WO2002017698A2

WO2002017698A2 - Distributed thermal management system for electronic components

Info

Publication number: WO2002017698A2
Application number: PCT/CA2001/001167
Authority: WO
Inventors: Howard R. Harrison; Jeffrey R. Brown
Original assignee: Hb Innovation Ltd.
Priority date: 2000-08-22
Filing date: 2001-08-22
Publication date: 2002-02-28
Also published as: WO2002017698A3; WO2002017698A8; AU2001281624A1

Abstract

A distributed thermal management system for electronic components has a heat pipe, a first heat pump in thermal communication with the evaporator end of said heat pipe and a second heat pump in thermal communication with the condenser end of said heat pipe, said first heat pump adapted to withdraw heat from an electronic component and said second heat pump adapted to deliver said heat to an external heat sink. A controller is adapted to control the system such that said electronic component, said heat pipe, and said heat sink operate at below, approximately at, and above ambient temperature respectively. Also taught is an interface plate adapted to cool multiple electronic components using a single external heat sink. Also taught is a thermal diode located between the heat sink and the electronic component(s), and adapted to block the reverse flow of heat.

Description

Distributed Thermal Management System for Electronic Components

BACKGROUND OF THE INVENTION

Field of the Invention

This invention relates to the selective cooling of electronic components within a cabinet using a unique combination of thermoelectric, heat pipe, and thermal diode technologies. Components may be cooled to sub-ambient temperatures, and the removed heat is transferred through the cabinet at ambient temperatures to be dispersed into the air surrounding the cabinet.

Acknowledgement of Prior Art

Recent trends in electronic components, particularly CPUs, include higher power in smaller packages. As a result, newer components produce a much higher average heat per unit area. Further, there is a much greater variation in heat flux since this heat is not spread evenly across the surface of the component. Adding to the problem of thermal management is the fact that systems are becoming more densely packed in order to accommodate the increased number of components that currently comprise a PC, workstation, or server.

There has been a tremendous amount of inventive activity in this field as a result of the above noted trends. This is clearly indicated by the numerous component cooling designs that have been disclosed over the last few years including US patent 6,249,428 issued 6/01 to Jeffries, et al (assigned to Dell USA), US patent 6,229,704 issued 5/01 to Hoss, et al (assigned to Dell USA), US patent 6,226,184 issued 5/01 to Stolz, et al (assigned to Sun Microsystems), US patent 6,226,178 issued 5/01 to Broder, et al, (assigned to Dell USA), US patent 6,034,870 issued 3/00 to Osborn, et al (assigned to Sun Microsystems), US patent 5,960,863 issued 10/99 to Hua, US patent 5,921,087 issued 7/99 to Bhatia, et al (assigned to Intel), US patent 5,896,917 issued 4/99 to Lemont (assigned to Lemont Aircraft), US patent 5,880,524 issued 3/99 to Xie (assigned to Intel), US patent 5,671 ,120 issued 9/97 to Kikinisi (assigned to Lextron Systems), US patent 5,637,921 issued 6/97 to Burward-Hoy (assigned to Sun Microsystems), and US patent 5,615,086 issued 3/97 to Collins, et al (assigned to Tandem Computers). In the majority of these recent examples, heat is removed from the CPU only to be dissipated within the cabinet. This compounds the problem since the heat must then be moved through the cabinet to the outside air, potentially heating other components along the way. It also requires a reasonable amount of plenum space around the components to accommodate the air flow, thus precluding the addition of more components within a system and / or the further reduction of system size. This fundamental approach is taught by all of the above patents except US patent 6,226,184 to Stolz, et al, US patent 6,226,178 to Broder, US patent 5,615, 086 to Collins, et al, and possibly 5,896,917 to Lemont.

In the exceptions noted above, methods are taught whereby heat may be directed to the outside of the cabinet. However the method taught by 6,226,178 may not be effective since the thermal interface between the CPU and the heat sink, i.e. the most critical thermal interface in the system, is not permanent and may be detached by removing a panel on the cabinet. Further, the method taught by 5,615,086 uses a recirculating coolant which may be prone to leaks and which may cause servicing difficulties.

It should also be noted that in the majority of these recent examples, electronic components may not be cooled to below ambient temperatures where they operate most effectively and reliably. This applies to all of the above patents except US 5,637,921 to Burward-Hoy and US 5,615,086 to Collins, et al. However, as previously noted, 5,637,921 teaches that the heat be dispersed inside the cabinet, and the method taught by 5,615,086 may be prone to leaks and servicing problems. Further, neither of these patents deals with the overall and related problem of condensation within the cabinet due to the lower than ambient temperature of the heat transfer components.

One of the patents that does provide below ambient cooling, US 5,637,921 to Burward-Hoy, requires a large amount of power since it employs up to three thermoelectric modules to achieve the desired temperatures. This is inefficient and places a tremendous thermal load on the other components within the cabinet since the combined heat from the CPU plus the excess heat produced by the thermoelectric modules is dispersed within the cabinet rather than being moved directly to the outside air.

Further, it should be noted that all of the active solutions (i.e. those employing a fan, thermoelectric module, or some other non-passive device), with the possible exception of US 6,226,178 to Broder and US 5,637,921 to Burward-Hoy, have a single point of failure. US

6,226,178 teaches a forced convection and a natural convection heat sink connected to the same heat pipe and it is likely that the natural convection heat sink would allow the system to operate, under certain circumstances, after a failure on the forced convection side. US 5,637,921 teaches three thermoelectric modules connected in series, and it is possible that the system could survive the failure of one of these modules since the defective module may still allow the through flow of heat.

Finally, it should be noted that the existing solutions may be difficult to mass produce since each solution must be custom designed for a particular CPU and cabinet configuration. This is primarily because a method for conveying heat from the components to the outside of the cabinet in an efficient, safe, and modular manner has not yet been identified.

Summary of the Invention

The current invention discloses a modular system which allows for the mass production of universal cooling systems comprised of thermoelectric, heat pipe, thermal diode, and forced convection heat sink components. A thermoelectric module is located at each end of the heat pipe to allow for precise control of heat flow rates and temperature as heat is pumped from the thermal components) to the heat sink located on the outside of the cabinet. The modularity allows one or more electronic components to be cooled using a single external heat sink. Composite heat / power/ monitoring signal conduits are used to make the required connections.

Further, an embodiment of the invention discloses a method for point cooling which allows selected components to operate at below ambient or even below freezing temperatures by using a thermoelectric module in thermal communication with the electronic component. A thermal dispersion plate may be mounted between the electronic component and the thermoelectric module in order to smooth out the heat flux variations across the surface of the electronic module. Condensation problems may be avoided by insulating and hermetically sealing the small chilled area around the electronic component, dispersion plate, and thermoelectric module to prevent the ingress of heat and humidity.

Further, an embodiment of the invention teaches that a first thermoelectric module pumps heat from the electronic component and delivers it into the evaporator end of a heat pipe. The heat is then efficiently delivered through the heat pipe to the condenser end where a second thermoelectric module then pumps the heat into an external heat sink. Temperatures are controlled such that the heat pipe operates at substantially ambient levels along its entire length, thus preventing any condensation or "hot conduit" problems within the cabinet.

