US 3276944 A
Description (OCR text may contain errors)
Oct. 4, 1966 M LEVY 15,276,944
NON-WOVEN SHEET OF YNTHETIC ORGANIC PGLYMERI FILAMENTS AND METHOD OF PREPARING SAME Filed Aug. 50, 1963 INVENTOR MARTIN RICHARD LEVY ATTORNEY United States Patent 3,276,944 NON-WOVEN SHEET OF SYNTHETIC ORGANIC POLYMERIC FILAMENTS AND METHOD OF PREPARING SAME Martin Richard Levy, Wilmington, Del., assignor to E. I. du Pont de Nemours and Company, Wilmington, Del.,
a corporation of Delaware Filed Aug. 30, 1963, Ser. No. 305,875 15 Claims. (Cl. 161150) This application is a continuation-in-part of my application Serial No. 220,564 filed August 30, 1962, now abandoned.
This invention relates to a novel self-bonded sheet material and to a novel method for bonding the fibrous elements of a nonwoven web.
Many techniques have been used for bonding filaments together in the manufacture of woven, knitted, and nonwoven textiles. The bonding techniques used in the past included fusion of elements by heat-bonding or through application of solvating chemicals which dissolve or plasticize the polymeric filaments suificiently to fuse them together. Application of various adhesives or resins to the filaments to glue them together is another conventional bonding method. These prior techniques are either uneconomical and/or fail to achieve the realization of maximum strength and other physical and chemical properties inherent in synthetic organic polymeric filaments.
It is an object of this invention to provide a novel selfbonded sheet of oriented synthetic organic polymeric filaments of textile denier, said sheet having a combination of high tensile strength, a strip tensile greater than 3 lbs./in.//oz./yd. and high tear resistance, a tongue tear strength greater than 1.5 lb.//oz./yd. Another object is to provide a novel process for bonding oriented synthetic organic polymeric filaments in a nonwoven web without the need for solvating agents or foreign binders which remain in the final bonded product. A further object is to provide novel nonwoven sheets which have a combination of properties that lead to superior performance when employed in such applications as rugbackings, bagging materials, and in other industrial and apparel uses. Other objects will become apparent from the description of the invention given below.
The novel products of the present invention comprise nonwoven self-bonded sheets of oriented synthetic organic filaments and are characterized by an average edge bond count of between 30 and 85 bonds per hundred filaments and a uniformity ratio of below 1.3. Within the sheet, the filaments overlap and intersect and in general are disposed in random fashion. The filaments are bonded to each other at a multiplicity of these intersection points. These bonds occur throughout the three dimensions of the sheet. A unique feature of the novel sheets is the fact that the polymeric filaments therein possess a birefringence that is at least 40% of the calculated maximum birefringence of fibers of that polymer.
The process of the invention comprises exposing a nonwoven web of oriented synthetic organic filaments to a nonsolvating fluid atmosphere at a temperature within the range of from about 45 C. below up to the crystalline melting point of the filaments while restraining the web under a force suflicient to prevent filament shrinkage of more than about 20%. The selected exposure temperature at the selected restraint must he suilicient to self-bond the filaments at a plurality of spaced intersection points while confining the drop in filament birefiringence to less than 50% and preventing a drop in birefringence to a level below about 45% of the maximum birefringence.
The filaments in the web to be heated in accordance with the novel method of the present invention are dis- "ice posed in random fashion throughout the web and overlap and intersect throughout the thickness of the sheet. They are of textile denier, varying from about 1 to about 15. Higher deniers up to about 25 may be used for special applications. Textile denier is intended to include the entire grouping. The filaments may be crimped or straight and may be round in cross section or of d-iiferent s-hapessuch as trilobal, elliptical, etc.
Oriented synthetic organic polymeric fibers that are crystalline or crystallizable are employed in the instant process. Suitable polymers include isotactic polypropylene, linear polyethylene, polyethylene terephthalate, polyhcxamethylene adipamide, polycaproamide, copolyester of ethylene glycol with a 90% 10% mixture of terephthalic/isophthalic acids, and crystalline copolyrners of at least propylene or ethylene with up to 25% of other copolymerizable monomers or mixtures of 75% isotactic polypropylene or linear polyethylene with up to 25 of other polymers (e.'g., atactic polypropylene, branched polyethylene, .polyisobutylene).
The filaments that are incorporated in the web prior to bonding should be highly oriented, that is, they must have a birefringence of at least about 40% of the maximum birefringence in order to obtain bonded sheets having high tear and tensile. While the instant process limits the birefringence drop of the filaments-employed to less than 50%, it is desirable that initial fiber orientation (prior to bonding) be at such level which will permit the orientation of the filaments in the bonded sheet to be at least 40% of the calculated maximum birefringence. This latter value, i.e., maximum birefringence, is determined in accordance with H. M. Morgans Correlation of Molecular Orientation Measurements in Fibers by Optical Birefringence and Pulse Velocity Methods, Textile Research Journal, 866, October 1962. The maximum birefringence for fibers of common polymers is given below:
Isotactic polypropylene 0.04 Linear polyethylene 0.06 Polyethylene terephthalate 0.26 Polyhexamethylene adipamide 0.08 Polycaproamide 0.07
Preferably isotactic polypropylene filaments are used in present invention having a crystallinity index of at least about 40%, and a birefringence of between 0.016 and 0.040. Best results are obtained when the melt flow rate (M.F.R.) of the polypropylene filaments is within the range of 1.7 to 43. The M.F.R. of the polypropylene filaments is determined using ASTM method D-1238 at 230 C. with a loading of 2:16 kilograms.
While an important consideration is that the filaments in the web to be bonded must be substantially identical to each other both chemically and physically, the fibers may be homopolymers, copolymers or mixtures of polymers. Extraneous filaments or other materials which act as adhesives at or below the bonding temperatures employed for the self bonding process should be avoided where possible. Small amounts however, of under 5% may not significantly affect the tear and overall sheet properties of the product.
In addition to the filaments heretofore described, up to 25 by weight of the total web may comprise other filaments which will not self-bond under the bonding conditions employed. These other filaments may be termed inert filaments and may be made of glass, metal or other inorganic or organic filaments. There may also be present small amounts of adjnvants such as fillers, delusterants, pigments, coloring materials, stabilizers and the like. For example, stabilizers may be incorporated into the polymer before spinning into filaments in applications where self-bonded webs will be exposed to ultraviolet light.
