribution from the die along its length. When a drilled hole die having no auxiliary cross channel is used and when three of the bores between positions 0 and 0.2 are clogged at their entrances, disrupted flow results at the discharge face. No flow exits from these clogged bores. The graph line labeled “0-Cross Channels” in FIG. 51 illustrates the flow from the die when these blockages are present.

[0528] When one or four cross channels are added to the die and where the dimensionless cross channel viscous number is Nvcm=0.5, the defect from clogging is completely healed. This is illustrated by the curves labeled “1-Cross Channel” and “4-Cross Channels,” which fall on top of each other. When the channels are employed flow exits from every orifice on the die face, and the distribution of flow from the bores along the die is substantially improved.

[0529] For the die design illustrated in FIG. 49a and the parameters noted for FIG. 51, the flow maldistribution from the die is large (UI is large) without the use of auxiliary channels even if there are no bore blockages. With no channels and no blockages, the ratio of local flow to average flow is at one end of the cavity 1.33 and 0.87 at the other. For perfectly uniform flow this ratio would be 1.0 all along the die. An observation of interest is that the introduction of auxiliary channels lowers the ratio to 1.09 and 0.93 at the ends. With no channels the uniformity index for the die is poor at 0.46, and with channels, the uniformity index is greatly improved to 0.16. This is a 65% improvement in the flow distribution uniformity. The uniformity index would be 0.0 for perfectly uniform flow from the orifices.

[0530] For the die design illustrated in FIG. 49a, and the parameters noted for FIG. 51, and with clogging of three bores, and no auxiliary cross channels, the uniformity index is poor at 1.33. When this clogging exists but the auxiliary channels are present, the uniformity index is excellent at 0.16.

[0531] It can be seen that clogging of bores by trapping particles has the negative effect of disrupting a die's outflow uniformity if not corrected by auxiliary channels. However, bore clogging also accomplishes the task of filtering material from the fluid. This can be a desirable result if those particles degrade a product being manufactured. The presence of auxiliary channels in the fluid distribution device allows the device to perform filtration of the fluid and simultaneously improve the distribution of fluid. Filtering is another useful function of the invention.

[0532] 5. Designing for Filtering

[0533] A filtering action is useful because even pre-filtered fluids and the purest of fluids are easily contaminated. Incomplete cleaning and accidental contamination of process equipment is an ongoing and continuous problem in manufacturing. It is therefore highly useful that the

[0534] CLEAN VERSION invention may be used to remove contaminants from the fluid stream during distribution. The presence of auxiliary channels of this invention allows contaminate removal and simultaneous improvement of the uniformity of flow distribution from a distribution device. This is an advancement over prior art.

[0535] The filtration utility is illustrated in FIG. 52. The die design illustrated in FIG. 49a is used to filter a large number of particles from the fluid, and the die parameter has a value Nvcm=1. Here forty-five percent of the bores have been randomly clogged by trapping contaminants carried by the fluid. Examples of possible contaminants are fibers, solid particles, soft particles, viscoelastic particles and even gas bubbles. The contaminants are trapped at the junction of the bores and the cavity.

[0536] in FIG. 52 the line labeled “no channels” shows the flow distribution for a multi-orifice die with no auxiliary channels. While the die usefully traps particles and does not allow them to pass to the orifices, the flow is totally disrupted. Flow does not issue from 45% of the orifices. Additionally, from the other orifices where there is flow, the rates are on the order of twice the desired average. When one auxiliary channel is added, the performance totally changes. With only the single auxiliary channel the distribution is continuous with flow from every orifice. The flow distribution has uniformity on the order of plus or minus 18 percent. With four channels all having the same flow resistance and where Nvcm=1, the distribution is further smoothed and the variation is diminished to only 3 percent. With four channels and Nvcm=1, the flow distribution is very uniform and a large number of particles have been filtered from the fluid. A multiplicity of auxiliary channels generally produces improved results over a single channel.

[0537] While the graph illustrates the results of clogging of the bores at their junction with the cavity, contaminants may also be trapped at the junctions of bores and auxiliary channels prior to arriving at the fourth channel, and there still will be flow from all orifices. FIG. 52 illustrates results for only one of a multitude of possible mechanical designs. The number of channels, their positions, the channel viscous flow resistances, and the cavity flow resistance may be chosen in combination to achieve a wide variety of flow distributions while meeting filtering challenges.

[0538] Those with ordinary skills in the art will recognize that while uniform flow distribution is a common need, other types of prescribed flows may be achieved with the invention. If a positions, the channel viscous flow resistances, and the cavity flow resistance may be chosen in combination to achieve a wide variety of flow distributions while meeting filtering challenges.

[0539] Those with ordinary skills in the art will recognize that while uniform flow distribution is a common need, other types of prescribed flows may be achieved with the invention. If a predetermined distribution that varies as a function of position is desired, it may be achieved by many methods including variation of orifice size, of bore flow resistance, of bore hydraulic radius, of bore length, and of bore properties along their lengths. Such design results may easily be engineered using the flow modeling described here.

[0540] When the bores 544 of FIG. 49a all intercept the channels 547 and 548, clogging primarily occurs at the cavity to bore entrance. Additional clogging may occur at the entrances of the bores from the auxiliary channels. Clogging at these has a much lower probability of occurring and it is substantially lower for each successive channel. It is also found that when there are multiple channels the flow defects from clogging at a first channel may be eliminated by the presence of a second channel or a number of additional auxiliary channels.

[0541] The addition of at least one channel to a multi-orifice die produces fluid distribution performance benefits. It enables improved designs and operational advantages. FIG. 53 illustrates another useful internal channel flow passageway design. Cavity 780 supplies fluid to bores 782. Bores 782 feed into channel 784 which extends down the length of the die and parallel to the cavity 780. Channel 784 transfers fluid to bores 786 which transport it to the exterior of the die. The bores 782 and 786 do not need to have their centerlines aligned.

[0542] Here bores 786 are of larger diameter than bores 782. This dramatically increases the probability that any clogging will occur at the cavity 780 and bore 782 junction and not at the channel 784 and bore 786 junction. This is advantageous.

[0543] 6. Incorporation of Filter Media

[0544] FIG. 54 illustrates still another improved design. Here cavity 790 feeds a series of bores and channels. Bores 792, 798 and 799 are in series with channels 794 and 796. To facilitate the collection of contaminates that may clog bores, filter media is placed in channel 794. Another teaching is to introduce the media into the die at one end and remove it from the other. This may be done intermittently or continuously. The media in channel 794 may be a rope, a thread, or a rod of material that may be pulled or pushed through the die while it is being used. This concept of continuous or intermittent filter media replacement may also be used when all or some of the 792, 798 and 799 bores are replaced by die slots. By this means final filtering of the fluid may be accomplished just prior to distribution from the die or device.

[0545] FIG. 55 illustrates still another improved design. Die body 730 contains cavity 732, channel 734 and bores 736. The bores terminate at discharge orifices 738 on the face 740 of the die. The channel 742 connects the bores 736 and the cavity 732. The opening 744 between the cavity and the channel provides for flow from the cavity to the channel. The opening 746 between the bores and the channel provides for flow from the channel 742 to the bores. A tubular filter element 746 is paced within the channel so as to provide filtering of the fluid flowing into and out of the channel. The open area of the filter tube leaves a portion of the area 748 of the channel free for unimpeded fluid flow parallel to the cavity.

[0546] 7. Profile Control

[0547] With prior art drilled multi-orifice dies, the flow distribution along the length of the die is a function of the dimensionless viscous number (Nvsm). This is illustrated by the graph in FIG. 56. When the viscous number for a specific die and fluid is high, generally greater than 0.5, the flow distribution along the length will be poor. With a fixed design, there is no way to improve this without changing the die dimensionless viscous number or the fluid characteristics. Changing the die design to change the die's dimensionless viscous number requires remachining or a new die.

[0548] The flow distribution from a multi-orifice die with at least one auxiliary channel may be drastically and easily modified. It is a teaching to adjust flow distribution by blocking the flow from a cavity into at least one bore or from a channel into at least one bore for the beneficial improvement of the flow distribution. This is illustrated by FIG. 57. Presented are results for a distribution die with one auxiliary channel with the dimensionless viscous number equal to 1.0 and the dimensionless channel viscous number equal to 0.5. Here the quantity, the local flow at a position divided by the average flow from the bores, is graphed as a function of the dimensionless position down the length of the die body.

[0549] The X's on the dashed line give the flow distribution from the 201 die bores with an auxiliary channel present and the selectively blocking a number of cavity to bore entrances. The solid line is the distribution from the die with no selective bore blockages. The flows are blocked to bore numbers 1, 3, 6, 9, 1.2, 15, 18, 21, 24, 27, 30, 33, 37, 41, 45, 50, 55, 61, 67, 75, 83, 91, 99, 108, and 118 where the first bore is at dimensionless position 0 and the last at position 1.0. The distribution from the die with no selective blockages has a uniformity index of 0.162. This is beneficially changed to 0.058 for a 64.4% improvement when selective blockages are employed.

[0550] For control here a simple on and off flow control is used. It may be accomplished manually or by using bore entrance blocking mechanisms that are activated by a control system.

[0551] A plurality of actuators that are independently operable may be used to produce local changes in flow restrictions into bores across the width of the die. The actuators can be located in the cavity or external to it. The overall apparatus may be operated automatically to control the flow distribution. A sensor would be used to measure the flow distribution profile across the width die and send this information to a controller. The controller compares the measured profile to a known target profile. Corrective adjustment signals may be produced by the controller and sent to the actuators spaced down the length of the die. The actuators manipulate the flow restrictions to change the profile. Simple on-off control of the flow is preferred.

[0552] a. Optimizing the Cross Channel Location with No Clogging:

[0553] FIGS. 58 and 61 are graphs of the uniformity index for a device of the invention using one auxiliary cross channel. Uniformity index is plotted versus the position of the channel along the bore. Curves are plotted for various values of the auxiliary cross channel dimensionless number Nvcm. They illustrate that the optimum positional placement of a single auxiliary channel is at the halfway point along the length of the bores when there is no bore clogging.

[0554] b. Optimizing the Cross Channel Location with Clogging:

[0555] When there is clogging at the junction of a bore and the cavity, improved performance is obtained when the auxiliary channel is placed at less than the halfway point along the bores. This is illustrated by FIGS. 59 and 60. In these everything is comparable to FIGS. 58 and 61 except that the last two bores at the end of the die have total blockages. This is at the end of the die furthest from the cavity flow entrance position.

[0556] In the case of these blockages without a cross channel, no fluid flows from the affected bores and the flow distribution is grossly non-uniform. With a die having at least one channel, fluid flows from all the bores, and the flow distribution is adequately uniform when the die parameters for the die design are well chosen. When a single channel is used, it is preferred that the cross channel be placed at a dimensionless position between 0 and 0.4 along the bores.

[0557] The die of FIG. 49a illustrates a die with two channels and bores with a constant hydraulic diameter extending from the cavity to the orifices. Extensive flow modeling studies were performed on this geometry. These show improved distribution may be obtained when two or more channels are employed. Also, the optimum position for the first channel is at a distance of between the 0.1 to 0.3 times the length of the bores. The optimum position for the second channel is at a dimensionless position between 0.4 to 0.5. Positions are measured from the bore entrance at the cavity.

[0558] When more than one cross channel is employed to distribute flow or filter and distribute flow, we refer to the sequential order of the channels from the cavity to the orifices. The first channel is the first channel which fluid encounters flowing from the cavity. Investigation of the flow in and from the cavity, bores and channels of our improved multi-orifice flow distributing system shows that improved performance is obtained by optimizing the position of the channel. It is preferred that the channels of a multi-channeled system be positioned so that the n.sup.th channel is between the (n−1).sup.th channel and the halfway position along the otherwise uninterrupted bore length between the cavity and the (n−1).sup.th channel.

[0559] c. Improvements Using Intersecting Bores:

[0560] Known multi-orifice coating dies like those described by McIntyre, use a simple set of parallel drilled hole bores to communicate between the cavity and the discharge orifices. The presence of auxiliary channel has been shown above to be useful. Their utility arises because they allow fluid exchange between the bores. Other methods and geometries that promote fluid exchange between bores are also useful and are a teaching of this invention. Investigation has shown that performance enhancement is obtained by using intersecting bores, and by adding auxiliary cross channels to the intersecting bore geometries.

[0561] FIGS. 50a, 50b and 50c illustrate types of intersecting bore geometries that improve the distribution of flow from the orifices. These also improve the tolerance to clogging, and the ability to simultaneously filter and distribute fluid.

[0562] FIG. 50a illustrates an improved design for the flow passageways of a fluid distribution device. The heavy lines 701 and 702 represent two sets of bores. If the device is a metal die, the bores may be produced by repeated drilling at two angles other than ninety degrees to the cavity 700. It is preferred that the acute angle that bores 701 make with the cavity equals the negative of the acute angle that bores 702 make with the cavity. The bores 701 and 702 in FIG. 50a connect the discharge orifices 703 on face 704 to the distribution cavity 700. Fluid discharges from the face 704 from the bores through the orifices 703. Bores 701 and 702 intersect at least one other bore. For a uniform flow distribution from the bores along face 704, it is preferred that each individual bore intercept at least one other bore and preferably more than one. By this means, at least one flow path exists that allows fluid to flow from one bore to another bore as it flows from the cavity to the die face.

[0563] As the auxiliary cross channel modeling illustrates, the ability to exchange fluid between bores after leaving the cavity and before exiting the discharge orifices is beneficial and is a teaching of this invention. More than one exchange means are preferred.

[0564] FIG. 62 illustrates the utility of the intersecting bore design of FIG. 50a. Here two side by side bores near the center of the die have been blocked at their junction with the cavity. With a prior art die this would result in total disruption of the discharge distribution with no fluid flowing from the die face orifices-of the blocked bores. With the bore geometry of FIG. 50a the disruption is healed. The plot of local bore flow for the case where a bore intersects just on other bore between the cavity and the die face is designated “1 intersection”. Flow at the exit of the blocked bores is found to have a dimensionless flow of 0.5. For the case of three intersections the flow is 0.8, and for seven intersections the flow is 0.94. All are improvements over the prior art.

[0565] d. Intersecting Bores and Cross Channels:

[0566] FIG. 63 illustrates the utility of the intersecting bore and channel design of FIG. 50b. Here two bores have been blocked near the center of the die. With a prior art die this would result in total disruption of the outflow distribution with no fluid flowing from the blocked bores. With the bore geometry of FIG. 50b the disruption is healed. All curves show improvements over the prior art.

[0567] Further improvements to the apparatus of FIG. 50b may be achieved by adding additional bores to the geometry. An example is illustrated in FIG. 50c. Here a set of additional bores 713 are placed at a third angle of intercept with the cavity. These bores have a shorter path from the cavity to their exit orifice. Generally, the flow resistance of these bores will be lower than the others to achieve uniform flow along the die face 715.

[0568] 8. Stripe Coatings

[0569] Multi-orifice coating dies are known for their utility in coating stripes on webs. Bores are present in areas where coating is desired and absent where no fluid is desired. FIG. 50d illustrates a device where adjacent bores 717 and 718 intersect. The bores are not equally spaced. They are arranged to provide lengths 719 and 720 along the device face 722 where the discharge of fluid is respectively present and not present. When used in coating, the device produces down web stripes. Again, the bore intersections provide for more uniform flow distributions and a tolerance to clogging.

[0570] 9. Further Examples Using Layers of Bores

[0571] When lines of bores are stacked on top of each other, improvements may be obtained. FIGS. 49b, 49c and 49d illustrate examples of possible multilayered bore geometries with auxiliary intra-layer and/or inter-layer channels. FIGS. 67a and 67b illustrate improvements may be obtained with auxiliary channels, with multiple layers of bores, and with inter-layer auxiliary channels. A die of 102 bores distributes flow between two rows, one above the other has been studied. The bores are numbered from the inlet end of the cavity sequentially by alternating between the top and bottom layers. That is the number one bore is at the first position of the bottom layer, and the number two bore is at the first position of the top layer. interlayer auxiliary channels are present connecting the intersections of the bores and cross auxiliary channels in one layer to those of the adjacent bore layer. Thus in the case with 51 bores per layer, 2 layers, and 1 cross auxiliary channel per layer, there are 51 interlayer auxiliary channels.

[0572] Twenty-three bores are blocked at their cavity end in the bottom row, and twenty-three are blocked in the top row. The blocked bores have been randomly chosen. The grids of bores and auxiliary channels are present. The dominant auxiliary channels are cross channels that run perpendicular to the bores and intersect all bores in a layer. These auxiliary channels are uniformly spaced between the supply manifold and the die exit orifice face.

[0573] The bore parameters are chosen so that the ratio of the incremental flow resistance in the manifold from bore to bore to that of a bore increment length equals 0.002. Here the bore increment length is the distance between the cavity and the first auxiliary channel. The ratio of an auxiliary cross channel flow resistance per increment to the flow resistance per increment in the manifold equals 2.0 with the incremental flow resistance in the auxiliary being also from bore to bore. One auxiliary cross channel is used for the results plotted in FIG. 67a. Four auxiliary cross channels are employed for the results plotted in FIG. 67b.

[0574] FIG. 67a plots the dimensionless bore flow rate as a function of bore number with and without interlayer auxiliary bores present for the case of one auxiliary cross channel. There are 46 blocked bores that are the result of filtering and trapping contamination at the manifold bore entrances. Without the auxiliary cross channels the bore flow rates would oscillate in values between 0 and approximately 2.0. This is totally unacceptable for most coating operations. With one auxiliary cross channel in each bore layer the flow distribution is improved. The flow per bore is always greater than zero.

[0575] When interlayer auxiliary bores are present the flow distribution is further improved. In the plot the data for the presence of interlayer auxiliary channels is indicated by the plotted squares. With interlayer auxiliary channels present the data is indicated by diamonds. When only one cross channel is used, the standard deviations of the bore flows at the orifice exits are 0.54 and 0.37 respectively. The interlayer auxiliary channels improve the flow distribution,

[0576] FIG. 67b plots the dimensionless bore flow rate as a function of bore number with and without interlayer auxiliary channels present for the case of four auxiliary cross channels. The flow resistances of all auxiliary cross channels in the example of these two figures are equal.

[0577] Additionally, the same 46 bores that were blocked for FIG. 67a are blocked here. With auxiliary cross channels in each bore layer the flow distribution is significantly improved over the results in FIG. 67a.

[0578] When interlayer auxiliary bores are present the flow distribution is further improved countering the negative effects of the blockage of 45 percent of the bore entrances. In the plot the data for the presence of interlayer auxiliary channels is again indicated by the squares. With no interlayer auxiliary channels present, the data is indicated by diamonds. The interlayer auxiliary channels improve the flow distribution. The standard deviation of the bore flow rates at the orifice exits when there is no inter-layer flow equals 0.24. When there is interlayer flow, the deviation is reduced to 0.22.

[0579] When filtering a fluid in a die where all bores have equal flow resistance, it is observed that most of the trapped particles are located at the intersection of the bores with the cavity. With time more and more of the bores will be blocked, and if the flow rate to the die is constant, the pressure in the cavity will increase. In the case when operation is limited to a doubling of the supply pressure, production will be stopped when fifty percent of the bores are blocked at the cavity with a prior art die.

[0580] With prior art drilled dies, clogging of a bore causes drastic flow uniformity defects at the exit from the die. Some post coating improvement methods are known to hide this problem. These all include methods of redistributing the fluid after it has left the die. However, when the filtering of particles clogs bores to the point of reaching pressure limits, production must stop. Because of the flow maldistribution prior art dies have not been successfully used to both filter and distribute flow.

[0581] The following additional examples of modeling investigations further illustrate the utility of using multi-orifice devices with auxiliary channels.

EXAMPLE 1

Auxiliary Inter-Layer and Intra-Layer Channels and Distributed Filtration)

[0582] A coating die is used to distribute fluid to a die face. It consists of a cavity feeding three layers of bores stacked upon top of each other. The cavity is four units long and has a flow resistance per unit of 0.002. From the cavity the bores direct fluid to the die face in each bore layer four bores connect to the discharge orifices on the die face. The bores are interconnected by auxiliary intra-layer cross and inter-layer cross channels with a geometry as illustrated in FIG. 49d. The four rows of cross channels are used, and the dimensionless parameters for the passageway layout are Nvm=0.024, Nvim=1.2E+08 and Nvcm=0.006.

[0583] A fixed volume of fluid is processed through the die at a constant rate. This fluid although carefully prepared contains eleven contaminant particles that are larger than some of the bore sizes. The probability of capturing the eleven unwanted particles of a fixed diameter from a set volume of fluid is adjusted for each segment of each bore. Additionally, the die is set so all eleven particles are captured within the die with a probability of 1. That is the probability of a particle escaping from the die is 0.

[0584] After flowing the test volume of fluid through the die, four particles are trapped at cavity to bore intersections. Their distribution is one at the first layer of bores, two at the second layer, and one at the third. With the bore location with respect to distance along the cavity numbered as 1 through 4, the particle locations are at positions 1, positions 2, and position 3 for the three layers respectively.

[0585] Four particles are trapped at the first auxiliary cross flow channel and bore intersections. Their distribution is two at the first layer of bores, one at the second layer, and one at the third. The particle locations are at positions 3 and 4, position 1, and position 2 for the three layers respectively.

[0586] Three particles are trapped at the second auxiliary cross flow channel and bore intersections. Their distribution is one at the first layer of bores, one at the second layer, and one at the third. The particle locations are at position 1, position 2, and position 3 for the three layers respectively.

[0587] Just after the last particle is trapped the dimensionless pressure drop through the die is 0.37. The flow from the first layer of bores is 0.000, 0.480, 0.450 and 0.417 respectively for bore locations 1 through 4. From the second layer of bores the flow is 0.445, 0.000, 0.450 and 0.417 respectively. From the third layer of bore orifices the flow is 0.445, 0.479, 0.000 and 0.417 respectively.

[0588] The dimensionless sum of the flows from all three layers at each of the four bore locations along the cavity is 0.890, 0.959, 0.900 and 1.25 respectively. The standard deviation of this population is 0.15.

EXAMPLE 2

[0589] A coating die is used to distribute fluid to a die face. It consists of a cavity feeding three layers of bores stacked upon top of each other. The cavity is four units long and has a flow resistance per unit of 0.002 k. From the cavity the bores direct fluid to the die face. In each bore layer, four bores connect to the discharge orifices on the die face. The bores are interconnected by auxiliary intra-layer cross and inter-layer channels with a geometry as illustrated in FIG. 49d. The rows of cross channels is four, and the dimensionless parameters for the passageway layout are Nvm=0.024, Nvim=1.2E+08 and Nvcm=0.006. The same volume of fluid used in example 1 is used here and it also contains 11 unwanted particles of a fixed diameter. However in this example, the bore diameters are fixed a value smaller than that of the particles. In this case, the bore cavity intersections act as absolute filters for the particles with a zero probability that any will flow into the bores. After flowing the test volume of fluid through the die, the eleven particles are trapped at cavity to bore intersections and block eleven of the twelve bore entrances.

[0590] Just after the last particle is trapped, the dimensionless pressure drop through the die is 1.31. The flow from the first layer of bores is 0.360, 0.513, 0.287 and 0.173 respectively for bore locations 1 through 4. From the second layer of bores the flow is 0.360, 0.513, 0.287 and 0.173 respectively. From the third layer of bore orifices the flow is 0.360, 0.513, 0.287 and 0.173 respectively.

[0591] The dimensionless sum of the flows from all three layers at each of the four bore locations along the cavity is 1.08, 1.53, 0.86 and 0.52 respectively. The standard deviation of this population is 0.37.

EXAMPLE 3

Potential Prior Art

[0592] A coating die is used to distribute fluid to a die face. It consists of a cavity feeding one line of twelve bores. The cavity is twelve units long and has a flow resistance per unit of 0.002. From the cavity the bores direct fluid to the discharge orifices on the die face. No auxiliary channels are present.

[0593] A fixed volume of fluid is processed through the die at a constant rate. This fluid contains eleven contaminant particles that are larger than the bore sizes. The probability of not capturing within the die the eleven unwanted particles from the set volume of fluid is zero.

[0594] After flowing the test volume of fluid through the die, the eleven particles are trapped at cavity to bore intersections. This leaves only the last bore open. All others are clogged. The dimensionless flow from the last bore orifice is 12, and just after the last particle is trapped the dimensionless pressure drop through the die is 3.00. The standard deviation of flow from this row of bores is 3.31.

EXAMPLE 4

Auxiliary Intra-Layer Channels and Distributed Filtration

[0595] Using identical conditions to Example 1 except where no interlayer flow was allowed, the test was run. Here then the parameter Nvim=0 was used for the filtration modeling experiment. As with Example 1, the particles were trapped at the intersections of channels and bores, and the cavity and the bores. They were trapped in the same locations found in Example 1.

[0596] In this case just after the last particle is trapped the dimensionless pressure drop through the die is 0.502. The flow from the first layer of bores is 0.000, 0.637, 0.174 and 0.058 respectively. From the second layer of bores the flow is 0.293, 0.000, 0.551 and 0.599 respectively. From the third layer of bore orifices the flow is 0.626, 0.367, 0.000 and 0.692 respectively.

[0597] The dimensionless sum of the flows from all three layers at each of the four bore locations along the cavity is 0.92, 1.00, 0.73 and 1.34 respectively. The standard deviation of this population is 0.22.

[0598] From the examples it is found that when the filtration, the trapping of particles, is distributed within the system of bores and auxiliary channels rather than concentrated at the cavity to bore, unctions, the pressure drop is lower. Also the standard deviation of the flow at the die face is lower. This is an improvement. Another finding is that multiple layers perform better than a single layer of bores. It is preferred that multiple layers of bores are employed with auxiliary inter-layer channels. More preferred are devices including multiple layers of bores with both inter-layer and intra-layer auxiliary channels.

[0599] It is a teaching that for any line of bores intersecting a cavity or an auxiliary channel, the probability of trapping a contaminant at the intersections may be adjusted by design. It is often beneficial to use a population of bores with non-uniform capture probabilities. A simple method of adjusting the probability for a bore is to manipulate the dimensions of the bore. This may be accomplished while simultaneously maintaining the flow resistance at a desired level. Additionally, it has been found that flow resistance variations may be made unimportant with the use of auxiliary channels, and acceptably uniform flows at the die face may be achieved to meet a desired precision.

[0600] When extremely high value products are coated, the presence of even one particle may make a large surface area defective. An example is a large format liquid crystal video display screen. One defective pixel makes the whole screen defective. A single particle in one of the multiple layers comprising the screen may create this defect. It is preferred when coating such a high value product to first double filter the fluid before entrance to a coating die. Additionally, the die should act as an absolute final filter as well as a flow distribution device.

[0601] F. Improved Slots for Fluid Distribution

[0602] I. Slot and Cavity Systems

[0603] Although slots have significant negative features, they are widely used for fluid distribution. This invention identifies desirable improvements. Existing slots have a gap that extends in two dimensions. The gap generally is constant or slowly varies along an axis.

[0604] It has been found that having the flow resistance vary in two directions may be used to improve distribution. FIG. 89 illustrates a cross-section of such a slot. Slot 890 conducts fluid from a cavity where the fluid enters as indicated by arrow 893. Fluid exits the slot at a slot orifice as indicated by arrow 894. The slot is modified by grooves 892 that extend across the dominant flow path from the cavity to the orifice. These small grooves provide dramatically reduced flow resistance to fluid movement in a direction parallel to the orifice and cavity centerlines extending down their lengths.

[0605] As shown the grooves are placed uniformly at a high frequency along the flow path from cavity to orifice. Preferred spacing is on the order of a fraction of the nominal slot gap 895 or the average slot gap to upwards to ten times the gap. It is preferred that they are uniform in the direction perpendicular to the dominant flow direction.

[0606] A key feature of the improve slot geometry is that its flow resistance in the dominant flow direction is larger than that in the direction of the grooves. This is characterized by the ratio of the resistance in the orifice centerline direction to the resistance in the depth or nominal flow direction from the cavity to the orifice. This ratio is symbolized by Rw/Rd or the equivalent conductance ratio Cd/Cw.

