KINETIC HEAT SINK HAVING CONTROLLABLE THERMAL GAP

Abstract

An apparatus and method for cooling a heat-generating component includes a stationary base structure and a rotating structure. The stationary base structure is mountable at the first heat-conducting surface to or near the heat-generating component. The stationary base structure has a first and second heat-conducting surface to conduct heat therebetween. The rotating structure is rotatably coupled to the stationary base structure. The rotating structure has a heat-extraction surface facing the second heat-conducting surface across a spatial gap. As a result of the rotating structure rotatably moving, the rotating structure substantially transfers the heat from the second heat-conducting surface to a thermal reservoir communicating with the rotating structure. In addition, as a result of the rotating structure rotatably moving, at least two surfaces of the rotating structure and/or the stationary base structure generate a thrust and an opposing thrust to vary and/or maintain the spatial gap.

Description

TECHNICAL FIELD

The present invention relates to heat-extraction and dissipation devices and methods and, more particularly, kinetic heat sinks for use with electronic components.

BACKGROUND ART

During operation, electric circuits and devices generate waste heat. To operate properly, the temperature of the electric circuits and devices has to be within a certain limit. The temperature of an electric device often is regulated using a heat sink physically mounted near or to the electric device. FIGS. 1A, 1B, and 1C show several examples of conventional heat sink assemblies 100. The heat sink assemblies 100 generally extract heat 102 (see FIG. 1C) from the thermal source 104 (i.e., the electric circuit or device) to a heat sink 106, and then rejects (i.e., dissipates) the heat into a thermal reservoir 108, which is generally ambient air. To improve heat dissipation, the heat sink 106 may have fins 110 and be coupled with a fan 112 to increase the airflow across the fins 110.

One relatively new type of heat sink assembly, known as a “kinetic heat sink,” has a thermal mass with integrated fluid-directing structures (such as fins and/or fan blades) that rotates with respect to a stationary base mounted on or near the heated electric device. A fluid refers to a substance that may flow and includes a liquid, a gas, or a combination of liquid and gas. To the knowledge of the inventors, those in the art have developed several topologies for transferring the heat between the stationary base and the rotating heat sink.

A first type of kinetic heat sink design provides a thermal transfer medium, including thin films or various thermal fluids (such as oils), between the stationary base and the rotating heat sink. These designs undesirably often have problems relating to sealing, risk of leaks, or high frictional energy losses. To avoid this problem, a second type of kinetic heat sink design employs a fluid gap (e.g., air) between the rotating heat sink and the stationary base. Although air is generally considered a thermal insulator (i.e., not an efficient thermal transfer medium), when the fluid gap is made sufficiently small (i.e., in the order of tens of micrometers), the thermal resistance over the fluid gap may become negligible. Current designs of this type of kinetic heat sink known to the inventors, however, do not adequately control the noted small fluid gap. For example, the gap may not be properly regulated over a wide-range of rotational speeds or at arbitrary orientations. Moreover, kinetic heat sinks may have significant start-up resistances that impede effectiveness.

SUMMARY OF ILLUSTRATIVE EMBODIMENTS

In accordance with illustrative embodiments of the invention, a kinetic heat sink apparatus has a stationary member that includes a base structure having both a first heat-conducting surface and a second heat-conducting surface to conduct heat therebetween. The first heat-conducting surface is fixably mountable to a heat-generating component. The kinetic heat sink also has a rotating structure rotatably coupled with the base structure. This rotating structure has a movable heat-extraction surface facing the second heat-conducting surface across a spatial gap. The rotating structure is configured to transfer heat from the second heat-conducting surface to a thermal reservoir communicating with the rotating structure. The rotating structure and stationary member has at least two pairs of opposing surfaces configured to form a first thrust and second opposing thrust to maintain the spatial gap within a pre-specified range during rotation of the rotating structure. At least one surface of each of the pair of opposing surfaces may have at least one surface with hydrodynamic features thereon to generate the thrust.

Each pair of the two pairs of opposing surfaces may form a hydrodynamic bearing (e.g., a spiral groove bearing, a step bearing, a sector step bearing, a Rayleigh step bearing, a thrust inward-pumping bearing, a thrust outward-pumping bearing, and a herringbone thrust bearing).

The kinetic heat sink also may have an electric motor assembly. The stationary member may include stationary motor component fixably affixed thereto. The rotating structure may include a rotating motor component. The stationary motor component and the rotating motor component may include the pair of opposing surfaces to generate thrust.

The stationary motor component may include, or form, a heat insulating structure to thermally insulate the electric motor assembly from heat. The electric motor assembly may include cooling channels to direct fluid (air, gas, or liquid, etc.) through the assembly when the rotating motor component is rotatably moving.

The rotating structure includes a plurality of flow-directing members (e.g., fin or fan blade) to channel cooling fluid therethrough when the rotating structure is rotatably moving. The flow may exceed ten meters per second. The plurality of flow-directing members may fixably extend from a rotating base-plate. The rotating base-plate may define the heat-extraction surface, which forms the spatial gap with the stationary member. The spatial gap may vary between approximately 5 and 10 μm when the rotating structure is at rest and between 10 and 20 μm when the rotating structure is rotatably moving.

The stationary base structure and the rotating structure may be made of different thermal-conducting thermal material (e.g., copper, aluminum, silver, nickel, iron, zinc, and combinations thereof).

The kinetic heat sink apparatus may include a sensor to provide a feedback signal to a controller. The sensor may sense a distance of the spatial gap or a thermal quantity of one or more of the heat-generating component, the stationary base structure, and the rotating structure. The controller may vary the rotating speed of the electric motor assembly based upon the feedback signal.

In an alternate embodiment, the kinetic heat sink may include an active pump regulator to introduce fluid into the spatial gap. In another alternate embodiment, the kinetic heat sink may include an active spatial gap regulator. The regulator may generate thrust to maintain or assist in maintaining the spatial gap. The regulator may generate forces based on an electric field and/or magnetic field.

In accordance with another embodiment, the stationary base structure may form in part, or be part of, a vapor chamber and/or a heat pipe.

In accordance with another embodiment, a kinetic heat sink includes a base plate and an impeller. The base plate is adapted to fixably mount to a heat-generating component. The impeller is adapted to rotate in relation to the base plate across a gap of less than 20 micrometers. The impeller has an impeller plate and a plurality of fins extending therefrom. The impeller plate forms the gap with the base plate. The fins form channels therebetween.

