Two-phase heat exchanger for high heat flux chipsets

Abstract

A two-phase heat exchanger includes a manifold assembly configured with an input manifold for a fluid inlet flow and an output manifold for a two-phase output flow, and a heat transfer matrix coupled to the manifold assembly. The heat transfer matrix includes a plurality of microchannels in fluid communication with the input manifold and the output manifold. Each of the plurality of microchannels includes a flow restriction configured to inhibit redistribution of the fluid inlet flow and to balance pressure drops between certain ones of the plurality of microchannels. Pairs of the plurality of microchannels are in a counterflow configuration with an inlet of one microchannel adjacent to an outlet of an adjacent microchannel, where the counterflow configuration increases packing density and reduces pressure drop in the heat transfer matrix.

Description

Technical Field

[0001] The present invention relates to heat exchangers, and more particularly to two-phase cooling plates comprising a flow restrictor and a counterflow diversion channel geometry to passively limit flow redistribution. Background Technology

[0002] Semiconductor devices in data centers or supercomputers generate significant amounts of heat during operation, requiring cooling. The transistors and active components on the semiconductors of the central processing unit (CPU) or graphics processing unit (GPU) of a computer or server consume substantial amounts of power, which is dissipated as heat. Active cooling is necessary to keep the temperature of these active semiconductor devices below their rated maximum temperatures. For example, the maximum operating temperature limit for silicon semiconductor devices is typically between 80°C and 95°C, while the maximum operating temperature for silicon carbide semiconductor devices is approximately 175°C. Exceeding these respective rated maximum temperatures increases the likelihood of semiconductor device failure, thus necessitating effective cooling to ensure the proper functioning of data centers or supercomputers and prevent this from happening.

[0003] The U.S. Department of Energy (DOE) has recognized the need to overcome technological hurdles associated with developing high-performance, energy-efficient cooling solutions for data centers and has announced up to $42 million in funding to address this issue. According to the DOE, data centers, which house computers, storage systems, and computing infrastructure, account for approximately 2% of total electricity consumption in the United States, while data center cooling can account for 40% of total data center energy consumption. Reducing the amount of energy used for cooling in data centers will help lower the operational carbon footprint associated with data center power supply and cooling.

[0004] The most common cooling method in today's data centers is using air as a coolant. Cool air is forced into a cooling device, i.e., a heatsink, by a fan and heated in the process. This hot air is then cooled by the heatsink, which removes heat from the air and exhausts it into the ambient air outside the data center. These heatsinks can be, for example, radiators, water-cooled towers, or compressors / coolers. Heatsinks connected to the CPU / GPU are made of conductive materials, typically aluminum or copper, and have fins that are away from the CPU / GPU surface. These fins increase the surface area of ​​the device that can transfer heat into the air, improving heat dissipation performance. These heatsinks use thermal interface materials (e.g., thermal grease or thermally conductive flexible pads) attached to the silicon to form a low-resistance thermal bond between the silicon and the heatsink. These thermal interface materials are necessary because neither the silicon nor the heatsink surfaces are perfectly flat, and the air gap between the two devices results in high thermal resistance and poor cooling performance.

[0005] Supercomputers or data centers performing more intensive computing (often referred to as "high-performance computing") cannot be effectively cooled by air because the thermal loads in CPUs / GPUs are far higher than in traditional data centers. For these applications, single-phase liquid coolants are used to dissipate heat directly from the CPU / GPU. Traditionally, water blocks or cold plates are mounted directly on the CPU / GPU, with cold water pumped into them and hot water discharged. The hot water is then cooled by a heatsink, which dissipates the heat into the ambient air.

[0006] The semiconductor industry predicts that within the next five years, future server chipsets will generate 200% more heat per unit area than current devices. For example, the most advanced graphics processing units currently available, such as NVIDIA's H100 GPU, with 80 billion transistors and two types of cores, are designed to be nine times faster than their predecessors. At peak performance, the H100 generates 125 W / cm² on its processor cores. 2 The average heat flux output exceeds 1100 W. However, the average heat flux of the next-generation device is expected to be 235 W / cm². 2 The output power will exceed 4000W. The significant increase in total power and power density is expected to test the limits of single-phase liquid cooling, as the increased flow rate approaches the microchannel erosion limit and increases pumping power requirements.

[0007] The aforementioned options are currently deployed in high-performance computing (HPC) data centers and have achieved varying degrees of success with current-generation CPUs / GPUs. However, two factors render these cooling technologies insufficient for future demands. First, CPUs / GPUs will generate more heat, increasing the performance requirements of cooling systems. Second, data centers need to be more energy-efficient, and new guidelines mean that cooling systems need to remove more heat while consuming less power. These two combined factors mean that future cooling systems need to reduce thermal resistance, decrease pump power consumption, and improve overall cooling performance to meet the needs of next-generation semiconductor devices.

[0008] Modern data centers currently employ single-phase cold plates as the most advanced solution to eliminate waste heat from high-performance, high-power computing and graphics processing units. Single-phase liquid cooling utilizes the sensible heat capacity of a liquid to store and transfer heat. Single-phase liquid cold plates have low thermal resistance but require significant flow rates, large input flow rates, and a large temperature difference, or a combination thereof, between the working surface to remove substantial amounts of heat.

