Cooling Devices and Systems

Abstract

An embodiment of the present application relates to an apparatus that includes means for obtaining a plurality of input values ​​including physical parameters of at least a portion of a cooling system; using the obtained input values ​​to estimate a dimensionless ratio that provides an indication of at least one of heat transfer of at least a portion of the cooling system and pressure loss of at least a portion of the cooling system; using the dimensionless ratio to determine at least one performance indicator of at least a portion of the cooling system; and using the at least one determined performance indicator to control one or more physical parameters of at least a portion of the cooling system.

Description

[Technical field]

[0001] TECHNICAL FIELD Embodiments of the present application relate to devices and systems for cooling. Some embodiments relate to devices and systems for controlling a two-phase cooling system. [Background technology]

[0002] Cooling systems are used to enable the functioning of devices that generate unwanted heat during use, such as electronic equipment. Increasing the efficiency of these cooling systems allows for more heat to be transferred, improving the functionality of the device and / or reducing its size. Summary of the Invention

[0003] According to various (but not necessarily all) embodiments of the present application, there is provided an apparatus including means for obtaining a plurality of input values ​​including physical parameters of at least a portion of a cooling system, estimating a dimensionless ratio using said obtained input values ​​providing an indication of at least one of heat transfer of at least a portion of the cooling system and pressure loss of at least a portion of the cooling system, determining at least one performance indicator of at least a portion of the cooling system using said dimensionless ratio, and controlling one or more physical parameters of at least a portion of the cooling system using the at least one determined performance indicator.

[0004] The cooling system may include a single phase or a two phase cooling system.

[0005] The physical parameters may be controlled for the entire cooling system or for parts of the cooling system.

[0006] The measure of heat transfer within the cooling system may include a Nusselt number, and the measure of pressure loss within the cooling system may include one or more of a friction pressure gradient, a fanning friction coefficient, and a pressure drop.

[0007] The dimensionless ratio may be estimated using machine learning.

[0008] The dimensionless ratio may be estimated using a non-linear regression algorithm.

[0009] At least one of the determined performance indicators may be used to control one or more physical parameters of the cooling system during use.

[0010] At least one of the determined performance indicators may be used to control a flow path within a cooling system.

[0011] At least one of the determined performance indicators may be used to control a liquid level in a downcomer of a passive two-phase cooling system.

[0012] At least one of the determined performance indicators may be used to control a pump in an active two-phase cooling system.

[0013] The plurality of inputs may include information obtained from one or more sensors within the cooling system.

[0014] At least one of the determined performance indicators may be used to modify one or more physical parameters during design of at least a portion of the cooling system.

[0015] The physical parameters of the cooling system may include one or more of parameters related to the geometry of the cooling system, the working fluid in the cooling system, the thermodynamic conditions in the cooling system, and the flow conditions in the cooling system.

[0016] The dimensionless ratio used may be estimated from physical measurements of the cooling system.

[0017] The dimensionless ratio used may be selected based on at least one of the geometry of the cooling system and the operating conditions of the cooling system.

[0018] The one or more determined performance indicators may be determined by solving for conservation of one or more of mass, momentum, and energy using the estimated dimensionless ratio.

[0019] According to various (but not necessarily all) embodiments of the present application, there is provided a method including obtaining a plurality of input values ​​including physical parameters of at least a portion of a cooling system, estimating a dimensionless ratio using the obtained input values ​​providing an indication of at least one of heat transfer of the cooling system and pressure loss of the cooling system, determining at least one performance indicator of at least a portion of the cooling system using the dimensionless ratio, and controlling one or more physical parameters of at least a portion of the cooling system using the at least one determined performance indicator.

[0020] According to various (but not necessarily all) embodiments of the present application, there is provided a computer program comprising computer program instructions that, when executed by a processing circuit, perform the following: obtaining a plurality of input values ​​including physical parameters of at least a portion of a cooling system; estimating, using the obtained input values, a dimensionless ratio providing an indication of at least one of heat transfer of at least a portion of the cooling system and pressure loss of at least a portion of the cooling system; determining, using the dimensionless ratio, at least one performance indicator of at least a portion of the cooling system; and controlling, using the at least one determined performance indicator, one or more physical parameters of at least a portion of the cooling system.

[0021] Some embodiments will now be described with reference to the accompanying drawings. [Brief description of the drawings]

[0022] [Figure 1]FIG. 1 is a schematic diagram illustrating a two-phase cooling system according to an embodiment. [Diagram 2] FIG. 2 is a schematic diagram illustrating a two-phase cooling system according to another embodiment. [Diagram 3] FIG. 2 is a schematic diagram showing a method according to an embodiment. [Figure 4] FIG. 11 is a schematic diagram showing a method according to another embodiment. [Diagram 5] FIG. 11 is a schematic diagram showing a method according to another embodiment. [Figure 6] FIG. 11 is a schematic diagram showing a method according to another embodiment. [Figure 7] FIG. 2 is a schematic diagram illustrating a two-phase cooling system according to another embodiment. [Figure 8A] FIG. 11 is a schematic diagram showing a method according to another embodiment. [Figure 8B] FIG. 11 is a schematic diagram showing a method according to another embodiment. [Figure 9A] FIG. 2 is a schematic diagram showing the results obtained in the examples of the present application. [Figure 9B] FIG. 2 is a schematic diagram showing the results obtained in the examples of the present application. [Figure 10A] FIG. 2 is a schematic diagram showing the results obtained in the examples of the present application. [Figure 10B] FIG. 2 is a schematic diagram showing the results obtained in the examples of the present application. [Figure 11] FIG. 1 is a schematic diagram showing an apparatus according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] The present application relates to an apparatus and method for controlling a cooling system or for controlling the design of a cooling system. The present application uses methods such as machine learning to control one or more physical parameters of a cooling system, such as a two-phase or single-phase cooling system. The physical parameters may be parameters that are controllable during use of the cooling system or a portion of the cooling system to allow management of the cooling system. In some embodiments, the parameters can be controlled during the design of the cooling system to allow design and manufacture of more efficient cooling systems.