The kinetic heat sink further includes a spindle motor adapted to cause the impeller to rotate. The spindle motor has a stationary portion affixed to the base plate and a rotating portion affixed to the impeller. Portions of the stationary portion and rotating portion form fluid-dynamic bearings. During the rotation of the impeller, the base plate conductively draws heat from the heat-generating component and transfers a substantial portion of the drawn heat across the gap to the impeller. The gap is regulated, in part, by thrusts formed in opposing direction by the fluid-dynamic bearings, and the impeller receives the transferred heat and rejects the transferred heat to ambient fluid communicating with the impeller by directing the ambient fluid through the channels.

The kinetic heat sink may be configured to operate with substantially low-friction between the heat-extraction surface and the second heat-conducting surface when at rest. In an embodiment, a portion of the heat-extraction surface may form a low-friction contact with the second heat-conducting surface via a low-friction insert disposed therebetween. The surfaces or insert may include a low-friction coating (i.e., having a friction coefficient less than 0.5). The low-friction surface may be non-abradable. Examples of coatings, includes ceramics, such as metal-doped ceramics, Teflon, and graphite (e.g., diamond like carbon (DLC)). The insert may form, or be part of, a roller bearing or a ball bearing. In another embodiment, the kinetic heat sink has a first portion of the heat-extraction surface forming the spatial gap with the second surface when at rest, while a second portion of the heat-extraction surface is in contact with the stationary base structure when at rest.

In accordance with another embodiment, a method of dissipating heat from an electronic device is described. The method includes providing a stationary structure having a first and second heat-conducting surface. The stationary structure is thermally coupled to the electronic device at the first heat-conducting surface to drawn heat from the electronic device. The stationary structure conducts the drawn heat from the first heat-conducting surface to the second heat-conducting surface. The method then includes rotating a rotating structure having a heat-extraction surface facing the second heat-conducting surface across a spatial gap. The rotation substantially transfers heat from the second heat-conducting surface to the rotating structure and rejects the transferred heat from the rotating structure to a thermal reservoir in communication with the rotating structure. The method then includes generating thrusts on the rotating structure as a result of the rotation of the rotating structure. The thrusts are in opposing direction to maintain the spatial gap within a pre-specified range.

The method may include varying the rotation of the rotating structure, including a rate of rotation, to maintain the spatial gap within a pre-specified range.

The method may also include measuring a thermal quantity associated with at least one of the stationary structure, the stationary structure, the spatial gap, and the electronic device and using the measured thermal quantity to vary the rotation of the rotating structure.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing features of embodiments will be more readily understood by references to the following detailed description, taken with reference to the accompanying drawings, in which:

FIGS. 1A-C show various examples of conventional heat sinks;

FIGS. 2A-2E schematically show examples of a spatial gap of an apparatus for dissipating heat according to a preferred embodiment;

FIG. 3A schematically shows the apparatus of FIGS. 2A-2E with a motor assembly according to an embodiment;

FIG. 3B schematically shows the apparatus of FIGS. 2A-2E with alternate embodiments of the motor assembly;

FIG. 4 schematically shows a detailed diagram of the motor assembly and apparatus of FIG. 3A according to an embodiment;

FIG. 5A schematically shows a cross sectional view of a kinetic heat sink according to an embodiment.

FIG. 5B schematically shows a perspective view of the kinetic heat sink of FIG. 5A;

FIG. 6A schematically shows a diagram of the fluid dynamic bearing of the kinetic heat sink of FIG. 5A in accordance with illustrative embodiments of the invention;

FIG. 6B schematically shows a picture of the fluid dynamic bearing of FIG. 6A disassembled;

FIG. 6C schematically shows a picture of a cross-sectional view of the rotor housing of FIG. 6B;

FIG. 6D schematically shows an exploded view of the fluid dynamic bearing of FIG. 6A;

FIG. 7A schematically shows cut-views of fluid dynamic bearings according to an illustrative embodiment;

FIG. 7B schematically shows a portion of a fluid dynamic bearing according to an illustrative embodiment of the invention;

FIG. 8A schematically shows an apparatus with hydrodynamic surfaces in the spatial gap according to another embodiment;

FIG. 8B schematically shows a diagram of rotating structure having included a hydrodynamic step-sector bearing;

FIG. 9 is a diagram of a pressure profile for the step-sector bearing of FIG. 8B;

FIG. 10A illustratively shows a diagram of stationary base structure having included a spiral groove hydrodynamic bearing;

FIG. 10B illustratively shows a zoomed-in image of FIG. 10A;

FIG. 11 schematically shows an exploded diagram of the kinetic heat sink of various embodiments of the invention;

FIG. 12A schematically shows an apparatus according to an alternative embodiment;

FIG. 12B illustratively shows a picture of one implementation of the apparatus of FIG. 12A;

FIG. 13 a method according to a preferred embodiment of the invention; and

FIG. 14 schematically shows a diagram of a resistance model of the apparatus according to the preferred embodiment of the invention.

DETAILED DESCRIPTION OF SPECIFIC EMBODIMENTS

Illustrative embodiments facilitate high-density heat transfer with a kinetic heat sink. Such embodiments provide more robust thermal control with fewer active components.

FIG. 2A schematically shows an example of a kinetic heat sink 200 (also referred to as “apparatus 200”) for dissipating heat 202 according to various embodiments. The apparatus 200 includes a stationary member, having a base structure 204 (also referred as a “stationary base structure 204”), and a rotating structure 206 that cooperate to transfer heat from a heat generating component. To that end, the stationary base structure 204 has a first heat-conducting surface 208 and a second heat-conducting surface 210 to conduct heat therebetween. The stationary base structure 204 may be mountable at the first heat-conducting surface 208 to a heat-generating component 212. The first heat-conducting surface 208 and the second heat-conducting surface 210 may form the footprint of the apparatus 200.

The rotating structure 206, which is rotatably coupled to the stationary base structure 204, has a heat-extraction surface 216 facing the second heat-conducting surface 210 across a spatial gap 218. The heat-extraction surface 216 may face some or all portions of the second heat-conducting surface 210. When moving, the rotating structure 206 rotates with respect to the stationary base structure 204 to transfer heat from the second heat-conducting surface 204 to a thermal reservoir 220 (e.g., the environment) communicating with the rotating structure 206. The term “thermal reservoir” (or “thermal energy reservoir” and “thermal bath”) refers to a fluid having a heat capacity intended for heat to flow. The rotating structure 206 may have additional flow-directing structures 214 (e.g., blades or fins) to reject the heat to the thermal reservoir 220.

