Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD OF FORMING A THERMAL-INTERFACE
Document Type and Number:
WIPO Patent Application WO/2016/170299
Kind Code:
A1
Abstract:
There is provided a method of forming a thermal-interface with high thermal conductivity between a first and a second metal body. The method comprises selecting a first body (1) having a first surface (5) prepared smooth with micro roughness and defining first asperity contacts (9); selecting a second body (8) having a second surface (6) prepared smooth with micro roughness and defining second asperity contacts (10); introducing a thin film of cold pressure welding agent between the first and second surfaces; and pressing together the first and second surfaces, by positioning a third clamping body (11) around the second body, thereby (i) deforming the first and second asperity contacts; and (ii) trapping at least some of said cold pressure welding agent between said first and second deforming asperity contacts, to thereby momentarily reduce the deformation threshold of the asperity contacts and cause extra asperity deformation and cold pressure welding there between sufficient to raise the thermal conductivity between the first and second bodies. The third clamping body has a lower coefficient of thermal expansion than the second body.

Inventors:
LINZELL GEOFFREY ROBERT (GB)
Application Number:
PCT/GB2016/000087
Publication Date:
October 27, 2016
Filing Date:
April 19, 2016
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
BALL BURNISHING MACHINE TOOLS LTD (GB)
International Classes:
B23K20/02; B23K20/16; F16B11/00
Domestic Patent References:
WO2011101616A12011-08-25
WO1991019589A11991-12-26
Foreign References:
EP2536528A12012-12-26
US6324060B12001-11-27
Attorney, Agent or Firm:
LINZELL, Geoffrey Robert (Brookmans ParkHatfield, Herts. AL9 7QJ, GB)
Download PDF:
Claims:
Claims

1. A method of forming a thermal-interface with high thermal conductivity between a first and a second metal body, the method comprising:

(a) selecting a first metal body having a first surface prepared smooth with micro roughness and defining first asperity contacts;

(b) selecting a second metal body having a second surface prepared smooth with micro roughness and defining second asperity contacts;

(c) introducing a thin film of cold pressure welding agent between said first and second surfaces; and

(d) pressing together said first and second surfaces, by positioning a third

clamping body around the second metal body, thereby

(i) deforming said first and second asperity contacts; and

(ii) trapping at least some of said cold pressure welding agent between said first and second deforming asperity contacts, to thereby momentarily reduce the deformation threshold of the asperity contacts and cause extra asperity deformation and cold pressure welding there between sufficient to raise the thermal conductivity between the first and second metal bodies, wherein said third clamping body has a lower coefficient of thermal expansion than the second body.

2. A method as claimed in claim 1 , wherein the first and second surfaces are pressed together by positioning the third stiffer clamping body that deforms and maintains the first surface in high pressure clamped contact with the second surface.

3. A method as claimed in either of claims 1 or 2, wherein the material of the second body has a higher thermal conductivity than the materials of the first and third bodies, and thereby acts as a heat sink.

4. A method as claimed in either of claims 1 or 2, wherein the material of the second body has a higher coefficient of thermal expansion than the materials of the first and third bodies.

5. A method as claimed in claim 4, whereupon warming causes differential expansion that raises the contact pressures between the bodies sufficient to contribute a further rise in thermal conductivity between the first and second metal bodies as their temperatures rise.

6. A method as claimed in any of claims 1 to 5, wherein the material of the third body is stiff and resists bulk deformation when pressing together the first and second surfaces.

7. A method as claimed in claim 4, whereupon on warming the three bodies they expand, and the third clamping body acts to resist the greater amount of thermal expansion of the second body in the direction normal to the thermal-interface.

8. A method as claimed in claim 4, wherein the larger thermal expansion of the second body causes the contact pressure at the thermal-interface between the first and second bodies to rise as the temperature of the bodies rises.

9. A method as claimed in any of claims 1 to 8, wherein the heat conduction across the thermal-interface is improved by increasing contact pressure as the temperature of the bodies rise by increasing the number and size of asperity contacts between the first and second bodies

10. A method as claimed in any of claims 1 to 8, wherein the heat conduction across the thermal-interface is improved by converting at least some of said deforming asperity contacts into cold pressure welds by the action of the trapped cold pressure welding agent.

11. A method as claimed in any of claims 1 to 10, wherein the first and second surfaces are first machined to size and shape, then cold worked to reduce roughness by flattening asperity peaks.

12. A method as claimed in any of claims 1 to 11 , wherein the first and second surfaces have uniform asperity contact spacing.

13. A method as claimed in claim 12, wherein said uniform asperity contact spacing is less than 25 micron.

14. A method as claimed in any of claims 1 to 13, wherein the directions of roughness between the first and second surfaces are aligned to maximise asperity contacts there between.

15. A method as claimed in any of claims 1 to 14, wherein the thermal-interface is formed between either flat or non-flat matching first and second surfaces.

16. A method as claimed in any of claims 1 to 15, wherein the method comprises incrementally raising thermal conductivity between the first and second bodies by maximising contact area and contact pressure with temperature.

17. A method according to any of claims 1 to 16, wherein said trapping step increases the area of contact between the first and second surfaces thereby increasing thermal conductivity.

18. Assembly obtainable by the method as claimed in any of claims 1 to 17, in which assembly the first body is a heat source, the second body is a heat sink and the third body is a clamp.

19. Assembly as claimed in claim 18, wherein the thermal-interface forms part of a thermal path within a cylindrical electrical machine.

20. Assembly as claimed in claim 19, wherein the first body is a stator with an outer prepared first surface, and the second body is a tubular heat sink sized to slide freely over the stator and with an inner prepared second surface, and the third body is a interfering fit tubular outer clamp.

21. Assembly as claimed in claim 20, wherein the third body acts as part of the machine housing.

22. Assembly as claimed in either of claims 20 or 21 , wherein the first body is an iron cored stator and the second body is an aluminium based heatsink.

23. Assembly as claimed in claim 22, wherein the second body (the heatsink) is tightly clamped to the first body (the stator) by the third body (the clamp).