[0009] As power levels continue to rise, the increasing flow rates and near-temperature requirements associated with single-phase cooling will face practical limitations. Two-phase flow in microchannels has proven to be the next logical approach for cooling next-generation semiconductor devices. Two-phase cold plates utilize the latent heat of phase change, absorbing heat by vaporizing all or part of the liquid coolant into a gas phase. Since the latent heat is typically several orders of magnitude greater than the sensible heat capacity of a single-phase liquid, two-phase flow rates are generally significantly lower than single-phase alternatives, thus saving energy in terms of pumping power. However, two-phase microchannel heat sinks are prone to flow, pressure, and temperature oscillations, which are related to, but not limited to, highly non-uniform local heat flux and non-uniform local heat transfer coefficients within the channels themselves. Despite research to date and ongoing, several challenges remain to be overcome before the widespread adoption of this technology.

[0010] Prior art disclosures regarding two-phase microchannel cold plates involve placing microchannels near an actively heated surface where the working fluid boils to remove heat. Two-phase microchannel cold plates have not been widely adopted in industry, primarily due to instability and difficulties in managing flow redistribution. Unlike single-phase flow, where a higher heat load reduces fluid viscosity and increases flow velocity towards channels with higher heat loads, two-phase flow generates vapor, increasing the pressure drop in these channels and forcing liquid into channels with lower heat loads. This inherent negative feedback mechanism in two-phase physics causes flow to redistribute undesirably, thus reducing the performance of many prior art two-phase cold plate designs. Summary of the Invention

[0011] This disclosure describes a two-phase heat exchanger. In one embodiment, the two-phase heat exchanger includes a manifold assembly and a heat transfer matrix. The manifold assembly is configured to have an input manifold for fluid inlet flow and an output manifold for two-phase output flow. The heat transfer matrix is ​​coupled to the manifold assembly. The heat transfer matrix includes a plurality of microchannels in fluid communication with the input and output manifolds. Each of the plurality of microchannels includes a flow restriction configured to suppress the redistribution of fluid inlet flow and to balance pressure drops between certain microchannels. In one embodiment, pairs of microchannels in the plurality of microchannels are configured in a counter-current configuration, with the inlet of one microchannel adjacent to the outlet of an adjacent microchannel. This counter-current configuration increases the packing density and reduces the pressure drop in the heat transfer matrix. Attached Figure Description

[0012] This disclosure can be best understood from the following detailed description when read in conjunction with the accompanying drawings. It should be emphasized that, by convention, the various features in the drawings are not drawn to scale. Instead, for clarity, the dimensions of the various features have been arbitrarily enlarged or reduced.

[0013] Figure 1This is a perspective view of a two-phase heat exchanger according to this teaching.

[0014] Figure 2 It is based on this teaching Figure 1 The diagram shows a partial exploded view of a two-phase heat exchanger.

[0015] Figure 3 It is attached to the semiconductor device according to this teaching. Figure 1 A perspective view of a two-phase heat exchanger.

[0016] Figure 4 It is based on this teaching Figure 3 The diagram shown is a partial exploded view of a two-phase heat exchanger attached to a semiconductor device.

[0017] Figure 5 This is an enlarged view of a partial cross-section, illustrating the application of this teaching. Figure 2 The internal flow path within a two-phase heat exchanger.

[0018] Figure 6 This is a top planar cross-sectional view of the manifold spacer layer according to this teaching.

[0019] Figure 7 This is a top plan cross-sectional view of the manifold distribution layer according to this teaching.

[0020] Figure 8 This is a top plan cross-sectional view of the port layer of the inlet and outlet regions of the heat transfer channel, in accordance with this teaching.

[0021] Figure 9 This is a top planar cross-sectional view of a heat transfer channel layer with countercurrent expansion microchannels according to this teaching.

[0022] Figure 10 This is a schematic diagram illustrating the profile of an alternative expansion channel according to this teaching.

[0023] Figure 11 This is a schematic diagram depicting the typical spatial relationship between the high-bandwidth memory (HBM) and CPU / GPU cores of a server chipset.

[0024] Figure 12 This is a schematic diagram of a two-phase heat exchanger for cooling a single heat-generating region including a first flow path layout, in accordance with this teaching.

[0025] Figure 13 This is a schematic diagram of a two-phase heat exchanger for cooling a multi-zone heating zone according to this teaching, the multi-zone heating zone including a second flow path layout with minimum flow redistribution. Detailed Implementation

[0026] Reference will now be made in more detail to embodiments, examples of which are shown in the accompanying drawings. Wherever possible, the same reference numerals will be used throughout the drawings and description to refer to the same or similar parts.

[0027] As used herein, the terms “determine” and “identify” or any variations thereof include selecting, determining, calculating, searching, receiving, establishing, obtaining or otherwise identifying or determining using one or more apparatuses and methods shown and described herein.

[0028] As used herein, the terms “example,” “implementation,” “aspect,” “feature,” or “element” are used as examples, instances, or illustrations. Unless expressly indicated, any example, embodiment, implementation, aspect, feature, or element is independent of each other and may be used in combination with any other example, embodiment, implementation, aspect, feature, or element.

[0029] The term “or” as used herein is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless otherwise stated or the context clearly indicates otherwise, “X includes A or B” is intended to mean any natural inclusive arrangement. That is, if X includes A; X includes B; or X includes both A and B, then “X includes A or B” holds true in any of the foregoing cases. Furthermore, the pronouns “a” and “an” as used in this application and the appended claims should generally be interpreted as “one or more” unless otherwise stated or clearly indicated from the context as singular.

[0030] As used herein, unless otherwise expressly stated, any singular term may include its plural form. For example, "computer that stores data and runs software" may include a single computer that stores data and executes software, or it may include two computers—a first computer that stores data and a second computer that runs software. Furthermore, "computer that stores data and runs software" may include multiple computers that together store data and execute software. At least one of the multiple computers stores data, and at least one of the multiple computers runs software.

