Active heat sink designs   

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application is related to U.S. patent application Ser. No. ______ (Docket No. Hernon 2-25-5-44-6) to Hernon, et al., entitled “Flow Diverters to Enhance Heat Sinks”, and U.S. patent application Ser. No. ______ (Docket No. Hernon 3-26-6-45-7) to Hernon, et al., entitled “Monolithic Structurally Complex Heat Sink Designs,” both of which are commonly assigned with the present application and hereby incorporated by reference as if reproduced herein in their entireties.

FIELD OF THE INVENTION

The present invention is directed, in general, to heat sinks.

BACKGROUND OF THE INVENTION

Heat sinks are commonly used to increase the convective heat transfer surface area of an electronic device to reduce the operating temperature of the device. A heat sink typically consists of a base and number of parallel fins or pins. A cooling fluid, typically air, is caused to flow over the heat sink to remove heat from the fins or pins, thereby cooling the electronic device.

SUMMARY

One embodiment is a heat sink that has a surface. A first active element is connected to the surface and configured to move from a first position relative to the surface to a second position relative to the surface. The heat transfer characteristics of the heat sink are altered by the movement.

Another embodiment is a method that includes providing a heat sink having a surface. A first active element is formed on the surface. The active element is configured to move from a first position relative to the surface to a second position relative to the surface different from the first position. The heat transfer characteristics of the heat sink are altered by the movement.

Another embodiment is a system that includes a device configured to produce heat and a heat sink having a surface in thermal contact with the device. The heat sink includes an active element connected to the surface and configured to move from a first position relative to the surface to a second position relative to the surface. The active element is configurable to change a direction of flow of a cooling fluid in response to the heat produced by the device.

BRIEF DESCRIPTION OF THE DRAWINGS

Various embodiments are understood from the following detailed description, when read with the accompanying figures. Various features may not be drawn to scale and may be arbitrarily increased or reduced in size for clarity of discussion. Various features in figures may be described as “vertical” or “horizontal” for convenience in referring to those features. Such descriptions do not limit the orientation of such features with respect to the natural horizon or gravity. Reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:

FIG. 1 illustrates a prior art heat sink;

FIGS. 2A-2C illustrates an embodiment of a bilayer active element;

FIGS. 3A-3C illustrates an embodiment of a bilayer active element configured to form an opening;

FIGS. 4A-4C illustrates an embodiment of an electrostatic active element;

FIGS. 5A-5C illustrates an embodiment a MEMS active element;

FIGS. 6A and 6B illustrates an embodiment of a heat sink with multiple active elements;

FIG. 7 illustrates an embodiment of a device and a heat sink with multiple active elements;

FIGS. 8A and 8B illustrates an embodiment of a heat sink with multiple active elements configured to form a channel;

FIGS. 9A and 9B illustrates an embodiment of a heat sink with multiple active elements;

FIGS. 10A and 10B illustrates an embodiment of a heat sink with multiple active elements configured to obstruct air flow;

FIGS. 11A and 11B illustrates an embodiment in which air flow through a heat sink in a system is increased by decreasing flow through another heat sink in the system;

FIG. 12 illustrates an embodiment in which active elements are configured to increase rate of air flow;

FIG. 13 illustrates an embodiment in which active elements move in a coordinated fashion to increase rate of air flow; and

FIG. 14 illustrates an embodiment in which an active element is configured to move cool air through a channel from one portion of a heat sink to another portion.

DETAILED DESCRIPTION

Embodiments described herein reflect the recognition that active heat sink elements may be used in heat sink designs to selectively alter flow of a coolant fluid within the heat sink to alter the heat transfer characteristics of the heat sink. The active heat sink elements may be actuated in response, e.g., to a transient heat output by an electronic device, or by a local thermal gradient within the heat sink. As discussed further below, active elements may be either commanded or uncommanded.

For brevity in the following discussion, air is discussed as the cooling fluid while explicitly recognizing that the invention can be practiced with other cooling fluids, including other gases and liquids.