24. Assembly as claimed in claim 23, wherein said third body is formed of a material selected for its strength, elasticity and low weight when securing the assembly.

25. Assembly as claimed in claim 24, wherein, the third body material is a stainless steel, a spring steel, titanium, or a metal or polymer matrix fibre reinforced composite with metal terminations made with similar technology as used in carbon composite drive shafts.

26. Assembly as claimed in any of claims 20 to 25, wherein the tubular heatsink is divided into two the division sized to allow for the greater axial expansion of the heatsink over the heat source.

27. Assembly as claimed in claim 26, wherein fluid flow ducts are provided to the outer surface of the tubular heatsink.

28. Assembly as claimed in claim 26, wherein said division is sealed by inserting a compressible seal that seals said fluid flow ducts.

29. Assembly as claimed in any of claims 18 to 28, wherein the clamp is positioned and secured by press-fitting or heat shrinking.

30. Assembly as claimed in any of claims 18 to 29, wherein the method provides a second body with thermal mass selected to provide thermal smoothing by rapidly absorbing thermal transients.

31. Assembly as claimed in any of claims 18 to 30, wherein the volume of the second body is selected to provide the desired thermal mass to absorb thermal transients.

Description:
Method of forming a thermal-interface

Technical field

The present method is in the technical field of forming a thermal-interface with high thermal conductivity contacts between first and second metal bodies.

Background

Thermal conductivity is the property of a material to conduct heat and is evaluated primarily in terms of Fourier's Law for heat conduction. Heat flow occurs at a higher rate across materials with high thermal conductivity. A thermal-interface is a coupling or junction between thermal conductors in a thermal path or system.

It is desirable to provide an improved method for forming a thermal-interface with high thermal conductivity between metal bodies. In particular, it is desirable to provide means of improving the thermal conductance across a junction of dissimilar metals where the coefficients of thermal expansion cause an unfavourable drop in contact pressure as temperatures rise. For example it is common for electrical machines to employ electromagnetic stators with iron based cores that are heat- shrunk into an aluminium machine housing that provides a thermal path for cooling the stators, whereupon if heated uniformly the aluminium will expand more than the iron causing both friction grip and thermal conductance at the thermal-interface between the stators and the housing to fall as temperatures rise. Current practice is to ensure the housing is sufficiently cooled to prevent this occurring, but this imposes a cost penalty and a serious limit on the overload tolerance of current machines.

The method herein overcomes the above limitation by raising thermal conductivity at a thermal-interface between a first and a second body by first introducing a cold pressure welding agent at the thermal-interface, then second, pressing and holding the first and second bodies together. The applied pressing force activates the cold pressure weld agent which enlarges cold pressure asperity welds. Upon heating the dissimilar metal bodies they thermally expand causing the contact pressure to rise further thereby further increasing thermal conductivity with temperature. Initial assembly is done at ambient temperature where a cold welding agent is applied to one or both first and second surfaces, which are then pressed together by forcing an interfering clamp into place to apply the initial pressure, the first force. This force activates the asperity cold weld agent to increase thermal conductivity as it clamps and holds the first and second bodies together like a three layer sandwich.

Upon passing heat though the sandwich, the second body, a heat sink expands faster than the heat source which is the first body; and the clamp, which is the third body; thereby raising the contact pressure between the heat source and the heat sink, to create a second force that increases with temperature, this second force further raises thermal conductance by causing further cold welding between source and sink due to the action of the cold welding agent as temperature rises.

The interface behaviour is counter intuitive because thermal conductance within and between metals naturally fall as their temperature rises. Thus in the event of a thermal overload the contact pressure at the thermal-interface rises causing more cold pressure asperity welds to form which increase thermal conductivity, which welds accumulate with successive overloads. Upon cooling the welds remain providing the first force is maintained, thus thermal conductance rises over the life.

The sequence of events is thus;

Upon initial assembly the introduced cold pressure welding agent stimulates the formation of an array of cold pressure asperity welds as it is compressed by positioning a third body, a clamp, that is sized to be interference fit over the second body, positioned by either pressing or heat shrinking onto the second body thereby forcing and holding the second body pressed hard against the first body.

In normal use, upon warming the contact pressure rises moderately at the interface due to differential thermal expansion between the three bodies. The initial cold pressure welds sufficient to cool the assembly for normal operation.

Upon overload the thermal-interface temperature rises causing differential thermal expansion which raises the contact pressure, which further activates the cold pressure welding agent thereby increasing welding and raising thermal conductivity. The second body acts as a heat sink and is sized to rapidly absorb and smooth thermal transients. Upon cooling, the additional welds remain intact providing the first static contact pressure (compressive stress) applied during the initial assembly is maintained. Hence thermal conductivity at the thermal-interface tends to increase over time as a result of successive thermal overloads.

Thus, the method herein comprises selecting a first body, with a first coefficient of thermal expansion, with a first surface prepared smooth with micro roughness and defining first asperity contacts; selecting a second body with a second coefficient of thermal expansion, with a second surface prepared smooth with micro roughness and defining second asperity contacts; introducing a thin film of cold pressure welding agent between the first and second surfaces; and pressing together the first and second surfaces with a static clamp, which clamp is a third interfering body forced against the second body, which forces and holds the second body in close contact with the first body, thereby (i) deforming the first and second asperity contacts; and (ii) trapping at least some of the cold pressure welding agent between the first and second deforming asperity contacts, to thereby activate part of the agent causing a momentary reduction in the deformation threshold of the deforming asperity contacts, providing extra asperity deformation and cold pressure welding there between sufficient to raise the thermal conductivity between the first and second bodies. Said cold welds are accumulative and if agent is available further welds will develop in service if contact temperature and pressure rise due to overload. Upon cooling, welds remain intact providing the initial static clamping force is maintained.