[0607] FIG. 93 illustrates the dramatic improvement in the uniformity of flow from the slot orifice of a cavity fed slot. It is plotted for Newtonian fluids. The dimensionless parameter Nsp here is defined as the width of the die (length of the slot orifice) divided by the distance from the cavity to the slot orifice all times the square root of the ratio of the slot flow resistance parallel with the orifice to the flow resistance toward the orifice. Here the uniformity index, UI, is plotted versus the parameter Nsp for various values of the ratio of the viscous flow resistance down the cavity to the flow resistance thru the slot, Nvs. A small UI is desirable. It is improved if the ratio Cd/Cw is less than one when all other variables are constant. Reducing the parameter Nsp by reducing the ratio produces desirable orders of magnitude reductions of the UI.

[0608] Modifying grooves may be placed in both walls of the slot. Grooves are generally produced by subtractive, material removal techniques.

[0609] Material additive techniques may be used to produce two dimensional flow properties for the. FIG. 91 illustrates one of many such modifications with top and side views. A section of one slot wall surface 912 is shown having a width in a direction parallel the centerline of the orifice exit of the slot and a depth in the dominant flow direction from the cavity to the orifice. Arrow 910 indicates the direction of flow. From the wall surface 912 projections 911 extend into the flow. These projections have lengths that extend across the dominant flow direction, and much smaller dimensions in the dominant flow direction. They are designed to obstruct flow more in the depth direction than in the width direction. By this means the ratio Rd/Rw is reduced below one.

[0610] Any technique that reduces the ratio Rw/Rd for the slot is useful. Any wall or internal modification that reduces the ratio Rw/Rd is useful. The slot wall modifications may be produced by machining, knurling, electric discharge machining, etching, plating processes, diamond turning, micromachining, photometric replication processes, micro-replication process, and other techniques. Additive or subtractive techniques may be used. Slots may be modified by placing material within them.

[0611] In dies and coating devices where a slot is fed by a constant area cavity, it is taught that improved outflow uniformity is achieved when the dimensionless number, Nvs, is kept below 1, preferably below 0.1, and most preferably below 0.04; and additionally the profiled slot has a dimensionless parameter value Nsp of less than 1.

[0612] Both types of modifications also improve the particle trapping ability of the slot. Surprisingly, both also, in terms of orifice out flow uniformity, improve the tolerance to clogging of the slot by contaminants. This allows these improved slots to be used for final filtration of the fluid just prior to discharge from the slot orifice.

[0613] 2. Point Fed Slots Systems

[0614] Slots terminating with a slot orifice on the face of a distribution device are useful. In the preceding section the feeding of fluid to the slot from a cavity through an inlet slot edge is described. A system of improved simplicity uses a slot which terminates at a slot orifice and which is fed fluid at a single point at an inlet edge. It has been found that modifying the slot flow resistances in the depth and width directions independently is useful for obtaining discharge uniformity. It is preferred to have a ratio Rd/Rw less than one.

[0615] FIG. 94a illustrates a top view of a slot fluid distribution system. The slot 9400 profile is a straight sided polygon. Fluid is fed into it at an apex as indicated by arrow 9430, and it exits at a slot orifice at an exit edge 9410. The exit flow is indicated by the arrows 9420. Boundary edges 9412 and 9414 do not intersect the outlet edge.

[0616] FIG. 94b illustrates a top view of a slot fluid distribution system. The slot 9441 profile is a rectangle with a width, W, and a depth, d, indicated by double ended grey arrows. Fluid is fed into it at an apex on the boundary opposite the outlet edge as indicated by arrow 9446 or it is fed at a point between the opposite boundary and the outlet edge as indicated by arrow 9447. Alternatively, fluid may be fed into the slot at any point in the interior of the rectangle as illustrated by arrow 9448. Fluid exits at a slot orifice at an exit edge 9442. The exit flow is indicated by the arrows 9440.

[0617] When feeding is at an interior point the rectangular slot profile may be divided into two rectangular sub-regions both with a depth of d. The left sub-region rectangle has a width WW and the right has a width WWW indicated by double ended grey arrows.

[0618] Simple rectangular slot geometries without feed cavities have utility. A sampling of extensive point fed slot modeling results is graphed in FIG. 90a. Results for a rectangle where the width to depth ratio is 3 to 1 are labeled “rect(3×1)”. Results for a rectangle where the width to depth ratio is 1 to 1 are labeled “squ(1×1)”. Results for right triangles where the width to depth ratio is 1/1, 1/2, 1/3, and 1/4 are labeled “1/1 tri”, “1/2 tri”, “1/3 tri” and “1/4 tri” respectively.

[0619] Here the UI of the flow from the exit edge is plotted. It is found that all rectangular slot systems fed from an inlet edge corner may be described by a single curve of UI versus Nsp where NSP=(W/d) (sqrt(Rw/Rd)) and contains the ratio Rw/Rd. The same is true for all right triangles where the slot exit is not along the hypotenuse side. For the triangles the feed point is the apex opposite the exit side; the width is the exit length and the depth is the perpendicular distance of the entrance point to the exit edge.

[0620] Most slots covering a polygonal area and when fed at an apex have UI results that fall between the triangle and rectangle results. In many cases, the results for polygons may be estimated from combinations of triangle and rectangle results.

[0621] In all cases improvement in UI is obtained over conventional slots when the ratio Rd/Rw is less than one. Useful UI's are obtained when the parameter Nsp is less than one. More preferred are values less than 0.5

[0622] When conventional slots where Rd=Rw are employed, useful flow discharge uniformity is achieved using by selecting geometry variables so that the parameter Nsp is less than one. More preferred are Nsp values less than 0.8. This is especially true for corner fed rectangular slots.

[0623] Multi-point feed geometries may be used to span a wide width. This design strategy is described in the later description section: “Improved Metering Sheets”

[0624] Similar outflow uniformity variation curves are obtained when the slot is fed at an interior point between a boundary edge and the outlet edge. UI may be conservatively estimated using a revised parameter, Nspi, in place of Nsp and the FIG. 90a graph. Here the value of Nspi is calculated from the equation

Nspi=(WWW/dd)sqrt(Rw/Rd)

[0625] where WWW is the larger of the two dimensions, WWW and WW.

[0626] When the feed point at a ratio dd/d=0.1, the uniformity is predicted by the graph in FIG. 90b. This graph again illustrates that changing a slot's flow resistance from omnidirectional where Rw/Rd=1 to bidirectional where Rw/Rd is less than 1 improves the outflow uniformity. In addition, it illustrates again that uniformity of outflow may be obtained when the dimensionless parameter, Nspi is less than one.

[0627] G. Improved Porous Sheets

[0628] 1. Convention Porous Sheets

[0629] Useful for investigating porous sheets are fluid dynamic simulations that describe the operational characteristics. We have developed models that describe the flow distribution characteristics of porous media in three dimensions.

[0630] Conventional porous sheets are composed of a large collection of pores. They allow for the transport of fluid and the filtering of fluid. They are characterized as having in any individual sheet a wide distribution of resistances to fluid flow from point to point. That is there is great variation in the pore lengths, sizes, and the inter-pore connectivity. FIG. 1a illustrates the typical flow distribution die using a known porous sheet in place of a slot. Fluid is fed into the die cavity 1 by a means not shown. The die has a top plate 5 and bottom plate 6 which are assembled with a slot gap 3 extending from the cavity 1 to the exterior of the die. The gap 3 is filled with porous sheet 2. Fluid flows through the cavity then through the porous sheet and exits the die along a line.

[0631] FIG. 1b is a photograph of a cross section of commercially available porous media. The pores have a distribution of sizes and shapes with random connectivities. The black areas 7 are pores. The grey areas 8 are solid material.

[0632] Using flow modeling one may investigate the point to point uniformity for flow from the cavity to the discharge face of the die and from one end to the other along the length of the cavity for a die like illustrated in FIG. 1a. FIG. 1c is a grey scale contour graph of the ratio of the local average flow rate at a point to the average flow rate for the whole sheet in the depth direction from a feed cavity to die exit. In this case, a distribution of pore sizes is present where the ratio of the flow resistance down the length of cavity to the total sheet flow resistance is 0.01. This graph points to the failing of common porous media. It has non-uniformities that create large local flow variations.

[0633] The failing is even more apparent in FIG. 1d where the point to point flow at the exit edge of the porous sheet is graphed. The ratio of the local flow to average flow is plotted as a function of the distance along the die exit face. The flow is very non-uniform. Variations greater than plus or minus 40 percent are present. In coating this will often create a defective product.

[0634] However in some cases, this point to point non-uniformity is not material but average uniformity over length on the order of centimeters or multiples of the sheet thickness is significant.

[0635] 2. Improved Conventional Porous Sheets

[0636] a. Porous Sheet and Cavity Systems

[0637] Conventional porous sheet material made from sintered metal powder has nominally no directional variation of flow resistance in the plane of the sheet. The flow resistances in any two directions parallel to the major surfaces are on the average not different. It is a teaching of this invention to improve the sheet's ability to distribute flow uniformly by making the resistances different.

[0638] FIG. 92 illustrates an example of an improved sheet. The sheet 9200 is made from granular material and is indicated by the texturing. The partial section has flow across it from front to back in the depth direction as indicated by the arrow 9230. Periodically placed along the top major surface are volume regions 9240 which serve to modify and make different the flow resistances in the depth and width directions. These regions generally extend into the sheet. It is also a teaching that they extend through the sheet. The spacing is indicated by the distance L1. It is preferred that it range from a fraction of the sheet thickness to multiple times the thickness. It is preferred that the width dimension of the region in the depth direction range from a fraction of the sheet thickness to multiple times the thickness.

[0639] If the regions 9240 are void of material, the flow resistance ratio Rw/Rd will be less than one. If the regions 9240 contain non-porous material, the flow resistance ratio Rw/Rd will also be less than one. The same effect may be obtained by using less porous or more porous structure of appropriate dimensions.

[0640] If the feature spacing distance L1 is many times smaller than the total sheet depth dimension and the flow from the exit distribution edge is averaged over lengths on the order of the sheet thickness, the effect of parameters on out flow uniformity is shown in FIG. 86. This figure illustrates the case of a flow distribution device using a cavity to feed fluid to the inlet edge of a rectangular sheet of the improved conventional porous media.

[0641] FIG. 86 illustrates the dramatic improvement in the uniformity of flow from exit of a cavity fed media. It is plotted for Newtonian fluids. Here the uniformity index, UI, is plotted versus the parameter Nsp for various values of the ratio of the viscous flow resistance down the cavity to the flow resistance thru the media, Nvs. A small UI is desirable. It is improved if the ratio Rw/Rd is less than one when all other variables are constant. Reducing the parameter Nsp by reducing the ratio produces desirable orders of magnitude reductions of the UI.

[0642] When the porous media is fed by a constant cross-sectional area cavity it is preferred that the parameter Nvs have a value below one.

[0643] b. Point Fed Conventional Porous Media Sheet Systems

[0644] Slots filled with porous media and terminating with a slot orifice on the face of a distribution device are useful. In the preceding section the feeding of fluid to the media from a cavity through an inlet slot edge is described. A system of improved simplicity uses a media sheet which terminates at an outlet edge and which is fed fluid at a single inlet point. It has been found that a sheet where the flow resistance in the depth and width directions is different is useful for obtaining flow discharge uniformity. It is achieved by having a ratio Rd/Rw less than one.

[0645] FIG. 94 illustrates a top view of a slot fluid distribution system. It also may represent a porous sheet fluid distribution system. The sheet 9400 is a straight sided polygon. Fluid is fed into it at an apex as indicated by arrow 9430, and it exits at a slot orifice at an exit edge 9410. The exit flow is indicated by the arrows 9420.

[0646] A simple rectangular porous sheet geometry exemplifies the utility and the design of point fed systems. A sampling of extensive modeling of results are graphed in FIG. 90a. Here the UI of the flow from the exit edge is plotted. It is found that all rectangular slot systems may be described by a single curve of UI versus Nsp where Nsp contains the ratio Rd/Rw. The same is true for all right triangles where the slot exit is not along the hypotenuse side. For the triangles the feed point is the apex opposite the exit side; the width is the exit length and the depth is the perpendicular distance of the entrance point to the exit edge.

[0647] Most sheets, whose top side view is a polygon, have UI's falling between the triangle and rectangle results. In many cases, the results for polygons may be estimated from combinations of triangle and rectangle results.

[0648] In all cases improvement in UI is obtained over conventional media sheets when the ratio Rd/Rw is less than one. Useful U's are obtained when the parameter Nsp is less than one. More preferred are values less than 0.5.

[0649] Multi-point feed geometries may be used to span wide widths. This design strategy is described in the later description section: “Improved Metering Sheets”.

[0650] The flow in and from porous distribution sheets is analogous to flow in slots. The conclusions reached for slot apply to porous media sheets.

[0651] H. Improved Metering Sheets

[0652] A fluid metering sheet of this invention provides a desired fluid flow distribution out from and along an edge of the sheet. The fluid is forced through the sheet from one edge to another. The fluid flows within the sheet between its top and bottom surfaces. Or in the case when the sheet itself does not confine the flow, flow is confined by top and bottom confining surfaces or capping films or walls. Flow may be within the flow passageways defined by positioning the sheet between confining solid surfaces. The flow from an edge is on average normal to this edge. This sheet may be used as a substitute for fluid passageways such as precision metering slots and drilled holes. The sheets are useful in conjunction with distribution devices and coating dies.

[0653] While it is commonly desired that the outflow from a sheet edge be uniform along its length, it is within the scope of this invention to provide non-uniform, prescribed distributions. It is a teaching that the internal passages be designed and located to achieve a desired outlet edge flow distribution. The flow distribution from the edge may be controlled by the internal flow passageway dimensions or any parameter that changes the local effective resistance to flow. Passages may be designed to provide desired flows using various techniques including but not limited to flow simulation, designed experiments, flow tests, trial and error testing, and fluid flow computer studies. Each individual passageway is designed and specified.

[0654] The flow discharge edge of the sheet may be along a straight line or a curved line. The edge may also be described by a combination of segments of straight and curved lines.

[0655] The fluid metering sheet is bounded by the discharge edge and at least one additional edge. The dominant flow is into an inlet edge and then out a discharge edge. The ratio of the inlet edge length to the outlet edge length may range from greater than one to near zero. Sheets may be designed to distribute a prescribed outflow distribution from the outlet edge even when the inlet is a single point. Commonly in coating devices, the desired outflow distribution is uniform outflow.

[0656] A fluid metering sheet of this invention is useful for distributing flow along a width (length) and across a depth of sheet. The width is commonly the nominal width of a distribution device such as a coater. For distribution of fluid along a straight line, the sheet will generally be rectangular. The sheet has edges, and it has two major surfaces. In the case where a sheet is oriented horizontally, the surfaces constitute its top side and its bottom side.

[0657] In the following description “uniform flow” from a fluid metering sheet along its discharge edge is understood to mean that the average flow rate per unit length along the discharge edge is uniform within acceptable tolerance limits. The appropriate sampling length is one to several times the largest dimension of a sheet base unit cell which is defined below or the sheet thickness.

[0658] “Uniform flow resistance” in a metering sheet is understood to mean that if the pressure gradient is uniform, then the resulting average flow rate of a fluid is uniform within acceptable tolerance limits. The appropriate sampling length is at least as large as the largest dimension of a base unit cell or the sheet thickness.

[0659] The term passageway flow conductance” refers to the reciprocal of the passageway flow resistance. At a given flow rate through the passageway, the resistance equals the flow rate divided by the pressure gradient from one end to the other of the passageway.

[0660] In the following description, the terms “metering sheet base unit cell”, “base unit cell”, “unit cell”, and “cell” refer to a small fraction of the total metering sheet volume. It has a thickness that equals the thickness of the sheet, or if the sheet is multilayered it has a thickness equal to the layers. The surface area is a fraction of the sheet surface area, and this area has a shape which in most cases is chosen from a group including but not limited to quadrilaterals, triangles, and hexagons. Its solid material and void space geometries are the basis for constructing portions of the sheet. For a sheet with multiple layers of 2D passageway grids, the base unit cell's three dimensional geometry repeats throughout functional areas of the sheet.

[0661] For purposes of clarity the term “sheet width” refers to the length of the fluid inflow or out flow edge for a rectangular sheet. When the sheet is rectangular and used within a coating die, the sheet width will generally be the length of the feed cavity. It nominally equals the width of coating deposited upon the substrate being processed.

[0662] The term “sheet depth” refers to the distance from the sheet inflow edge to its outflow edge.

[0663] When a sheet is composed by a collection of quadrilateral unit cells, each cell is surrounded by four other similar cells. Each of the four sides of the cell is intersected by at least one void, a flow passageway, which interconnects with flow passageways of the abutting cells. Each flow passageway in a cell interconnects with at least one other void volume flow passageway within the cell. It is preferred that it connects with all other passageways in the cell. It is preferred that each cell edge intersects a passageway of the cell.

[0664] In the case where the sheet is constructed from a collection of triangular unit cells, each cell is surrounded by three other similar cells. Other arrays of repeating shapes are also possible.

[0665] 1. Sheets Using 2-Dimensional Cell Structures

[0666] In FIG. 2 an active fluid metering sheet 10 of the invention is illustrated which is made of a layer of polymeric material that has an internal structured void space (not shown). This sheet is rectangular with a width indicated by the arrow 14 that is longer than its depth which is indicated. by the arrow 16. The thickness of the sheet is indicated by the arrow 11. When used in a coating die the width 14 of the sheet will approximate the width of the active portion of the coating die. The depth dimension 16 will span from an inlet edge to an outlet edge. In the case of coating dies the depth is generally from the die cavity to the die exit face.

[0667] One embodiment of sheet 10 of this invention will be composed of a continuum of multiple, uniform and identical base unit cells. It is a useful sheet and may be designed to produce uniform outflow. Illustrated in FIG. 3 is an example of a single unit cell 22. Shown are top, front and side views. These cells need not be discrete physical elements, but they are the smallest subdivision of the volume of the sheet that can be used to characterize the flow paths in the sheet. The sheet consists of a repeating internal structure. These cells repeat throughout the width and depth of the sheet 10 of the FIG. 2.

[0668] The cell 22 is composed of a block of material through which two intersecting fluid flow passageways 24 and 26 pass. The flow passageways intersect. The block has eight sides which are a top and bottom, left and right, and front and back. It has a top surface and a bottom surface and four edges (left, right, front, and back). The thickness of the cell equals the thickness of the sheet. Throughout the sheet adjacent cells are orientated edge to edge; front to back and left side to right side. In this manner, each of the flow passageways of a cell intersects and connects with the flow passageways of the two adjacent cells orientated along the passageway axis. Each passageway is also connected to the other adjacent cells via an angled intersecting passageway. This results in a sheet where flow cells are fluidically connected to the surrounding adjacent cells. This allows fluid from any unit cell to flow edgewise through the sheet to any other cell. It allows fluid to flow from any flow passageway to any other flow passageway with in the sheet.

[0669] Preferred are grids of passageways that confine the flow within the sheet and that do not allow flow through the major sheet surfaces.

[0670] An example sheet consisting of the cell structure of FIG. 3 may be characterized by a first set of uniformly spaced flow passageways extending parallel to the width direction and a second set of equally spaced flow passageways extending parallel to the depth direction. The flow passageways of the first set intersect flow passageways of the second set. Flow through the sheet is parallel to the top and bottom of the sheet. In general flow is directed from an input edge to an output edge.

[0671] FIG. 4a shows a magnified top view representation of a region of this sheet taken halfway through the sheet's thickness. The sheet is composed of many unit cells 44 of the type shown in FIG. 3. The bold lines 42 and 43 represent intersecting flow passages created by the repeating unit cells. Flow passageways 42 run parallel to the depth edge 38 and intercept the width edge 37 of the sheet. In a like manner, the flow passageways 43 intercept the end of the sheet along a depth edge 38. The flow of fluid is into a first input widthwise edge 37 as indicated by arrow 35. The flow exiting at the discharge edge of the region is indicated by arrow 36. The grid of flow passageways in this figure is also referred to a “square grid” flow passageway layout or square grid.

[0672] When a simple metering sheet design is desired for uniform flow distribution along a straight line, the square grid design is preferred. The utility of a metering sheet with a square grid and of the cells illustrated in FIG. 4a is that when fluid is force into it across its width, the sheet facilitates the production of uniform flow from a discharge edge. It will do this even if it filters contaminants from the fluid and traps them at the inflow edge or excludes them from passage through the sheet. Surprisingly, it will accomplish uniform outflow even when a large portion of the passageways are clogged.

[0673] Preferred characteristics of a fluid distribution sheet are:

[0674] The flow may proceed in at least two directions in the plane of the sheet in the cells. Stated in another way the sheet contains void volume flow passageways which allow flow toward each of the four edges of the sheet.

[0675] 2. Cells in an active flow area have identical flow passageways within the limits of machining precision.

[0676] 3. Individual flow passageways of a cell interconnect with at least one of an adjacent cell.

[0677] 4. No flow passageway is dead ended.

[0678] 5. The conductance of passageways and the grid layout of the passageways are fixed by design and specified.

[0679] For the purpose of metering fluid flow uniformly, it is also preferred that the sheet is manufactured precisely with the flow passages in each unit cell having a specific orientation having a uniform resistance to fluid flow. The cell characteristics described produce the desired precision flow metering characteristics for a useful metering sheet. Methods of manufacture are described in a later section.

[0680] When the sheet is not uniform in depth but uniform flow from a lengthwise edge is desired, the internal flow passageway flow resistance may be adjusted or prescribed to produce uniform flow from the flow exit edge. Here key preferred characteristics are:

[0681] 1. The flow may proceed in at least two directions in the plane of the sheet in the cells.

[0682] 2. Cells have mathematically similar flow passageways.

[0683] 3. Individual flow passageways of a cell interconnect with at least one of an adjacent cell.

[0684] 4. No flour passageway is dead ended.

[0685] 5. The conductance of passageways and the grid layout of the passageways are designed and specified.

[0686] All materials are elastic and deform in response to applied forces. External forces applied to the sheet may be used to deform the internal dimensions and change the flow resistance locally in the sheet. Energy may be applied locally to the sheet to change the flow resistance locally in the sheet. Sheets may be fabricated from compressible materials or with compressible structures. Their internal fluid flow resistances may be locally adjusted through the application of forces.

[0687] It is a teaching of the invention to adjust the local internal flow resistance by design prior to or during its manufacture. It is a teaching to adjust its local internal flow resistance by applying forces or energy during its use.

[0688] When the sheet is for the purpose of providing a prescribed flow variation from a discharge edge, the internal flow passageway conductance may be controlled to produce the desired flow from the flow exit edge. Here key preferred characteristics are:

[0689] 1. The flow may proceed in at least two directions in the cells in the plane of the sheet.

[0690] 2. Cells have similar flow passageways.

[0691] 3. Individual flow passageway of a cell interconnects with one of an adjacent cell.

[0692] 4. No flow passageway is dead ended.

[0693] 5. The conductance of passageways and the grid layout of the passageways are designed and specified.

[0694] FIG. 4b shows a magnified top view representation of a section for a sheet taken halfway through the sheet's thickness and wherein the unit cells are not equal. The sheet is composed of many unit cells of the type similar to those in FIG. 3 and delineated by the dashed lines 45. The bold lines 46 and 47 represent intersecting flow passages created by the multiple base cells. Each cell is a quadrilateral in cross section, but not equal in detail in size or shape. Such a sheet may be useful in distributing flow along a discharge edge or filtering particles from a flow.

[0695] FIG. 4c illustrates another metering sheet structure of this invention with top and side views of the sheet. Here the top and bottom solid surfaces are held in fixed proximity to each other by internal cylindrical columns 49. Again the unit cells are delineated by the dash lines 48.

[0696] The sheets of FIGS. 4b and 4c are each designed so that the flow passageways void space is specified and connect so that fluid may flow from one to any other. No dead ended flow passageways exist when fluid flows through the sheet edges. Of course, if any sheet edge is blocked by any exterior placed object, there will be local dead ended flow passageways at that edge. But the whole sheet has local flow conductances that are predetermined by their design, and fluid may move from any local void volume to any other in response to pressure distributions and forces.

[0697] FIG. 5 illustrates the internal flow passageways of a die similar to FIG. 1, but here the known porous sheet is replaced by an improved metering sheet 56 using the flow passageway pattern of FIG. 4a and the cell structure of FIG. 3. A line of flow passageways 50 is exposed along the sheet exit edge 54 and along the depthwise edge 55 of the sheet. Fluid enters the end of the cavity 58 as shown by the arrow 59. It then flows down the length of the cavity and simultaneously enters the fluid metering sheet 56 through an inlet edge. It exits through the exit edge 54 from exposed flow passageway ends as indicated by arrows 52. Multiple cross flow passageways intersect the discharge flow passageways extending across the sheet depth. Their ends 51 intersect at the edges 55 and 53 of the sheet. Flow from the ends of the sheet is prevented by solid die structure (not shown) confining the ends of the sheet when it is assembled in the coating die. Sealing of these ends by various means is also a teaching of the invention.

[0698] FIGS. 3 and 4a illustrate only one useful cell structure and resulting sheet flow passageway configuration. Many others are possible. Examples are shown in FIGS. 6a and 6b which illustrate a magnified top view representation of a section of:the sheet taken halfway through the sheet's thickness. The sheet is composed of many unit cells. The bold lines in FIGS. 6a, 61 and 62, represent intersecting flow passages created by the repeating unit cells. A single unit cell is identified by the dotted lines 63. Flow passageways 61 and 62 intersect at acute angles. The bold lines in FIGS. 6b, 67 and 68, represent intersecting flow passages created by the repeating unit cells. A single unit cell is identified by the dashed lines 60. Flow passageways 67 and 68 intersect at right angles. Both the patterns of these figures are characterized as having at least two planes of symmetry.

[0699] Arrow 64 indicates the direction of flow through the depth of the sheet. Away from depth wise edges the resistance to flow across the sheet from side 66 to 65 in FIG. 6a is uniform. Also away from edges, the resistance to flow across the sheet from side 70 to 69 in FIG. 6b is uniform.

[0700] It is preferred that all flow passageways of a metering sheet are interconnected. With this compensation for any obstruction in any flow passageway within the sheet can occur. That is flow can divert around any point of obstruction in a flow passageway of the sheet. Additionally, clogging at an inflow sheet is overcome by the sheets ability to direct fluid around a clogged section. By this means, the sheet can provide filtering while producing an effectively uniform distribution of fluid from its discharge edge.

[0701] FIGS. 7a and 7b illustrate magnified top and side views of a preferred metering sheet structures created by columnar posts 76 extending from the bottom of the sheet. The sheet 77 is composed of unit cells one of which is delineated by the dotted lines 75. These cells repeat throughout the depth and width of the sheet. The cell has eight sides which are a top and bottom, left and right, and front and back. The thickness of the cell equals the thickness of the sheet. Throughout the sheet, adjacent cells are arranged front to back and left side to right side. In this manner, for each cell the void volume of a cell connects with the voids of all adjacent cells. This results in a sheet where the flow cells are connected, and this allows fluid from any cell to flow to any other cell. It allows fluid to flow from any flow passageway to any other flow passageway. The entire sheet consists of uniformly spaced cells with equal void volumes.

[0702] While uniformity in void volume and equal spacing of cells has utility in producing uniform metered flows, non-uniform spacing and volumes are also a teaching of this invention. As in the case of linear grids of flow passageways illustrated in FIG. 4b, the quadrilateral base cell shape and size need not be uniform throughout the sheet.

[0703] The sheet in FIG. 7a is one which is intended to be covered (capped) by a solid surface on its top side to contain fluid within its thickness. The sheet of FIG. 7b has an integral top impervious surface so that fluid flow is contained within it. Both types are sheets of this teaching.

[0704] FIGS. 8a, 8b and 8c illustrate other flow passageway grid configurations for metering sheets. Each is intended for use as a device to distribute flow along a discharge length. The arrows 82, 85 and 86 indicate the intended direction of flow of fluid through the sheets. The basic unit cells are identified by the dashed lines 81, 84 and 87. In each case two or more flow passageways extend across the cell. The void volume of a cell interconnects with the volumes of all adjacent cells.