[0024] FIG 1 is a schematic diagram of a two-phase cooling system 101. In the embodiment of FIG 1, the two-phase cooling system 101 includes a passive two-phase cooling system that is driven by gravity rather than a pump. In the embodiment of FIG 1, the passive two-phase cooling system 101 includes a thermosiphon loop. Other types of two-phase cooling systems can be used in other embodiments of the present application.

[0025] The thermosiphon loop shown in Figure 1 includes an evaporator 103, a condenser 105, a downcomer 107 and a riser 109. Provided within the thermosiphon loop is a working fluid 113. When the thermosiphon loop is in use, the working fluid 113 circulates through the components of the thermosiphon loop.

[0026] The evaporator 103 is located at the bottom of the thermosiphon loop such that the working fluid flows by gravity through the downcomer 107 into the evaporator 103 as shown by arrow 115. The height and inner diameter of the downcomer 107 can be selected such that the static end of the fluid in the downcomer 107 causes the fluid to flow through the evaporator 103, the riser 109 and the condenser 105. When in the downcomer 107, the working fluid 113 is in a liquid phase 117.

[0027] The evaporator 103 includes any means for transferring heat from a heat source 111 to the working fluid 113. The evaporator 103 is thermally coupled to the heat source 111. A thermal interface material may be used to enable the evaporator 103 to be thermally coupled to the heat source 111. The heat source 111 may include electronic equipment that generates unwanted heat during use. The electronic equipment may be servers, routers, network switches, storage equipment, or any other suitable type of equipment. In some examples, these heat sources may include a data center, a communications equipment room, or multiple electronic equipment that may provide a network, a communications room, a computer room, a network room, or any other suitable arrangement.

[0028] Heat is transferred from the heat source 111 to the working fluid 113 in the evaporator 103, as indicated by arrows 119. This heat transfer causes evaporation of a portion of the working fluid 113 in the evaporator 103, converting the working fluid 113 from a liquid phase 117 to a mixture of liquid and vapor phases. In particular, the evaporator 103 converts a portion of the working fluid 113 to a vapor phase 121 while allowing a portion to remain in the liquid phase 117, resulting in a two-phase mixture exiting the outlet of the evaporator 103. The two-phase mixture may include droplets of vapor entrained within a liquid or other flow state, depending on the thermosiphon loop design, heat load, fill factor, and other suitable parameters.

[0029] The vapor degree of the working fluid 113 in the two-phase cooling system is the amount of vapor generated during the phase change process and is calculated as the ratio of the mass of vapor divided by the total mass of fluid (liquid and vapor). The vapor degree of the fluid discharged from the outlet of the evaporator 103 is important for the functioning of the two-phase cooling device 101. If the vapor degree of the fluid discharged from the outlet of the evaporator 103 is too high, the two-phase cooling system 101 will operate in a friction center state with reduced thermal performance. Therefore, it is beneficial to ensure that the vapor degree is maintained below a threshold value to maintain sufficient thermal performance. The threshold value may be about 0.6 or any other suitable value.

[0030] The evaporator 103 is connected to a riser 109 such that working fluid discharged from the evaporator 103 flows from the evaporator 103 into the riser 109. The working fluid comprises a two-phase mixture in which the vapor phase 121 is less dense than the liquid phase 117. As shown by arrow 123, the working fluid 113 in the thermosiphon loop rises through the riser 109 as a result of a balance between buoyancy (associated with the two-phase mixture exiting the evaporator 103) and gravity (associated with the working fluid 113 exiting the condenser 105).

[0031] The evaporator 103 may include structures that enable efficient heat transfer from the evaporator 103 to the working fluid 113. For example, the evaporator 103 may include a wick structure, microchannels, an array of fins, a serpentine arrangement of macro / microchannels, or any suitable combination of these features.

[0032] The condenser 105 is located at the top of the thermosiphon loop. The condenser 105 is disposed above the evaporator 103 such that the working fluid 113 flows upward from the evaporator 103 to the condenser 105.

[0033] The condenser 105 is coupled to a riser 109 such that the working fluid 113 in the two-phase mixture (vapor phase 121 and liquid phase 117) flows from the riser 109 into the condenser 105. The condenser 105 may include any means for cooling the working fluid 113. For example, the condenser 105 may be air-cooled or liquid-cooled. A liquid-cooled condenser 105 may include a tube-in-tube configuration, a shell-and-tube configuration, a plate heat exchanger, or any other suitable configuration or arrangement. An air-cooled condenser 105 may include a louvered fin flat tube configuration, a tube-and-fin heat exchanger, or any other suitable type of configuration or arrangement.