FIG. 1 illustrates a prior art heat sink 100. The heat sink 100 includes a base 110 and fins 120. The heat sink 100 is characterized by structural uniformity. For instance, the fins 120 are about the same height and thickness, and are spaced from each other by substantially equal amounts. Even if an electronic device is in contact with only a portion of the heat sink 100, air flow through each space between the fins 120 is substantially equal. While some variations are known, such as use of round or elliptical pins, or variation of fin height, the heat sink 100 captures the general uniformity of prior art heat sinks.

FIGS. 2A-2C illustrate an embodiment 200 of a portion of a heat sink fin 210 and an active element 220 attached thereto. FIG. 2A shows a plan view, and FIG. 2B shows a view through section A-A of FIG. 2A. FIG. 2C illustrates the view through section A-A when the active element 220 is in an actuated position. As used herein, an active element refers to a heat sink element that is configured to move from one position relative to a surface of a heat sink to another position relative to the surface. Movement may be in response to, e.g., a change of temperature of the element, or by a command provided by an electrical signal. The surface of the heat sink comprising the fins 210, e.g., includes the surface of the fins 210 and the surface of a base to which the fins may be attached.

In the illustrated embodiment, the active element 220 includes a first metal layer 230 and a second metal layer 240. The first metal layer 230 has a greater coefficient of thermal expansion (CTE) than that of the second metal layer 240. As illustrated in FIG. 2C, the active element 220 deforms away from the fin 210 at a temperature determined in part by the thickness of the first and second metal layers 230, 240, the difference in CTE, and the zero-stress temperature of the active element 220. (The zero-stress temperature is discussed further below.) The active element 220 may then divert a flow 250 of a cooling fluid, e.g. air, from a path parallel to the fin 210.

The active element 220 is formed from two layers, each having a different CTE. While two layers are used to illustrate the embodiment, the principles discussed may be extended to three or more layers. When such a combination of layers is formed from two dissimilar metals, it is often referred to as a bimetal strip. However, such a combination of layers with different CTEs may be formed with other combinations of materials, such as, e.g., metal and polymer or polymer and polymer. Unless stated otherwise, the following discussion refers to a bilayer metallic (bimetallic) element while recognizing that other materials and more than two layers may be used.

In the context of active elements, actuation of such bimetal strips is referred to herein as uncommanded, as actuation occurs independent of any external command to the active element. In a bimetal strip, the differential CTE results in torque on the bimetal strip when the temperature is greater or less than a zero-stress temperature at which the bimetal strip assumes an a fully relaxed position. The bimetal strip will deform from its relaxed position at a temperature higher or lower than the zero-stress temperature.

In some embodiments one or more of the layers also have a thermal conductivity of about 200 W/mK or greater. Thus, when actuated the active element 220 may increase the convective heat transfer surface area of the heat sink in addition to diverting the flow of air. The layers may be metal layers bonded together by, e.g., cold rolling or electroplating. The metals may include those in the group copper, aluminum, copper, silver, and gold. In some cases, a material with a relatively poor thermal conductivity may be used, such as the metals titanium, steel or nickel, or even some polymers such as Kapton H®, manufactured by E.I. du Pont de Nemours and Co., Circleville, Ohio, USA. It may be desirable in such cases to minimize the thickness of the poorly conducting material so that the resulting active element has sufficient thermal conductivity to increase the convective surface are of the heat sink. In other cases, thermal conductivity may be less important, such as when the active elements 220 are used to obstruct air flow. In such cases, constraints on the thermal conductivity of the materials used to construct the active elements 220 may be relaxed.

In some embodiments, the active element 220 includes a copper layer and an aluminum layer. Copper has a CTE of about 17 E-6° C.−1, while aluminum has a CTE of about 24 E-6° C−1. Rolling of copper and aluminum to form a bimetal sheet is well known. The active element 220 may be cut or stamped from a formed sheet. The active element 220 may then be bonded to the fin 210 by soldering, welding, or bonding with an adhesive. When copper and aluminum are used, the active element 220 could be configured such that the first metal layer 230 is aluminum and the second metal layer 240 is copper so that the active element 220 deforms away from the surface of the fin 210 when actuated.

