Abrasion-resistant Composite Sheet BACKGROUND OF THE INVENTION

Field of the Invention This invention relates to an abrasion-resistant composite sheet and to a process for making the sheet. More particularly, the invention concerns such a composite sheet in which pile-like groups of fibers are immobilized by a resin in a position that is generally vertical to the surface of the sheet. Description of the Prior Art It is known to laminate various woven or nonwoven fabrics to resin layers to form composite sheets intended for use in Λeπnofoiming and molding processes. For example, such composite sheets are disclosed in Miyagawa et al, U. S. Patent 4,298,643, Zafiroglu, U. S. Patent 5,075,142, and in Japanese Patent Application Publications 63-111050 and 63-162238. Moldable composites have been utilized in many applications. However, such composites are in need of improvement when intended for use in articles that are subject to strong abrasion, for example, in athletic shoe parts, luggage surface layers, protective work clothes, heavy duty sacks, etc.

Pile fabrics and pile-like fabrics, such as velvets, velours, terry cloths, moquettes, and tufted-pile fabrics have a surface layer in which fibers are generally vertical to the surface of the fabric. Also, Zafiroglu, U. S. Patents 4,773,238 and 4,876,128, disclose certain stitchbonded fabrics in which fibrous layers are contracted by means of elastic threads to form pile-like groups of fibers. Generally, such fabrics are not disclosed for incorporation into resin-impregnated composite sheets. However, Japanese Laid-open Patent Applications 64-85614 and 64-85615 disclose a floor mat which includes a tufted-pile fabric to which a rubber resin is added. The tufted pile has an 8-mm pile height and a 0.08-g/cm3 pile fiber concentration. The combination of pile fiber and resin is 62 % by weight pile fiber and 38% by weight resin. The combination has an average over-all density of 0.13 g/cm*--*. The mat has a pile parameter, P, as defined hereinafter, of only 0.1 g/cm3. The present inventor found that (a) such a low pile fiber concentration in combination with such a low over-all layer density and low pile parameter does not provide the mat with high abrasion resistance and (b) even if such relatively high piles are have a dense layer of resin at the outermost tips of the pile, (e.g., with all the resin in the top 1 mm of the pile) such a resin/pile layer is stretched and torn apart by strong abrasion. Increases in the abrasion resistance of such floor mats would improve their utility.

Watt et al, U. S. Patent 4,808,458, discloses flocked pile fabrics wherein a foamed resin is located primarily in the bottom 75%, preferably the bottom 50%, ofthe pile leaving the remainder ofthe pile nearly free of resin, for the purpose of obtaining a suede effect. Such products are ineffective in resisting severe surface abrasion.

Zafiroglu, International Application Publication WO 94/19523 discloses an abrasion-resistant resin-impregnated nonwoven fabric. The fabric is made by contracting a nonwoven fibrous layer to cause groups of fibers buckle out ofthe plane ofthe layer and form "inverted U-shaped" loops that project generally vertically from the layer and then impregnating the contracted layer with resin. Typically, within the nonwoven fibrous layer, the fibers are positioned in all directions. The individual fibers in the loops ofthe buckled fibrous layer therefore are not positioned in a generally perpendicular direction to the contraction, but are rather positioned in all directions within the loops ofthe buckled nonwoven fibrous layer. Even though the groups of fibers form vertical U-shaped loops, a large fraction of the fibers within the loops still are not oriented perpendicular to the plane of the layer. Resin impregnation of such contracted fibrous layers provides a composite sheet that has an abrasion resistance that is superior to a resin- impregnated flat (not contracted or buckled) nonwoven layer, but further improvements in wear resistance are desired.

An aim ofthe present invention is to provide a composite sheet that has a very high resistance to severe abrasive wear.

SUMMARY OF THE INVENTION The present invention provides an abrasion-resistant composite sheet.

The sheet has an upper surface and a lower surface and comprises a resin, groups of pile-like fibers and a planar fibrous network. The fibrous network is located between and substantially parallel to the upper and lower surfaces ofthe composite sheet. The groups of pile-like fibers are located between the upper and lower surfaces and are mechanically connected to and protrude generally perpendicularly from the planar fibrous network. The composite sheet has a stretchability in any direction that is no more than 25%. The groups of pile-like fibers have an effective pile concentration, c _e ff, of at least 0.10 g/cn_3, preferably in the range of 0.15 to 0.4 g/cn_3, and are surrounded and immobilized by the resin in the generally perpendicular position.

Typically, the resin amounts to 30 to 90 %, preferably at least 50%, ofthe total weight ofthe composite sheet. The groups of pile-like fibers and planar fibrous network are composed of fibers of textile decitex (i.e., of 0.7 to 20 dtex). Further, the pile fiber/resin layer has a thickness in the range of 0.5 to

3 mm, preferably 1 to 3 mm, an over-all density, d, of at least 0.4 g/cn_3, preferably in the range of 0.5 to 1.2 g/cw , a stretchability in any direction of no greater than 25%, preferably no greater than 10%, a vertical compressibility of no greater than 25%, preferably no greater than 10% and a pile parameter P, calculated by the equation, P of at least 0.3 g/cm--- ^** , preferably at least 0.35. Typically, the pile fiber/resin layer has a unit weight that is in the range of 500 to 2,500 g/n_2. The surface ofthe composite sheet is abrasion resistant and is abraded by no more than 50 microns per 1,000 cycles of 40-grit Wyzenbeek abrasion testing. The invention also includes a process for making the abrasion- resistant composite sheet, comprising (a) providing a fabric in which fibers form or can form pile-like fiber groups in a surface layer of 0.5-mm to 3 mm thickness in which the pile-like groups of fibers are positioned generally perpendicular to the surface ofthe fabric and an end of each pile-like group of fibers is mechamcally attached to, or is protruding through, a generally horizontal fibrous network, (b) contracting the area ofthe fabric by a factor of at least two and usually not more than by a factor of 15, preferably by a factor in the range of 3 to 12, to buckle and vertically orient groups of fibers on the surface and to increase the concentration of generally perpendicular groups of pile-like fibers to a concentration of at least 0.10 g/cm--* ^* , preferably in the range of 0.15 to 0.4 g/cm^, (c) immobilizing the pile-like groups of fibers in the generally perpendicular position by incorporating a resin in the surface layer and providing the layer with an over-all density of at least 0.4 g/cn_3 and a pile parameter, P, of at least 0.3 g/cia . the resin constituting in the range of 30 to 90%, ofthe total weight ofthe surface layer, and (d) optionally further stabilizing the dimensions ofthe composite by attaching inelastic elements to the composite sheet.

The invention further provides a shaped article having the abrasion- resistant composite sheet attached to at least a portion ofthe surface ofthe article. The composite sheets provide surface layers that are capable of withstanding Wyzenbeek 40-grit abrasion-testing (as described hereinafter) with a loss in thickness of no more than 50 microns/1,000 cycles. BRIEF DESCRIPTION OF THE DRAWINGS The invention will be better understood by referring to the attached drawings. Figure 1 schematically represents an idealized magnified cross- section of an abrasion-resistant surface layer ofthe invention, in which pile¬ like groups of fibers are in the form of generally vertical inverted U-shaped loops 10 of height H and base B, which loops 10 are immobilized in resin 15 between upper surface 17 and lower surface 18 of the resin-impregnated

fibrous layer. The loops 10 are generally perpendicular to upper surface 17 ofthe layer and contracted elements 20 are generally parallel to surfaces 17 and 18. Surface 17 is the surface that is intended to be exposed to the abrasive conditions. Fig. 2 represents a segment 11, of a yarn or of a bundle of fibers in a nonwoven fibrous layer, before buckling or contraction of the yarn or layer, the segment being located between stitches or fixed points 12 and 13, which are a distance of S apart. Fig. 3 and 4 respectively represent segment 11 would appear after the fabric or fibrous layer in which it was located is contracted by a factor of two (Fig. 2) or three (Fig. 3) in the direction ofthe length ofthe segment. Note that the greater contraction is accompanied by a greater verticality of the fiber bundles or yarns. DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

The following description of preferred embodiments is included for the purposes of illustration and is not intended to limit the scope ofthe invention; the scope is defined by the appended claims.