Cold pressure welding agents are described in our earlier application EP2536528 which teaches means of enhancing friction grip therewith for the purpose of making mechanical assemblies. It does not at all mention forming a thermal-interface for conducting thermal energy between contacting metal bodies or the use of materials with different rates of thermal expansion to increase contact pressure and thereby improve conductivity with temperature. Neither does it anticipate the assembly of electrical or mechanical machines incorporating the method of enhancing thermal conduction between metal bodies therein.

US 6,324,060 B1 describes improving thermal conductance by use of thermal paste. Therefore the invention provides a method of forming a thermal-interface suitable for improving cooling of machines, especially electrical machines, by incorporating enhanced cold pressure asperity welds between pressed together first and second dissimilar metal bodies; whereupon further heating of said first and second bodies increases the thermal conductivity at the thermal-interface as temperature rises,

Summary of Invention:

A method of forming a thermal-interface with high thermal conductivity between a first and a second metal body, the method comprising:

(a) selecting a first metal body with a first coefficient of thermal expansion and having a first surface prepared smooth with micro roughness and defining first asperity contacts;

(b) selecting a second metal body with a second coefficient of thermal expansion and having a second surface prepared smooth with micro roughness and defining second asperity contacts;

(c) introducing a thin film of cold pressure welding agent between said first and second surfaces; and

(d) pressing together said first and second surfaces, by positioning a third

clamping body against the second metal body, thereby

(i) deforming said first and second asperity contacts; and

(ii) trapping at least some of said cold pressure welding agent between said first and second deforming asperity contacts, to thereby

momentarily reduce the deformation threshold of the asperity contacts and cause extra asperity deformation and cold pressure welding there between sufficient to raise the thermal conductivity between the first and second metal bodies, wherein said third clamping body has a lower coefficient of thermal expansion than the second body. In a further aspect the force pressing the first and second surfaces together is the sum of two forces, first, a static force applied cold upon assembly by positioning the third interfering clamping body and second, a dynamic force the result of the differential thermal expansion between the second and the first and third bodies upon heating.

Description

For optimum thermal conduction between the first and the second metal body a homogeneous metal junction is optimum. To form such a junction requires high energy (temperatures) sufficient to either fuse or diffuse body materials together. The method herein provides a compromise by introducing a chemical cold pressure welding agent that when trapped between deforming asperities (asperities means high spots within a surface roughness), the cold welding agent is activated and releases small amounts of chemically active elements, such as single atoms of hydrogen, that acts as a reducing agent and also rapidly diffuses into the deforming asperities, where it appears to momentarily lower the asperity yield strength, thereby increasing the amount of plastic deformation between each pair of contacting asperities causing enlarged cold pressure welds to form. These enlarged cold welds raise the thermal conductivity of each contact. Upon cessation of deformation the hydrogen appears to diffuse out of the metal and the metal recovers its full strength and ductility with no evidence of hydrogen embrittlement.

In embodiments, the term 'thermal-interface' defines a region or area wherein multiplicities of thermally conductive cold pressure asperity welds are formed as the bodies are pressed together in the presence of a cold pressure welding agent.

Thus the method provides as a first step, the preparation of the surfaces to provide a multiplicities of bridging asperity contacts; a second step is the introduction of a cold pressure welding agent and a third step is positioning a clamp that presses and holds the prepared surfaces in high pressure contact to activate the cold weld agent.

All surfaces have residual roughness due to practical manufacturing limitations, for example cut-marks left by tools. The roughness typically takes the form of

microscopic high spots (asperities) that resemble microscopic mountains with valleys running there between. The tops of the asperities may be flattened (plateaued) to maximise their individual contact areas (the cold pressure weld sites).

Each asperity contact will transfer heat between the contacting bodies. Therefore, the more asperities there are in contact the greater the thermal conductivity. Thus the thermal-interface surfaces are prepared with micro-roughness to provide a dense array of flattened microscopic contacts, with a labyrinth of valleys carrying residual cold pressure welding agent between the flattened asperities.

The term "smooth" is used herein to describe the general appearance of a surface to the naked eye, such a surface appears shiny and without obvious roughness. The term micro-roughness defines a surface that appears smooth but actually possesses microscopic roughness that is only visible through a microscope or with a profile measuring machine. Upon pressing the surfaces together, the number of contacting asperities increases, therefore thermal conductance increases with contact pressure. Thermal conduction will be increased further by the introduction of said cold welding agent as it increases contact area. What appears to be a good physical contact between touching surfaces may in fact be poor, with as little as 1 or 2 % true contact. The effect of introducing the cold welding agent is to increase the actual contact area to between 5 and 15%, rising to an estimated 50% over time due to overloads.

Thus surface roughness is mainly caused by imperfect machining when shaping and sizing parts. A machined surface will typically have two roughness profiles that are superimposed. First is referred to herein as micro roughness, which is a low amplitude high frequency wave like pattern the result of successive passes of a cutting tool. Second is a low frequency deviation of larger amplitude commonly referred to as waviness or macro roughness, which is usually attributed to machine imperfections such as bearing wear, frame resonance or work-piece instability like de-lamination of laminated stacks during machining.

By way of a practical example, in the case of an electrical machine, the first body will typically be an iron or steel based material such as used in an electromagnet and the second body is typically a high thermal conductivity material such as aluminium alloy that acts as a heat conductor referred to hereinafter as a heat sink. Said first and second bodies are pressed and held together by a stiff band that acts as a clamp, referred to hereinafter as the third body, which may be part of the machine housing.

In embodiments, the material of the first body is selected primarily for its magnetic properties but should also have good thermal conductance. In embodiments, the materials of the first and third bodies have lower coefficients of thermal expansion than the material of the second body.

The material of the second body selected to provide high thermal conductivity and is sized to provide adequate thermal mass to absorb anticipated thermal transients. in embodiments, the thermal-interface between a stator and a heatsink is sized to provide sufficient conductivity (conductive area) to transfer predicted fault energy from the stator into the heatsink.

The material of the third body may be a metal or non-metal such as a carbon reinforced composite providing it maintains its strength at the temperatures likely to be experienced under fault conditions and it has a suitably low coefficient of thermal expansion.

