Description Technical Field

[0001 ] The invention relates to a coil of sawing wire possibly wound on a carrier.

The carrier can be a spool or a mandrel. The wire is treated locally or totally with an adhesive. The sawing wire can be a wire suitable for use in a loose abrasive sawing process (loose abrasive sawing wire) or it can have an abrasive fixed on its surface (fixed abrasive sawing wire) or it can be used for Electric Discharge Machining (EDM wire).

Background Art

[0002] The first patents on wire saws for cutting hard and brittle materials such as silicon, quartz, germanium or the like are probably due to Hayward and were issued in the 50's of the previous century (see. GB 771622, GB 717874). Silicon ingots had then a diameter of 1 1/8 inch (28.5 mm). A loose abrasive process is described wherein a tungsten wire is used to carry a slurry containing an abrasive powder (carborundum i.e. silicon carbide) into the cut. Sawing occurred through three body abrasion (the work piece, the abrasive and the wire) and the wire used is referred to as 'loose abrasive sawing wire'. The wire is guided over four wire guides with parallel grooves thus forming a plane of parallel wires that is called the wire web. The wire is moved in a reciprocal to-and-fro manner whereby constantly a part of the wire is removed at the web exit, while fresh wire is being added at the entrance of the web. This machine has all features of what is currently called a 'multi wire sawing machine', which is actually a misnomer in that there is only one single wire present but this single wire is guided into multiple loops. Motorola Inc. seems to have been the first to exploit the technology on a large scale to cut silicon wafers for the semiconductor industry (see e.g. GB 1397676). Steel wires were then used to carry the slurry with the abrasive into the cut.

[0003] Nowadays the multi wire saw technology has matured and has completely overtaken the previous technology of cutting silicon ingots by means of circular internal diameter saw blades. Such an internal diameter saw was not longer cost effective due to the large kerf loss (ab. 300 μιτι). Moreover, cutting the current state-of-the-art silicon ingots of 300 mm diameter posed insurmountable technological barriers to this technology which since has been abandoned almost completely. At present all 300 mm wafers for semiconductor applications are cut with wire saws operating in a to-and-fro mode.

Also in the field of solar cell manufacturing, wire saws have taken a leading position in that practically all current crystalline solar cells are cut from single crystalline or polycrystalline silicon ingots by means of wire saws. As the geometric requirements on solar cell wafers are less stringent than for semiconductor wafers, solar cells are cut without reciprocal wire movement i.e. in unidirectional mode. This allows for much faster cutting as no time is lost in wire direction reversal (ramping-up and slowing-down).

[0004] Not only the size of the ingots has changed, also the diameter and length of the wire used has evolved. Where initially steel sawing wires of 180 micron were first used (leading to a kerf loss of about 200 μιτι) this has gradually diminished over sizes, 175, 160, 150, 140, down to 120 and even 100 μιτι. Inroads are being made to go as low as 80 μιτι with an associated kerf loss of about 90 μιτι. As the tension used in the sawing process has not been reduced over time the tensile strength of the wire had to increase in order to maintain sufficient strength of the wire. While initially the tensile strength was around 3400 N/mm ² for 180 μιτι wire, the tensile strength of the current 120 μιτι wire is over 4000 N/mm ² . The length of a single sawing wire has increased from 30 km up to about 900 km nowadays. Such a length must be provided without any splice or flaw.

[0005] Recently an alternative sawing wire technology is on the rise that seeks to eliminate the cumbersome slurry preparation, control, handling and disposal. By fixing the abrasive material to the wire, the need for a slurry can be dispensed with. Only a coolant is needed to remove the swarf from the cut and to clean the wire. In addition, the cutting efficiency is higher because abrasive particles do not longer roll-stick between wire and workpiece, but directly carve the material out of the workpiece. This wire is generally referred to as 'fixed abrasive sawing wire'.

[0006] Another wire saw technology that is used for cutting electrically conducting materials is 'Wire Electro Discharge Machining' or 'Wire EDM'. In this cutting technology the cut progresses by drawing electric discharges in a dielectric medium (an oil for example) between the work-piece and a conductive wire that is constantly renewed and defines the track of the cut. Such technology is making its first inroads for cutting semi-conducting materials such as silicon (see e.g. WO 2006/027946 A1 ).

