Clipper seams are used to join the ends of belts or other industrial fabrics such as papermachine clothing and filtration media to form endless structures for passing continuously about drums and rollers in related machinery. Fabrics of woven material may be seamed by extending machine direction yarns beyond the last cross-direction yarns and weaving the ends back into the fabric to produce a series of loops on each fabric end which are interdigitatable to form a tunnel through which a pintle wire can be passed to join the ends of the fabric. In the case of nonwoven or heavily coated or impregnated fabrics however, for example as often used in dryer fabrics and filter fabrics, extendable yarns are either absent or unusable due to the impregnation of the fabric. In these cases, a seaming tunnel is formed by an array of metal clips in the form of U-shaped wire loops one or both ends of which are embedded in the material of the end region of the fabric, and the loops are interdigitatable with those on the other end of the fabric to form a seaming tunnel for passage of a pintle wire. An embodiment of such a clipper seam is shown in US Patent 3,576, 055 (GISBOURNE).

When such an arrangement is used in a papermachine fabric, such as a dryer belt, the seam is often protected by a flap of the fabric, or buried in a multilayer structure of the fabric, or the belt is thicker than the depth of the wire loops. However, filter belts or pulp dewatering belts are much less bulky than papermachine belts, and the metal loops stand proud of the belt. In almost all filtering operations it is necessary to remove the solid cake from the belt surface by using a scraper blade set just above the belt surface. The scraper blade as a result has a tendency

to catch the metal hooks every time the seam passes below the scraper leading to wear of the meal hooks, which causes the latter to weaken and fail unfastening the belt, and also the constant snagging of the metal hooks on the blade pulls at the point where the hooks puncture the belt.

Failure in the latter case is due to formation of a tear or hole in the belt.

A recent modification of this design is to apply a thickness of an adhesive material to the fabric edge regions on the side or sides from which the hooks engage into the fabric of the belt. The adhesive sits proud of the surface and wears sacrificially instead of the metal hooks, and tears and holes caused by the metal hooks pulling at the belt are reduced because the adhesive immobilises the hooks within the belt so that tension forces are distributed over as great an area as possible.

It has been found however that despite the improvements which are offered by the use of adhesive in this way the adhesive layer or body does not have permanent resistance to abrasion caused by the scraper blade. In consequence the adhesive eventually wears down exposing the hooks, or delaminates from the belt in response to repeated scraping.

The use of harder more resistant adhesives is desirable but not possible as these are increasingly brittle with increasing hardness and will facture very easily during use under flexure stress, as when passing around guide and drive rollers. Such fractured adhesive will wear or fall off from the belt very quickly, possibly contaminating the product.

Accordingly, an object of the invention is to provide a clipper seam construction which will achieve extended belt life and substantially reduce or overcome the problems noted above.

The present invention provides a clipper seam construction comprising a row of loop defining members provided on each end of a

fabric which can be interdigitated to form a tunnel for a pintle joining member, each of the loop defining members having at least one hook for engaging in the respective end of the fabric, and each end of the fabric having an adhesive layer covering the hooks of the loop defining members, characterised in that the adhesive layer comprises and adhesive material which incorporates a hard material in the adhesive material.

The term"hard"in the context of this description means a material which is at least harder than the matrix in which it is embedded. The hard material may comprise particles which are preferably of stainless steel, but can be any of HASTELLOY (Trade Mark) metal, or alumina, zircon, silica (e. g. as sand) or glass balatini or microspheres.

The mean particle size may about 100, um, and all the particles preferably pass through a 300, um mesh, the particles being sieved to remove coarse material.

The hard particulate material may comprise up to 50% by weight of the total adhesive layer. A more typical proportion of the hard particulate material may be in the order of 20% by weight of the adhesive material.

The hard material may comprise a fibrous material such a metal fibres, e. g. steel or HASTELLOY (Trade Mark), ceramic fibres crystalline or amorphous mineral fibres, mineral wool, or glass fibres. Ceramic fibres can be alumina, alumino-silicate, calcium silicates, or other silicate materials. A mix of fibres and particulate material may be used as the hard material.

Preferably, a thixotropic agent may be added to modify the viscosity of the adhesive to prevent settling of the hard particles from the adhesive due to gravity. The thixotropic agent may be a powdered magnesium silicate.

The loop defining members may be in the form of a generally U- shaped wire or rod members, the ends of the limbs of the U-shape being turned inwardly to form barbs which engage in the fabric, as shown in US 3576055 noted above, or may have a generally P-shaped form, with one limb of the member turned in to form a loop, and the other provided with an inturned barbed end to engage in the fabric. In the former case, both sides of the fabric may be provided with an adhesive layer to cover the limbs of the loop defining members, at lest the adhesive layer on the upstream (filter cake accumulating) side of the fabric having a hard particulate material therein, preferably both. In the second case, the adhesive layer with hard particulate material therein is provided on the side of the fabric overlain by the longer barbed limb of the P-shaped loop defining member.