[0034] The condenser 105 is thermally coupled to a coolant 125. A thermal interface material may be used to enable the condenser 105 to be thermally coupled to the coolant 125. In other embodiments, the coolant 125 may be directly integrated into the condenser 105 with a wall interface separating the flow of working fluid 113 from the flow of the coolant 125. The wall interface may include a highly conductive metal or metal alloy, such as copper, aluminum, brass, or any other suitable metal. In some embodiments, the wall interface may include a highly conductive ceramic, such as aluminum nitride (AlN), or a polymer, such as a filled polymer composite. The condenser 105 may be a liquid-cooled condenser or any other suitable type of condenser. The condenser 105 may include any suitable shape that allows for efficient removal of heat from the working fluid.

[0035] The condenser 105 allows for heat transfer from the working fluid 113 to the coolant 125, as indicated by arrows 127. This heat transfer causes the working fluid 113 to at least partially condense and return to the liquid phase 117. The working fluid 113 at the outlet of the condenser 105 is in the liquid phase 117.

[0036] The condenser 105 is coupled to the downcomer 107 such that the working fluid 113 in the liquid phase 117 flows by gravity down the downcomer 107 and back to the inlet of the evaporator 103 .

[0037] 2 illustrates another embodiment of a two-phase cooling system 101 that includes one or more parameters that may be controlled in an embodiment of the present application. In this embodiment, the two-phase cooling system 101 includes another thermosiphon loop that may be configured to enable cooling of multiple electronic devices 203 in a data center or any other suitable environment. Other two-phase cooling systems 101 that may be used to cool other types of devices and equipment may be used in other embodiments of the present application.

[0038] The thermosiphon loop includes a number of sensors 223 distributed throughout the thermosiphon loop to allow a number of inputs to be obtained, including physical parameters of the thermosiphon loop, which can then be used to estimate dimensionless ratios, such as the Nusselt number and the friction pressure gradient, along with additional parameters, such as the geometry of the two-phase cooling system 101 and the heat load from a heat source.

[0039] In the embodiment of FIG. 2, the thermosiphon loop includes a plurality of evaporators 103, each of which is thermally coupled to an electronic device 203. The electronic device 203 may be a server, a router, a switch, or other optoelectronic device. The evaporator 103 is coupled to the electronic device 203 to allow heat from the electronic device 203 to evaporate the working fluid 113 in the evaporator 103. Although six electronic devices 203 and corresponding evaporators 103 are shown in the thermosiphon loop of FIG. 2, it should be understood that other numbers of electronic devices 203 and evaporators 103 may be used in other embodiments of the present application.

[0040] 2, the evaporators 103 are arranged in series in the thermosiphon loop such that the working fluid 113 discharged from a first evaporator 103 is fed to the inlet of the next evaporator 103 in the series. In this series arrangement, the working fluid 113 passes through each of the evaporators 103 in sequence. As the working fluid 113 passes through the evaporators 103, the amount of the working fluid 113 in the vapor phase 121 increases because more heat is absorbed in the evaporation process.

[0041] The example thermosiphon loop of FIG. 2 includes an accumulator 205 in the downcomer 107. The accumulator 205 may be a liquid accumulator or receiver. The accumulator 205 may include any means capable of storing the working fluid 113 in the liquid phase 117. The accumulator 205 stores the working fluid 113 in the liquid phase 117. The working fluid 113 inside the accumulator 205 may include a portion of liquid, and the remaining volume is a saturated two-phase mixture. The accumulator 205 is advantageous in preventing the working fluid 113 in the liquid phase 117 from entering the condenser 105 due to the large cross-sectional area of ​​the accumulator 205 relative to the downcomer 107. This avoids overflow of the condenser 105 due to a decrease in the thermal performance of the thermosiphon loop.

[0042] In this embodiment, the condenser 105 includes a compact heat exchanger. Other types of condensers 105 may be used in other embodiments of the present application. The type of condenser 105 used may be selected based on the expected amount of heat to be removed from the thermosiphon loop. This may be based on the number of electronic devices 203 to be cooled by the thermosiphon loop. In some embodiments, if the two-phase cooling system 101 is used to cool a rack of servers or other similar arrangement, the condenser 105 may be located directly above the rack (overhead condenser 105) to provide sufficient driving force for a passive gravity-driven two-phase cooling system.

[0043] In the embodiment shown in FIG. 2, a secondary cooling system 201 is coupled to the condenser 105 to allow heat to be transferred from the thermosiphon loop and the data center environment. In this embodiment, water is the coolant and is provided at an inlet 207, as shown by arrow 211. The condenser 105 is configured such that heat is then transferred from the working fluid 113 of the thermosiphon loop to the water of the secondary cooling system 201. The heated water is discharged through an outlet 213, as shown by arrow 215. In other embodiments, the secondary cooling system 201 may use a different working fluid. For example, this may be a boiling dielectric fluid in a heat transfer process, and the heat exchanger will operate in a condensation-evaporation mode.

[0044] 2, the condenser 105 has a counter-flow arrangement, with inlets for the working fluid 113 and the water located on opposite sides of the condenser 105. In other embodiments, the condenser 105 may be configured to have a parallel flow arrangement, a cross-flow arrangement, or any other suitable type of arrangement.