Further, an embodiment of the invention discloses a method whereby heat from the one or more components being cooled is dissipated outside the cabinet through an external heat sink. A second thermoelectric module in thermal communication with the condenser end of each heat pipe and the external heat sink (i) contributes to the amount of heat that may be pumped from the electronic modules through the heat pipe and (ii) elevates the temperature of the heat sink above the ambient level of the heat pipe so that the heat may be dispersed into the ambient air through forced or natural convection. Again, It should be noted that the heat is transferred through the cabinet at ambient temperatures, and that there are no related condensation or "hot conduit" problems. This approach does not present any additional thermal loads within the cabinet, and in fact allows for a denser packing of components since much less plenum space is required.

Further, an embodiment of the invention discloses an extremely reliable solution since the only moving component is the external fan. The thermoelectric modules are inherently reliable, and the heat pipes are passive. In addition, the system has been configured such that each electronic component is served by two thermoelectric modules such that at least one would be available in the unlikely event of a failure.

Further, an embodiment of the invention teaches that a thermal diode be placed between the heat sink and the condenser end of the heat pipe to prevent the reverse flow of heat from the heat sink back into the electronic components, possibly causing damage to the electronic components, in the event of a power failure or power down situation. The current invention discloses various means to build a thermal diode that may be used to fulfil this function.

Finally, an embodiment of the invention teaches that the control system for the distributed thermal management system for electronic components may be used to externally display the temperature of the chilled components, and may be used to deliver the relevant signals to the system software such that the user may be kept informed of the status of the chilled components and such that the disclosed system may be interfaced with other software based thermal management systems.

Accordingly, this invention relates to a distributed thermal management system for electronic components that may be used to cool the electronic components to sub-ambient temperatures and deliver the heat to an external heat sink where it may be dispersed into the ambient air. The invention comprises (a) a thermal dispersion plate in thermal communication with an electronic component; (b) a first thermoelectric module in thermal communication with the thermal dispersion plate; (c) a means to insulate and hermetically seal the chilled area around the electronic component, thermal dispersion plate, and the first thermoelectric module; (d) a heat pipe with an evaporator end in direct thermal communication with the first thermoelectric module; (e) a second thermoelectric module located outside the system cabinet and in thermal communication with the condenser end of the same heat pipe, and configured to be in thermal communication with multiple such heat pipes; (f) a heat sink in thermal communication with the second thermoelectric module; (g) a thermal diode connected between the heat sink and the condenser end of all connected heat pipes such that heat may only flow from the heat pipes and into the heat sink: (h) a control system to monitor and control the function of all components within the thermal management system, and; (i) composite heat / power / monitoring signal conduits to integrate the necessary heat pipe, electrical power, and monitoring signal connections required between the heat sink and each electronic component.

This invention teaches a distributed thermal management system consisting of a) a heat pipe having an evaporator section and a condenser section; b) a power source; c) a first heat pump in thermal communication with a thermal load and in thermal communication with the evaporator section of the heat pipe; d) a second heat pump in thermal communication with the condenser section of the heat pipe and in thermal communication with a heat sink; wherein the heat pipe and the heat pumps are configured to withdraw heat from the thermal load and deliver the heat to the heat sink where it may be dispersed into the ambient air.

The distributed thermal management system may further include a dispersion plate located between the first heat pump and the thermal load.

The distributed thermal management system may further include a heat flux dispersion plate, a quantity of phase change material, and a further dispersion plate located between the first heat pump and the thermal load.

The distributed thermal management system may further include a composite conduit containing the heat pipe, the power supply wires for the first heat pump, and all signal monitoring wires connected to the first heat pump and the thermal load. This invention also teaches that the first heat pump and the heat pipe may be located inside an enclosed cabinet, and that the second heat pump and the heat sink may be located on the outside of the enclosed cabinet.

The distributed thermal management system may further include a control system; wherein the control system is adapted to monitor and control all aspects of the distributed thermal management system, provide visual information regarding the performance of the distributed thermal management system to a user, and interface with operating software that may be associated with the thermal load to provide further indications to the user and to interoperate with other thermal management systems that may be associated with the thermal load.

This invention also teaches that the thermal load may be controlled to operate at below ambient temperature, the heat pipe may be controlled to operate at approximately ambient temperature, and the heat sink may be controlled to operate at above ambient temperature.

This invention also teaches that the thermal load, all thermal components in thermal communication with the thermal load and operating at below ambient temperature, may be insulated and hermetically sealed.

This invention also teaches that the first heat pump may be one or more thermoelectric modules.

This invention also teaches that the second heat pump may be one or more thermoelectric modules.

The distributed thermal management system may further include a fan to disperse heat from the heat sink through forced convection.

The distributed thermal management system may further include an interface plate located between the second heat pump and the condenser end of the heat pipe to allow the detachable thermal connection of multiple heat pipes to a single second heat pump and heat sink assembly.

This invention also teaches that the thermal load may be an electronic component. This invention also teaches an enhanced distributed thermal management system consisting of a) a heat pipe having an evaporator section and a condenser section; b) a power source; c) a first heat pump in thermal communication with a thermal load and in thermal communication with the evaporator section of the heat pipe; d) a second heat pump in thermal communication with the condenser section of the heat pipe and in thermal communication with a heat sink; e) a thermal diode located between the heat sink and the condenser section of the heat pipe; wherein the heat pipe and the heat pumps are configured to withdraw heat from the thermal load and deliver the heat to the heat sink where it may be dispersed into the ambient air, and wherein the thermal diode is configured to prevent the reverse flow of heat.

The enhanced distributed thermal management system may further include a dispersion plate located between the first heat pump and the thermal load.

The enhanced distributed thermal management system may further include a heat flux dispersion plate, a quantity of phase change material, and a further dispersion plate located between the first heat pump and the thermal load.

The enhanced distributed thermal management system may further include a composite conduit containing the heat pipe, the power supply wires for the first heat pump, and all signal monitoring wires connected to the first heat pump and the thermal load.

The enhanced distributed thermal management system may further include a control system; wherein the control system is adapted to monitor and control all aspects of the distributed thermal management system, provide visual information regarding the performance of the distributed thermal management system to a user, and interface with operating software that may be associated with the thermal load to provide further indications to the user and to interoperate with other thermal management systems that may be associated with the thermal load.

The enhanced distributed thermal management system may further include a fan to disperse heat from the heat sink through forced convection.

The enhanced distributed thermal management system may further include an interface plate located between the second heat pump and the condenser end of the heat pipe to allow the detachable thermal connection of multiple heat pipes to a single second heat pump and heat sink assembly.

This invention also teaches that the thermal diode may be virtually implemented by using one or more thermoelectric module(s) as the second heat pump, and by applying a small forward voltage to the thermoelectric module(s) after main power has been removed.

This invention also teaches that the thermal diode may be a modular spacer block component consisting of a dispersion plate on each of the outside thermal communication surfaces, an insulating layer between the dispersion plates, and an integrated heat pipe in thermal communication with the dispersion plates; wherein the integrated heat pipe may be attached to diagonally opposed comers of the dispersion plates and configured to be of sufficient length.