The geometry of the filamentary web to be selected for self-bonding may vary to a certain extent, although for best results the filaments in the web should have a high degree of random positioning. It is preferred for most applications that the filaments should be separate and independent from each other except at crossover points. Particularly good results are obtained by using as the filamentary web a random continuous filament structure which is substantially free of filament aggregates. Another form of random filamentary web may be obtained by depositing staple filaments from a Rando Weber. Still other forms of suitable webs are ones which are formed by cross-lapping of unidirectional filamentary webs.
Randomness is -a measure of the isotropic nature of the arrangement of the individual filaments in the web. A suitable test for determining whether the web is random involves selecting a representative section of sheet approximately one half inch square for measurement and placing the same in epoxy resin. The epoxy resin is polymerized. The embedded section is then microtomed transversely to the face of the sheet along each of the four sides of the square sample. The microtomed sections are deposited on a microscope slide in immersion oil and extended without stretching so that no wrinkles are present. The total number of fibers extending through the face of one-half inch length of the microtomed section is counted for each of the four sides of the sample square. The measure of randomness, randomness ratio, is calculated as fol-.
A/B=Randomness ratio A=Total number of filaments on side having the largest count B=Total number of filaments on side having the smallest count In a highly random sheet, the number of filaments that will be counted along any side of the square will not exceed the number of filaments terminating at any other side of the square by more than 20% (randomness ratio below 1.2) regardless of the location or orientation of the square within the plane of the sheet. For the purpose of this invention, a sheet will be considered random if the number of filaments that will be counted along any side of the square will be less than twice the number of any other side of the square. This is equivalent to a randomness ratio of under 2.0.
It is preferred that the filaments be separate and independent from each other along their lengths thus indicating a relative freedom from aggregation or ropiness. By this is meant that between filament crossover points there will be a substantial absence of strands of multiple filaments.
The term bunching coeflicient can be used to de scribe the degree of filament aggregation. The bunching coefiicient designated BC is defined as the ratio of the number of fiber spaces occupied by fibers relative to the total number of fiber spaces available. In this measurement the term fiber space represents the average space occupied by a fiber and is calculated by dividing a unit distance of the nonwoven sheet structure by the total number of fibers oriented in a single direction in that unit length.
The term bunching coeflicient can be used to describe the degree of filament aggregation. The bunching coefiicient designated BC is defined as the ratio of the number of fiber spaces occupied by fibers relative to the total number of fiber spaces available. In this measurement the term fiber space represents the average space occupied by a fiber and is calculated by dividing a unit distance of the nonwoven sheet structure by the total number of fibers oriented in a single direction in that unit length.
The bunching coefiicient concept is based on the premise that where the individual fibers disposed in the same direction are uniformly spaced from each other, each fiber space will contain one fiber and the bunching coefficient of such a structure will be unity. A nonwoven sheet which contains bunched fibers, some of the fiber spaces will contain bundles of fibers while others will be unoccupied. The bunching coefiicient of such a structure will be less than one. The greater the filament aggregation the lower the bunching coeflicient.
The bunching coetficientis measured on thin layers of uniform thickness taken from a representative 6" x 6" unbonded web or unbonded sheet at random, parallel to the length and width direction and sampled through the thickness. The sample layers to be measured must have sufiicient thickness so that at least 50 fibers are counted per thin layer determination (to be described later) and thin enough so thatthe fibers through the thickness are distinguishable. Thus, thickness will generally be less than 5 mils.
For each determination on a thin layer, at least two areas are selected at random, each of which encompasses at least a /s" square. These areas are photographed at 28 magnification. Each photomicrograph is measured as follows: A straight line (called the first line) is cast from one corner of the photomicrograph along an edge and a second movable straight line (or indicator) placed perpendicular to the first and extending across the view. The second line is moved along the first line in this perpendicular position across the face of the view. All fibers which intersect the first line and run parallel (:2 around the point of intersection with the first line) with the second line for a distance of at least 5 mm. are counted as fibers. This number is then equal to the number of fiber spaces for the measurement. The first line is then divided into equal spaces, equal in number to the number of fiber spaces measured. The number of fiber spaces occupied" for the measurement is then equal to the number of the equal spaces which are intersected by one or more fibers meeting the parallelism requirements with the second line. The first line is then rotated 15 across the face of the view from the pivot point in the corner of the view, and a similar measurement made moving the perpendicular second line along the first as before. This, is continued at 15 intervals until the 45 position has been measured. Following the 45 measurement the second line is extended on the opposite side of the first and measurements continued every 15 to the position. The pivot point of the first line is then moved to anadjacent corner of the view and measurements'made from 0 to 90 in 15 intervals around this pivot point in a similar fashion. The procedure described is repeated on the remaining photomicrographs. This completes the determination on one thin layer. The number of fiber spaces and number of fiber spaces occupied are totaled for all the measurements. The bunching coefiicient for this layer is calculated by dividing the totaled number of fiber spaces occupied by the totaled number of fiber spaces.
A sufiicient number, at least two, of separate bunching coefiicient determinations are made on separate random layers of a sheet sample and the results averaged so that the confidence limit (see Statistical Methods for Chemists by W. J. Youden, Publ. John Wile) & Sons, 1951) of the average is no greater than :15% of the average.
The bonded sheet of this present invention will have a bunching coefiicient that is substantially equal to that of the unbonded web since the bonding process does not affect filament aggregation. Accordingly, an unbonded web is selected for bonding that has the bunching coefficient that is desired in the bonded sheet.. Since it is simpler to measure bunching coefiicient on bonded samples, the results below are reported for the bonded sheet only. A bunching coefiicient of at least about 0.5 is required for the products of the present invention. A
high bunching coetficient is desirable where the sheet is to be used for drapable textile purposes.
Webs of various thicknesses may be submitted to the bonding process of the present invention. Sheets having a basis weight from as low as /2 oz./yd. to as high as 20 oz./yd. or even higher may be obtained.
The conditions of self-bonding the filamentary webs in accordance with this invention are highly critical and must be carefully controlled in order to preserve the orientation of the filaments. The significant bonding conditions for any particular web include exposure temperature, heating fluid, heat-up rate' and restraining force applied during the exposure of the webs to the elevated temperatures.
The heating fluid must be a nonsolvent fluid for the filament being bonded. By this is meant, that the vapor or liquid used as the heating fluid must not have any solvent effect on the filaments at the bonding temperature used. The fluid must not reduce the orientation of the filaments to any substantial degree at the bonding temperature. Although hot inert gases such as air may be employed, it is preferred that a saturated vapor atmosphere be employed as the heat-bonding means to insure proper temperature control and to obtain a high degree of uniform disposition of bonds throughout the sheet.