As used herein, the terms "pile-like groups of fibers" or "pile-like fibers" includes buckled yarns, inverted-U-shaped loops formed from buckled nonwoven layers of textile fibers, tufted yarns and the like. The fibers within each of these pile-like groups of fibers, as well as the fibers of the nonwoven fibrous layers, are of conventional textile decitex, namely, of 0.7 to 20 decitex.

The highly abrasion-resistant composite ofthe invention has a surface layer in which pile-like groups of fibers are crowded together and immobilized by a resin within the surface layer. The pile-like fibers protrude generally perpendicularly from a fibrous network that is also located within the surface layer, for example at the mid-plane or at the base ofthe layer. The fibrous network can be a nonwoven fibrous layer, a knitted fabric, a woven fabric, or the like. Typically, as shown in Fig. 1, the fibrous network 20 is located no further than 3 mm from outer surface 17 ofthe layer that is to be exposed to abrasion. The resin 15 and the fibrous network 20 prevent the pile-like fibers 10 from moving from side to side or from collapsing into the layer when the surface ofthe composite is subjected to lateral and normal forces during repeated cycles of abrasion or rubbing. The composite has a stretchability and the resin/pile-fiber layer has a compressibility (measured as described hereinafter), each of which is no greater than 25%, preferably no greater than 10%. Stretchability and compressibility, respectively are measures of how much the fibers can be moved from side to side and how much the fibers can be collapsed from their perpendicular position when the composite is subjected to conditions of severe abrasion.

In accordance with the invention, the surface layer ofthe abrasion- resistant composite has a thickness in the range of 0.5 to 3 mm. Thicknesses of greater than 3 mm are avoided because it is difficult to immobilize and stabilize pile-like layers if the to-be-abraded surface is further than 3 mm away from the horizontal fibrous network. The effective concentration, c _e f , ofthe generally vertical pile-like fibers within the surface layer is in the range of 0.1 to 0.5 gram/cm-**, preferably in the range of 0.15 to 0.4 g/cm^. In the surface layer ofthe composite, the resin constitutes 30% to 90%, preferably at least 50% and most preferably at least 70%, ofthe total weight ofthe layer. The over-all density, d, ofthe surface layer is at least 0.4 g/cm-*- ^* . For high resistance to abrasion and wear, the surface layer ofthe composite has a "pile parameter", P, ,that is at least 0.3 g/cn_3. The pile parameter, P, as defined herein is the square root ofthe product ofthe pile¬ like fiber concentration and the overall density ofthe surface layer. The pile parameter is expressed by the formula, P = [(c)(d)] ^.

Typically, the surface layer of a composite sheet ofthe invention is not completely filled with resin and fiber. The layer can contain many small voids. An over-all density ofthe surface layer of at least 0.4 g/cm**- ⁵ , and generally not more than about 0.9 g/cπ_3, automatically implies the presence of voids in the surface layer amounting to at least 10% and as much as 65% or more ofthe total volume of the layer, because most fibers and resins suitable for use in the invention have densities of at least 1.0 g/cn_3. However, overall densities of as high as 1.2 g/cm* , and accordingly resin densities of greater than 1.0 g/cm3 are contemplated for use in the invention. For lighter weight and more flexible composite sheets, surface layer void volumes of 25 to 75% are preferred. Usually, high amounts of resin are employed in the composites that have low concentrations of vertical fibers in the surface layer. For example, in composite sheets ofthe invention having effective pile-fiber concentrations near the lower O.l-g/cm- limit, a layer constituting 70 to 90% resin is preferred. For relatively high effective pile- fiber concentrations, lower percentages of resin can be used (e.g., 30-50%).

Various types of resins are suitable for immobilizing the fibers or fiber bundles in the generally vertical position. Particularly useful polymeric resins include polyurethanes, epoxies , synthetic rubbers, polyesters, polyacrylates, polyethers, polyetheresters, polyamides, copolymers and mixtures thereof and the like. The resins can be thermoplastic or thermosetting. Very soft resins, for example soft rubber latexes or resins that are highly foamed, generally are not suitable for use in the present invention. Resins suitable for use in the invention adhere well to the fibers and usually

are well distributed throughout the pile-fiber layer. However, if the resin distribution is not fully uniformly distributed throughout the pile-fiber layer, the resin preferably is concentrated nearer the surface that is to be abraded than to the end mechanically attached to the horizontal fibrous network. The resins can be applied to the pile-fiber layer in any of several conventional ways, as for example, by dipping, spraying, calendering, applying with a doctor knife, or other such techniques. The resin may be applied from a solution, dispersion or slurry or by melting a layer ofthe resin and forcing it into the layer of vertical fibers. The resin also can be introduced as adhesive particles or as binder fibers that are activated by heat or chemicals. Conventional coagulation and or foaming techniques also may be employed in applying the resin. In most instances, the resin or binder can be introduced into the fibrous layer before, during or after the contraction step which is required according to the process ofthe invention for obtaining the desired density of vertical fibers in the surface layer. However, when forming the surface layer and incorporating resin, care must be taken to avoid deflecting the fibers from their vertical position and to avoid immobilizing the fibers before the fibers have been positioned vertically. After incorporation into the pile-like fiber layer, the resin is dried and/or cured by conventional methods. If during resin application, the pile-like fiber layer is compressed in thickness by as little as 25%, the pile-like fibers sometimes can be deflected significantly from verticality and result in the composite sheet having decreased abrasion resistance.

The surface layer of a composite sheet ofthe invention is much more abrasion resistant than a 100% resin layer containing no vertical fibers or a resin/fiber surface layer in which the fibers are not in a vertical position. For example, composites ofthe invention having vertical fibers encased in a layer of relatively soft polyurethane resin can be 50 to 150 times more resistant to abrasion than a layer made from 100% ofthe same resin. When harder, relatively more wear-resistant resins are employed, the advantage of the fiber/resin layer ofthe invention over a layer of 100% ofthe same resin is not as great. However, compared to surface layers containing no fibers or containing primarily horizontal fibers, surface layers of composites ofthe invention still are very much more abrasion resistant. In the process ofthe invention, the first step provides a fabric that has, or has the capability of forming, a pile-like fiber layer. As used herein, the term "pile-like fiber layer" means a surface layer of a fabric within which fibers are positioned generally vertical to the surface ofthe fabric.

In accordance with certain embodiments ofthe process ofthe invention, the generally vertical fiber layer is derived from a substantially nonbonded fibrous nonwoven layer which is subjected to a contraction step that causes fibers or groups offibers to buckle out ofthe flat plane ofthe fibrous nonwoven fabric to form the pile-like layer. The generally vertical fibers ofthe pile-like layer are depicted in Figure 1 and often appear as inverted U-shaped loops, of height H and base B. Such loops, when formed in part from a nonbonded fibrous nonwoven, typically have an average spacing (i.e., base B) in the range of 0.1 to 2 mm, and a height-to-base ratio of at least 0.5. Loop spacings as small as 0.1 mm and height to base ratios of as large as 15 can be achieved when the layer is highly contracted (e.g., by a factor of 10 to 15) and additional elements capable of foiming pile or pile¬ like fibers are included in the structure (e.g., other non-elastic yarns in the stitching patterns). Practical ways to determine the H and B dimensions of the loops, are described below in the paragraphs on test methods.

A typical nonwoven fibrous layer for use in an embodiment ofthe process ofthe invention is a thin, supple, substantially nonbonded web of staple fibers or continuous filaments of textile decitex. These fibrous materials are referred to collectively herein as "fibers". The fibers are naturally occurring or formed from synthetic organic polymers. Fibers that are smaller than 5 dtex and longer than 5-mm are preferred. Preferred fibrous layers are capable of buckling over a relatively short interval (e.g., as small as 1 mm) and typically weigh in the range of 15 to 100 grams/square meter, preferably less than 60 g/mΛ Suitable materials for the starting nonwoven fibrous layer include carded webs, air-laid webs, wet-laid webs, spunlaced fabrics, spunbonded sheets, and the like. Generally, thick lofty webs, felts, adhesively or thermally bonded webs, or the like are not suitable; such materials usually are difficult to buckle over short intervals.