[0007] In what follows 'sawing wire' is to be understood as referring to any one of a wire for loose abrasive sawing or a wire for fixed abrasive sawing or a wire suitable for use in Wire-EDM.

[0008] It will be clear from the previous that in the sawing process, be it with a loose or a fixed abrasive, or in Wire EDM, the wire plays a central role: the cutting stops instantly once the wire fractures. Wire fractures should be prevented by all means as they do not only lead to a loss of material - as in many cases, the workpiece is lost due to quality issues - but also to a loss in up-time as it takes some hours to clean up and rethread the web (that for solar cell cutting can contain more than 1000 loops).

[0009] One source of wire fractures in the sawing process is the winding of the wire. As the sawing wire is made in long lengths it is spooled on reels or bobbins. Alternatively it can be spooled on a mandrel that may or may not be removed out of the coil afterwards. In either case a coil of sawing wire will form on the spool or mandrel. In what follows consistent reference will be made to a 'sawing wire coil' thereby taking abstraction of the carrier of the coil: it is not relevant to the invention that the wire is on a certain type of carrier or not.

[0010] For reference purposes it is assumed that the winding axis is horizontal, that the wire moves towards the observer during winding and that the wire arrives on the spool above the axis i.e. the spool rotates towards the observer when released from its axis (see Figure 1 ). This frame of reference is by no means limiting as winding wire can equally well be done on a vertically organised wire winder, or on a winder where the spool is standing still while the wire is wound on the spool through a rotating flyer coaxially organised to the spool and moving up and down axially.

[001 1 ] The windings are organised in layers during the winding of the spool as the wire is led in a reciprocal manner from the left side of the spool to right side of the spool and back. For the purpose of this application, a 'layer' is defined as that series of windings that form when the wire travels from one side of the spool to the other side. Another layer is formed if the wire travels from the other side back to the first side. For the purpose of this application, the convention will be adhered to that odd numbered layers are layers wherein the wire is led from left to right during winding while even numbered layers are layers wherein the wire is led from right to left during winding.

[0012] A 'winding' within one layer takes the form a helix having a radius, a pitch length and pitch direction (right hand screw, 'Z' direction for even numbered layers with pitch P _eV en, left hand screw, 'S' direction for odd numbered layers with pitch P ₀ dd)- The radius is equal to half the diameter of the coil already formed and increases during winding as successive layers build on one another. The pitch length is the axial displacement that a wire shows over just one full turn around the coil axis. The ratio of pitch to circumference of the coil is equal to the tangent of the winding angle. Thus - when winding at constant pitch - the winding angle will decrease with increasing coil diameter.

[0013] From the above it will be clear that windings of odd and even layers will always cross somewhere during winding and the number of crossings over the width 'W of the coil 'C is:

C = W(1/P _odd + 1/Peven)

[0014] The 'Wire Coil Density' (WCD) is defined as the ratio of the volume

occupied by the wire divided by the overall volume of the coil and can be expressed as a percentage. A higher WCD allows to put more length of wire in the same spool volume.

[0015] There are many alternatives in selecting the pitch for winding of wires: - Using a pitch equal to the diameter 'D' of the wire is a necessary condition to achieve 'perfect winding'. In that case each layer is closed and odd and even layers build-up on top of one another. The density of the wire coil is maximal as the wires will assume a hexagonal packing (except where they cross). Perfect winding is difficult to achieve at high winding speeds, but allows extremely high unwinding speeds such as needed for missile guidance by means of optical fibres (see e.g. US 4 950 049, US 5 064 490).

- Using a pitch that is equal to a multiple of 'D' will result in side by side layers that leave no gap in between. Again such winding is difficult to achieve at high winding speed and with a large number of layers.

- By using large pitches the number of crossings over the spool width can be reduced but then the overall length per layer will diminish. This results in a smaller than desired wire coil density.

[0016] Within the sawing wire industry the following types of winding can be

identified.