Saturated steam is the preferred heating fluid. When employing saturated steam, the bonding temperature range for isotactic polypropylene should be between 140 C. and 180 C. This is equivalent to using a saturated steam pressure of 38 p.s.i.g. to 130 p.s.i.g. The steam will condense at any cold spots in the sheet and immediately bring them up to temperature, thus accounting for uniform bond disposition.
When using a saturated vapor atmosphere for bonding it is preferred to operate in a closed system. In such a system, the pressure drop across the sheet being treated is essentially nil, and the temperature of condensing steam on both sides of the web is essentially the same. If the steam is merely passed through the web in an open system, precautions must 'be taken to avoid a substantial pressure drop across the web, which would cause the temperature on one side of the web to be considerably higher than the temperature on the other side, producing overbonding at the high temperature side and underbonding at the low temperature side. In accordance with the instant process, the temperature of the heating fluid at any point throughout the web difiers from the temperature at any other point by less than 5 C., preferably less than 2 C. For certain applications where an extremely high degree of sheet uniformity is required as in isotactic polypropylene primary carpet backings, the variation in temperature may not exceed 0.5 C.
The critical bonding temperature lies in the range between 45 below up to but not including the crystalline melting point of the polymer filaments. The crystalline melting point can be determined by the method outlined in Preparative Methods of Polymer Chemistry by W. R. Sorensen and T. W. Campbell, Interscience Publishers, Inc., 1961, pp. 44-47. In applying the procedure to filaments it is necessary to immobilize the filament on the slide by securing its ends with a suitable heat-resistant adhesive or tape. In some instances temperatures just under the crystalline melting point may be too high for the process because of too rapid a heat-up rate.
The bonding temperature to be used for any particular system is at least that necessary to result in sheets having a strip tensile strength greater than 3 lbs./in.//oz./yd. or an edge bond count of at least 30 bonds per hundred filaments. The upper temperature limit is set by the requirement that the birefringence drop must be less than 50% and not fall below about 45% of the calculated maximum birefringence. In practice, one may determine the bonding temperature (i.e., the highest temperature to which the web is exposed) by running a bonding profile on a series of samples at two degree intervals upward from the lowest temperature permitted, i.e., 45 below the crystalline melting point. The remainder of the system (apparatus, heating fluid, heat-up rate, etc.) is kept constant during the profile. Samples are tested for strip tensile strength and birefringence. When a sample having a tensile of 3 lbs./in.//oz./yd. is obtained, bonding may be performed at or above the temperature used for bonding the said sample. The upper bonding temperature limit is reached when a sample is obtained having either a birefringence below 45% of the calculated maximum birefringence or 50% or less of the birefringence of the starting material.
As mentioned previously, the filamentary batt to be bonded must be held under a positive restraining force in the bonding system to prevent shrinkage. It may be held under compressional restraint in the direction perpendicular to the faces of the sheet, but it may also be confined in the width and length directions.
It has been discovered that a critical amount of restraint of the filamentary web is required in order to maintain the high level of orientation inherent in the individual filaments which level is related to the sheet tensile strength. Restraint also permits the control of the degree of bonding as measured by edge bond count (defined below), which is related to tear strength. It has been found that if the filaments in the web are allowed to shrink beyond a critical amount as defined below, the resulting bonded fibrous article loses a substantial amount of its achievable strengths and toughness.
The restraining force found to be effective is that sufficient to prevent individual filaments from shrinking more than about 20% in length and preferably it must prevent shrinkage of greater than 5%. More highly drawn polymeric filaments will generally require a greater restraining force than those filaments that have been drawn to a lesser extent since fibers with a high degree of orientation have the greatest tendency to shrink. Also, as sheet thickness increases, the restraining force must be increased. Since restraint prevents filament and hence web shrinkage, it must be maintained during heat-up to the bonding temperature, at the bonding temperature and during the cooling off period where shrinkage may take place. Generally, no further shrinkage occurs when the sheet temperature drops more than 20 C. below the bonding temperature.
The minimum amount of restraint employed in the process of this invention is about 0.25 lb./in. preferably applied to the web surface for each ounce per square yard of unbonded filamentary web. This minimum value is suitable when the sheet to be treated is sandwiched between canvas cloths. Smooth retaining surfaces require greater restraining forces as will be apparent to those skilled in the art. A preferred amount of restraint for rug-backing applications wherein polypropylene is used as the fibrous material, is approximately 1 lb./in. oz./yd. on the Web when using a batch bonding process (confined in a press) in a saturated steam atmosphere. Approximately 0.3 lb./in. //oz./yd. of restraint web is preferably used for continuously bonding (confined between moving belts) such webs in a saturated steam atmosphere. The webs to be bonded may be restrained around the edges as Well as by compression during bonding. In the case of small samples, edge restraint may be used alone.
The heat-up rate, fiber crystallinity and bonding temperature are interrelated parameters. While applicant does not wish to be bound by any particular theory of operation, it is believed that bonding occurs when the crystalline form of the polymer as it exists in the fiber begins to melt. A slow heat-up rate permits the crystallinity to increase so that bonding can take place only at a higher temperature corresponding to the melting point of the higher crystalline form of the polymer. This bonding temperature will neither equal nor exceed the crystalline melting point referred to earlier. In the case of rapid heat-up rate, the bonding temperature is reached before the polymer in the fiber has an opportunity to substantially increase in crystallinity and hence a lower temperature can achieve bonding. This is i1- lustrated below.
Starting out with the same nonwoven web of isotactic polypropylene filaments, the faster the heat-up rate, the lower will be the bond-ing temperature necessary to achieve a desired level of tongue tear and strip tensile properties and edge bond count. Conversely, if one maintains the same bonding temperature and changes the rate of heating the initial filamentary web, the sheet bonded at the lower rate will have the lower edge bond count. In operating one typical process batch-wise with saturated steam, the steam pressure in the closed chamber or autoclave is increased such that the heat-up rate of the web to be bonded may be of the order of 30 C./minute. When operating one typical process in a continuous manner, the web is fed continuously through a bonding chamber essentially closed to the atmosphere in which the chamber is pressured with saturated steam at a constant pressure so that the heat-up rate of the web may be of the order of 250 C./ second. Using saturated steam with isotactic ploypropylene filamentary webs, the heat-up rate when operating continuously may be as much as 500 times greater than when operating batch-wise. The bonding temperatures employed in the continuous process can therefore be as much as C. below that used in batch bonding.
While there is some latitude permitted in the process conditions, these should be controlled to limit the drop in filament birefringence to less than 50%.