This invention also teaches that the thermal diode may be a modular spacer block component consisting of a dispersion plate on each of the outside thermal communication surfaces, an insulating layer between the dispersion plates, and an integrated heat pipe in thermal communication with the dispersion plates; wherein the integrated heat pipe may have two evaporator sections connected through a wickless u-tube that only allows the passage of vapor.

This invention also teaches that the thermal diode may be a modular spacer block component comprised of a dispersion plate on each of the outside thermal communication surfaces, an insulating layer between the dispersion plates, and an integrated heat pipe in thermal communication with the dispersion plates; wherein the integrated heat pipe may have an externally heated reservoir section attached to a condenser section; and wherein the heated reservoir section may contain an inert gas which may expand, when heated, to displace operating vapor contained within the condenser section.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention are described by way of example with reference to the drawings in which: Figure 1 is a tabular representation of the internal and external subassemblies within a distributed thermal management system,

Figure 2 is a diagrammatic representation of the push / pull heat principle,

Figure 3 is a diagrammatic representation of a passive thermal persist heat pipe,

Figure 4 is a diagrammatic representation of a semi-active thermal persist heat pipe with the control element "off,

Figure 5 is a diagrammatic representation of a semi-active thermal persist heat pipe with the control element "on",

Figure 6 illustrates the basic construction of a thermal diode spacer block,

Figure 7 provides a side view of a thermal diode spacer block with "Z" geometry,

Figure 8 provides a top view of a thermal diode space block with "Z" geometry,

Figure 9 provides a simplified view of the major components in the distributed thermal management system for electronic components,

Figure 10 provides a detailed view of the components in the distributed thermal management system for electronic components,

Figure 11 provides further detail regarding a dispersion plate, and,

Figure 12 provides further detail regarding a thermal diode spacer block.

DETAILED DESCRIPTION OF AN EMBODIMENT

The distributed thermal management system for electronic components has six major components as follows;

Heat Flow Component Heat Out Fan Module

Module

Heat In Electronic Component

* Heat pipe in this case refers to a phase change heat pipe, and this will continue to be the case throughout this document unless otherwise stated. A distributed thermal management system for electronic components may also be built with heat conduits (which use straight conductance to convey heat from source to destination) providing that the resulting increase in temperature differential is acceptable for the application. The latter approach may become more practical as the length of the heat conduit becomes shorter. The general method of operation can be most easily understood if one considers the flow of heat through the system. The internal thermoelectric (TE) module serves as a heat pump to remove heat from the electronic component (e.g. a CPU) and "push" it into the bottom of the heat pipe. The heat pipe then serves as a conduit to transport the heat to the outside of the cabinet. The external thermoelectric module then "pulls" heat out of the top of the heat pipe and dissipates it through the heat sink into the ambient air. A fan may be used to aid this process through forced convection.

In addition to heat flow, one must also understand the implications regarding temperature. The basic underlying principles are (1) the thermoelectric modules maintain a temperature differential between their hot and cold sides and (2) the heat pipe remains at a relatively constant temperature along its length while conducting heat. These principles may be applied to the distributed thermal managements system for electronic components with the following results.

Temperature Component

Fan

Above Ambient Heat Sink

External TE Module

Ambient Heat Pipe

Internal TE Module

Below Ambient Electronic Component

The fan and heat sink must necessarily operate at above ambient temperatures to facilitate the dissipation of heat away from the cabinet. The external thermoelectric module is designed to maintain a similar temperature differential relative to ambient, but in the reverse direction, such that the heat pipe remains at approximately ambient temperature along its length. This is extremely important since it allows the internal thermoelectric module to then provide a further temperature differential such that the electronic component operates at below ambient temperatures.

This design allows the electronic component to operate at very low temperatures while allowing heat to flow through the cabinet safely at ambient conditions. It is possible that the temperature of the electronic component may reach below freezing levels, especially if the internal thermoelectric module is a cascaded device, i.e. comprised of multiple thermoelectric modules connected in series. From a manufacturing standpoint, the various components can be grouped into two sub- assemblies, internal and external, as depicted in Figure 1. Note that the interface plate, while not a major component, has been included for clarity. The interface plate is designed to allow the connection of multiple internal subassemblies to a single external subassembly.

With reference to Figure 1 , the external subassembly is comprised of the fan, heat sink, external thermoelectric module and the interface plate. The external subassembly is designed to be attached to the outside of the cabinet, and its primary function is to quickly dissipate the heat that has been generated inside the cabinet. The heat pipe(s) carrying this heat may be thermally and mechanically attached to the interface plate located at the bottom of the external subassembly (i.e. on the cold side of the external thermoelectric module). Note that more than one heat pipe may be connected to a single interface plate.

Again with reference to Figure 1, the major elements of each internal subassembly include a heat pipe, an internal thermoelectric module, and an electronic component. The internal thermoelectric module is in thermal communication with the input (evaporator) end of the heat pipe and the electronic component. These junctions may be insulated and hermetically sealed to prevent the ingress of heat and moisture, thereby preventing condensation. The heat pipe, together with the power wires for the thermoelectric module and any associated monitoring and control wires, may be housed in a composite conduit so that only one physical "pipe" is required to facilitate all thermal and electrical communications between the internal and external subassemblies.

In some cases it may be desirable to place a dispersion plate between the electronic component and the internal thermoelectric module. The primary functions of this dispersion plate are to (i) disperse or "smooth out" any heat flux variations that might occur in the electronic component and (ii) present the heat to the thermoelectric module across a surface area that matches that of the thermoelectric module and (iii) add to the stability of the temperature of the electronic component as it produces heat at various rates. The dispersion plate may be constructed of any material with suitable high thermal conductance. Further, the thermal mass of the dispersion plate may be enhanced by adding a thermal gel, or, in some cases, a quantity of phase change material (PCM). The advantage of a PCM is that once "frozen" by the thermoelectric module, it is capable of absorbing a tremendous amount of heat from the electronic component, despite variations in heat flux, and it will do so at a relatively constant temperature. Once the electronic component has been placed on the printed circuit (PC) board, and the other elements of the internal subassembly are in place, the output (condenser) end of the heat pipe is then adjusted to mate with the interface plate on the external subassembly. Alternatively, the heat pipe could be pre-formed with the correct geometry so that the components will mate when assembled.

One of the key aspects of this design is that more than one internal subassembly may be attached to a single external subassembly. In other words, the distributed thermal management system for electronic components may be used to selectively cool a number of components within the cabinet while using only one external heat sink. This may be accomplished by configuring an interface plate on the cold side of the external thermoelectric module, and sizing it such that it will accept the thermal connection of more than one heat pipe.

From a practical standpoint, the external subassembly may be sized according to the amount of heat that they can absorb at the interface plate while maintaining a safe operating temperature. This may be represented, for example, as a 60 watts input rating. The internal subassemblies, on the other hand, may have output ratings that represent the total amount of heat that they will extract from an electronic component at the required operating temperature plus the excess heat that is generated by the internal thermoelectric module. Multiple internal subassemblies may be connected to the interface plate as long as the total of their output ratings does not exceed the input rating for the external subassembly.