[0705] FIG. 9a illustrates portion of a useful metering sheet 922 where the base structure is described by a more internally complex unit cell indicated by the dashed lines 920. This sheet is designed for metering flow in the direction indicated by arrow 930. Unit cells fill the width and depth of the sheet except near the edges 932 and 934 where incomplete unit cells exist. Again here the flow passageways are indicated by heavy dark lines 940.

[0706] FIG. 9b illustrates another useful metering sheet grid layout of this invention. The sheet is intended for use as a device to distribute flow along its discharge edge 910 which is perpendicular to the direction of flow indicated by arrow 902. The basic unit cell is identified by the dashed lines 901. In each cell in the sheet at least two flow passageways extend across the cell and interconnect with adjacent cells. Each cell consists of flow passageways which are angled to each other. They allow flow both toward the depthwise and widthwise edges. Any point blockage of a flow passageway in the bulk of the sheet may be overcome by adjacent flow passageways diverting flow around it.

[0707] Additional supplemental flow passageways 903, 904, 905, and 906 extend. perpendicular to the direction of flow 902. These may be added to the sheet in a systematic or random manner. These do not interfere with the connectivity of the cells of the sheet. They improve the connectivity of the cells’ void volume in a direction perpendicular to the direction of flow. Uniform spacing of these supplemental flow passageways is preferred, but not required. Uniform spacing and intersection of these flow passageways with the cells is most preferred. These flow passageways improve the uniformity of flow from outlet edge 910.

[0708] Any random supplemental flow passageway not perpendicular to the direction of flow will produce local non-uniform in flow from the sheet along its discharge edge. However, when averaged over a significant length it may not be consequential. Any periodic placement of flow passageways not perpendicular to the direction of flow may produce non-uniform flow from the sheet along its length if they are not present in every cell along the length of the sheet.

[0709] It is a teaching of this invention to construct a metering sheet with a base unit cell structure throughout the sheet to which are added one or more supplemental auxiliary flow passageways. These generally enhance or modify flow characteristics of a sheet and may improve average flow uniformity at the discharge edge

[0710] FIG. 9c illustrates a metering sheet 911 consisting of two distinct flow regions 914 and 915 represented by two different cell geometries delineated by the dashed lines 912 and 913 respectively. The flow direction is indicated by the arrow 919. It is often advantageous to have different flow resistance, flow passageway size, unit cell size, materials of construction, flow passageway orientation, flow passageway spacing, and fluid transport characteristics in the regions near the sheet entrance and sheet flow exit. This has utility in coating and filtering technology.

[0711] FIGS. 9d and 9e illustrate coating flow metering sheets with flow directions indicated by the arrows 965 and 955. In FIG. 9d the density of flow passageways becomes less from the point where the flow enters the inlet edge 961 of the sheet 960 until roughly halfway across the sheet. From there to the discharge edge 962 the flow passageway density remains constant. The high density of flow passageways at the inlet edge can facilitate filtration of contaminants from the fluid.

[0712] In FIG. 9e the density is larger at the discharge edge 952 than at the entrance edge 951. In between there is a transition in the flow passageway density from low to higher. This allows a fluid discharging from the sheet as individual streams to more easily merge together into one continuous ribbon of fluid.

[0713] FIG. 9f illustrates still another flow passageway grid in the fluid metering sheet 972 that is useful in promoting uniform flow from a discharge edge 974 where the arrow 970 indicates the flow direction. Again there is not uniformity in the base unit flow cell geometries throughout the sheet in the flow direction. The unit cell is constant with depth position.

[0714] The flow passageway structures of FIGS. 9d, 9e, and 9f illustrate a more general form of improved metering sheets for promoting uniform flow from the sheet outflow edge. In these no flow passageway is dead ended, and the cell structure of the sheet is uniform at a fixed distance from the inflow edge of the sheet

[0715] FIG. 9g illustrates another useful sheet internal flow passageway grid 980 arrangement wherein the repeating unit cell is triangular. The flow passageway grid geometry is referred to as “hexagonal”.

[0716] It is a teaching to construct a metering sheet with a base unit cell structure throughout the sheet and to add one or more additional solid structures locally obstructing passageways. It is also a teaching to construct a fluid metering sheet where the base cell structure is interrupted by solid structures, dams, non-flow areas and etc. Additionally, compound sheets containing areas of differing base cell structure are within the scope of the invention.

[0717] The unit cells of metering sheets of this invention may be characterized by having a flow passage or passages with a perimeter and a hydraulic diameter. These are scalar quantities defined by a flat plane normal to the sheet and passing through the cell. The hydraulic diameter of the void space passageway of the cell at the plane is defined as four times the open cross sectional area of the void divided by the perimeter of the surface of the void at the plane. For a flow passageway that is cylindrical and normal to the plane the void cross sectional area is pi times the diameter divided by four. The perimeter is pi times the diameter. The hydraulic diameter is then cylinder diameter.

[0718] For a slot normal to the plane (a void space contain by only an upper and a lower solid surface at the plane), the area is the height of the slot times the cell width intersected by the plane. The perimeter (the wetted perimeter) is the sum of the length of the slot wall lengths intersected by the plane. The hydraulic diameter equals two times the void height.

[0719] FIG. 10a illustrates a fluid metering sheet useful when interfaced with a internal die cavity in metering flow from a coating die used to produce a down web striped coating. The sheet 102 consists of areas 104 which allow flow. These are interspersed with areas 106 which do not allow flow. Arrow 108 indicates the direction of fluid flow from one edge of the sheet to the exit edge. This sheet may be used to communicate fluid dynamically between the internal cavity of a coating die and the external transfer area where fluid is applied to a substrate. Patterned metering sheets of this type are of great utility in the process of stripe coating of adhesives on webs. Many different stripe patterns may be coated with the same die by changing only the metering sheet.

[0720] FIG. 10b illustrates a fluid metering sheet flow passageway grid useful in metering flow from a small die cavity and distributing it along a line much larger than the cavity width. This sheet has an inlet edge region along the edges 122a and 122b. It has an outlet region edge 124. Spanning the area in between the inlet and outlet are two different grids of flow passageways that interconnect: grid 126 (indicated by black lines) and grid 128 (indicated by grey lines). The edges 122a and 122b are meant to abut a supply cavity or connect to a fluid source. The grey lines indicate a grid that serves dominantly a distribution function. The black lines 126 illustrate a grid that serves mainly a metering function and if desired a filtration function.

[0721] Cross channel 127 is delineated by the heavy grey line. It is where the grids 126 and 128 meet. It is preferred that the passages of this set be designed so that they have a near zero probability of being clogged by the target contaminant. The same is desired for the distribution grid 126.

[0722] Flow enters the inlet edge as indicated by the arrows 134, and exits the outlet edge as indicated by arrows 136. Fluid is distributed across the width of the sheet and along the line of the outlet face 124. The grid 128 serves mainly to distribute fluid across the width of the sheet to the upstream side of the metering grid 126. The metering grid serves mainly to meter the flow and filter target contaminants from the fluid. The combination of grids works together. They produce a specified desired flow profile of filtered fluid from the outlet edge 124. This is accomplished by designing and specifying the flow resistance of every passageway in both grids.

[0723] By using this type of distribution sheet in a coating die, the internal distribution cavity may be quite small and span only a percentage of the die outflow exit edge width.

[0724] For uniform out flow from the sheet, it is preferred that the flow passageways 130 and 132 and those in-between have identical flow resistances and be uniformly spaced. This will produce equal inflow into channel 127 along its length. It is preferred that all the passageways normal to the outlet edge 124 in grid 126 have equal flow resistances.

[0725] All passageways of the grid 128 may be made to have a uniform flow resistance by designing their hydraulic diameter variation appropriately.

[0726] For Newtonian fluids, the flow resistance through a tube is proportional to its length and inversely proportional to the fourth power of the hydraulic diameter. Therefore, equal resistance in passageways 132 and 130 may be obtained by choosing diameters to compensate for the difference in lengths.

[0727] If passage 132 were 81 times longer than passage 130 then a hydraulic diameter ratio of 3.0 between the two would be required for equal flow resistance. If passage 132 were 256 times longer than passage 130 then a diameter ratio of 4.0 between the two would be required for equal flow resistance. Such variations are easily accomplished.

[0728] The type of distribution sheet illustrated may be used in coating dies with small cavities to apply fluid to wide substrates. The ratio of the cavity size in the die width direction to die width may be smaller than 0.5, 0.2 or 0.1. This is helpful in designing some styles of simple or inexpensive dies.

[0729] FIG. 10c illustrates another sheet grid design using two different grid designs combined. It is useful for accepting flow through one input flow edge 140, and distributing the flow along a lengthy outlet edge 141. The arrows indicate the flow into and from the sheet. The channels of the grids may be all specified and designed to achieve a large number of distribution purposes including uniform flow from the outlet edge 141. The first grid region 142 may be designed to make it clog tolerant.

[0730] An unfulfilled need in the science of coating is the combination of filtration of target contaminants along with the uniform out flow from the outlet edge. Sheets illustrated here may easily achieve this goal with proper grid designs.

[0731] A designed porous distribution sheet may be used to distribute flow from a very small inlet region to a large outlet region. In the extreme we have found that we can design grids to take flow from a single input corner point and distribute it uniformly along a discharge edge. FIG. 10d illustrates such a rectangular sheet useful for distributing a uniform flow indicated by arrows along a linear distribution edge 150. The internal flow grid 148 is designed and engineered to take input flow from passage 149 and distribute it to the output edge.

[0732] Multiple sections of grids like in FIG. 10d may be combined to form simple but effective sheets. FIG. 10e illustrates a sheet 143 design where sixteen sections of corner feed sheet grids of the FIG. 10d design have been merged together along a length to form a composite sheet grid region 144b. A single point input grid region 144a interfaces with the region I 44b at an interface channel 145 which spans the width of the sheet 143.

[0733] Flow enters the sheet through a single passageway 147 in the branching region 144a The branching passageways of this region serve to distribute flow uniformly through eight passageways into the interface channel 145. Uniformity is achieved by designing this grid with all 29 passageways illustrated with uniform flow resistance. Designs may also achieve uniformity while simultaneously minimizing pressure drop.

[0734] Sheet like those illustrated by FIGS. 10a, b, c, d and e may have a complex mix of grid passageway conductance's to obtain a specific outflow distribution, tolerance to clogging, and filtration of target contaminants. The micro-replication techniques referenced above are well suited to the mass production of these sheets. The sheets are well suited for distribution and filtering of low viscosity liquids, and most gasses.

[0735] FIGS. 11a, b, and c illustrate useful in metering sheet structures produced from column structure combinations.

[0736] Metering sheets of this invention may consist of regularly spaced columnar structures. These columns may extend from one surface of a sheet base as illustrated in FIG. 11a. Here the sheet 110 is formed with a base 113. From this the column structures 112 extend for a fixed distance. These columns 112 may be cylindrical with constant diameters or other forms. The columns are arranged across the surface of the sheet in a regular, repeating pattern of unit cells one of which is indicated by gray dotted line 111. When this sheet is sandwiched between two confining solid surfaces the void volume becomes the fluid transport means.

[0737] As shown in FIG. 11b, the sheet itself may consist of a top and bottom bases 115 and 116 with columns 117 spanning between the two. The elliptical columns here are preferred to circular cross section columns. With the columns 117 as shown, the unit cell passageway in the direction of flow indicated by the arrow will have a higher flow resistance than the perpendicular passageway of the unit cell. It is preferred that the ratio of the cell passageway flow resistances, Rd/Rw, be greater than one. Rd is the passageway resistance in the depth direction and Rw is the resistance in the width direction.

[0738] Still another columnar based metering sheet structure of our teaching is shown in FIG. 11c. Here a placement of spheres 118 in a square grid arrangement makeup the inventive sheet. These spheres are interconnected at their mid-sections. When the sheet is confined between two surfaces 119, the enclosed void volume 120 provides precision fluid flow distribution flow passageways. A sheet made using spheres deformed by elongating them all in the rectangular sheet one direction and with spatial locations chosen so that the resistance ratio Rd/Rw is greater than one would be more preferred. This will promote generally more uniform outlet distribution of flow.

[0739] Many other useful spaced column profiles are possible for forming metering sheets. Three additional examples are shown in side view profile in FIG. 11d.

[0740] Those knowledgeable in the art will recognize that useful composite metering sheets may be formed by layering two or more metering sheets between two confining die walls. In this manner the total fluid conductance may be increased. It may be doubled by using two sheets of the identical design or tripled by using three sheets.

[0741] 2. Sheets Using 3-Dimensional Cell Structures

[0742] The sheets described above all use planar grids of flow passages. It is also a teaching of this invention to use three dimensional grids of flow passages often referred to as “3D grids”. One example is illustrated in FIG. 16. Here a cubic like grid of passageways 161 is used to distribute flow in three dimensions. It may be thought of as two planar arrays of cells stacked on top of each other with additional passages extending through each plane and intersecting the other.

[0743] In the cell of FIG. 3, the intersection of the two passages, the nodal point of the cell, has a coordination number of four. Four passageways extend from each node. The passages extend in two dimensions. In the pore structure of FIG. 16, the nodal points have a coordination number of six. The higher coordination number and the three dimensional interconnection of passageways can produce improved fluid flow performance over a single sheet or stacked sheets with only planar flow grid geometries.

[0744] For the sheets with the geometry illustrated in FIG. 11c, it should be noted that the grid of passages are three dimensional when top and bottom capping layers are used. When multiple sheets of this geometry are stacked, thick three dimensional arrays of passages may be formed.

[0745] It is also a teaching of this invention to create flow distribution and filtering sheets of granular, particulate, open cell foams or other porous materials by using capping films on one or both major surfaces. These confine the available flow paths to within the sheet when in use. Such sheets may not be highly accurate in flow distribution. There is variation of the internal pore sizes. However, fluid distribution from one sheet edge to another without loss through a major surface has utility in many applications.

[0746] 3. Modeling Metering Sheets

[0747] The use of Darcy's Law to describe flow in metering sheets has limited value. It has been found a more detailed analysis of the flow in each individual flow passageway and modeling the actual connectivity of a flow passageway is useful for design of devices employing metering sheets.

[0748] Flow models may be used to accurately describe the distribution of flow rates from metering sheet dies. Using known flow modeling principles, a model of flow through multi-passageway media was developed. With this model each passageway of the media is considered as a discrete element and the flows and pressure drops for all are simultaneously calculated. The Reynolds number for each passage is very low, and the effects of gravity and inertia are not generally important. A Stokes flow model is assumed for our work. The models may be used to design the performance of sheet arid to design devices constructed using the sheets. Of course with enough computing power, one need not make the Stokes flow assumption, but little is added to the accuracy of the predicted flow from the outlet edge of a sheet.

[0749] Many different fluids exist. They may be characterized by their rheology. The standard fluid for all first investigations and modeling fluid devices is a Newtonian fluid. These are characterized as having a constant viscosity that is independent of flow rate. All simulation results in this disclosure are specifically for Newtonian fluids. Newtonian fluids are generally used for defining design principles, illustrating preferred embodiments of the invention, and determining certain features of the elements, apparatus and methods of the invention. Persons with ordinary skills in the art will recognize that the teachings pertain to all fluids in general.

[0750] We have modeled sheets that have both uniform and non-uniform internal flow structures. For a distributing die using a cavity to feed a metering sheet and when all flow passageways in the metering sheet have equal conductances, it is found that the flow uniformity is dominated by one variable. Uniformity is primarily a function of the ratio of the viscous flow conductance down the length of the cavity to the total composite viscous flow conductance through the whole sheet from the cavity to the sheet exit edge. This ratio is referred to as the “dimensionless sheet die viscous number” and identified with the symbol Nvs. For any sheet grid geometry, the total grid composite conductance equals the volumetric flow through the sheet divided by pressure drop from inlet to outlet edge of the sheet. This may be calculated or measured experimentally.

[0751] In the case where incremental cavity flow resistance changes along its length, it is appropriate to define Nvs in terms of an average cavity flow resistance. An example is the length average. Other definitions of “average” are envisioned and are dependent upon the geometry and fluid properties along the cavity.

[0752] We have modeled and studied many sheet grid geometries including those that have uniform structures. Sample results have been obtained for the sheet flow passageway geometries of FIGS. 4a, 6a, 8a, 9a and 9g. These geometries are referred to as the square, diamond, triangle, rectangle, and hexagon grid geometries respectively. These and other grid geometries may be fed from a cavity of constant cross-sectional area and shape along its length.

[0753] Three studies defining the utility of metering sheets have been undertaken. The first considers the uniformity of flow from devices where a cavity directs flow into a rectangular distribution sheet. The second investigates the ability of a sheet's design to overcome blockages of internal passageways and still produce uniform outlet flow. The third study investigates the ability to uniformly distribute flow when the sheet inlet edge is not feed by a long cavity but feed at a single point.

[0754] Cavity fed rectangular sheets are considered in the following. The cavities are assumed to have constant cross-sectional areas.

[0755] a. When all Grid Passages Have Equal Conductance

[0756] For fabrication purposes, it is useful but not necessary to create a flow passageway grid structure where all flow passageways have identical viscous flow conductances. For this case, it is found that conservative estimates of the out-flow uniformity index for a cavity and sheet device is given by the equation:

UI 2 A {0 0.995 A−0.3334 A.sup.3+0.13334 A.sup.5−0.05397 A.sup.7}

[0757] where A=0.5{(Nvs).sup.0.500 },

[0758] and where Nvs is <1,

[0759] and the sheet depth to width ratio is less than 0.5,

[0760] and the depth of the sheet is greater than 9 unit cells.

[0761] A simpler, somewhat less accurate design expression is given by the equation:

UI 0.46 Nvs.

[0762] The uniformity index, UI, is defined as the maximum local outflow minus the minimum local outflow rate along the width of the sheet divided by the average rate.

[0763] Based on our studies, a fluid distributing die with a cavity and a metering sheet should have a ratio of cavity to sheet viscous conductance less than 1. This will achieve a UI value below 0.5. For shear thinning fluids it will be larger.

[0764] It is more preferred that this ratio of conductances be below 0.5. Here the achievable “Flow Uniformity Index” will be below 0.25 giving a more uniform flow distribution. Still more preferred is a ratio of viscous conductance less than 0.1. This will produce a “Flow Uniformity Index” below 0.05.

[0765] Most preferred are die designs where the geometry results in a viscous ratio less than 0.04. This achieves a “Flow Uniformity Index” less than 0.02. For shear thinning fluids the uniformity index will be larger so that these values of the conductance ratio, Nvs, provide an upper limit for the conductance ratio which should not be exceeded.

[0766] b. Grid Passage Conductance Unequal in Two Directions

[0767] The passageways for the grid need not have uniform conductance. In the case of a square grid, it is useful to have passages in the depth and the width directions with different conductance.

[0768] When a distribution sheet with a square grid of internal passageways is fed from a cavity, in the general case the passages in the width direction may have a conductance different from those in the depth direction. The performance of the distribution system may be improved with proper choices of conductance.

[0769] The sheet composite conductance through the sheet across its width and perpendicular to the depth direction with all the internal grid passageways acting together is Csw. It may be calculated or measured experimentally. The dimensionless sheet viscous number in the width direction Nvsw is the ratio Csw to the conductance of the cavity.

[0770] Extensive modeling has shown that the uniformity index is a very strong function of Nvs, and a function of Nvsw. It may be approximated by the equation:

UI=0.466 Nvs.sup.0.98(1−Nvsw}

[0771] where Nvsw is less than 1, where w/d is greater than 2, and where Nvs is between 0.01 and 1. When sheet output uniformity is desired, and Nvsw is greater than 0.00000001, a value of less than 0.5 is preferred, a value of less than 0.25 is more preferred, and a value of less than 0.1 is most preferred.

[0772] For all values of Nvs significant improvements in the sheet out flow uniformity are obtained when dimensionless sheet viscous number in the width direction is greater than 0.1 This is illustrated by tabulated uniformity data in FIG. 88.

[0773] i. Overcoming Sheet Passageway Clogging

[0774] Besides enabling good flow distribution, metering sheets provide a tolerance to clogging of internal passageways. The presence of grid passageways which allow flow in the direction parallel to the supply cavity provides a means to heal any flow up sets caused by clogging passages within the sheet.

[0775] The ability to a redistribute flow around a clogged passage has been extensively investigated. A significant characteristic of the grids in a metering sheet is the flow uniformity downstream from a clog. If the local flow distribution is uniform at a distance downstream from a clogged passage, the local flow uniformity index equals zero. Immediately downstream from the clogged point there is no flow in the direction of flow through the grid toward the outlet edge, and the local index is along a line perpendicular to the direction of flow is equal to 1.0. There is no flow in the clogged passage leading away from the obstructed point. As one moves downstream from this point successive cross flow passages allow flow to be reestablished in the passage directly in line with and leading towards the outlet edge.

[0776] After each successive crosswise passage the uniformity of the local flow distribution improves. This may be quantified by the local uniformity index after each successive cross passageway. It is found that the improvement in uniformity, the healing of the flow defect, is a direct function of and dominated by the ratio of the cross flow passage resistance to the depth direction passage resistance, Rw/Rd. The local uniformity improves as the ratio decreases and as the number of cross flow channel increases.

[0777] For countering clogging problems, it is preferred that the ratio be less than 100. It is more preferred that the ratio be less than 10. It is most preferred that the ratio be less than 1.0. With a sufficient number of cross channels, it is found that in all cases a usefully uniform flow from the sheet may be achieved.

[0778] ii. Feeding Distribution Sheet at a Single Point

[0779] Another characteristic of the metering sheets are their ability to take flow inputted at one point at an inlet sheet edge and distribute it with adequate uniformity along an outlet edge. Such a sheet could produce uniform flow at the outlet edge when all but one of the flow pathways into the sheet input edge is clogged. This design could perform well when the fluids are highly contaminated.

[0780] Rectangular metering sheets have been extensively studied for this single input case where the inflow is introduced at one edge of the sheet. It is found that the uniformity of the flow from the outlet edge of a sheet is dominantly dependent upon a single dimensionless parameter, Nsp. This parameter equals the sheet width divided by sheet depth and all times the square root of the quantity the resistance of an individual passageway in the width direction divided by the resistance of an individual passageway in the depth direction.

Nsp={w/d} {sqrt(Rw/Rd)}

[0781] When the parameter increases the uniformity index decreases.

[0782] FIG. 88 graphs our findings for the variation of the uniformity versus the parameter is Nsp.

[0783] The data is accurately fitted by the equation:

UI=e.sup.x

[0784] where x−1.3537 Nsp.sup.4+3.4267 Nsp.sup.331 3.6685 Nsp.sup.2+3.0575 Nsp0.4899.

[0785] It is found that for square grids flow from all passageways exiting the outlet edge of the metering sheet may be obtained if the parameter Nsp is less than 1.5. Better uniformity may be obtained if the parameter Nsp is less than 0.75. And still better uniformity may be obtained if the parameter Nsp is less than 0.6

[0786] If a sheet inlet edge is fed by an inlet cavity that flow from the cavity to the sheet through more than one grid passageway, the flow uniformity is improved over that results obtained by a single feed point as describe in the previous paragraph

[0787] Surprisingly it is found that just a limited number of single feed points to the metering sheet can reduce the uniformity index from a totally unacceptable value to a desired very low value. For example if the sheet width to depth ratio is 10, and the ratio resistance ratio Rw/Rd is 4.0, then feeding the sheet at a single point at an end of the sheet width will result in an uniformity index of about 2. If the sheet is fed at two points, one at the one-quarter point along the width direction and the other at the three-quarter point, the uniformity index is reduced to a very desirable 0.04 value. If the sheet is fed at four points, at the one-eighth, three-eighths, five-eighths and seven-eighths points along the width direction the uniformity index is reduced to an extremely desirable 0.0004 value.

[0788] FIG. 10e illustrates a metering sheet that employs the design principle described in the proceeding paragraphs. It schematics pictures a sheet with a composite grid geometry where eight spaced final distribution grid entrance points are used.

[0789] Metering sheets of this invention produce uniform fluid flow from their discharge edge when the design parameters are properly selected. They are also generally insensitive to clogging by particles in the fluid. Solid matter can collect at the inlet edge of the sheet as the fluid flows into it from a feed cavity. When 80 percent of the flow passageway entrances are clogged, one would expect an unacceptable distribution of flow at the sheet exit and throughout the whole sheet. FIG. 12 illustrates that this is true for conventional porous sheets where the pore sizes are distributed about an average.

[0790] This figure is a grey scale contour graph the local average flow rate at a point. Each position on the graph represents a location looking down on the sheet. The grey scale plotted at that position indicates the ratio of the local flow rate in the direction toward the exit edge divided by the average flow rate. As indicated by the legend, black color indicates greatly reduced or excessive flows. A white color indicates flow at or near the average.

[0791] FIG. 13 graphs the flow distribution for an improved metering sheet for the same 80 percent blockage condition shown in FIG. 12. The disruption from the blockage at the inlet side of the sheet is substantially mitigated by the sheet design. The flow at the exit is shown to he better than the range of plus or minus ten percent of the average. The flow passageways of the sheet are arranged in a square grid.

[0792] More detailed results are shown in the graph of FIG. 14 where the actual local flows at the sheet outlet are plotted. Here the variation in the exit flow rate is plotted both for the simulated commercial porous sheet (labeled “Porous Sheet”), and for the improved metering sheet. For the commercial sheet the variation is plus 97 percent and minus 40 percent. For the improve sheet the variation is plus or minus 5 percent. This is a vast improvement. A metering sheet also offers performance improvements over distributing slots when clogging conditions exist.

[0793] Of course, additional actions may be employed to improve the uniformity of a coating applied from a coating die. U.S. Pat. No. 5,262,194 to Louks, et. al., discloses applying ultrasonic energy to excite the line of initial contact between the coating fluid and the substrate to provide a coated material of increased crossweb uniformity than would be otherwise present without the ultrasonic energy. However, the method does require additional process equipment and can result in increased complexity to the coating operation. The new concept disclosed in this invention corrects potential coating non-uniformities prior to being coated on the substrate.

[0794] 4. Filtration Using Improved Metering Sheets

[0795] Filtration is the term used to describe the removal of contaminants from flowing fluid. It is an extremely important industrial process with uses ranging from pollution abatement, mineral recovery, polymer processing and a multitude of other material processes in chemical, biological and petroleum industries. When a porous medium acts as a filter and the suspended particles in the fluid being processed are larger than the pore or restrictive flow regions of the pores (pore throats), the process is commonly referred to as screening or straining. It is the filtration process considered here.

[0796] Straining large particles is essential for polymer processing, and coating. Large particles obstruct flow and can create point defects, functional anisotropic regions, and a multitude of other problems within products. In optical products the malfunctioning of even one area as small as one pixel can be cause for rejection of the product. Generally these large particles range from ten to thousands of micro-meters in hydraulic diameter.

[0797] In an embodiment of our invention, the fluid distribution and metering sheets function both as a final filter to remove contaminants, and to distribute fluid. FIG. 13 illustrates ability of a very simple sheet to both trap targeted particles and distribute fluid. Here the particles challenging the sheet are approximately the size or the passages or larger. In this example all of the passages directing flow from the inlet supply manifold have a probability of stopping the contaminant particles equal to 1. By definition a probability of capture of 1 means that any particle presented to a passage will be trapped, and it will clog the passage. Therefore, all the particles are collected at or in the entrance passageways at the inlet face or edge of the sheet. As such, the filtration ability here is generally proportional to the inlet face area.

[0798] Such a sheet design can be adequate for low particle concentrations. But the number of particles that may be filtered is limited. As more and more particles challenge the sheet enough of the passages at the entrance edge may become clogged to destroy the outlet uniformity. As still more particles challenge the sheet, the pressure drop through the sheet may increase beyond limits for a given flow rate, or the flow may be totally stopped. Improved sheet passageway grid designs have been developed to substantially postpone these problems and to allow filtration of many times more particles than in the FIG. 13 example.