As a schematic drawing, FIG. 2A and others simply show two planar surfaces 210 and 216. It is contemplated, however, that various embodiments may configure those surfaces 210 and 216 with non-planar surfaces, including grooves, channels, and recesses (shown in subsequent figures). Similarly, it is contemplated that the various embodiments may configure those surfaces 210 and 216 as an angled region 224 defined by angle 226 (as shown in FIG. 2B, for example). The angle region may be over a portion of the two planar surfaces 210 and 216 and may be located on the outer circumferential portion of the planar surfaces 210 and 216 (as shown) or, alternatively, in the inner circumferential portion. Various other planar topologies are also contemplated to be within the scope of the invention and are provided in FIGS. 2C-2E for illustrative purposes.

A In accordance with illustrative embodiments of the invention, during operation, surfaces of the rotating structure 206 and the stationary base structure 204 generate a (axial) thrust and an opposing (axial) thrust to vary and/or maintain the spatial gap 218 within a prescribed distance. As known by those in the art, thrust is a tangible force resulting from fluid flowing over a tangible surface. Such surfaces may be part of the heat-extraction surface 216 and/or the second heat-conducting surface 210. In an embodiment, the rotating structure 206 may be configured with features at differing portions that may generate different thrust vectors when the rotating structure 206 is rotatably moving. For example, a first portion of the heat-extraction surface and/or the second heat-conducting surface may generate the thrust in a first direction, while a second portion of the heat-extraction surface 216 and/or the second heat-conducting surface 210 may generate an opposing thrust in a second direction. The heat-extraction surface 216 and/or the second heat-conducting surface 210 may be configured with a dividing structure to isolate the first portion and the second portion. The rotating structure 206 and the stationary base structure 204 may include other structures, as discussed below, that include the surfaces to generate the thrust.

The rotating structure 206 transfers heat from the second heat-conducting surface 210 to the thermal reservoir 220 as a result of enhanced heat transfer properties resulting from the rotating structure 206 rotatably moving with respect to the stationary base structure 204. A portion of the heat at the second heat-conducting surface may dissipate directly to the thermal reservoir through convection and radiation mechanisms (i.e., thermal leakage). As such, the term “substantially transfers” refers to transferring a non-negligible amount of heat that is beyond the thermal leakage mechanisms. In certain embodiments, a majority of heat is transferred from the second heat-conducting surface 210 to the thermal reservoir 220 through the rotating structure 206. The enhanced heat transfer results, in part, due to the thermal resistance of the spatial gap 218 being reduced as a result of flow (e.g., turbulent) being developed in the spatial gap 218 from the rotating structure 206 rotatably moving. The heat transfer also depends on the size of the spatial gap 218. The spatial gap may be configured based upon the fluid medium and the gap distance to have a thermal resistance 25 percent or less of the total thermal resistance. The heat-extraction surface 216 and/or the second heat-conducting surface 210 may be configured with surface topographic features to induce localized turbulence and mixing. The flow may be viscous-dominated to generate hydrodynamic lift.

The spatial gap 218 may include a region characterized as a fluid gap. For example, the spatial gap 218 may be approximately 10 micrometers (μm) or less when at rest. The spatial gap 218 preferably is configured to remain within a pre-specified range when operating. For example, the spatial gap 218 may vary in distance between 10 μm and 20 μm when the rotating structure 206 is rotatably moving. In some implementations, a reasonable magnitude of the thermal resistance for the spatial gap includes 0.05° Celsius/Watt or 10° Celsius·cm²/Watt. This thermal resistance value, however, depends on the application. For example, for a small heat source (i.e., heat watt output), a higher thermal resistance value may suffice. For such small heat sources, a small kinetic heat sink may be employed. Conversely, for larger heat sources, a low thermal resistance may be preferred to prevent the temperature at the heat source from getting too hot. In general, a thermal resistance across the gap of 25 percent or less of the total thermal resistance should provide satisfactory results. The thermal resistance of the spatial gap is inversely proportional to surface area of the spatial gap.

To enhance fluid flow, the rotating structure 206 may include fluid-directing structures 214 including, for example, fins and blades, to reject heat into the thermal reservoir 220. The rotating structure 206 may be configured as an impeller or other members that radiate heat with an enhanced surface area.

In one embodiment, the rotating structure 206 includes a fluid-directing structure to direct air. Compared to a conventional heat sink (i.e., a block of metal with or without fins and possibly coupled with a fan blowing air on it), the rotating structure 206 may have a higher fluid velocity flowing through it, thereby enhancing its heat transfer properties. Stated differently, energy may be expended directly to pump fluid for heat transfer rather than to move the fluid. From testing, it was observed that the rotating structure 206 rotatably moving at a comparable rotation speed to a fan may have up to ten times higher air velocity with respect to the fins. It should be apparent to those skilled in the art that similar advantages are contemplated when various types of fluids are employed with the various embodiments.

Accordingly, due to its increased heat-dissipation density, the kinetic heat sink may have a smaller form factor as compared to conventional blower with heat sinks. This also means that the kinetic heat sink 200 may rotate at a slower rate, thus saving energy, while providing comparable thermal dissipation performance. Moreover, in addition to requiring less energy, illustrative embodiments also reduce unintended acoustics resulting from the rotation, and should reduce the accumulation of dust that typically gathers on conventional heat sink designs.

The kinetic heat sink 200 may be used with any of a variety of different heat-generating components 212. For example, the heat-generating component 212 may include, among other things, a circuit board having electric components generating and/or transferring heat thereon, a semiconductor device, a thermal mass (i.e., a heat sink), a power circuit including a power switch such as a IGBT, and/or an electric circuit packaging surface (ceramic or plastic).

In alternative embodiments, heat-extraction and heat-spreading devices may be employed between the first heat-conducting surface 208 and the second heat-conducting surface 210. Heat-extraction and heat-spreading devices are generally utilized to spread/transport heat from one area to other rather than to dissipate the heat. Heat-extraction and heat-spreading devices may include, for example, heat pipes and fluid pumping systems.

FIG. 3A shows another heat sink embodiment, this time identified by reference 300. This figure shows more specific components of the device than those shown in FIG. 2A, including a motor assembly 302, 304. The stationary base structure 204 has heat-conducting surfaces to conduct heat from the heat-generating component 212 to the rotating structure 206. The rotating structure 206 is rotatably coupled to the stationary base structure 204 though the motor assembly. The motor assembly includes a motor-stationary component 302 and a motor-rotating component 304. The motor-stationary component 302 may include a stator (i.e., electrical windings and armature) and a housing. The motor-rotating component 304 may include a rotor shaft and components attached thereon, including, for example, permanent magnets (in some embodiments).