Some examples will clarify this modular architecture.

Modular Design -Advantages

This modular design combined with thermoelectric technology allows one to split the required heat pumping capacity and place it at different locations within the system, in this case at either end of the heat pipe, to create a distributed thermal management system. In existing solutions, all of the heat pumping capacity has been placed in one location with the following associated problems;

A - Place entire heat pump outside the cabinet

This means that the cold side of the heat pump must be substantially below the required electronic component temperature to allow heat to flow away from the electronic component. As a result, the heat conduit operates at very cold temperatures and must be insulated to prevent condensation. More seriously, if a true heat pipe is used to produce very low component temperatures, the heat pipe is prone to bursting if it is allowed to warm up too quickly due to the rapid expansion of internal gasses

B - Place entire heat pump inside the cabinet

In most cases this design places the heat pump right on the electronic component. As a result the heat conduit can become dangerously hot as it operates substantially above ambient temperature. It must be insulated to prevent the flow of heat out to other components as it passes through the cabinet. An excessively hot heat conduit can actually contribute to thermal management problems rather than resolving them.

In addition to heat problems, there is also a power problem associated with this design. Since the entire heat pump is located at the electronic component, all of the power must be carried through the cabinet to the same location. Due to the fact that most designs use thermoelectric components, this power is delivered predominately at 12V DC. The resulting current levels could easily exceed 10A.

^' Note that it is possible to implement a distributed thermal management system for electronic components using different types of heat pumps / refrigeration units at either or both ends of the heat pipe. There may be situations where, for example, one might utilize a number of internal thermoelectric modules and connect them through discrete heat pipes to a conventional refrigeration unit. This may be more feasible because of available components, or a very large aggregate cooling requirement at the external end of the heat pipe.

Regardless of the components chosen, the concept is the same - modularizing the thermal system allows one to provide selective cooling and provide a safe and efficient conduit for the heat to flow to an area where it can be safely and easily dissipated. The heat pump or refrigeration device at the input end 'pushes" the heat through the heat pipe. The heat pump or refrigeration device at the output end "puffs" the heat out of the heat pipe and drives it into the ultimate heat sink - be that the ambient air or any other suitable destination.

Heat Pipes - Push / Pull Efficiency

There are many existing applications that take advantage of the efficiency of a heat pipe by either pushing heat into one end or pulling it out of the other end. The "pushing" applications can be simulated by placing one end of the heat pipe in boiling water. Heat is rapidly "pushed" to the other end, which can become uncomfortable to hold within a few seconds. Alternatively, the "pulling" applications can be simulated by placing one end of the heat pipe in an ice bath. In this case, heat will be "pulled" from the other end as evidenced by the fact that it rapidly becomes very cold.

The combined "push / pull" aspect of the distributed thermal management system for electronic components actually contributes to the performance and responsiveness of the heat pipe. Unlike a standard heat conduit, which relies solely on a thermal gradient to start the heat flow, a heat pipe relies on a phase change to initiate operation. The "push" at one end of the heat pipe causes the contained fluid to vaporize as it absorbs the heat. This expanding vapour then travels the length of the heat pipe where it is cooled by "pulling" heat away from it, resulting in condensation. Note that in the case of a distributed thermal management system for electronic components, the "pull" is accomplished by actively extracting heat from the output end of the heat pipe using the external thermoelectric module, thereby creating a temperature differential to condense this vapour very quickly and a pressure differential to "draw" the vapour more rapidly to this end of the pipe. This distributed heat pump architecture accelerates the phase change cycle more rapidly than could be achieved by simply applying heat to one end of the heat pipe. The current invention takes full advantage of the phase change characteristic of a heat pipe by adding heat at the input end while simultaneously removing it from the output end. The heat "pushed" into the pipe plus the heat "pulled" from the pipe can only act on the fixed amount of combined vapor and condensate that exists within the heat pipe. Bearing this in mind, the "push / pull" cycle operates as follows;

Push Heat added to the input (evaporator) end serves to vaporize the internal fluid, and this same amount of vapor must be then condensed at the output end as the heat is released.

Pull Additional heat removed at the output (condenser) end serves to condense even more of the vapor that exists within the heat pipe, and this same amount of incremental condensate must then be vaporized at the input end, thus absorbing a commensurate amount of incremental heat.

The important principle here is that the total heat traveling through the heat pipe equals the heat "pushed" into the input end Plus the heat "pulled" out of the output end. This has been illustrated diagrammatically in Figure 2.

This principle allows the distributed thermal management system for electronic components to achieve tremendous efficiency since the heat pumped by both thermoelectric modules contributes to the heat removed from the CPU or other component to be cooled. This is in part due to the fact that (i) thermoelectric modules act passively as a thermal conduit, allowing the through flow of heat as driven by the thermoelectric module at the other end of the heat pipe, and (ii) that the application of power will add incrementally to this through flow of heat and produce the desired temperature differential across the thermoelectric module. The heat pipe will continue to operate effectively until the total heat flow capacity of the pipe has been reached.

This process may be controlled by first activating the thermoelectric module at one end of the heat pipe, and then by activating the thermoelectric module at the other end of the heat pipe to provide incremental heat pumping capacity until the net heat flow capacity of the heat pipe has been reached. For example the internal thermoelectric module could be turned on first, followed by adding power in a controlled fashion to the external thermoelectric module to remove incremental heat from the electronic component as required to keep it's temperature within a set-point range, and to keep the heat flowing through the heat pipe at ambient temperatures. Care must be taken to avoid overpowering the external thermoelectric module as this may freeze some of the condensate in the heat pipe, substantially reducing its rated capacity.

Heat Pipe Selection - Rate and Direction of Flow

Now that the overall operation of the distributed thermal management system for electronic components has been discussed, the focus will turn to the selection of the heat pipe. This is a critical component since it controls the flow of heat between the internal and external subassemblies. Note that this flow has two major characteristics - rate and direction.

During normal operation, heat flows from the internal thermoelectric module to the external thermoelectric module. It is driven by the former, and absorbed and accelerated by the latter. The heat is then pumped into the heat sink where it can be dissipated into the surrounding air. As a result of this process, a large amount of heat becomes stored or trapped in the heat sink as its temperature rises above ambient. This is necessary for normal operation since it ensures that heat will continue to flow in the correct direction, i.e. from the heat sink into the ambient air. However this becomes a technical liability once power is removed from the thermoelectric modules since at that point they will become bi-directional thermal conduits and, the heat will then tend to flow in the opposite direction, i.e. from the heat sink back into the electronic components.

The underlying problem is that a thermoelectric module becomes a thermal short rather than a heat pump while in a passive state - i.e. as soon as power has been removed. This unfortunately provides an easy alternative path for the heat stored in the heat sink. The heat will flow "backwards" through the heat pipe, across the Internal thermoelectric module, and back into the electronic component where it could cause some severe damage. For a short period of time this "backwards" flow of heat actually becomes the favoured path since there is such a large temperature differential between the heat sink and the electronic component. The problem is compounded by the fact that the thermal mass of the heat sink is substantially greater than that of the electronic component.