The novel product of the invention is a self-bonded, nonwoven sheet. By self-bonded is meant that the crystalline and oriented filaments in the bonded sheet are fused to each other at crossover or intersection points. No foreign material is present to achieve bonding. The arrangement of the filaments in the bonded sheet is substantially the same in the unbonded web except for the presence of bonds.
Because of the restraining forces applied to the webs during the bonding process of this invention, some distortion or flattening of the filament cross-section between bonds may occur in the self-bonded sheet products. The exact geometry of the distortion may vary, but sometimes the cross-section of the filament will assume an elliptical shape. Usually distortion occurs at the bond sites, whereas between bonds some of the filaments may be distorted and others may have the same cross-sectional shape as the original filaments in the unbonded Webs.
Microscopic studies show that the major cross-sectional dimension of each filament at the bond site is on the average throughout the sheet, less than twice that of the unbonded filament segments midway between the bond sites. In the preferred polypropylene sheets this value is on the average about 1.35. Essentially no polymer flows to the bond sites when the present process is employed as distinguished from that which usually occurs during heat-bonding of thermoplastic filaments. It is also noted that the bonds are weaker than the filaments as evidenced by the fact that exertion of a force tending to disrupt the sheet, as in tufting, will fracture bonds before breaking filaments.
The sheetsof the invention have substantially uniform birefringence along their width and length. Microscopic studies on the bonded filaments show that the molecular orientation, as measured by birefringence through the filaments at the bond and through the filaments between bonds, is substantially the same. 7
For a better understanding of the concepts involved, reference may be made to the drawings in which:
FIGURE 1 is an enlarged schematic representation of a portion of a bonded sheet product of the invention. The sample illustrated is a thin layer of such a sheet. In the figure, sheet 1 is made up of randomly disposed filaments 2 self-bonded at crossover points through bonds 3.
FIGURE 2 is a schematic illustrative of the break occurring when a ribbon of the bonded sheet of this invention is tested for tensile in accordance with ASTM D- 1682, modified as described below. In the figure fibrous ribbon 4 having broken ends 6 exhibits a large number of separated fibers 5. This evidence showing that the products of the invention fluff up and delaminate during the tensile test establishes that the bonds are weaker than the fibers.
FIGURE 3 is a schematic illustrative of the break occurring when a ribbon of an overbonded sheet is subjected to the tensile test. In the figure fibrous ribbon 7 having broken ends 8 shows only ragged edges and little or no lateral separation of fibers at the break. The bonds in this sheet are clearly stronger than the fibers.
The bonded sheets will exhibit an edge bond count (E.B.C.) of between 30 and bonds per hundred fila ments. Lower levels of bond-ing result in sheet materials having relatively low tensile strength. Exceeding an E.B.C. of 85 results in a substantial loss of sheet tear properties. This loss of resistance also occurs where fiber birefringence is less than about 40% of maximum birefringence. By achieving high filament orientation along with the indicated degree of bonding, the unique result of both high tear and tensile strength is obtained.
The bonded sheet materials of this invention will contain self-bonds which are distributed throughout the textile product both randomly and uniformly. However, it should be realized that certain modifications may be made either during thebonding process or thereafter to modify this distribution to some extent. Thus, the uniformity of bond distribution or bond density in the final product may be affected if two types of filaments are laid down in a nonwoven structure, only one of which is thermoplastic. Where the web is bonded having one surface against a smooth metal sheet and the other surface against a canvas layer it is found that the product has one smooth side and one rough side. The smooth side has reduced pickiness and is less porous than the side adjacent the canvas layer. Another modification may be made in which the bonded sheet may be after-treated by compressing its surfaces while applying heat to form a glazed or embossed surface or pattern.
Other modifications may be made by using standard embossing techniques on the bonded products to prepare various designs uniformly or randomly distributed over the surface of the bonded products, which in some instances may be in the form of patterns which actually form depressions in the sheet products. It is also possible to join or seam two or more pieces of web to make wider sheets by first overlapping the edges of the Webs and then self-bonding the overlapped pieces by the process of this invention. I
The method employed for determining the average edge bond count of bonded samples is described as follows:
The number of bondsper filaments, as counted on the edge face of a bonded web is used as a measure of the amount of self-bonding. This count is based on the probability that the appearance of a bond at an edge face is dependent upon (1) the number of bonds, (2) the size of the bond, and (3) the distribution of these bonds.
The edge bond count measurements are made on samples taken from a representative 6-inch square .test piece of bonded sheet. Four samples are taken from the test piece, one from near each of the four. corners'in such a fashion that the cross sections to be counted later will come from edges parallel to the four edges of test piece. These samples are fastened in a frame and embedded in an epoxy resin which is then polymerized. The embedded specimens are trimmed so that at least a one-quarter-inch cross section of the sheet is exposed .for microtoming. These samples are microtomed transversely to they face of the sheet at a thickness of about 6 microns, and the, resulting sections are deposited in immersion oil on the micrometer stage of a polarizing Projectina No. 4014. This apparatus is a microscope (distributed by Hudson Automatic Machine and Tool Company, 137-139 38th St., Union City, NJ.), which projects a View of the section onto a screen from which the filaments and bonds are counted. A 10X eyepiece and a 20X objective lens are used. The polarizer and an R-I red filter are used to improve filament definition. In the projected view of the section of the sheet, the filaments will appear in various shapes depending on the angle at which they were cut by the microtome. Thus all the variations from a transverse cut perpendicular to the filament axis to a longitudinal cut parallel to the filament axis may be present. Each filament in the projected view is counted once. A bond occurs and is counted whenever there is no complete separation between any two filaments. Thus, while a filament is counted only once, that filament may participate in from to or more bonds depending on the number of other filaments it contacts. For example, if a filament contacts two other filaments, the filament count is 3 and the bond count is 2. If the other two filaments also contact each other, the filament count is 3 and the bond count is 3. At least one full one-quarter inch slice is counted and not less than 200 filaments from each embedded sample. The number of bonds per 100 filaments is averaged from the four samples and reported as the average edge bond count.
In the examples, the strip tensile strength of the sheet was measured in lb./in.//o/z./yd. according to the method of ASTM D-1682, except for the fact that sample Width is 0.5 inch, the distance between the jaws on the tensile machine is 5 inches and the jaw speed is 100% per minute. The tongue tear measurement was in accordance with ASTM D-39 except that specimens were 2 x 2.5 inches.
A measure of the crystallinity index of the synthetic organic polymeric filaments can be obtained by means of X-ray diffraction techniques.
