This application claims the benefit of U.S. Provisional Application No. 60/591,254, filed on Jul. 26, 2004 and U.S. Provisional Application No. 60/560,382, filed on Apr. 7, 2004. The entire teachings of the above applications are incorporated herein by reference.
FIELD OF INVENTION
The present invention relates to a thermal management system, especially electronics cooling and computer arrangement, and more particularly, to high performance electronics thermal management device that achieves high heat dissipation rates and provides a compact arrangement in which the electronic components are disposed with maximized space and weight efficiency.
BACKGROUND OF THE INVENTION
High power electronic components continue to have an increasing demand for higher power dissipating waste heat generated during computer operating within a relatively confined space. A major thermal management challenge is to adjust the constantly increasing heat emission from the electronic components to the constantly reducing weight and size of the computer arrangement. Conventionally used electronics cooling materials and technologies often become beyond of demands of modern computer idea. The further computer miniaturization requires the development of fundamentally new thermal management technologies and materials able to meet a combined requirement of high thermal conductivity, miniature performance, high heat dissipating rate, and low cost.
A variety of thermal management techniques are known in the prior art. In general, the thermal management cycle for electronics comprises two phases: the heat removal phase and heat dissipation phase. The heat removal stage can be performed by the following heat transfer methods:
a conductive heat transfer in solids
a convection heat transfer in fluids
a mixed phase change—convection heat transfer
a heat transfer accompanied thermal electric effects
These heat removal methods are used in the prior art and provided with the corresponding design embodiments such as solid-state high conductivity heat spreaders (often coupled with heat sink), close-loop circulation systems for fluids, phase change heat pipes, and thermal electrical cooling devices. The purpose for all of the above listed heat removal methods is the dislocation of the heat energy from a heat emitting component to a heat dissipating device, which mostly is an air cooled heat sink. Functional and designing division onto the heat removal and heat dissipating parts becomes typical for the modern computer architecture because the currently used cooling material and technologies can no longer manage the increased heat flux being coupled in one thermal management device. Commonly used heat pipes in notebooks, heat spreaders, and recently developed liquid cooling system are examples of such a functional dividing.
Unfortunately, the currently used solutions are not free from the drawbacks. The solid-state conductivity heat removal devices are heavy and their thermal conductivity often is beyond of what is required. Another disadvantage of the commonly used aluminum and cooper heat spreaders is the CTE mismatch with CTEs of semiconductor materials by the factor 2-3. The liquid cooling close-loop systems are heavy, roomy, and expensive in manufacturing. Heat pipes are often bulky and fragile. Thermoelectric cooling devices are heavy and bulky; their energy consumption is in a range that puts them outside of battery-driven applications.
The finned heat sinks are the most usable heat dissipating devices in computers. FIGS. 1, 2A and 2B illustrates a prior art device. Typically, heat sink 10 includes a plurality of fins 12 that extend outward from the plate 11. Thus, air flow 14 from the fan (not shown) flows over and between the fins 12 and cools the individual components. An electronic device 15 disposed on the ceramic substrate 16 are mounted on the plate 11 of heat sink 10 through an interface material 13. Distribution of air streams within the finned structure is uneven due to turbulent nature of air flow and formation of the flow stagnation areas. Areas adjusting to plate 11 tend to form air trapped zones 17, which prevent the air circulation and heat exchange. As air is a very low thermal conductivity substance, the stagnation and trapping areas prevent heat removal from the plate 11, which is the hottest part of heat sink 10. Instead, the most intensive heat exchange takes place in the remote part of fins 12 having a significantly lower temperature relative to temperature of plate 11. As shown in FIG. 2B, the difference ΔT between fin temperature and air temperature is the driver for heat exchange. Since ΔT1 is lower than ΔT2, then general heat removal capacity of heat sink is reduced.
Conventional heat sink-fan tandems produce a large amount of high velocity air flow that is the main source of computer noise. A large amount of air pumped through the computer causes deposition of contaminations on the surface of electronic components that reduces heat exchange additionally. Use of air filtration devices for fan air movers is limited because of the air pressure developed by air fans do not meet the pressure drop conditions in the air filtration systems. A large volume air flows produced by fans are used in inefficient way in computer systems. A simple calculation made on the basis of energy balance shows that the same amount of waste heat energy can be removed from a computer by an air flow that constitutes only 4-6% of the flow produced by conventional fans. Thus, the heat accumulating capacity of air flow is used only on 4-6% in the thermal solutions of the prior art. To achieve such energetic efficiency a condition of a quasi—total entropy heat exchange must be provided. Under this condition temperature of exhausting air should be close to temperature of heat sink. In turn, this high efficiency condition requires use of high thermal conductive materials and heat exchangers with extremely high specific surface. Unfortunately, the prior art does not disclose solutions able to meet these requirements. Still another disadvantage of the common design heat sinks is that they occupy too much physical volume to be practical in extremely confined applications. In view of the foregoing, it would be highly desirable to provide a compact electronics cooling device with high heat dissipating capacity, low noise, low energy consumption, harsh environment protected, and a low vertical profile to insure its compatibility with compact electronic equipment.