- Wire winding pitches can be dictated by the unwinding station of the wire saw. Some wire saws are equipped with a take-off sheave that moves at a constant traverse speed. If the pitch of the sawing wire coil does not follow this unwinding pitch closely, the wire will be unwound at an unacceptable large angle relative to the axial perpendicular plane of the wire.

- For fixed abrasive sawing wire a pitch of between 'D' and '2xD' has been suggested in US 2007/0023027.

- Currently wire coils for loose abrasive sawing are spooled at a pitch that is larger than 2xD but smaller than 200xD. This proves to be a good compromise to achieve a sufficiently high WCD and unwinding quality.

The following problems can occur during use of a sawing wire:

[0017] Whenever the wire is used in to-and-fro manner, the problem of 'self- damage' of the sawing wire exists. As used wire is repeatedly wound back on the coil under a high tension, abrasive particles on windings that shift relative to one another may severely damage the wire. In case of loose abrasive sawing the abrasive particles stick in the slurry on the wire surface when wound back on the fresh wire. In case of fixed abrasive sawing wire the problem is even more pronounced as the abrasive particles are present on back spooled wire as well as on fresh wire. Such damage can lead to premature fracture of the wire.

[0018] There is also the problem of 'clamped wire'. As mentioned before sawing wires have evolved from relatively thick (180 μιτι) to - literally - 'hair thin' (80 μιτι). As the stiffness of a wire scales with the fourth power of the diameter the fine sawing wire does not resist bending. As the tensile strength has also increased over the years, a sawing wire is very difficult to plastically deform. Both trends have resulted in a 'springy' kind of wire that easily deflects - even under its own weight - but that is difficult to give a permanent bend. During manufacturing of the wire the windings are successively spooled next to one another and thereafter on top of one another. When during unwinding the wire leaves the spool under the same angle as it was spooled on, no unwinding problems will occur. If, due to a disturbance of the windings, one or more later laid windings arrive between the coil and windings of the same or another layer that were laid down earlier then later windings, clamping of the wire can occur during unwinding.

[0019] To make this problem better understandable reference is made to Figure 2. Part I of Figure 2 shows a single layer 212 at the outside of the coil 210. The layer 212 comprises different windings 'a', 'b', 'c', 'd' and 'e' which have been spooled upon the coil in that order. The visible half of the winding is indicated with a full line, while the not visible half is shown with a dashed line. Due to a disturbance, loop 216 of winding 'e' got trapped between the previously laid winding 'd' and the coil 210. Upon unwinding by pulling off the wire in direction 214, the windings 'e' and 'd' tension and 'd' clamps the loop 216 of winding 'e' against the coil 210. During unwinding a transient high force (a 'snag') occurs which can lead to wire breakage.

The problem is aggravated if the operator takes the loop 216 as the end of the wire and pulls out the free end out of winding 'e'. Over the full width of the coil unwinding will go difficult till the end of the coil is reached.

In part II of Figure 2, the winding ' f ' has become trapped under two previously laid windings ' e' ' and ' d' '. The interpretation of the figure remains identical.

'Clamped wire' occurs whenever a first laid winding is under-crossed by a later laid winding. These crossings are indicated with the dashed circles 218 in Figure 2.

[0020] It will be clear from the above that when the arrangement of the spool does not change and the wire is correctly unwound, no problems can occur. However the following disturbances frequently happen:

- During doffing of the spool the arrangement of the windings can be changed due to attaching the end of the wire (usually by making a knot), by falling windings when changing the orientation of the spool (e.g. by putting it vertical), or by paper wrapping.

- During the handling and transport of the spool (vibrations) the windings can be disturbed.

- As many sawing machines have the axis of the spools oriented in the vertical direction, there is a genuine risk of falling windings - and hence clamped wires - when mounting the spool.

- If the wire - during unwinding - is pulled from the spool under a large 'unwinding angle' relative to the axial perpendicular plane, there is a risk that top windings move and are pulled sideways over the underlying windings, thereby disturbing the order of the underlying windings. This risk is particularly eminent for sawing machines that in the return cycle of the to-and-fro wire movement use primitive winding systems that do not take into account the position of the last unwound fresh wire.