[0799] A target contaminant particle generally is irregular in shape and has a nominal hydraulic radius dependent upon its orientation in a flow field. Flow passages of porous media may have uniform flow resistances per passage on the average. However, there are variations of their hydraulic radii from point. Because of this, the capture probability of a particle of a particular hydraulic radius passing through a passage of the same radius may be one or less than one. Additionally, a collection of particles of a nominal size will actually have a distribution of sizes about the nominal. This also adds to the statistical nature of the particle screening, trapping process with which we are concerned. Therefore, the capture probability for the transit through a passageway for the target particles is a variable of importance. Generally it is a design parameter that can be fixed for a passageway on the average when a large number of target particles are considered.

[0800] 5. Further improvements in Metering Sheet Filtering

[0801] Improved filtration using a sheet may be accomplished in a number of inventive ways. These have been found and defined by the use filtration process models.

[0802] Consider the case where the concentration of target particles is low. So low that in a specific unit of time only one particle passes into the inlet sheet edge. In this situation, the probability of particle capture in a specific passage is the product of the probability of the particle being in the fluid flowing through the specific passage multiplied by the probability of capture of the particle.

[0803] The probability of the particle being in the volume of fluid flowing through a passage equals the ratio of the flow rate through the passage to the total flow rate. The probability of the particle being captured while flowing into or through a passage is a function of multiple variables. When filtering is accomplished by sieving commonly one variable is the ratio of the particle hydraulic diameter to the passage hydraulic diameter. Another is the flow rate through the passage.

[0804] Filtration modeling may consider contaminant particles where the sizes are all uniform and the capture probability is known for that size. More complex modeling may consider: populations of contaminants having a distribution of sizes; pore capture probabilities that are dependent upon particle size, pore flow, pore size, and past history; populations of pores having a distribution of sizes; and many other parameters.

[0805] Consider a distribution sheet with only one plane containing a two dimensional square grid of passages as illustrated in FIG. 4a. Here eleven rows of passages span the depth of the sheet. The depth is measured from the bottom where the flow enters as indicated by arrow 35 to the edge where the flow exits as indicated by arrow 36. The width of the sheet contains thirty four channels directing flow to the sheet outlet. Further consider the case where the probability of capture of a target particle in each passage is less than 1.0 and even much less than 1.0. Surprisingly, it is found that nearly 100 percent of the particles entering the grid are captured.

[0806] A filtration improvement is to modify the grid of passages so that the probability of capture of a target particle in the entrance edge passages, the “entrance composite probability”, is less than one. Preferred is to have the entrance composite probability substantially less than 1.0 while simultaneously having the probability of capture by the remaining passages near or at 1.0. More preferred is to have the flow past through successive regions with each have these characteristics.

[0807] While much effort in the past has been devoted to modeling flow through porous media using a continuum approach, network models allow investigation of the microscopic details of the filtration and flow processes. They have been used for this teaching.

[0808] A network model employs a regular or random array of pores and throats with specified geometry and topology. Statistical variation of these may be used to model known granular porous filter media. A network model can deal with pore scale behavior of fluid flow and particle collection. Such models have also been used in the study of infiltration and two phase flow transport problems in oil and gas extraction and also soil percolation. We have used this technique to develop unique media and devices. The modeling is discussed in greater detail in a later section.

[0809] 6. Multilevel Distribution Grids

[0810] Those with ordinary skills in the art of fluid flow and distribution will recognize that three dimensional flow grids are a natural extension of the two dimension planar grids discussed above. Expanding the grids with channels in the direction of the metering sheet's thickness may improve its ability to distribute flow and to reduce the pressure drop through the sheet. Three dimensional grids are quite helpful for improving filtration.

[0811] 7. Fluid Distribution Along Non-straight Lines

[0812] The preceding discussion has dealt with fluid distribution along a straight line. Those with ordinary skills in the art will recognize that the invention may be used to distribute fluid along a line that is not straight. The invention may also be used to distribute fluid with a controlled variation along the distribution line. The controlled distribution variation may be achieved by designing the sheet flow passageway structure with controlled variation of the viscous conductance. It may also be achieved using a deformable sheet where deformation adjusts the local conductance of the sheet. Compressible sheets where the flow passageway volume may be manipulated are useful.

[0813] It is common in industrial fluid distribution to require fluid to be discharged along a lie that is circular. FIG. 15 illustrates such a sheet with an inner circumference 152 and an outer circumference 151. On this sheet are located protruding columns 155 a portion of which are shown between radial lines 153 and 154. These and the columns not shown are spaced along radial and circumferential lines so as to provide the desired flow distribution along the outer circumference 151.

[0814] In each of the above uses of improved metering sheets, it is preferred that the internal flow passageways and flow grids of the sheet are designed so that first there are no dead ended passageways in the bulk of the sheet and away from edges or confining surfaces. Additionally, the flow passageways are interconnected so fluid may flow from any flow passageway to any other flow passageway. The passageway flow conductance of the flow passageways are specified by the sheet design, and the location of the passageways and flow passageway intersections are fixed by the sheet design.

[0815] 8. Methods of Manufacturing Metering Sheets

[0816] Metering sheets may be fabricated from many materials by many different methods. The techniques of making structured surfaces are applicable to making improved metering sheets. The making of structured surfaces on a polymeric layer such as a polymeric film is disclosed in U.S. Pat. Nos. 5,069,403 and 5,133,516, both to Marentic et al. Structured layers may also be continuously replicated using the principles or steps described in U.S. Pat. No. 5,691,846 to Benson, Jr. et al.

[0817] Structured polymeric metering sheet media produced in accordance with such techniques can be replicated and micro-replicated. The provision of replicated structured layers and metering media is beneficial. They can be mass produced without substantial variation from piece to piece and without complicated processing techniques. The replicated surfaces and the replicated flow passageway defining surfaces preferably are produced such that the flow passageway features retain individual feature fidelity during manufacture from piece-to-piece.

[0818] Metering fluid transport sheets for any of the invention embodiments can be formed from a variety of polymers or copolymers including thermoplastic, thermoset, and curable polymers. As used here, thermoplastic, as differentiated from thermoset, refers to a polymer which softens and melts when exposed to heat and re-solidifies when cooled and can be melted and solidified through many cycles. A thermoset polymer, on the other hand, irreversibly solidifies when heated and cooled. A cured polymer system, in which polymer chains are interconnected or crosslinked, can be formed at room temperature through use of chemical agents or ionizing irradiation.

[0819] Polymers useful in forming metering sheets in articles of the invention include, but are not limited to, polyolefins such as polyethylene and polyethylene copolymers, polypropylene, ethylene/vinyl acetate polymers, ethylene/ethyl acrylate polymers. Other useful polymeric materials include vinyl polymers (e.g., polyvinyl chloride, polyvinyl alcohol, vinyl chloride/vinyl alcohol copolymers, polyvinylidene chloride, polyvinylidine diflouride (PVDF)), acrylate polymers (e.g., polymethyl methacrylate), polycarbonate polymers, polyesters (e.g., polyethylene terephthalate), polyamides (e.g., Nylon), polyurethanes, polysaccharides (e.g. cellulose acetate), polystyrenes (e.g., polystyrene/methyl methacrylate copolymer), polysiloxane polymers (e.g., polysiloxane and organopolysiloxane polymers). Metering sheets can be cast from curable resin materials (monomer and prepolymer mixtures) such as acrylates or epoxies and cured through free radical polymerization pathways promoted chemically, by exposure to heat, electromagnetic radiation or electron beam radiation. Plasticizers, fillers or extenders, antioxidants, ultraviolet light stabilizers, surfactants, and the like may be utilized within the polymers for the invention.

[0820] The metering sheet could also be made from materials other than polymers if desired. Metals, ceramics, super cooled liquids, organic and inorganic materials may all be used.

[0821] Polymeric materials including polymer blends can be modified through melt blending of plasticizing active agents such as surfactants or antimicrobial agents. Surface modification of the structured surfaces can be accomplished through vapor deposition or covalent grafting of functional moieties using ionizing radiation. The polymers may also contain additives that impart various properties into the polymeric structured layer. For example, plasticizers can be added to decrease elastic modulus to improve flexibility.

[0822] Distribution and transport of fluids, including gasses, liquids, super critical fluids, and combinations is central to many unit operations. These operations can include, for example, heat transfer, mass transfer, ion exchange, reactive chemistry, and coating. Additionally, the present invention provides an apparatus and methods of distribution of fluids for active thin film reaction in applications such as chemical or radiant reactors.

[0823] Preferred embodiments of the invention may use thin flexible sheets that have metering sheet cell topographies on their surface. For purposes of this invention, a “film” is considered to be a thin (less than 5 mm thick) generally flexible sheet of polymeric material. The economic value in using inexpensive films with highly defined metering sheet film surfaces and structures is great. Flexible films can be used in combination with a wide range of capping materials and can be used unsupported or in conjunction with a supporting body where desired. The flow passageways formed from such structured surfaces and caps may be flexible for many applications, but also may be associated with a rigid structural body where applications warrant.

[0824] Similarly, the capping material may be a rigid metallic. The capping material may be a confining wall of a mounting fixture. The metering sheet material may be a sheet with complex three dimensional structures on both sides. In this font' capping material is required on both sides to confine the fluid flow within the sheet. This capping material may be flexible films or the solid confining walls of a mounting structure. FIG. 11c is an example of a metering sheet that requires capping on two sides. Here capping could be provided by confining the sheet within in a slot of a fluid distribution die.

[0825] Another meter of constructing a metering sheet is to align side by side a row of precision tubes on a flat surface. These may be fused or bonded together to form a sheet. This sheet may be bonded to a plane flat sheet of material or between two sheets of material. In this form one has a composite sheet with many bores or passages running parallel with no auxiliary channels running at an angle to the bores. Such channels may be added in many ways.

[0826] When the tubes are assembled and bonded to a single plane sheet, many cross channels may be created by removing lines of material at an angle to the bores. Machining, laser ablation, electric discharge machining, grinding and thermoforming are some of the many methods of creating flow channels between the bores. Alternatively many rows of very short length tubes may be assembled side by side and in sequential rows on a plane flat sheet and bonded to that sheet. The spacing between the rows then provides the auxiliary cross channels.

[0827] The high volume production of the fluid metering sheets by repeated or continuous replication processes allows the metering sheet to be manufactured at low cost. This then allows the sheets to be disposable. By this means, the expensive precision parts that require cleaning and reuse are eliminated, and manufacturing using the sheets is simpler and less costly.

[0828] 9. Methods of Assembly of Metering Sheets

[0829] The fluid metering sheets can be assembled from layers of multiple sheets using a variety of techniques, including, but not limited to external force clamping devices, bonding using thermal means, bonding using curing layers, using adhesive layers, or bonding using internal mechanical interlocking structures and the like.

[0830] Clamping methods can include constraints on the external major surfaces of the sheets that apply a mechanical clamping force, or other external clamping forces such as air pressure, gravitational, electrostatic or magnetic forces.

[0831] Thermal bonding methods can include heat transfer devices, sonic, ultrasonic or electromagnetic radiation, such as focused infrared or radio or microwave radiation.

[0832] Pressure sensitive adhesion concepts can also be envisaged.

[0833] Mechanically interlocking layers can also be utilized. These may have molded, machined, or formed internal sheet structures.

[0834] Additionally, tie layers of one or more materials suitable for use with the sheets can be used. Generally the tie layers can be a meltable, curable, or chemically bondable layer.

[0835] Additionally, the layers can be adhered using adhesive forces. These layers can be pre-applied including application just prior to the assembly and bonding step. For example, a lower melting temperature polymer can be applied to non-meltable or higher melting temperature sheets to aid in thermal or mechanical bonding means. A nonwoven film can also he used as a tie layer.

[0836] A curable tie layer may be utilized. The curing step may be accomplished by thermal means such as ultrasonic, infrasonic, or electromagnetic radiation such as focused infrared or radiation. Component layers that initiate curing on contact may be used. The curable layer can contain a component that absorbs the applied energy at a greater rate than the sheet material, thereby only melting the tie layer.

[0837] These bonding means can be used to only bond at sheet internal intersecting surfaces (interstices) where the energy is focused.

[0838] 10. Three Dimensional Sheets

[0839] Consider a 2-dimensional grid lattice laid out in a square pattern as shown in FIG. 17 where the nodes 171, the intersections of the lattice lines 170, are connected by the lattice. The lines of this lattice represent flow passages. The nodes are the intersection points of the passageways. These passageways may be simple flow channels. They may also be irregular shaped pores. The collection of interconnecting passages may also be referred to as a grid of passages, pores, or simply a grid.

[0840] A two 2-dimensional grid is a layer, and multiple layers may be stacked together. When the two layers are interconnected by channels spanning between nodes in the bottom and top layers, a cubic lattice or 3-dimensional grid of passages may be created. An example is shown in FIG. 18. A cubic lattice of channels 180 is shown. Here the bottom grid of channels 181 is connected to the top layer of channels 184 by interconnection channels 183. The shaded vertical planes 182 are added solely to illustrate and emphasize the three dimensional character of the lattice of pore flow passageways.

[0841] The grid of pores in FIG. 18 is embedded within solid material such as metal, plastic, or ceramic. This material may be a sheet of material with an inlet edge and an outlet edge. These have exposed passageways where flow may enter and exit. FIG. 19 schematically illustrates such a sheet. Flow enters the front edge as indicated by the arrows 190, flows through the depth of the sheet, and exits the back edge as indicated by arrows 191. The shading of the vertical planes is added only to emphasize the 3-D nature of the media. This is a basic sheet type for a fluid distribution or metering device. It may also function as a fluid filtering sheet device.

[0842] FIG. 16 illustrates the cube like grid composed of eight nodes 162, 163, 164, 165, 166, 167, 168, and 169. The node points are the intersections of channels 161 that extend in width, depth and sheet thickness directions from each node.

[0843] 1. Improved Filtration Media

[0844] 1. Modeling Filtration

[0845] Modeling the flow of fluid in the passages and the capture of particles within a flow grid is extremely valuable in understanding the filtration ability of such grids. The modeling has been used to identify significant parameters and to develop improved filtration methods and devices.

[0846] When the flow resistance of each channel in the filter media is known, one may calculate the flow rate and flow direction in each grid passageway. With an applied pressure gradient from the front to back side across the depth of a sheet or block of media, the total flow rate through the media and the individual channel flows may be calculated. If individual channels are clogged by trapped particles, the redistribution of flow around the obstructed passages may be determined. This is the basis for various approaches to modeling the filtration of a dilute concentration of contaminant particles from a flow of fluid.

[0847] In a grid of nodes, the connectivity of all the channels to nodes is known. When the flows and flow directions are calculated, the directional connectivity is determined. Knowing the contaminant capture probability for each channel, and knowing the probability for a contaminant to take a specific flow path at each node juncture, allows us to perform statistical experiments following the paths and to determining the point of capture of large numbers of particles.

[0848] 2. Filtration: Following a Particle

[0849] a. Particle Paths

[0850] In one variation of the modeling the filtration performance of a fluid distribution sheet, a volume of fluid that contains a particle of a known size is allowed to pass completely through the sheet. The model determines if the particle is trapped within the sheet and its location. This is appropriate for low concentrations of contaminants where only one particle is present in the flow grid at a time. In a great many industrial processes there is a very sparse population of particles in the fluid at certain steps. However, the demand for quality requires that they must be removed. Details of this approach are described below.

[0851] At higher particle concentrations two or more particles may be passing through the grid simultaneously. In this case the general approach to simulation is the same. In the simplest case, the probability of a particle entering the grid at each increment in time is calculated from the concentration, the flow rate, and volume of the pores. The simulation marches forward in time with the particle paths being individually calculated as time proceeds.

[0852] In FIG. 20 a node point is illustrated. The arrows 200 indicate flows into and out of the node 201. When a particle is carried by the flow into the node, it is assumed to exit in the flow from the node. The channel by which the particle leaves is determined by chance. The probability that particle is in the flow in a particular channel equals the ratio to the flow in a channel divided by the total flow rate leaving the node. The probabilities for all flow leaving channels are thus known, and the sum is equal to one.

[0853] Given these probabilities one may “roll the dice” and determine by chance which channel carries the particle leaving the node. The flow model for the grid gives the directional connectivity. Therefore, at any node it is known which channels contain entering flows and which contain exiting flows. Once a particle is in a channel, the node to be encountered is known. At the next node the dice are rolled again to determine its exiting route. The stochastic processes of the movement of many particles through the filter grid may be followed one by one.

[0854] b. Particle Capture

[0855] The process of capture of a particle as it flows through or into a channel can be complex. Extensive modeling of the specialized process of straining has been done. In straining if a particle is captured in a channel, it is assumed the channel is plugged and thereafter allows no flow. This is accomplished by immediately setting the particular channel's flow conductance to zero in the flow model. If the particle is captured, the flow distribution through all the channels is recalculated with the newly clogged channel taken into account. For a fixed fluid supply pressure, the total flow in the sheet will be reduced. At some point when many particles have been captured, the total flow will be reduced to zero, and the sheet is totally clogged. In reality filters are often replaced when the pressure drop across them exceeds a limit, or the flow is reduced to a limit value when the upstream pressure is maintained constant.

[0856] In this modeling, it is assumed that the capture probability for each channel and each particle is known. This probability may be controlled by many factors. Significant ones include cross sectional shape, hydraulic diameter, length and size variations. These may be set for the filter media, the sheet, by design.

[0857] The particles entering the filter media may have a distribution of properties like size. It is assumed that the capture probability for every particle with respect to every pore is known or may be calculated from known parameters. In the following discussion on filtering simulation, results represent the cases where capture probability varies only from pore to pore, and not from particle to particle. This approach simplifies the analysis of results but does not diminish the utility of the teachings.

[0858] Once a particle enters the flow going into a particular channel, we determine if it is captured by the channel using the capture probability. Knowing the probability we “roll the dice” to determine if capture occurs. If the probability is 0.001, then on the average only one particle in one thousand will be captured by the channel. In the simulation all “roll the dice” decisions are made using a random number generator.

[0859] In simulation of filtration multiple numerical experiments are used to determine the average or most expected trends and effects. An estimated of a mean equals the average of a limited number of experiments plus or minus two times the standard deviation of those experiments divided by the square root of the number of experiments. A large number of experiments are required for very accurate estimates of means. However, the estimates are independent of each other so they may all be calculated simultaneously using parallel processing computer hardware and software.

[0860] Particles are generally captured sequentially and with each event the flow in the grid redistributes and needs to be recalculated. The recalculation involves the solution of a large set of equations, and this is the most resource and time intensive step of the model algorithms employed. Commonly up to 90 percent of the computer simulation time is consumed by the step. Flows are often calculated for filter media containing pores ranging from hundreds to hundreds of thousands of interconnected pores. When the number is large, the particle paths and capture events for a subset of the total particles processed may be estimated without updating the flow distribution. That is the movement of a subset of particles may be approximated by a simultaneous calculation rather than a sequential one. This allows the use of parallel processing algorithms and results in proportional reductions in computation times. Improvements in the accuracy of the obtained estimates may be obtained by limiting the use of this approach to the cases where the particles traverse pores in the media that are significantly removed from each other,

[0861] 3. Characteristics of the Filtration Sheets

[0862] a. Filtering Flow Distribution Grids

[0863] An example geometry where the sheet grid of nodes is designated as 5×26×2 is used in the following example. This designation means the test sheet consists of 2 layers. In a layer there are 5 columns of nodes extending from an entrance edge (inlet) to an outlet or exit edge, and there are 26 rows of nodes between the inlet and outlet. The rows of nodes extend parallel to the inlet face, and the columns extend perpendicular to the inlet and outlet faces. Filtering flow grids have channel, geometries or unit cells that repeat throughout the sheet or media. In filtering geometries, channels direct flow in a number of primary directions. For cubic channel grids there are three: parallel to the depth direction, parallel to the width direction, and parallel to the direction from layer to layer. When only a single layer of with a rectangular layout of channels is used there are just two primary directional orientations of the channels.

[0864] In filtering grids the flow resistances in channels in each primary direction of the grid are generally held constant in the following simulations. However, the channel flow resistances may differ between the primary directions, but in general the filtration results depend upon the capture probability for the channels that have a directional component orientated from inlet to outlet edge.

[0865] In commercial porous metal the flow passages are randomly distributed in space and are organized with a multitude of mostly random orientations and flow resistances.

[0866] FIG. 21 graphs representative examples of the filtration process for various channel capture probabilities. Each curve shown plots the dimensionless total flow rate through the cubic 5×26×2 grid of channels as a function of the number of particles that have entered the sheet. The capture probabilities of the channels differ for each curve. All channels in the grid have the same designated probability for the experimental trial. Each line of graphed results represents a single experimental trial. In each case the process continues until the sheet is clogged and the flow rate abruptly drops to zero. These results illustrate that capture probabilities less than 0.9 are preferred increasing the number of contaminants contained. It is found that generally channels that have a direction component orientated from inlet to outlet are most significant in influencing the nature of the results.

[0867] The curves of FIG. 21 represent single simulation experiments. Repeat experiments at the same conditions would illustrate the same trends but vary in the exact numeric results. Because the particle path and particle capture are a stochastic processes, duplicate trials would result in a different curves, but on the average the endpoint results would be the same,

[0868] When the probability is high, only a few particles are processed before all flow is blocked. The graphs are plotted versus the independent variable which is the number of particles passing into the sheet. Some of the particles are captured and some pass through the sheet and exit from it. The graph does not show the number of particles that escape from the sheet. FIG. 21 illustrates that the filtration of many more particles may be accomplished if the capture probability is significantly less than one. It also indicates that the improved filtering will be achieved if the entrance composite capture probability for the sheet is not one.

[0869] The average results of many trials for a two layer stack of a 5×26×grid are summarized in 22. All data points represent the average of at least one hundred trial simulations and thousands of particles passing into the media. This is done because of the statistical nature of the capture process. In this series of experiments only filtering passages were used. Additionally, the capture probabilities for channels in any direction were equal. The base grid was by our designation a 5×26×1 grid, and two layers were used. We designate this combination grid as 2×(5×26×1). This is equivalent to a 5×26×2 grid where the layer interconnect channels have an infinite flow resistance.

[0870] Three curves are plotted: one curve for the total particles processed before all flow stops, one curve for the particles trapped before all flow stops, and one curve for the particles escaped from the filtering sheet before all flow stops. The curves are plotted as a function of the capture probability. As the probability increases the number of particles processed is decreased before all flow through the sheet stops. At the same time, the number of particles that escape from the sheet also decreases. It tends to zero as the probability approaches 0.2, and becomes zero at a probability of 1. As it happens for this double layer 2×(5×26×1) grid, approximately 0.2 percent of the particles escape when the capture probability is 0.2, but there is no simple correlation between these variables.

[0871] The results for the two layered grid 2×(5×26×1) of FIG. 22 and the results for a 5×26×2 grid are approximately equal when the capture probability is greater than 0.02.

[0872] Increases in the numbers of particles processed may be achieved by expanding filter media volume by increasing the width dimension of the sheet or the numbers of layers in the sheet. Surprisingly, increasing the depth dimension does not improve filtration. We have documented the effect of increasing the depth dimension from 26 rows of nodes to 52 and 98 rows. The findings are illustrated in FIG. 23. Here all the results for the grids 5×26×2, 5×52×2 and 5×98×2 plot closely together and are approximated by one curve for the particles processed versus probability, trapped particles versus probability and escaped particles processed versus probability. Also plotted are the results for the 2×(5×26×1) grid.

[0873] It is found that there is no substantial variation in filtration and no substantial improvement in filtration by increasing the filtration media size in the depth dimension. However, the pressure drop through the sheet is increased when this is done, and the mechanical strength is increased.

[0874] Also investigated was the effect of different filtration end points. It is found that there is no substantial difference in the plotted curves for the number of particles processed, trapped, or escaped when the end point for filtration is a 50 or a 100 percent reduction in flow through the filter. In FIG. 23 we have aggregated this data together. These results are only for probabilities of 0.02 and higher. When the grid depth is reduced below 26, at some point the results begin to become depth dependent.

[0875] The finding of a parameter region of depth independence and end-point independence is counterintuitive and not expected. But, it is significant and useful. These results indicate that for a given set requirements an optimum filter media depth exists. Beyond this depth, increasing the depth increases cost without benefit of improved filtration. In general, it appears that on the average it may only take a very limited number of particles to reduce the flow from a value of 50% of the initial start-up value to a no flow condition when uniform grids of constant capture probability are used.

[0876] In general, we find that regularly spaced designed grids in a flow device may serve as useful particle filters. Preferred are cubic grids where node points are laid out in rows, columns, and layers. The columns are through the depth direction where the depth direction corresponds to the direction of flow through the filtration sheet or media. The grids extend in planes. The rows in a plane are generally normal to the direction of flow through the filtration media. Layers generally consist of one plane of rows and columns of node points. Flow passageways connect adjacent and near nodal points and form a lattice of interconnected passageways. It is preferred the filtering flow passageways have similar characteristics to the flow distribution passageways described prior to this section.

[0877] In its simplest form a useful filtering sheet or block of media flow device may be a two dimensional grid of passageways or layers of 2D grids similar to a simple improved fluid distribution sheet described earlier. Micro-channel grids of flow paths on the surface of a thin sheet of material may be manufactured by known micro-replication techniques and stacked together to form filtration media.

[0878] Devices for filtration or filter media elements may be assembled of multiple sheets stacked together. Filtration media may consist of material containing multiple layers or planes of two dimensional grids. Filter media may also be three dimensional grids of flow passageways, or assemblages of two and three dimensional grids of passageways. Preferred are passageway assemblages which repeat within the volume of the filter media. Preferred are passageway assemblages which repeat within the area of a plane of a filter media sheet.

[0879] Our findings, some of which are exemplified by graphed results in FIG. 23, indicate that improved filtration may be obtained with grids of more than 5 rows of nodes when the capture probability is below 1, preferably below 0.5, more preferably below 0.2, and most preferably between 0.02 and 0.2. Additionally, we have found that the pressure drop through the filter may be minimized without significantly sacrificing particle capture. This is accomplished by using grids with snore than 5 rows and less than 52 and preferably less than 27 rows of node points.

[0880] b. Perfect Filtration—No Particle Escapes Using Filtering Grids

[0881] Often manufacturing processes require totally particle free fluid flow from a filter. This may be accomplished by designing a filter with an overall capture probability of 1.0 for the target particle. That is a filter element where the combination of flow passages acting together as a whole create a particle capture probability of 1.0. A filter sheet with only one or several rows of nodes, and where all individual passageways have a capture probability of 1.0, may be used. However, the number of particles that can be filtered from the inlet flow before flow is stopped or significantly reduced is relatively small. It is generally limited to the number of node columns times the number of layers in the case of a design fabricated in a grid of 3D cubic forms. It is generally limited to the number of passageways directly exposed at the media entrance edge. Therefore, the volume of fluid that may be processed is low. In the prior art this limitation is dealt with by expanding the filter inlet area by various means. We have found that this is not necessary and that improvements may be obtained by using filter sheets with designed variations (predetermined) in the internal structure and functioning of the sheet. This is also found with filter media in general and not just sheets. One key element is to design the row by row capture probabilities of arrays of internal channels so that they have a value less than 1.0 for the majority of the rows.

[0882] Commercial porous media is often constructed using an assemblage of spherical particles. One type of assemblage may be constructed using close packed spheres of equal sizes. This media may be modeled by a modified 3D cubic grid of flow passages with flow resistances and particle capture probabilities.

[0883] Irregular media particles of a known size classification are often used to construct a filter media. Cubic grids are useful to model the flow and particle capture for this media. In general, the filter may be presented with a filter challenge consisting of target particles of a uniform size or of a known distribution. The pores may have a known distribution of flow resistances and particle capture probabilities.

[0884] i. Bimodal Distributions—“bimodal total capture distributions”

[0885] In the following examples, target particles of uniform properties are used for purposes of illustrating new principles of filter design. A uniform capture probability for all channels throughout the bulk of the media are used except for passageways near media exit. Designing the sheet with a probability of one for just the exit channels prevents particles from escaping. We refer to these distributions as bimodal total capture distributions. Here the passageway capture probabilities in the bulk of the filter are at a constant value less than 1.0, but the exit or near exit channels have a value equal to 1.0. One method of obtaining a capture probability of 1 is generally to set the flow passageway hydraulic diameter to a value less than 80 percent of the target particle hydraulic diameter. Simulation results for these distributions are shown in FIG. 24 It indicates there is an optimum capture probability for the channels leading up to the exit node row. For the sheet grid example presented the optimum probability is 0.08. The upper two curves indicate on the average a volume of fluid containing 80 particles may be processed by the grid geometries 5×26×2 and 5×52×2, and the curves for all practical purposes are identical. These results represent the average hundreds of experiments.