It is a continuing need in the prior art to provide low thermal impedance, stress avoiding heat spreaders and substrates that are thermal-mechanical compatible with the semiconductor materials. Commonly used ceramic substrates have thermal expansion coefficients similar to that of the semiconductor chip. However, recent remarkable progress in the semiconductor industry has promoted larger heat emission. Under such conditions, ceramic substrate becomes the most thermal resistive component in the heat transmitting path “semiconductor-substrate-heat sink” because of relative low thermal conductivity of ceramics. With the condition of close thermal-mechanical properties, high thermal conductivity metal matrix composite materials can be used in a beneficial method of the direct chip attachment to the high conductivity heat spreader becomes practicable.
In order to satisfy requirements of low thermal expansion the variety of metal matrix composites have been developed. The U.S. Pat. No. 5,167,697 discloses W—Cu, W—Ag, Mo—Cu and Mo—Ag materials and substrates of these materials made by sintering powders of tungsten or molybdenum with following infiltration of sintered preforms with copper or silver alloys. The heterogeneous microstructure of such composite materials combines the specific properties of both its components—high thermal conductivity of copper and silver and low thermal expansion of tungsten and molybdenum. The substrates with relative low CTE are provided by use of aluminum-silicon carbide composite materials. However, the mentioned metal matrix composite materials do not provide exact harmony with the CTE of such semiconductor materials as silicon or gallium arsenide. Besides, the conventionally used sintering and die casting manufacturing technologies do not provide production of the net-shape, net-size components with fine details. Additional machining is requires for theses difficult-to-machine materials that adds to their cost. A manufacturing process is still in need able to produce net-shape and net-size parts and micro-parts of metal matrix composite with adjustable thermal-mechanical properties.
Generally, computer systems are comprised of a cabinet or housing, that contains a plurality of components or subsystems, such as processors, memory, power supply, video cards, audio cards, disk drivers, and the like. Each of these components generates some heat, and collectively, the computer system can be considered as a thermal device requiring a fine thermal management.
Further, individual components, such as a microprocessor, may produce significant heat emission in very local areas. This waste heat is typically dissipated by a heat sink mounted directly on the individual component. As computer systems have become more complex and powerful, the individual components generate more heat. Increasing the number and size of the fins located on the heat sinks has generally provided increased cooling. Unfortunately, as the number of fins has increased, airflow provided by the fans has been inadequate to penetrate the now relatively dense fin structure, limiting the ability to cool the component. Increasing air flow to a sufficiently high level has proven problematic because fans tend to be noisy, consume substantial electrical power and increase their size consuming valuable real estate in the computer housing. As consequence, modern computer systems have added a small auxiliary fan adjacent to the overheating component. Quantity of auxiliary fans in modern computers reaches 9 pieces at the time of composing this patent application. These auxiliary fans have provided some relief but they have created additional problems. For example, increase of the number of auxiliary fans adds to the cost and complicity of the computer system. Further, with the growing complexity of modern computer systems, more and more individual components require additional cooling capacity. As the number of components grows, the problem associated with installing additional auxiliary fans is compounded. A plurality of auxiliary fans create a plurality of separate air flows within compressed space of a computer case. This creates an additional problem of air flows matching, which often becomes an obstacle for further miniaturization of computer systems.
Conventionally, computer components, such as those of “desktop” computers, are disposed within an exterior housing such that various computer components are stacked one over the other in a generally horizontal manner within the housing. Depending on the system requirements, heat dissipation is controlled using typically air fan—heat sink thermal management structures. The fans take up valuable space within the housing, increase costs and noise, and are particularly susceptible to failure due to their mechanical nature. The fans and heat sinks also complicate the general layout of the components within the housing, hampering overall size of the computer. To resolve this problem, conventional arrangements have relied on a less compact layout, such the components, while taking up less space, still provide sufficient heat removal.
SUMMARY OF THE INVENTION
It is the object of the present invention to provide a low weight, compact, low vertical profile thermal management system for removing heat from electronic components. The thermal management system comprises of a plurality of “flow-through” type cooling devices, a source of filtered pressurized fluid suitable for use as a coolant, and a fluid delivering device that supplies the cooling devices with fluid. A preferred fluid is air, such as a filtered pressurized air. The cooling device is comprised of a high conductivity metal matrix composite heat spreader, a miniature heat sink, preferably of the same material, a permeable heat exchanger with high specific surface, and a closure that provides structural integrity of the cooling device.
It is the further object of the present invention to provide a source of pressurized filtered air for use in the flow-through cooling devices. The source of pressurized air (air port) is comprised of a miniature air mover, an air filter, an air distribution device, and a plurality of flexible hoses that connect the air port with a plurality of the flow-through cooling devices.
It is the further object of the present invention to provide a self-cleaning air filtering system in which the self-cleaning effect is provided by use of two air filters and a cyclical air flow directivity change. Within this system the filters are self-cleaning by turn by means of expulsion of the filter by exhausted air.
It is the further object of the present invention to provide a miniature air mover capable of overcoming the hydraulic resistance of the air filtration device and a plurality of the flow-through cooling devices.
It is the further object of the present invention to provide a Coefficient of Thermal Expansion (CTE) controlled heat spreader applicable for direct and non-direct chip attachment. More particularly, an improved thermal conductivity heat spreader with an adjustable CTE within 4.0-6.5 PPM is provided.
It is the further object of the present invention to provide a high conductivity, CTE controlled heat spreader that is integrated metallurgically with a ceramic substrate, on which an electronic device can be placed using conventional technique. Thus, an ideal thermal contact between the ceramic substrate and a heat removal structure is created, and use of the low thermal conductivity interface materials is avoided.