The motor-stationary component 302 preferably is fixably coupled to the stationary base structure 204 and thus, may be considered part of the stationary member. The motor-rotating component 304 may be fixably coupled or coupled via a gear to the rotating structure 206. The motor-stationary component 302 and the motor-rotating component 304 preferably are generally concentrically located between the rotating structure 206 and the stationary base structure 204.

Any number of different motor configurations may be used. For example, the apparatus 300 may include a controller 306 to regulate the rotation speed of the rotating structure 206 by regulating the current or voltage provided to the electrical winding. In an illustrative embodiment, the electrical winding is part of the motor-stationary component 302. However, it is noted that it should be apparent to those skilled in the art that various motor topologies may be employed, including designs having the electrical winding being part of the motor-rotating component 304. The controller 306 may include a control circuit, a driver, and corresponding signal processing circuitries. The controller 306 may be mounted within or on the stationary base structure 204. The control circuit may be configured to provide pulse-width modulation, frequency, phase, torque, and/or amplitude control.

The apparatus 300 may also include a sensor 308 to provide feedback signals for the controller 306. The feedback signals may be based upon the speed, temperature, or clearance of the spatial gap. The speed may be of the rotation speed of the rotating structure 206 and/or of the motor. The temperature may be of the heat-generating component 212, the stationary base structure 204, the rotating structure 206, spatial gap 218 and/or the motor. The sensor 308 may be a capacitive-based sensor, a thermocouple, and/or an infrared detector and may output an electrical signal that is un-scaled or offset and merely have some correlation to the temperature value. It should be apparent to those skilled in the art that various controllers and control schemes may be utilized to regulate the heat dissipating apparatus based upon temperature, rotation speed, and clearance gap.

It should also be apparent to those skilled in the art that a portion of the motor-stationary component 302 (e.g., the electrical winding) may be placed in various locations that are concentric the axis of rotation. For example, FIG. 3B shows another example of an apparatus with a motor assembly according to an alternative embodiment. Rather than the motor assembly being located proximal to or near the axis of rotation, the motor-stationary component 302 (having the electrical windings) is located distally to the rotor axis. Similarly, it is contemplated that parts of the motor-stationary component 302 (e.g., electrical winding) may be located on top of the rotating structure 206 or within the stationary base structure 204.

It should be further apparent to those skilled in the art that various direct-current and alternating-current based motor may be employed. For example, examples of direct-current (DC) based motors may include brushed DC motors, permanent-magnet electric motors, brushless DC motors, switched reluctance motors, coreless DC motors, universal motors, and examples of alternating-current (AC) based motors may include single-phase synchronous motors, poly-phase synchronous motors, AC induction motors, and stepper motors. The motor assembly may include an integrated motor controller, such as a servo motor. The motor may operate based upon pulse-width modulation scheme or direct current control.

FIG. 4 schematically shows detailed diagrams of the motor assembly of FIG. 3A (now identified by reference number 400 in this figure). The motor assembly 400 rotates the rotating structure 206 with respect to the stationary base structure 204. In an illustrative embodiment, the motor assembly 400 includes a rotor 404 seated, via bearings 410, in a rotor housing 402. The rotor 404 is fixably attached to the rotating structure 206. The attachment may be at the fluid-directing structures 214 (as shown) and/or at a rotating base 222.

In an illustrative embodiment, the stator motor winding 406 forms a portion of the stationary base structure 204, and the permanent magnet 408 forms a portion of the rotating structure 206. The permanent magnet 408 may be attached at the rotating base 222 (as shown) and/or the fluid-directing structures 214.

The bearings 410 maintain the centricity of the rotor 404 with respect to the axis of rotation of the rotating structure 206. The bearings 410 may be mechanical-based bearings (including roller bearings or ball bearings) and hydrodynamic bearings.

Hydrodynamic bearings (i.e., self-acting bearings) provide a thin film of fluid (including gas, such as air, or liquid, such as oils and thin films) as lubrication. The thin film of fluid is created by the relative motion of two mating surfaces separated by a small distance. A subset of hydrodynamic bearings, where the thin film of fluid comprises, mainly, of gas (such as air), is referred to as an aerodynamic bearing. As the speed of the motion increases, a velocity-induced pressure gradient is generally formed across the clearance between the two mating surfaces. Hydrodynamic bearings generally include topographical features formed onto at least one of the mating surfaces. Among other things, the topographical features may be characterized as spiral grooves of a step bearing. Spiral groove bearings may be configured as a thrust inward-pumping bearing, a thrust outward-pumping bearing, and a herringbone thrust bearing.

In an illustrative embodiment, journal bearings maintain the centricity of the rotor 404, and the thrust control the gap thickness. Among other things, thrust forces may be generated by hydrodynamic bearings of differing types. For example, the apparatus may employ one type of opposing hydrodynamic bearings to control the overall gap thickness, and employ a second type of hydrodynamic bearing features along the surface of either the stationary base structure 204 or the rotating structure 206 for additional stiffness to prevent tilting of the rotating structure 206 and/or to provide stiffening/dampening force when there is a shock.

FIGS. 5A and 5B show another example of a kinetic heat sink/apparatus 500 according to an illustrative embodiment of the invention. The apparatus 500 includes a base plate spreader 502 adapted to mount onto the heat-generating component 212. A fluid-dynamic bearing spindle motor 504 is affixed to the base plate spreader 502 and a kinetic heat sink impeller 506 to rotate the kinetic heat sink impeller 506 with respect to the base plate 502. The kinetic heat sink impeller 506 includes the rotating base 222 (referred to as an impeller plate) and a plurality of flow-directing structures 214 (referred to as impellers). The plurality of flow-directing structures 214 are configured to direct air flow through channels formed therebetween when the kinetic heat sink impeller 506 rotates to dissipate heat into ambient air.

The fluid dynamic bearing spindle motor 504 drives the kinetic heat sink impellers 506 and simultaneously controls the spatial gap 218 (also referred to as a gap thickness) between the baseplate spreader 502 and the rotating heat sink impeller 506. The spatial gap 218 may vary to have a clearance of less than some predefined amount, such as about 20 μm, over a portion of the surfaces between the impeller plate 222 and the base plate 204 when the kinetic heat sink impeller 506 is rotating. The surfaces may be treated or configured with a low friction contact when the kinetic heat sink impeller 506 is at rest.