[0031] As used herein, unless otherwise expressly stated, the terms fluid and / or coolant fluid may refer to, but are not limited to, electronic coolant liquids containing perfluorinated compounds (PFCs), water, and / or water-glycol mixtures (salt water).

[0032] Furthermore, for the sake of simplicity, although the accompanying drawings and descriptions may include sequences or series of steps or stages, the elements of the methods disclosed herein may appear in various orders or simultaneously. Additionally, the elements of the methods disclosed herein may appear together with other elements not explicitly presented or described herein. Furthermore, not all elements of the methods described herein are required to implement the method according to this disclosure and claims. Although aspects, features, and elements are described herein in specific combinations, each aspect, feature, or element may be used independently or in combination with other aspects, features, or elements.

[0033] Furthermore, the accompanying drawings and descriptions provided herein have been simplified to illustrate aspects of the described embodiments that are relevant to a clear understanding of the processes, machines, and / or manufacturers disclosed herein, while other aspects that may be found in typical similar apparatuses, systems, and methods have been omitted for clarity. Therefore, those skilled in the art will recognize that implementing the apparatuses, systems, and methods described herein may require or be expected of other elements and / or steps. However, because such elements and steps are not conducive to a better understanding of the disclosed embodiments, a discussion of these elements and steps may not be provided herein. However, based on the discussion herein, this disclosure is considered to inherently include all such elements, variations, and modifications to the described aspects that are known to those skilled in the art.

[0034] Two-phase instability and flow redistribution management remain among the challenges for the widespread adoption of high heat flux two-phase cold plate technology in industry. One of the main challenges facing two-phase heat exchangers is thermal fluid instability, which can lead to localized drying and associated increases in thermal resistance during operation.

[0035] This paper describes a two-phase cooling plate heat exchanger with reduced thermal resistance for heating or cooling high-performance computing chipsets, data center servers, automated intelligent supercomputers, power electronics, laser modules, and other high-power, high-heat-flux heating or cooling applications. The teachings described herein mitigate the instability and flow redistribution management challenges of traditional two-phase microchannel cold plates. The two-phase heat exchanger described provides efficient, high-cooling-capacity thermal management and includes design features to mitigate the effects of uneven flow distribution.

[0036] In embodiments, the disclosed two-phase cooling plate comprises a microchannel cooling plate pressed against a heating device. The two-phase cooling plate includes the function of passively limiting the flow redistribution of spatially and temporally varying heat loads common in high-performance computing applications. The teachings disclosed herein effectively shape the pressure drop and flow rate characteristics of the cooling plate to minimize instability regions within the application operating range of the two-phase cooling plate.

[0037] In one implementation, the two-phase cooling plate includes flow restrictions at multiple inlets of the microchannel array to suppress flow redistribution from one channel to another, while the counter-flow channel geometry reduces the overall pressure drop and increases the packing density.

[0038] In this implementation, the two-phase cold plate includes flow limiting and diffusion channel geometry, as well as mechanisms to further mitigate uneven flow distribution and instability for more complex loads, i.e., loads involving multiple regions, each with its own heat generation rate, which may differ spatially and temporally, such as server microprocessor chipsets used in data center servers handling complex loads. Such chipsets have one or more processing cores surrounded by additional regions dedicated to high-bandwidth memory. The implementation employs a serial-parallel flow architecture, distributing channels in the cold plate close to the processor cores to suppress flow redistribution under various operating conditions. The processing cores typically do not appear as monolithic entities but rather exhibit localized hotspots within the processing footprint as different regions of the core are activated to perform various computations.

[0039] Figure 1 This is a perspective view of a two-phase heat exchanger 10 according to this teaching. Figure 2 This is a partial exploded view of the two-phase heat exchanger 10 according to this teaching. Figure 3 This is a perspective view of a two-phase heat exchanger 10 attached to a semiconductor device 24 in accordance with this teaching. Figure 4 This is a partial exploded view of a two-phase heat exchanger 10 with attached semiconductor devices 24, in accordance with this teaching.

[0040] In an embodiment, the two-phase heat exchanger 10 may include, but is not limited to, a heat transfer matrix or cooling plate 12, a manifold assembly 14, an inlet manifold 16, at least one inlet port 18, and at least one outlet port 20. The heat transfer matrix or cooling plate 12 includes a heat transfer surface 22. The heat transfer matrix or cooling plate 12 is supplied with flow via the manifold assembly 14. The inlet manifold 16 links the manifold assembly 14 to at least one inlet port 18 and at least one outlet port 20.