The ideal solution to this problem would be to use a heat pipe that only allows heat to flow one way, from the electronic component to the heat sink, and not in the reverse direction. Controlling the Reverse Flow of Heat

This section identifies three methods of controlling the reverse flow of heat. They can be used discretely or in combination.

These are referred to as thermal persist technologies since they maintain the desired thermal relationships even after power has been removed from the system. In a distributed cooling system for electronic components, thermal persist technologies will provide further protection for the electronic components.

As illustrated in Figure 3, passive thermal persist utilizes "water trap" heat pipes to automatically block the flow of heat when the thermal gradient is reversed. A water trap is connected to the evaporator end of the heat pipe through a wickless "U" tube. This allows the flow of vapor, but the absence of a wick prevents the return flow of any liquid. As more water becomes trapped in this manner, the vapor pressure in the heat pipe drops to the extent that it ceases to operate in the normal manner.

During normal operation, heat applied to the evaporator section (includes main evaporator plus second evaporator / water trap) causes the liquid inside to boil, producing vapor. This increases the pressure and causes the vapor to flow to the condenser tube at the other end of the heat pipe. Heat is then removed from the vapor, producing condensate that returns to the evaporator section through the wick by capillary action. Under normal operation an excess of liquid may accumulate at the condenser tube, however it can always return to the evaporator section since the wick is continuous between the condenser tube and the main evaporator tube.

When the thermal gradient is reversed, evaporation begins to take place in the condenser tube and condensation begins to take place in both evaporator tubes. Operation is identical to the "normal" or forward mode, including the fact that excess liquid tends to accumulate where the condensation takes place (now in the evaporator tubes). However in this case the water can not return to the other end since there is no continuous wick between the second evaporator / water trap and the other end of the heat pipe. Ultimately the vapor pressure will drop to the extent that the heat pipe ceases to operate. Note that this is a passive operation that occurs automatically upon the reversal of the thermal gradient and requires no further inputs.

What happens when the thermal gradient returns to normal? Evaporation once again takes place in the evaporator section, including the water trap. Since vapor can easily flow out of the water trap (the absence of a wick only prevents the flow of liquid), the trap soon dries out and the heat pipe returns to normal operation.

The careful observer will note that although the heat pipe ceases to operate normally when the thermal gradient is reversed, there is still a conductive flow of heat through the metallic heat pipe envelope. This can be minimized by inserting thermal insulator rings somewhere between the evaporator and condenser ends of the heat pipe. The critical part of this approach is to ensure that the wick remains continuous throughout the insulator section. One way to provide some degree of insulation would be to scribe a ring or spiral into the sides of the heat pipe, sufficiently deep as to reduce the cross sectional area but not to the extent of creating a hole in the pipe (as this would not only affect the continuity of the wick, but also the vapor pressure within the heat pipe) The material removed could then be replaced with another material that provides the required insulation and preserves the structural integrity of the heat pipe. Care must be taken to avoid any groove that may have been scribed on the inside of the heat pipe to aid in the wicking process.

Unlike passive thermal persist, which is automatic and does not require any input to stop the reverse flow of heat, semi-active thermal persist does require an input to initiate this process. This has been illustrated in Figure 4.

In this case, an inert vapor such as Argon or Helium is kept in a reservoir connected to the condenser tube. (The inert vapor is selected so as to not mix with the working vapor in the heat pipe.) When the reservoir is heated, the inert vapor expands until the pressures within the heat pipe equalize. When sufficient heat is applied, the entire working vapor is pushed out of the condenser section, effectively stopping normal heat pipe operation as illustrated in Figure 5. The input required to stop heat pipe action can be as simple as heating the reservoir. This may be accomplished through resistive heating, consuming only a few watts of power. Insulating the reservoir may reduce the actual energy requirement even further. In this case the heat flow applied to the reservoir could be reduced after the desired temperature had been reached.

Although the energy requirement for semi-active thermal persist is small, it must be met in some manner. One method is to use re-chargeable batteries to store energy during normal operation, and then use these batteries to drive the resistive heating element during a power interruption. It would also be advisable to use the rechargeable batteries to keep the fan running as long as possible in order to dissipate any heat contained in the heat sink, reducing its temperature to as close to ambient as possible. This will minimize the reverse thermal gradient and therefore reduce the risk of damaging the electronic components should the thermal persist batteries run out before main power is restored to the system.

In each of the above scenarios (passive thermal persist and semi-active thermal persist), the reverse flow of heat is stopped by "shutting down" a specialized heat pipe. Active thermal persist, on the other hand, does not require the use of a specialized heat pipe, but rather changes the characteristic of the thermoelectric modules to block the reverse flow of heat. This effectively "disconnects" the heat sink and prevents it from conducting heat back into the electronic components.

Active thermal persist takes advantage of a common characteristic of all thermoelectric modules. The temperature differential between the two surfaces of a thermoelectric module will generally increase as the input voltage is increased. Conversely, this differential will decrease as the input voltage is reduced, even to the point where it becomes negative when the voltage drops below a certain point - the barrier threshold voltage. Active thermal persist takes advantage of this principle by only supplying the barrier threshold voltage in order to prevent the reverse flow of heat through the thermoelectric module(s), thermally "disconnecting" the electronic component from the heat sink. Net energy requirements may be reduced if the input voltage is pulsed, perhaps at a slightly higher voltage to compensate for the time between pulses.

If active thermal persist is used to prevent the reverse flow of heat in a distributed thermal managements system for electronic components, one must be careful to consider whether it is required at one or perhaps both ends of the heat pipe. Intuitively the one end approach consumes less power, but active thermal persist may be required at both ends depending on the thermal masses within the system, the allowable temperature rise within the cabinet, etc.

As in the previous example, power for active thermal persist may come from a set of internal rechargeable batteries. Alternatively, as in the case of a PC or workstation, it may come from an external UPS as part of a standard shutdown sequence. Regardless of source, active thermal persist will require more power than the semi-active approach.

When to use Passive, Semi-Active, or Active Thermal Persist

With at least three possible approaches, one must be able to decide which approach is best for a particular installation. The following table will provide some useful information.

As can be seen from the above table, passive thermal persist offers reasonable response time and consumes no power, but it does require a specially designed heat pipe. Semi-active thermal persist, on the other hand, offers an improved response time but does consume some power and still requires a specially designed heat pipe. Finally, active thermal persist consumes more power but the response is immediate and it can be applied to standard heat pipes or even heat conduits. (The latter use straight conductance to convey heat from source to destination. The distributed thermal management systems for electronic components may be built with heat conduits providing that the resulting increase in temperature differential is acceptable for the application. This becomes more practical as the length of the heat conduit becomes shorter.) Thermal Diodes

One component that has been long sought after by thermal designers is a truly effective thermal diode - a component that will theoretically allow the free flow of heat in one direction while completely blocking the reverse flow of heat. Modular thermal diodes may in fact be produced using the thermal persist technologies described in the previous above. A modular design lends itself to mass production, and may be used as a universal component in a variety of thermal systems.