[0045] 2, only a portion of the secondary cooling system 201 is shown. The secondary cooling system 201 may be an active cooling system driven by a pump, or any other suitable type of cooling system.

[0046] The thermosiphon loop further includes a number of sensors 223. The sensors 223 are disposed within the thermosiphon loop and configured to enable measurement of one or more parameters of the thermosiphon loop. The monitored parameters may relate to the efficiency of the thermosiphon loop and the amount of heat transferred by the thermosiphon loop. The measurements obtained by the sensors 223 may be used as inputs to a control system that may process inputs to determine at least one performance indicator of the thermosiphon loop. The determined performance indicator may then be used to control one or more parameters of the thermosiphon loop. Figures 3-6 show examples of methods in which measurements from the sensors 223 may be used as inputs to determine at least one performance indicator of the thermosiphon loop or other two-phase cooling system 101.

[0047] The sensor 223 may be used to measure any suitable parameter that is used to provide suitable input values ​​used to determine the performance index. For example, the sensor 223 may include a temperature sensor, a pressure sensor, a mass flow sensor, or any other suitable type of sensor 223. The sensor 223 may be positioned at a number of different locations within the thermosiphon loop such that various parameters can be measured at different locations within the thermosiphon loop.

[0048] 1 and 2, the two-phase cooling system 101 includes a passive thermosiphon loop. It should be understood that in other embodiments of the present application, other types of two-phase cooling systems 101 may be used. For example, in some embodiments, the two-phase cooling system 101 may include an active two-phase cooling system 101 that includes one or more pumps configured to pump the working fluid through the system.

[0049] Figure 3 shows an embodiment of a method that can be used to facilitate control of a cooling system, such as the two-phase cooling system 101 shown in Figures 1 and 2. It should be understood that in some embodiments of the present application, variations of this method may be used. For example, block 303 may be omitted and the input values ​​obtained in block 301 may be used directly as inputs for determining at least one performance index of the cooling system (block 305).

[0050] The method includes obtaining a plurality of input values ​​including physical parameters of a cooling system at block 301. The input values ​​may include any values ​​that provide measurements of one or more physical parameters of the cooling system.

[0051] In some embodiments, the plurality of inputs may include measurements obtained by one or more of the sensors 223 in the cooling system. In such embodiments, the obtained inputs may include information obtained from one or more sensors in the cooling system. For example, the sensors 223 shown in FIG. 2 may be used to obtain measurements of temperature, pressure, mass flow, or any other suitable parameter during use of the cooling system.

[0052] In some embodiments, the plurality of input values ​​may include known values ​​of physical parameters that do not change during use of the cooling system. For example, parameters related to the geometry of the cooling system, such as flow path widths and downcomer heights, may be fixed so as not to change during use. In such embodiments, these parameters do not need to be measured, but may be stored and retrieved as needed.

[0053] In some embodiments, the plurality of input values ​​may include values ​​relating to a thermal load distribution, the thermal load distribution being measured during use of the cooling system.

[0054] In some embodiments, the plurality of input values ​​may include proposed values, for example if the method is used to enable the design of a cooling system, some initial input parameters for a proposed geometry of the cooling system may be used, which may then be adjusted as part of the design process.

[0055] In block 303, the method includes estimating a dimensionless ratio using the obtained input values. The dimensionless ratio may provide a performance indicator for the cooling system.

[0056] In some embodiments, the dimensionless ratio can be derived from physical measurements of the cooling system. For example, the dimensionless ratio can be derived from inputs obtained from one or more sensors in the cooling system. In some embodiments, the dimensionless ratio can be derived from stored values ​​associated with the cooling system. In some embodiments, the dimensionless ratio can be derived from a combination of physical measurements and stored values.

[0057] In some embodiments, the dimensionless ratio may provide an indication of heat transfer within the cooling system. In such embodiments, the dimensionless ratio may include a Nusselt number, or a value related to or derived from a Nusselt number. The Nusselt number represents the ratio of convective heat transfer through a boundary to conductive heat transfer. A Nusselt number of 1 represents heat transfer by pure conduction. A higher Nusselt number represents more effective convective heat transfer. Thus, the Nusselt number represents the enhancement of heat transfer through the fluid layer by convection relative to conduction through the fluid layer.

[0058] The Nusselt number Nu for a pipe with a non-circular cross-sectional area is calculated by the following formula:

number

[0059] where h represents the thermal conductivity and dh represents the hydraulic diameter, and k fluid represents the fluid thermal conductivity, and q eff represents the effective heat flux, and T wall represents the wall temperature, and T fluid represents the fluid temperature. For other systems different from piping, the Nusselt number can be obtained by multiplying the heat transfer coefficient by the characteristic length of the given system.

[0060] It should also be understood that the temperature difference from the wall to the fluid represents the evaporation process, and the temperature difference from the fluid to the wall is used to calculate the heat transfer coefficient for the condensation process.

[0061] In some embodiments, the dimensionless ratio may provide an indication of pressure loss within the two-phase cooling system 101. This may provide an indication of pressure change as the working fluid evaporates. In some embodiments, the pressure loss may be due to interactions between a working fluid and components of the two-phase cooling system 101. In such embodiments, the dimensionless ratio may be derived from a frictional pressure gradient, which is the frictional pressure loss per unit length.