[0041] In embodiments, the heat transfer matrix 12, manifold assembly 14, and / or combinations thereof may include microchannel geometries, for example, as shown and described in U.S. Patent 8,474,516, published July 2, 2013, entitled "Heat Exchanger Having Winding Micro-channels," assigned and / or owned by the applicant, which is incorporated herein by reference as if proposed ("'516 Patent"). In summary, the '516 Patent heat exchanger includes a heat transfer component having wound microchannels, a manifold, and a cover. The wound design of the microchannels is defined by a nonlinear flow axis with multiple short-pitch and small-amplitude undulations that cause changes in flow direction, and two or more large-amplitude bends that cause flow reversal. In low flow rate per unit area applications, the wound microchannels allow users to customize the pressure drop to promote good flow distribution, achieve improved heat transfer uniformity, and keep the pressure drop above the bubble point of the heat transfer structure to prevent gas blockage. The wound microchannels also increase the heat transfer coefficient. The heat transfer component includes one or more heat transfer layers, each having multiple inlet openings and corresponding outlet openings. Each wound microchannel is in fluid communication with at least one inlet opening and at least one corresponding outlet opening, allowing cooling fluid to enter through the inlet opening, flow along the microchannel, and exit through the outlet opening. The openings are arranged in rows through each layer, each extending from a first surface to a second surface of each heat transfer layer. A manifold supplies fluid to each inlet opening of the heat transfer component and receives fluid from each outlet opening of the heat transfer component. The manifold distributes and collects fluid throughout the effective heat transfer area to promote uniform heat transfer across the entire area. Fluid enters the heat exchanger through an inlet port fluidly connected to an inlet manifold that distributes fluid along the y-axis of the manifold. The fluid is then fed to inlet channels fluidly connected to the inlet manifold, such that the fluid is distributed along the x-axis of the manifold by the inlet channels. Multiple outlet channels intersecting the inlet channels collect outlet fluid along the x-axis of the manifold and deliver it to an outlet manifold that collects fluid along the y-axis of the manifold and delivers it to the outlet port. The functions of distributing and collecting fluid to the heat transfer surface, and transferring heat between the fluid and the surface, are achieved by the manifold and the heat transfer components, respectively. This functional separation allows for the selection of the channel geometry in each component to fulfill its respective function. The configuration of the wound microchannels can be modified according to specific applications, but the microchannel axis remains nonlinear in all applications.

[0042] Return to Figures 1-4 To cool the device, the heat transfer surface 22 of the heat transfer matrix 12 is pressed against the device 24 to be cooled. In an embodiment, a thermal interface material 26 may be placed between the heat transfer matrix 12 and the device 24 to improve thermal contact and reduce interfacial thermal resistance.

[0043] Figure 5 This is an enlarged view of a partial cross-section, showing the internal flow path within the two-phase heat exchanger 10 according to this teaching. Figure 6 This is a top plan sectional view of the manifold spacer layer and / or plate 28 in the manifold assembly 14 according to this teaching. Figure 7 This is a top plan sectional view of the manifold distribution layer and / or plate 34 in the manifold assembly 14 according to this teaching.

[0044] In an embodiment, the inlet and outlet fluid manifolds in manifold assembly 14 can distribute flow to and receive flow from the parallel microchannel array in heat transfer matrix 12. The inlet and outlet fluid manifolds can be, for example, suitable portions of one or more manifold spacer layers and / or plates 28 and / or suitable portions of one or more manifold distribution layers and / or plates 34. That is, manifold assembly 14 may include one or more manifold spacer layers and / or plates 28 and / or one or more manifold distribution layers and / or plates 34, wherein certain or defined regions and / or structures are used for the inlet of the fluid flow (inlet manifold), while certain or defined regions and / or structures are used for the outlet of the two-phase flow (outlet manifold). In an embodiment, the inlet and outlet fluid manifolds can distribute flow to multiple parallel arrays, as is known to those skilled in the art. The fluid and / or coolant fluid (composed of...) Figure 5 (Indicated by the light gray arrow in the diagram) fluid can enter the inlet manifold 16 through the inlet pipe 18. The inlet manifold 16 delivers fluid to the liquid side of the manifold assembly 14.

[0045] Each of one or more manifold spacers and / or plates 28 in the manifold assembly 14 may provide an open region 30 to allow inlet flow to diffuse to the inlet of a microchannel located near the active heat transfer surface (e.g., heat transfer surface 22). In an embodiment, the inlet flow diffuses laterally in a direction parallel to the heat transfer surface. The maximum lateral flow resistance in one or more manifold spacers and / or plates 28 is approximately 10% to 25% or less of the pressure drop across the microchannels of the heat transfer matrix 12 to promote a uniform mass flow distribution within the active microchannels, with variations between individual microchannels less than 20%, or more preferably less than 10%. Each of one or more manifold spacers and / or plates 28 also includes a hole through the laminate to form an outlet channel 32 for exiting the two-phase flow (i.e., at least as...) from the heat transfer matrix 12. Figure 5 The steam (indicated by the dark gray arrow) passes through manifold assembly 14 to avoid mixing with the inlet flow fluid. The term two-phase flow means or implies that the flow includes both steam and liquid.

[0046] Each of one or more manifold distribution layers and / or plates 34 may include a plurality of small liquid distribution holes 36 in a region adjacent to an open area or a liquid lateral flow area 30 on one or more manifold spacer layers and / or plates 28. The smaller liquid flow area across the plane of one or more manifold distribution layers and / or plates 34 (as shown by the plurality of small liquid distribution holes 36) increases the flow resistance through the manifold assembly 14 relative to the lateral flow path in one or more manifold spacer layers and / or plates 28 (e.g., as shown by the open area 30), thereby facilitating flow distribution to all microchannels in the heat transfer matrix 12. These liquid distribution holes 36 in one or more manifold distribution layers and / or plates 34 are positioned such that flow from upstream channels impinges on the solid regions of one or more manifold distribution layers or plates 34.

[0047] In embodiments, multiple manifold spacer layers and / or plates 28 may be stacked together to form a higher channel with lower lateral flow resistance, while multiple manifold distribution layers and / or plates 34 may be added to increase flow resistance through the plates. When the manifold distribution layers and / or plates 34 are used in a single manifold assembly 14, the liquid distribution holes 36 in the manifold distribution layers and / or plates 34 are preferably offset from each other to force the fluid to be laterally diverted before entering the distribution holes 36 on the adjacent downstream distribution plate. In embodiments, a series of stacked plates may be brazed together to create a flow geometry. However, as those skilled in the art will know, other methods can be used to construct, arrange, and manufacture the manifold assembly 14 to achieve the desired pressure drop and flow uniformity characteristics. These other methods include, but are not limited to, machined parts and 3D-printed manifold assemblies, which are considered alternative embodiments of the manifold assembly.