- Whenever the tension on the end of the sawing wire is lost, there is a risk for disturbance of windings. The wire-end must therefore be held tensioned at all times.

[0021 ] It will be clear from the above that many things can go wrong of which most of them are outside the control of the sawing wire producer. Various solutions have been sought for e.g. to have the wire coil fixed in a shrink wrap, to tape the outer windings or to better indicate the wire coil end by means of a clearly visible end tag. However, none of these solutions turns out to be satisfactory as a shrink wrap causes considerable disturbance of the outer windings during unpacking, a tape causes the windings to come loose during its removal. A clearly visible end-tag helps only if the sawing machine operator manages to keep the wire taught during threading of the machine which requires considerable skill and ability from the operator.

[0022] The inventors therefore set themselves the task to find a full proof solution to the 'clamping' and 'self-damage' problem.

Disclosure of Invention

[0023] The main object of the invention is therefore to eliminate the 'clamping' and 'self-damage' problem for any kind of sawing wire. More particularly the inventors wanted to eliminate these problems for fixed abrasive sawing wires, loose abrasive sawing wires and EDM wires. The inventors solved the problems on two levels: on the level of the sawing wire coil and on the level of the sawing wire on itself. Also a method to prevent clamped wires is offered.

[0024] According a first aspect of the invention a sawing wire coil is claimed with sawing wire wound in layers, each of said layers comprising a plurality of windings. From the previous section it is clear what is meant with a 'sawing wire coil' (paragraph [0009]), what is to be understood with a 'layer' (paragraph [001 1 ]) and what is meant with a 'winding' (paragraph [0012]).

The sawing wire within the context of this application (in addition to what was already mentioned in paragraph [0007]) is a metallic wire. Preferably the sawing wire comprises a steel core and one or more coating layers possibly with the abrasive incorporated in it or free of abrasive. The steel core is made of plain carbon steel (with a minimum carbon content of 0.70 wt% carbon) or a stainless steel. The coatings can be brass (the preferred variant for loose abrasive sawing wire), zinc (in case of EDM Wire) or copper topped with a nickel coat (in case of fixed abrasive sawing wire, the abrasive particles are held in the copper coat)

In addition preferred diameters and tensile strength of the sawing wire are if diameter is smaller than ... then tensile strength is larger than ...

250 pm 2900 N/mm ²

150 pm 3600 N/mm ²

140 Mm 3700 N/mm ²

120 Mm 3900 N/mm ²

[0025] Particular about the sawing wire coil is that on at least an area of the outer layers of the coil an adhesive is present (see Figure 3 for clarification).

- With the Outer' layers is meant those layers that are last wound on the spool (and thus the layers to be unwound first). Any layer with windings visible from the outside is considered to be an Outer layer' (indicated with 302 in Figure 3).

- With 'area' is meant a part or the whole of the surface of the coil . The area can be distributed over non-contiguous sub-areas, the area can be convex or concave, all this is not limiting the invention.

- With 'adhesive' should be understood a substance that has the

capability of holding the windings of sawing wire fixed. Synonyms to 'adhesive' is 'glue' or 'gum' or 'tacky substance' or 'sticky substance'. The way in which the glue is distributed on the wire is immaterial: it can be by means of a brush, a sponge, a spray, a tape dispenser or any other means known in the art. However with adhesive is not meant a tape with an adhesive substance on (such as a sellotape, scotch tape or the like) as this has been proven not to solve the problem.

By preference this area 306 at least includes the spot where the sawing wire 304 ends (see Figure 3a).

[0026] The function of the adhesive is to just hold the wire windings sufficiently so that they do not move easily. This 'holding force' can be determined by pealing of the wire in a direction perpendicular to the outer surface of the coil. The holding force of the adhesive is determined by the specific force of the adhesive and the contact surface area between adhesive and wire: the larger the contact surface area, the stronger the holding force will be. At least the holding force must be able to overcome the gravity force on at least one winding to prevent falling windings during transport and manipulation. This amounts to about 200 to 300 M N. By preference the adhesive area must be able to hold some windings in order to increase certainty that no windings will drop. Hence it is preferred if the holding force is larger than 1 mN.