FIG. 6A schematically shows a detailed diagram of the fluid-dynamic bearing 600 within the fluid dynamic bearing spindle motor 504 or within other similar embodiments. The fluid dynamic bearing 600 includes an axial control component 602 and may include a radial control component 604. The axial control component 602 may be incorporated into the rotor 404, the rotor housing 402, and the rotor cap 610.

The axial control component 602 controls the position of the rotor 404 situated on or along the direction of the axis 612. The axial control component 602 may include a first control surface 614 to generate a thrust in a first direction 616, and a second (opposed) control surface 618 to generate a thrust in a second direction 620. The first control surface 614 may be a part the rotor housing 402. The second control surface 618 may be a part of the rotor cap 610. It should be apparent to those skilled in the art that the rotor cap 610 and the rotor housing 402 may be configured as a single unitary component. Clearance between rotor 404 and/or the rotor housing 402 and rotor cap 610 at the axial control component 602 may vary during normal operation. In an illustrative embodiment, the clearance may be about 5-10 μm or, more specifically, between about 2-4 μm. Again, the application may dictate the required clearance.

The radial control component 604 controls the clearance, and thus the position, of the rotor 404 with respect to the rotor housing 402 along the radial direction 622. The radial control component 604 may be a fluid bearing or a mechanical-based bearing. In an illustrative embodiment, two journal bearings 624, 626 (upper and lower) in the upper and lower regions of the rotor 404 and rotor housing 402 are employed. Clearance between rotor 404 and the rotor housing 402 may vary during normal operation. In an illustrative embodiment, the clearance may be approximately 10-20 μm or, in other embodiments, approximately 3-5 μm.

It should be apparent to those skilled in the art that the clearance values provided are merely illustrative. For example, clearance values greater than about 5-10 μm (for the axial clearances) may be used for larger scale applications where the apparatus is larger than a few inches, and for low cost cooling applications having generally high variability in tolerances.

FIG. 6B shows a picture of the fluid dynamic bearing 600 of FIG. 6A, including the rotor 404, the rotor housing 402, and the rotor cap 610. The rotor housing 402 includes the first control surface 614 having included a fluid bearing with herringbone groove patterns. The rotor cap 610 includes the second control surface 618, also having included a fluid bearing with herringbone groove patterns. The rotor 404, in this embodiment, includes a shaft portion 628 embedded with a ring-shaped portion 630 at one end. The shaft portion 628 and the ring-shaped portion 630 have included the journal (i.e., portions in contact with the bearing).

In illustrative embodiments, the rotor cap 610 is assembled to be flush between the stationary base structure 204 and the rotating structure 206 (see for example, FIGS. 5A, 5B). The rotor cap 610 may be fixably attached to the stationary base structure 204 (such as, for example, by adhesives or by a frictional fit). The rotor cap 610 may preferably be made of a thermally conductive material to provide a parallel heat path between the stationary base structure 204 and the rotating structure 206.

FIG. 6C shows a photo of a cut-away of the rotor housing 402 of FIG. 6B, specifically of the radial control component 604. The rotor housing 402 has included the fluid bearing 624 and 626 with herringbone groove patterns.

FIG. 6D shows an exploded view of an alternate assembly of the fluid dynamic bearing 600 of FIG. 6A. The rotor housing 402 includes a casing component 632 and a bearing component 634. As a result, the rotor 404 may be in direct contact only with the bearing component 634. The bearing component 634 may be made of a different material and/or manufacturing process than the casing component 632. For example, a lower cost manufacturing process may be employed, such as casting, to form the casing component 632. The through-hole may be machined to achieve a defined finish and/or tolerance. In contrast, the bearing component 634 may be fabricated with high tolerance techniques. Similarly, the bearing component 634 may be made, in part, of high strength and high toughness material to minimize its wear.

In alternate embodiments, the fluid dynamic bearing 600 may be configured with a heat insulating structure embedded between the motor-stationary component and the stationary base structure, thereby insulating the electric motor assembly from the heat. The rotor cap 610 made of a thermal insulating material (i.e., a ceramic) may form the heat insulating structure. The rotor cap 610 may extend across both the motor-rotating component and the motor-stationary component, thereby providing thermal protection to the motor assembly from the heat being transferred.

Similarly, the motor assembly may be configured with cooling channels to direct air through the motor assembly when the motor-rotating component is rotatably moving.

FIG. 7A shows cut-views of fluid dynamic bearings according to an illustrative embodiment of the invention. The fluid dynamic bearings include an axial control component and a radial control component, both of which may be incorporated into the rotor 404, rotor housing 402, and rotor cap 610. The fluid bearing is shown as being a part of the rotor 404, rather than of the rotor housing 402 and the rotor cap 610. The first control surface 614 is configured as a spiral groove bearing.

FIG. 7B shows a portion of a fluid dynamic bearing having a first control surface 614 according to an illustrative embodiment of the invention. The first control surface 614 is configured as a herringbone thrust bearing.

It is noted that in any of the above described embodiments, among other things, the second control surface may be configured as one of a spiral groove bearing and a herringbone thrust bearing.

Referring to FIGS. 5A and 5B, the fluid dynamic bearing spindle motor 504 may include thrust fluid bearings 516 to both generate the gap thickness 218 and to confine the gap thickness 218 to a finite rage (less than 20 microns) for a wide range of rotational speeds, impeller weight/shape, and orientation of the apparatus. The fluid dynamic bearings independently or in combinations with other fluid dynamic bearings may be configured to minimize the axial and radial movement of the heat-sink impeller 214.

The lift and/or thrust force generated by a hydrodynamic bearing, including fluid bearings, may be expressed as:

$\begin{array}{cc}F=K^{\prime }\frac{\mathrm{\mu \omega }\left(R_{\mathrm{outer}}^4-R_{\mathrm{inner}}^4\right)}{h^2}& \left(\mathrm{Equation}2\right)\end{array}$

where K′ is a dimensionless function of the bearing topology based upon the type of features, the number of features, and the thickness and depth of the features; R_outer and R_inner are the radius of the bearing features; h is the clearance gap; ω is the angular velocity; and μ is the viscosity of fluid and/or medium.

From Equation 2, it is noted that hydrodynamic bearings may form varying thrusts based on a gap h between the bearing surface and an opposing surface. Generally, the thrust (i.e., opposing pressure) increases as the gap becomes smaller due to the reverse relationship with the resulting thrust force F. As a result, hydrodynamic bearing may be characterized as having an inherent feedback control loop. When two opposing hydrodynamic bearing are employed, they may be arranged to maintain a pre-defined gap thickness.