[0048] In this implementation, the heat transfer matrix 12 includes a plurality of microchannels 44 and fluid ports. These references Figure 8 and Figure 9 As shown, Figure 8 This is a top planar cross-sectional view of a stackable port layer 48 in a heat transfer matrix 12 according to this teaching, which accesses the inlet and outlet regions of heat transfer channels (e.g., microchannels 44) in the heat transfer matrix 12. Figure 9 This is a top plan longitudinal section view of the heat transfer channel layer 41 in the heat transfer matrix 12 with countercurrent expansion microchannels 44 according to this teaching.

[0049] In one implementation, the stackable port layer 48 may include inlet ports 50 and outlet ports 52, which are arrays of holes and / or channels providing fluid pathways from the manifold assembly 14 to the microchannels 44. Conductive fins are formed around the structure of the inlet ports 50 and outlet ports 52, which facilitate heat dissipation within the active heat transfer matrix region 12. Figure 8In the middle, the outline of the microchannel 44 in the adjacent layer is projected onto the stackable port layer 48 to show the connectivity with the microchannel 44 in the heat transfer channel layer 41.

[0050] The microchannel geometry in heat transfer matrix 12 includes many features described below to address the problem of uneven flow distribution within the cold plate under non-uniform heating conditions. For example... Figure 9 As shown, flow restriction 38 is located at the inlet 40 of each microchannel 44 to help balance the pressure drop between high-heat-load microchannels and low-heat-load microchannels. In embodiments, flow restriction can be implemented at the inlet, through the microchannel, at the microchannel, and / or combinations thereof. Flow restriction 38 is a geometric change in the flow path that increases the pressure drop in these regions by increasing friction and dynamic pressure loss characteristics. As known to those skilled in the art, flow restriction 38 can be implemented using one or more of the following: channel cross-sectional area, orifice, slot, orifice plate, pin fin, perforated layer, and / or combinations thereof, but is not limited thereto. In embodiments, flow restriction 38 can be used in manifold channel geometry (e.g., in one or more manifold spacer layers and / or plates 28 and / or one or more manifold distribution layers and / or plates 34) to further balance flow redistribution.

[0051] In this implementation, the geometry of the microchannels and / or heat transfer channels 44 expands from inlet 40 to outlet 42 (referred to herein as "expanded microchannel geometry"), i.e., increases, to reduce the pressure drop associated with the two-phase regions in the heat transfer matrix 12. The transition of the microchannels 44 from a smaller cross-sectional area near inlet 40 to a larger cross-sectional area near outlet 42 reduces the velocity of the two-phase flow. During operation, the generation of vapor in the microchannels 44 leads to flow acceleration, which increases the frictional pressure drop within the microchannels 44. The slower flow velocity associated with the expanded microchannels reduces the pressure drop characteristics of the microchannels located in the higher vapor mass region, thereby achieving a more balanced flow distribution under non-uniform heating conditions. Figure 10 As shown, the axial transition from the small cross-section near inlet 40 to the large cross-section near outlet 42 can take many different forms, including but not limited to straight lines 54, linear lines 56, piecewise straight lines 58, hyperbolic lines 60, and / or customizable profiles, provided that the disclosed functionality is achieved. In other words, various profiles can be used to expand the microchannel geometry. For high heat flux applications, the channel expansion or expanded microchannel geometry profile can be customized for a specific application based on various factors, including flow rate, heat load, and fill efficiency, to design the heat transfer and pressure drop characteristics of the cold plate for that application.

[0052] In one embodiment, the microchannels 44 are oriented in a counter-current configuration. In this configuration, the microchannels 44 are grouped in pairs, with the inlet 40 of each microchannel 44 adjacent to the outlet 42 of its paired neighboring microchannel. The counter-current configuration increases the packing density for the expansion channel geometry, effectively reducing pressure drop and thermal resistance due to increased flow and heat transfer area. The counter-current configuration also spatially averages the non-uniform heat transfer coefficients in the two-phase microchannels 44, thereby improving temperature uniformity at the heat transfer surface 22 between the heat transfer matrix and / or the cooling plate 12 and the semiconductor device 24. In embodiments where lower heat flux or layout prevents a counter-current configuration, the microchannels 44 may be arranged in a co-current or cross-current configuration, as is known to those skilled in the art.

[0053] In one embodiment, the microchannels 44 are arranged in one or more parallel configurations. Repeating elements constitute thermal pixels 46, which are arranged on the heat transfer matrix and / or cold plate footprint to form a microchannel array within the heat transfer matrix 12. For countercurrent microchannels, the thermal pixels 46 include a countercurrent pair. This microchannel layout utilizes multiple shorter microchannels to prevent excessive pressure drop in the heat transfer matrix 12 and limit saturation temperature variations along the length of the microchannels. In another embodiment, multiple heat transfer layers (heat transfer channel layer 41 and / or stackable port layer 48) can be stacked to create a three-dimensional heat transfer matrix 12, which increases the internal heat transfer surface area and flow cross-sectional area. As is known to those skilled in the art, fluid and / or coolant are supplied to and removed from the multiple microchannels via the manifold assembly 14.