It is the further object of the present invention to provide an electronic device with an integrated cooling system within which a silicon die integrated metallurgically into the flow-through cooling structure to form a single whole thermally uninterrupted body.
It is the further object of the present invention to provide a flow-by type electronics cooling device that unites a CTE controlled high thermal conductivity heat spreader and a miniature metal matrix composite heat sink in one thermally interrupted item.
It is the further object of the present invention to provide a computer arrangement in which one or more electronic components can be arranged in any desirable order and orientation without concern of air flow directivity, components proximity, gravitation, and electro magnetic influence. New computer arrangement comprises the encapsulated modules with electronic components and the thermal management system on the basis the flow-through cooling devices and a supply of filtered pressurized air. New computer arrangement provides freedom in layout for electronic devices and protection from exterior contamination factors.
It is the further object of the present invention to provide micro composite and macro composite high thermal conductivity materials applicable for the electronics cooling devices.
It is the further object of the present invention to provide a method for manufacturing the components of “flow-through” cooling devices in high volume and low cost.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a cross-sectional view of a prior art device.
FIG. 2A is a more detailed cross-sectional view of a single set of fins of the prior art device of FIG. 1.
FIG. 2B is a graph demonstrating the difference ΔT between the fin temperature and the air temperature at the location joined by the dotted lines with FIG. 2A.
FIGS. 3A and 3B depict of cross-sectional and top views, respectively, of one embodiment of a micro heat spreader of the invention.
FIGS. 4A and 4B depict of cross-sectional and top views, respectively, of one embodiment of a macro heat spreader of the invention.
FIGS. 5A and 5B depict of cross-sectional and top views, respectively, of an alternative embodiment of a macro heat spreader of the invention.
FIGS. 6A and 6B depict of cross-sectional and top views, respectively, of another alternative embodiment of a macro heat spreader of the invention.
FIG. 6C is a graphic depicting the thermal conductivity range of embodiment of the invention in FIGS. 6A and 6B.
FIG. 7 depicts an alternative arrangement of a macro composite heat spreader of the invention with a single flexible heat connector.
FIG. 8 depicts another alternative arrangements of a macro composite heat spreader of the invention with a dual flexible heat connectors.
FIG. 9 depicts one embodiment of a flow-through cooling device which unites a heat spreader, a heat sink and a heat exchanger.
FIG. 10A depicts a cross-section across the air flow of one embodiment of the invention.
FIG. 10B depicts a cross-section along the air flow of one embodiment of the invention.
FIG. 11 depicts one embodiment of the invention having a single body with uninterrupted heat flux.
FIG. 12 is a bar graph representing different performance of commonly used air fans and sources of pressurized air.
FIG. 13 is a graph comparing the air supply requirements for different cooling solutions.
FIG. 14 depicts Heat dissipation versus air flow of a finned heat sink versus a Flow-through heat sink of the invention.
FIG. 15 is a graphic which compares pressure and device size versus specific heat exchange surface for three different device combinations (i) fan plus heat sink, (ii) vacuum pump plus flow-through heat sink and (iii) compressor and flow-through heat sink.
FIG. 16 is an illustration of a flow-by heat sink of a micro composite material according to one embodiment of the invention.
FIG. 17 is an illustration of a flow-by micro composite heat sink of the invention having a spreader and a plurality of primary fins.
FIG. 18 is an illustration of a flow-by brush-type heat sink of the invention.
FIG. 19 is a nested and cut away views of a macro composite article of the invention.
FIGS. 20A and 20B are a front view and a cross-section of the CTE controlling core shown in FIG. 19.
FIG. 20C is a graphic depicting the relative magnitude and directivity of compressive and tensile stresses of an embodiment of the invention.
FIG. 21 is a cross-section of an integrated substrate of one embodiment of the invention.
FIG. 22 illustrates an arrangement of an integrated substrate such as in FIG. 21 couple with a flow-through cooling device of the invention.
FIG. 23 shows a schematic of an exemplary general-purpose computer arrangement in which the flow-through cooling device embodiment of the invention is employed (vertically oriented).
FIG. 24 shows a schematic of another exemplary general-purpose computer arrangement in which the flow-through cooling device embodiment of the invention is employed (horizontally oriented).
FIG. 25 shows a schematic of an exemplary general-purpose computer arrangement of the invention in which the vacuum pump acts as an air mover (vertically oriented).
FIG. 26 shows a schematic of another exemplary general-purpose computer arrangement of the invention in which the vacuum pump acts as an air mover (horizontally oriented).
FIG. 27 is a graph which depicts a typical surface area and heat load distribution in a cooler of the present device.
FIG. 28A-D depict additional embodiments of the device. FIGS. 28A and B show cut away views of devices where the CTE is present or absent and where the CTE has the same dimensions. FIGS. 28C and D illustrate an embodiment where the heat spreader and pedestal are used.
FIG. 29 depicts an CTE controlled substrate according to one embodiment of the invention.
FIG. 30 illustrates an arrangement of an CTE controlled internally cooled substrate in which the functions of CTE controlling and heat removal are provided by the same component.
FIG. 31 illustrates an arrangement of an CTE controlled internally cooled substrate in which the functions of CTE controlling and heat removal are provided by separate components.
DETAILED DESCRIPTION OF THE INVENTION
A description of preferred embodiments of the invention follows.