The material of the third body is selected for its stiffness (strength). The term 'stiff means the material has a high yield point and good elastic range, ideally behaving like spring steel. Its function is to constrict (that means to limit) the radial expansion of the first and second bodies during thermal overload without yielding. The term 'stiffness' relates to the rigidity of said clamp; this is the extent to which the third body resists the thermal expansion of the first and second bodies.

Therefore the three bound together bodies form a sub-assembly with a thermal coupling between the first and second bodies where they touch, held in contact by the third clamp body. The assembly forming part of a machine through which heat is passed (conducted) from a first body to a second body, where the first body is a heat source and the second body is a heatsink. The assembly includes a thermal junction with at least some inter metallic blending of the joined materials at the junction, which is known in the literature as a cold pressure weld. Typical welded materials in the case of an electrical machine being magnet iron and aluminium, (Fe and Al). This material combination is known to be suitable for cold pressure welding, for example it is used for cold roll bonding aluminium onto steel.

Thermally coupled bodies are collectively referred to herein as a part of a 'thermal path'. Thus the method herein provides means of improving the thermal conduction at a thermal-interface within a thermal path. Such paths are common in machines where waste heat requires conducting from a source to a means of dissipation. Thus the method is applicable to, but is not limited to uses in, internal combustion engines, steam turbines etc. hydraulic and pneumatic pumps, linear or rotary electrical and mechanical components like solenoids for control valves etc. electric actuators for door openers etc., maglev (magnetic levitation devices for transportation), chemical devices such as batteries, electrical power distribution terminals, circuit breakers, bus bar terminals as well as electrical machines such as motors or generators.

While in principle the method can be applied to improve durability in any of the above machines, the method is particularly useful for improving the durability of electrical machines especially those used in electric vehicle drives where large externally generated overloads occur in service which impact on drive train reliability.

In a particular aspect, the method herein provides an electrical machine, wherein the first body is an electromagnetic stator which needs to be cooled, the second body is an aluminium heatsink that is clamped to the stator for the purpose cooling the stator via a high conductance thermal-interface between the first and second bodies; the third body is a stiff band clamp that maintains the heatsink clamped to the stator.

Typical electrical machines used in vehicle drives are electronically commutated ac synchronous motor/generators referred to herein after as 'brushless machines'.

There are two basic magnetic field configurations employed in brushless machines and these configurations influence both the machine shape and their cooling.

First is an axial flux machine, so called because the flux fields are axially disposed between rotating discs. This provides a short shaft machine with relatively large diameter end faces through which heat is often removed. These machines are commonly referred to as pancake machines and may be conveniently shaped and sized to mount into vehicle wheel hubs where they are either air or fluid cooled. Second are radial flux machines, so called because the flux fields are arranged along the axis of the machine shaft. This provides a relatively long axial machine with a smaller diameter cylindrical body that resembles the traditional cylindrical profile of common electric motors. The main cooling area is the outer cylindrical housing which may be hollow to facilitate air or water (fluid) cooling. Means of cooling of these machines can become very complex, extending to the costly need of passing coolant through both rotor and stator.

Similar thermal-interfaces are used in the cooling paths of radial and axial flux machines wherein iron cored electromagnets are heat shrunk into a high thermal conductivity aluminium housing that also acts as a heat sink. Said electromagnets are the primary heat source in brushless machines due to l 2 R loss in the

electronically commutated windings and iron losses within their iron cores.

For ac operation laminated iron based cores are used, formed by stacking thin insulated laminations of suitable cold rolled silicon iron. These provide optimum magnetic penetration, minimal hysteresis and iron loss with minimal eddy currents and acceptable thermal conductivity to the thermal-interface at the outer edge of the stack. Alternatively for dc operation where flux direction does not alternate, the iron cores are often pre-formed with powdered iron or iron/ceramic compounds and sintered to form near solid cores. If these cores are electrically conductive they will develop high iron losses due to induced eddy currents, hence they are less suitable for ac operation. Newer materials based on polymer insulated iron powder particles have lower iron losses in ac operation. Providing the core materials have thermal coefficients of expansion suitably less than aluminium they may in principle be used with the method herein; however the cold weld agent would not be expected to work as well on electrical insulators because they lack the high concentration of free conduction electrons able to move freely, which are typical of metals, and which are available to transfer heat as they move through a lattice. Lattice vibration (Phonon- based) conduction occurs in non-metals and non-welded thermal contacts.

At the welded asperities within the thermal-interface conduction will be mainly electronic and at non-welded asperity contacts conduction will be mainly by phonon. In embodiments, means are provided herein for minimizing the effect of both micro and macro roughness. First micro roughness is reduced and bearing area ratio is increased by cold working the surfaces with means such as roller or ball burnishing or shot peening to flatten asperities. Second the asperity contact pattern caused by macro roughness is reduced by applying a uniform deforming force with the third body as it presses the softer second body into the waviness of the harder first body. In embodiments, wherein the first and second surfaces are pressed together by positioning the third stiffer clamping body that deforms and maintains the first surface in high pressure clamped contact with the second surface.

In embodiments, the outer clamp (the third body) is selected to be resilient and able to adapt to waviness errors across the contact area between the second and third bodies which ensures pressure is evenly applied to all the touching asperities within the thermal-interface.

In embodiments, the third clamping body has a lower coefficient of thermal expansion than the second body.

In embodiments, the first and third bodies have lower coefficients of thermal expansion than the second body.

In embodiments, the first and second metal bodies are dissimilar metals.

In embodiments, warming causes differential expansion that raises the contact pressures between the bodies sufficient to contribute a further rise in thermal conductivity between the first and second metal bodies as their temperatures rise.

In embodiments, the material of the third body is stiff and only deforms elastically when pressing together the first and second surfaces.

In embodiments, whereupon warming the three bodies expand, and the third clamping body acts to resist the greater amount of thermal expansion of the second body in the direction normal to the thermal-interface.