In some cases, the active element 220 may be configured to move into an air stream when the temperature of the bimetal strip exceeds the zero-stress temperature. Such a configuration may be desirable, e.g., to reconfigure air flow through a heat sink to provide greater air flow to another portion to accommodate a transient heat output. For example, the bimetal strip may be attached to a heat sink fin or base in such a manner that below the zero-stress temperature the bimetal strip remains flat against the surface it is attached to.

In other cases, the active element may be configured to project into an air stream when the active element is at its zero-stress temperature. The layers 230, 240 may be configured such that when the temperature of the active element 220 exceeds the zero-stress temperature, the active element 220 moves toward the surface of the fin 210. Such a configuration may be advantageous, e.g., to reduce air flow resistance in a path of the heat sink 200 as the temperature of that portion of the heat sink increases.

It should be noted that the active element 220 may be attached to a surface of a heat sink base with results that are similarly advantageous to those configurations in which the active element 220 is attached to heat sink fins. Moreover, the beneficial effects of the use of active elements may be realized when used with unconventional heat sink designs such as, e.g., those disclosed in U.S. patent application Ser. No. ______ (Hernon 3).

Movement of the active element 220 alters the heat transfer characteristics of the heat sink. For instance, diverting the flow 250 from a path parallel to the fin 210 may increase exchange of heat between the fin 210 and the air flow. Without limitation by theory, diversion of air flow may disrupt a boundary layer of air adjacent to the fin 210. When a fluid flows over a surface, a boundary layer of some thickness is formed. The boundary layer may have, depending on several factors, laminar, turbulent or transitional characteristics. In general, the laminar boundary layer is relatively less effective removing heat from the surface than other flow regimes. The thickness of the boundary layer is also thought to increase as distance of flow along the free-stream flow increases. It is therefore generally beneficial to reduce the thickness of the boundary layer to improve the rate of heat flow from the heat sink to the cooling fluid.

It is thought that the thickness of the boundary layer may be reduced by impinging a free-stream turbulent flow, e.g., the diverted flow 250, onto the laminar boundary layer. Such a reduction may be caused by generating any flow which sets up secondary flows normal to the surface that, e.g., compresses or thins the boundary layer to increase the heat transfer. Examples of these flow types include vortices and eddies, and transitional, turbulent, unstable, chaotic or resonant air flow. These aspects are discussed in greater detail in U.S. patent application Ser. No. ______ (Hernon 2).

Moving to FIGS. 3A-3C, illustrated is an embodiment 300 of a fin 310 with an active element 320. FIG. 3A shows a plan view, and FIG. 3B shows a view through section B-B of FIG. 3A. FIG. 3C illustrates the view through section B-B when the active element 320 is in an actuated position. The fin 310 is a multi-layer structure formed from at least an inner layer 315 and a first outer layer 317 and a second outer layer 319. A window 330 is formed in the second outer layer 319 by a conventional method such as, e.g., mask and etch. The active element 320 is separated from the inner layer 315 and the first outer layer 317 by a cut 340 formed, e.g., by stamping or other conventional method. Preferably at least one of the inner layer 315 and the outer layers 317, 319 are formed of a material with a thermal conductivity of about 250 W/mK or greater, e.g., Al, Cu, Ag, or Au. The inner layer 315 has a CTE different than the CTE of the outer layers 317, 319. Preferably the CTE of the first outer layer 317 is about equal to the CTE of the second outer layer 319. Thus, a portion of the fin 310 in which all layers, e.g., the layers 315, 317, 319, are all present would not deform since the torque developed by the CTE mismatch between the first outer layer 317 and the inner layer 315 is balanced by the torque developed by the CTE mismatch between the second outer layer 319 and the inner layer 315.

However, within the perimeter of the window 330, the CTE mismatch between the first outer layer 317 and the inner layer 315 causes the active element 320 to deflect when the temperature of the active element 320 changes. Again, the position of the active element 320 at a particular temperature is determined in part by the zero-stress temperature of the active element 320. When the CTE of the inner layer 315 is greater than that of the first outer layer 317, the active element 320 deflects in the direction of the first outer layer 317, as illustrated in FIG. 3C. Conversely, when the CTE of the inner layer 315 is less than that of the first outer layer 317, the active element 320 deflects in the direction of the inner layer 315.