Illustrative embodiments of the thermal management system for electronics and computer arrangement according to the present invention are shown in FIGS. 3-15. As will be readily apparent to those skilled in the upon a complete reading of the present application, the present cooling device and electronic system arrangement are applicable to a variety of computer systems (such as servers, personal computers, notebooks, and the like) other than the embodiment illustrated herein, and moreover to electronic and electrical devices other than computer systems, including, but not limited to, power supply, plasma TV, automotive electronics, airborne electronics, and the like. Several non-limiting embodiments are provided and described in more detail herein. Embodiments include, but are not limited to, (a) high conductivity micro composite heat sinks, (b) graphite fiber brush-type heat sinks, (c) CTE controlled substrates and heat spreaders, (d) heat spreaders with flexible heat cable(s), (e) micro composite impingement cooling device, (f) macro composite flow-through heat sink, (g) macro composite super conductivity CTE controlled cooling device, (h) air station enabling computer architecture and combinations thereof.
The present invention provides a high thermal conductivity micro composite and macro composite heat spreaders that absorb waste heat generated by an electronic component distribute heat onto larger surface area, and transfers heat to a heat dissipating structure. According to the present invention, the heat spreader can be provided as a separate item or in a combination with a heat dissipating unit. Turning to FIGS. 3A and 3B, a heat spreader 20, preferably a micro composite heat spreader, is shown. A heat emitting electronic component 21 is mounted on the surface of a heat spreader 20 through a conventionally used interface material. A macro composite material 19 used in a heat spreader 20 is another embodiment of the present invention and disclosed in more detail in the following corresponding example in this description. In a preferred embodiment, the thermal conductivity of micro composite substrate constitutes 450-550 w/mK that provide rapid and uniform heat distribution within heat spreader 20. CTE in micro composite heat spreader is close to that of typically used silicon semiconductor material typically in the range of 5-9 PPM, preferably about 6 to about 8 PPM. Both thermal conductivity and CTE in micro composite heat spreader are isotropic.
A heat spreader, preferably macro composite super thermal conductivity heat spreader, is shown on FIGS. 4A and 4B. In the embodiment shown in FIG. 4, a macro composite heat spreader comprising a micro composite material 19 and inserts 23, preferably pyrolitic graphite inserts, having thermal conductivity in the range 1,400-1,700 w/mK at that micro composite material 19 which serves a matrix phase in the macro composite heat spreader. Micro composite material 19 envelops inserts 23 and provides structural integrity of macro composite spreader. Inserts 23 are oriented within heat spreader in such a way that their most heat conducting axis's coincide with directions of heat flow 25 from electronic component (Z direction) and direction of heat flow within heat spreader (X direction). Volume of inserts 23 can reach 85% in the most thermally loaded areas, which coincide with the projecting area of electronic component 21. Thus, bulk thermal conductivity in these areas can reach 1,200 w/mK. Other areas of macro composite heat spreader are less saturated with pyrolitic graphite inserts 23, and the most remote areas consist of micro composite material 19 only and provide local thermal conductivity within 450-550 w/mK. In this embodiment, CTE within macro composite heat spreader is non-uniform and depends on the volume of pyrolitic graphite inserts 23. The most CTE-sensitive area within location of electronic component 21 has CTE that is very close to that of electronic component, for example around about 3 to about 5 PPM, preferably 4 PPM. Thus, this macro composite heat spreader unites two important advantages in one product: the highest known thermal conductivity and CTE matched with semiconductor materials.
Another arrangement for macro composite heat spreader according to the present invention is shown on FIGS. 5A and 5B. Macro composite heat spreader comprises of a graphite fabric insert 26 of a pitch-based fiber and a micro composite material 19. Micro composite material 19 envelops the graphite fabric insert 26 and provides structural integrity for macro composite spreader. In a preferred embodiment, the graphite fabrics used in the present invention have thermal conductivity in the range of about 600 to about 1,100 w/mK that provides a good bulk thermal conductivity for macro composite heat spreader. Highest thermal conductivity in macro composite heat spreader coincides with direction of tows and rows in graphite fabric insert 26. Heat transfer in the transverse direction in macro composite heat spreader is provided by the micro composite material 19 having thermal conductivity of about 450 to about 550 w/mK.
Still another arrangement for macro composite heat spreader according to the present invention is shown on FIGS. 6A and 6B. Macro composite heat spreader comprises of a net of pitch-based graphite fiber 27 and a micro composite material 19. Micro composite material 19 envelops graphite fibers 27 and provides structural integrity of macro composite heat spreader. Graphite fibers used in this embodiment of the present invention have thermal conductivity in the range of about 600 to about 1,100 w/mK and CTE that is close to negative value. The bulk CTE and thermal conductivity depends on volumetric ratio between micro composite material 19 and graphite fiber 27. In this embodiment, the volume of graphite fiber is different in different areas of heat spreader. Maximal fiber saturation takes place in the heat collecting zone that coincides with the projecting area of electronic component 21. Volumetric content of fibers in this zone constitutes approximately 80% of that which provides bulk thermal conductivity approximately 900 w/mK. In the heat spreading area beyond of the heat collecting area the bulk thermal conductivity gradually falls to about 450 w/mK (FIG. 6C).