The holding force should not be larger than the force by which the wire is pulled from the spool. In general the holding force is then lower than 25 N. By preference the holding force is below 5 N which is about the transient force occurring at a clamped wire. By preference this holding force must be lower than 1 N in order that unwinding at the adhesive area does not provoke transient force peaks. In general the adhesive will be a relatively weak adhesive that may hold the wires by pure mechanical anchoring (curing around sawing wire thereby holding wire in place) or by weak chemical interaction (Van Der Waals forces, polar interactions, but not covalent binding) or both. Tests have shown that a holding force between 20 to 260 millinewton suffices to hold the wire in place and does not obstruct the unwinding of the wire.

[0027] It is further preferred that at least some windings should be held by the adhesive. All windings of the outer layers 302 can be fixed by the adhesive if the area 308 extends from one end of the coil to the other end on the outer layers of the coil (see Figure 3b). Every winding of the outer layers will thus be held for at least a part of the winding. In this way the tension on the sawing wire end can never be lost during mounting and threading of the wire saw: always the wire is held and no windings can drop.

[0028] Alternatively the adhesive can be present in a circular band area 310 that encircles the axis of the coil (see Figure 3c). The band should be wide enough that at least enough windings are present to thread the saw machine as once the winding progresses outside the band the windings are not longer held.

[0029] A combination of both is also possible when the area takes the shape of a helix band that extends from one end of the coil to the other end while it encircles the axis of the coil (embodiment not shown in figures).

[0030] The adhesive can also be present on the whole surface 312 of the outer layers which certainly allows a sufficient number of windings to be held while the saw machine is threaded (see Figure 3d).

[0031 ] The adhesive can also be present throughout the whole coil i.e. all layers are covered with adhesive (embodiment not shown in figures). This is particularly preferred if the machine is working in to-and-fro mode as then the coil of fresh wire is still covered with adhesive which acts as a cushion for re-wound used wire that is coated with abrasive. This embodiment is particularly preferred for fixed abrasive sawing wire.

[0032] Sawing wire is usually wound with pitches that are between 2 and 200

times the diameters of the wire Ό'. More preferable the pitch is between 4 and 20 times the diameter of the wire Ό'. It is generally neither desirable nor possible to wind sawing wires in 'perfect winding' as is e.g. done in case of fiber optic wire packs (US 4 950 049, US 5 064 490). As explained (paragraph [0018]) sawing wires behave 'springy' due to their very low bending stiffness and high tensile grade level and do not lend themselves to perfect winding at the winding speeds that are customary for sawing wire winders.

[0033] It is also preferred that the pitches (Podd and P _eV en) are not an integer

multiple of the wire diameter 'D'. If the pitches are an integer multiple 'm' of 'D' the layers will be generally closed (except on those spots where a crossing occurs). With 'closed' is meant that windings of 'm' different layers lay one next to the other with the same radius. The 'm+1 'th layer will then start to form on top of these 'm' layers. If the pitches are not a multiple of 'D' a gap will be present between the wires thus forming an 'open' net through which adhesive can penetrate to a still lower radius layer, thereby enhancing the fixing of the coils. Hence, open layer winding is preferred.

[0034] According a second aspect of the invention a sawing wire on itself is

disclosed whereof the sawing wire coil according the first aspect of the invention is comprised.

[0035] The sawing wire is characterised in that it comprises an adhesive on at least a part of its surface. It has been clarified what is understood with a 'sawing wire' (paragraph [0024]) and an 'adhesive' (paragraph [0025]) in the context of this application. With 'on at least a part of the surface of said sawing wire' is meant that somewhere over the length of the sawing wire some adhesive can be found adhering to the wire.

[0036] By preference the adhesive is soluble in a polar medium. Examples of polar media are: - Water, that is generally used as a coolant agent in fixed abrasive wire sawing.

- Poly-ethyleneglycol (PEG) or specific compounds di-ethyleneglycol (DEG), tri-ethyleneglycol (TEG), tetra-ethyleneglycol. If the number of number of ethyleneglycol monomers is larger than four one generally refers to PEG. PEG - and to a lesser extent DEG - is particularly used as an abrasive carrier for loose abrasive sawing.