In determining average uniformity ratio, the bonded sample to be tested is divided into quadrants and a quarter inch square section taken from the approximate center of each quadrant. The quarter inch sections are divided into at least three layers of 1-2 filament thickness each. A layer is selected at or adjacent each sheet face and one from about halfway through the sheet thickness. Birefringence measurements are made as described below. The ratio of the highest average birefringence to the lowest for the three layers of each quarter inch specimen is the uniformity ratio. The average ratio for the four specimens of the sheet under test is the average uniformity ratio. For the products of the invention this ratio may not exceed 1.3.
In order to determine the birefringence of filaments from either the unbonded webs or the bonded sheet structures the following techniques are used.
For determination on fibers in a bonded sheet a sample of the sheet measuring A" by A is selected for testing. A thin layer selected from about halfway through the thickness of the sheet can be obtained either by careful delamination or by sectioning with a razor blade. This is necessary since light must pass through the single fiber being tested without interference from other fibers, Measurements should be made on at least ten fibers in the test sample and the results averaged and reported as the average birefringence. The birefringence of round fibers may be determined using a polarizing microscope with a Berek compensator.
Specifically, the following steps are carried out on each filament of the web or sheet. The sample is placed on a 1" x 3" slide with the filament to be examined parallel to the 1" side and a drop of cedar oil deposited therein. The slide is placed on the stage of the microscope with the illuminator (blue light) turned on. The barrel Nicol prism is withdrawn and the substage prism set to zero. The slide is turned to a N-S position and the microscope focused on the filament. The barrel Nicol prism is inserted and the Berek Compensator put in the 30 position. The stage is set to that N-S point where the filament is least visible and the angle of the stage noted. The stage is rotated 45 counterclockwise. The compensator drum is then rotated above 30 and stopped at that point where the black hand between the pointed color segments just crossesthe center of the ocular micrometer scale and its position is noted. This step is repeated except that the drum is rotated below 30. The ditferen-ce between the two compensator readings is determined for each filament measured, and divided by two to give (i) and the values are converted to retardation in millimicrons by the formula:
Retardation=C sin i(1+-0.2040 sin i+0.0709 sin i) where C is a constant associated with the particular compensator employed. Filament diameter is also measured and the birefringence determined by the equation:
retardation (millimicrons) fiber diameter (mm) X 10 In the event of minor fiber flattening between. bonds, a graphical correction may be made as follows. taining the corrected birefringence, a linear plot of the Berek data for each fiber is made with apparent birefringence on one axis and the apparent [fiber diameter on the other. A best-fit line is drawn through the points. The corrected birefringence is read from the plot where the line crosses a point representing the average initial or round fiber diameter. The average initial or round fiber diameter is found by averaging fiber diameter data from at least ten fibers between crossovers in an underbonded or unbonded (less than 3 lbs./in.//oz./yd. tensile) web sample.
Another method for determining the birefringence of nonround fibers using a Berek compensator is as follows. The irregular shaped fiber is carefully sectioned along its length while avoiding stretching (fiber may first be mounted in an embedding medium, and the sectioning done with a microtome) and a longitudinal section of known thickness obtained. The birefringence is then determined on this section using the Berek compensator.
Another satisfactory method for determining the birefringence of fibers with irregular shapes or cross sections where sample thickness is not known involves the use of an interference microscope. This technique is described in Measurement with Phase and Interference Microscope, by Oscar W. Richards, ASTM Special Technical Bulletin, No. 257 (1959). Preparation of thin sample layers from sheets for determination of birefringence using interference microscopy is similar to that described )for the Berek method.
It is to be understood that some of the measurements, such as birefringence, result in destruction of the samples. In such instances, other measurements, such as edge bond count, are made on samples of the same sheet adjacent to those consumed in the prior test.
As mentioned previously, the bonding process of this invention may be carried out either batchwise or continuously, using any suitable form of apparatus for restraining the filamentary article to be bonded. The restraining surface in contact with the filaments may be made of metal, ceramic, glass, plastic, or the like. The restraining means on one side of the web must permit access of steam to and throughout the web and hence must be porous such as a screen or perforated plate. The specific material comprising the restraining surface in contact with the web may be selected on the basis of the frictional forces between the restraining surface and the filaments; thus, a fibrous cloth may be preferred over a coarse screen. For batchwise bonding of the filamentary mass the combined heat-up time and the contact time at the bonding temperature will normally vary from a few seconds up to as much as 30 minutes or more, depending on the bonding temperature, heat capacity of the fluid,
Specific Birefringence= 'For oband of the apparatus, and initial temperature of the apparatus and the choice of filaments used. When operating the bonding process on a continuous basis, combined heatup and contact times vary from a fraction of one second up to one minute or more, depending upon the same process variables as well as the speed at which the filamentary mass passes through the heating fluid.
A typical method for carrying out the bonding process of this invention batchwise follows. A nonwoven web was prepared by air-deposition of continuous isotactic polypropylene filaments of about 3 denier and having a melt-flow rate (M.'F.R.) of 12. The filaments were airdeposited on a moving support in random array. A web containing 3.0 oz./yd. and measuring 20" x 30" was cut and selected for bonding. The web was placed between two cloths and confined by two porous plates at 3 lb./in. normal pressure. The assembly containing the confined web was placed in a closed steam chamber. The steam line was opened and saturated steam at 100 p.s.i.g. was admitted within 35 seconds. under these conditions for 2 minutes. This steam pressure was equivalent to a temperature of 170 C. inside the steam chamber. The polypropylene filaments used had a crystalline melting point of 176 C. The chamber. was then vented, and allowed to cool, following which it was opened and the confined sheet removed from the two plates. The bonded nonwoven fibrous sheet had a tensile strength of 10.4 lb.'/in.//oz./yd. and a tongue tear of 4.3 lb.//oz./yd. The sheet still weighed 3.0 oz./yd. and was approximately 20 x 30" in dimension, indicating that the area of the sheet had not undergone any measurable amount of shrinkage. Upon X-ray and microscopic analyses, it was found that the filaments in the web had a crystallinity index of 63% and a birefringence of 0.024.
This invention is applicable to the production of a wide variety of self-bonded random filamentary articles. The products may be of any thickness, such as single layers of filaments or multiple layers of filaments. The bonding process is useful for producing flat sheets or three-dimensional shapes such as hollow cones, cylinders, etc. It permits the production of many types of apparel andindustrial products, such as clothing, hats, cfelts, heat, electrical and sound insulation, pipes, coated and laminated structures, wall boards, roofs, bagging, acoustical and mechanical fabrication, fume ducts, filters, reinforcing tapes, molded sheets (face masks, brassieres), furniture foundation sheets, shoe fabrics, wall paper, primary and secondary rug backings, floor tiles, other types of floor coverings, mattress pads, siding, industrial packaging, plywood laminates, upholstery, curtains and draperies, window shades, tents and tarpaulins, bookbinding, and the like.