The most practical modular design may be in the format of a spacer block. Many thermoelectric systems utilize spacer blocks to increase the distance between hot and cold surfaces of the thermoelectric module to allow for better insulation between the two surfaces while still facilitating the required heat flow. A modular thermal diode spacer block has the same characteristics except that it will only allow heat to flow in one direction.

As illustrated in Figure 6, thermal diode spacer block 6 may be constructed by placing insulating layer 64 between upper dispersion plate 62 and lower dispersion plate 60. The dispersion plates are then thermally connected through integrated heat pipe 70, allowing heat to flow between the two outside surfaces of thermal diode spacer block 6. Since integrated heat pipe 70 represents the only thermal channel between the two dispersion plates, it may be used to control the flow of heat through thermal diode spacer block 6. This being the case, any of the previously described thermal persist technologies, passive, semi-active, or active, may be used to ensure that the heat can only flow in one direction. The resulting component is extremely modular, and it can be used in a thermal system in the same manner as one would use a standard aluminum (or other thermally conductive material) spacer block. The most notable difference is that in this case the "spacer block" is actually a thermal diode.

One of the challenges faced in the design of thermal diode spacer blocks is that the distance between the two dispersion plates is extremely short given that spacer blocks are normally in the area of Y∑" thick. The result is that integrated heat pipe 70 can't really be directly connected between the two dispersion plates as depicted in Figure 6. Rather, integrated heat pipe 70 needs to be connected between the two diagonally opposed corners of the dispersion plates, resulting in a length that approximates the much larger diagonal dimension of the dispersion plates. This "Z" like geometry can easily be seen in Figures 7 and 8. Figure 7 presents a side view of thermal diode spacer block 6 with "Z" geometry. In this case integrated heat pipe 70 is connected to the right corner of upper dispersion plate 62, at upper connection point 70a, and connected to the left corner of lower dispersion plate 60, at lower connection point 70b. The result is a distinctive "Z" geometry that substantially extends the length of integrated heat pipe 70 as it thermally connects the two dispersion plates.

Figure 8 presents a top view of thermal diode spacer block 6 with "Z" geometry. This figure illustrates that integrated heat pipe 70 may be connected to the rear right comer of upper dispersion plate 62, at upper connection point 70a, and connected to the forward left corner of lower dispersion plate 60 (not shown as it is directly underneath upper dispersion plate 62), at lower connection point 70b. This connection of heat pipe 70 to diagonally opposed corners of the two dispersion plates further extends the length of heat pipe 70.

Now that a suitable design has been developed, one that accommodates a full heat pipe between the two dispersion plates, passive or semi-active thermal persist technologies can be applied to the device to perform the thermal diode function.

Note that in a passive thermal persist configuration, no input is required to stop the flow of heat in the reverse direction. Once power is removed from the thermoelectric module, the thermal gradient will quickly reverse, automatically triggering the correct response in the thermal diode.

In the semi-active thermal persist configuration, a small "turn off voltage must be applied to the control input to stop the flow of heat through the thermal diode. Although this does consume a small amount of power, the faster response time will provide advantages in applications where the reverse flow of heat must be more closely controlled. It may also be used in temperature control circuits where power is intentionally removed from the TE module when a low set point has been reached. The semi-active thermal diode spacer block may be used to block the flow of heat back into the cold area during the "off cycle, effectively reducing the parasitic load on the system. This approach may save on net energy usage, depending on the system.

A further advantage of thermal persist diode spacer blocks is that they have less thermal mass than a standard aluminum spacer block while offering the same or better forward thermal conductance. This will contribute to the responsiveness of the system during normal operation. Figure 9 provides simplified view of the major components in the distributed thermal management system for electronic components.

Dispersion plate 16 may be iri direct thermal communication with electronic component 18 in order to smooth out the impact of any heat flux variations within electronic component 18, and to more closely match the area of the top surface of electronic component 18 with that of the bottom surface of internal thermoelectric module 14.

Internal thermoelectric module 14 may be in direct thermal communication with dispersion plate 16 such that the heat produced by electronic component 18 may be extracted through dispersion plate 16. Internal thermoelectric module 14 pumps this heat into the evaporator end of heat pipe 12. Internal thermoelectric module 14 is controlled such that the evaporator end of heat pipe 12 is maintained at ambient temperature levels. The normal temperature differential across internal thermoelectric module 14 will then ensure that electronic component 18 is operating at below ambient temperatures.

Electronic module 18, dispersion plate 16, thermoelectric module 14, and the evaporator end of heat pipe 12 may be encased within insulated and hermetically sealed area 20 to prevent condensation problems. Note that heat pipe 12 does not need to be insulated along its length since it will be operating at ambient or near ambient temperatures.

The condenser end of heat pipe 12 protrudes through cabinet wall 10 where it may then be in direct thermal communication with external thermoelectric module 8. External thermoelectric module 8 contributes to the amount of heat that may be extracted from electronic component 18 through dispersion plate 16, internal thermoelectric module 14, and heat pipe 12. External thermoelectric module 8 also provides a temperature differential between heat pipe 12 and external heat sink 4 such that heat sink 4 operates at above ambient temperatures. Ultimately this provides a temperature gradient between heat sink 4 and the outside air, allowing heat to be dispersed to the outside air. Fan 2 may be used to aid this process through forced convection.

Thermal diode spacer block 6 may be inserted between heat sink 4 and the condenser end of heat pipe 12, ideally between external thermoelectric module 8 and the condenser end of heat pipe 12, in order to prevent the reverse flow of heat back through the system and into electronic component 18 in the event of a power outage or power down situation. During normal operation, thermal diode spacer block 6 allows the free flow of heat from electronic component 18 out to the surrounding air through heat sink 4. Thermal diode spacer block 6 may be based on passive or semi-active thermal persist technologies. Alternatively, thermal diode spacer block 6 may be a virtual device based on active thermal persist technology, in which case the physical component represented as thermal diode spacer block 6 will become a passive spacer block.

The external sub-assembly, consisting of external thermoelectric module 8, thermal diode spacer block 6, heat sink 4, and fan 2, may be configured to accept thermal input from more than one internal sub-assembly, consisting of heat pipe 12, internal thermoelectric module 14, dispersion plate 16, electronic component 18, and insulated and hermetically sealed area 20. In this manner, multiple electronic components may be cooled using a single external heat sink and fan.

Figure 10 provides a more detailed view of the components in the distributed thermal management system for electronic components.

The connection of heat pipe 12 to internal thermoelectric module 14 and external thermoelectric module 8 may be accomplished through internal interface plate 32 and external interface plate 34 respectively. The interface plates serve to increase the effective surface area of heat pipe 12, at the ends, so that it more closely matches that of the thermoelectric modules. This increases the effectiveness of the thermal communication between the heat pipe and the thermoelectric modules.

Internal interface plate 32 may be manufactured as an integral part of the internal sub- assembly to save cost and improve thermal performance. External interface plate 34, on the other hand, serves as the thermal junction point between the internal and external sub- assemblies and as such must be configured to accept the thermal attachment of one or more heat pipes 12.