- Alcohol in all its variations for example methanol, ethanol, n-propyl alcohol, iso-propyl alcohol.

Mixtures of the above mentioned polar media can also be used and in certain cases will develop during use. For example: PEG is known to be hygroscopic and will absorb water during use.

[0037] A concern is that when using an adhesive on the sawing wire this

adhesive may interfere with the composition of the slurry in case of loose abrasive sawing. Indeed, the adhesive on the wire will be rubbed off in the first loops of the web and carried away by the slurry. In order to prevent clogging of tubing or conglomeration of abrasive on adhesive remnants, it is preferred that the adhesive dissolves in the slurry.

In case of fixed abrasive sawing, the coolant must be able to easily wash away the adhesive so that the abrasive particles in the coating can readily interact with the work piece.

[0038] It is therefore a preferred embodiment to use adhesives which are soluble in at least one of the above mentioned polar media. Possible adhesives which are preferred to this end are those selected out of the group comprising: polyvinylpyrrolidone (the 'sticky' component in hairspray); polyvinyl acetate, methylcellulose, polyvynil alcohol, or ethylene/vinyl alcohol copolymers, polyethyloxazolines, or mixtures thereof.

[0039] Alternatively other types of adhesives that are not necessarily soluble in polar media can be used. In that case the remnants of the adhesive will be removed in the sawing process through abrasion anyhow. In this embodiment the particles will be carried away by the slurry or the coolant. It is therefore preferred that when such adhesives are used, the amount of adhesive that enters the coolant or slurry is reduced to a minimum by e.g. minimising the area on which the adhesive is applied. Alternatively when the adhesive does not interact chemically with the wire and only mechanically anchors the wire, a mechanical cleaning can be provided prior to entry to the sawing web. This can e.g. be done by bending the wire or leading it through a brush.

[0040] Adhesives not necessarily soluble in polar media can be selected from the group comprising thermoplastic adhesives, hot melt adhesives, or thermosetting adhesives such as epoxy resins.

[0041 ] Alternatively the adhesive can be a bioadhesive selected from the group comprising caserne, starch derivates, hydrogels, poly saccharide or protein based glues.

[0042] As the adhesive comes into contact with the sawing wire it should not induce corrosion of the sawing wire. In order to prevent this a corrosion inhibitor can be added to the adhesive. Exemplary types of corrosion inhibitors are: phosphates, silicates, silanes, carbonates or carbonic acids, sulfides or mercaptoderivates, amines or sulfonates or combinations thereof.

[0043] The presence of an adhesive can easily be ascertained by checking the holding force of the wire end. Alternatively, the wire shows a certain 'tackiness' that can be felt by hand. The type of adhesive can be inferred from infra-red spectroscopy which can be performed directly on the wire.

[0044] A minimum amount of adhesive must be present on the sawing wire

before a beneficial effect is noticeable. The amount of adhesive can be best determined through determination of the organic carbon residue content on the surface of the wire by means of carbon pyrolysis. In this test only a limited sample (1 to 2 grams) is needed. The sample is heated to 480°C till the organic residues (but not the carbon in the steel) on the sample decompose into carbon monoxide and carbon dioxide. In a catalyser at 850°C all carbon monoxide is converted to carbon dioxide. The total amount of carbon is calculated from the infra-red absorption of the carbon dioxide. The total amount of carbon remaining must at least be larger than 400 g carbon residue per gram of sawing wire. It should not be larger than about 3 000 g carbon residue per gram of sawing wire. As the adhesive is not only composed of carbon (but also hydrogen and oxygen) the numbers are less than those determined by other methods such as the double weighing method. Also it must be assessed that the carbon residue is indeed due to the adhesive (and not some other organic compound) something which can easily be assessed through IR

spectroscopy.

[0045] According a third aspect of the invention a method is provided to prevent clamped sawing wires that occur during unwinding on a wire saw. The method comprises the step of winding a coil of sawing wire. In a first preferred embodiment, the wire is provided with an adhesive during coiling. In a second preferred embodiment the outer surface of the sawing wire coil is provided with an adhesive after the coil has been fully wound. The adhesive is at least applied on an area of the outer surface of the sawing wire coil.