Further another arrangement for macro composite heat spreader 22 combined with a flexible heat connector 28 is shown on FIGS. 7 and 8. The flexible heat connector 28 is made of a high conductivity graphite fiber or fabric, which combines good heat transferring capacity with flexibility. Macro composite heat spreader 22 comprises a graphite fabric insert 26 enveloped by a micro composite material 19. Graphite fabric insert 26 is exposed to sides and continued to form a heat connector 28, preferably flexible heat connector. In one embodiment, waste heat 29 a is transferred to a desirable destination by the high conductivity heat connectors 28 and attached to a heat dissipating device using fasteners 29 made of a micro composite material 19.
Suitable materials for manufacturing the coolers of the invention include graphite foam (about 70 w/mK), pitch graphite fabric (up to about 1,100 w/mK), and pyrographite (1,400 w/mK). The conduits for controlling fluid flow can be made of, for example, steel, PVC or other material. The device can be encapsulated in a shell made of, for example, a macrocomposite material.
Turning to FIG. 9 a flow-through cooling device 30 of the present invention is shown. The flow-through cooling device 30 unites a heat spreader 32, a heat sink 33, and a heat exchanger 34, preferably a permeable heat exchanger, into one, low vertical profile and thermally uninterrupted body (an exploded view is shown on FIG. 9 for clarity). A component to be cooled 31, preferably an electronic component, is mounted onto heat spreader 32 using conventional technique and interface materials. High density heat load emitted (shown by arrows) by the electronic component 31 comes to the heat spreader 32 that is made of a high thermal conductivity metal matrix composite material according to the present invention. The heat spreader 32 is significantly larger than electronic component 31. Therefore the specific heat load in heat spreader 32 is reduced by several times comparing to the heat load in electronic component 31. Heat spreader 32 is made of high thermal conductivity composite materials according to the present invention; due to high thermal conductivity heat rapidly stretches within heat spreader 32.
While the relative sizes of the components can vary substantially depending upon the materials and fluids selected and the volume of air flow employed, in one example, the electronic component can have a surface area of about 50 mm2; the heat spreader can have a surface area of about 5 cm2; the micro heat sink can have a surface area of about 120 cm2; and the permeable heat exchange of up to about 50 m2 or more.
Further heat comes to heat sinks 33, for example a net of miniature heat sinks, which are metallurgically connected to the heat spreader 32 forming a single body with uninterrupted heat flux (FIGS. 10A, 10B and 11). Heat sinks 33 have extended thermal contacts with permeable heat exchanger 34. Heat transfer path continues further within permeable heat exchanger 34. A cooling fluid 37 enters through an inlet channel (see arrow on FIG. 11) is forced through permeable heat exchanger 34 using inlet channel 38 and outlet channel 39. The cooling fluid 37 can be in form of either gas or liquid. Cooling fluid 37 provides intensive heat removal from the permeable material of heat exchanger 34 and transfer of heat outside of the flow-through cooling device. Further the cooling fluid 37 now hot fluid as it exits the outlet channel (see exiting arrow on FIG. 11) can be collected and transferred either to an exhausted point (typically used in air embodiments), or to be returned for next cooling cycle (typically used in liquid embodiments). A heat exchanger 34 is lodged into a closure 40, preferably a hermetically sealed closure, which provides structural integrity of cooling device 30 and prevents leakage of cooling fluid 37. Additionally cooling device 30 is provided with a fluid delivering and fluid exhausting systems through inlet 44, outlet 45, and connecting hoses (not shown).
In one embodiment, the present invention provides a source of filtered and pressurized air for the flow-through cooling devices. The principal advantage of the pressurized air supply is that it enables miniaturized heat exchangers with extremely high heat dissipating capacity, but having a significant hydraulic resistance and pressure drop. By this, use of pressurized air enables basically new electronics cooling concept, one core principle of which is the use of miniature permeable heat exchangers with extremely large heat exchanging surface. Another advantage is that pressurized air has enough power to overcome hydraulic resistance of air filtering devices that produce clean contamination—free air flow and prevents contaminating impurity of the permeable heat exchanging media. FIG. 12 illustrates a principal difference in performance of commonly used air fan and sources of pressurized air according to one embodiment of the present invention. Computer air fans produce high air flow at several hundred liters per minute. However, air pressure developed by fans is very low, typically in the range from about 0.005 to about 0.01 atm. Air pressure at this level is not enough to overcome a hydraulic resistance of heat exchanging structures with well developed heat exchanging surface. Low pressure made by fans prevent the use of reliable air filtering devices in computers as the pressure drop in these devices are beyond of what can be produces by fans. Still another inevitable consequence of low air pressure made by fans is the large size of commonly used heat sinks. Main heat sink designing parameters such as fins size and distance between fins are selected as a compromise between a possible largest heat exchanging surface and admissible pressure drop. Since air pressure developed by fans is limited, the further improvement of heat sinks is limited too. Therefore the further increase of heat dissipating capacity on basis of fans is available in an extensive, but not in an intensive way. The last development in modern computer technology confirms this conclusion by the fact that number of fans in computers constantly grows.
In contrast to fans, compressors and vacuum pumps produce high air pressure sufficient to overcome hydraulic resistance in both, flow-through cooling devices and air filtering devices (See FIG. 13 which shows a graph of Air Supply Requirements for different cooling solutions). A moderate amount of air flow developed by air compressors and vacuum pumps is sufficient for removal heat from the heat generating electronic components with calculation of sharply increased heat removal efficiency in the flow-through cooling devices. These circumstances create a basis for new thermal management concept in the computer cooling area. A core idea of this concept is that commonly used “heat sink-fan” cooling structures can be replaced by the space saving and more energetically effective “flow-through cooling device—source of pressurized air” structures.