In embodiments, the heat conduction across the thermal-interface is improved by converting at least some of said deforming asperity contacts into cold pressure welds by the action of the trapped cold pressure welding agent. In embodiments, the first and second surfaces have uniform asperity contact spacing. In embodiments, the asperity contact spacing is less than 25 micron, preferably less than 15 micron.

In embodiments, the thermal-interface is formed between either flat or non-flat matching first and second surfaces.

In embodiments, the method includes incrementally raising thermal conductivity between the first and second bodies by maximising asperity contact area and contact pressure with temperature.

In embodiments, the trapping step (i.e. step (d)(ii) ) increases the area of contact between the first and second surfaces thereby increasing thermal conductivity.

According to another aspect of the claimed invention there is provided an assembly obtainable by the method as described herein.

In embodiments, in the assembly the first body is a heat source, the second body is a heat sink and the third body is a clamp.

In embodiments, the thermal-interface forms part of a thermal path within a cylindrical electrical machine.

In embodiments, wherein the first body is a stator with an outer prepared first surface, and the second body is a tubular heat sink sized to slide freely over the stator and with an inner prepared second surface, and the third body is a interfering fit tubular outer clamp.

In embodiments, the third body acts as part of the machine housing.

In embodiments, the first body is an iron cored stator and the second body is an aluminium based heatsink.

In embodiments, an assembly is claimed wherein the second body (the heatsink) is tightly clamped to the first body (the stator) by the third body (the clamp).

In embodiments, the third body is stiff, formed of a material selected for its strength, elasticity and low weight when securing the assembly. In embodiments, the radial thermal expansion of that part of the aluminium heatsink (the second body) that frictionally retains the stator (the first body) is constricted (that means it is selectively restricted) by surrounding said part of the heatsink with a stiffer (stronger) external clamp (the third body), which clamp has a lower coefficient of thermal expansion than said heatsink.

Upon warming, the aluminium (the second body) tries to expand radially by a greater amount than the stiffer clamp (the third body) and the stiffer stator (the first body).

In embodiments, the second body is constricted by the stiffer first and third bodies causing the interface pressures between the bodies to rise with temperature and the second body is forced to extrude axially to accommodate its greater expansion.

The optimal thermal-interface within a thermal path would be a perfectly smooth continuous gap free metal to metal contact, where thermally conducting electrons could pass unimpeded from the first body to the second body. But it is impractical to make suitable microscopically flat surfaces. Therefore practical thermal-interfaces are prepared with surfaces made as smooth as possible then treated with a thin smear of thermally conducting grease, which surfaces are pressed and held together to force the paste to flow and fill air gaps. This can result in a thin film of paste that may prevent metal to metal contact and inhibits the formation of cold pressure welds.

US 6,324,060 B1 describes improving thermal conductance by use of thermal paste.

In embodiments a thermally conductive paste may be combined with a cold pressure welding agent in sparing amounts, this can be beneficial providing it does not hold the surfaces apart and prevent asperity welds developing at the interface. In embodiments, a thermal paste may be applied to one surface at a thermal-interface and a cold welding agent to the other surface prior to pressing them together to form an enhanced thermal-interface. Again good asperity contact is essential.

In embodiments, differential rates of thermal expansion between the first, second and third bodies occurs as their temperatures and contact pressures rise, which when combined with the introduction of the cold welding agent between the pressed together first and second bodies causes cold pressure welds to form at deforming asperity contacts there between. In embodiments, the material of the second body has a higher thermal conductivity than the materials of the first and third bodies, and thereby acts as a heat sink, a suitable material for the second body being copper or aluminium or alloys thereof. The term "heat sink" describes a body able to absorb and rapidly distribute heat evenly throughout its mass, the material selected to have high thermal conductance, high specific heat and low specific gravity. The volume (dimensions) of the heat sink selected to absorb the peak overload energy (that is the maximum thermal dissipation within the object being cooled).

In embodiments, the method provides a second body with thermal mass selected to provide thermal smoothing by rapidly absorbing thermal transients. In embodiments the volume of the second body is selected to provide the desired thermal mass to absorb predicted thermal transients.

In embodiments, heat is transferred by conduction from the first body via a thermal- interface made by the method into the heatsink. Heat is removed from the heat sink via a cooling medium by conduction via either pumped fluid or blown air flowing through channels cut into the heat sink. The rates of heat transfer in and out of the heatsink are selected to limit the overall temperature rise of the insulation on the windings within the stator when an overload occurs therein.

In embodiments, the material of the second body has a higher coefficient of thermal expansion than the materials of the first and third bodies.

In embodiments, the first and third bodies are stiffer (stronger) than the second body.

In embodiments, upon warming the bodies, the second body deforms aniso-tropically due to the restraining force exerted by stiffer (stronger) first and third bodies, this raises the contact pressure between the first and second bodies with temperature.

In embodiments, a method is provided whereupon on warming the three bodies they expand, and the third clamping body acts to resist the greater thermal expansion of the second body in the direction normal to the thermal-interface.

In embodiments, the larger thermal expansion of the second body causes the contact pressure at the thermal-interface between the first and second bodies to rise as the temperature of all the bodies rise . In embodiments, the heat conduction across the thermal-interface is improved by increasing contact pressure as the temperature of the bodies rise by increasing the number and size of asperity contacts between the first and second bodies

In embodiments, the first and second surfaces are first machined to size and shape, then cold worked to reduce roughness by flattening asperity peaks.

In embodiments, the directions of roughness between the first and second surfaces are aligned to maximise asperity contacts there between.

In embodiments, the third body material is selected and conditioned to be stiff and strong enough to deform the second body (the heat sink) when either press (force) fitted or shrink fitted over the second body; and to have a coefficient of thermal expansion less than that of the second body (usually aluminium), suitable materials include metals or fibre reinforced metals or fibre reinforced non-metal composites.

Alternatively a composite material such as a fibre reinforced thermoplastic such as carbon fibre may be imbued with highly beneficial anisotropic features.