The loop defining members may be arranged with alternating longer and shorter limbed forms, to reduce stress acting on a single zone of the fabric and thus reduce the risk of tearing.

An embodiment of the invention will now be described by way of example, with reference to the accompanying drawings wherein: Figure 1 is a plan view of a short length of a clipper seam in accordance with the invention; Figure 2 is a sectional view on line II-II of Figure 1; and Figure 3 is a much enlarged diagrammatic cross-section of an adhesive matrix forming part of the clipper seam.

As shown in the drawings, a clipper seam is used to join the ends of a belt or filter fabric 10,11 to form an endless member for usa for example in filtration. The seam is formed by an array of P-shaped loop defining members 12 on each fabric end. The members 12 of the arrays

are interdigitatable with those of the opposed array as shown in Figure 1, and have outer ends 13 which are turned back to form an almost complete loop. These loops are interdigitated to form a tunnel through which a pintle 14 comprising a robust monofilament or multifilament yarn of e. g. nylon is passed. The members 12 also each comprise a shank 15 which terminates on a barbed end 16 which engages on the fabric 10, 11. the members 12 are provided with shanks 15 of two differing lengths, and members with longer and shorter shanks 15 are alternated in each array, as shown in Figure 1 so that the barbed ends 16 engage in the fabric in two different zones, spreading the load acting on the fabric and reducing the risk of tearing of the fabric.

As shown in Figure 2, the ends of the belt 10, 11 are each covered on the face of the fabric overlain by the shanks 15 of the members 12, by an adhesive layer 17, which covers and encapsulates the shanks 15 of the member 12. This adhesive layer 17 protects the shanks 15 against wear and pulling out from the fabric by a scraper blade acting on the fabric at intervals to remove accumulated filter cake during operation of the filter. In accordance with the invention, this adhesive layer 17 includes particles 20 of a harder material included in the mean matrix 21 of the adhesive material. The matrix 21 is in the preferred embodiment polyurethane based, and the harder particles are of steel. The inclusion of harder particles in the adhesive markedly reduces wear of the adhesive and thus helps to prevent premature failure of the belt.

EXAMPLE OF ADHESIVE COMPOSITION The adhesive matrix is formed by mixing four components: first polyurethane chosen to give optimum second polyurethane flexibility or other properties isocyanate hardener pigment according to colour required.

A typical laboratory trial batch mix comprises the following quantities and stages: 1. Mix 5kg first polyurethane with 4.5kg second polyurethane 2. Separately mix 5kg hardener with pigment 3. Mix the two products in proportions of 25. 3g of the second mix, per lOOg of the first (polyurethane) mix.

Once the basic adhesive mixture has been prepared, stainless steel powder is added thereto. The stainless steel is 316L grade, with a mean particle size of 100, um. The particles are sieved through a 300µm mesh to remove coarse material before being added. All mixing was carried out in a laboratory turbine mixer.

To test the effects of adding stainless steel particles to the mixture a plurality of batches were made up with different proportions of stainless steel particles, as set out below in Table 1.

TABLE 1 Batch Polyurethane/9 Hardener/g Stainless % Steel by Steel/g Weight 100 25. 3 0 B 100 25.3 10 7.3 C 100 25. 3 20 13. 7 D 100 25.3 120 48.9

To study the effect of the addition of stainless steel on the abrasion resistance and mechanical properties, films of each formulation were cast onto MELINEX film, using the laboratory K-bar 500 to give a nominal wet film thickness of 0.5mm. This resulted in cured polymer films of between 0.4 and 0.5mm thickness.

After curing the samples were peeled from the MELINEX film and employed for abrasion testing. The results of the abrasion tests are shown in Table 2 below. This compares the wear (as decrease in thickness) of a 0.5mm film when cycling against 240 grit silicon carbide paper. The results have been normalized against a particle free film in the last column.

TABLE 2 Batch Stainless Steel % Mean wear after Normalised mean wt 800 abrasion wear life cycles (mm) compared to 0% steel A 0 0. 112 1 B 7. 3 0. 052 2. 16 C 13.7 0.054 2.08 D 48.9 0.027 4.11

It is thus demonstrated that a 7.3 or 13. 7% by weight addition of stainless steel powder doubles the life of the adhesive film and a 48.9% addition of steel particles extends life by at least four times.

It is thus expected that useful results will be obtained with additions of steel particles from less than 10% by weight to in excess of 50% by weight.

It is also expected the relatively hard particles other than steel will give useful results, including zircon, alumina, sand (silica or silicate powder), glass, microbeads or balatini, and HASTELLOY metal. The choice of hard particles is determined by the conditions which the belt is likely to encounter. The normal choice will be stainless steel particles.

However if oxidising conditions are expected the more inert HASTELLOY metal may be used. If a metal is not suitable then one of the inorganic minerals or materials may be used.

Polyurethane or other adhesive formulations other than as set out in the examples may be used to form the matrix of the adhesive which contains the relatively hard particles.