[0054] Figure 11 This is a schematic diagram depicting the typical spatial relationship between high-bandwidth memory (HBM) and CPU / GPU cores in a server chipset. High-performance semiconductor chipsets typically combine various functional components, such as computing processing units (CPUs), graphics processing units (GPUs), high-bandwidth memory (HBM), etc., into a single chip or chipset. These functional components are often spatially grouped into regions, providing different thermal loads for heat transfer channels and / or microchannels within heat transfer matrix 12. Each region (e.g., a high-power CPU / GPU 62 or high-bandwidth memory 64) generates its own waste heat, which may vary spatially and temporally. Non-uniform thermal loads between parallel flow paths within the cooling plate and / or heat transfer matrix 12 can cause flow to be redistributed from microchannels with higher thermal loads to microchannels with lower thermal loads unless the pressure drop characteristics of the flow paths are reasonably balanced under various heating conditions. As described herein, each region is cooled by microchannels specifically designed for localized thermal loads. One or more manifold components can be connected in series, in parallel, or in combination to connect different regions, and their configuration allows the two-phase heat transfer and pressure drop characteristics of various parallel paths to be balanced under a desired set of operating conditions.

[0055] Figure 12 This is a schematic diagram of the flow connectivity for cooling a single heat-generating region including a first flow path layout, according to this teaching. In an embodiment, the flow connectivity shown is for a cooling plate that cools a device having a single heat-generating and / or functional area 62 with a time-varying but spatially uniform heat load. An inlet manifold 66 divides the incoming flow into multiple parallel channels 68. The inlet manifold 66 and the outlet manifold 70 disperse and collect the flow connected to the inlet / outlet 18 and the outlet 20 via an access header 16. This embodiment can be used as a standalone cooling plate or integrated as a sub-component into more complex flow structures, such as... Figure 13 As shown.

[0056] In many semiconductor devices, heat output varies with time and location. This is especially true for multi-functional chips, where memory cells 64 are located near the processor core 62, such as... Figure 11 As shown. In such devices, the memory 64 typically has lower cooling requirements compared to CPU or GPU cores. Significant variations in the thermal load of the various microchannels in a purely parallel array can lead to substantial flow redistribution in this versatile semiconductor device.

[0057] Figure 13 This is a schematic diagram of the flow connectivity of a two-phase heat exchanger for cooling a multi-zoned heating region according to this teaching, the multi-zoned heating region including a second flow path layout with minimal flow redistribution. In an embodiment, the flow connectivity shown is used for coupling a high-power region 62 (e.g., processor core 62, CPU, or GPU) to a low-power region 64 (e.g., memory 64) on a single semiconductor device. In this embodiment, the cooling plate has multiple main flow paths that have substantially the same flow and thermal load characteristics. For clarity, Figure 13 Two primary flow paths are shown, but those skilled in the art will know that there is no limitation on the number of primary paths that can be employed. Each primary flow path includes a high-power region 62 or a portion thereof connected downstream of one or more lower-power regions 64. A top inlet manifold 72 divides the flow into primary paths. Each primary flow path has a parallel subarray of cooling microchannels located above one or more lower-power regions 64. Each primary flow path is positioned to cool a subset of the lower-power regions. As fluid leaves the lower-power regions 64, intermediate manifolds 74 collect and mix the fluid from multiple channels on each lower-power channel subarray. This mixing homogenizes any localized heating variations in the upstream region, resulting in better uniformity entering subsequent regions. These intermediate manifolds 74 deliver the mixed flow to the downstream high-power region 62. After passing through the high-power region, the high-quality vapor exiting the microchannels is collected in a top outlet manifold 76, which connects to a cooling outlet 18 located on the inlet manifold 16.

[0058] In this implementation, both primary flow paths utilize fluidly distinct arrays of microchannel subarrays to cool the same high-power region 62. Flow from the primary paths is fluidly connected only at the top-level inlet manifold 72 and outlet manifold 76. In the high-power region 62, microchannels from the various primary flow paths are staggered among microchannels from other alternative primary flow paths. In this staggered arrangement, the microchannels are spatially adjacent but fluidly separated, ensuring that each primary flow path is exposed to nearly identical thermal load conditions, even under spatially and time-varying load conditions. This approach minimizes variations between parallel flow paths, thereby mitigating uneven flow distribution.

[0059] A two-phase heat exchanger is described. In one embodiment, the two-phase heat exchanger includes a manifold assembly and a heat transfer matrix. The manifold assembly is configured to have an input manifold for fluid inlet flow and an output manifold for two-phase output flow. The heat transfer matrix is ​​coupled to the manifold assembly. The heat transfer matrix includes a plurality of microchannels in fluid communication with the input and output manifolds. Each of the plurality of microchannels includes a flow restriction configured to suppress the redistribution of fluid inlet flow and to balance the pressure drop between certain microchannels. Pairs of microchannels in the plurality of microchannels are arranged in a counter-current configuration, with the inlet of one microchannel adjacent to the outlet of the adjacent microchannel. The counter-current configuration increases the packing density and reduces the pressure drop in the heat transfer matrix.