[0886] If the probability for all sheet channels were 1 for this example, only the fluid volume containing 10 particles could he processed. The two level gradients in probability, this bimodal distribution, can provide an 8 fold improvement. The exit edge interface channels are preferred for the location of the capture probability of 1.0. If the channels at the inlet sheet edge for the 5×24×2 and 5×50×2 grids have a probability of 1, only the volume containing 10 particles may be processed before total clogging occurs and flow from the filter stops.

[0887] Unexpectedly, the bimodal total capture distributions do not provide ever improving filtration when more depth is added to the sheet. As FIG. 24 shows, 52 rows of nodes will not capture significantly more particles than a sheet with only 26 rows. Additionally, if the near exit channels with a capture probability of 1.0 for the 5×52×2 grid were to include the all the channels from the 27.sup.th to the 52.sup.nd rows, the filtration results are about equal to the upper curves of FIG. 24.

[0888] The inclusion of these rows in the filter design do not improved the filtration when the rows 27 to 52 have a capture probability ranging between the two values of the bimodal distribution. They only provide a redundancy factor in case there is some defect in the preceding channels. This finding on the influence of filter media depth is similar to that for filter media where the capture probability is single valued. This is also true when the filter is challenged with distributions of particle sizes, and it is true when the individual pore capture probabilities vary randomly about a mean value equal to the nominal average of the bulk of the media and final exit pores have a probability of one.

[0889] FIG. 24 also plots a curve for a filter sheet consisting of two 2D grids 2×(5×24×1). The reduced total number of filtering channels reduces the particles captured. The results illustrate that 3D grids are preferred over 2D grids if only the number of captures is considered.

[0890] When designing filter media with bimodal channel particle capture probabilities, it is preferred that one probability be at or near 1.0 and the other be between 0.05 and 0.2. It is more preferred that the average probability be between 0.06 and 0.1. It is also preferred that the high probability be at or near the exit edge. For bimodal probability filter sheets, after the depth exceeds about a dimension so that a particle flows through about 15 pores sequentially, the benefit of adding additional passageways for extending the sequence is very limited while the pressure drop for forcing flow through the media increases proportional to the number of passageways. Preferred are micro-replicated sheets with 5 to 50 rows of nodes. Most preferred are sheets with 20 to 40 rows of nodes.

[0891] The results in FIG. 24 are for a simple grid of filter passages similar that are illustrated in FIG. 4a. Similar trends and results are obtained with other grid layouts and types. Examples include those illustrated in FIGS. 6, 7, 8, 9, and 11. We find the general results and conclusions also apply to other more complex grids, and to like grids when they comprise subunits of a collection that acts together to form a filter or filter element. While we have discussed the simple filter grid results in FIGS. 21, 22, 23, and 24, and they have been discussed in terms of them as representing filter media, a filter may be composed of multiple units of filtering grids. These may be arranged and interconnected in series and parallel and in complex patterns. These units may be directly connected to an inlet and outlet face or edge or remotely fluidically connected to other inlets and outlets.

[0892] ii. Multimodal Distributions

[0893] Further studies have shown that gradient, multimodal capture distributions improve results over the bimodal and single value capture distributions. Improvements in filtration may be obtained using linear, parabolic, exponential and other multimodal gradients where the passageway capture probability increases from the inlet to the outlet edge of the sheet. FIG.25 shows a useful gradient for a 5×26×2 grid which allows the fluid volume containing approximately 125 particles to be processed. This is a fifty percent improvement over the best the results for the bimodal distributions in FIG. 24

[0894] In all the grids used in the experiments illustrated by FIG. 24, and with the capture probability distribution shown in FIG. 25, the designs allow no particles to escape the filter.

[0895] Using a sheet design where the capture probability advances from a low value to high value in the depth direction, from inlet to outlet, improves filtration performance. When for any row of nodes in a micro-replicated sheet has a capture probability of one for the channels conveying flow away from the row, no particles will escape from the sheet. As before, if this row of nodes is placed towards the outlet edge the filter media, filtration may be improved. Most preferred is to have this row of nodes at the outlet edge. In a like manner for porous media, it is preferred that the capture probability advance from the low to high in the flow direction through the filter depth.

[0896] Improved filtration generally results when multimodal total capture distributions with gradients in probability from inlet to outlet are employed. Furthermore, improvements that are proportional to the increased number of rows in the depth direction may be obtained. If we double the depth of the sheet we can double the number of particles trapped by the sheet in these cases. It is preferred that three or more probabilities are used for sequential passageway regions in the depth flow direction. Most preferred are five or more, or a continuous variation with depth direction.

[0897] FIG. 26 shows another useful multimodal total capture distribution for a sheet with a depth of 48 rows. While the 26 node row sheet with probabilities illustrated in FIG. 25 captures on the average approximately 125 particles, the 48 row sheet defined by FIG. 25 captures approximately 250 particles. This illustrates that gradients in probabilities allow the volume of the media to be used more efficiently and particle capture rates to be proportional to the volume of the media and not the inlet area.

[0898] We have investigated regular two and three dimensional grids of filter channels. Multimodal capture probability distributions where all flow at some point passes through channels with a capture probability of 1.0 are preferred. This results in the capture of all particles and prevents the passage of particles beyond the filter.

[0899] Many grids are possible for constructing micro-replicated filter media. The examples illustrate results for two layer cubic grids and single layers in sheets. Other experiments have been performed for multiple layers, various sheet dimensions, and various supply channel arrays. The simple layouts of channel grids described above are square or rectangular lattices of channels in layers or planes. Other single and multilayer grids of channels are useful. Single and multilayer grids of pores of distributed sizes are useful. Our findings and conclusions hold in general for all of these.

[0900] When porous media is assembled of particulate material, a construction which increases the average particle capture probability locally in the direction of flow is preferred. This increases the total number of particles that may be captured.

[0901] The conclusions and findings are valid for filter media in general and not just sheets or media with regular repeating geometric grid assemblies. The conclusions are valid for filter media with random pores, a distribution of pore sizes, or the filtering of particles with a distribution of sizes. When porous media is assembled of particulate material, a construction which increases the local average particle capture probability in the direction of flow is preferred. This increases the total number of particles that may be captured.

[0902] c. Improved Filtration Using Auxiliary Channels

[0903] We have shown that it is preferred to have multimodal capture probability distributions for the filtering flow grids. It has been also found that even more efficient filtering may be achieved by more complex types of grids. Unexpectedly, we have found that allowing particles to pass more freely toward the outlet edge of the sheet or media may improve filtration, It has also been found that allowing fluid to pass more freely from within the media to the outlet face or edge may improve filtration. In the past, it has been thought that “short circuiting” the filter media is detrimental. One embodiment of the invention is to employ auxiliary channels and auxiliary micro-channels within the media to allow fluid containing particles to pass easily from the inlet edge to regions within the depth of the sheet. Auxiliary micro-channels are positioned within the body of filtration media, and preferably throughout the bulk of the media. While enhancing particle movement toward the outlet of a filtration sheet might he expected to degrade the filtering performance, experiments show dramatic improvements actually result.

[0904] Auxiliary channels enhance flow in the width, depth and layer thickness directions within the porous media. Just a few channels dramatically and unexpectedly reduce pressure drops and improve particle filtration.

[0905] A “non-filtering auxiliary flow grid” is a special grid defined as a grid of passages having particle capture probabilities much lower than the filtering channels present. Preferably, these are at or near zero for the particles or particle size distribution being targeted. Their properties may be chosen so that they generally never become clogged. Complex types of compound flow grids employing the use of non-filtering auxiliary flow grids are beneficial. These serve to distribute flow to multiple points and regions within the filter media. Auxiliary channels of this teaching are channels with hydraulic diameters generally ranging from microns to a few millimeters. Micro-channels are preferred for the auxiliary channels.

[0906] It is useful that auxiliary channels be substitute for some of the primary filter grid flow pores, or be superimposed upon a single layer filtering micro-replicated grid. It is preferred that the auxiliary channels have different flow resistances and particle capture probabilities from the filtering grid. It is preferred that their capture probabilities be much less than the base filtering grid. More preferred are probabilities near zero. Examples of 2D micro-replicated combination grids including a base filtering grid and non-filtering auxiliary channels are shown in FIGS. 27, 28a, 28b, 28c, 30, 31, 32, 33, 35a, 35b and 34c. Only representative sectional areas are shown.

[0907] FIG. 27 illustrates schematically a section of a micro-sheet with a basic filtering square 2D grid 270 in one plane with one flow grid layer. The thinner lines represent a filtering grid, and the bold lines represent a non-filtering auxiliary grid 271. Arrow 272 indicates the direction of flow through the sheet. It is from an inlet edge represented by the bold grey line 273 to an outlet edge represented by the bold grey line 274. Replacing a few of the filtering grid channels are the channels of a non-filtering auxiliary grid. Here the auxiliary channels 271 of the non-filtering auxiliary grid type provide alternative flow paths into the depth of the filter sheet. With proper choices of the particle capture probabilities of the channels of the filtering grid, and of the relative flow resistances of the two types of grids, improved filtration may be obtained relative to the filtering grid alone without auxiliary channels.

[0908] In FIG. 27 the filter media is subdivided into sub regions 276, 277, 278, and 279 by the placement of auxiliary channels among the filtering channels. These sub regions contain filter channel grids that are in the plane parallel of the layer surface and are surrounded by auxiliary channels or the edge or the sheet. FIGS. 27, 28a, 28b, 28c and 32 illustrate the use of auxiliary channels to form sub regions. It is preferred that the filter performance of these be improved by varying the channel capture probabilities within them. Generally, it is preferred that the probabilities within a sub unit increase in the flow direction indicated by the arrow 272. Bimodal capture probability distributions for the filtering grid are preferred. Multimodal capture probability distributions for the grids are most preferred for these sub units. When 3D grids of filter passageways and 2D and 3D grids of auxiliary channels are present, the same types of probability variations are preferred.

[0909] In general, when auxiliary channels arc present, the flow from the inlet face to the outlet face may take place through numerous paths or sequences of filter. Each path will have a cumulative capture probability dependent upon the many individual channels in the path. It is preferred, but not required, that the sequence of capture probabilities have increasing values along the flow path.

[0910] FIG. 28a illustrates another sheet section with a basic filtering square 2D grid 283 in one plane with one flow layer. This filtering grid would be designated as an 11×15×1 grid indicating it has 11 columns of nodes across the width, 15 rows of nodes through its depth, and only nodes in one plane or layer. Arrow 280 indicates the direction of flow through the sheet from an inlet edge 281 to an outlet edge 282. Replacing some of the filtering grid passageways are the channels of a non-filtering auxiliary grid shown as bold lines. Here the auxiliary channels 284 of the non-filtering auxiliary grid provide alternative flow paths into the depth of the filter sheet. With proper choices of the particle capture probabilities of the auxiliary channels, of the filtering grid, and of the relative flow resistances of the two grids, improved filtration may be obtained relative to an un-enhanced filtering grid. Here a single valued capture probability equal to or near 1.0 for the filtering grid is useful.

[0911] FIG. 29 graphs filtration experiment average results for a filter sheet of the style illustrated in FIG. 28a with the filtering grid having a 5×22×2 geometry. That is it has 5 columns, 22 rows and 2 layers of nodes. All the filtering cross width channels and all interlayer filtering channels have been replaced with non-filtering auxiliary grid channels. Additional non-filtering auxiliary channels have also been added to complete the serpentine path from the inlet sheet edge to near the outlet edge. Our simulations assumed that all the filtering grid channels have a particle capture probability of 1.0. The ratio of the flow resistance per unit length of each channel increment the non-filtering auxiliary grid, to that of the filtering grid was varied from 0.001 to 1.0. This ratio is the dimensionless auxiliary channel flow resistance. Curve 1 plots the average particles trapped by the filter sheet before all flow is stopped by clogging of the filter. Curve 2 plots average particles trapped by the filter sheet before the flow through the filter drops below 50 percent of the initial flow rate. The pressure drop from the inlet to outlet side is maintained constant to generate curve 2.

[0912] The sheet grid geometry for the FIG. 29 results has 170 active filtering channels in the filtering grid with the non-filtering auxiliary channels in place. By selectively choosing the dimensionless auxiliary channel flow resistance ratio to be greater than 0.03, the number particles that may be captured before the flow stops approaches the number of active filtering channels. This results because the filter geometry tends to fill and clog sequentially from the inlet to outlet edge for parameters chosen.

[0913] In many filtration tasks one needs to stop the process when the flow through the filter drops below 50 percent of the initial flow rate. Curve 2 illustrates results for this end point. Curve 2 data shows that for the filter design corresponding to the FIG. 29, when the dimensionless auxiliary channel resistance is 0.03, the filter design still allows on the average an amazing 152 particles to be captured. Stating this another way, the filter design allows 93 percent of all active filter channels to be used to trap particles while only reducing the flow rate by only 50 percent. This results because of the presence of the auxiliary channels.

[0914] For all results in FIG. 29 no particles escaped from the filter because of the high capture probability chosen for the filtering grid.

[0915] The FIG. 29 results illustrate in general there will be a range of preferred values for the dimensionless auxiliary channel flow resistance. A ratio of 0.002 to 0.2 is preferred. A ratio of 0.01 to 0.1 is more preferred. A ratio of 0.02 to 0.06 is most preferred.

[0916] For porous media where the pores have a size distribution and a grid of non-filtering auxiliary is employed, it is preferred that the ratio of the flow resistance of the auxiliary non-filtering channels to the average pore flow resistance per unit length be in the range of 0.001 to 0.1

[0917] In FIGS. 27, 28a, 28b, 28c, 30, 31, 32, 35a, 35b, and 35c the thin grid lines represent the filtering channels. The bold black lines represent the auxiliary non-filtering channels. The bold grey lines represent inlet and outlet edges. While these illustrations are all for variations of square and rectangular grids in a single plane, flow device designers with ordinary skills will recognize that the same principles may be extended to other grid layouts including those formed from hexagons, triangles, diamonds, quadrilaterals and other geometries and tessellations. The principles may be extended to filter media where there are a large number of layers of grids in planes parallel to the general flow direction. Generally, the layout of the auxiliary channels will be in a repeating pattern for purposes of uniformity of action and ease of fabrication. This however is not necessary for them to function to improve filtration.

[0918] FIGS. 27, 28a, 28b, 28c, 30, 31, 32, 35a, 35b, and. 35c illustrate types of filter grids and auxiliary channel geometries. They are meant to represent design concepts. As such single thin lines may also be considered to represent a multiplicity of passages. Also the areas enclosed by the filter passageway lines or auxiliary channel lines may he considered to contain additional filter passages and grids of filter passages that are not shown.

[0919] It is also a teaching that flow passages which are referred to by the terms filtering channels, flow distribution channels, auxiliary channels, filtering pores, flow distribution pores, and auxiliary pores in filtering media do not need to be uniformly and regularly spaced or be in regular repeating geometric patterns. The spacing and patterns may be irregular and be randomly variable. The grid or lattice of passageways in two dimensional planes may be composed of a multiplicity of differing polygons and curve sided and straight sided geometries. The grids may be Voronoi tessellations in 2 and 3 dimensions. The grids may be formed by random processes or a tessellation process or any tiling process.

[0920] FIG. 30 illustrates a section of another sheet with a basic filtering, square 2D grid 303 in one plane in one flow grid layer. Arrow 305 indicates the direction of flow through the sheet from an inlet edge 301 to an outlet edge 302. Replacing some of the filtering grid channels are the channels of a non-filtering auxiliary grid. Here the auxiliary channels 304 of the non-filtering auxiliary grid provide alternative flow paths into the depth of the filter sheet. These are distinguished as bold lines. In this style of non-filtering auxiliary grid implementation, the auxiliary channels extend both from the inlet edge and the outlet edge. With proper choices of the particle capture probabilities of the channels of the filtering grid and the dimensionless auxiliary channel flow resistance, improved filtration may be obtained relative to the filtering grid alone.

[0921] FIG. 31 illustrates a section of a micro-channeled sheet with a basic filtering, square 2D grid 313 in one plane in a sheet. Arrow 315 indicates the direction of flow through the sheet from an inlet edge 311 to an outlet edge 312. Here the auxiliary Channels illustrated as bold lines 314 provide alternative flow paths into the depth of the filter sheet. In this style of auxiliary channel grid implementation, the auxiliary channels extend both from the inlet edge and the outlet edge, and do not replace any of the filtering grid passageways. With proper choices of the particle capture probabilities of the filtering grid channels and the dimensionless auxiliary channel flow resistance, improved filtration may be obtained relative to the filtering grid with no auxiliaries.

[0922] FIG. 32 illustrates a section of another sheet with a basic filtering, square 2D grid 323 in one plane with one flow grid layer. Arrow 325 indicates the direction of flow through the sheet from an inlet edge 321 to an outlet edge 322. Replacing some of the filtering grid channels are the channels of a non-filtering auxiliary grid. The auxiliary channels 324 of the non-filtering auxiliary grid provide alternative flow paths into the depth of the filter sheet. In this style of non-filtering auxiliary grid implementation the auxiliary channels enclose regions 326 containing only filtering channels. With proper choices of the particle capture probabilities, and the dimensionless auxiliary channel flow resistance, improved filtration may be obtained relative to a simple filtering grid. Preferred capture probability distributions in the enclosed regions are bimodal and multimodal. Preferred are distributions where the probability increases in the general flow direction.

[0923] 4. Improvement of Known Porous Media with Auxiliary Channels

[0924] FIG. 33 illustrates a section of a conventional porous sintered metal sheet modified with auxiliary micro-channels 332 according to the teaching of this invention. An edge view is shown along with a mid-plane, sectional top view. The porous metal material 330 comprises the bulk of the sheet. The individual pores are not illustrated. Auxiliary micro-channels 331, 332 and 334 comprising part of a non-filtering auxiliary grid present within the sheet are shown. Micro-channel 334 extends from the inlet face 336 of the sheet. Arrow 335 indicates the general direction of flow from the face edge to the outlet face of the sheet. Some auxiliary micro-channels 334 extend in the depth direction. Others 332 extend in the width direction. The rectangular grid defines regions of influence. They are delineated and outlined by the micro-channel closed loops in the plane. An example is region 311. Region 333 adjacent to the outlet edge of the sheet contains no auxiliary micro-channels.

[0925] The non-filtering auxiliary grids geometries of FIGS. 27, 28a, 28b, 28c, 30, 31, 32, 35a, 35h, and 35c all may be used to improve the flow through the internal structure of conventional granular porous media including, but not limited to granular, particulate, fibrous and various combinations. This media may be in the form of a sheet, layered sheets, and three dimension blocks.

[0926] While this illustration of the invention is of a sheet of granular porous media, and includes only a one layer 2D grid of auxiliary channels, porous media containing regular 3D grids of auxiliary channels are also part of the teaching of this invention. Random placement and orientations of the passageways of one or more auxiliary grids are also a teaching.

[0927] FIG. 34 illustrates a sectional view looking down on a portion of a sheet containing micro-channels in a single plane. FIG. 34 is a sectional view of conventional sinter metal porous sheet containing multiple planes or layers of auxiliary micro-channels through its thickness. It contains in one plane channels 340. It also contains inter-plane channels 341 that allow flow to pass between the grids of channels on different levels.

[0928] Careful studies of conventional porous media simulations have been performed. In conventional porous media the filtering pores may consist of the void volumes in beds of the media formed from assemblages of particles. The pores may be regular or irregular shapes. The pores may have various size distributions. The particles may be of uniform size. They may be obtained from conventional screening processes or maybe produced by other methods.

[0929] It is also a teaching that non-filtering auxiliary channel grids may be used to improve porous media when incorporated into the internal matrix of the porous media. The media includes assemblages of micro-channel sheets and granular porous materials. Conventional media, as exemplified by sintered porous metal, has an array of particles with a distribution of sizes. Other conventional media materials also have randomness. We have found that filter media may be improved by the placement of auxiliary micro-channels within the media. Commonly, making a block of conventional porous media more porous improves its ability to flow fluid through it while decreasing its ability to filter out contaminating particles. This invention allows improved flow and simultaneously improves filtering by the media. Using micro-channel auxiliary grids to modify the filtering porous internal structure and connectivity of conventional media improves filtration and is a teaching.

[0930] Micro-channel and auxiliary channel modification of the internal structure of known porous media allows more effective use of the internal pore structure and reduces flow resistance. It has been found the inclusion of auxiliary channels within the porous media substantially improves filtration. Preferred channels include internal repeating geometries of flow grids.

[0931] Improve filtration may also he achieved with the random grids or non-regular grids of auxiliary channels. The filtration improvements are obtained without substantial reduction in the strength of a filter element and its ability to resist shear, compressive, tensile and bending forces.

[0932] Consider the case of a replaceable filter element which fits into the volume space which is a cube. Further consider the case where the flow through this element is from the front to the back face and where the other faces are not permeable to flow. If the filter media is a fused granular matrix, the strength of the element depends only on the strength of the matrix. The maximum element strength is obtained when the media totally fills the cube of volume. The same is true if the filter media is a layered structure containing grids of channels where the layers are fused together. Any removal of portions of media to create a structure of greater inlet or outlet surface area on the media within the cube volume produces an element of lower strength. It also lowers the mass of filter media per unit volume.

[0933] We have found placing grids of auxiliary channels, or a plurality of micro-channels within the filter element base media material more porous and improves the filtering action. This allows the mass of filter media to totally fill the volume for maximum strength. These channels simultaneously promote the use of all internal volume of the media for capturing particles while lowering the pressure drop through the media.

[0934] These very small auxiliary channels do not significantly reduce the strength of the element. These very small auxiliary channels do not significantly reduce the active volume or mass of filter media within the limited space of the volume of the element.

[0935] Without the auxiliary channels the performance of the filter element is dependent upon the exposed inlet edge face area. With auxiliary channels the performance of the filter element may be made dependent upon the volume of the media in addition to the area of an inlet face or faces. Conventional filter media tends to concentrate the entrapment of particles at the faces of the filter element. Auxiliary channels promote the use of the entire volume and improve filtration. The auxiliary channels are fluid transport pores designed so as to avoid contaminant capture within them. They are pores designed and located to enhance fluid transport and contaminant flow into the media. And again in general, the addition of auxiliary channels does not significantly change the mechanical strength of the media.

[0936] Any known porous media may be used for the bulk of the new inventive media with auxiliary channels and auxiliary transport pores. The bulk media described as conventional porous media may be comprised of granular materials, fibrous materials, and combinations. The chemical compositions may be singular or mixtures of compounds or elements. The individual grains or fibers may be bonded together to form a matrix. They may be simply packed together or loosely associated in a confined volume.

[0937] The conclusions and findings are valid for filter media in general and not just sheets or media with regular, simple, repeating geometric grid assemblies. The conclusions are valid for filter media with random pores, or a distribution of sizes, or the filtering of contaminates with a. distribution of sizes.

[0938] The following examples and simulation results illustrate the improved filter media that may be obtained using the teachings of this invention.

[0939] In the prior art one method of improving filter performance is to increase the inlet edge area through which the fluid flows. One illustration of this teaching is U.S. Pat. No. 7,125,490. Here a folded sheet structure is used. This is schematically illustrated in FIG. 36. Shown is a cross section of a segment 360 of the structure. The media is shown as the crossed hatched area. Fluid flows to an inlet side indicated by arrows 362 and from an exit side indicated by arrows 364. The folding of the media creates increased area within the volume using the large inlet side indents 361 and outlet side indents 363. The view is a cross sectional view where the pleated filter geometry extends in the direction normal to the section. The total volume enclosing the filter media in cross section is outlined in this view by dashed line box 365. The volume is roughly twice the actual volume of the media. Half the total volume of the filter element is used to enable fluid to access the extended surface area of the folds.

[0940] A teaching of this invention is to fill the total volume indicated by line 365 with filter media modified by auxiliary flow channels where these facilitate penetration of particles into the total volume. The auxiliary flow channels have hydraulic diameters slightly larger to larger than the nominal or average pore size of the media. It is preferred that the auxiliaries are distributed within the media with a repeating, predetermined geometry. It is preferred that a multiplicity of the auxiliaries provide auxiliary flow paths from the inlet edge of the media and into the internal volume of the media. It is preferred that the auxiliary channels have a low probability of capturing the particles being filtered. It is most preferred that their probability of capturing the particles being filtered is near zero. It is preferred that the auxiliary channels be micro-channels.

[0941] Tests have shown that using media of this invention with the auxiliary channels allows the filter to remove three to more than six times more particles than the conventional folded structure of FIG. 36. The media of this invention will have more than twice compressive strength of the folded structure, more than 7 times the resistance to shear, and more than 120 times the resistance to bending.

[0942] FIG. 37 illustrates the device of Kelly et al. in U.S. Pat. No. 7,361,300. The cylindrical element 370 filters fluid flowing from top to bottom. The entering flow indicated by arrows 372 enters the top inlet face and into the large holes 371 extending into the volume of the element. The holes increase the filter inlet face area. Holes also extend from the bottom outlet face into the element. These are not shown. Fluid flows out the bottom of the element as indicated by arrow 373. The illustration figures of the Kelly et al. patent illustrate hole placements that remove approximately half the media volume and mass of the element. Although the sintered metal media described is strong, the elimination of half the mass reduces the compressive strength by roughly one-half Our modeling studies indicated the volume of the element is inefficiently used.

[0943] Modeling of the use of micro-auxiliary channels indicates a two to eight times improvement in filtration may be achieved by filling the entire volume taken up by the element with media and using a multiplicity of the auxiliary channels internally. Investigation of the detailed reasons for the improvement indicates that the improvement is first a result of having more mass of media present in the element and secondly that the auxiliaries allow a greater percentage of the internal pores to trap particles.

[0944] A simple useful auxiliary channel 3D grid geometry is illustrated in cross section in FIG. 38. Auxiliary channels penetrate the media volume from the inlet and outlet faces. The flow is from an inlet face to a parallel outlet face. FIG. 38 illustrates a small sub-cross-section 385 of a cross section of the filter media taken parallel to the inlet face. This subsection is bounded by the line 380 that defines the cross sectional area of the subsection. This is a region which abuts edge to edge with like regions that fill the total cross section of media. Running perpendicular to the cross-section are inlet auxiliary passageways that intersect the inlet face. These are indicated by the small squares 383. Small circles 382 indicate outlet auxiliary passageways that parallel the inlet auxiliary passageways. The outlet passageways intersect the outlet face of the element, but not the inlet face. Likewise the inlet auxiliary passages do not intersect the outlet face. The size of the circles and squares are not necessarily proportional to their actual size. They only serve to locate the relative positions.

[0945] The volume of the media is filled with a regular cubic grid of filtering passages that run parallel to the three major axii of the element. They intersect each other and totally fill the volume. These are the pores that filter the fluid passing through the media. These individual pores are not shown in this schematic.

[0946] Each outlet auxiliary channel is surrounded by a performance region of influence indicated by the large dashed square areas 384. If an inlet auxiliary channel lies within the region, the probability of a particle escaping from the filtering grid and passing into the outlet flow in an outlet auxiliary channel is greater than a target level. The target level depends upon the design specifications of the filter. For example it might be 1 percent of the particles processed, 0.1 percent or some other value. A preferred grid of auxiliaries is shown in FIG. 38. It consists of eight outlet auxiliaries located on the perimeter of the subregion illustrated. Further, all the inlet auxiliaries are confined to an area in the form of a cross indicated by the grey area 381 within the subregion 385.

[0947] One useful space placement of auxiliaries is shown. It has been found that each inlet auxiliary is surrounded by a small square zone of influence indicated by the small dashed lined squares 386. Only two are shown, but all inlet auxiliaries are surrounded by these zones of influence. It is found that when two inlet auxiliaries exist within a zone of influence, they restrict the filtration performance.