[0062] The pressure loss gradient is calculated by the following formula:

number

[0063] Here, Δp fri, represents the friction pressure loss, and L port represents the total length, f represents the fanning friction coefficient, G represents the mass flux, and d h represents the hydraulic diameter, and ρ fluid represents the fluid density. For other systems different from piping, the hydraulic diameter can be replaced by the characteristic length of the given system.

[0064] In other embodiments of the present application, other dimensionless ratios providing an indication of the pressure loss may be used, for example, fanning friction factor, pressure loss or other ratios. For two-phase cooling systems, the total pressure loss is the sum of static, momentum and friction contributions. Other parameters and dimensionless ratios may be generated to characterize the pressure loss contributions.

[0065] It should be understood that in other embodiments of the present application, other dimensionless ratios may be used, for example, the dimensionless ratios may include the Bond number, the Weber number, the boiling number, the Reynolds number, or any other suitable ratios or values ​​related to these ratios.

[0066] The dimensionless ratio can be estimated using machine learning. In some embodiments, the dimensionless ratio can be estimated using a non-linear regression algorithm. In other embodiments of the present application, other types of algorithms can be used, such as support vector machines (SVM), stochastic gradient descent (SGD), nearest neighbor method (NNR), Gaussian process regression (GPR), Gaussian naive Bayes, decision tree analysis random forest or neural network.

[0067] In some embodiments, the machine learning algorithm can be configured to select which dimensionless ratios can be used to provide an optimal or substantially optimal output. If there are multiple different dimensionless ratios available, there will also be multiple different combinations of the dimensionless ratios that can be used to implement the method. The machine learning algorithm may select the type and / or number of dimensionless ratios to be used. The type and / or number of dimensionless ratios to be used can be selected based on factors such as the geometry of the cooling system, the size operating conditions of the cooling system, the size of the cooling system, the type of the cooling system, or any other suitable factors.

[0068] At block 305, the method includes determining at least one performance index of the cooling system using the dimensionless ratio by solving for conservation of one or more mass, momentum, or energy, or any other suitable parameters of the cooling system using the estimated dimensionless ratio.

[0069] In some embodiments, the performance index may be determined by solving one or more conservation equations using machine learning, or any other suitable process. The equations may be solved for a portion of the cooling system or for the entire cooling system. In other embodiments, instead of solving the equations, estimated dimensionless ratios may be directly related to output values ​​using machine learning or any other suitable process. This direct relationship may reduce the amount of computation. In such embodiments, the input values ​​obtained in block 301 may be used to determine the performance index without directly calculating any dimensionless ratios (block 305).

[0070] In some embodiments, numerical methods can be used to determine the temperature at a finite number of points in the cooling system over time. In other embodiments, elements of the cooling system can be treated as a single entity for temperature calculations. For example, a plate heat exchanger can be treated as a single entity rather than a series of finite points.

[0071] At block 307, the method includes controlling one or more physical parameters of the cooling system using the at least one determined performance index.

[0072] In some embodiments, at least one determined performance index can be used to control one or more physical parameters of the cooling system during use. For example, the determined performance index may be used to control the flow paths within the two-phase cooling system 101 or to control the liquid level in a downcomer of a passive two-phase cooling system 101. This may be controlled by controlling one or more valves or other suitable means. In embodiments where the cooling system includes an active two-phase cooling system 101, the determined performance index may be used to control a pump. This can provide fast and accurate control of the cooling system.

[0073] In some embodiments, the physical parameters controlled during use of the cooling system can include flow circulation of a working fluid through the cooling system, or flow circulation of a working fluid through one or more components of the cooling system. In some embodiments, the physical parameters can be controlled to optimize flow circulation of a working fluid. The method used to control flow rate can be determined by the type of cooling system used. For example, if the cooling system is a passive two-phase cooling system 101, the flow rate circulation can be controlled by controlling one or more valves in the two-phase cooling system 101. If the cooling system is an active system, the flow rate circulation may be controlled by controlling pump speed, voltage, or any other suitable factor.

[0074] Control of the working fluid can be applied to different parts of the cooling system, or to the entire cooling system. For example, the thermosiphon loop shown in FIG. 2 can be coupled to a secondary cooling system 201. Such a cooling system can be used to cool data centers or other types of electronic equipment. In embodiments of the present application, circulation of the working fluid can be controlled in the thermosiphon loop, the secondary cooling system 201, or both.

[0075] In some embodiments, at least one determined performance index can be used to modify one or more physical parameters during design of the cooling system. For example, a simulation of the cooling system may be generated and physical parameters of the simulated system may be adjusted based on the determined performance index. The physical parameters of the cooling system may include one or more of parameters related to the geometry of the cooling system, the working fluid in the cooling system, the thermodynamic conditions in the cooling system, the flow conditions in the cooling system, and any other relevant parameters.

[0076] In some embodiments, the determined performance index can be used to modify features within the cooling system. For example, the cooling system can include wick structures and / or microchannels that allow for the flow of working fluid through the cooling system. Parameters related to these structures (such as diameter or length) can be adjusted during the design process. In some embodiments, the heat exchanger can include fins or other structures to improve heat transfer, and these structures can also be adjusted during the design process based on the determined performance index.

[0077] It should be understood that the method illustrated in Figure 3 can be used to control portions of the cooling system or the entire cooling system. For example, in some embodiments, the cooling system may include multiple cooling loops and the control may be applied to each individually or to the entire system.