[0060] In one embodiment, the flow restriction is a reduction in the cross-sectional area at each inlet of the plurality of microchannels. In another embodiment, each of the plurality of microchannels has an expanded microchannel geometry extending from the inlet configured with the flow restriction to the outlet. In another embodiment, the profile of the expanded microchannel geometry is based on at least one of flow velocity, heat load, and microchannel filling efficiency. In another embodiment, the profile of the expanded microchannel geometry is one of linear, piecewise linear, and hyperbolic. In another embodiment, the heat transfer matrix further includes one or more stackable port layers configured to provide fluid pathways from the manifold assembly to the inlets and outlets of the plurality of microchannels, and one or more heat transfer channel layers, each comprising the plurality of microchannels. In another embodiment, the manifold assembly further includes one or more manifold spacer layers, each defining an open region for the diffusion of fluid inlet flow to the inlets of the plurality of microchannels, and each manifold spacer layer defining an outlet channel for the two-phase outlet flow discharged via the outlets of the plurality of microchannels. In another embodiment, the open region allows the fluid inlet flow to diffuse laterally through the plurality of microchannels in a direction parallel to the heat transfer surface of the heat transfer matrix. In one embodiment, the manifold assembly further includes one or more manifold distribution layers, each manifold distribution layer including multiple liquid distribution regions in a region adjacent to an open region within a manifold spacer layer to facilitate the distribution of fluid inlet flow to all multiple microchannels. In one embodiment, when the manifold distribution layers are stacked together in the manifold assembly, the liquid distribution regions of one manifold distribution layer are offset from the liquid distribution regions of another manifold distribution layer. In one embodiment, one or more of the open region, outlet channel, or multiple liquid distribution regions include flow restrictions. In one embodiment, the manifold assembly further includes an input manifold configured to split the fluid inlet flow into multiple fluid inlet flows, each fluid inlet flow associated with a low-power cooling section of the multiple microchannels; an intermediate manifold configured to collect and mix the two-phase output flow from each low-power cooling section, wherein mixing homogenizes localized heating variations and provides the mixed two-phase output flow to the high-power cooling sections of the multiple microchannels; and an output manifold configured to collect the two-phase output flow from the high-power cooling sections.

[0061] A heat transfer matrix is ​​described. In one embodiment, the heat transfer matrix includes a plurality of microchannels in fluid communication with a manifold for inputting flow into and outputting two-phase flow from the microchannels. Each of the microchannels includes a flow limit configured to suppress flow redistribution and balance pressure drops between certain microchannels. The pairs of microchannels are configured in a counter-current arrangement, with the inlet of one microchannel adjacent to the outlet of the next microchannel. This counter-current arrangement increases the packing density and reduces pressure drops within the heat transfer matrix.

[0062] In one embodiment, the flow restriction is a reduction in the cross-sectional area at each inlet of the plurality of microchannels. In another embodiment, each of the plurality of microchannels has an expanded microchannel geometry extending from the inlet configured with the flow restriction to the outlet. In another embodiment, the profile of the expanded microchannel geometry is based on at least one of flow rate, heat load, and microchannel filling efficiency. In another embodiment, the profile of the expanded microchannel geometry is one of linear, piecewise linear, and hyperbolic. In another embodiment, the heat transfer matrix includes one or more stackable port layers configured to provide fluid pathways from the manifold to the inlets and outlets of the plurality of microchannels. In another embodiment, the heat transfer matrix includes one or more heat transfer channel layers, wherein each heat transfer channel layer includes a plurality of microchannels. In another embodiment, the plurality of microchannels constitute a microchannel array of thermal pixels, wherein each thermal pixel is a pair of microchannels of the plurality of microchannels in a countercurrent configuration.

[0063] While this disclosure has been described in conjunction with certain embodiments, it should be understood that this disclosure is not limited to the disclosed embodiments. Rather, this disclosure is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims, which should be interpreted in the broadest possible sense to include all such modifications and equivalent structures permitted by law.

[0064] Those skilled in the art will understand that the various embodiments of this disclosure illustrated in the accompanying drawings constitute non-limiting examples, and additional components and features may be added to any of the embodiments discussed above without departing from the scope of this disclosure. Furthermore, those skilled in the art will understand that elements and features shown or described in one embodiment may be combined with elements and features of another embodiment to achieve any desired result without departing from the scope of this disclosure, and other features and advantages of the subject matter of this disclosure will be understood based on the provided description. Variations, combinations, and / or modifications to any embodiments and / or features of the embodiments described herein, within the capabilities of those skilled in the art, are also within the scope of this disclosure, as are alternative embodiments resulting from combining, integrating, and / or omitting features from any of the disclosed embodiments.

[0065] The use of the term “optional” for any element of a claim means that the element may or may not be included, and both alternatives are within the scope of the claim. Furthermore, the use of broader terms such as “comprising,” “including,” and “having” should be understood to support narrower terms such as “consisting of,” “substantially consisting of,” and “truly consisting of.” Therefore, the scope of protection is not limited by the foregoing description, but is defined by the following claims and includes all equivalents of the subject matter of the claims.

[0066] In the foregoing description, reference may be made to the spatial relationships between the various structures and their spatial orientations shown in the accompanying drawings. However, those skilled in the art will recognize upon a complete reading of this disclosure that the structures described herein can be positioned and oriented in any manner suitable for their intended purpose. Therefore, the use of terms such as “above,” “below,” “upper,” “lower,” “inner,” “outer,” “left,” “right,” “upward,” “downward,” “inward,” “outward,” “horizontal,” and “vertical” should be understood as describing the relative relationships between structures and / or their spatial orientations. Those skilled in the art will also recognize that the use of these terms may be provided within the context of the illustrations provided in the corresponding figures.

[0067] Furthermore, terms such as “approximately,” “roughly,” and “basically” should be understood to allow for variation in any numerical range or concept associated with them, including variations of approximately 25% (e.g., allowing for manufacturing tolerances and / or design deviations). For example, the term “roughly parallel” should be understood to refer to a configuration in which related components are oriented to define an angle between them equal to 180° ± 25% (e.g., an angle in the range of (approximately) 135° to (approximately) 225°). Therefore, the term “roughly parallel” should be understood to include a configuration in which related components are arranged in a parallel relationship.

[0068] Although terms such as “first,” “second,” and “third” may be used herein to describe various operations, elements, components, areas, and / or portions, these operations, elements, components, areas, and / or portions should not be limited by the use of these terms, as these terms are used to distinguish one operation, element, component, area, or portion from another. Therefore, unless expressly stated otherwise, a first operation, element, component, area, or portion may be referred to as a second operation, element, component, area, or portion without departing from the scope of this disclosure.