[0948] When a first inlet auxiliary exists in the center of its zone of influence, it is found that on the average a given number of particles will flow from the auxiliary into the filtering pores and be trapped. The number will depend at least on the average capture probability of the pores, the flow in the auxiliary, the relative resistances to flow in the pores and the auxiliaries, and the desired filtration end point. When another inlet auxiliary is placed within the influence zone of the first, the particles filter will not double. The two auxiliaries will interact to restrict the filtration performance expected by two single isolated auxiliaries.

[0949] For a given set of filtration and media parameters, there will tend to be an optimum number and an optimum distribution of the auxiliaries.

[0950] The filter pores need not be arranged in cubic grids. Grids of pores created in the sintering granular materials are more random. The filtration performance of granular metal and plastic media may also be greatly improved by the placement of auxiliary channels and micro channels within the media. Generally, this does not significantly change the number of pores available for filtering. Conventional filter element designs limit capture of contaminants to pores near the inlet face. The auxiliary channels of this teaching allow the pores within the whole element volume to function to capture particles.

[0951] If the volume of an element is filled with sintered granular media, a simple useful 3D grid of auxiliary channels consists of two sets of passageways that penetrate the inlet and outlet faces. Flow is from an inlet edge or face to an outlet face. FIG. 39 illustrates a small sub-cross-section 395 of a cross section of the filter media taken parallel to the inlet face. This subsection is bounded by the line 390 that defines the area of the subsection. This is a region which abuts edge to edge with like regions that fill the total cross section of media. Running perpendicular to the cross-section are inlet auxiliary passageways that intersect the inlet face. These are indicated by the small squares 393. Small circles 392 indicate outlet auxiliary passageways that parallel the inlet auxiliary passageways. The outlet passageways intersect the outlet face of the element, but not the inlet face. Likewise, the inlet auxiliary passages do not intersect the outlet face. The size of the circles and squares are not necessarily proportional to their actual size. They only serve to locate the relative positions.

[0952] By nature granular and other known media, their volume is filled with a random grid of interconnected passages. Passages intersect each other throughout the volume. These pores filter the fluid passing through the media. These individual pores are not shown in this schematic.

[0953] Each outlet auxiliary channel is surrounded by a performance region of influence indicated by the area delineated by the large dashed circle 394. If an inlet auxiliary channel lies within the region, the probability of a particle escaping from the filtering grid and passing into the outlet flow in an outlet auxiliary channel is greater than a desired level. The target level depends upon the design specifications of the filter. Therefore, the size of this circle depends upon the characteristics of those chosen for the filter and the media. For example the desired level of particle escapes might be 1 percent of the particles processed, 0.1 percent or some other value. A preferred grid of auxiliaries is shown in FIG. 39. It consists of eight outlet auxiliaries located on the perimeter of the subregion illustrated. Further, all the inlet auxiliaries are confined in an area in the form of a cross indicated the grey area 391 within the subregion 395.

[0954] One useful spatial placement of inlet auxiliaries is shown. It has been found that each inlet auxiliary is surrounded by a circular zone of influence indicated by the small dashed circles 396. Only one is shown, but all inlet auxiliaries are surrounded by these zones of influence. It is found that when two inlet auxiliaries exist within a zone of influence, they restrict the filtration performance.

[0955] When a first inlet auxiliary exists in the center of its zone of influence, it is found that on the average a given number of particles will flow from the auxiliary into the filtering pore and be trapped. The number will depend at least on the average capture probability of the pores, the flow in the auxiliary, the relative resistances to flow in the pores and the auxiliary, and the desired filtration end point. When another inlet auxiliary is placed within the influence zone of the first, the particles filter will not double. The two auxiliaries will interact to restrict the filtration performance expected by two single isolated inlet auxiliaries

[0956] For a given set of filtration and media parameters there will tend to be an optimum number and a distribution of optimum placements.

[0957] FIGS. 40 and 41 illustrate small subsections of cross sections of the filter media closely related to that illustrated in FIG. 39. These subsections are bounded respectively by the lines 400 and 410 that define the area of the subsections. These are regions which abut edge to edge with like regions that fill the total cross section of media. Running perpendicular the cross-section are inlet auxiliary passageways that intersect the inlet face. These are indicated by the small squares 403 and 413. Small circles 402 and 412 indicate outlet auxiliary passageways that parallel the inlet auxiliary passageways. The outlet passageways intersect the outlet face of the element but not the inlet face. Likewise the inlet auxiliary passages do not intersect the outlet face. The size of the circles and squares are not necessarily proportional to their actual size. They only serve to locate the relative positions.

[0958] Again the volume of the element is filled with a random grid of pores. Some of them are interconnect. These pores filter the fluid passing through the media. The individual pores are not shown in this schematic.

[0959] Each outlet auxiliary channel 402 and 412 is surrounded by a performance region of influence indicated by the area delineated by the dashed circles 404 or 414. Here to, if an inlet auxiliary channel lies within the region, the probability of a particle escaping from the filtering grid and passing into the out flow passages is greater than a desired target level. The target level depends upon the design parameters for the filter. Therefore, the size of this circle depends upon the characteristics chosen for the filter and the media. Preferred arrangements of auxiliary channels are shown in FIGS. 40 and 41.

[0960] In FIG. 40 the closest outflow auxiliaries are arranged in a square pattern. In FIG. 41 they are arranged in a triangular pattern. They are located on the perimeter arid internal of the subregion illustrated. Further, all the inlet auxiliary channels are confined to an area outside the performance areas delineated by the dashed circles 404 and 414.

[0961] Our studies show it is quite beneficial for the number of inlet auxiliary channels to outnumber the outlet channels by a factor of more than 1.2 to 1. It is preferred that the ratio of the number of inlet to outlet auxiliary channels be greater than 2. More preferred is a ratio greater than 3.

[0962] a. Further Filtration Examples

[0963] The performance of a conventional pleated filter media like illustrated in FIG. 36 has been modeled. The depth thickness was sufficient to achieve a capture efficiencies greater than 98.2 percent for a particle capture probability of 0.3, greater than 99.9 percent for a pore capture probability of 0.5, and greater than 99.99 percent for a probability of 0.7 for the target particles. In the simulations using a fixed pressure across the filter, the number of particles captured was followed up until the flow was reduced to 50 percent of the initial value. The ratio of the pleated media thickness before folding, to the thickness of the Z-fold element in the depth direction as indicated by arrow 366 was approximately 0.14. The results for these conditions and geometry are labeled Z-fold in the tables of results in FIGS. 44 and 45.

[0964] Contrasted with the Z-fold results are results for elements using solid blocks of improved filter media employing micro-auxiliary channels. The solid blocks all had the same depth dimension as the Z-fold element.

[0965] The improved filter media with an auxiliary inlet and outlet micro-channels in the pattern of FIG. 38 was tested. Its results are labeled Micro-auxiliary in the following tables.

[0966] Improved filter media with auxiliary inlet and outlet micro-channels positioned as illustrated in FIGS. 42 and 43 were also tested. The results for these geometries are labeled Micro-auxiliary2 and Micro-auxiliary3 respectively. The filtering pores of the media were identical for all four geometries.

[0967] A representative sample of the Z-fold filter element occupies a volume which has a rectangular cross section indicated by dashed rectangle 365 in FIG. 36. In one case, the Z-fold filter is compared with blocks of the improved Micro-channel media where the blocks fill the same total volume as the Z-fold element sample. In a second case, the performance of improved media is compared to Z-fold elements on an equal mass basis.

[0968] In FIG. 44, Table 1 compares the total number of particles captured by the filter elements by the time the filtration process end point is reached. Z-fold media is compared to the improved media Micro-auxiliary for three pore capture probability levels. The results are compared for both the equal total mass case and for the equal total volume case. In every situation the improved media containing micro-auxiliary channels is vastly superior. It is superior on an equal volume basis, and on an equal mass basis. Additionally, the performance varies by more than 100 percent as the particle capture probability varies. The micro-channel media of this invention shows improved operational flexibility indicated by only a less than 25 percent variation in performance.

[0969] In FIG. 45, Table 2 compares the ratio of particle captures of the improved media to the Z-Fold element. The improved auxiliary channel containing media is compared on both an equal volume and equal mass basis. In all cases a significant improvement is found when using the media of this invention. Because the improvement is observed on both a mass and volume comparison, the inventive media will generally allow replacement of conventional media elements to improve the contaminant capture performance, or the pressure profile characteristics during filtration, or both simultaneously.

[0970] b. Methods of Producing Auxiliary Channels

[0971] In still another aspect of the invention, a method for manufacturing the filter element and media is taught. The method generally comprises a step of charging a porous filter media precursor composition into a mold. Commonly, the precursor is a pourable granular material. The grain size and distribution is chosen as to create on the average the pore sizes and distribution desired after bonding the material into a rigid element. The mold is configured to provide an element having a desired size and shape.

[0972] Fugitive auxiliary micro-channel masters are placed within the mold. These are constructed of materials that may be removed leaving behind passageways within the bed of granular material. For example, a self-supporting wire frame master structure may be placed within the mold. The wires of the frame may be of wax like material. Granular precursor is poured into the mold filling. It surrounds the wire frame. The precursor is then bonded together without disturbing the wire frame structure. Afterwards the wire frame may be dissolved, liquefied or converted to gas and removed from the bonded granular material. Upon removal of the wire frame master material, auxiliary channels will be left behind within the pore structure of the media. A solid structure is formed with interconnecting pores being the void spaces between the granular material and with an internal grid of auxiliary channels.

[0973] The wire frame may be a single piece or multiple pieces. It may he formed by various techniques including hut not limited to molding, injection molding, net spinning and 3-dimensional printing. The material of the frame may be polymeric, organic, inorganic, a meltable salt, ice, a meltable material, absorbable material, absorbable material, or vaporizable material. The removal of the auxiliary micro-channel masters is accomplished by a means appropriate for its composition and the media's composition. The auxiliary channels will be interconnected with the pores adjacent to their positions leaving functional auxiliary channels within the porous media.

[0974] Removal of the wires leaves behind void spaces that provide the auxiliary flow passageways within the media. After forming the auxiliary channels or during their formation, the precursor can be sintered or bonded together with known techniques.

[0975] J. New Coating Methods Using Fluid Distribution Metering Sheets

[0976] Coating dies perform two functions. First they distribute and meter fluid across the width of substrate to be coated. And secondly, they apply fluid onto the substrate after the distribution and metering step.

[0977] The use of fluid distribution metering sheets is beneficial in coating devices because it allows precise fluid distribution without resorting to precision steel die slots. Secondly, the enclosed edgewise fluid transport through the sheet allows the functional parts of a coating device to be physically separated and replaced with less complex, easier to maintain and often disposable subcomponents. The utility of metering sheets is further illustrated in the following discussion.

[0978] . Physical Separation of the Metering and Application Coating Device Functions

[0979] a. Improved Blade Coating

[0980] Die coating methods in general use the slot and cavities of the die to premeter the needed film of liquid required for producing a coating upon a substrate. The die technology creates the film of liquid at the slot exit. The geometry of the die lip and its position next to the substrate is responsible for the uniform transfer of a continuous liquid film onto the substrate translating past the die. With most die coating methods the die operating position is required to be very close to the substrate. This close positioning is needed for uniform coating transfer without air entrainment. Generally, the slot exit is required to be positioned at a distance away from the substrate that is no larger than 1 to 5 times the wet coating caliper on the substrate. This presents operational complexity, difficult positioning requirements, and it creates an elevated probability of scrap generation.

[0981] In the process of blade coating, excess liquid is applied to the substrate and the excess is removed by a flexible blade or a stiff blade bearing against the prewet surface. Here the amount of coating left on the substrate is controlled by the forces exerted on the blade and the response of the fluid under the blade.

[0982] Coating blades vary widely in shape, sizes, and material. Some are easily bent plastic strips. Others are stiff metal ones. Steel blades range from 0.1 to 1 millimeters in thickness and 10 to 200 millimeters in width equal to the substrate width. Loading of the blade against the liquid wet substrate may be accomplished in several ways. It may be pivoted about a blade clamp to a working angle that differs from the installation angle. It may be deformed by an air bladder bearing against a portion of its length. Various mechanical devices can be used to transmit force normal to the blade surface. The force may be generated by hydraulic, pneumatic, electrical or mechanical means.

[0983] Known blade coating methods are difficult to employ in that they do not allow a premetered fluid to be ducted directly to the blade tip, and applied to the substrate without using excess fluid. FIG. 69a illustrates a novel blade coating device using a fluid distribution sheet device of this invention that overcomes prior art deficiencies. This device allows a coating liquid to be applied onto the substrate when the die body containing the distribution cavity is physically set back from the substrate surface. The metering sheet meters and transfers liquid from the die cavity to the blade tip proximity, and the blade facilitates the transfer onto the substrate.

[0984] Shown as a cross sectional illustration, the die consists of a top plate 1111 and a bottom plate 1112. A flexible blade 1118 extends through the die and is sandwiched between the die plates. Additionally, a distribution sheet extends from the internal die cavity 1114. A die slot indicated by the arrows 1116 is present between the two die plates 1111 and 1112. The blade and the distribution sheet pass through and extend from the die slot.

[0985] Web 1128 translates past the die in the direction indicated by arrow 1130. The substrate may be a free span of web, or it may be supported by a means (not shown) such as a roll.

[0986] Liquid exits from the exit edge 1124 of the distribution sheet 1122. The sheet transports the liquid from the cavity 1114 to the web 1128. The blade forces or promotes transfer of liquid from the sheet end onto the web. A liquid coating 1126 is applied to the web by the blade. The combination of the metering sheet and the blade allow the die slot exit 1118 and the adjacent die lips 1132 and 1134 to be positioned at a distance from the web surface that is more than 20 times greater than the wet coating 1126 caliper.

[0987] The distribution sheet 1122 may extend into the cavity 1114 as shown, or it may extend only a limited distance into the die slot 1116. Extension into the cavity is preferred. Doing so tends to enhance the cross web uniformity of flow from the die.

[0988] When the coating blade is flexible or the web is unsupported (free span), or the web is supported by a deformable support means, the mechanics of the web, web support, and blade deformation allow uniform coatings to be applied. The blade—distribution sheet combination facilitates coating with the die lips positioned at large distances from the substrate.

[0989] The setback die lip position of this coating method allows web splices to pass by the die without retracting the die. The blade force on the liquid and the web allows premetered coating to be applied without air entrainment.

[0990] The blade and the distribution sheet of this invention are shown as two separate items. This may sometimes be convenient. The blade and distribution sheet may also be fabricated as a single compound element. In addition they may be laminated together by various means to form a compound element. The two may be combined so they flex or bend together or separately. The position of the blade end 1120, and the sheet fluid exit edge 1124, with respect to the web and each other is determined by coating trials. The positions are dependent upon the coating speed, coating caliper, the web surface and fluid flow properties.

[0991] FIG. 69b illustrates and economical construction of the blade type applicator of this invention. It illustrates a simplified coating die assemble 1200 using a fluid distribution metering sheet 1210 and a tubular element 1202. This element is pierced by a slot aperture 1204 which terminates at a die face 1206. Coating fluid is supplied by a metering and pressurizing means not shown. Fluid is transferred to a substrate 1208 which translates past it. The slot 1204 contains the fluid distribution metering sheet 1210. This type of sheet is described in the preceding sections.

[0992] The tubular element 1202 is constructed as a cylindrical body which is open at both ends and has a cylindrical bore 1212. The slot 1204 is formed, molded or machined in the tube. There is thus as defined by the bore 1212, a confined space connected to a discharge slot 1204. Also provided are end caps (not shown) which operate to close off the slot 1204 at opposite ends of the tubular element as well as the ends of the cylindrical bore 1212. The die 1200 is mounted in clamp 1236 which includes a plurality of clamping bolts 1218.

[0993] The distribution sheet 1210 is cantilevered from the tube 1202. It ducts flow from the cavity 1212 by way of its inlet edge 1216 to its discharge edge 1214. The die is mounted so as to bring the distribution sheet against the web 1208 translating by it. Fluid 1213 issues from the sheet discharge edge 1214 and is applied to the web. The web speed, the flow rate of fluid, and fluid properties determine the structure and stiffness of the sheet needed for coating. An extruded tubular die element forming the bulk of a die body combined with a fluid distributing metering sheet make up a low cost coating device requiring little or no precision machining to fabricate. The die may be constructed from polymeric materials and may be disposed of by incineration.

[0994] The distribution sheet may be a compound sheet composed of a sheet layer which ducts flow from an inlet to an outlet edge along with a structure support layer to modify the composite mechanical properties.

[0995] FIG. 70a illustrates another novel blade coating device variation of this invention. This device allows the coating die to be used to deliver liquid onto the substrate when the die slot exit is retracted from the substrate surface. A metering sheet transfers liquid 1318 from the die to the blade tip 1320 proximity, and the blade 1319 forces it onto the substrate.

[0996] Shown as a cross sectional illustration, the die consists of a top plate 1311 and a bottom plate 1312. A flexible blade 1319 extends through the die. It is sandwiched between the die plates 1311 and 1313. Additionally a metering fluid distribution sheet 1318 extends from the internal die cavity 1314. A die slot indicated by the arrows 1316 is present between the two die plated 1311 and 1312. The blade and the metering sheet extend from the die slot exit area.

[0997] Web 1328 translates past the die in the direction indicated by arrow 1330. The substrate may be a free span of web, or it may be supported by a means (not shown) such as a roll.

[0998] Liquid exits from the exit edge 1324 of the metering sheet. The sheet transports the liquid from the cavity 1314 to the web 1328 and the blade tip region. The blade threes or promotes transfer of liquid from the sheet end onto the web. A liquid coating 1326 is applied to the web by the blade. The combination of the metering sheet and the blade allows the die slot exit and the adjacent die lips 1332 and 1334 to be positioned at a distance from the web surface that is more than 20 times greater than the wet coating 1326 caliper.

[0999] Again, the setback die lip positions of this coating method allow web splices to pass by the die without retracting the die. The blade transfers coating liquid from the discharge edge of the metering sheet to the web.

[1000] FIG. 70b illustrates still another novel blade coating device variation of this invention. It exemplifies the principle that the fluid metering sheet allows separation of the functions of a coating die into separate component devices. The device 1400 contains the fluid distribution cavity 1402, and the device 1406 performs the function of transfer of the fluid coating 1410 to the substrate 1412.

[1001] Coating fluid is supplied by a means not shown to a die body device 1400 and into cavity 1402. Here the fluid is distributed across the width of the substrate 1412 and flows into the inlet edge of the distribution sheet 1404. Fluid proceeds internally inside the sheet and exits at the outlet edge 1414 near the end of the application blade 1408. The blade is held and positioned by the separate means 1406. The blade tip transfers liquid onto the substrate producing a coating 1410.

[1002] The advantages of separating the functions of the coating applicator into separate devices are that individual components are less complex, that the distribution cavity may be position remote from the substrate, that individual components may be smaller and lighter, that single components may be replaced without disturbing the other components, and that there is more flexibility in the mounting, design and adjustment of components.

[1003] Improved Slot Coating

[1004] One widely use coating method is slot coating” as described in chapter 11a of the Liquid Film Coating book. In this technique a die with a distribution chamber (cavity), a metering slot, and die lips is brought into very close proximity of a substrate translating past it. The lips transfer the coating onto the substrate.

[1005] Such lips are also known as knives. They are commonly spaced at a precise distance away from the surface of the translating surface of the substrate. It is also common to have the substrate assume a hydrodynamically created distance away from the knife surface in this case, it is common for the substrate speed, the fluid rheology, and the pressure distributions acting upon the fluid between the substrate and knife to determine the required distance between the knife and substrate.

[1006] Improvements are illustrated in FIG. 70c. Coating fluid is supplied by a means not shown to a die body device 1420 and into cavity 1422. Here the fluid is distributed across the width of the substrate 1430 and flows into the inlet edge of the distribution sheet 1424. Flow proceeds internally inside the sheet and exits from the outlet edge 1432 near the end of the knife blade 1426. This blade is positioned close to the substrate 1430 to effect a coating 1428.

[1007] As shown only one knife blade 1426 is in use on the out running side of the application point. It is also a teaching to confine the outlet end of the distribution sheet between an in-running and an out-running knife blade.

[1008] Further improvements of slot dies are disclosed and discussed in the improved casting die specifications.

[1009] d. Improved Slide Coating

[1010] Another useful coating device is the slide coater as described in chapter 11b of the Liquid Film Coating book. In this technique a die with a distribution chamber (cavity), a metering slot presents liquid onto an inclined slide. The fluid flows by gravity down the side and transfers to the substrate at a lip at the end of the slide.

[1011] An improved slide coater is illustrated in FIG. 70d. Coating fluid is supplied by a means not shown to a die body device 1440 and into cavity 1442. Here the fluid is distributed across the width of the substrate 1450 and flows into the inlet edge of the distribution sheet 1444. Flow proceeds internally inside the sheet and exits at the outlet edge 1452 onto the inclined slide surface of slide device 1446. This slide is positioned close to the substrate 1450 to accomplish a coating 1448.

[1012] e. Improved Curtain Coating

[1013] Still another useful coating technique is curtain coating. It is widely used for high speed coating of water based materials. It is also described in chapter 11c of the Liquid Film Coating book. In this coating method, a die containing a cavity and a metering slot is used to produce a flowing liquid film on an exterior surface of the die body. From this surface or a lip the liquid then free falls under the influence of gravity and impacts the substrate and coats the substrate translating past the die. The free fall distance is generally large on the order of multiple inches.

[1014] An improved curtain coater is schematically illustrated in FIG. 70e. Coating fluid is supplied by a means not shown to a die body device 1460 and into cavity 1462. Here the fluid is distributed across the width of the substrate 1474 and flows into the inlet edge of the distribution sheet 1464. Flow proceeds internally inside the sheet and exits at the outlet edge 1466 onto the surface of the curtain support device 1468.

[1015] The fluid flows by gravity to the distal end of the curtain support 1468 then free falls forming a fluid curtain 1470. The curtain impacts upon the translating substrate 1474, displacing the air and forming a layer coating 1472.

[1016] The curtain extends across the substrate from one edge to the other. In order to maintain the curtain edges in fixed positions mechanical edge guides bridge from curtain support device 1468 to the substrate. These are not shown.

[1017] Many useful variations of the metering sheet improved curtain coater are possible. Multiple die body devices 1460 with a metering Sheet 1464 may be positioned on each side of the curtain support 1468 to produce multilayer coating.

[1018] Also multiple die body devices may be easily positioned adjacent to the curtain support 1468 because of the positional flexibility afforded by the use of metering sheets. The coating operator may use one until it becomes clogged or until a formulation change is desired. At that time the flow may be stopped from the current die body and instantaneously started from a second die body without interrupting production.

[1019] Additionally, it is a teaching that the metering sheet alone may be used as the curtain support.

[1020] 2. Cast Coating Methods and Apparatus Using Organic Die Lips

[1021] Die coating methods in general use the slot and cavities of the die to pre-meter the needed film of liquid required for producing a coating upon a substrate. The die technology creates the film of liquid at the slot exit. The geometry of the die lip and its position next to the substrate is responsible for the uniform transfer of a continuous liquid film onto the substrate translating past the die. With most die coating methods the die operating position is required to be very close to the substrate. This close positioning is needed for the coating uniformity and transfer without air entrainment. Generally, the slot exit is required to be positioned at a distance away from the substrate that is no larger than 1 to 5 times the wet coating caliper on the substrate. This presents operational complexity, difficult positioning requirements, and it creates an elevated probability of scrap generation.

[1022] a. Cast Coating Improvement Needs

[1023] The production of embossed sheeting and the casting of free pressure sensitive adhesive (PSA) films prior to lamination to a backing have common problems. It is the damage to critical surfaces of critical elements (examples: roll, drum, belt, mold, web or plate) when using a slot die liquid applicator. Generalizing, the improvements in coating of a liquid onto casting surface exemplified by rolls, drums, belts, molds, webs, or plates using a close proximity slot die are a need of industry.

[1024] Lippert in U.S. Pat. 5,067,432 describes an improved slot die useful for casting a coating onto a web or mold surface. The improvement comprises a means of removably attaching the lips to the coating die. While this allows easy replacement of die lips damaged by clashing with rolls, drums, belts, webs or plates, it does not prevent the damage to those casting rolls, drums, belts, webs or plates. Improvements are in the casting coating process needed.

[1025] b. Disposable Lip Casting Die

[1026] FIGS. 71, 72 and 73 illustrate three existing methods of using slot orifice coating dies. A coating station for producing embossed sheeting is shown in FIG. 71. This station illustrates the process for casting an embossed web such as a substrate with a micro-replicated surface texture. A roll 1801 is provided with a patterned surface 1812. The pattern surface 1812 contains the negative image of the embossed pattern desired on the resulting web 1814.

[1027] A coating die comprised of a top metal plate 1802 and a bottom metal plate 1804 contains an internal manifold 1806 and slot 1808. These direct fluid to the end of the die slot between lips 1810 and 1811. The slot terminates in a slot orifice 1813 where the coating fluid emerges. The trailing lip 1810 forces casting fluid into the pattern on the roll surface.

[1028] In combination with the fluid feed system, the applicator must function to coat the entire patterned surface with a preset rate of fluid supply. Generally, the gap between the trailing lip 1810 and the pattern surface 1812 will have to be carefully and precisely adjusted to achieve a smooth and continuous coating. Even slight variations in the gap clearance will produce non-uniformities in the cast sheeting. Precise machining of the metal die and the patterned roll 1801 is required to achieve uniform sheeting.

[1029] The applied coating is transformed into solid sheeting 1814 as the fluid progresses around the roll 1801. Various means may be used to accomplish this including chemical reactions and phase changes. Reactions may be facilitated using electromagnetic radiation, heat or other methods. The sheeting is stripped from the roll 1801 and leaves with a negative image of the pattern on the roll.

[1030] If the lip 1810 touches the roll surface the embossing pattern on it is damaged. The metal lips 1810 and 1811 also will be damaged.

[1031] An existing free span coating of a web or a belt is illustrated in FIG. 72. Here coating is forced onto the surface of the substrate as it wraps around and translates past the metal slot die 1900. The die is comprised of a top metal plate 1922 and a bottom metal plate 1924. It contains an internal manifold or cavity 1926 and a slot 1928. These direct fluid to the end of the die slot between lips 1920 and 1921. Fluid exits from the slot through the slot orifice opening 1923. The trailing lip 1920 forces coating fluid onto the substrate surface. If the metal lip 1920 touches the surface the lip is damaged. If the substrate is a belt with an embossing pattern on its surface the embossing pattern is damaged by contact with the die lip 1920. If the substrate is a smooth belt, contact with the die lip 1920 may damage it.

[1032] FIG. 73 illustrates an existing transfer coating station. Here coating is forced onto the surface of the transfer roll 2060 as it translates past the slot die 2067. The die is comprised of a top metal plate 2068 and a bottom metal plate 2066. The die contains an internal manifold 2069 and slot 2070. These direct coating fluid to the end of the die slot between lips 2062 and 2064. The trailing metal lip 2064 forces coating fluid onto the roll surface. Fluid is forced through the die slot orifice into the gap between the lips 2062 and 2064 and the roll 2060 surface. Fluid is coated onto the roll and then is transferred to the web 2074 which is nipped to the roll 2060 by forcing roll 2072. If the lip 2064 touches the surface of roll 2060, the surface they will be damaged.

[1033] In combination with the fluid feed system, the die 2067 must function to coat all of the fluid delivered at preset rate supply. Generally, the gap between the trailing lip 2064 and the roll 2060 surface will have to be carefully and precisely adjusted to achieve a smooth and continuous coating. Even slight variations in the gap clearance will produce non-uniformities in the coating. Precise machining of the metal die 2067 and its lips and the roll 2060 is required to achieve uniform coating upon the roll.

[1034] If the fluid applied by die 2067 is a hot melt and the roll 2060 is internally cooled, the solidified fluid transfers to the web 2074 at the nip between rolls 2060 and 2072. If the fluid is liquid at this nip only some portion will transfer to the web 2074. In another variation of this coating method the web may be wrapped around roll 2060. In this mode, the fluid will be applied directly to the web in the gap between the roll 2060 and the die lips 2062 and 2064.