Contraction and buckling ofthe fibrous layer can be accomplished in any of several ways. For example, a contractible element, or array of contractible elements, can be intermittently attached to the fibrous layer. The spacing between attachment locations is typically at least 1 mm to allow for efficient buckling. Then, the element or array of elements is caused to contract so that the area ofthe fibrous layer is decreased significantly and groups offibers buckle out ofthe plane ofthe layer. Before the contractible elements are attached, additional gathering or contraction can be imparted to the fibrous starting layer by overfeeding the layer to the apparatus being employed to attach the contractible elements.

Many types of contractible elements are suitable for use in the present invention. For example, the nonwoven fibrous layer can be stitchbonded with elastic yams under tension. Textured stretch yams, covered or bare spandex yams and the like are suitable yams for contractible element stitching. After the stitching, the tension can be released to cause the desired contraction and buckling ofthe fibrous layer. Instead of stitching, extended elastic elements in the form of warps, cross warps, films or the like, can be intermittently attached to the fibrous layer by hydraulic entanglement, adhesive or thermal point bonding, or the like. Thereafter, tension on the extended elements can be released to cause layer contraction and buckling. Other types of contractible elements, which shrink on being treated with heat, moisture, chemicals or the like can be attached intermittently to the fibrous layer without initial tension or extension in the elements. After attachment, the contraction ofthe contractible elements can be activated by appropriate treatment.

Still another way of accomplishing the contraction and buckling ofthe fibrous layer involves intermittently attaching the fibrous layer to a stretchable substrate that necks-in in a direction that is transverse to the direction in which the substrate is pulled. For example, certain substrates, when stretched by 15% in one direction, can automatically experience substantially irreversible contraction (i.e., neck in) in the transverse direction by an amount that is two or three times the amount of stretch. Thus, appropriate intermittent attachment of a fibrous layer to the stretchable substrate before the stretching and necking-in operation, and then applying the stretching forces to the combined layer and substrate, can significantly decrease the area ofthe fibrous layer and cause buckling of groups offibers as required by the process ofthe invention.

In other embodiments ofthe invention, the pile-like layer offibers is derived from conventional yams in a knit or woven fabric which is constructed with contractible elements. When the contractible elements contract, the area ofthe fabric decreases significantly and causes the conventional yams ofthe fabric to gather and buckle and project in a vertical direction from the horizontal plane ofthe gathered fabric. In another embodiment, the pile-like layer includes loops offibers that project from wrapping yams that were loosely wrapped around the axis ofthe contractible core of an elastic combination yam. Generally, yams that can be contracted and buckled provide denser pile-like layers than do contracted and buckled nonwoven fibrous layers. After resin is applied according to the invention,

the resultant composite sheets made with buckled yarns possess a higher abrasion resistance than those made with buckled nonwoven fibrous layers.

In still other embodiments ofthe invention, a pile-like layer can be derived from a combination of a contracted substantially nonbonded fibrous nonwoven fabric, loose wrapping yams of a contracted combination yam and/or a buckled non-elastic yam. In those embodiments wherein the pile¬ like layer is formed partially or totally from buckled yams derived from a knit or woven fabric, the knit or weave is sufficiently coarse permit satisfactory yam buckling. Typically, the buckled elements, before buckling, have a flat length of at least 1 mm. In yet another embodiment ofthe invention, a tufted pile fabric is contracted to increase the density ofthe pile tufts, for use in a composite sheet ofthe present invention.

As used herein, a combination yam means a yam having a contractible core (e.g., provided by an elastic or shrinkable yam) surrounded by a non-contractible conventional "wrapping" yam or "covering" yam. The wrapping or covering yam may be of any natural or synthetic fiber. The wrapping may be combined with the elastic core while the elastic core is under tension by conventional wrapping, winding, plying, covering, air jet entangling or intermingling, or the like. The core may be a yam or monofilament of any elastic material. Cores of spandex yam are preferred. If the wrapping yam is combined loosely (e.g., fewer than 3 turns/inch) with a tensioned and extended elastic core, when the tension is released, the core contracts and the wrapping yam contracts and buckles perpendicular to the core. When a fabric is knit, woven or stitchbonded with a combination yam under tension, when the tension is released from the combination yarn, the yam contracts and the wrapping yam buckles contributing pile-like fibers to the surface layer. However, if the wrapping yam is wound too tightly around the elastic core, the combination yam cannot provide pile-like fibers to the composite sheet In the contraction step ofthe process ofthe invention, the area ofthe fabric from which the vertical fibers are derived is contracted by a factor of at least 2, preferably in the range of 3 to 10 and sometimes by a factor as high as 15. The contraction step is employed before or during application of the resin. The fabric cannot be contracted after the resin has become set. As a result ofthe contraction step, the concentration of vertical pile¬ like fibers in the surface layer of the fabric is significantly increased. Then, the fibers are immobilized in place by adding a resin to the surface layer, in an amount in the range of 30 to 90% ofthe total weight ofthe resin- containing layer (i.e., weight of resin and pile-like fibers). Preferably the

resin amounts to at least 50% and most preferably at least 70% ofthe total weight ofthe layer. Typically, the resin is distributed uniformly throughout the layer of pile-like fibers. However, as long as the pile-like fibers are immobilized in a substantially vertical position, the distribution of resin can be somewhat non-uniform and there also can be a fairly large void fraction in the layer. The voids can amount to as much as 75% or more ofthe total volume ofthe layer. Ridding the layer of voids to completely fill the layer with resin is unnecessary. In fact, techniques for this purpose are avoided because the techniques often excessively crush the fibers and deflect the fibers from a vertical position. Vertical pile-like fibers in the resin/fiber layer are essential for the improvements in abrasion resistance provided by the composite sheets ofthe invention.

In addition to the pile-like fabrics described above, composite sheets according to the invention also can be produced from other types of pile fabrics, such as tufted pile fabrics, velvets, moquettes, and velours so long as the fabrics have a pile height and a pile fiber concentration within the requirements ofthe present invention and the fabrics are capable of being combined with resin to produce a layer having a pile parameter of at least 0.3 g/cm-3. Such starting fabrics typically have pile-fiber concentrations in the range of 0.05 to 0.15 g/cn_3. In composite sheets made with such fabrics, the resin typically amounts to at least two-thirds ofthe total weight ofthe pile¬ like fiber/resin surface layer. The over-all density ofthe pile-like fiber/resin layer is generally at least 0.4 g/cm-^, preferably in the range of 0.5 to 0.9 g/cm-3. Higher, rather than lower, pile densities are preferred because the higher density piles are more resistant to compression and pile fiber deflection during the resin- impregnation step and ultimately can lead to a more abrasion resistant layer.

The abrasion-resistant surface ofthe composite sheet ofthe invention is resistant to lateral stretch, and to vertical compression. The stretchability and compressibilty ofthe composite sheet can be controlled in several ways. The stretchability ofthe composite sheet is affected greatly by the horizontal fibrous network to which the pile-like fibers are attached and from which the pile-like fibers protrude. An inherently non-stretchable fiber network, located within about 3 mm ofthe outer surface ofthe composite, can impart non-stretchability to the resin-fiber surface layer. For low stretchability and low compressibility, a hard resin, rather than a softer one, is preferred. Lateral stability ofthe composite sheet in any linear direction also can be achieved by the attachment of strong, substantially non-stretchable strips, films, sheets, webs, cross-warps or the like to the back surface ofthe

abrasion-resistant layer. The attachment may be made by any convenient means, such as gluing, thermal bonding or the like.