The temperatures and heat flow rates throughout the distributed thermal management system for electronic components are controlled by controller 26. Controller 26 accepts input signals from temperature sensors 36, 38, and 40 located on dispersion plate 16, internal interface plate 32, and external interface plate 34 respectively. Controller 26 then controls the operation of power supply 28 accordingly. Power supply 28 is connected to internal thermoelectric module 14 and external thermoelectric module 8. Power supply 28 is intentionally located outside the cabinet wall to prevent any unnecessary thermal loads inside the cabinet.

The internal sub-assembly is connected to controller 26 and power supply 28 through internal signal wire 22 and internal power wire 24 respectively. Internal signal wire 22, internal power wire 24, and heat pipe 12 are all combined within composite conduit 30 which runs through the cabinet to cabinet wall 10. This facilitates the mass production of internal sub-assembly components and addresses cable management problems within the cabinet. Note that insulated and hermetically sealed area 20 forms a seal with composite conduit 30 as it enters insulated and hermetically sealed area 20 to prevent the possible ingress of heat and humidity.

In some cases controller 26 and power supply 28 may be connected to thermal diode spacer block 6 to control the operation thereof. This depends on the type of thermal diode spacer block 6 which has been implemented, and whether or not its operation requires input power.

Also, in some cases controller 26 may communicate with the software associated with the electronic components contained within the cabinet. This allows for software monitoring of the distributed thermal management system, and a possible integration with other software controlled thermal management solutions.

Figure 11 provides further detail regarding dispersion plate 16 and the use of phase change materials to enhance thermal stability and performance.

Electronic component 18 may be mounted on circuit board 50 with pins 52 protruding through circuit board 50. Internal thermoelectric module 14 may pump heat from electronic component 18 through heat flux dispersion plate 54, thermal ballast 56, and dispersion plate 16. The heat then flows through heat pipe 12 to the external subassembly as previously described.

Electronic component 18, heat flux dispersion plate 54, thermal ballast 56, dispersion plate 16, and internal thermoelectric module 14 may be encased with insulation 20 to prevent this assembly from absorbing heat from the air surrounding circuit board 50. Further, the small space between insulation 20 and circuit board 50 may be sealed with hermetic seal 58 to prevent the ingress of moisture into, and therefore the formation of condensation within, the assembly. Internal thermoelectric module 14 will initially pump heat from thermal ballast 56 through dispersion plate 20. In the case of a thermal ballast 56 comprised of phase change material, a substantial amount of latent heat may be removed from thermal ballast 56 as it changes phase from a liquid to a solid. Thermal ballast 56 will remain at a relatively constant temperature during this process, the temperature being dependent upon the characteristic freezing point of the phase change material contained therein.

Electronic component 18 is a fluctuating thermal load that may vary widely over time. Furthermore, the heat flux density across the top surface of electronic component 18 will also vary widely over time, causing potentially damaging "hot spots" to form.

Heat flux dispersion plate 54 will serve to even out the heat flux density as seen by thermal ballast 56, allowing the heat to be absorbed across a greater surface area of thermal ballast 56. The net effect is that the heat may be absorbed more rapidly by thermal ballast 56.

Thermal ballast 56 will continue to absorb heat from electronic component 18 through heat flux dispersion plate 54 at varying rates as determined by the workload being handled by, and therefore the heat generated by, electronic component 18 at any one point in time. Again, in the case of a thermal ballast 56 comprised of phase change material, this process will occur at a relatively constant temperature as determined by the characteristic freezing point of the material contained therein. This will serve to keep electronic component 18 operating at a relatively constant temperature regardless of workload.

Internal thermoelectric module 14 will continue to pump heat away from thermal ballast 56 on a steady basis while electronic component 18 is operating. Thermal ballast 56 acts as "cold battery" that is able to compensate for the difference between the steady rate at which internal thermoelectric module 14 may pump heat from thermal ballast 56, and the variable rate at which electronic component 18 may produce heat and deliver it to thermal ballast 56.

When thermal ballast 56 is comprised of phase change material, internal thermoelectric module 14 will continue to pump heat away from thermal ballast 56 until such time as all of the latent heat has been removed and thermal ballast 56 is once again in a completely frozen state. Although internal thermoelectric module 14 pumps heat away from thermal ballast 56 at a much slower rate than thermal ballast 56 may absorb heat from electronic component 18, the system works because the total amount of heat removed from thermal ballast 56 over an extended period of time is equal to or greater than the total amount of heat that may be absorbed by thermal ballast 56 over the same period of time.

Figure 12 provides greater detail regarding thermal diode spacer block 6. It should be noted that this figure illustrates a passive thermal persist embodiment of the thermal diode, however it is also representative of the semi-active embodiments. The active thermal persist embodiment is not implemented as a physical thermal diode, but rather as a virtual thermal diode that is created by applying a small forward voltage to external thermoelectric module 8 after main power has been removed (reference Figure 10).

Thermal diode spacer block 6, in this embodiment, is constricted of lower dispersion plate 60 and upper dispersion plate 62 separated by insulation 64. Lower dispersion plate 60 contains evaporator 66a and evaporator / water trap 66b, and upper dispersion plate 62 contains condenser 68. Evaporator 66a is connected to condenser 68 through integrated heat pipe 70. In this manner, integrated heat pipe 70 forms the only thermal connection between lower dispersion plate 60 and upper dispersions plate 62.

During normal operation, heat flow 72 will flow from lower dispersion plate 60 to upper dispersion plate 62 through integrated heat pipe 70. Heat applied to lower dispersion plate 60 will produce vapor in evaporator 66a and evaporator / water trap 66b. This vapor will flow through integrated heat pipe 70 where it will condense in condenser 68. Excess liquid may accumulate in condenser 68, however it can always return to evaporator 66a since the wick, located on the inside surface of integrated heat pipe 70, is continuous between condenser 68 and evaporator 66a. The wick is not continuous between evaporator 66a and evaporator / water trap 66b since they are only connected through wickless u-tube 72.

When the thermal gradient is reversed, evaporation begins to take place in condenser 68, and condensation begins to take place in evaporator 66a and evaporator / water trap 66b. Operation is reversed but otherwise identical to "normal" mode including the fact that excess condensate tends to accumulate where the condensation takes place - now in evaporator 66a and evaporator / water trap 66b. However in this case the excess condensate can not return to the other end of integrated heat pipe 70 since there is no continuous wick between evaporator / water trap 66b and condenser 68 as evaporator / water trap 66b is only connected by wickless u-tube 72. Ultimately the vapor pressure within integrated heat pipe 70 will drop to the extent that integrated heat pipe 70 ceases to operate. Note that this is a passive shut down operation that occurs automatically upon the reversal of the thermal gradient.

When integrated heat pipe 70 is shut down in this manner, the only thermal connection between upper dispersion plate 62 and lower dispersion plate 60 is the conductive flow of heat through the walls of integrated heat pipe 70. This conductive flow of heat may be minimized by the installation of one or more insulating ring(s) 74. Insulating ring(s) 74 may be constructed of insulating material, and scribed into the wall of integrated heat pipe 70 such that any remaining wall thickness is minimized, and the integrity of the heat pipe is maintained - i.e. there is no loss in vapor pressure and the wick remains continuous along the length of integrated heat pipe 70.