[0078] 4 illustrates another method embodiment according to an embodiment of the present application. The method embodiment can be used in single-phase (vapor and liquid) cooling systems as well as active cooling systems including mechanical drivers and passive two-phase cooling systems 101 such as the system shown in FIG. 1. The method may also be used in parts or components within the cooling system. For example, in some embodiments, the method may be used to design an evaporator 103 or a condenser 105.

[0079] The method provides system geometry inputs 401 and physical condition inputs 403. The system geometry inputs may provide a complete system geometry for the cooling system, or may provide a partial system geometry. The complete system geometry inputs 401 can be used in embodiments where the cooling system is in use, such that the entire geometry of the system is known. The partial system geometry inputs may be used in embodiments where the cooling system is being designed.

[0080] The system geometry values ​​401 may relate to dimensions such as height and width of one or more components of the cooling system. The system geometry values ​​401 may be fixed values ​​that do not change during use of the cooling system. In some embodiments, the system geometry values ​​401 may include one or more values ​​that can be altered during use. For example, one or more valves may be used to change the flow paths within the cooling system to change the effective length and / or diameter of a portion of the cooling system. The system geometry values ​​401 may be known values ​​obtained from a database or other register, or may be measured as needed.

[0081] The physical condition inputs 403 may include inputs related to one or more fluid properties or thermodynamic flow conditions within the cooling system. In embodiments where the cooling system is controlled during use, at least some of the physical condition inputs 403 may be obtained from one or more sensors 223 within the cooling system. In embodiments where simulation is used to design the cooling system, at least some of the physical condition inputs 403 may be estimated from experimentation or using any other suitable process.

[0082] The physical condition inputs 403 may relate to the working fluid, for example to the temperature and / or pressure within the working fluid, or may relate to other parameters of the working fluid, such as heat capacity or the type of working fluid used.

[0083] In some embodiments, the physical condition inputs 403 may include inputs related to one or more thermodynamic conditions within the cooling system. These conditions may be temperature and / or pressure at one or more locations within the cooling system.

[0084] In some embodiments, the physical condition inputs 403 may include flow conditions, such as mass flow rate, or any other suitable parameters related to the flow of the working fluid within the cooling system. Such values ​​may be measured by one or more sensors 223 or estimated from experimental data.

[0085] In the embodiment shown in FIG. 4, the system further obtains one or more dimensionless ratios 405. In this embodiment, the dimensionless ratios 405 include the Nusselt number and the friction pressure gradient. It should be understood that in other embodiments of the present application, other dimensionless ratios 405 may be used. The dimensionless ratios 405 may be estimated using machine learning. Machine learning, or any other suitable process, may be used to select which of the available dimensionless ratios 405 are used to calculate the heat load and pressure drop, which may then be used to determine the performance index.

[0086] In block 407, the equations governing the conservation of mass, momentum and energy are solved. These equations can be solved for localized regions of the cooling system. For example, the cooling system can be divided into portions and the equations can be solved for specific portions of the cooling system.

[0087] As noted above, in some embodiments, the equations may be solved so that the cooling system performance index is calculated. In other embodiments, machine learning or other suitable processes may be used to directly convert the input values ​​401 and the input values ​​403 into an output performance index without actually solving the equations.

[0088] Once the equations are solved, one or more performance indicators 409, 411 are provided as output. In this embodiment, a heat exchange performance indicator 409 and a pressure drop performance indicator 411 are provided. The heat exchange performance indicator 409 may provide an indication of the total heat exchanged by the cooling system or by the particular portion of the cooling system being analyzed. The pressure drop performance indicator 411 may provide an indication of the total pressure drop through the entire cooling system or the particular portion of the cooling system being analyzed. It should be understood that in other embodiments of the present application, other performance indicators may be used.

[0089] Once the performance indicators 409, 411 have been obtained, they can be used to control one or more physical parameters of the cooling system. For example, during use of the cooling system, the flow path of the working fluid may be controlled by opening or closing a valve. This can control the pressure within the cooling system, which in turn can control the amount of heat transferred to the working fluid. For active cooling systems, the flow of the working fluid may be controlled by controlling a pump speed, voltage, or other suitable parameter.

[0090] In an embodiment where the method is used to design a cooling system, the performance index may be used to adjust one or more system geometry values ​​401. For example, it may be used to adjust a physical parameter such as a length or width of a portion of the cooling system. The adjusted system geometry value 401 may then be used and the process may be repeated until a sufficient output performance index is achieved. The sufficient output performance index may be an optimal value, or a substantially optimal value, for a heat exchange performance index 409 and / or a pressure drop performance index 411, or a obtained value for the heat exchange performance index 409 and / or the pressure drop performance index 411 that exceeds a threshold value.

[0091] FIG. 5 shows an example of a method that can be used to estimate the dimensionless ratio.

[0092] In block 501, a number of input values ​​are obtained, including physical parameters of the cooling system. The input values ​​may be obtained by making measurements within the cooling system, or from experimental data, or from other suitable sources.

[0093] At block 503, the input values ​​are converted to dimensionless parameters. Any suitable dimension reduction process can be used to convert the input values ​​to dimensionless ratios. The dimensionless ratios may include the Reynolds number, Bond number, density ratio, or any other suitable dimensionless parameter.