Abrasion-resistant composite sheets ofthe invention are suitable for use in many different articles. The sheets can be molded into various shaped articles, can be used as single or multiple layers, or can be attached by various means to the surface or portions ofthe surface of various shaped articles to provide the articles with abrasion resistance. For example, composite sheets ofthe invention are suited for use in shoe uppers, work gloves, automotive engine timing belts, leather-like apparel, indoor athletic protective pads, women's pocketbooks, bags, luggage, saddles, seating surfaces, etc. The more abrasion-resistant composite sheets ofthe invention are especially suited for articles that are subjected to more demanding abrasion conditions, such as toe, heel and/or sole portions of shoes, bottoms of industrial bags that are often dragged on concrete floors, bearing surfaces of interacting mechanical parts, soccer balls, heavy duty work boots, gloves, motor-cyclist apparel pads, and the like. Test Methods

The following methods and procedures are used to measure various characteristics ofthe resin-impregnated fabrics of the invention. In composite sheets ofthe invention, which have vertical pile-like fibers formed by the buckling of a nonwoven fibrous layer or by the buckling of yam segments over short intervals, inverted U-shaped loops are formed from buckled groups offibers or from buckled yams. The height H and the base B ofthe U-shaped loops of buckled groups offibers are determined from magnified (e.g., 15-20X) photomicrographs of cross-sections ofthe loops taken through the loops in a plane perpendicular to the plane ofthe fibrous layer. The data are then used to calculate an H/B ratio. A low magnification microscope with strong top and/or back lighting on the sample permit direct measurement of H and B. Usually the average loop height H is equal to the thickness ofthe contracted fibrous layer. Alternatively, the average loop height H can be measured directly with a "touch" micrometer having a 1/4-inch (0.64-cm) diameter flat cylindrical probe which applies a 10-gram load to the contacted surface. A digital micrometer, model APB- 1D, manufactured by Mitutoyo of Japan is convenient for the measurement. In addition to the above-described method, "verticality" of pile-like fibers can be determined by examination of a magnified cross-section ofthe fiber/resin layer. If loops are "crashed" or excessively "pushed down" during resin application, a relatively long flat portion ofthe inverted U is seen near the outer surface ofthe fiber/resin layer. Deflection of straight

fibers or yarns from a vertical position also is readily observable. Such severe deflection of pile fibers can occur during resin application. A 30% decrease in pile-fiber layer thickness during resin application can decrease the abrasion resistance ofthe final composite sheet, especially when the pile fiber concentration is near the lower end of the range of concentrations suitable for use in the present invention.

Stretchability, S, is determined by: (a) cutting a specimen measuring 2-cm wide by 10-cm long sample from the composite sheet; (b) marking a standard length, L ₀ , on the specimen parallel to the long dimension; (c) suspending a 1.0-kilogram weight from specimen for 2 minutes; (d) with the weight suspended, re-measuring the "standard length", the re-measured length being designated Lf, and (e) calculating the percent stretchability, %S, by the formula, %S = 100 (Lf - L ₀ ) L ₀ .

Compressibility, C, is determined by measuring the change in thickness of the surface pile-fiber/resin layer of the composite sheet (a) under no pressure, t ₀ , and (b) under a pressure of 351 kiloPascals (51 lb/in ² ), tf A thickness gage is employed which imparts a 2.5-pound (1.14-Kg) load onto the pile fiber/resin composite through a cylindrical foot of 1/4-inch (0.64-cm) diameter. Then, the percent compressibility, %C, is calculated by the formula, %C = 100 (t ₀ - tf)/^. To avoid possible errors in these determinations, caused by the presence of a compressible horizontal fibrous network within the impregnated layer and to assure that it is the characteristics ofthe pile fiber/resin layer that are being measured, the horizontal fibrous network is carefully removed by sanding until only the pile/resin layer remains.

The unit weight of a fabric or fibrous layer is measured according to ASTM Method D 3776-79. The density ofthe resin-impregnated fabric is determined from its unit weight and its measured thickness. The void fraction ofthe layer can be readily determined from the measurements ofthe over-all density ofthe layer and the weights and densities ofthe fiber and resin in the layer.

Over-feed ratio, contraction ratio and total gather are parameters reported herein which are measures of how much an initial fibrous layer contracts or gathers as a result ofthe operations to which the layer is subjected. The over-feed ratio, which applies only to the embodiments of the invention which employ a buckled nonwoven fibrous layer, is defined as the ratio ofthe initial area of a starting fibrous nonwoven layer to the area of the layer immediately up-stream of a first processing step (e.g., a stitchbonding step). Over-feed causes buckling, gathering or compression of

the nonwoven layer in the direction in which it is being fed to the operation. The contraction ratio is a measure ofthe amount of further contraction the nonwoven layer undergoes as a result ofthe specific operation to which it is subjected (e.g., stitchbonding and release of tension from yarns to which the fibrous layer was intermittently attached). The contraction ratio is defined as the area ofthe fibrous layer as it enters the specific operation divided by the area ofthe fibrous layer as it leaves the specific operation. The total gather is defined as the product ofthe over-feed and contraction ratios. The fraction of original area is the reciprocal ofthe total gather and is equivalent to the ratio of the final area of the fibrous layer to the initial area of the starting fibrous layer.

The effective pile fiber concentration is determined from the concentration offibers within the surface layer ofthe composite sheet which are in a vertical (or pile-like) position with respect to the surface that is to be exposed to abrasion. For fabrics in which the pile-like fibers are derived from buckled hard yams (e.g., as in a contracted knit fabric) the effective pile fiber concentration is the weight ofthe hard yams in a unit of area divided by the thickness ofthe layer. Similarly, if the pile-like yams are provided by buckled yams that had been loosely wrapped around a stretched elastic yam that had then been permitted to contract, the total weight of the wrapping yams are included in the calculating the concentration ofthe pile ya s, but the weight ofthe elastic core is not included. For composite sheets in which the pile-like yams are formed from a contracted-and-buckled nonwoven fibrous layer, only 50% ofthe weight ofthe nonwoven fibrous layer is included in the calculation ofthe pile-fiber concentration. The present inventor found empirically, that the abrasion-resistance-versus-pile- parameter data for composites ofthe invention correlate much better when only half the weight ofthe buckled nonwoven fibrous layer is employed rather than the full weight said layer. This reflects the fact that pile-like fibers formed from buckled hard yams and tufted pile fibers, for example, are more effective in providing abrasion resistance to the composite sheet than are pile-like fibers formed from contracted-and-buckled nonwoven fibrous layer. Thus, the effective pile fiber concentration, c _e f , = 10 ^-4 kw/t, wherein k is 0.5 for pile-like fibers provided by buckled nonwoven fibrous layers and 1.0 for buckled yams or tufts, w is the unit weight of the pile-like yams in grams per square meter and t is the thickness ofthe surface layer in centimeters.

To determine the abrasion resistance of samples a Wyzenbeek "Precision Wear Test Meter", manufactured by J. K. Technologies Inc. of

Kankakee, Illinois, is employed with a 40 grit emery cloth wrapped around the oscillating drum ofthe tester. The drum is oscillated back and forth across the face ofthe sample at 90 cycles per minute under a load of six pounds (2.7 Kg). The test is conducted in accordance with the general procedures ofASTM D 4157-82. The thickness ofthe sample is measured with the aforementioned thickness gage before and after a given number of abrasion cycles to determine the wear rate in microns lost per 1,000 cycles. To provide adequate wear resistance to the composite sheets ofthe invention, a wear rate of no more than 50 microns/1000 cycles is considered to be satisfactory.

EXAMPLES In the following Examples, the fabrication and abrasion resistance of various composite sheets ofthe invention are illustrated and compared to similar composite sheets that are outside the invention. The composite sheets ofthe invention are much more abrasion resistant than are the comparison composite sheets. Samples ofthe invention are designated with Arabic numerals and comparison samples with upper case letters. Conventional wa -knitting nomenclature is used to describe the particular repeating stitch patterns that were employed to prepare the various knit or stitchbonded fabrics ofthe Examples. A table accompanies each example and records fabrication details, weights, composition and characteristics of each composite sheet, as well as the abrasion performance ofthe sheet.