In embodiments, the polymer matrix is selected to provide thermal stability and low coefficient of thermal expansion. In embodiments the stiffness of the third body is varied with direction by adjusting the direction (lay) and the density (that means the number of the arranged fibres). Suitable reinforcement fibre's include carbon, glass or aramid. A suitable polymer is an epoxy, vinyl ester or polyester.

In embodiments, the second body is divided into two, the division sized to allow for the greater axial (lateral) expansion of the second body (the heat sink) over the first body (the heat source) upon warming the bodies.

In embodiments, the division of the second body is sealed by inserting a

compressible seal thereby to seal said fluid flow ducts.

In embodiments, the clamp is positioned and secured by press-fitting or heat shrinking. List of diagrams

Figure 1- a schematic diagram showing the components of a basic thermal-interface.

Figure 2-a schematic diagram showing an assembled section of a thermal-interface.

Figure 3- schematic diagram with a section cut away showing a radial flux electrical machine with a cylindrical thermal-interface made by the method herein.

Further description with reference to the diagrams

Figures 1 and 2 are schematic diagrams that illustrate the forming of a thermal- interface by the method; wherein surface roughness 5, 6 on bodies 1 , 8 is shown exaggerated to illustrate the nature of the surface roughness, especially the flat top asperities 9, 10. As a guide, the areas of the rectangular bodies shown are roughly equivalent to a 1mm square section of a larger thermal-interface.

Figure 1 shows three aligned pre-assembled sections of bodies 1 , 8, 11 ready for assembly by applying force 12 and 13 in Figure 2, wherein 1 is a section of a first metal body; 8 is a section of a second metal body and 11 is a section of a third body.

Said third body may be a metal, or a non-metal, or a composite with a polymeric matrix, a metal matrix composite or a carbon-carbon composite, or any other suitably reinforced fibre composite.

The first surface 5 on body 1 and the second surface 6 on body 8 have been prepared with arrays of plateaued asperities 5, 9, 10 to maximise interfacial contact between 5, 6.

Figure 2 is a post assembly view after the three bodies 1 , 16, 11 are pressed and held together by applying external clamping forces 12, 13 to create the thermal- interface at 15. It will be appreciated that although shown flat these bodies may equally be small incremental sections of a cylindrically shaped rotary machine.

By way of an example, Figure 3 illustrates the method applied to improve thermal cooling within a cylindrical electrical brushless machine wherein the interface 25 between an electromagnetic stator 22 and its mounting 28, 29 is a thermal-interface made by the method. Specifically Figure 1 illustrate three aligned sections of bodies 1 , 8, 11 , wherein 1 is a stack of bonded laminations 2 forming part of an iron core of an electro magnet, held compressed by force arrows 3a and b. This represents a small section of a typical laminated stator. For purposes herein a stator is an electromagnet with a wound copper or aluminium coil therein the coil wound on a stack of iron alloy laminations 2. Each lamination is insulated and usually bonded across their contacting faces and the stack may be seam welded at the outer face to minimise the risk of delamination when machining, thus the laminated stack 1 ideally behaves like a block of mild steel when machined.

The thermal-interface surfaces 5, 6 between a lamination stack 1 , and heat sink 8 are typically prepared by metal working processes that ideally provide a cross hatch pattern such as grinding or honing, preferred because they have minimal textural directionality. Shaping by boring, turning or milling will also provide micro roughness 4, 7 but roughness left by these processes tends to have directionality. Surfaces prepared by all the afore mentioned methods benefit from subsequent cold-working by burnishing with a roller or ball burnishing tool, shot peening or planishing will also flatten asperities and produce a regular array of plateaued contacts as shown at 5, 6. The amplitude of the micro-roughness 4, 7 is shown exaggerated. The combined roughness at the thermal-interface 15 in Figure 2 is shown less than roughness at 4, 7 to illustrate the interaction between the prepared surfaces 5, 6 as they are pressed together by external forces 12, 13. It will be appreciated that roughness 6 on the softer body 8 which is usually aluminium will deform more than the roughness on the harder material of the laminated steel body 1.

Cold pressure welding occurs when the oxide on the contacting deforming asperities are damaged to reveal highly reactive metal that will spontaneously weld upon coming into contact with other exposed reactive material. The introduction of the cold welding agent accelerates this process resulting in enlarged cold pressure welds.

For optimum heat transfer the spacing between asperities should be as uniform as possible, as shown with horizontal spacing distance 9 will be approximately equal to vertical spacing distance 10.

It is difficult to provide precise guidance on measured surface texture. Texture being the term used to describe combined micro and macro roughness. R a is the most common roughness parameter but this only provides short term micro roughness data, which can be misleading because roughness measurements can vary with direction. As a guide the amplitude, which means the peak to trough deviation about their mean, of the micro roughness, preferably should be less than 5 micron. This micro roughness is superimposed onto the macro roughness, which itself should preferably not exceed 25 micron deviation. The peaky roughness traces shown on interface 15 are typical of those observed with an industrial roughness profile-meter, which generally use a condensed horizontal scale that exaggerates the actual roughness. Without this lateral distortion the average inclined roughness angles are typically about 30 ° .

The actual thermal-interface consists of two pressed-together prepared surfaces 5 and 6. The first surface 5 is on the first body 1 which is typically the heat source, usually iron or steel. The second surface 6 is on the second body 8 which is made with a less strong but superior heat conducting material like an aluminium alloy that is sized to acts as a heat sink that absorbs heat from body 1 , smooths any transients then dissipates the heat to air or transfers it to a circulating coolant fluid.

Figure 2 shows softer body 16 clamped (sandwiched) between harder bodies 1 and 11 , the clamp force sufficient to deform body 16 shown bulging at 14 due to axial deformation caused by clamp force 12, 13, the first force according to the method. Upon heating the assembly the softer sandwiched body 16 expands faster than bodies 1 and 11 , raising the compressive force (pressure) across the clamped thermal-interface 15, thereby increasing thermal conductivity with temperature, this second increase in pressure is the second force according to the method.