Illustrative embodiments may use a number of different types of bearings, such as Rayleigh step thrust bearings, step-sector bearings, and spiral groove bearings (e.g., inward, outward, and/or herringbone). Hydrodynamic bearings may include Rayleigh step thrust bearing and spiral groove bearings. Rayleigh step bearings are generally angularly symmetric, and generate a repulsive force regardless of the direction of rotation. A step-sector bearing employs a non-angularly symmetric design (see, e.g., FIG. 12B) and generally generate a repulsive force when rotating in the intended direction. Despite this, the force generated by a step-sector bearing is generally greater than that produced by a Rayleigh step bearing. Step-sector bearings generally have a deep groove right after the land section, as compared to a Rayleigh bearing, to ensure that the pressure is equalized to the surrounding fluid pressure at that location.

Spiral groove bearings may be configured as a thrust inward-pumping bearing, a thrust outward-pumping bearing, and a herringbone thrust bearing. Inward-pumping bearings are generally characterized as having fluid drawn inward from the outer perimeter. Outward-pumping bearings are generally characterized as having fluid drawn from the inner diameter. Herringbone thrust bearings are generally characterized as having both inward pumping and outward pumping actions.

The fluid dynamic bearing spindle motor 504 may rotate the heat sink impeller at high speeds. In a test setup, fluid dynamic bearing spindle motor was rotated at up to 12,000 revolutions per minutes (RPM).

The fluid dynamic bearing may be configured to provide dampening and/or stiffening effects as well as to account for thermal expansion effects. In an alternative embodiment, the fluid dynamic bearing spindle motor 504 may include mechanical bearings, such as ball bearings and/or roller bearings.

The bearings may be used for both thrust and journal bearings (i.e., for radial alignment). The fluid dynamic bearings and the mechanical bearings may be used independently or in combination for either of the axial alignment and for thrust.

The baseplate spreader 204 may be composed in part of a copper and/or aluminum disk and/or a vapor chamber. Moreover, at least two journal bearings may be used to control tilting and radial displacement of the rotor in the fluid dynamic bearing spindle motor. In addition, two opposing thrust bearings, which may be spiral, groove, or herringbone journal bearings, may be used to control the axial positioning. The operating gaps between the rotor and the rotor casing in the fluid dynamic bearing spindle motor may be 2 to 3 microns in the radial direction, and 5 to 10 microns on each side of the thrust bearings. Any number of different fluids may be used, such as oil (e.g., ester oil), thin-film, ferro-fluid, and/or air may be used as the lubricant.

When operating, the thrust bearings in the fluid dynamic bearing spindle motor may provide an additional 5-10 micron gap. The thrust generated, and thus the resulting gap thickness, may depend in part on the operating characteristics of the heat sink impeller, including its rotational speed, effective weight, and orientation.

As described, the opposing thrust bearings provide an opposing thrust to confine the gap thickness to a pre-determined, preferably small range. The heat-sink impeller 214 may thus operate in arbitrary orientations over a wide range of operating speeds, as required for a robust heat-sink solution.

FIG. 8A shows a diagram of kinetic heat sink/apparatus 800 according to another preferred embodiment of the present invention. At least one of the heat-extraction surface 216 and the second heat-conducting surface 210 are configured with a plurality of surface features 802 configured to maintain the spatial gap 218 when the rotating structure 206 is rotatably moving. The plurality of surface features may form a hydrodynamic bearing (e.g., a Rayleigh step bearing and/or a spiral groove bearing).

FIG. 8B shows a diagram of rotating structure 206 having a hydrodynamic bearing, specifically, a Rayleigh step bearing 804. A Rayleigh step bearing is generally characterized as having topographic surface features of alternating land regions 806 and groove regions 808. A profile 810 of the Rayleigh step bearing is provided, where h₀ is the spatial gap 218 and S_h is the depth of the groove region.

Step bearings, such as Rayleigh step bearings, generally operate through the principle of fluid being compressed resulting from a motion between two parallel bodies. As the two parallel bodies move, the fluid is forced to move between the groove and the land regions; the forced fluid results in a force that may be configured to repel in either clockwise or counterclockwise direction of movement.

FIG. 9 is a diagram of a pressure profile 902 for the Rayleigh step bearing of FIG. 8B. The geometry of the Rayleigh step bearing 804 is characterized by B₁ (the length of the land region), B₂ (the length of the groove region), B (the length of the each step bearing), h₀ (the spatial gap between the rotating and the stationary structure), and h₁-h₀ (the depth of the groove region). In this example, at position 904, the pressure from the forced fluid is normalized to be zero. As the rotating structure 206 rotatably moves with respect to the stationary base structure 204 at speed U (910), a repulsion force is created by the compressed fluid, which linearly increases the pressure. The pressure reaches a maximum pressure, P_max, at the end of the groove region at position 906. Then, the pressure profile linearly drops back to the initial starting pressure, P₀, at position 908.

FIG. 10A shows a picture of the stationary base structure 204 having included a hydrodynamic bearing, specifically, a spiral groove bearing 1002 on the second heat-conducting surface 210. FIG. 10B shows a zoomed-in image of the stationary base structure 204 of FIG. 10A. With the rotating structure rotatably moving in a counter clockwise direction with respect to the stationary base structure 204, the spiral groove bearings 1002 between land regions 1004 act as an outward pumping bearing. The average groove depth may vary up to 7 μm.

Referring back to FIGS. 5A and 5B, when at rest, the fluid dynamic bearing may be configured to provide a very small clearance between certain portions of the baseplate 204 and the impeller plate 222. Some other portions of the opposing components 204 and 222, however, may have no such clearance at rest. If there is no clearance, then there is some degree of friction during startup, and also during operation (e.g., if subjected to a shock). Illustrative embodiments therefore minimize the contact region of the fluid dynamic bearing to other components in the spindle motor 504, effectively reducing start-up friction. This consequently improves the efficiency and the operating performance of the kinetic heat sink 200.

To that end, the apparatus 500 may include a low-friction insert 518. Although the fluid dynamic bearing spindle motor 504 may control the tilting of the heat-sink impeller 214 relative to the baseplate spreader 204 (as a baseplate spreader) during operation, the rotating and stationary surfaces of the heat sink may make contact at low rotational speed and/or when a large external shock is introduced. In addition, when the heat transfer surface area is sufficiently large, the contact may result in a high start-up power/torque in order to rotate the heat-sink impeller 214.