The nonwoven sheet structures of this invention can be laminated to metallic foils and to polymeric films. Particularly useful products are obtained through bonding of the nonwoven polypropylene webs to biaxially oriented polypropylene films or to aluminum foil with an adhesive layer such as low density polyethylene resin or an ethylene copolymer. The nonwoven webs may also be coated with other plastic resins to form composite structures.
The chief advantage of the present invent-ion is that it permits the self-bonding of oriented polymeric filaments to produce nonwoven fibrous articles having a combination of high tensile strength and resistance to tear and good abrasion resistance without significant loss of the physical properties inherent in such polymeric filaments. This unique bonding process takes place without the use of conventional solvents, adhesives, binders, low-melting binder fibers, and other diluents which in former methods of bonding have in one way or another detracted from the physical properties of the filaments.
The combination of high tensile strength and high tear resistance in the products of the invention has been The chamber was held found particularly advantageous for the preparation of 7 primary carpet backings from propylene polymers. In fact, the nonwoven self-bonded fibrous articles produced in accordance with this invention provide a superior primary carpet backing to those obtainable from conventional material, such as woven jute backings. The carpet backings made from sheets of this invention which have bee-n lubricated cause less deflection of the needles during tufting of pile yarn and permit a greater uniformity of tufts per inch when compared to tufting into jute backings. It addition there is a weight saving of approximately 50% in the use of the nonwoven backings of the present invention over woven jute backings.
The present invention represents an economical way of preparing nonwoven fibrous articles which have a combination of excellent tensile strength and tongue tear resistance.
The following examples are illustrative of the present invention, but are not to be construed as limiting the scope of the invention.
The filaments used in Examples I-XI were prepared from isotactic polypropylene having 4% atactic' content. The polymer used for Examples II-X was made by thermal cracking of polypropylene having an M.F.R. of about 1.7 in the presence of an organic peroxide to increase the M.F.R. to the indicated values.
Only one sample was taken from the test piece for edge bond count determination in Examples I-IX, XVI- XXI and XXV. All others were performed as prescribed earlier. 1
The bunching coefficient of Examples XI, XII, and XXII are O.69:5%, 0.72 -4%, and 0.71il0%, respectively. The randomness ratios for Examples XI, XIV, and XXII are 1.12, 1.14, and 1.04, respectively.
In the examples, tensile and tear measurements were made in only one direction except where otherwise indicated. The values for the machine and cross-machine directions will be substantially identical where isotropic (highly random) sheets are involved. The sheets of the invention will have a strip tensile greater than 3 lbs./in.//oz./yd. and a tongue tear greater than 1.5 lbs.//oz./yd. in both machine and cross directions.
Example 1 Two nonwoven webs of the same polypropylene filaments were bonded under different process conditions to produce two self-bonded sheets. The polypropylene filaments used were drawn 2.8x after spinning from a polymer having a M.F.R. of 14. The filaments were 4 denier per filament and had a tenacity of 3.0 g.p.d. The webs were prepared as follows: The polypropylene filaments were spun through a 30-hole spinneret at a rate of 18 grams/ minute total. Each spinneret hole. was 0.015 inch in diameter and the temperature of the spinneret was 234 C. The filaments were led to a heated feed roll operated with a surface temperature of C., and then around the rolladvancing the yarn by means of an idler roll canted with respect to the heated roll. A total of 5 wraps were used on the heated feed roll, which was operated with a surface speed of 612 yards/minute. The yarn leaving the heated feed roll was then passed 5 wraps around an idler roll/ draw roll system operated cold with a surface speed of 1732 yards/minute. The drawn filaments were then charged with a corona discharge device passed into a draw jet for separation, and subsequently deposited on a moving belt to 'form a nonwoven web of randomly distributed continuous filaments. The birefringence of the filaments in the webs was 0.026. Each of the filamentary webs was bonded by passing the web at a speed of 30 ft./min. while under restraint between one porous metal plate and one solidplate, each faced with cloth for a distance of 30 inches through a steam chamber in which saturated steam was maintained .at superatmospheric pressure. The basis weight for each of the bonded nonwoven polypropylene sheets is given in Table I. The fibrous sheet of Sample 1 was very tough and resisted tearing whereas Sample 2 was overbonded and tore readily by hand. Details of bonding the webs and characterization of the bonded sheets appear in Table I.
TABLE I Example 11 A nonwoven web of continuous polypropylene filaments was prepared in the same general manner as that described in Example I. The filaments drawn 4X were made from a polymer having a M.F.R. of 9.6. The filaments were spun from a 30-hole spinneret at a rate of 18 g./min. total. Each spinneret hole was 0.020 inch in diameter and the spinneret was at 250 C. The resulting fiber had a tenacity of 4.4 grams/denier, an elongation of 107%, a modulus of 30.5 grams/ denier, a denier of 8.2, a crystallinity index of 43%, and a birefringence of 0.031. The heated feed roll was operated with a surface speed of 203 yds./min. and the idler roll/draw roll system was operated cold with a surface speed of 800 yds./min. The filaments in the web were randomly deposited on a reciprocating table. The web was bonded by passing it under restraint between one porous metal plate and one solid metal plate, each faced with cloth, through a steam chamber for the same length of travel and speed as in Example I. The steam chamber contained saturated steam at 70.5 p.s.i.g. on the porous plate side and 72 p.s.i.g. on the solid plate side (corresponding to a saturated steam temperature of 158 C. and 158.7 C., respectively). The difierence in pressure on either side of the solid plate served to restrain the web and filaments against shrinkage during bonding.
A very tough nonwoven bonded fibrous article suitable for use as a primary carpet backing was obtained. Its physical properties and structural parameters are given as follows:
Basis weight: 2.8 oz./yd.
Tensile strength: 8.2 lb./in.//oz./yd. Elongation: 40 percent.
Tongue tear: 6.4 l'b./oz./yd.
Edge bond count: 60 bonds per 100 filaments. Crystallinity index: 52 percent. Birefringence: 0.028.
Example III A nonwoven random fibrous web of 3.5 denier filaments was prepared from polypropylene continuous filaments by the same procedure as that described in Example II, with the exception of the draw ratio. In this example, the filaments were drawn 136x after spinning from the 9.6 M.F.R. polymer. The filaments had a tenacity of 2.1 grams/ denier, an elongation of 355%, a modulus of 14.5 grams/denier, a birefringence of 0.028, and a crystallinity index of 44%. This web was bonded by passing it at a speed of 10 yards/min. through the steam chamber of Example II containing saturated steam at 53 p.s.i.g. on the porous plate side and 54 p.s.i.g. on the solid plate side (corresponding to a temperature of 149.3 C. and 149.8 C., respectively).