If the surfaces 5, 6 have significantly differing hardness more asperity deformation occurs at the softer surface and it is beneficial to make the harder material smoother. If the roughness has directionality due to machining then the directions of roughness 4, 7 between the first and second prepared surfaces 5, 6 should preferably be aligned to maximise asperity contacts.

According to the method, one or both surfaces 5, 6 are wetted with a thin film of cold pressure weld agent sparingly applied by brush or spray onto clean (oil free) dry surface 5, 6, The fluid is surface active and rapidly spreads like a low viscosity cleaning solvent leaving a wetted layer only a few molecules thick, this layer is incapable of maintaining metal to metal separation at asperity contacts. Surplus agent should be allowed to drain off before pressing the faces together. Stray fluid may interfere with subsequent surface finishing (painting) and the assembled parts should therefore be solvent washed after assembly. The agent is safe to handle and is readily dispensed by hand or an automatic metered applicator onto a dry clean surface, preferably immediately after the surfaces has been lightly brushed with a mildly abrasive cleaning brush to ensure the surfaces are free of particulate matter such as dust or rust. The cold weld agent is an insulator and there is no evidence this will affect coil insulation, it is advisable to avoid necessary contact.

A suitable cold pressure welding agent is a 30C/s methylhydrogenpolysiloxane, a siloxane polymer carrying hydrogen side groups some of which are thought to be released as single atoms of hydrogen when trapped between pressed together plastically deforming metal asperities. The released hydrogen is believed to rapidly absorb into the first few molecular layers of plastically deforming asperities especially when surface oxide is disrupted. Hydrogen appears to momentarily lower the yield point of the deforming asperities increasing the amount of plastically deforming material resulting in enlarged cold pressure welds that increase thermal conduction. Upon cessation of deformation the hydrogen appears to diffuse out of the metal which recovers its previous strength.

Each cold pressure asperity weld is formed and held compressed between asperities under elastic compression due to forces 12, 13. If this compressive stress is relaxed the welds spontaneously fracture, allowing disassembly. Surface marks due to fractured asperity welds are very small and can only be seen under a microscope.

Detail 15 on Fig. 2 shows the abutting roughness profiles at the edge of the thermal- interface, wherein asperities on the second aluminium body 16 are more heavily deformed than those on the harder material of the first and third bodies 1, 11.

Figure 3 illustrates the method by way of an example with a cut away view 20 of an electrical machine wherein a high thermal conductance coupling is provided between a first body, a laminated stator 23 carrying electronically commutated windings 31, the principle heat source to be cooled and aluminium heat sinks, 28, 29 the second bodies in the thermal cooling path of the machine. The general arrangement of Figure 3 is typical of electronically switched brushless machines employing interactive magnetic flux fields; the first field is provided by fixed coils 31 that are electronically commutated and second typically a rotating assembly of permanent magnets 35. Machine details such as end plates, shaft bearings, tie rods (if used) between end plates, electrical and coolant connections etc. are omitted to simplify the diagram. A symbolic rotor 35 is included in the diagram for

completeness; the rotor shown is not representative of any particular rotor

construction or function.

Specifically, view 20 in Fig. 3 shows a machine with a segment cut away to reveal its internal construction, which incorporates a high conductivity thermal-interface 25 made by the method; in which, the first body is the laminated stator 23, with a first prepared surface 22 thereon; and the second body is an aluminium heat sink 24, which divides into 28, 29, at 26 to allow for the higher rate of axial thermal expansion of the aluminium of the heatsinks and with the second prepared surfaces 36, 37 within the bores of 28, 29 (heat sinks). The first and second prepared surfaces 22, 36, 37 treated with a thin film of cold pressure welding agent. Said bores of heat sink bodies 28, 29, sized to pass over the stator 22. The outside diameter of the heat sinks (second bodies 28, 29) sized to be an interfering fit with a third body 32 the clamp body. Upon forcing the third body 32 over the second bodies 28, 29, the heat sinks are compressed by body 32, bringing the second prepared surfaces 36, 37 into cold welding contact with the first prepared surface 22 on the first body 23 thereby forming a high conductivity thermal-interface 25. The third Body 32 is a stiff elastically extended band holding the first and second bodies in welded contact.

The coefficient of thermal expansion of heat sinks 28, 29 is selected to be greater than the sandwiching stator 23 and the interfering out cylinder 32, therefore upon warming the volume of heat sink 24 expands more than the bounding stator 23 and housing 32 causing the pressure at the thermal-interface 25 to rise with temperature. 24 is therefore divided at 26 into 28, 29 to allow axial expansion. A suitable elastomeric seal 19 is inserted into the split 26 to absorb the extruded differential expansion of 28, 29. Said seal 19 prevents coolant escaping from grooved ducts 30 formed in the outer surface of the heat sinks. When running the first, second and third bodies expand by slightly differing amounts, the overall expansion being limited by the stiffness of the outer clamping cylinder 32. Upon overload contact pressure within 32 rises due to the second aluminium body expanding more rapidly, which allows further cold pressure welds to develop and thermal conductivity rises The material of the third body 32 selected to be stiff with an elastic range greater than the thermal expansion range of the first body 23 (the stator). Body 32 also selected to be strong enough to constrain the radial thermal expansion of the aluminium heat sinks, forcing them to extrude axially. The clamp body 32 maintains the clamping force by elastic extension holding the thermal- interface together under all conditions. Upon cooling clamp body 32 thermally and elastically contracts by a sufficient amount to maintain the cooled thermal-interface in compression, thus newly formed cold pressure welds remain intact. Therefore any improved thermal conductance at the interface due to additional cold pressure welding will remain after cooling, providing the clamp 32 remains in elastic extension.

The method provides means of making a high conductivity thermal-interface between the first body 23 and second bodies 28, 29 made by either heat shrinking, press fitting, electromagnetic, explosive or hydraulic deforming or by some other suitable means of compressive forming to position a stiff elastically extended band 32 around the assembly to maintain the thermal-interface 25 in compression over an operating temperature range -40 to +250 ° C.