The insert 518 may be configured to minimize wear or damage and may lower the high start-up power/torque. The insert 518 may be made from a variety of low friction materials, and/or may be placed closer to the axis of rotation than shown (the figure shows it at the maximum diameter), provided that the tolerances are such that the rotating and stationary surfaces do not make contact anywhere except for the region where the insert is present. FIG. 5B shows a cross sectional view of the apparatus 500 of FIG. 5A.

The insert 518 or the surfaces surrounding the gap thickness 218 (e.g., the surface on the baseplate 204 and the surface on the impeller plate 222) may include a coating. The coating may be hard (i.e., non-abradable) and may provide a low-friction contact. The coating may be composed in part of at least electro-less nickel, a diamond-like coating, hardened steel, and/or dry film lubrication (e.g., tungsten disulfide, WS₂).

The low-friction insert may be placed either on the rotating surface and/or on the stationary surface such that a portion of the insert extends out with respect to either of the surfaces. When placed at a perimeter of a 4-inch diameter rotating structure, the insert may be configured to protrude up to some amount, such as about 7 μm from the underlying surface.

FIG. 11 shows an exploded diagram of the apparatus 500. Further shown are the components of the motor assembly 504, including the rotor 404, rotor cap 610, rotor housing 402, stator windings 406, permanent magnets 406, and a retaining cap 1102. In an embodiment, the retaining cap 1102 fixably retains the rotor 404 with the impellers 214.

FIG. 12A shows an apparatus 1200 according to an alternative embodiment. The apparatus 1200 includes the various aspects of apparatus 800 (FIG. 8A) and further includes an active control schemes to help regulate the spatial gap distance during operation and at start-up. Specifically, the stationary base structure 204 includes an output 1202 to introduce actively pumped fluid (including, e.g., air) into the spatial gap 218. The actively pumped fluid may be introduced at inlet 1204 and may be provided by a fluid pump 1206.

The actively pumped fluid may have a pressure as low as 0.5 psig. FIG. 12B shows a photo of the apparatus 1200 of FIG. 12A. The apparatus 1200 may be configured to introduce the actively pumped fluid only at the startup to reduce the frictional contact at the spatial gap 218. As a result, upon the heat sink impeller reaching a sufficient speed for the fluid bearings 802 to produce self-regulating thrust, the actively pumped fluid may be stopped. Similarly, the actively pumped fluid may be introduced during shutdown.

FIG. 13 shows an illustrative process of operating the kinetic heat sinks discussed above. In general, the process begins by providing the heat sources, such as a heat-generating component 212. The kinetic heat sink is secured to the heat generating component 212 (step 1302), which may be, for example, a package of an electronic device or a printed circuit board. Various types of securing and mounting mechanisms known in the art may be used for these purposes. Among other things, those mechanisms may include screws, clips (e.g., z-clip, clip-on), push pins, threaded standoffs, glue, thermal tapes, and thermal epoxies.

When at rest, the rotating structure 206 may be seated against the base structure 204 or on a low-friction startup surface (e.g., insert or coating). The spatial gap 218 may have a distance of 5-10 microns.

To begin cooling, the controller 306 energizes the motor assembly 302, 304 (step 1304) causing the motor-rotating component 304 and rotating structure 206 to rotate. The power may be in V_DC (e.g., 12V, 5V, etc), V_AC, or PWM. As the rotating structure 206 rotates, fluid in the spatial gap 218 and between the control surfaces 614, 618 begin to move. At some rotational speed, such as over 1000 RPM, the fluid movement causes (i) the first control surface 614 (e.g., of the rotor housing 402) to generate a first thrust thereby lifting the rotating structure 206 with respect to the base structure 204 and (ii) the second control surface 618 (e.g., of the rotor cap 610) to generate a second thrust. The second thrust is in an opposing direction of the first thrust. This first thrust and second thrust balances the rotating structure 206 at a location along the axial axis with the spatial gap between the desired range, such as between approximately 10 and 20 microns. At such spatial gap and rotation speed, the thermal resistance characteristic of the spatial gap 218 is reduced allowing for heat to more readily transfer from the base structure 204 to the rotating structure 206. If a change of speed or a sudden shock is introduced to the system, the opposing thrusts correspondingly increase, thereby attenuating or absorbing the shock.

While rotating, the fluid-directing structure 214 (e.g., impeller) also rotates causing the fluid in the channels between the fluid-directing structures 214 to move. As the fluid moves, heat from the fluid-directing structure 214 is rejected to the moving fluid and dispels into the thermal reservoir 108. As a result, heat is drawn from the heat-generating component 212, spread across the base structure 204, transferred from the base structure 204 to the rotating structure 206 across the spatial gap 218, and spread from the heat-extraction surface 216 to the fluid-directing structures 214.

FIG. 14 schematically illustrates this step by showing a diagram of a heat-flow model of the apparatus according to the preferred embodiment of the invention. The heat generated by the heat-generating component 212 is characterized as Q_chip. As heat is spread across the base structure 204, the resistance of the base structure 204 may be characterized as a R_{base, linear} and R_{base, spread}. The movement of fluid from the rotation of the rotating structure 206 may be model as a sheering loss that generates heat (Q_shear). The motor assembly 302, 304 also generates heat (Q_motor) from the rotation. The heat from the motor (Q_motor) is modeled to be a part of the base structure 204 to be transferred to the rotating structure 206. Heat from the sheering loss is introduced at the spatial gap (R_airgap). With laminar flow formed in the small (i.e., 10-20 um) distance of the spatial gap 218, the thermal resistance is sufficiently small allowing heat to transfer across the spatial gap 218 to the rotating structure 206. As heat is spread across the rotating structure 206, the resistance of the rotating structure 206 may be characterized as P_platten and R_fins. The rejection of the heat from the fluid-directing structures 214 to the fluid may be characterized as R_rejection. Parasitic heat loss and leakage may be characterized as R_leak.

Referring back to FIG. 13, at block 1306, the controller 306 determines whether to continue to cool the heat-generating component 212. This may be based on a control signal or power being applied to the kinetic heat sink. The controller 306 may vary the rotation speed of the motor-rotating component 304 or the power output thereto based on temperature (e.g., at the heat-generating component 212 or various components of the kinetic heat sink) or distance of the spatial gap 314. Of course, various control schemes may be employed to rotatably move the rotating structure with respect to the stationary base structure. The process concludes at step 1308, in which the kinetic heat sink is de-energized. To that end, the controller 306 may reduce power to the motor assembly 302, 304, or power may be removed to the controller. The rotating structure 206 slows to a stop forming a contact with a portion of the base structure 204 or de-energize to a low-friction state.