When the active element 320 is not coplanar with the fin 300, as illustrated in FIG. 3C, the active element 320 may deflect air flow passing over the fin 310. In addition, an opening 360 is formed in the fin 310 through which air may pass from one side of the fin 310 to the other side of the fin 310. For example, in the illustrated embodiment air flow 350 flows adjacent to the first outer layer 317, and air flow 355 flows adjacent to the second outer layer 319. A low-pressure region may be formed downstream of the active element 320 adjacent to the first outer layer 317. Air from the air flow 355 may be diverted by the low-pressure region through the opening 360. In this way, air from a relatively cool region of the heat sink may be drawn into a relatively warm region of the heat sink to augment cooling of, e.g., an electronic device in thermal contact with the heat sink.

Turning to FIGS. 4A-4C, illustrated is an embodiment 400 of a fin 410 with an active element 420. FIG. 4A shows a plan view, and FIG. 4B shows a view through section C-C of FIG. 4A. FIG. 4C illustrates the view through section C-C when the active element 420 is in an actuated position. In this embodiment, an electrostatic field between the fin 410 and the active element 420 actuates the active element 420. The electrostatic field may be formed, e.g., by placing a first voltage potential on the fin 410 and a second voltage potential on the active element 420. A dielectric layer 430 between the fin 410 and the active element 420 serves to insulate the fin 410 form the active element 420. In an advantageous embodiment, the dielectric layer 430 is formed from a material having a relatively high thermal conductivity, such as Kapton® MT, also manufactured by E.I. du Pont de Nemours and Co. It may be desirable to form the dielectric layer 430 with a thickness small enough to provide sufficient thermal conduction from the air to the fin 410. Of course, other dielectric layers with lower thermal conductivity may be used, though the thickness of the dielectric layer 430 may need to be reduced accordingly.

The active element 420 may be formed by conventional methods. In a nonlimiting example, a sacrificial layer is formed over the dielectric layer 430. A portion of the sacrificial layer is removed to expose the dielectric layer 430. A metal layer is deposited onto the sacrificial layer and the exposed dielectric layer 430. The metal layer is selectively removed to leave the active element 420. The sacrificial layer is then removed. The portion of the active element 420 deposited onto the dielectric layer 430 adheres thereto, while the portion deposited onto the sacrificial layer is free to move.

A voltage potential may be placed on the active element 420 via a control line 440 formed over the dielectric layer 430. The control line 440 may be connected to a control system that may actively control actuation of the active element 420 in response to temperature of an electronic device connected to the heat sink. An active element that is actuated in response to an external command is referred to herein as a commanded element. In another embodiment, the control system senses temperature on the electronic device or in one or more regions of the heat sink via, e.g. one or more thermocouples or thermistors, and actuates one or more active elements 420 in response to the measured temperature. In this manner, e.g., active elements 420 in the heat sink may be selectively actuated to enhance heat flow in a portion of heat sink smaller than the entire heat sink.

FIGS. 5A-5C illustrate an embodiment 500 of a fin 510 with an active element 520. FIG. 5A shows a plan view, and FIG. 5B shows a view through section D-D of FIG. 5A. FIG. 5C illustrates the view through section D-D when the active element 520 is in an actuated position. The active element 520 is a portion of a micro-electrical-mechanical system (MEMS) device 530. The term “MEMs” as used herein includes electrical-mechanical systems built with feature sizes on the order of a few millimeters or smaller. The term includes devices sometimes referred to as nano-electrical-mechanical systems or mini-electrical-mechanical systems. The MEMS device 530 also includes a substrate 540 and control circuitry 550. A control line 560 provides a control signal to the control circuitry 550. The control line 560 may be electrically isolated from the fin 510 by a dielectric layer 570 when the fin 510 is formed of an electrically conductive material.