Where hooks engage into the belt from both sides, both sides of the ends of the belt may be provided with an adhesive strip, or only the side of the belt engaged by the scraper blade may be so provided.

A problem which has been encountered with the use of relatively dense hard particles, for example, metal such as steel, is that the particles of a metal powder tend to settle to the bottom of the adhesive matrix before the matrix cures or hardens and behaves at this time as a liquid.

To overcome this it has been proposed to add a thixotropic agent to the matrix to increase its viscosity and thus hold the steel particles suspended in a more uniform dispersion throughout the matrix. Increasing the viscosity of the already relatively viscous adhesive however tends to reduce the penetration of the adhesive into the structure of the belt which would be expected to result in a less effective bond between the adhesive and the belt.

It has been found however that addition of a thixotropic agent as a rheology modifier, together with stainless steel powder results in an

improved bond, between the adhesive and many substrates. The preferred thixotropic agent is a powdered magnesium silicate, for example.

For the purposes of testing, a polyurethane adhesive composition was used, and formulations were prepared of the adhesive alone, without thixotropic agent or hard particles ; a second sample of the adhesive with addition of thixotropic agent, but without hard particles ; and a third sample of the adhesive with both hard particles and the thixotropic agent.

The formulations were then tested in turn as a range of substrates, for strength of adhesion using a Peel test.

The composition of the adhesive matrix was as described in the above example.

The thickener used is a thixotropic powder which is usually used as a thixotropic rheology modifier in adhesives, in an amount of 6.2g per 100g of polyol component.

The hard particulate material used comprises steel powder, in an amount of 20g per 100g of polyol component.

The samples thus have formulations as follows (weight %) :- Formulation 1 Formulation 2 Formulation 3 Polyol 80% 76% 66% Isocyanate 20% 19 % 17 % DT 5039 5% 4. 00% Steel Powder 1 3. 00 % Testing samples were prepared by attacmng ivitLvtx t i raae Mark) film to specimen pieces of different belt material using double sided tape half-way down an A4 piece. The adhesive formulation (1,2, or 3) was then poured equally into the belt specimen and the MELINEX film.

After curing of the formulation, the MELINEX is removed to produce tagged Peel Test samples 25mm wide. These were then tested on a testometric tensile test machine, and the degree of penetration of the adhesive into the material of the belt was assessed qualitatively.

The results are summarised in the following table : - Belt Material Formulation 1 Formulation 2 Formulation 3 M2-0070 Penetration :- Penetration :- Just Penetration :-Almost polyester mono and Almost complete. through complete. multi-filaments Peel Load :- ON thickness. Peel Peel Load :-1 13ON Load :-110N Y9-0620 Complete Just penetrated Complete. Coarse polyester Peel Load :- 70N through Peel Load :- 60N mono-filaments thickness. Peel Load 90N M9-0651-C Complete. Complete. Complete. Very coarse nylon and Peel Load :- 80N Peel Load :-110N Peel Load :- 78N polyester mono- filaments E3 0430 Complete. Complete. Complete. Medium polyester Peel Load :- 25N Peel Load :- 35N Peel Load :- 32N mono-filaments E2 0120 Complete. Almost complete. Almost complete. Fine polyester Peel Load :- 12. 5N Peel Load :- 20N Peel Load :- 11. 5N monofilaments

Formulations 1 and 2 differ in viscosity, due to the presence of the thixotropic in formulation 2. In general the higher viscosity compositions (formulation 2) penetrates the belts less effectively than the lower viscosity compositions, with only the more dense or tightly woven cloths

holding the adhesive back. It might thus be expected that the lower viscosity compositions would provide the better adhesion. However the high viscosity composition shows higher peel force except in one case (the coarsest most open belt). In all cases the mode of failure is predominantly adhesive (adhesive parts from substrate) with some cohesive failure (separation of bodies of adhesive) as adhesive is left within the weave structure.

Comparing data for formulations 1 and 3 illustrates the effect of adding stainless steel powder on the adhesion. Comparing the peel force data, adding both steel and thixotropic results in similar data with a minor decrease in one case and an improvement in two cases.

Much more variance is due to the nature of the belt substrate. With coarse and very coarse mono-filament substrates, i. e. those with the smallest relative surface area, the thixotropic agent is unable to give a stronger peel force with formulation 3 than formulation 1. However, without use of the thixotropic agent, the peel strength of an adhesive with steel particles would be so low as to preclude it from use on these types of substrates. It is thus evident that counter-intuitively, thickening of the adhesive has a generally beneficial effect on adhesion, whilst the further addition of steel particles is broadly neutral in its effects on the adhesive strength.

Instead of hard particulate material, hard fibrous material may be used, or a blend of particulate and fibrous materials. The fibres may be selected from metal fibres such as steel or HASTELLOY, inorganic fibres such as ceramic fibres, crystalline or amorphous mineral fibres or wools, and glass fibres. The ceramic fibres may comprise alumina, alumino- silicate, calcium silicates or other silicate materials.