[0094] At block 505, machine learning is performed on the dimensionless ratios to obtain estimates for the particular dimensionless ratio. Converting the input values ​​to dimensionless parameters reduces the dimensionality of the feature vectors of the machine learning algorithm, thereby reducing the complexity of the machine learning algorithm.

[0095] 5, the machine learning algorithm used includes random forest regression. Other types of machine learning algorithms can be used in other embodiments of the present application.

[0096] In block 507, dimensionless ratios are provided that provide a measure of heat transfer in a cooling system and a measure of pressure loss in said cooling system. These dimensionless ratios may be the output of the machine learning algorithm used in block 505.

[0097] 5, the dimensionless ratios provided include the Nusselt number and the friction pressure gradient. It should be understood that in other embodiments of the present application, other dimensionless ratios may be used.

[0098] The dimensionless ratios obtained by the method shown in FIG. 5 can then be used as inputs to solve the equations for the conservation of mass, momentum and energy in the cooling system, as shown in FIG.

[0099] FIG. 6 shows examples of methods and formulas that can be used in some embodiments of the present application.

[0100] At block 601, a number of inputs are provided, including physical parameters of the cooling system.

[0101] These input values ​​may include one or more values ​​related to system geometry. These values ​​may relate to the geometry of one or more parts or components within the cooling system. These may be measurements or known values ​​of an existing cooling system, or may be selected values ​​for the cooling system being simulated.

[0102] These input values ​​may also include values ​​related to the system requirements. In some embodiments, the values ​​related to the system requirements may be related to the working fluid used. In some embodiments, the values ​​related to the system requirements may be related to thermodynamic and flow conditions, for example they may be related to the temperature in the cooling system or the flow rate of the working fluid or any other suitable parameter.

[0103] In block 603, the input values ​​are used to estimate dimensionless ratios that provide an indication of heat transfer in the cooling system and pressure loss in the cooling system. In the example of FIG. 6, the dimensionless ratios include the Nusselt number and the friction pressure gradient. These dimensionless ratios can be calculated by solving the evaporating flow and condensing flow equations using the input values ​​obtained in block 601. Machine learning can be used to obtain the Nusselt number and the Fanning friction factor, which can then be used to solve the following equations:

[0104] For evaporating flow, the Nusselt number can be given by:

number

[0105] For condensing flow, the Nusselt number can be given by:

number

[0106] The pressure gradient of the evaporating and condensing flows can be given by:

number

[0107] In block 605, the output obtained in block 603 is used to solve the equations governing the conservation of mass, momentum and energy. The dimensionless ratios can further be used to solve the equations governing heat transfer.

[0108] The equation governing the conservation of mass is given by:

number

[0109] The equation governing conservation of momentum is given by:

number

[0110] The equation governing the conservation of energy is given below:

number

[0111] The equation governing heat conduction is given by:

number

[0112] Here, u represents the velocity, θ represents the angle defined with respect to the horizontal, H represents the enthalpy, and α represents the thermal diffusivity.

[0113] The fluid flow equations provided herein are one-dimensional. In these equations, z is the dimension along the flow direction. It should be understood that equations covering three dimensions may be used in the examples of this application.

[0114] The equations governing heat transfer provided here are in three dimensions, in other embodiments they may be solved in two or one dimensions, or at steady state.

[0115] The equations shown in block 605 can be solved in a steady state or transient mode. In some embodiments, additional equations may be used that take into account flow maldistribution, flow instabilities, cold start, unbalanced heat load conditions, and other physical conditions that can affect a cooling system.

[0116] At block 607, equation outputs are obtained. These outputs include performance metrics related to physical parameters of the cooling system. In the example of FIG. 6, these outputs include total heat exchanged and total pressure drop across the cooling system. These metrics can be obtained for parts or components within the cooling system, or for the entire cooling system.

[0117] At block 609, an overall performance index is determined. In this embodiment, the overall performance index is given by the ratio of heat exchanged to total pressure drop. Other performance indexes may be used in other embodiments of the present application.

[0118] The overall performance index can then be used for the control of one or more physical parameters of the cooling system. For example, if the cooling system is in use, the performance index can be used to control the flow of working fluid through the cooling system. If the cooling system is a simulated system analyzed for design purposes, the cooling system can then be controlled by adjusting any of the input values, such as values ​​related to system geometry. This process can be repeated until the simulated cooling system provides a sufficient overall performance index. The sufficient overall performance index can be a value of the ratio of heat exchanged to total pressure drop that is within a predefined threshold range.

[0119] 7 shows another embodiment of a two-phase cooling system 101 that can be used in some embodiments of the present application. The evaporator 103 and electronics 203 are not shown in FIG. 7 for clarity. Multiple evaporators 103 may be provided in series or parallel or a combination of both. As shown in FIG. 2, the two-phase cooling system 101 may further include one or more sensors 223 for measuring physical parameters of the two-phase cooling system 101.

[0120] 7, the downcomer 107 includes multiple branches 701A, 701B. Each branch 701A, 701B provides a line or conduit between a condenser 105 and one or more evaporators 103. Each branch 701A, 701B is configured to provide a flow path for working fluid 113 in a liquid phase 117.

[0121] In the embodiment shown in Figure 7, two branches 701A, 701B are provided, where the first branch has a larger diameter than the second branch. It should be understood that other numbers of branches 701A, 701B having different diameters may be provided in other embodiments of the present application.