As is known to those skilled in the pertinent art, a MEMS device is typically formed using conventional and specialized semiconductor processing to form moving parts on a semiconductor substrate such as silicon. The moving parts may be integrated with control logic or other electronics on the substrate, and may be actuated using electrostatic fields, e.g. In some MEMS devices, such as micro-mirrors, a planar feature is attached to torsion springs so that the planar feature may be displaced from a rest position when actuated, and may then return to the rest position when not actuated. Actuation may be static, e.g., between two or more equilibrium positions, or dynamic, e.g., continuous motion between two limits at a frequency ranging from hertz to kilohertz or higher.

The active element 520 may be actuated, as illustrated in FIG. 5C, to a partial or full extent of its possible travel. As was described with respect to the embodiment 400, each of several active elements 520 may be actuated in response to a measured temperature in one or more portions of the heat sink or an electronic device attached to the heat sink.

Turning now to FIGS. 6A and 6B, illustrated is an embodiment 600 of a heat sink fin 610 on which multiple active elements 620 are attached. The fin 610 is attached to a base 630. FIG. 6A illustrates the active elements 620 in their rest position, and FIG. 6B illustrates the active elements 620 in their actuated position. The active elements 620 are configured such that an axis of rotation thereof is about parallel to the base 630. In other embodiments, the axis of rotation may be arbitrary with respect to the base 630. In the illustrated embodiment, the active elements 620 are bilayer elements, while other embodiments may employ electrostatic or MEMS elements, such as the active element 420 or the active element 520, respectively. The illustrated configuration may be desirable, e.g., to selectively increase available surface area of a heat sink fin without significantly increasing the pressure drop of an air stream across the heat sink. Thus, the heat transfer characteristics of the heat sink are altered by the movement of the active element 620.

FIG. 7 illustrates without limitation a heat sink 700 with fins 710 connected to a base 720. A plurality of active elements 730 are connected to the fins 710. An electronic device 740 is connected to the base 720. This example describes a case, e.g., in which the thermal conductivity of the heat sink 700 is relatively low, the lateral extent of the heat sink 700 is large, or both. Thus, heat from the electronic device 740 flows to a subset 750 of the fins 710 to a greater extent than to the fins 710 at the extremities of the base 720. In the illustrated example, the active elements 730 on the subset 750 of fins 710 are actuated in response to the temperature of the fins 710 in the subset 750. Thus, the surface area of the fins 710 in the subset 750 is increased while the surface area of the fins 710 outside the subset 750 is not. Because the axes of rotation of the active elements 730 are configured about parallel to the base 720, the active elements 730 present a small cross-sectional area to air flowing between the fins 710. Thus the surface area of the fins 710 is increased by the actuation of the active elements 730 without significantly increasing the pressure drop across the heat sink 700.

FIGS. 8A and 8B illustrate a portion of a heat sink 800 in which active elements 810 are placed on facing surfaces of fins 820. The active elements 810 are not actuated in FIG. 8A, and are actuated in FIG. 8B. As illustrated in FIG. 8B, the active elements 810 are configured to form channels 830 bounded by the active elements 810 and the fins 820 when actuated. Thus, the active elements 810 may be used to selectively configure the heat sink 800 to include channels to guide the flow of air. In some embodiments, the channels 830 are configured to redirect air from one portion of the heat sink 800 to another portion. In some cases, the air flow is redirected to cause air to flow from a cooler region of the heat sink 800 to a warmer region of the heat sink 800. In some cases, the active elements 810 are configured to provide openings between adjacent channels 830 to, e.g., enhance mixing of air between channels 830 or to increase vortices or eddies, or transitional, turbulent, unstable, chaotic or resonant air flow.

Turning now to FIGS. 9A and 9B, illustrated is an embodiment 900 of a heat sink fin 910 on which multiple active elements 920 are attached. The fin 910 is attached to a base 930. FIG. 9A illustrates the active elements 920 in their rest position, and FIG. 9B illustrates the active elements 920 in their actuated position. The active elements 920 are configured such that an axis of rotation of the active elements 920 is about perpendicular to the base 930. Thus, when the active elements 920 are actuated, they present a large cross sectional area to the air flow and thus act to obstruct or generate unsteady flow of air between adjacent fins 910.