The flow-through cooling device in a combination with the source of filtered pressurized air represents a number of advantages before the commonly used finned heat sink-fan structures. This new electronics cooling solution provides a significantly greater heat dissipating capacity than solutions of the prior art. See FIG. 14 for a comparison. It is known that the heat dissipating limit for the heat sink-fan structures lies in the range of about 25 to about 35 w/ cm2 It also known that heat emitted by modern microprocessors approaches about 50 to about 75 w/cm2, and will reach about 150 to about 250 w/cm2 in near future. The heat removal capacity for flow-through cooling devices of the invention constitutes from at least about 100 to at least about 150 w/cm2 for air cooling media, and at least about 250 to about 300 w/cm2 for liquid cooling media.
Expected air requirements for different computer applications include an air flow between about 3-6 l/min at about 0.01-0.015 atm for a mobile personal computer; between about 5-10 l/min at about 0.015-0.03 atm for a desktop personal computer; and between about 10-20 l/min at about 0.025-0.05 atm for a performance computer;
An especially important advantage is low weight and miniature arrangement for cooling devices. Comparing to conventional solutions based on finned heat sinks, the flow-through cooling devices of the invention are lighter by 4-8 times and smaller by 3-5 times (FIG. 15). The size reduction in flow-through cooling devices is directed in vertical direction that makes the cooling devices thinner and better adapted to the smaller electronic devices, such as but not limited to a computer notebook layout. According to the present invention, the permeable heat exchanger 34 (FIGS. 10A, 10B and 11) is made of a material having high specific surface area and a relative high thermal conductivity. It is preferable that the materials have porosity in the range from at least about 30 to at least about 70%, more particularly from about 50 to about 60%. In one embodiment, the greater part of the porous portion of the exchanger should be interconnected to provide a through pass, for example for a cooling fluid. A number of commercially available materials can be used for making permeable heat exchangers according to the present invention. Suitable material include but are not limited to graphite foams, metal foams, ceramic foams, graphite fabrics, porous metals, carbon nanotubes and the like. In an alternative embodiment, a labyrinth type heat exchanger can made as an integral micro structure of a high thermal conductive micro composite material according to the present invention and by a micro manufacturing process according to the present invention. Such a labyrinth type heat exchanger comprises a highly branched, high specific surface and specific volume micro composite ribs which manage air flow and produce high heat transfer in a relative small volume.
FIGS. 16 and 17 illustrate still more embodiments of the invention. In these embodiments, a flow-by heat sink 70 comprises a micro composite material 79. The heat sink 70 comprises a heat spreader 72 and a system of primary fins 74. Primary fins 74 might be preformed as straight or wavy ribs, broken or continues ribs, or other shapes. Primary fins 74 manage air flow within heat sink 70 and provide the macro level heat exchange. A secondary micro-fin structure is formed on the surface of primary fins 74. The second micro-fin structure (not shown) turbulizes airflow, prevents formation of air stagnation zones and provides the micro level heat exchange. A flow-by micro composite heat sink 70 with a side fins location is shown on FIG. 17.
A flow-by brush-type heat sink 77 is shown on FIG. 18. Brush-type heat sink 77 comprises a heat spreader 72 of a micro composite material 79 and a plurality of fibers 78, preferably graphite fibers. One end of fibers 78 is embedded into micro composite material of heat spreader 72 (not shown in FIG. 18) and is located significantly close to the heat receiving surface of heat spreader 72. The non-embedded end of the fibers 78, in this example graphite fibers, exit and are exposed to air flow. The embedded portions of graphite fibers 78 collect heat and transfer it to the exposed part of graphite fibers 78, where heat exchange with surrounding air takes place. Brush-type heat sink 77 provides a beneficial combination of an extremely high heat removal capacity, low weight, and small overall size.
The present invention provides a CTE controlled heat spreaders and substrates that can be made as a separate item or as a joint item that comprises a heat spreader, a heat sink, a ceramic substrate, and a permeable heat exchanger. The CTE controlled heat spreader and substrate 50 is a macro composite article comprising a CTE controlling core 51, which has CTE close that of the electronic component, and a shell 57 that three-dimensionally envelopes the core 51 (FIG. 19). In a preferred embodiment the shell is a metal, a metal alloy or metal matrix composite (further metal shell). The shell 57 provides structural integrity for the macro composite heat spreader 50 and high thermal conductivity due to specifically selected components of the metal matrix composite material. Core 51 has a number of through holes 55 placed in a predetermined manner onto its footprint. The manner in which the holes are configured depends upon the device to be cooled, the materials selected and other factors as described herein. All mating surfaces 54 on the core 51 are rounded to reduce stress concentration in shell 57, particularly the metal matrix composite shell of the embodiment shown.