When the normal thermal gradient has been restored, evaporation once again begins to take place in evaporator 66a and evaporator / water trap 66b. Since vapor can easily flow through wickless u-tube 72 (i.e. the absence of a wick only prevents the flow of liquid), evaporator / water trap 66b soon dries out, sufficient vapor pressure is restored, and integrated heat pipe 70 returns to normal operation.

The present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. Certain adaptations and modifications of the invention will be obvious to those skilled in the art. Therefore, the above-discussed embodiments are considered to be illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Claims

CLAIMSWe claim;

1. A distributed thermal management system comprising: a. a heat pipe having an evaporator section and a condenser section; b. a power source; c. a first heat pump in thermal communication with a thermal load and in thermal communication with the evaporator section of said heat pipe; d. a second heat pump in thermal communication with the condenser section of said heat pipe and in thermal communication with a heat sink; wherein said heat pipe and said heat pumps are configured to withdraw heat from said thermal load and deliver said heat to said heat sink where it may be dispersed into the ambient air.

2. A distributed thermal management system as claimed in claim 1 , further comprising a dispersion plate located between said first heat pump and said thermal load.

3. A distributed thermal management system as claimed in claim 1 , further comprising a heat flux dispersion plate, a quantity of phase change material, and a further dispersion plate located between said first heat pump and said thermal load.

4. A distributed thermal management system as claimed in claim 1 , further comprising a composite conduit containing said heat pipe, the power supply wires for said first heat pump, and all signal monitoring and control wires connected to said first heat pump and said thermal load.

5. A distributed thermal management system as claimed in claim 1, wherein said thermal load, said first heat pump, and said heat pipe may be located inside an enclosed cabinet and wherein said second heat pump and said heat sink may be located on the outside of said enclosed cabinet.

6. A distributed thermal management system as claimed in claim 1 , further comprising a control system; wherein said control system is adapted to monitor and control all aspects of said distributed thermal management system, provide visual information regarding the performance of said distributed thermal management system to a user, and interface with operating software that may be associated with said thermal load to provide further indications to the user and to interoperate with other thermal management systems that may be associated with said thermal load.

7. A distributed thermal management system as claimed in claim 1 , wherein said thermal load may be controlled to operate at below ambient temperature, said heat pipe may be controlled to operate at approximately ambient temperature, and said heat sink may be controlled to operate at above ambient temperature.

8. A distributed thermal management system as claimed in claim 1, wherein said thermal load, and all thermal components in thermal communication with said thermal load and operating at below ambient temperature, may be insulated and hermetically sealed.

9. A distributed thermal management system as claimed in claim 1 , wherein said first heat pump may be one or more thermoelectric modules.

10. A distributed thermal management system as claimed in claim 1, wherein said second heat pump may be one or more thermoelectric modules.

11. A distributed thermal management system as claimed in claim 1 , further comprising a fan to disperse heat from said heat sink through forced convection.

12. A distributed thermal management system as claimed in claim 1, further comprising an interface plate located between said second heat pump and said condenser end of said heat pipe to allow the detachable thermal connection of multiple said heat pipes to a single said second heat pump and said heat sink assembly.

13. A distributed thermal management system as claimed in claim 1 , wherein said thermal load may be an electronic component.

14. An enhanced distributed thermal management system comprising: a. a heat pipe having an evaporator section and a condenser section; b. a power source; c. a first heat pump in thermal communication with a thermal load and in thermal communication with the evaporator section of said heat pipe; d. a second heat pump in thermal communication with the condenser section of said heat pipe and in thermal communication with a heat sink; e. a thermal diode located between said heat sink and said condenser section of said heat pipe; wherein said heat pipe and said heat pumps are configured to withdraw heat from said thermal load and deliver said heat to said heat sink where it may be dispersed into the ambient air, and wherein said thermal diode is configured to prevent the reverse flow of heat.

15. An enhanced distributed thermal management system as claimed in claim 14, further comprising a dispersion plate located between said first heat pump and said thermal load.

16. An enhanced distributed thermal management system as claimed in claim 14, further comprising a heat flux dispersion plate, a quantity of phase change material, and a further dispersion plate located between said first heat pump and said thermal load.

17. An enhanced distributed thermal management system as claimed in claim 14, further comprising a composite conduit containing said heat pipe, the power supply wires for said first heat pump, and all signal monitoring and control wires connected to said first heat pump and said thermal load.

18. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal load, said first heat pump, and said heat pipe may be located inside an enclosed cabinet and wherein said second heat pump and said heat sink may be located on the outside of said enclosed cabinet.

19. An enhanced distributed thermal management system as claimed in claim 14, further comprising a control system; wherein said control system is adapted to monitor and control all aspects of said distributed thermal management system, provide visual information regarding the performance of said distributed thermal management system to a user, and interface with operating software that may be associated with said thermal load to provide further indications to the user and to interoperate with other thermal management systems that may be associated with said thermal load.

20. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal load may be controlled to operate at below ambient temperature, said heat pipe may be controlled to operate at approximately ambient temperature, and said heat sink may be controlled to operate at above ambient temperature.

21. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal load, and all thermal components in thermal communication with said thermal load and operating at below ambient temperature, may be insulated and hermetically sealed.

22. An enhanced distributed thermal management system as claimed in claim 14, wherein said first heat pump may be one or more thermoelectric modules.

23. An enhanced distributed thermal management system as claimed in claim 14, wherein said second heat pump may be one or more thermoelectric modules.

24. An enhanced distributed thermal management system as claimed in claim 14, further comprising a fan to disperse heat from said heat sink through forced convection.

25. An enhanced distributed thermal management system as claimed in claim 14, further comprising an interface plate located between said second heat pump and said condenser end of said heat pipe to allow the detachable thermal connection of multiple said heat pipes to a single said second heat pump and said heat sink assembly.

26. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal load may be an electronic component.

27. An enhanced distributed thermal management system as claimed in claim 14, therein said thermal diode may be virtually implemented by using one or more thermoelectric module(s) as said second heat pump, and by applying a small forward voltage to said thermoelectric module(s) after main power has been removed.

28. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal diode may be a modular spacer block component comprised of a dispersion plate on each of the outside thermal communication surfaces, an insulating layer between said dispersion plates, and an integrated heat pipe in thermal communication with said dispersion plates; said integrated heat pipe attached to diagonally opposed comers of said dispersion plates and configured to be of sufficient length.

29. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal diode may be a modular spacer block component comprised of a dispersion plate on each of the outside thermal communication surfaces, an insulating layer between said dispersion plates, and an integrated heat pipe in thermal communication with said dispersion plates; said integrated heat pipe having two evaporator sections connected through a wickless u-tube that only allows the passage of vapor.

30. An enhanced distributed thermal management system as claimed in claim 14, wherein said thermal diode may be a modular spacer block component comprised of a dispersion plate on each of the outside thermal communication surfaces, an insulating layer between said dispersion plates, and an integrated heat pipe in thermal communication with said dispersion plates; said integrated heat pipe having an externally heated reservoir section attached to a condenser section; said heated reservoir section containing an inert gas which may expand, when heated, to displace operating vapor contained within said condenser section.