[0069] Each claim is incorporated into the specification as further disclosure and represents an embodiment of this disclosure. Furthermore, the phrases "at least one of A, B, and C" and "A and / or B and / or C" should be interpreted as including only A, only B, only C, or any combination of A, B, and C, respectively.

Claims

1. A two-phase heat exchanger, comprising: A manifold assembly configured to have an input manifold for fluid inlet flow and an output manifold for two-phase output flow; as well as The heat transfer matrix, which is coupled to the manifold assembly, The heat transfer matrix includes multiple microchannels that are in fluid communication with the input manifold and the output manifold. Each of the plurality of microchannels includes a flow limit configured to suppress the redistribution of fluid inlet flow and to balance pressure drops between certain microchannels. In this configuration, pairs of microchannels are arranged in a counter-current configuration, with the inlet of one microchannel adjacent to the outlet of the next microchannel. This counter-current configuration increases the fill density and reduces the pressure drop in the heat transfer matrix.

2. The two-phase heat exchanger according to claim 1, wherein, The flow restriction is a reduction in the cross-sectional area at each inlet of the plurality of microchannels.

3. The two-phase heat exchanger according to claim 1, wherein, Each of the plurality of microchannels has an expanded microchannel geometry extending from the inlet configured with the flow restriction to the outlet.

4. The two-phase heat exchanger according to claim 3, wherein, The profile of the expansion microchannel geometry is based on at least one of flow rate, thermal load, and microchannel filling efficiency.

5. The two-phase heat exchanger according to claim 3, wherein, The geometry of the expansion microchannel is one of linear, piecewise linear, and hyperbolic.

6. The two-phase heat exchanger according to claim 1, wherein, The heat transfer matrix also includes: One or more stackable port layers configured to provide fluid pathways from the manifold assembly to the inlet and outlet of the plurality of microchannels; and One or more heat transfer channel layers, wherein each heat transfer channel layer includes multiple microchannels.

7. The two-phase heat exchanger according to claim 1, wherein, The manifold assembly also includes: One or more manifold spacers, Each manifold spacer layer defines an open area for the fluid inlet flow to diffuse to the inlets of the plurality of microchannels, and Each manifold spacer layer defines an outlet channel for the two-phase outlet flow discharged via the outlets of the plurality of microchannels.

8. The two-phase heat exchanger according to claim 7, wherein, The open area allows the fluid inlet flow rate to diffuse laterally through multiple microchannels in a direction parallel to the heat transfer surface of the heat transfer matrix.

9. The two-phase heat exchanger according to claim 7, wherein, The manifold assembly also includes: One or more manifold distribution layers, wherein each manifold distribution layer includes multiple liquid distribution regions in a region adjacent to an open area within a manifold spacer layer to facilitate the distribution of the fluid inlet flow to all multiple microchannels.

10. The two-phase heat exchanger according to claim 9, wherein, When the manifold dispensing layers are stacked together in the manifold assembly, the liquid dispensing area of ​​one manifold dispensing layer is offset from the liquid dispensing area of ​​the other manifold dispensing layer.

11. The two-phase heat exchanger according to claim 9, wherein, One or more of the open area, the outlet channel, or the plurality of liquid distribution areas include flow restrictions.

12. The two-phase heat exchanger according to claim 1, wherein, The manifold assembly also includes: The input manifold is configured to divide the fluid inlet flow into multiple fluid inlet flows, each fluid inlet flow being associated with a low-power cooling section of the multiple microchannels; intermediate manifold, the intermediate manifold being configured as follows: The two-phase output flow from each low-power cooling section is collected and mixed, where mixing homogenizes localized heating variations; and A mixed two-phase output flow is provided to the high-power cooling section of the multiple microchannels; The output manifold is configured to collect two-phase output flow from the high-power cooling section.

13. A heat transfer matrix, comprising: Multiple microchannels are connected in fluid communication with a manifold, which is used to input flow into the microchannels and output two-phase flow from the microchannels. Each of the plurality of microchannels includes a flow limit, which is configured to suppress flow redistribution and balance voltage drops between certain microchannels. In this configuration, the microchannels in pairs are arranged in a counter-current configuration, with the inlet of one microchannel adjacent to the outlet of the next microchannel. This counter-current configuration increases the filling density and reduces the pressure drop in the heat transfer matrix.

14. The heat transfer matrix according to claim 13, wherein, The flow restriction is a reduction in the cross-sectional area at each inlet of the plurality of microchannels.

15. The heat transfer matrix according to claim 14, wherein, Each of the plurality of microchannels has an expanded microchannel geometry extending from an inlet configured with the flow restriction to an outlet.

16. The heat transfer matrix according to claim 15, wherein, The profile of the expansion microchannel geometry is based on at least one of flow rate, thermal load, and microchannel filling efficiency.

17. The heat transfer matrix according to claim 15, wherein, The geometry of the expansion microchannel is one of linear, piecewise linear, and hyperbolic.

18. The heat transfer matrix according to claim 13, further comprising: One or more stackable port layers are configured to provide fluid pathways from the manifold to the inlets and outlets of the plurality of microchannels.

19. The heat transfer matrix according to claim 13, further comprising: One or more heat transfer channel layers, wherein each heat transfer channel layer includes multiple microchannels.

20. The heat transfer matrix according to claim 13, wherein, The plurality of microchannels constitute a microchannel array of hot pixels, wherein each hot pixel is a pair of microchannels of the plurality of microchannels in the reverse flow configuration.

Images