[1035] With the transfer coating technique in FIG. 73 and the casting process of FIG. 71 a gap must be maintained between the die lip and the roll. This gap must be larger than the combined positional machining, adjustment and environmental tolerances of the coating station to prevent clashing. Additionally, the variability of the gap must be less than the allowable variability in the mass per unit area of the product being produced. A ten percent variation in the gap will produce product variation of about 10 percent. If the absolute accumulative variation of the gap is plus or minus 0.02 millimeters and the layer being cast is 0.2 millimeters in thickness, the produce will have a variation of about plus or minus 10 percent. However, this same gap variation when trying to produce coating of 0.02 millimeters results in product variation of plus or minus 100 percent, and there is the almost certain probability of clashing of the die lips and the roll. This makes thin coatings or thin cast sheets difficult to manufacture,

[1036] Replacing the expensive, precisely ground metal lips on coating dies with inexpensive, non-metallic disposable lips allows precision coating at low cost. Additionally, thin coatings may be successfully achieved that are beyond the capabilities of metal lipped dies. Unlike metal lipped dies, organic and polymeric die lips may be operated with the die positioned so that the lip would clash with the coating roll or the substrate.

[1037] The coating station in FIG. 74 illustrates the method and apparatus of this invention for coating a hot melted fluid with the disposable lip contact die of this invention. Die 2180 coats fluid across the width of transfer roll 2191. Fluid hot melt is pumped into the die cavity 2186 and exits from slot 2184. The melt is forced onto the surface of roll 2191 by the lip 2196. Roll 2191 is cooled internally by the flow of chilled water. This cooling removes heat from the fluid as it moves with the surface of roll 2191. The rotation brings cooled adhesive to the nip between rolls 2191 and 2192, and web 2190 is transported through the nip on roil 2192. At the nip, solidified fluid is laminated to the web.

[1038] The surface of roll 2191 is covered with a thin layer of a release material. Upon cooling of the hot melt as it travels around roll 2191, the adhesive may be totally transferred to the surface of web 2190. When the web is an open highly porous nonwoven material, this allows placement of solidified fluid on its surface and reduces or eliminates strike through to the opposite side.

[1039] Die 2180 consists of a bottom plate 2106 and top plate 2182. Disposable trailing lip insert 2196 and a disposable lead-in lip insert are mounted at the discharge end of the bottom plate 2106 and top plate 2182 respectively. These are held in place by small plates 2197 and 2108, and a plurality of bolts 2111 and 2110. It will be appreciated that horizontal surfaces of the lip inserts 2188 and 2196 cooperate to form an extension of the slot 2184 which connects to a slot orifice 2120 between the disposable lip inserts 2188 and 2196.

[1040] Mounted on the plate 2106 are a plurality of lip adjusting blocks 2198, which are attached to the discharge end of plate 2106. The lip adjusting blocks 2198 are spaced apart longitudinally along plate 2106 and each threadedly receives therein an adjusting screw 2112, which is also engaged with an adjusting nut 2102 functionally attached to plate 2106. By adjustment of the adjusting screws 2112, the force applied through the adjusting blocks 2198 can be varied, serving to vary the deflection of the distal end of plate 2106 and its attached lip insert 2196. The deflection is about the narrow portion 2104 of plate 2106. This deflection may be either toward or away from the discharge end of plate 2182, thereby adjusting the average thickness of the slot 2184 and slot orifice opening 2120 dimensions.

[1041] Die 2180 is mounted by conventional means and brought into coating position by pneumatic cylinders. When in coating position, the disposable lip 2196 will be in contact with the roll 2191 if no adhesive is flowing and the roll is not moving. Because of this, the coating method is referred to as contact die coating. The exact die position is adjusted by wedge blocks. These may be moved to adjust the degree of engagement of the lip with the roll. Many other useable gap setting mechanisms are known to those skilled in the art for positioning the die to the roll.

[1042] The disposable die lip 2196 is extruded or molded from polymeric and organic materials. These processes do not have the exacting precision of machining and grinding steel. While steel may be ground to a tolerance of about plus or minus 0.0050 millimeters, polymeric and organic materials may be molded or extruded only with a tolerance of about plus or minus 0.050 millimeters a factor of ten less precise. Surprisingly, it has been found that molded or extruded polymeric and organic materials, as exemplified by polymers, may be used to replace high precision metal die lips. The substitution allows quality coatings.

[1043] Polymers and organic materials are generally more resilient than metal. Incidental impact of the lips and their edges that would dent, nick or permanently deform metals are resisted by the polymeric and organic materials. Additionally when polymeric and organic lips clash with substrates and rolls, they are much less prone to damage the substrate or lip. Running an organic lip into a stationary metal belt or roll will generally not damage it or the lip. Both will be damaged if the lip is metal. If embossed or patterned sheeting such as a cube corner reflective sheet is being manufactured, clashing a metal lip to the patterned roll or belt will destroy this most expensive piece of tooling.

[1044] Additionally, the polymeric and organic lips may be mass produced inexpensively in great volume. This may be accomplished by injection molding, continuous profile extrusion, or other forming processes. When damaged the polymeric and organic lips are inexpensive to replace.

[1045] Casting of embossed webs is shown in FIG. 71. During preparation of the casting station for the manufacture of micro-replicated, surfaced sheeting is usual to precisely align the casting die lips parallel to the casting roll axis of rotation. Also it is desirable that the clearance between the die lips and the roll surface is monitored. The limiting position of clashing contact with the roll should be known. This is the position where steel lip would begin to damage the roll.

[1046] A surprising finding is that resilient polymeric and organic lips may be used differently than metal lips. It is found that when polymeric and organic resilient lips are used in place of steel lips, thinner caliper coatings of casting resin may be applied upon the roll to produce thinner sheeting. This is also true when coatings are cast upon a roll then transferred to a substrate as illustrated in FIG. 72. Thin coatings may be achieved with the die positioned closer to the roll than limiting position of clashing contact. Additionally, changes in coating flow rate made to adjust the coating mass deposited often require no adjustment in die position. With rigidly mounted all steel dies any change in coating fluid flow rate requires a change in die to substrate gap.

[1047] While the detailed reasons for this phenomenon have not been studied, it is believed that the lips allow a deflection response which prevents lip and substrate damage and enables thinner coatings. That is the hydrodynamic pressure at the lip forces the lip to move away from a damaging position. Stiff steel lips do not allow this to happen.

[1048] As a further explanation, consider the adhesive coating process in FIG. 73. FIG. 75 shows an enlarge view of the lips of the coater. Here resilient lip insert 2258 terminates with a profiled lip geometry 2254. A segment of the roll 2270 is shown. The roll rotates in a direction indicated by arrow 2272. With a resilient lip, the die may be positioned in a roll clashing contact position. Unlike a steel lip, when the lip 2254 contacts the rotating roll, it is believed that it resiliently deforms by stretching downward in response to the shear and tractive forces exerted by the rotating roll. It is further believed that the deformation lessens the normal force between the lip and the roll surface. Also the flexible, elastic, and resilient character of the lip material reduces the probability of “biting” into the roll surface.

[1049] The ability of polymeric and organic materials to undergo deformation in response to applied forces is in general much greater than tool steel. This is another factor. Young's modulus for steel is on the order of 200 giga-pascals while for resilient polymeric and organics the modulus may be as low as 0.01 to 0.1 giga-pascals.

[1050] During coating, adhesive is forced between the trailing lip 2254 and the roll surface 2291 by the shear forces created by the roll rotation. When coating fluid is flowing from the slot orifice 2274 the pressures and forces generated by the flow also produces deformation of the lip. It is believed these forces open a passageway between the lip and the roll surface allowing coating to pass even when the die is positioned beyond the clashing contact limiting position. When the lip position is not yet at the clashing contact limiting position, the metered flow from the slot orifice 2274 creates fluid force that deflects the lip to allow the flow to pass between the lip and the roll. A steel lip will not do this. It is not self-adjusting.

[1051] It is preferred that the trailing lip design and the positioning of this lip with respect to the roll centerline, a tangent to the roll at the horizontal plain of the slot orifice 2274, and the die slot angle with respect to horizontal, and the gap between the lip 2254 and the roll 2270 surface all be chosen so that the flow of the fluid and the rotation of the roll 2270 both tend to resiliently deform the lip to open the gap between the lip and the roll when the roll 2270 rotates.

[1052] It is preferred the that flow exiting from the slot orifice 2274 be confined by the lip inserts 2260 and 2258, and most preferred that the flow be confined between roll 2270 and the trailing profile 2254 or both the lip termination profiles 2254 and 2261.

[1053] With stiff metal die lips, a precise gap must be set between the lip and the roll surface. The &formation of the lip in response to forces created by the flowing fluid and moving roll is negligibly small with respect to the gap required to allow the desired coating caliper. With metal lips the coating caliper produced is proportional to the gap. Thin coatings require gaps so small that the limitations on machine tolerances introduce unacceptable coating variations. An additional problem is that the probability of clashing the die lip with the roll becomes high. Clashing metal lips to the roll damages both.

[1054] Numerous organic lip widths and profiles are possible for the insets 2258 and 2260.

[1055] The choice of a lip is made experimentally as follows. With the web speed and adhesive flow rate set, the lip is brought in towards the roll. At first discontinuous bands of adhesive will be coated onto the roll commonly orientated at an angle to the web edge. These bands are often called “tiger stripes”. As the lip is moved closer to the roll, the bands will become wider and wider. At some point, a continuous coating will be achieved.

[1056] When continuous coating is established, further forcing the lip against the roll will still result in uniform coatings until an upper limit is reached. At this point, the adhesive may either accumulate behind the lip or be forced beyond the two ends of the die. Just before either of these occurs a measurable increase in coated width on the roll occurs.

[1057] The operability range is quantifiable if the die position relative to the roll surface is measured. The difference in position between that for the initial achievement of a continuous coat, and the position where the coating width increases more than a few percent is the operational window. The coating method will be less sensitive to upsets if the organic lip resilience, and die orientation are chosen to give the widest possible operational window. Well-chosen organic lips generally have much larger operational ranges than steel lips.

[1058] Coating with an organic lip has limitations. These relate to the modulus of the lip material, its geometry, and the coating weight and the product of the speed times the viscosity. For a given lip, viscosity, and coating weight, there will be a minimum and maximum coating speed. Generally, the operation between these limits is easy to achieve. One sets the flow rate and the web speed, and then one adjusts the gap between the die and the roll to achieve stable coating. Commonly, lip profile and modulus are chosen based on experimental trials. The sensitivity of the quality of the coating to die position and roll gap is much less than that which is achieved with steel lips.

[1059] One consideration is that as the product of the speed times the viscosity for the coating process increases, a point is reached where stable flow cannot be established. Coatings of high viscosity at high speeds can be difficult.

[1060] Referring to FIG. 74, the cooled roll 2191 is covered with a thin coating of a silicone rubber to provide for the release of the PSA after cooling. The diameter of the roll must be large enough to allow the PSA sufficient dwell time for solidification. For high speed coating the diameter required may be on the order of one meter.

[1061] Casting dies offer the ability to coat both stripes and patches of adhesive on a web. For down web stripe coating the die may be fitted with internal deckles. These block flow at the desired cross web positions. In the uncoated web positions the organic lip should be cut back. The lip should only engage the roll where adhesive is applied.

[1062] Patch coating requires both deciding and modulation of the adhesive flow rate.

[1063] Referring again to FIG. 75, the die slot 2263 is formed between the horizontal flat surfaces of metal plates 2250 and 2264. Their distal ends terminate with seat areas 2252 and 2256 to which the organic lips 2258 and 2260 are attached by a plurality of bolts 2266 and 2267. A die slot extension 2262 is formed between the horizontal flat surfaces of the organic lips 2258 and 2260. It is preferred that this extension have a gap height larger than the slot 2263 height. Surprisingly it is found that although the tolerances achievable for organic lip inserts are larger than steel lip, adequately uniform flow may be achieved From the slot orifice across the width of the substrate being coated or cast. The substitution of resilient polymeric organic lips of lower precision in place of high precision steel lips need not destroy the precision of the coating. It may be that the organic lips resiliently deform to compensate or partly compensate for non-uniformities.

[1064] In one mode of operation, a die position is chosen so that the lip 2254 will contact the roll surface 2291 when no fluid is flowing. This is a clashing position. In this position when the fluid is then forced from the die orifice 2254, and when the substrate is translating, good coatings are observed. When the flow and the substrate are stopped, clashing is again observed. Although it is difficult to observe, the flow and motion appear to cause elastic lip deformation opening a flow passage between the lip and the roll surface.

[1065] We believe that when a metered flow of adhesive is distributed behind the lip, the lip deforms in response to pressure and shear allowing coating. The lip floats on the flow of coating liquid between it and the roll. With a proper adjustment of the forces holding the lip against the roll, the lip deforms to allow the metered flow to be applied as a uniform and continuous coating to the roll surface 2291. The lip riding position off the roll surface is self-adjusting. The lip deformation self-compensates for the flow rate, and the coating uniformity is insensitive to mechanical precision. Operation in this mode is often desired to achieve very thin coatings, but there is the possibility of damaging the lip 2254 if the flow is interrupted. This is another reason for using inexpensive, disposable organic lips.

[1066] Useful polymeric and organic lips have resilience as measured by ASTM™ Shore A hardness ranging between 10 and 90 durometer, or Shore O or Shore OO durometers less than 100. Preferred materials for coating at elevated temperatures include heat resistant materials, fluoropolymers, silicone polymers and fluorosilicone compounds.

[1067] A general feature of an improved coating or casting die includes the use of replaceable die lips. It also includes the extension of the flow distributing slot surfaces with a replaceable disposable organic or polymeric inserts where these extend the slot to direct the flow to the die lips.

[1068] FIG. 76 illustrates a polymer or organic lip 2300 insert which contains a means of adjusting the resilience of the lip termination end 2304. The internal structure may be modified by a region 2302 of a second material with different material properties than the base material of the insert 2300. By this means, the effective resilience of the lip termination end 2304 can be substantially different than the base lip material. The region may be a cylindrical bore containing a pneumatic fluid or a hydraulic fluid. The internal pressure of the fluid may be adjusted and controlled to manipulate the coating performance of the lip.

[1069] FIG. 77 illustrates a polymer or organic lip insert 2405 which contains a means of modifying the resilience of the lip termination end 2409 and a base subsection 2403. The external structure may be modified by a region 2407 of a second material with different material properties than the base material of the region 2403. Again by this means the effective resilience of the lip termination end 2409 can be substantially different than the base lip material. It is a teaching of this invention that the lip insert or the lip termination portion of the die lip is able to deform or move in response to fluid pressures or forces acting upon it during coating or casting. Multi-component lips are within our teachings.

[1070] FIG. 78a illustrates a polymer or organic lip 2505 insert which contains a means of modifying the resilience of the lip termination end 2509 and a base subsection 2503. The external structure may be modified by a region 2507 of a second material with different material properties than the base material of the region 2503. The region 2507 of differing material may be a ribbon of material attached to the lip insert 2505 by an adhesive or other means. The effective resilience of the lip termination end 2509 can be different than the base lip material and have differing surface properties. It is a teaching of this invention that the lip insert or the lip termination portion of the die lip is able to deform or substantially move in response to fluid pressures or forces acting upon it during coating or casting. Compound lips are within our teachings.

[1071] FIG. 78b illustrates a polymer or organic lip 2506 insert nearly identical to that FIG. 78a. However, here modifying material 2508 is only partially attached to the base 2502. When this lip inset is used in a die, and when the lip is spaced from the web in a coating position, the modifying material 2508 will be flexed and deformed to assume a shape more similar to that of material 2507 in FIG. 78a. The material 2508 may be a polymeric sheet or film.

[1072] Those skilled in the art of polymer and organic part design will recognize that many different materials may be combined to modify the local stiffness or resilient response of a part. The lip in of FIG. 77 may even be fabricated with external structure 2407 consisting of a material that a high Young's modulus such as a metal. In this ease the flexibility, resiliency, and deformability of the lip termination 2409 are provided by the deformable base subsection 2403. Those ordinarily skilled in the art will recognize that many different geometries and combinations of materials will allow resilient response of the die lips or their termination ends. All are within the scope of this invention.

[1073] c. Simplified. Disposable Dies and Non-metallic Dies

[1074] A simplified preferred coating die using the metering sheet of this invention is illustrated in FIG. 79a. Considering in further detail the coating head assembly 1500, there is included in this assembly a tubular element 1502. This element is pierced by a slot aperture 1504. It terminates at a discharge face 1590 at the seats 1514 and 1515 to which the lip inserts 1506 and 1508 are attached. The gap between the lip inserts terminates at an elongated slot orifice 1564. Coating fluid supplied by the metering and pressurizing means exits this orifice. It is transferred to a substrate which translates past it. The slot 1504 and the gap between the lip inserts contain a fluid distribution and filtering, metering sheet 1510. This sheet is described in preceding sections.

[1075] The tubular element 1502 is constructed as a cylindrical body which is open at both ends and has a cylindrical bore 1512. The slot 1504 is formed or machined in the tube 1502 along with lip insert seats 1514 and 1515. There is thus, as defined by the bore 1512, a cavity connected to a discharge slot 1504. A plurality of bolts 1520 pass through the wall of tube 1502 and are thread into the opposite wall section locations 1516. By adjusting these screws, the gap of the elongated slot 1504 may be varied.

[1076] Also provided are end caps (not shown) which are used close off the slot 1504 and bore 1512 at opposite ends of the tubular element. The slot 1504 in the tubular element 1502 is extended by the opposing horizontal surfaces of the lip inserts 1506 and 1508. These have termination ends 1530 and 1532 which may be of any desired shape. Casting die terminations are shown. Terminations, usable for curtain coating, slide coating, blade coating, slot coating and any method where fluid is supplied through a slot orifice are teachings of this invention. The lip inserts 1506 and 1508 are attached to the tubular body 1502 by a plurality of bolts 1536 and 1534 passing through attachment elements 1538 and 1540.

[1077] Fluid is supplied to the bore 1512 of tubular element 1502 by one or more entrance ports (not shown) in its wall. Generally a metered supply rate is used that is equal to that required to achieve a desired coating upon the substrate being coated. A rate in excess of this rate may also be used in some cases. The flow is distributed along the length of the bore 1512 and enters a metering sheet 1510. The metering sheet ducts the flow from the inlet end 1560 to the discharge end 1562. Flow exits the sheet 1510 and is discharged from the slot orifice opening 1564 located between the lip terminations 1530 and 1532. The metering sheet may be bent as shown to accommodate a long length. In the case where clamping bolts 1520 are used, the length may be longer than half the length of any chord of the circular bore 1512. If the bolts are replaced by an external clamping means, the length may be longer than any chord of the bore.

[1078] Bending and coiling the sheet within the cavity allows the sheet dimension in the direction of flow to be much longer than the slot 1504. This allows a substantially increased flow path for fluid leaving the cavity. The longer flow path compared to a slot of prior art dies produces a much improved flow distribution at the slot orifice 1564.

[1079] The use of the sheet allows useful filtration to be accomplished along with fluid distribution.

[1080] The bolts 1520 are used to adjust the slot 1504 width, and thus may be used to bold the metering sheet 1510 in place. Alternatively, the lip inserts may be used to hold the metering sheet in place.

[1081] The great advantage of these dies is that major components including the lip inserts, the tubular element, and the metering sheet may individually or all be made from polymeric or organic materials. The individual elements may also be constructed from metals or ceramics. All these possible combinations allow optimization for each particular situation. It is within the scope of this invention to use combinations of two or more materials for the construction of the die assembly. These materials may include but are not limited to organic, inorganic, polymeric, metallic, natural, man-made, and ceramic materials.

[1082] The tubular element 1502 and lip inserts 1506 and 1508 may be extruded, cast, molded, or formed from inexpensive polymeric or organic materials. The slot 1504 in the tube 1502 may be formed during the step of making the tube. The slot may be cut or machined in the tube after the tube is formed.

[1083] The attachment means for mounting the lip inserts 1506 and 1508 to the tube 1502 need not be by bolts. Other known means of attachment, including adhesives, may be used. Alternatively, the inserts need not be separate elements from the tube 1502. They may be formed as integral parts of the tube during the tube forming process.

[1084] The metering sheet facilitates the uniform distribution of fluid to the die slot orifice but may not always be necessary. It is a teaching of this invention to coat with the die 1500 with and without the metering sheet 1510. Coating without the sheet but with the disposable and resilient lip elements is a teaching,

[1085] FIG. 79b illustrates a metallic die where lip inserts 1600 and 1602 are organic. The tubular element 1606 is metallic. The metering sheet 1604 is also metallic. A slot is placed to connect the bore of the tubular element 1606 with the exterior. Its walls 1610 and 1612 extend from the interior to the exterior of the tube 1606. In this construction the metering slot is fabricated with a slot height that is less than the thickness of the metering sheet 1604. To assemble the die, forces are exerted to open the slot gap prior to inserting the metering sheet. After insertion the forces are released and the wails 1610 and 1612 of the slot clamp onto the metering sheet 1608. By this means the metering sheet is clamped within the slot without using clamping bolts.

[1086] Those skilled in the art of coating will recognize that the coating die illustrated may be used at room temperature or at elevated or depressed temperatures. The temperature of the dies may be controlled by external or internal heating or cooling elements.

[1087] FIG. 79a illustrates a coating die construction that, with the possible exception of the clamping bolts 1520, may he constructed from polymeric, organic or other inexpensive materials. When the elements are manufactured by mass production methods such as extrusion, molding, casting and etc., the cost may be drastically reduce compared to precision machined individual metal parts. This enables the coating device to be economically disposable. As a whole, the die may be discarded after a period of use to avoid the expense of cleaning or the expense of refurbishing damaged pieces. Also individual elements of the coating die may be discarded after a period of use to avoid the expense of cleaning or the expense of refurbishing damaged surfaces.

[1088] The coating die or individual pieces of the die are “economically disposable” if their cost of is insignificant, or lower than using and maintaining the alternative conventional coating die or die elements.

[1089] FIG. 79c illustrates a coating die 1820 that is constructed with lip inserts 1824 and 1822, a tubular element 1834 and a metering sheet 1832. All or a portion of these may be made from polymer or organic or other economically disposable materials. It is preferred that the fluid contacting items be made from low cost materials. It is preferred that these items be fabricated using efficient low cost methods. Preferred are lip inserts and metering sheets of polymeric or organic materials. Most preferred are lip inserts, metering sheets and tubular elements of polymeric or organic materials. Preferred methods for producing these items are fabrication methods such as casting, molding, injection molding, extrusion, and micro-replication.

[1090] The die 1820 is mounted in clamp 1836 which includes a plurality of clamping bolts 1818. The lip inserts are attached to the tube element 1834 by means of a mechanical interlocking design. The interlocking is accomplished by the foot protrusions 1826 and 1828 which are inserted into matching cavities in the wall of the tube 1834. The matching the protrusions, and the cavities into which they fit, both run the length of the inserts and tube. They are easily produced during a forming step (extrusion, molding, etc.) for the inserts and tube. This construction reduces the number of parts making up the coating die. An extruded tubular die element forming the bulk of a die body, combined with extruded organic or polymeric lip inserts, and further combined with a fluid distributing metering sheet, may be assembled to construct a low cost coating device. Additionally, assembly by use of mechanical interlocking techniques further lowers the cost.

[1091] FIG. 80 illustrates an improved tubular casting die consisting of a tube element 2714 which contains a slot 2708 connecting the internal bore 2715 to the lip region. Lips 2702 and 2704 terminate the flow passage from the tube. The termination is at the slot orifice 2703. A metering sheet 2712 is contained within the slot 2708, the tube bore 2715 and between the lips 2704 and 2702.

[1092] The tube 2714 may be made from polymeric, organic non-rigid or nonmetallic materials. It is held precisely and rigidly in position by metal mounting plates 2716 and 2706. These are held together by a plurality of bolts 2719. With this construction all the wetted parts of the die may be made from economically disposable, inexpensively fabricated parts. All the wetted parts may be formed by extrusion or casting processes. All the wetted parts may be formed without precision metal working techniques. Additionally, the die lips and the tube may be formed in one step such as simultaneous co-extrusion. Still further, they may be made from the same material.

[1093] FIG. 81 illustrates an improved tubular casting die 3500 consisting of a tube element 3502 which contains a slot 3504 connecting the internal bore 3512 to the lip region. Lips 3562 and 3566 terminate the flow passage from the tube. The termination is at the slot orifice 3564. A metering sheet 3514 is contained within the slot, the tube bore, and between the lips. In this case only one replaceable lip insert 3530 is present, and the metering sheet extends to the vicinity of the lip 3562.

[1094] In summary, it is well known that the cost of fabricating a distribution die is proportional to its size and mass. Lower size and mass means less material is needed for fabrication and generally less time and labor in fabricating critical surfaces. The inventive use of metering sheets allows smaller overall die dimensions. The sheet allows long length of precision metering slots to be replaced by the coiled or curved sheets. By this means the size of the distribution dies may be dramatically reduced. The use of metering sheets avoids precision machining of die slots.

[1095] When die parts may be destroyed by incineration, waste disposal is less expensive. The solid mass is reduced in incineration, and energy is recovered in the process. The polymeric die elements of this teaching are ideal for this process. The use of disposable polymeric lips and metering sheets to replace machined metal components improves manufacturing economics.

[1096] K. Mist Collection

[1097] 1. Fundamentals: General Observations on Silicone Fluid Misting

[1098] The functional silicone coatings on release liners and pressure sensitive adhesive backings are very thin. Solventless silicone roll coating processes represent a special thin regime of roll coating. Detailed observations of the fluid dynamics at the roll nip have resulted in significant findings.

[1099] The separation point in the nip between two counter-rotating rolls is the point at which the respective surfaces make the transition from contacting to non-contacting.

[1100] When applied by a 5-roll coater, the liquid wet caliper on the high speed transfer roll (the roll that transfers coating to the substrate) is generally thinner than 10 microns. Often, the thickness is much thinner. In the coating process, a portion of the liquid on the transfer roll is transferred to the substrate as it is carried through the transfer nip on a backing roll. On the in-running side of the nip, liquid on the transfer roll is brought into contact with the substrate. On the out-running side of the nip, the liquid splits between the surfaces. Some liquid stays on the transfer roll and a portion leaves the nip on the substrate.

[1101] For this ultra-thin coating range, the liquid film split location occurs very near the separation point of the rolls. The film split is chaotic, 3-dimensional, and random in nature. Liquid surface and air interface perturbations produce filaments, septums, waves, and other disturbances. These are formed as the two wet roll surfaces separate. These generate mist as surface tension acts to forth droplets from the unstable perturbations. It is an observation of this teaching that for silicone coating the mist is actually generated in the region normally less than five millimeters from the roll separation point. Of course without confinement, mist rapidly spreads throughout the entire nip region and into the coater room.

[1102] The process at the out-running side of the roll nip is shown in magnified detail in FIG. 82.

[1103] At the out-running side of the roll nip, there are three regions. The roll contact zone is indicated by arrow 812, and it is filled by fluid 813 (indicated by the solid grey shading). The mist generation region is indicated by arrow 811. The mist 814 is indicated by the fine textured gray shading. The mist dilution zone is indicated by arrow 810, and the diluted mist 815 is indicated by the coarse darker grey shading. While these regions are shown with sharp transitions in the FIG. 82, in reality, they blend into each other. As coating passes through the nip, it reaches the separation point 818. Here the surfaces of the rolls diverge, and the liquid flow diverges with major portions attached to both roll surfaces. Perturbed liquid interfaces are created, and mist is formed.

[1104] If the wet caliper of the silicone being coated is “x”, then the silicone film thickness in the nip roll contact zone 112 is approximately 2x when roll speeds and surface textures are equal. The length of the mist generation zone is quite small relative to the roll diameter. Compared to the coating caliper, x, is on the order of 100x to 10,000x. The dimension is variable and depends upon coating rheology, speeds, surface textures, coating caliper, etc. ‘The exact dimension is so small it is difficult to measure. In general, the mist generation region will extend less than a fraction of a centimeter from the separation point. Arrows 816 and 817 indicate the direction of motion of the separating roll surfaces 821 and 820. The solid grey thin films of liquid 822 and 823 are carried on the surfaces 820 and 821 and moving with them.