Due to different properties of the core 51 and shell 57, a thermal mechanical conflict takes place within the CTE controlling substrate 50 during manufacturing process and over time following service in an electronic device. During fabrication a residual tensile stress 59 and a residual compressive stress 58 are purposely created in the shell 57 by braking shrinkage according to the manufacturing process of the present invention. Magnitude and directivity of residual tensile stress 58 and residual compression stress 59 in shell 57 can be specified and managed by a proper disposition of through holes 55 in core 51 (FIG. 20). According to the present invention, arrangement of and distances between holes 55 are configured in a such way that residual stresses 58 and 59 developed in the local fields between holes 55 do not reach the limiting tensile stress 61 and compressive stress 62, inherent to a particular metal matrix composite material (or combination of materials) employed in shell 57. One skilled in the are will appreciate that this approach provides formation of residual stresses with controlled magnitude and prevents from cracking and integrity failure in the shell 57, for example a metal matrix composite shell, during the manufacturing process according to the present invention. The residual stresses have opposite directivity relative to thermal stresses generated by operation of the semiconductor device. The interaction between differently oriented stresses produces absorption and mutual cancellation of thermal stresses. The result of such interaction is that the source for substrate expansion disappears and no substrate expansion is produced, except an expansion, which is inherent for the TBS core material. Residual stresses 58 and 58 disappear with time due to the stress self-relaxation process, which can be accelerated by corresponding heat treatment.
Generally, the core 51 is subjected mostly to compressive stress 59, while metal matrix composite shell 57 is subjected to tensile stress 58. Generally, materials better withstand against compressive loads than against tensile load. Therefore the shell 57 is not able to compress the core 51 when temperature comes down. Further, the shell 57 cannot withstand against expansion of core 51 when temperature rises. Being a weaker structure, shell 57 follows to thermal mechanical behaving of the core 51. Since the CTE of the core material 51 is selected similar to the CTE of electronic component, the CTE controlled substrate 50 becomes thermal-mechanically matched with the electronic component 52. The thermal-mechanical matching method according to the present invention provides the electronic supporting and cooling components with useful properties as it shown in the following embodiment examples.
Turning to FIG. 21 an integrated substrate 70 for electronic component 71 is shown. Integrated substrate 70 comprises a CTE controlled heat spreader 73 in which a ceramic substrate 74 is integrated metallurgically. A heat spreader 73 is comprised of a CTE controlled core 75 and a metal matrix composite shell 77. Arrangement of core 75 and shell 77, as well as the character of interactions between them are similar to that described in the embodiment for CTE controlled heat spreader. Heat spreader 73 contains an alighting deepening 76 that matches to the configuration of ceramic substrate 74. CTEs for core 75 and ceramic substrate 74 are selected to be the same or substantially close. Thermal-mechanical interaction within integrated substrate 70 takes place between core 75 and shell 77. The relative thin ceramic substrate 74 is protected from creation of stresses by the larger and stronger core 75. By this, ceramic substrate obtains a gentle support from the CTE controlled heat spreader 73 that prevents deformation or cracking. Detention of ceramic substrate in the CTE controlled heat spreader is provided by a combination of mechanical and metallurgical action. The integrated substrate 70 provides an ideal thermal contact between ceramic substrate 74 and heat spreader 73 and eliminates a need for use of interface materials. FIG. 22 shows an arrangement of an integrated substrate 70 that is coupled with a flow-through cooling device 79, which is described in more details in the following embodiments of the present invention.
Turning now to FIGS. 23 and 24, the views of two embodiments of a general-purpose computer arrangement 100, such as a personal computer or server that may advantageously employ one or more aspects of the present invention are shown. FIGS. 23 and 24 show the vertical and horizontal components layout, accordingly. Generally, the computer arrangement 100 is comprised of a case 115, a plurality of electronic devices, such as a microprocessor 101, a power supply 102, a hard disk drive 103, a video card 104, and the like. These components (e.g., power supply, CPU, video card, and hard drive) are coupled together via an architecture, which allows the components to communicate with one another and potentially with external devices (not shown), such as other computer systems, scanners, etc. The architecture may take on any of a variety of forms without departing from the scope of the instant invention.
The electronic components 101, 102, 103, and 104 lie on the flow-through cooling device 107 which is connected with the air port 106 by mean of flexible hoses 105. Although only four cooled electronic devices shown on FIG. 1, any of a wide variety of electronic devices may be readily employed without departing from the scope of the instant invention. The air port 106 comprises of a micro compressor 110, an electro motor 109, an air filter 111, and an air distributor 114.
The air compressor 110 sucks cold air 116 through the air filter 111 entrapping contamination particles and preventing the pollution deposition within the flow-through cooling devices 107 and clogging the porous heat exchange media. The flow path of slightly pressurized cold air starts in the air filter 111 and the micro compressor 110 is powered by an electrical motor 109. Further purified air passes the air distributor 114, in which air flow is divided on a plurality of flows destined to cool electronic components 111, 112, 113, and 115 by flowing through cooling devices 107, on which the electronic devices are mounted. Within the permeable media of the cooling devices 107 the heat exchange between cold air and a heat exchanging structure takes place as it explained in more details below. Cold air removes heat from the heat exchanging structure and passes out of the cooling devices 107. Further hot air 117 exiting different cooling devices 107 is collected in the air collector 121, transferred to a definite location beyond the bounds of the computer case 115 and then exhausted to atmosphere. Alternatively, exiting hot air 117 can be exhausted into atmosphere in the destined points within the computer case 115 by connecting the air outlets of the flow-through cooling devices 107 and desirable exhausting points with the flexible hoses 105 (not shown).