In the examples, fabrics that were made with elastic yams were in sequence (1) removed from the fabric forming machine, (2) allowed to achieve an initial contraction, (3) subjected to a "boil-off" treatment by being immersed in boiling water (100°C) for 1-2 minutes, (4) dried and then (5) heat set on a tenter frame for 1-1.5 minutes at 380°F (193 °C). The particular amounts of stretching in the longitudinal and transverse directions during the heat setting were used control the final amount of contraction experienced by the fabric.

Two different polyurethane resins were used for impregnating the pile-like layer ofthe composite samples. "ZAR", a clear polyurethane finish sold by United Gilsonite Laboratories of Scranton, Pennsylvania, designated herein as "PU-1", was used for the samples of Examples 1 and 3. PU-2. a softer polyurethane resin, sold by K. J. Quinn & Co., Inc. of Seabrook, New Hampshire was used for all the remaining samples. PU-2 is a two-part formulation that was mixed, applied to the pile fibers and then cured. The samples were resin-impregnated by conventional dipping techniques. The applied resin was smoothed with a doctor blade and dried and/or cured with

the pile fibers facing downward for at least 12 hours in a hot-air oven. Oven temperature was maintained at 65°C for fabrics impregnated with PU-1 and at 95°C for fabrics impregnated with PU-2. The compressibility, density and 40-grit abrasion wear rates (in microns/1000 cycles) and Shore A hardness of a 5-mm-thick layer of each resin containing no fibers were as follows:

Shore A % Com- Density Abrasion Resin hardness pressibility g/cm-i wear rate PU-1 70 0 1.1 900

PU-2 53 5 1.0 4,500 Example 1

This example compares two samples of resin-impregnated composite sheets ofthe invention with a sample that is outside the invention. In each sample, pile-like fibers are formed of Kevlar® aramid fibers (sold by E. I. du Pont de Nemours & Co.). In Sample 1, the pile-like fibers are formed by the buckling of Kevlar® yams in a knit fabric. In Sample 2, the pile-like fibers are formed from buckled Kevlar® stitching yams and a buckled nonwoven fibrous substrate of Kevlar® in a stitchbonded fabric. In comparison Sample A, which is outside the invention because of its low effective pile-fiber concentration and low pile parameter, the pile-like groups offibers are formed only from a buckled nonwoven layer of Kevlar® fibers. Samples 1 and 2 the invention are shown have between about 3 to 5 times the abrasion resistance of comparison Sample A.

The starting fabric for the composite sheet of Sample 1 was a two bar knit fabric that was prepared with a "Liba" machine operating at 10 gauge (10 rows per inch or 4 per cm) and 22 courses per inch (8.7 per cm). The back bar was threaded with 400 den (440 dtex) filament Kevlar®-29 yarn foπriing a repeating pattern of 1-0,4-5 stitches. The front bar was threaded with a combination yam that consisted of a 280-den (320-dtex) Lycra® spandex elastic core, around which was tightly wound, at about 7 turns/inch (2.8/cm), a 70-den (78-dtex) 34-filament textured polyester yam. Upon removal ofthe knit fabric from the LIBA machine, the as-knit fabric weighed 159 g/m ² . The as-knit fabric was then boiled-off and heat set. As a result the fabric contracted by a factor of 2.6 and increased its weight to 413 g/m ² . The back bar yam buckled to form groups of inverted U-shaped pile-like fibers. The tight wrapping of the combination yarn of the front bar yarn did not contribute to the formation of pile-like groups offibers. Then, the fabric was impregnated with polyurethane resin PU-1 and dried and cured in the hot-air oven. Characteristics ofthe resultant composite sheet, Sample 1, are recorded in Table I below.

The starting fabric for the composite sheet of Sample 2 was a two-bar stitchbonded fabric that was prepared with a 140-inch (3.6-meter) wide, two- bar "Liba" machine adapted for stitchbonding a nonwoven fibrous layer. Each bar was threaded to 14 gage (14 rows/inch or 5.5/cm) and inserted 9 stitches per inch (3.5/cm) in each row. A 34-g/m ² Type Z-l 1 Sontara® spunlaced fibrous substrate (made by E. I. du Pont de Nemours & Co.) of Kevlar®29 aramid fibers of 1.5 den (1.7 dtex) per filament and 2.2 cm length was fed to the LIBA with 47% overfeed. The back-bar and front-bar stitching threads were the same as those used for Sample 1, but formed opposing 2-course Atlas stitch pattems. Upon removal ofthe stitchbonded fabric from the LIBA, the fabric contracted in area and increased its weight to 184 g/m ² . The stitchbonded fabric was then boiled off, heat set and resin impregnated in the same manner as for Sample 1, except that the area of Sample 2 in the boiling and heat setting contracted by a factor of 2.9 and the nonwoven fibrous layer, which had been overfed to the stitchbonding step, experienced a total gather of 4.3. Further details ofthe fabrication and characteristics of Sample 2 are listed in Table I below.

A comparison composite sheet, Sample A, was prepared from the same type and weight of Kevlar® spunlaced starting fabric as was used for Sample 2. A one-bar stitchbonding machine, threaded at 12 gauge (12 needles /inch or 4.7/cm) and inserting 14 stitches/inch (5.5/cm), was used for Comparison Sample A. The nonwoven fibrous layer was overfed 25% and stitched with the same combination yam as was used in preparing Samples 1 and 2. A 1-0, 2-3 repeating stitch pattern was employed. Sample A was resin impregnated, boiled off and heat set in the same way as Samples 1 and 2. Details of comparison Sample A are summarized in Table I.

The data of Table I clearly demonstrate the superior abrasion resistance of composite sheet Samples 1 and 2 over comparison composite sheet of comparison Sample A. Samples 1 and 2 respectively were 2.7 and 3.1 times as abrasive resistant as the comparison Sample A.

Table I (Example 1) Sample Identification 1 2 1 Starting Materials Nonwoven wt., g/m ² 0 34 34 Over-feed ratio na 1.47 1.25

Hard yarns wt., g/m ² 135 104 0 Contractibles wt., g/m ² 24 30 44 Total wt., g/m ² 159 184 87 Gathering Contracted wt., g/m ² 413 537 361 Contraction ratio 2.6 2.9 4.15 Nonwoven wt., g/m ² 0 145 176 Hard yarn wt., g/m ² 351 302 0 Nonwoven total gather na 4.3 5.2 % of original area 38 35 24 Resin application Resin wt., g/m ² 306 566 510 % pile height loss 0 0 0 Surface layer characteristics Total wt. , g/m ² 657 1013 686 Thickness, mm 1.1 2.0 1.4 Density, g/cm ³ 0.59 0.51 0.49 Pile fiber weight, g/m ² 351 445 176 Pile fiber cone, g/cm ³ 0.32 0.23 0.13 Effective pile cone, g/cm ³ 0.32 0.19 0.065 Pile Parameter, P, g/cm ³ 0.44 0.32 0.18 Loop base, B, mm 0.7 0.5 0.4 Loop H/B ratio 3.6 5.0 3.5 Wt. % resin 46 56 74 % voids 51 58 59

% stretchability 10 10 10 % compressibility 10 10 5 40-grit abrasion resistance Test duration, 10 ³ cycles >5 >5 >5 Wear, microns/10 ³ cycles 23 36 110 % normalized wear* 21 33 100 Note; na = not applicable; * = Normalized to Sample A.

Example 2

In this example, several samples of composite sheets ofthe invention and comparison samples similar to those of Example 1 are prepared, but with softer polyurethane resin PU-2 replacing polyurethane resin PU-1 of Example 1. Each sample and comparison sample has groups of pile-like fibers formed from Kevlar® aramid fibers.