[1105] Note that FIG. 82 is not drawn to scale.

[1106] The “nip air volume” is the air volume between the roll surfaces 820 and 821 It is defined as the air volume on the outrunning side of the nip between the roll cylindrical surfaces, the two planes containing the roll ends, and a plane simultaneously tangent to both roll cylindrical surfaces at a distance of approximately one roll radius outward from the separation point when the rolls are of equal diameter. The volume of the zone where mist is actually generated will be a very small percentage of the nip air volume.

[1107] In the small mist generating zone, stresses generated by the diverging rolls form the mist. Once the liquid on the rolls passes out of this generation zone, little or no additional mist is formed. It is not generated from the roll surfaces outside the critical generation zone. The shear created by rapid movement through air is generally not sufficient to rip droplets from the surface liquid films 822 and 823. The viscosity and the very thin liquid caliper prevent this. Only the divergent liquid flow near the contact point 118 creates mist.

[1108] Droplets formed will initially have a velocity outward. The magnitude at first will be on the order of the roll surface speed. This tends to concentrate mist near the midline between the rolls (assuming equal roll speed). However, the mist particles quickly decelerate by transferring momentum to the surrounding air.

[1109] The dilution zone 815 features air flow in and out of the “nip air volume”. The air mixes with and dilutes the mist. The droplets are entrained by the air currents and carried from the dilution zone to the coating room. The standard method to control mists from process equipment is to enclose the area around the roll coater and to ventilate it. Multiple air replacement volumes for the large enclosures are required for success. Capital investment, operating and maintenance costs are high.

[1110] 2. An Improved Strategy for Controlling Mists

[1111] Consider a control volume which contains the generation zone 814 that is only moderately larger than the generation zone but small compared to the total air nip volume. All the mist formation is contained within this volume. If we continuously remove the mist from this volume and replace it with clean air, the mist is contained and removed. If the volume is properly flushed, mist escaping into the coater room can be drastically reduced. It may be reduced to near zero.

[1112] The method and apparatus disclosed here accomplishes the extraction of the mist from the nip. The contaminated air flushed from the control volume is not expelled into the coater room; it is captured and discarded by a vacuum system. Key to successful., economical mist collection is the design of an apparatus that confines the mist to a very small volume, and flushes it from that volume.

[1113] 3. Mist Extraction Apparatus

[1114] A new mist removal die of this invention is described below. One key to its successful and economic operation is that it confines the mist in a very small controlled volume deep within the nip of the rolls. A second feature is a practical die design allowing deep penetration into the roll nip. A third is that the die design uses a replaceable fluid conveying device such as a fluid distribution sheet, or an equivalent thin profile fluid duct. A thin profile for extending far into the nip is preferred.

[1115] The operation and utility of the mist removal device may be understood by referring to FIG. 83. Shown here is the mist extraction die 830 which is a fluid flow die. It is constructed and shaped so that it penetrates deeply into the nip region indicated by arrow 832. FIG. 83 is drawn scaled proportionally for roll diameters of approximately 16.5 inches with dimensions indicated by the one inch scale marker 834 in the figure. The large roll surfaces are indicated by the arcs 836 and 838.

[1116] The figure illustrates one variation, and the die's ability to penetrate to a position just outside the mist generation zone. The die tip position and the volume of air sucked into the die tip prevent mist escape. The mist confinement zone is indicated by the arrow 840. In this embodiment the close clearance between the die body and the surfaces of the rolls confine the mist. The close approach of the extending die tip element 842 to the roll surfaces also confines the mist. In other preferred apparatus not illustrated, the thin protruding die tip element 842 may extend substantially more than illustrated in FIG. 83. This allows the majority of the die body to have greater clearance with the surfaces of the rolls while the tip element still confines the mist.

[1117] The die is shown schematically in FIG. 84, and partial details of the bottom plate 840 and the protruding die tip element sheet 880 are shown in FIGS. 85 and 86 respectively.

[1118] Referring to FIG. 84, the die is composed of three pieces. They are the bottom plate 840, the top plate 842, and the sheet 880. A cavity 860 is machined into the bottom plate. Also, there is a recess 850 in the bottom plate down its entire length. This may be more clearly seen in the top and side view schematics of FIG. 85. The protruding die tip sheet 880 fits into the recess 850 and is held in place by the clamping of the top plate. The sheet connects and directs flow between the die exterior and the cavity 860. Attachment mechanisms for holding the three die pieces together are not shown. Additionally, the cavity is connected by process lines and entrance ports to a vacuum system. These also are not shown.

[1119] The protruding die tip sheet is a useful design element of the die. It is constructed from rigid materials. Inside this sheet multiple flow passages provide paths to conduct mist laden air from the confinement zone into the cavity 860 of the collection die. The sheet will both transport mist laden air and filter mist from the air. The sheet passageways may be designed to have a probability of capturing mist. Since the sheet will tend to become laden with mist material, low cost disposable sheets are preferred.

[1120] Mist capture fluid distribution sheets allow many design and operational advantages. The sheets may be single pieces or multiple pieces placed end to end across the die width. Dies can be designed with quick opening clamp mechanisms for rapid replacement of strips, Strips may be inserted from one die end and simultaneously removed from the other end,

[1121] In operation, the passages of the sheet provide precise and uniformly dimensioned flow passages. With a proper die design, this produces uniform air flow into the cavity along the length of the roll nip. The sheet design may be accomplished using the same flow modeling techniques described earlier. Additional inertial effects must be taken into account for the high speed flow of air in the die cavity. Inertia is discussed in the article “Inertia and gravitational effects in extrusion dies for non-Newtonian fluids”, W. K. Leonard, Polym. Eng. Sci., 25, 9 (1985), pp. 570-576.

[1122] The sheets may be constricted from plastics, metals, ceramics, and other solid materials. They may be sheets containing micro-replicated flow channels, and they may be assembled from multiple single sheets. Plastic materials are preferred to reduce the potential clashing damage to the rolls and to minimize costs. These active flow elements may be laminated to or sandwiched between plates that provide support or rigidity.

[1123] The mechanical design and fabrication of the die and associated mountings are not a demanding task compared to manufacture of precision extrusion or coating dies. While coating dies are commonly machined to tolerances of plus or minus 0.0005 or 0.001 inches, tolerances here may be an order of magnitude larger. Those with ordinary skills in the art of precision die design will have no difficulty in fabricating the extractor die, its positioning mechanisms, and its mounting fixtures.

[1124] The mounting and positioning of the die requires only standard technology used by coating equipment vendors. The exact details depend upon the design and operation of the multi-roll coater. The movement of the die into the nip during startup, shut down, and emergency stops are preferred to be automatically controlled. The accurate positioning of the die between the rolls may be controlled manually or electronically. A control system may include position sensors, feedback algorithms, adjustable mechanical stops, mist density sensors, etc. Piezoelectric and magnetostrictive actuators are useful for the automatic control of positioning to a roll. These are described in U.S. Pat. No. 5,409,732. All of these features are considered within the scope of this invention.

[1125] The metering sheet may be fabricated as a single piece typically by a casting or extrusion process. The sheet may be rigid, or flexible and supported by rigid plates.

[1126] It is preferred that if a particle clogs the sheet at a point the flow diverts around the clog and continues through the surrounding channels. The preferred sheet allows flow in both perpendicular and parallel to the length direction of the die.

[1127] The compactness of the mist extractor and its operating principles achieve mist collection without the need to process huge volumes of contaminated air. This approach is attractive because it does not depend on the coating formulation, and it lacks the negative features of a HVAC enclosure technology.

[1128] L. Expanded Embodiments

[1129] The preceding description, drawings, examples and claims represent embodiments of the present invention. However, it will be understood that various additions, substitutions, combinations and modifications may be made without departing from the spirit and scope of the present invention. It will be clear to those with ordinary skills in the art that the present invention may be embodied in other specific forms, arrangements, structures, proportions, combined with other elements and components and be constructed of other materials without departing from the spirit and scope of the present invention. The disclosed embodiments are to be considered as illustrative and not restrictive of the scope of the invention. All of the patents and patent applications cited above are incorporated by reference into this document in total.

[1130] Slot Distribution Device:

[1131] For fluids or flowable materials in a slot, flow distribution improvements result by having first direction resistances parallel to a dispensing slit orifice which differ from those in a second direction toward it. Consider the flow toward one edge of the slot where fluid is dispensed from a slit orifice. This flow geometry may be divided into multiple parallel lanes terminating at the dispensing slit. These may have equal or different resistances. Also the resistance may vary along the lanes. The same is true for the flow parallel to the slit. The important parameters will be the average resistances calculated by considering the total slot. The flow resistance of the slot as a whole is the important variable not any lane or specific location.

[1132] Consider a rectangular slot having width and depth length sides, if flow is forced from one depth side to the other, and pressure is uniform along the input side and uniform but lower along the output side, then the total slot has an average flow resistance in the depth direction. The total slot resistance equals the pressure difference divided by the total flow. The local total resistance in any lane is the sum of the resistances in series in that lane, and the total resistance of parallel lanes equals the sum of the reciprocals for the individual lanes. The average slot resistance for flow in the depth direction then is the total resistance times the ratio of rectangle's width to the depth. The average conduction is the reciprocal of the average of the total slot resistance.

[1133] If flow is forced from one width side to the other, and pressure is uniform along the input side and uniform but lower along the output side, then the total slot has an average flow resistance in the direction parallel to the depth side. Total slot resistance in this width direction equals the pressure difference applied in this width direction divided by the total flow in this direction. The average total slot resistance in this width direction then is the total times the ratio of rectangle depth to the width. When a slot gap varies, it is additionally useful to characterize it with a dimensionless slot viscous number. For rectangular slot geometries the Dimensionless Rectangular Slot Viscous Number is preferred. It is the ratio of the viscous flow resistance over the width to the viscous flow resistance over the depth. The number is calculated as the ratio of the width to depth times the square root of the ratio of the average total flow resistance in the width direction to the average total resistance in the depth direction.

[1134] FIG. 98 illustrates the dependence of slot outflow uniformity on the dimensionless viscous number for five representative non-uniform slot geometries. The series 1 data is for slots with uniformly spaced protruding ridges running in the width direction. Here the ratio of the average resistance in the width direction to that of the depth direction is always less than one, and such slots have more uniform outflows than equivalent smooth walled slots. Only variation of the width and depth lengths will cause the Dimensionless Rectangular Slot Viscous Number to be greater than one. The graph illustrates the specific case of feeding the slot from a point at a corner opposite the outflow edge of the rectangle geometry.

[1135] The series 2 data is for slots with periodic protruding ridges running in the width direction, and where the ridges are periodically interrupted along their lengths in the width direction. FIG. 91 illustrates an example where the bottom wall of a slot is enhanced with this type of protrusions 9110. This slot geometry has a resistance ratio less than one and improves distribution over smooth uniform slots.

[1136] Series 3 shows the performance of a single groove in the slot wall running across the width direction. Here the groove is placed close to the slot outlet. Series 4 shows the variation of uniformity index with the Dimensionless Slot Viscous Number for a single widthwise groove far removed from the slot flow exit. Using a widthwise groove placed substantially anywhere in a smooth slot, creates a favorable flow resistance ratio and improves outflow uniformity.

[1137] Series 5 plots the performance of the modification illustrated in FIG. 89 where periodically spaced grooves 8920 are present. These run parallel to the discharge edge of the slot. They create a favorable flow resistance ratio and improve outflow uniformity.

[1138] While the graph of FIG. 98 represents only one possible slot flow entrance port location, our studies find the beneficial use of a slot with flow resistance that differ in two directions is not unique to the geometry. Virtually all distribution slots may be enhanced using this teaching.

[1139] Healing the flow distribution defects caused by filtration of target particles is governed and improved by having differing average resistances in two directions. More specifically the healing is dominated by the Dimensionless Blockage Viscous Number. FIG. 97 illustrates efficacy of the criteria. When a particle logged in and slot blocks a “first length” of the slot parallel to the discharge slot orifice and is located away from the slot orifice at a “second distance”, the local flow defect is reduced by selecting a slot design with an appropriate Dimensionless Blockage Viscous Number.

[1140] The Dimensionless Blockage Viscous Number is defined by the product of two ratios: the “first length” to the “second distance,” and the square root for the fluid flow resistance parallel the orifice divided by the fluid flow resistance towards the orifice. Large benefits are possible by using appropriate ratios of differing resistances. If the flow resistances vary in direction and position throughout the slot, the average total slot resistances defined above should be used to characterize the slot's behavior.

[1141] The slot of this invention may also be used to beneficially filter target contaminant particles from fluid passing through it. This is often needed to remove contaminants before the fluid is coated or distributed to prevent quality problems. Additionally, using the inventive slot design allows filtration without disturbing the desired flow distribution from the slot. Particles may be trapped at locations clogging a portion of the slot disrupting the flow. However, an inventive design criteria employing slot flow resistances negates this problem. It is found so that this disruption may be substantially diminished, often eliminated, if the slot flow resistance are different in two directions.

[1142] As shown in FIG. 97 when a flow distribution has a desired deviation at the slot orifice of less than ten percent (Uniformity Index=0.1) then a Dimensionless Blockage Viscous Number less than 0.5 is preferred. When a deviation less than one percent is desired, a Dimensionless Number of 0.15 is preferred.

[1143] An inventive method of modifying a slot to enhance removal of unwanted target particles is to place woven mesh materials in the slot. These may be placed in the entrance region away from the discharge orifice. Metallic screen meshes create many restrictive flow passages within the slot for trapping particles. Spacing and sizing of the warp and weft wires or threads enable many different fluid flow and particle entrapment characteristics. Importantly, the flow grid paths created by these screens are also extremely useful for creating average flow resistances that beneficially differ in two directions. The weave can create large variations in flow resistance in at least two directions. The placement of woven and mesh-like materials into smooth surface slots is a simple means to modify slot flow properties. It is a teaching to create grids of periodically spaced flow passages within a slot by placing screen like mesh materials, woven materials, fiber materials, and wire materials in the slots for modifying flow properties or filtration properties or both. Entrance regions may be modified in this manner while leaving the exit region unmodified.

[1144] If a slot is poorly designed, fluid may not flow from the entire length of the discharge edge Often the pressure at one end does not exceed the surface tension pressure differential acting to prevent flow from the slot. Reducing the quotient of the slot flow resistance parallel and resistance perpendicular to the orifice may be used to solve the problem. Adapting the first and second resistances in combination to maintain a flow distribution through a slit orifice port along its entire length is a teaching of this invention.

[1145] Multiorifice Distribution Device:

[1146] In order to provide a uniform flow distribution from a line of orifices of a multi-orifice coating die, the device cavity and bore sizes have to be carefully chosen. This is highly significant if the device is used for precise coating. In a worst case, pressure may be high near the fluid entrance into the cavity and near zero far away from the entrance. The resulting flow is then high from orifices near the entrance and at or near zero from orifices away from the entrance. An exemplary device of the present invention provides apparatus to correct this deficiency. Additionally, the invention discloses preferred means for using bores and auxiliary channels to filter, correct bore plugging flow defects, and provide tolerance to clogging while providing desired cross-web flow distributions.

[1147] An exemplary die of the present invention is illustrated generally at 710 in 99. Die 710 may be used in a coating process where fluid is translated out of orifices 725 and onto a web (not shown) so as to form a coating. The cross-web direction of die 710 is perpendicular to sectional view shown. Web can be formed of a multitude of various materials including polymers or paper. The inventive die 710 may be utilized with many types of coating and extruding processes, including fixed gap coating, slot orifice die coating, curtain coating and slide coating, among others. Additionally, the shape of die 710 may vary according process employed and nature of the flowable material.

[1148] The exploded cross sectional view of one embodiment of inventive die 710 illustrates important features. Die 710 includes mounting block portion 720 and faceplate portion 723. Cavity 728 is present internally within block portion 720. The illustrated die 710 is formed of ten pieces. Eight shims, 721a through 721d and 722a through 722d, are shown. Any number of pieces may be used to form die 710. For example, shim portions 722a can be divided into first and second pieces (not shown) alternatively shims 721a and 722a could be construction as a unitary shim. Note that for purposes of this specification a shim is defined as a piece of material generally ranging between 0.05 millimeters and several centimeters in thickness and commonly having two parallel faces. Shims space two mechanical pieces apart from each other.

[1149] Faceplate portion 723 is generally of a wear resistance material that differs from the other die component materials. External face 729 is disposed on faceplate portion 723. Multiple tubes 724 extend into faceplate 723 and terminate at a multiple orifices 725. The orifices 725 extend along a line for a length on the face 729 in the cross-web direct for a functional width herein designated as “W”.

[1150] Referring the both FIGS. 99 and 100, flowable material is introduced into cavity 728, typically by a pump (not shown), such as an extruder or a positive displacement pump as known in the art. Pressure in cavity 728 created by the pump forces fluid out of orifices 725. As fluid emerges from orifices 725 when employed in a coating process the fluid streams from each orifice are transferred to the web substrate (not shown) and may be merged or remain as individual streams as desired and as the result various interactions of the die face 729, web positioning, face geometry, fluid properties, forces applied and other factors.

[1151] In FIG. 99, fluid material flows generally from cavity 728 toward the final distribution tubes 724 in the die faceplate portion 723. The orifices extend along a line across the width of the faceplate for the approximate desired width W of the applied coating on a substrate. Generally, the cavity extends for this same width W. Interposed between the cavity 728 and faceplate 723 is an assemblage of bore containing shims (721a, 721b, 721c, 721d) and auxiliary channel shims (722a, 722b, 722c, 722d). These in combination form a grid of bores 727a through 727d, and auxiliary channels 726a through 726d. The bores direct fluid toward the face plate 723, and the auxiliary channels provide flow paths generally perpendicular the fluid action of the bores 727 and tubes 724.

[1152] Preferred are auxiliary channels which allow flow in two orthogonal directions rather than just one direction. Useful hydraulic radii for the bores 727 range from one tenth to several millimeters. Useful hydraulic radii for the auxiliary channels 726 range from a few hundreds of a centimeter to several centimeters.

[1153] When assembled, the device block 720 and faceplate 723 force together the multiple shims positioned in between them. The shims provide separation. The forcing means may be external clamps or bolts extending between the block 720 and faceplate 723. Neither of these means is shown.

[1154] The cavity 728 provides fluid a flow path to preferably all of the bores 727a of shim 721a. Its conductivity of fluid down its length and along the width, W, may be large so as to provide some uniformity of flow into all the bores 727a of plate 721a. In the absence of other overriding design restrictions this may be helpful. However, if the eight shims 721a, 721b, 721c, 721d, 722a, 722b, 722c, and 722d are not present a problem may arise. A contaminant particle equal to or larger than the tube 724 diameter which enters into the die cavity 728 may clog and prevent flow through one of the tubes 724 transporting fluid to an orifice 725. This will locally disrupt the discharge of fluid. The following discussion and examples describe the use of a grid of bores and auxiliary channels to prevent this flaw.

[1155] FIG. 100 illustrates design details of shims 721b and 722b with a top view and a side sectional view. Shim 721b is perforated by a population of bores 727b. These span the width, labelled W, which nominally equals the cross-web coating width and the line of orifices 725 on the die 710 faceplate 723. These bores 727b have fluid flow resistances and have probabilities of capturing troublesome target contaminant particles occurring in the fluid. A side sectional view illustrating the bores at plane labelled A-A is shown. A side sectional view illustrating the channel 726b on shim 722b at plane labelled B-B is shown. The bores 727b and channel 726b use on their respective shims which encompasses the span indicated by the arrows 731 tunes the width W.

[1156] As seen in FIG, 99, shim 721b abuts to shim 722b and during coating fluid proceeds through the bores 727b into the auxiliary channel 726b.

[1157] FIG. 100 illustrates top and cross-section profiles of a preferred geometry for the auxiliary channel 726b. It is a slot like void when shim 722b is sandwiched between shims 721c and 721b, The preferred slot channel 726b geometry has a gap, indicated by arrow 730, equal to the thickness of the shim 722b. The channel 726b allows fluid to take a three dimensional path with directional components both perpendicular and parallel to the width W and gap 730. An important channel 726b property is its fluid flow resistance parallel to the faceplate 723.

[1158] An important property of slot 726b can be its probability of capturing target contaminant particles. It is found that the channel 726b particle capture probability approaches one hundred percent as the shim 726b thickness and slot gap 730 is reduced to the diameter of a contaminate particle.

[1159] The capture of particles by the bores 727b of shim 721b approaches one hundred percent as the bore hydraulic diameter approaches the contaminate particle diameter. Importantly, for the bores such an occurrence tends to prevent further flow through it. However for capture by the slot channel, flow around the particle is still easily accommodated in directions both perpendicular and parallel to the die face 729.

[1160] Flow in the slot auxiliary channel geometry has been fluid dynamically studied. It has been found that the flow and particle trapping characteristics offer improved die operational characteristics compared to cylindrical like auxiliary channels. This a direct result of the slots allowing fluid flow simultaneously in two directions.

[1161] The flow distribution from the orifices 725 of faceplate 723 depends dominantly upon the precision of the faceplate tubes 724. If the combined action of the cavity 728, of bores 727a through 727d, and auxiliary channels 726a through 726d produces a substantially uniform pressure in channel 726d, the flow through tubes 724 will generally be uniform. As a consequence, no one element need be precisely machined or maintained. The blockage of various shim bores or omitting entirely multiple bores may be accommodated with a good flow circuit design for the shims.

[1162] For uniform flow distribution through the line of orifices 725, common practice would suggest using a large cavity 728 where the pressure in it does not vary from end to end. This then means that a large volume of relatively stagnant fluid is present in the cavity 728. This may be avoided with the inventive design of shims 721a, 721b, 721c, 721d, 722a, 722b, 722c, and 722d described in the following example.

[1163] Consider a die 710 assembled without the shims 721a, 721b, 721e, 721d, 722a, 722b, 722c, and 722d wherein the cavity 728 cross-sectional area tapers from its feed point to a distal end if the area is small at the feed position, and the area is very much smaller at the end, the resulting cavity pressure may vary from the feed point to the end by a factor of two, four, or even 32 times. Normally, this is a very undesirable situation requiring a very special design of the faceplate 723 to compensate such pressure variations.

[1164] However compensation for the cavity pressure variation is may be simply accomplish in various ways with the design of the shims 721a, 721b, 721c, 721d, 722a, 722b, 722c, and 722d combined. Or, the compensation may be accomplished using only the design of the bores 727a of shim 721.a. FIG, 101 illustrates bores 7.27a in a modified layout on shim 721a. This design is capable of compensating for an example cavity 728 pressure variation of 32 times. The width of the flow from the faceplate 723 orifices 725 is indicated by the arrow labeled W on FIG. 101. Note that the bores 727a vary in size along the length W from left to right. These variations are capable of compensating for high pressure at the left end and low pressure at the right. Flow through a cylindrical bore depends upon the diameter to the fourth power. Therefore, a pressure change of sixteen times may be accommodated by a doubling of the bore diameter.

[1165] Another example of the device is illustrated in cross section in FIG. 102. A mounting block 752 is attached by bolts 771 to faceplate 754. This faceplate contains a modest size cavity 764 extending for its length. The faceplate 754 contains a notch 756 on its surface which defines a lip 758. A line of orifices 760 are present on the faceplate surface. Connecting the cavity 764 to die orifices 760 are tubes 762. Notably, the modest size of cavity 764 is not adequate to provide uniform flow from the orifices if fluid is introduced directly into it.

[1166] Mounting block 752 contains a large cavity 770. The large cavity 770 is fluidically connected to the faceplate 754 and cavity 764 by a multiplicity of bores 769, slot auxiliary channel 768, and a second multiplicity of bores 766. In the die of FIG. 99 the inventive bores and channels were present in shims. In the die of FIG. 102 no shims are present as shown, and the inventive bores are placed within the wall 772 of block 752 and in the wail 774 of faceplate 754. The auxiliary channel 768 is placed within the wall of faceplate 754. All the cavities 770 and 764, the bores 769 and 766, the auxiliary channel 768, the tubes 762, and the orifices 760 fluidically connect in a grid of flow paths. A fluid port (not shown) provides flow into the cavity 770.

[1167] Notably, functional shims of the type illustrated in FIG. 99 may be added the die 750 of FIG. 102 between the mounting block 752 and faceplate 754.

[1168] In operation the die 750 may be use to distribute a flowable material containing contaminating particles which are unwanted in the discharge flow from the orifices 760 or that may clog tubes 762. The bores 769 may be constructed to target and capture the contaminating particles without destroying the desired flow distribution from the orifices.

[1169] The faceplate 754 maybe constructed from tubing or gun drilled centerless ground rod. The mounting block may be constructed from metal or non-metallic tubing. This allows to a low fabrication cost.

[1170] Drilled hole, multi-orifice devices are commonly made by boring a line series of holes into the face of a die with a drill bit. However, those with ordinary skills in manufacturing and machining will recognize that there are many other methods of producing the holes or bores. They may be created by placing a line of precision tubes within a mounting. Here hypodermic tubing is commonly used. Bores may be made by inscribing a series of periodic grooves in a plate and mating it to a smooth plate. They may be made by burning holes into a block with electric discharge machining methods. Bores may be made using lost wax casting processes. Also embossing, threading and knurling process may be employed, and it is important to note that bores may have any cross-sectional shape and need not be circular.

[1171] Porous Sheet Distribution Device:

[1172] Devices of this invention employing porous media may also be used to filter target contaminant particles from fluid passing through it. Using these new teachings allows filtration without undesirable disturbance to flow distributions. In the case of flow through porous material, a particle may be trapped at an entrance edge. This locally clogs the media disrupting the flow. However, an inventive design criterion is found so that this disruption may be substantially diminished.

[1173] FIG. 103 illustrates efficacy of the criteria. Consider the case of a fluid distribution device containing a sheet of porous media having width, depth and thickness dimensions. Fluid is flowed from an entrance edge in an edgewise manner through the sheet height and on to a distribution edge. A contaminant particle may lodge at entrance blocking a “first length” of the sheet parallel to distribution edge width. A flow defect results in the discharged flow distribution. This may be reduced by selecting media with an appropriate Dimensionless Width Blockage Porous Viscous Number. FIG. 103 plots the uniformity index of the discharge flow in these cases.

[1174] The Dimensionless Width Blockage Porous Viscous Number is defined by the ratio of the “first length” to the depth times the square root for the quantity the average fluid flow resistance toward the distribution edge divided by the average fluid flow resistance parallel to the orifice. Large benefits are possible by using ratios of differing resistances. FIG. 14)3 indicates a desired low uniformity index values may be maintained when clogging occurs if the geometry and flow resistances are chosen to achieve a low Dimensionless Width Blockage Porous Viscous Number.

[1175] When the desired deviation at the discharge edge is less than ten percent (Uniformity Index—0.1) then the media design should have a Dimensionless Width Porous Viscous Number less than 0.435. When a deviation less than two percent is desired a dimensionless number less than 0.2 is indicated. Porous media especially that made from arcicular granular material may have resistances to flow that vary with direction. Woven porous media and fibrous media may be designed to achieve directional flow resistance variations isotropic porous media may be modified to achieve flow resistance that differ in two directions as previously not n reference to FIG. 92.

[1176] The embodiments above are chosen, described and illustrated so that persons skilled in the art will be able to understand the invention and the manner and process of making and using it. The descriptions and the accompanying drawings should be interpreted in the illustrative and not the exhaustive or limited sense. The invention is not intended to he limited to the exact forms disclosed. While the application attempts to disclose all of the embodiments of the invention that are reasonably foreseeable, there may be unforeseeable insubstantial modifications that remain as equivalents. It should be understood by persons skilled in the art that there may be other embodiments than those disclosed which fall within the scope of the invention as defined by the claims. Where a claim, if any, is expressed as a means or step for performing a specified function it is intended that such claim be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof; including both structural equivalents and equivalent structures, material-based equivalents and equivalent materials, and act-based equivalents and equivalent acts.