FIGS. 25 and 26 show the schematic views of a compact computer arrangement 100 according to the present invention, in which a micro vacuum pump is used as an air mover. Vacuum pump creates an overpressure on the sucking side that is lower than 1 atmosphere. This pressure is sufficient for use in the flow-through cooling devices according to the present invention. An additional advantage of vacuum pumps before the micro compressors is that this air mover does not increase temperature of air as it takes place in the compressors. Therefore heat removal in the flow-through cooling devices is better when a vacuum pump is used. It is noted that, while FIG. 25 depicts the use of hoses 105 to move air from the air intake filter 116 to the components 107, the use of such hoses can be, preferably, be omitted. Further, while the hoses to move the air from the components 107 to the vacuum pump 110 can also be omitted, efficiency can be expected with their use.
In both such embodiments, while the figures may suggest that a single cooling device is matched with an electronic component. While this configuration may be convenient, other configurations are also possible. For example, a single cooling device may service two or more electronic components or all of the electronic components. Indeed, even the units of the cooling devices can be combined. For example, a single micro heat sink can service two or more heat spreaders.
In one embodiment, the air mover is a vane machine, preferably a toroidal vane machine. Suitable vane machines include but are not limited to those embraced by U.S. Pat. No. 5,233,954, U.S. Published application 2003/0111040 and U.S. applications entitled “Improvements in Intersecting Vane Machines” U.S. Ser. No. 10/744,230 filed Dec. 22, 2003 as Atty. Docket number 4004-3001; “Improvements in Sealing Intersecting Vane Machines” U.S. Ser. No. 10/744,229 filed Dec. 22, 2003 as Atty. Docket No. 4004-3004; and “The Use of Intersecting Vane Machines in Combination with Wind Turbines” U.S. Ser. No. 10/744,232 filed Dec. 22, 2003. The entire teachings of the aforementioned patents and patent applications are incorporated herein by reference. The toroidal vane machine is capable of serving as a micro compressor or vacuum pump which can be scaled to meet a variety of dimensional requirements. In one embodiment the toroidal intersecting machine is a micro air mover capable of being used in small computer applications, such as but not limited to, in computer “notebooks”, desktop computers or servers. Other suitable air movers are known to those skilled in the art.
Turning to the following drawings, a vertical (FIG. 25), and a horizontal (FIG. 26) layout of electronic components within the computer arrangement 100 are shown. A micro vacuum pump powered by an electro motor is connected with a plurality of cooling devices 107, an air distributor 114, and air filter 111 by flexible hoses 105. When operating, the vacuum pump sucks cold air 116 through the air filter 111 and the cooling devices 107. As cold air 116 advances within the heat exchanging structure of the flow-through cooling devices 107, an intensive heat exchange takes place between the heat exchanging structure and air. Air removes heat from the cooling device 107 and transfers it to the vacuum pump and further hot air 117 exhausts outside of the computer case 115.
A thermal-mechanical matching method according to the present invention provides the electronic supporting and cooling components with useful properties as it shown in the following embodiment examples. Turning to FIG. 29, a CTE controlled substrate 70 for electronic component 71 is shown. Material for the perforated core 74 is selected with consideration to be close to CTE of the semiconductor material. Different types of solid and porous graphite and ceramic can comply with these requirements. Material for the metal shell 77 is selected with consideration of high thermal conductivity and manufacturability. A plurality of copper and aluminum alloys can fulfill these requirements. Manufacturing process for the CTE controlled substrate 70 can be provided by different metal forming technologies such as the insert die casting, the insert investment casting, powder sintering, infiltration, etc. In general the manufacturing process includes preparation of the perforated core of graphite or ceramic, and following encapsulation of the core by the metal shell using one of the above listed methods.
Turning to FIG. 29 a simple CTE controlled solution is shown. An CTE controlled substrate 70 is comprised of a perforated core 75 and a metal shell 77, which envelops the perforated core 75 and fills the perforation openings. Materials for the perforated core 75 and metal shell 77 are selected according to the above explained considerations. One of described above manufacturing methods can be used for fabrications of the CTE controlled substrate 70.
Turning to FIG. 30 a CTE controlled, internally cooled substrate 80 is shown. Arrangement of core 75 and shell 77, as well as the character of interactions between them are similar to that described in the embodiment for CTE controlled heat spreader. The perforated core 75 is made of a porous permeable material 81 such as graphite foam, porous graphite, porous ceramic, metal foam, etc. Beside of the CTE controlling function, the perforated core 75 provides the heat removal function by flowing the cooling media through the permeable media 81. Configuration of the perforated core 75 is designed in a way that provides a desirable path of the flowing through cooling media within the CTE controlled substrate 80. The CTE controlled internally cooled substrate comprises an inlet and outlet for the cooling media similarly to the above explained cooling device on FIG. 10 (not shown on FIG. 30).
Turning to FIG. 31, a CTE controlled, internally cooled substrate 90 is shown. Arrangement of core 75 and shell 77, as well as the character of interactions between them are similar to that described in the embodiment for CTE controlled heat spreader shown on FIG. 29. Internal cooling in the substrate 90 is provided by a separate permeable insert 91. Inlet 92 and outlet 94 provide a managed flow 93 of the cooling media within the substrate 90. Such a dividing of the heat removal and CTE controlling functions in the substrate 90 allowing use of the permeable materials having CTE different than the semiconductor material. A CTE mismatch between the semiconductor material 71 and the permeable core 91 is smoothed by the perforated core 75, which has CTE equal or close to what of the semiconductor material 71. One of described above manufacturing methods can be used for fabrications of the CTE controlled internally cooled substrate 90.
While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.