Sample 3, Sample 4 and comparison Sample E, respectively, contain the same fabrics as Samples 1 and 2 and comparison Sample A of Example 1. The composite sheet of Sample 3 includes a contracted fabric that had been knit with an elastic combination yam on one bar and a non-elastic yam on the second bar. The composite sheet of Sample 4 includes a contracted fabric that had been prepared by stitchbonding a fibrous layer with one bar of elastic composite yam and one bar of non-elastic yam. Sample E includes a contracted fibrous layer that had been prepared by stitchbonding the layer with a single bar of elastic combination yam. Additional composite sheet comparison Samples B and C respectively are made with the same starting fabrics as Samples 3 and 4, but with different amounts of resin applied. Further details on the fabrication, characteristics and abrasion performance ofthe resultant composite sheet samples are summarized in Table H Table π clearly shows the abrasion-resistance advantage of composite sheets made with pile parameters of high value . See for example, the abrasion-test results for Sample 3 versus comparison Sample B and Sample 4 versus comparison Sample C. Also, the abrasion wear results for these samples, and Sample E, versus ofthe corresponding samples in Example 1 show that as long as the resins immobilize the fibers and provide low compressibility and stretchability to the layer, increasing resin hardness apparently does not increase abrasion resistance ofthe composites ofthe invention. Increases in pile-fiber concentration and pile parameter are more effective .

In addition to the above-described composite sheets containing fabrics with pile-like groups of Kevlar® aramid fibers, two more composite sheets, Sample 5 and comparison Sample D, are prepared. Starting fabrics for these two samples are made by tufting a 1,000 den (1,100 dtex) Kevlar®-29 yam into a 119-g/m ² lightly bonded Reemay® spunbonded polyester nonwoven fabric (made by Reemay, Inc., of Old Hickory, Term.) at 16 tufts per inch (5.1/cm) at 14 gauge (14 tufting needles per inch or 5.5/cm). The tufted fabric was stretched 20%, with an accompanying necking-in of about 40%, to contract the area ofthe fabric by a factor of 2. Further details of the composite sheet construction and performance are summarized in Table II.

Table II (Example ² )

Sample Identification 3 B 4 I

Starting Materials Nonwoven wt. , g/m ² 0 0 34 34 119 119 34 Over-feed ratio na na 1.47 1.47 na na 1.25 Hard Yarns wt. , g/m ² 135 135 104 104 205 205 0 Contractibles wt.,g/m ² 24 24 30 30 0 0 44 Total wt. , g/m ² 159 159 184 184 314 314 87

Gathering Contracted wt. , g/m ² 413 413 537 537 636 636 361 Contraction ratio 2.6 2.6 2.9 2.9 2.0 2.0 4.15 Nonwoven wt. , g/m ² 0 0 145 145 0 0 176 Hard Yarns wt. , g/m ² 351 351 302 302 415 415 na Nonwoven total gather na na 4.3 4.3 na na 5.2 % of original area 38 38 35 35 49 49 24

Resin application Resin wt. , g/m ² 646 102 1263 192 1680 560 1020 % pile height loss 0 0 0 0 0 0 0

Surface layer characteristics

Total wt. , g/m ² 9 99977 453 1710 639 2095 975 1196 Thickness, mm 1 1..11 1.1 2.0 2.0 2.6 2.6 1.1 Density, g/cm ³ 0.91 0.41 0.86 0.32 0.81 0.38 1.08 Pile fiber weight g/m ² 351 351 447 447 415 415 176 Pile fiber cone g/cm ³ 0.32 0.32 0.22 0.22 0.16 0.16 0.16 Eff. pile cone, g/cm ³ 0.32 0.32 0.18 0.18 0.16 0.16 0.08 Pile Parameter, g/cm ³ 0.54 0.36 0.40 0.24 0.36 0.25 0.29 Loop base, B, mm 0.7 0.7 0.5 0.5 0.8 0.8 0.4 Loop H/B ratio 3 3..66 3.6 5.0 5.0 3.8 3.8 3.5 Wt. % resin 6 644 22 74 30 80 57 85 % voids 2 244 66 28 72 32 68 10

% stretchability 0 0 20 0 15 0 0 0 % compressibility 0 0 15 0 15 0 0 0 40-grit abrasion resistance Test duration, 10 ³ cycles 13 1.5 25 7.9 25 6 13.2 Wear, microns/10 ³ cycles 18 80 25 66 30 60 50 % Normalized wear* 36 160 50 132 60 120 100

Notes; * = normalized to Sample E; na = not applicable;

Example 3

This example further illustrates the strong effects of total gather and of concentration of pile-like fibers on the abrasion resistance of composite sheets ofthe invention. Composite sheet Samples 6, 7 and 8 and comparison Sample F each contain a layer in which the pile-like fibers are derived from both a buckled fibrous nonwoven layer and buckled non-elastic stitching yams. In comparison Sample G, the buckled non-elastic yams are not present.

To prepare the starting fabric for each sample, a fibrous layer of 26- g/n_2 Sontara® 8017 spunlaced fabric of nonbonded polyester fibers was overfed to a two-bar "Liba" stitchbonding machine. The front bar ofthe Liba formed a repeating pattern of 1-0,2-3 stitches with combination yam that was a 280-den (311-dtex) Lycra® spandex elastic core, tightly wound 9 turns/inch (3.5/cm) with 70-den (78-dtex) 34 filament polyester yam. Except for the fabric of comparison Sample G, which employed no back-bar, the back bar formed a repeating pattern of 3-4, 1-0 stitches with a 210 den (233 dtex) 34 filament high tenacity Type 62 Dacron® polyester yam. Lycra® and Dacron® are each sold by DuPont. The Liba, a 14-gauge (14 needles per inch or 5.5/cm) machine, inserted 14 courses per inch (5.5/cm). Different amounts of tension were imposed on the combination yam used to prepare each sample with a different amount of contraction when the fabric was removed from the Liba. The contraction ofthe combination yam caused a layer of pile-like fibers to develop. The pile-like fibers were formed by contraction and buckling ofthe nonwoven fibrous layer and buckling ofthe back-bar non-elastic stitching yams, which accompanied the contraction of the combination yam. Then after boil-off the contracted fabrics were heat treated on the tenter frame to set the final dimensions ofthe fabric. Thereafter, the fabric samples were impregnated with polyurethane resin PU- 1, the same resin as was used in Example 1, to immobilize the pile-like fibers. The impregnated samples were then dried to form the composite sheet sample. Sample details and abrasion test results are summarized in Table HI, below.

Table III (Example 3)

Sample Identification F Starting Materials Nonwoven wt., g/m ² 26 26 26 26 26 Over-feed ratio 1.2 1.2 1.2 1.2 1.3 Hard yarns, g/m ² 71 71 71 71 0 Contractibles wt., g/m ² 31 31 31 31 32 Total wt., g/m ² 133 133 133 133 66 Gathering Contracted wt., g/m ² 578 785 918 238 544 Contraction ratio 4.4 5.9 6.9 1.8 8.2 Nonwoven wt., g/m ² 137 183 215 56 277 Hard yarn wt. g/m ² 312 418 490 128 0 Nonwoven total gather 5.3 7.1 8.3 2.2 10.7 % of original area 19 14 15 45 12 Resin application Resin wt., g/m ² 1043 884 655 499 986 % Pile height loss 10 9 15 8 0 Surface layer characteristics Total wt., g/m ² 1489 1485 1360 683 1263 Thickness, mm 1.8 2.0 1.8 1.4 1.6 Density, g/cm ³ 0.83 0.74 0.75 0.50 0.79 Pile fiber weight, g/m ² 449 601 705 184 277 Pile fiber cone, g/cm ³ 0.25 0.30 0.39 0.13 0.17 Eff. pile cone, g/cm ³ 0.21 0.25 0.33 0.11 0.085

Pile Parameter, p, g/cm ³ 0.41 0.44 0.49 0.23 0.26 Loop base, B, mm 0.44 0.30 0.25 1.0 0.30 Loop H/B ratio 4.1 6.6 7.2 1.4 5.3 % resin 70 60 48 73 78 % voids 31 38 38 58 34

% stretchability 5 5 10 5 0 % compressibility 10 5 10 15 0 40-grit abrasion resistance Test duration, 10 ³ cycles 23 31 25 2.3 16 Wear, microns/10 ³ cycles 32 13 12 80 92 % normalized wear 35 14 13 87 100 Notes; *Normalized to Sample G

Example 4

In this example composite sheet samples are prepared from contracted fabric that was single-bar knit with an elastic combination yarn having a loosely wound non-elastic wrapping yam. The pile-like fibers are formed from the wrapping yam when the combination yam contracts.