The tough nonwoven fibrous article prepared from these polypropylene filaments can be used as secondary carpet backing or as bagging, wall covering or tent substrate. Its properties are listed below:
Basis weight: 3.5 oz./yd.
Strip tensile strength: 6.4 lb./in./ /oz./yd. Elongation: 54 percent.
Tongue tear: 4.6 lb./oz./yd.
Edge bond count: 61 bonds per 100 filaments. Crystallinity index: 57 percent. Birefringence: 0.025.
Examples I V-X I A series of randomly laid nonwoven webs of polypropylene filaments were prepared following the same general procedure given in Example II, except that the polymer M.F.R. varied and the filaments were given different amounts of draw after spinning. The products of Examples IV-X were bonded batchwise following the general procedure outlined above as the typical method for batch operation. Identification of the variables is made in Table 11. Some of the conditions for bonding thewebs were deliberately varied to show the critical effect on the properties of the bonded sheets. In Examples VII, VIII and IX, the webs were bonded under the same conditions except for different amounts of compressional restraint on the surface of the webs, which resulted in the bonded webs having a linear filament shrinkage of 0%, 25% and 50%, respectively. The bonded webs of Examples IV, VI, VIII, and IX fall outside the scope of this invention, as will be seen by an inspection of the edge bond counts and tongue tear strength data given in Table II. The crystallinity index of the filaments in the bonded sheet of Example XI which was prepared by the procedure of Example II, was 53% The above unbonded web of Example X may be wrapped around a mandrel in the form of a tube and steam bonded as above in this form. Other curved or shaped articles of the unbonded web can be bonded in accordance with this invention.
TABLE II Fiber S t Bonded Sheet a Polymer Basis Wt. Steam Bonding Example (M.F.R.) d.p.f. Tenacity Birefrin- (oz./yd. Press. Temp. Strip Tongue E.B.C.
(g./d.) gence (p.s.i.g.) C.) Tensile Tear (Bonds/100 Biref.
(lb/inl/ (lb/loz/ Filaments) ozJyd. yd.
1 Edge bond count not measured since sheets were stifi, boardy, 2 Birefringence was not measured because the sheet had fused to of non-uniform thickness and showed some areas of fusion. too great an extent to permit measurement.
1 Examples XII-XX V A series of randomly laid nonwoven webs were pre-. pared for bonding. Examples XII-rXV were bonded by essentially the same batch process described for Examples IV-X except the heat-up rate was 30 C. per minute while Examples XVI-XXV are samples bonded continuously at a rate of 30 feet per minute substantially in accordance with the steam bonding process of Example II. Table III below, identifies the constitution of the webs prior to bonding and the properties and structural characteristics of the webs after bonding. The bonding temperature and pressure of the steam is given in the table. It should be noted that since the bonding is effected in a substantially closed system, the heating fluid i.e., the saturated steam, is well within the 5 C. variation permitted by the process throughout the three dimensions of the web. Actually the temperature of the heating fluid at any point in the web is within 2 C. of the temperature at any other point within the web.
Polymer A is polypropylene (96% isotactic, 4% atactic). Polymer B is a composition of 90% isotactic polypropylene and branched polyethylene melt blended before filament extrusion. Polymer C is a melt blend (M.F.R. 3) of 85% isotactic polypropylene and solid polyisobutylene. Polymer D is composed of 88% isotactic polypropylene and 12% atactic polypropylene. Polymer E is composed of the same ingredients as polymer D but having 75% of isotactic polypropylene. Polymer F is a block copolymer of propylene with 2-5% ethylene having a Vicat softening point of 132 C., and polymer G is linear polyethylene.
The webs of Examples XII-XV were prepared following the general procedures of Example I with the following variations: Polymer was spun through a 15-hole spinneret having a hole size of 0.020 inch at a rate of 9 grams per minute and the filaments were drawn 4.2x. Feed roll surface speed and draw roll speed were 181 yds./1nin. and 763 yds./min. respectively, and the webs were laid down on a reciprocating table.
The webs of Examples XVI-XXI were obtained following essentially the spinning, drawing, charging and laydown procedures of Example I, however, machine settings and temperatures, e.g., polymer throughput, temperature,
feed and draw roll speeds, etc., were changed to accommodate the particular polymer or to produce a desired tenacity or denier.
The webs of Examples. XXII-XXIV were prepared following the general procedures of Example I with the following variations: Polymer was spun through a 30- hole spinneret having a hole size of 0.015 inch at a rate of 18 grams per minute and the filaments were drawn 4.2x. Spinneret temperature was 223 C. and the spinneret block temperature was 240 C. Feed roll temperature was 118 C. Feed roll surface speed and draw roll speed were 216 yds./min. and 900 yds./min., respecti-vely, and the webs were laid down on a moving belt.
The web used in Example XXV consisted of 4-inch staple fibers made by cutting crimped continuous filaments of isotactic polypropylene. The staple was carded and then garnetted, a crosser-lapper being employed to form the unbonded webs. A series of these unbonded staple webs were bonded at difierent temperatures to give a series of self-bonded staple sheets having dilferent degrees of bonding and randomness ratios varying from about 1.3 to about 1.8. The characteristics of one of the samples in this staple series having acceptable sheet properties are reported as Example XXVin Table HI.
The average uniformity ratio for the sheets of Examples XIV, XXII and XXIII were 1.07, 1.07, and 1.11, respec-' tively.
The bonded sheets of Examples XXII, XXIII and XXIV were tested as primary carpet-backings by tufting pi-le yarn into each sample on a commercial carpet tufting machine using methyl hydrogen polysiloxane as a lubricant during tufting. The number of fibers broken during tufting of the three carpets was, respectively, few, numerous and excessive. The carpet grab tensile in pounds was 198, 65. and 23, and the carpet tongue tear was 59, 12,. and 2, respectively. The small amount of fiber breakage in tufting the lubricated sheet of Example TABLE III Fiber Saturated Basis Wt. Steam Bonding Example (ea/yd?) Pressure Temp. C.)