FIGS. 10A and 10B further illustrate the embodiment in which air flow is obstructed by active elements. A portion of a heat sink 1000 is shown in which active elements 1010 are placed on facing surfaces of fins 1020. The active elements 1010 are not actuated in FIG. 10A, and are actuated in FIG. 10B. As illustrated in FIG. 10B, the active elements 1010 are configured to obstruct air flow between the fins 1020 when the active elements 1010 are actuated. Thus, the active elements 1010 may be used to selectively obstruct the flow of the cooling fluid in portions of the heat sink 1000. The illustrated configuration may be desirable, e.g., to selectively decrease the flow of the cooling fluid through a portion of the heat sink 1000 to, e.g., increase the temperature of an electronic device. The active elements 1010 may also be configured to obstruct air flow at a lower temperature and open an air path at a higher temperature. One heat sink may include some active elements 1010 configured to obstruct air flow as the temperature increases and others configured to open an air path as the temperature increases. Such example configurations may be used, e.g., as a component of a control system configured to maintain a temperature of a temperature-sensitive device such as a laser.

It should be noted that in general active elements will increase back pressure through a heat sink when the active elements project into an air stream. The increased back pressure has implications in systems design issues, as greater fan capacity may be necessary, with resulting greater power consumption and system heat dissipation.

FIGS. 11A and 11B illustrates an embodiment of a system 1100 in which heat sinks 1110a, 1110b, 1110c are reconfigured to dynamically redistribute air flow in a system. FIG. 1A illustrates the heat sinks 1110a, 1110b, 1110c mounted over a substrate 1120, e.g., a circuit pack. Air flow 1130 is shown flowing about equally through the heat sinks 1110a, 1110b, 1110c. The air flow 1130 through the heat sinks 1110a, 1110b, 1110c is coupled, meaning that a change of back pressure of one heat sink 1110 affects the flow of air to another heat sink in the same air flow local environment. For example, when air flow through two heat sinks 1110 is coupled, if the air flow through one heat sink is reduced due to an increase of back pressure through that heat sink, the air flow through the other heat sink increases. The heat sinks 1110a, 1110b, 1110c include active elements that in FIG. 11A are configured such as in FIG. 10A, e.g., to allow air to flow therethrough without obstruction.

FIG. 11B again illustrates air flow 1130 through the heat sinks 1110a, 1110b, 1110c. But in this case, the heat sinks 1110a, 1110c are configured such as in FIG. 10B, e.g., to obstruct the flow of air therethrough. This has the effect of causing air that would have flowed through the heat sinks 1110a, 1110c to be diverted to flow through the heat sink 1110b. Such a configuration may be desirable, e.g., when a transient power loading of a device cooled by the heat sink 1110b necessitates greater cooling by the heat sink 1110b.

Turning now to FIG. 12, illustrated is an embodiment in which active elements 1210 are configured to increase a rate air flow 1220 between heat sink fins 1230. The active elements 1210 are commanded by a controller (not shown) to move at a frequency and to an extent 1240 that is determined to increase the rate of air stream 1220. In some embodiments, asymmetry of the shape or the motion of the active elements 1210 is provided to enhance the operation thereof. For example, a streamlined edge 1250 may be formed to produce an asymmetrical aerodynamic drag on the active elements 1210 that may increase pumping effectiveness. In another example, the rate of motion of the active elements 1210 is greater in the direction of the air stream 1220, thus imparting greater momentum to the air stream 1220. In some cases, the rate and extent of movement is different at different velocities of incoming air. In some embodiments, the active elements 1210 may be actuated at frequencies ranging from hundreds to thousands of hertz.