Samples 9, 10 and 11 and comparison Sample H were each knit with a one-bar "Liba" machine forming a repeating pattern of 1-0,2-3 stitches with a combination yam that had a 280-den (311-dtex) Lycra® spandex elastic core loosely wrapped at about 1.5 turns/inch (0.6/cm) with a 70-den (78-dtex) 34- filament textured polyester yam. Each sample was made with the Liba operating at 14 courses/inch (5.5/cm) and the needle bar at 20 gage (i.e., 20 needles per inch or 7.9/cm), except Sample 11 which was made at 10 gage. The fabrics for each sample were knit with the combination yam under a different tension. Upon removal ofthe fabric from the knitting machine and tension in the combination yam was released and the as-knit area ofthe fabric contracted by a factor 11.5 to 14 times, with an accompanying buckling ofthe loosely wound wrapping yam to form the pile-like layer. Polyurethane resin, PU-2, as was used in Example 2, was applied to the pile¬ like fibers of each sample. Further details of sample fabrication and abrasion performance are listed below in Table IV, which also includes Sample G of Example 3 for further comparison purposes.

The abrasion test results listed in Table IV show that high pile concentrations and high pile parameters and accompanying high abrasion resistance can be obtained with fabrics having pile-like groups offibers provided by only buckled yams. The results also demonstrate the importance of avoiding excessive stretchability in the composite sheet. Note that the pile-like fiber/resin surface layer ofthe composite sheet of comparison Sample H, having a resin content of only 25 weight %, was readily stretchable and exhibited an 11 to 18 times greater abrasion rate than did composite sheet Samples 9-11 ofthe invention. Unless sufficient resin is incorporated in the fiber/resin layer, even if other characteristics ofthe layer are in accordance with the invention, the composite sheet will still lack resistance to stretching and abrasion. The abrasion-resistance data and other details ofthe samples are summarized in Table IV and clearly demonstrate the superior abrasion resistance ofthe composite sheets ofthe invention over the comparison samples. The data again show that a resin-impregnated pile¬ like fiber layer meeting the requirements ofthe present invention provides a composite sheet with highly effective abrasion resistance.

Table IV (Example 4)

Sample Identification 10 11 H Starting Materials

Nonwoven wt., g/m ² 0 0 0 0 26 Over-feed ratio na na na na 1.3.

Hard yarns, g/m ² 0 0 0 0 0

Wrap yarn weight, g/m ² 24 24 21 24 0

Contractibles wt., g/m ² 22 22 20 22 32

Total wt., g/m ² 46 46 41 46 66 Gathering

Contracted w ., g/m ² 557 557 476 557 544

Contraction ratio 11.5 11.5 14.0 11.5 8.2

Nonwoven wt., g/m ² 0 0 0 0 277

Nonwoven total gather na na na na 10.7 Wrap yarn wt. g/m ² 276 276 294 276 0

% of original area 8.7 8.7 7.8 8.7 12 Resin application

Resin wt., g/m ² 910 1009 987 94 986

% Pile height loss 17 17 13 17 0 Surface layer characteristics

Total wt., g/m ² 1186 1285 1281 370 1263

Thickness, mm 1.5 1.5 1.4 1.5 1.6

Density, g/cm ³ 0.79 0.86 0.91 0.25 0.79

Pile fiber weight, g/m ² 276 276 294 276 277 Pile fiber cone, g/cm ³ 0.18 0.18 0.21 0.18 0.17

Eff. pile cone, g/cm ³ 0.18 0.18 0.21 0.18 0.085

Pile Parameter, p, g/cm ³ 0.38 0.38 0.44 0.38 0.26

Loop base, B, mm 0.12 0.12 0.10 0.12 0.30

Loop H/B ratio 12.5 12.5 14.0 12.5 5.3 % resin 76 78 77 25 78

% voids 35 28 24 79 34

% stretchability 10 5 5 80 o

% compressibility 10 5 10 15 0 40-grit abrasion resistance Test duration, 10 ³ cycles 20 18 23 0.8 16

Wear, microns/10 ³ cycles 30 40 43 450 92

% normalized wear 33 43 47 489 100 Notes; *Normalized to Sample G

Example 5

In this example, resin-impregnated composite sheets were prepared from contracted two-bar warp-knit fabrics. The starting fabric for each of Samples 12-15 and Comparison Sample I was knit on a two-bar Liba machine, with one bar threaded with an elastic combination yam having a loosely wound non-elastic wrapping yam and the second bar threaded with non-elastic textile yam. The front bar formed repeating pattems of 1-0,2-3 stitches with a 280-den (311-dtex) Lycra® spandex around which was loosely wound, at one turn/inch (0.4/cm), a 70-den (78-dtex) 34-filament textured polyester yam. The back bar formed a repeating pattem of 3-4, 1-0 stitches with a 150-denier (167-dtex) conventional polyester textile yam. Each sample was made with the Liba operating at 14 courses/inch (5.5/cm), except Sample 12 and comparison Sample J, which were each made with 22 courses per inch (8.7/cm) and with the machine threaded at 20 gauge, and Sample 13 which was made with the machine threaded at 10 gage. Tension on the combination yam used to prepare the samples was adjusted so that when the knit fabric was removed from the Liba, boiled-off and heat set, the as-knit area ofthe fabric contracted by a factor of 1.9 to 7. The contraction ofthe fabric was accompanied by contraction and buckling ofthe loosely wound wrapping yam and buckling ofthe second bar yam. Polyurethane PU-2 (same as in Example 2) was applied to each sample.

Further fabrication and abrasion performance details ofthe samples are summarized in Table V below, in which Sample G of Example 3 is also included for further comparison. The abrasion wear results show that Comparison Samples G and I, exhibited between about 2.5 to about 20 times as much abrasion wear as did Samples 12-15 ofthe invention. The pile-like fibers ofthe fabric of Sample I apparently were deflected from the vertical position during application of resin, as indicated by the 31% loss in surface layer height and the insufficient contraction ratio. The results summarized in Table V showed that contraction ofthe fabric for Samples 12-15 ofthe invention resulted in substantially vertical pile-like fibers being formed. The pile-like fibers were derived from both the contracted wrapping yam and from the buckled second bar yam. In contrast, the fabrics of comparison Sample I did not provide a satisfactory pile-like layer and was not adequately resin impregnated and accordingly exhibited much inferior abrasion resistance. Sample G, despite it very high resin density did not provide high abrasion resistance because ofthe low effective pile density and low pile parameter.

Table V (Example 5)

Sample Identification 12 13 14 15

Starting Materials Nonwoven wt., g/m ² 0 0 0 0 0 26 Over-feed ratio na na na na na 1.3 Hard yarns wt., g/m ² 92 32 32 32 92 0 Wrap yarns wt,, g/m ² 33 23 23 23 26 0 Contractibles wt., g/m ² 28 22 22 22 26 32 Total wt. , g/m ² 153 77 77 77 154 66

Gathering Contracted wt., g/m ² 554 540 870 870 299 544 Contraction ratio 3.6 7.0 11.3 11.3 1.9 8.2 Nonwoven wt., g/m ² 0 0 0 0 0 277 Hard yarn wt., g/m ² 331 224 361 361 175 0 Wrap yarn wt., g/m ² 118 161 260 260 49 0 Nonwoven total gather na na na na na 10.7 % of original area 28 14 9 9 53 12

Resin application Resin wt. , g/m ² 826 1098 380 920 408 986 % pile height loss 16 4 15 0 31 0

Surface layer characteristics

Total wt. , g/m ² 1275 1483 1001 1541 632 1263.