The embodiments of the invention described above are intended to be merely exemplary; numerous variations and modifications will be apparent to those skilled in the art. All such variations and modifications are intended to be within the scope of the present invention as defined in any appended claims.

Claims

1. A kinetic heat sink apparatus comprising: a stationary member comprising a base structure having a first heat-conducting surface and a second heat-conducting surface to conduct heat therebetween, the base structure being fixably mountable at the first heat-conducting surface to a heat-generating component; anda rotating structure rotatably coupled with the base structure, the rotating structure having a movable heat-extraction surface facing the second heat-conducting surface across a spatial gap, the rotating structure being configured to transfer heat from the second heat-conducting surface to a thermal reservoir communicating with the rotating structure,the rotating structure and stationary member including two pairs of opposing surfaces, at least one surface of each of the two pairs of opposing surfaces having hydrodynamic features thereon, the two pairs of opposing surfaces of the rotating structure and the stationary base structure producing respective first and second thrust bearings to maintain the spatial gap within a pre-specified range during the rotation of the rotating structure.

2. The apparatus of claim 1, wherein the stationary member includes a stationary motor member, the rotating structure includes a rotating motor member, the stationary motor member and rotating motor member forming a motor assembly configured to rotate the rotating structure when energized,the stationary motor component and the rotating motor component include at least one of the two pairs of opposing surfaces to generate at least one of the two thrusts in opposing direction during the rotation of the rotating structure.

3. The apparatus of claim 2, wherein the stationary member includes a heat insulating portion between the stationary motor member and the base structure to thermally insulate the stationary motor member.

4. The apparatus of claim 2, wherein at least one of the stationary motor member and the rotating motor member includes channels to direct cooling fluid therethrough.

5. The apparatus of claim 1, wherein at least a portion of the heat-extraction surface and at least a portion of the second heat-conducting surface form one of the two pairs of opposing surfaces to generate one the two thrusts in opposing direction during the rotation of the rotating structure.

6. The apparatus of claim 1, wherein the rotating structure includes a shaft portion including at least two surfaces, each of the two surfaces of the shaft portion being a part of each of the two pairs of opposing surfaces, the stationary member including two corresponding surfaces facing the two surfaces of the shaft portion, each of the two corresponding surfaces of the stationary member being a corresponding part of each of the two pairs of opposing surfaces, the two surfaces of the shaft portion and the two corresponding surfaces of the stationary member generating the two thrust bearings during the rotation of the rotating structure.

7. The apparatus of claim 1, wherein each pair of the at least two pairs of opposing surfaces form a hydrodynamic bearing, the hydrodynamic bearing including at least one member of the group consisting of a spiral groove bearing, a step bearing, a sector step bearing, a Rayleigh step bearing, a thrust inward-pumping bearing, a thrust outward-pumping bearing, and a herringbone thrust bearing.

8. The apparatus of claim 1, wherein the rotating structure includes a plurality of flow-directing members to channel ambient fluid to dissipate heat from the rotating structure, the flow-directing members including at least one member of the group consisting of fins and fan blades.

9. The apparatus of claim 1, wherein the spatial gap varies between approximately 10 and 20 micrometers during the rotation of the rotating structure.

10. The apparatus of claim 1, wherein the base structure and the rotating structure are comprised of different thermal conducting materials.

11. The apparatus of claim 1 further comprising: a sensor configured to generate a feedback signal associated with at least one of 1) the distance of the spatial gap and 2) a thermal quantity associated with at least one of the heat-generating component, the base structure, and the rotating structure; anda controller configured to vary a speed of rotation of the rotating structure based on the feedback signal.

12. The apparatus of claim 1 further comprising: an active pump regulator configured to introduce fluid into the spatial gap.

13. The apparatus of claim 1 further comprising: a spatial gap regulator configured actively vary the spatial gap based on an electromagnetic field.

14. The apparatus of claim 1, wherein at least one of the base structure and rotating structure includes a coating having a coefficient of friction less than 0.5, the base structure and rotating structure being in contact at the coating when the rotating structure is at rest.

15. The apparatus of claim 1, wherein at least one of the base structure and rotating structure includes an insert that separates a portion of the heat-extraction surface from a portion of the second-heat conducting surface when the rotating structure is at rest.

16. The apparatus of claim 1, wherein the base structure forms, in part, at least one of a vapor chamber, a motor, and a heat pipe.

17. A kinetic heat sink comprising: a base plate adapted to fixably mount to a heat-generating component;an impeller configured to rotate in relation to the base plate across a gap of less than 20 micrometers, the impeller having an impeller plate and a plurality of fins extending therefrom, the impeller plate forming the gap with the base plate, the fins forming channels therebetween; anda spindle motor configured to cause the impeller to rotate, the spindle motor having a stationary portion affixed to the base plate and a rotating portion affixed to the impeller,at least two portions of the stationary and rotating portions of the spindle motor forming fluid-dynamic bearings,during the rotation of the impeller, the base plate conductively receiving heat from the heat-generating component and transferring at least a portion of the received heat across the gap to the impeller, the gap being regulated, at least in part, by thrusts formed in opposing directions by the fluid-dynamic bearings, the impeller receiving the transferred heat and transferring the transferred heat to ambient fluid communicating with the impeller by directing the ambient fluid through the channels.

18. A method of dissipating heat from an electronic device, the method comprising: providing a stationary structure having a first and second heat-conducting surface, the stationary structure being thermally coupled to the electronic device at the first heat-conducting surface to draw heat from the electronic device, the stationary structure conducting the drawn heat from the first heat-conducting surface to the second heat-conducting surface;rotating a rotating structure having a heat-extraction surface facing the second heat-conducting surface across a spatial gap, rotating transferring heat from the second heat-conducting surface to the rotating structure and rejecting the transferred heat from the rotating structure to a thermal reservoir in communication with the rotating structure; andgenerating thrusts on the rotating structure as a result of the rotation of the rotating structure, at least two of the thrusts being in opposing directions to maintain the spatial gap within a pre-specified range.

19. The method of claim 18 further comprising: varying the rotation of the rotating structure, including a rate of rotation, to maintain the spatial gap within a pre-specified range.

20. The method of claim 19 further comprising: measuring a thermal quantity associated with at least one of the stationary structure, the stationary structure, the spatial gap, and the electronic device; andusing the measured thermal quantity to vary the rotation of the rotating structure.