The material of the clamp 32 further selected first for its stiffness (yield strength) and second for its elasticity. Secondly the material of 32 is selected to withstand the radial thermal expansion forces of the first and second bodies 23 and 38, 39 as they expand. During positioning 32 extends elastically as it plastically deforms the second bodies 28 and 29 into compressive thermal coupling at 25 to create the high thermal conductivity coupling, the thermal-interface between the first body (the stator) 23, and second bodies (the heat sinks) 28 and 29.

In embodiments, suitable clamp materials 32 include alloys of steel, fibre reinforced metal matrix composites and non-metals such as carbon-carbon fibre composites. In embodiments, suitable steels include those conditioned for use in springs with high yield strengths for example ranging up to 1000 Pa (mega Pascal's) or more. Other suitable clamp metals include 301 stainless steel or commercial grade titanium or silicon carbide fibre reinforced titanium. In embodiments, the material of the third body is selected from a carbon fibre reinforced steel or aluminium matrix composites and carbo-carbon fibre materials whose stiffness, strength and thermal expansion properties are selected by varying the density, direction and the nature of the reinforcement fibres, including monofilaments, whisker or staple fibres particulates.

The outer ends of the divided heat sink 28, 29 are secured to either the housing 32 or optional steel tie rods (not shown) and the differential axial expansion of the heatsinks 28, 29 is absorbed by the seal 19 inserted in gap 26 between heat sinks. This limits the overall axial expansion of the machine to that of the steel housing.

Of particular concern in such machines is the need to limit the temperature of the insulated coating of the magnet wire 31 in the field coil windings of the stator 23, because magnet wire insulation suitable for winding electromagnetic stators is limited to about 250 C. Any excursion above this is likely to permanently degrade the structure of the insulation. Thus successive high temperature excursions may cause cumulative damage, resulting in a life limiting deterioration leading to premature failure of the insulation of windings 31.

In embodiments, the heat sink 24 (28 and 29) is sized to absorb and conduct heat and thereby 'regulate' or 'smooth's' transient peaks of heat into steady state heat flowing from the stator 23 across the interface 25 into the heat sink 28, 29; the smoothed heat flow is then transferred to a liquid coolant flowing in spiralled groves (ducts) 30, hence bodies 28, 29 also functions as a heat exchangers. The heat sink material is most commonly an aluminium alloy, but may include blends with carbon fibres, or copper or alloys thereof.

In a further embodiment a thin film of weld enhancement fluid is applied to surfaces 22 and/or surfaces 36, 37 so that upon assembly the film is trapped at the thermal- interface 25. The film strength of the thin film is selected to be low to ensure it is unable to hold the surfaces apart. Thus true metal to metal contact is established at asperity contacts. The cold pressure welding agent when trapped between two surfaces 22 and 36, 37 in asperity contact causes the asperity contacts to enlarge by additional plastic deformation due to the action of the enhancement fluid, the fluid becoming more active as contact pressure increases. As already explained, the additional deformation attributed to chemical-mechanical interaction between the deforming asperities and the thin film material.

At the ends of the stator body 23 the field winding end loops 31 are shown encased in a thermally conductive moulding 33, the moulding is typically a cured epoxy filled with thermally conductive particulate, which is standard current practice.

In summary, machine assembly Method starts by selecting a first body 23, the stator, which is made with a material selected for its electro-magnetic properties. A Typical material is silicon steel which unfortunately has relatively poor thermal conductivity, for example it is quoted as being in the range 20.5 to 25 W/mk compared with values of 43 for mild steel and 150 for annealed aluminium. Most of the heat to be removed is created by electrical resistive losses in the windings 31 and iron losses if the magnetic fields alternate (reverse direction as occurs in alternating current

operation). The aim therefore is to surround the stator with a tight fitting high thermal conductivity heat sinks 28, 29 that also acts also as a thermal conduits in which heat is transferred into a cooling fluid pumped through spiral groves 30 that flow into end chambers 21 and away from the machine (fluid entry and exits not shown).

In embodiments an electrical machine is provided wherein: i) the first body 23 is the stator, which is a stack of iron based laminations with copper or aluminium windings 31 therein forming an electromagnet whose field is arranged to rotate about the rotational axis 27 of the machine and with a first prepared smooth outermost surface 22 thereon; ii) the second body 24 is a high conductivity heat sink divided at 26 into 28 and 29 to allow for differential axial expansion and with prepared inner second surfaces 36, 37 the bores of the divided heat sink 28 and 29; 36,37 are sized to receive the first body 23 (the stator) as a free fit, and the outside of the second bodies 28, 29 sized to be an interfering fit with the third body 32; and iii) the third body 32 is a strong outer cylinder, sized elastically expand to

provide an interfering fit over the outer faces of the second bodies 28, 29. iv) The machine assembled by heat shrinking or press fitting the clamp 32 over the heat sinks 28, 29 to force the heat sink into high thermal conductance with the stator 23.

In embodiments, in the electrical machine, the coefficient of thermal expansion of the third body 32 is selected to be similar to the first body 23 and the coefficient of thermal expansion of the second bodies 28, 29 are similar and selected to be greater than the first 23 and the third body 32 and third body 32 selected to be strong enough to restrict radial thermal expansion of the second bodies 28, 29; and

In embodiments, upon warming the volumes of second bodies (heat sinks 28, 29), expand at near twice the rate of both the stator 23 and clamp 32. The clamp material is selected to be strong enough to withstand (constrain) said radial expansion of the heat sinks 28, 29 thereby forcing the heat sinks to extrude axially as shown in detail 14 in Figure 2. This second additional increase in pressure across the thermal- interface 25 is designated as the second pressure, which causes further asperity deformation that re-activates the cold pressure welding agent as the interface temperature rises and it enlarges the area of metal to metal contact between the first prepared surface 22 and second prepared surfaces 36, 37, thereby increasing thermal conduction between the first body 23 (the stator) and the heat sinks 28, 29 as interface temperature rises.