Thickness, mm 1.6 2.4 1.7 1.9 1.3 1.6

Density, g/cm ³ 0.65 0.74 0.59 0.81 0.49 0.79

Pile fiber weight, g/m ² 449 385 621 621 224 277

Pile fiber cone , g/cm ³ 0.28 0.16 0.37 0.33 0.17 0.17

Eff. pile cone, g/cm ³ 0.28 0.16 0.37 0.33 0.17 0.085

Pile Parameter, g/cm ³ 0.43 0.34 0.47 0.51 0.28 0.26

Loop base, B, mm 0.3 0.3 0.2 0.2 0.3 0.3

Loop H/B ratio 5.3 8.9 8.5 8.5 4.3 5.3

Wt. % resin 65 74 38 60 65 78

% voids 46 38 51 33 59 34

% stretchability 10 5 10 0 5 0

% compressibility 10 5 10 0 25 0 40-grit abrasion resistance

Test duration, 10 ³ cycles >25 >25 >25 >25 4 16

Wear, microns/10 ³ cycles 28 40 8 5 100 92

% normalized wear* 30 40 9 5 109 100 Notes; na = not applicable; = normalized to sample G

Example 6

In this example, composite sheet samples were prepared which had resin-impregnated fork-needled pile-like fibrous layers.

The starting fabrics for Sample J and 16 were prepared from a 272- g/m ² air-laid web of 1.5-den (1.7-dtex), 3-inch (7.6-cm) long Type 54 Dacron® polyester fiber (sold by E. I. du Pont de Nemours & Co.) that was placed upon and fork needled into a 119-g/m ² Reemay® spunbonded polyester fabric with a Dilo fork needier. The needier was a 14 gauge (14 fork needles per inch or 5.5/cm) machine that made about 130 insertions square inch (20/cm ² ). Loops of fiber were formed on the surface ofthe Reemay® that was opposite to the side the needles entered the layer. The loops projected about 2.5 mm above the surface ofthe Reemay®. Comparison Sample J was boiled off, heat set, impregnated with polyurethane, PU-2 and then dried. Sample 16 was treated the same way, except that prior to the heat-setting, resin-impregnation and drying, Sample 16 was stretched longitudinally by 30% with a corresponding decrease to about 37% of its original width to provide the sample with a contraction in area by a factor of 2.1. Further fabrication details, characteristics ofthe samples and abrasion wear test results are summarized in Table VI. The summary also includes data for comparison Sample G of Example 3 for further comparison purposes.

Table VI clearly shows that composite sheet Sample 16 ofthe invention, containing a pile-like fiber layer prepared by fork-needling and contraction, is 6.5 times as abrasion resistant as a similarly prepared composite sheet having fork-needled pile-like layer that was not contracted. Also, the composite sheet of Sample 16 was 4.5 times as abrasion resistant as the composite sheet of comparison Sample G of Example 3. These results, emphasize, as was shown in the preceding examples as well, that the pile parameter ofthe resin-impregnated pile-fiber layer is very important in providing a composite sheet with abrasion resistance. Sample 16 ofthe invention had a pile parameter of 0.37, while comparison Sample J and G respectively had pile parameters of 0.22 and 0.26.

TABLE VI (Example 6) Sample Identification J ___ G Starting Materials Nonwoven wt. , g/m ² 272 272 26 Over-feed ratio na na 1.3

Contractibles wt., g/m ² 119 119 32 Total wt., g/m ² 391 391 66 Gathering

Contracted wt., g/m ² 391 821 544 Contraction ratio 1.0 2.1 8.2 Nonwoven wt., g/m ² 272 571 277 Nonwoven total gather 1.0 2.1 10.7 % of original area 100 48 12 Resin application Resin wt., g/m ² 720 930 986

% pile height loss 0 0 0 Surface layer characteristics Total wt. , g/m ² 992 1501 1263 Thickness, mm 2.4 2.5 1.6 Density, g/cm ³ 0.41 0.60 0.79

Pile fiber weight, g/m ² 272 571 277 Pile fiber cone, g/cm ³ 0.11 0.23 0.17 Eff. pile cone, g/cm ³ 0.11 0.23 0.085 Pile Parameter, g/cm ³ 0.22 0.37 0.26 Loop base, B, mm 0.8 0.4 0.3 Loop H/B ratio 3.0 6.0 5.3 Wt. % resin 73 62 78 % voids 65 50 34

% stretchability 10 0 0 % compressibility 15 0 0

40-grit abrasion resistance Test duration, 10 ³ cycles 5 15 16 Wear, microns/10 ³ cycles 130 20 92 % normalized wear* 141 22 100 Notes; na = not applicable; * = normalized to sample G

Example 7

In this example, two composite sheet samples were prepared with resin-impregnated velour fabrics providing the pile-like fibers for the resin/fiber layers. The starting fabrics for Sample 17 and 18 were prepared on a 130- inch (3.3-meter) wide, 70 gauge (70 needles/inch or 27.6/cm), three-bar warp-knitting machine which formed 64 course per inch (25.2/cm). A type KS3P machine, manufactured by Karl Mayer of Frankfurt, Germany, was used. The first bar was threaded with a flat (i.e., not textured) 70-den (77- dtex), 24-filament polyester yam and formed a repeating pattem of 1-0, 1-1, 2-1 stitches. The second bar was threaded with a flat, 100-den (110-dtex) 40-filament polyester yam and formed a repeating pattem of 1-0, 0-0, 1-1 stitches. The third bar was threaded with a flat (i.e., not textured) 70-den (77-dtex), 24-filament polyester yam and formed a repeating pattem of 1-0, 0-0, 1-1 stitches. The velour fabric that was formed by the machine had a knit backing (i.e., base) layer that weighed 347 g/m ² and layer of looped yams that measured 1.5 mm high, weighed 231 g/m ² , had a effective pile fiber concentration of 0.18 g/cn_3 .

Sample 17, in the as-knitted condition, was impregnated with polyurethane resin PU-2. After resin curing, the sample had a pile parameter of 0.31 g/cm^. Sample 18, in the as-knit condition, was heat set at a temperature of 375°F (191°C) to stabilize the fabric dimensions. Then the loops ofthe heat-set Sample 18 were sheared to provide a pile thickness of 1.2 mm. Sample 18 also was impregnated with polyurethane resin PU-2, which after curing also provided the sample with a pile parameter of 0.31 g/cm**- ^* '.

As shown in Table VII, Samples 17 and 18 were composite sheets of the invention that performed quite well in the 40-grit abrasion wear test. Comparison Sample G of Example 3 was also included in the Table VH for further comparison. As compared to the composite sheets of Samples 17 and 18 of the invention with their resin-impregnated velour fabrics, the composite sheet of comparison Sample G with its resin-impregnated contracted-and-buckled nonwoven fibrous layer abraded 2.2 to 2.4 times as rapidly as Samples 17 and 18.

TABLE VII (Example 7) Sample Identification ,12 18

Starting Materials

Base layer weight, g/m ² 347 347 na Pile weight, g/m ² 231 231 na Pile height, mm 1.5 1.5 na After heat setting & shearing" ¹ " Base layer weight, g/m ² 347 347 na Pile weight, g/m ² 231 210 na Pile height, mm 1.5 1.4 na Resin application Resin weight, g/m ² 452 422 986 % pile height loss 15 15 0 Surface layer characteristics Total weight, g/m ² 683 682 1263 Thickness, mm 1.3 1.2 1.6 Density, g/cm ³ 0.53 0.53 0.79 Pile fiber weight, g/m ² 231 210 277 Pile fiber cone, g/cm ³ 0.18 0.18 0.17 Eff. pile cone, g/cm ³ 0.18 0.18 0.085 Pile Parameter, g/cm ³ 0.31 0.31 0.26 Base, B, mm 0.4 0.4 0.3 H/B ratio 3.3 3.0 5.3 Wt. % resin 66 67 78 % voids 45 44 34

% stretchability 10 0 0 % compressibility 15 15 0 40-grit abrasion resistance Test duration, 10 ³ cycles 16 15 16 Wear, microns/10 ³ cycles 41 38 92 % normalized wear* 45 41 100 Notes; ⁺ Only Sample 18 was sheared. "na" means not applicable, See Table III (Example 3) for further details about Sample G. * % wear normalized to Sample G.