Polymer (M.F.R d.p.f. Tenacity Bire- (p.s.i.g.)
g.p.d. tringence V A 12-15 9. 3 4. 0 2. 0 166. 2 A 12. 15 9. 3 4. 0 2. 0 169. 9 A 12. 15 9. 3 4. 0 2. 0 175. 1 A 12-15 9. 3 4. 0 2. 0 179. 8 B 13. 4 8 4. 0 3. 0 78 161. 3 C 3 8 2. 7 3. 2 87. 5 165. 2 D 15 8 4. 0 3. 0 88 165. 4 E 6 7. 9 3. 2 3. 7 73 159. 1 F 2. 5 11 3. 6 2. 5 58 151. 9 G I 15 7. 5 3. 4 4. 4 29 133. 7 A 11. 5 8 4. 0 4. 3 85 164.2 A 11. 5 8 4. 0 4. 3 90 166. 2 A 11. 5 8 4. 0 4. 3 100 169. 9 A 11. 5 7. 8 3. 8 2. 5 74 159. 6
TABLE HI.Cont1nued Bonded Sheet Example Birefringence Birei. Drop Crystallinity Strip Tensile Tongue Tear Thickness E.B.C. (percent) Index (1b./in.// (lb.//oz./yd. (111.) (Bouds/ z.Iyd. 100 fil.)
2. 1 2. 2 27 0. 031 4. 6 3. 3 37 0. 028 14. 7 4. 9 61 0. 020 15. 3 0. 4 92 0. 014 4. 5 4. 9 56 0. 024 11. 4 5. 3 66 0. 024 7. 5 7. 8 73 0. 017 5. 5 8. 0 72 0. 016 6. 2 8. 0 47 0. 022 6. 0 3. 8 49 0. 042 1 6. 0 1 6. 8 70 0.026 2 7. 1 2 5. 8 XXIII 1 10. 5 1 4. 7 79 0. 022 27 60 2 l0. 5 Z 4. 5 XXIV 1 l6. 9 l 0. 8 92 0. 015 50 2 18. 6 2 0. 5 XXV 3.1 2.6 58 0.028
1 MD 2 XD Example XXVI 30 ments are so disposed as to provide a bunchlng coefficient Using the batch process generally described in Examples IV-X, a randomly laid nonwoven web of continuous filaments was prepared for bonding. The filaments were composed of a copolyester of ethylene glycol with a 90%/ 10% mixture of terephthalic/isophthalic acids. The filaments had a denier of 3. The wab was placed in the autoclave between two confining metal screens and exposed to 150 p.s.i.g. saturated steam for two minutes. This represented a bonding temperature of 186 C. The assembly was removed from the autoclave after cooling. The product was found to be a strong, tough, bonded sheet having an average edge bond count of 40 bonds/ 100 filaments, an average birefringence of 0.132, a strip tensile of 6.2 lbs./in.//oz./yd. and a tongue tear of 3.4 lbs.//oz./yd. The basis weight of the Web after bonding was 2.4 oz./yd. The bonded sheet was useful for electrical insulation.
What is claimed is:
1. A tough nonwoven sheet having a tongue tear greater than 1.5 lb.//oz./yd. comprising randomly disposed crystalline and oriented synthetic organic polymeric filaments of textile denier, said filaments being selected from the group consisting of isotactic polypropylene, linear polyethylene, polyethylene terephthalate, polyhexamethylene adipamide, polycaproamide, copolyester of ethylene glycol with a 90% 10% mixture of terephthalic/isophthalic acids, crystalline copolymers of at least 75% propylene or ethylene with up to of other copolymerizable monomers and mixtures of at least 75% isotactic polypropylene or linear polyethylene with up to 25% of other polymers, and being self-bonded at a multiplicity of filament intersection points and possessing a birefringence that is at least 40% of their maximum birefringence, and said sheet having an average edge bond count of between about and 85 bonds per hundred filaments.
2. The sheet of claim 1 wherein the polymeric filaments are isotactic polypropylene.
3. The sheet of claim 2 having an average uniformity ratio of under 1.3.
4. The sheet of claim 1 wherein the polymeric filaments are linear polyethylene.
5. The sheet of claim 1 wherein the polymeric filaof at least 0.5 and a randomness ratio not greater than 2.
6. A laminar structure comprising the nonwoven sheet of claim 1 and a polymeric film.
7. The laminar structure of claim 6 wherein the film is an oriented polymeric film.
8. The laminar structure of claim 7 wherein the film is a biaxially oriented polypropylene film.
9. A laminar structure comprising the nonwoven sheet of claim 1 and a metallic foil.
10. The laminar structure of claim 9 wherein the metallic foil is an aluminum foil.
11. A method for preparing a self-bonded nonwoven sheet of high tensile and tear strength comprising uniformly exposing a web of randomly disposed oriented synthetic organic polymeric filaments selected from the roup consisting of isotactic polypropylene, linear polyethylene, polyethylene terephthalate, polyhexamethylene adipamide, polycaproamide, copolyester of ethylene glycol with a 10% mixture of terephthalic/isophthalic acids, crystalline copolymers of at least 75 propylene or ethylene with up to 25 of other copolymerizable monomers and mixtures of at least 75% isotactic polypropylene or linear polyethylene with up to 25% of other polymers of textile denier having a birefringence that is at least 40% of their maximum birefringence to a nonsolvating fluid atmosphere at a temperature, in the range of from 45 C. below up to the crystalline melting point of the polymer, that is sufiicient to self-bond the filaments at a plurality of spaced intersection points, while restraining the web to prevent filament shrinkage of greater than 20%, said temperature and restraint confining the birefringence drop to less than 50% and preventing a drop in birefringence to a level below about 45% of the maximum birefringence.
12. The process of claim 11 wherein the bonding is suflicient to produce sheets having a strip tensile strength greater than 3 lbs./in.//oz./yd.
13. The process of claim 11 wherein the temperature of the heating fluid at any point Within the web differs from the temperature at any other point by less than 5 C.
14. The process of claim 13 wherein the temperature variation is less than 2 C. and the filaments are isotactic polypropylene.
15. The process of claim 11 wherein the web comprises filaments of isotactic polypropylene having a crys- 19 20 tallinity index greater than 40% and a birefringence of 3,049,466 8/1962 Erlich 156-334 between 0.016 and 0.040 and the nonsolvating atmosphere 3,106,501 10/ 1963 Cobb et a1 156-180 is saturated steam. 3,148,101 9/1964 Allman et a1 161150 X References Cited by the Examiner 5 FOREIGN PATENTS UNITED STATES PATENTS 2,476,282 7/1949 Castellan. 2,905,585 9/1959 Hubbard et a1. 161-402 EARL BERGERT 2,920,992 1/ 1960 Hubbard 161150X 10 MORRIS SUSSMAN, ALEXANDER WYMAN, 2,956,723 10/1960 Tritsch. Examiners.
626,443 8/ 1961 Canada.