FIG. 13 illustrates another embodiment in which active elements 1310 are attached to heat sink fins 1320. The active elements are configured to create a dynamic pressure gradient 1330 that increases air flow 1340 between the fins 1320. The active elements 1310 may be controlled by a controller (not shown) configured to cause the active elements 1310 to move in a cooperative fashion. In the illustrated embodiment, the active elements 1310 are shown at varying degrees of actuation that form an actuation pattern. The controller may cause the actuation pattern to translate in the direction of air flow to increase a rate of air flow. Such cooperative movement of the active elements 1310 is similar to the cooperative movement of cilia in biological systems to transport fluids in one direction. In a similar manner as the streamlining of the active elements 1220, the shape of the active elements 1310 may be configured enhance the movement of air in the direction of flow.

Now turning to FIG. 14, illustrated is an embodiment of a heat sink 1400. The heat sink 1400 is an embodiment of the use of an active element to actively transport air within the heat sink 1400. The heat sink 1400 has fins 1405 and a base 1407, and also includes a number of intake channels 1410 with inlets 1415 that are connected to a manifold 1420. The heat sink also includes a number of output channels 1430 with outlets 1435 that are connected to the manifold 1420. Within the manifold 1420 is an active element 1440 adapted to create a pressure differential in the manifold 1420. The intake channels 1410 are connected to a low-pressure region of the manifold 1420, and the output channels 1430 are connected to a high-pressure region of the manifold 1420. Other active elements as described in the various embodiments herein may also be present on the heat sink 1400.

The inlet channels 1410 are configured to draw bypass air (or other cooling fluid) from an upstream location of the heat sink 1400. In many cases, the air will be cooler at the upstream portion than at a downstream location. The cooler air is directed by the inlet channels to the manifold 1120 and through the output channels 1430. The outlets 1435 are configured to output the bypass air at a location of the heat sink 1400 downstream of the inlets 1415.

The intake channels 1410 and output channels 1430 may be internal to the fins 1405 and the base 1407, may be external thereto, or may be partially internal and partially external. In an example embodiment, the intake channels 1410 and output channels 1430 are formed as integral structures of the heat sink 1400 by investment casting, as described in application Ser. No. ______ (Hernon 3). The channels 1410, 1430 may be thereby formed as passages wholly within the heat sink fins 1405 and/or the base 1407. Alternatively or in combination to internal passages, conduits along the surface, e.g., of the heat sink may be used to route air to and from the manifold 1420. When dimensions of the heat sink 1400 allow, the conduits may be formed separately of, e.g., tubing, and attached to the heat sink in the desired configuration.

In some cases, the bypass air will be cooler than the air traversing the path between the heat sink fins 1405. When the air is output at the outlets 1435, the cooler air may mix with the air stream, thereby cooling the air stream to increase heat transfer from the fins 1405 in the vicinity of the outlets 1435. Even when the air is not cooler than the air stream, the air output at the outlets 1435 may disrupt boundary layer flow between the fins 1405. Because boundary layers generally insulate the fins 1405, disruption of the boundary layers may increase heat transfer.

The active element 1440 may be any movable element configured to produce a pressure differential in the manifold 1420 that causes air to flow from the inlets 1415 to the outlets 1435. In one embodiment, the active element 1440 is a synthetic jet device. Synthetic jets are familiar to those skilled in the pertinent art, and may include, e.g., a membrane or a diaphragm configured to move air from one portion of the device to another portion of the device. The membrane or diaphragm may be driven, e.g., electromagnetically or piezoelectrically. Such a jet may be manufactured in a compact form that may be integrated within the base, e.g., of the heat sink 1400.

Each of the various embodiments presented may be used singly or in combination with other embodiments as part of a heat sink design. Thus active elements may be combined in one heat sink that have, e.g., different active element sizes, orientations, and actuation temperatures. Furthermore, some active elements may be uncommanded, and some may be commanded. Some active elements may be configured to alter air flow through the heat sink. For example, the active elements may induce unsteady, unsteady laminar, transitional, turbulent, unstable, or resonant air flow. Such air flow may, e.g., reduce a boundary layer thickness. In other cases, active elements may be configured to divert air flow from one portion of the heat sink to another.

Although the present invention has been described in detail, those skilled in the art should understand that they can make various changes, substitutions and alterations herein without departing from the spirit and scope of the invention in its broadest form.