Brief Description of Figures in the Drawings

[0046] Figure 1 shows the reference frame on how one can situate windings and layers on a spool.

[0047] Figure 2 shows in a detailed way how clamped wires arise.

[0048] Figure 3 shows four different embodiments 'a', 'b', 'c' and 'd' on how the invention can be implemented.

Mode(s) for Carrying Out the Invention

[0049] In a first attempt to overcome the clamped wire problem, the inventors were of the opinion that by oiling the wire, clamping could be suppressed as the sawing wire winding more easily shift over one so that they more easily find their equilibrium. Quite astonishingly the wire clamping increased. This led them to go into the completely opposite direction i.e. to make the wire sticky with an adhesive. This is counter-intuitive in that one does not expect a better unwinding by gluing the sawing wire together.

[0050] A series of trials were organised with a water based emulsion of

polyvinylacetate (on a 50/50 weight basis), called 'glue', hereinafter in different degrees of dilution in water.

[0051 ] In a first trial a Length 1 of about 8 km of 140 μιτι brass plated sawing wire was wound on a spool with core diameter of 156 mm at a tension of about 3 N and a pitch of 1 .5 mm. During winding a glue to water solution of 1 in 4 volume parts was constantly applied on the wire. After drying about 2 km was removed at a speed of 500 m/min and at unwinding tension of 9 N. Unwinding tension was monitored through a tension meter. Unwinding was done under a severe unwinding angle of about 45° to 85° with respect the plane perpendicular to the coil axis. Only small clamping forces (smaller than 0.1 N) were noticed that could be attributed to the too strong 'holding force' of the adhesive. Visually no shifted windings were observed.

[0052] A reference spool, LengthR, wound under equal circumstances as the first length spool but without administering any adhesive was made. The spool was also unwound under identical circumstances as the first length.

Visually, shifted wire windings could be clearly observed and heavy clamping was noticed throughout the spool.

[0053] In a second series of experiments different degrees of dilution of the glue in water (volume ratios) were tested. On 3 spools filled with 5 km of 120 μιτι sawing wire under a tension of 3N with a pitch of 1 .5 mm (12.5 x D) a circumferential band 30 mm in width was coated with an adhesive 10 mm from the end of the coil and at the other side followed by a 'control zone' (uncoated) of 30 mm. On the first spool a solution with a dilution of 1 :20 (Length2), on the second spool a dilution of 1 :10 (Length3) and on the third spool a dilution of 1 :5 (Length4) was applied in the circumferential band by means of a paint brush.

[0054] Again the spools were unwound at the same severe unwinding angle at an unwinding tension of about 14 N at a speed of 600 m/min. The following results were obtained (Table I):

Ratio Carbon residue Unwinding (glue:water) rating

1 :5 1515(132) No clamping

1 : 10 659(104) 2 occurences of

clamping

1 :20 375(12) Heavy clamping

throughout

Control zone 95(24) Heavy clamping

throughout [0055] Figures between brackets refer to standard deviation. Numerical results are an average of at least four individual values. 'Carbon residue' was determined by the pyrolisis method. Note that the 'control zone' shows some carbon residue due to carbon remnants on the uncoated sample. The unwinding rating confirms that a minimum amount of glue is necessary to prevent clamping. It can be concluded that a solution of at least 1 in 10 volumes of glue in water is sufficient to reach the desired holding force, while a solution of 1 in 4 volumes of glue in water already shows too much holding force.

[0056] The glue solution can be applied in a number of ways e.g. by means of spraying, dipping, painting or immersion. The glue solution can be applied continuously during winding (e.g. by running the sawing wire through a dip tank or a glue application felt), intermittently (e.g. after a number of layers has been wound an automatic spray gun applies some glue to the intermediately formed coil without stopping the winding process) or finally after winding, (e.g. when the sawing wire coil is doffed, the outer surface is painted with a band of glue, preferably before de-tensioning the wire).