[0122] The two-phase cooling system 101 further includes a plurality of valves 703A, 703B, 703C, 703D configured to control the flow of the working fluid 113 within the two-phase cooling system 101. The plurality of valves 703A, 703B, 703C, 703D may be any suitable type of valve. In some embodiments, the valves 703A, 703B, 703C, 703D may include spring constant valves that can be opened and closed using an applied electrical signal. The electrical signal used to control the valves 703A, 703B, 703C, 703D may be based on a determined performance metric of the two-phase cooling system 101. Other types of 703A, 703B, 703C, 703D may be used in other embodiments of the present application.

[0123] 7, the inlet to the evaporator 103 has a smaller diameter than the riser 109. This is configured to provide a pressure force within the thermosiphon loop. This provides a unidirectional flow within the thermosiphon loop, preventing bubble formation at the inlet of the evaporator 103.

[0124] A first valve 703A is provided in the first branch 701A. The first valve 703A can be used to control the flow of working fluid through the first branch 701A. A second valve 703B is provided in the second branch 701B to control the flow of working fluid through the second branch 701B. In some embodiments, the valves 703A and 703B can be controlled to allow the working fluid 113 to flow through either the first branch 701A or the second branch 701B. Controlling whether the branch 701A or the branch 701B is used to supply the working fluid 113 to the evaporator 103 allows for adjustment of the liquid level in the downcomer 107, thereby controlling the flow of working fluid 113 through the thermosiphon loop.

[0125] In some embodiments, the two-phase cooling system 101 may be configured such that the working fluid 113 can flow through both the branches 701A and 701B simultaneously. This allows three different configurations of the two-phase cooling system 101 to be used: in a first configuration only the first branch 701A is used, in a second configuration only the second branch 701B is used, and in a third configuration both the first branch 701A and the second branch 701B are used.

[0126] The valves 703A, 703B can be operated based on the calculated performance index. Control signals can be provided to the valves 703A, 703B to open or close the valves 703A, 703B based on the performance index. This can be used to increase or decrease the flow of working fluid through the evaporator 103.

[0127] The two-phase cooling system 101 further includes a third valve 703C in the downcomer 107. The third valve 703C is disposed between an accumulator 205 and a branch point where the downcomer 107 branches into two or more branches 701. The third valve 703C can be used to control the amount of working fluid 113 entering the branches of the downcomer 107. This valve 703C can be operated based on a calculated performance index.

[0128] 7, the two-phase cooling system 101 further includes a reservoir 705. The reservoir 705 may include any means that may be configured to store a working fluid 113. The reservoir 705 may be configured to store the working fluid 113 in a liquid phase 117.

[0129] The reservoir 705 is coupled to the accumulator 205 in the downcomer 107 such that the working fluid 113 can flow from the reservoir 705 to the accumulator 205. A fourth valve 703D is provided between the reservoir 705 and the accumulator 205 such that the flow from the reservoir 705 to the accumulator 205 can be controlled.

[0130] The fourth valve 703D may also be controlled based on a determined performance index of the two-phase cooling system 101. The performance index may provide an indication of whether additional working fluid is needed in the two-phase cooling system 101. The fourth valve 703D may be closed when additional working fluid 113 is not needed in the thermosiphon loop and may be opened when additional working fluid 113 is needed in the thermosiphon loop.

[0131] It should be understood that different variations of the two-phase cooling system 101 may have different configurations of different branches 701. For example, in some embodiments, more than one branch 701 may be provided. Also, the branches 701 may have different diameters. In such embodiments, the flow of the working fluid 113 may be controlled such that the working fluid 113 flows through one branch 701 or through a combination of different branches 701.

[0132] As described above, the valves 703 may be operated based on the determined performance indicators of the two-phase cooling system 101. For example, if it is determined that the heat load transferred to the working fluid 113 exceeds a threshold value, the first valve 703A and the third valve 703C may be opened to allow the working fluid 113 to flow through the first branch 701A, which has a larger diameter, to avoid high static pressure in the downcomer 107 that may flood the condenser 105.

[0133] When the heat load in the two-phase cooling system 101 is determined to be low, the second valve 703B and the third valve 703C can be opened to allow the working fluid 113 to flow through the second branch 701B having a smaller diameter, which provides sufficient static pressure for low heat loads.

[0134] 8A and 8B show an embodiment of another method that can be used to implement embodiments of the present application.

[0135] 8A illustrates an example of a method that can be used to design a cooling system. At block 801, a number of input values ​​are obtained. These input values ​​may include a number of initial design choices related to the geometric shape of the simulated cooling system. These input values ​​may further include one or more values ​​estimated from experiments regarding the expected performance of the cooling system having the chosen shape.

[0136] At block 803, a machine learning model is employed. The machine learning may include methods such as those shown in Figures 3-6, or any other suitable methods. At block 805, the machine learning provides as output a performance metric of the simulated cooling system. In the example shown in Figure 8A, the performance metric provides an indication of thermal performance of the cooling system. The performance metric may include heat exchanged, pressure drop, or any other suitable performance metric.

[0137] In block 807, the initial input values ​​selected for the cooling system geometry are adjusted to improve the performance index. Any suitable method can be used to adjust the input values. In the embodiment shown in FIG. 8A, a machine learning gradient descent method is used. In other embodiments of the present application, other types of processes can be used.

[0138] The adjusted input values ​​are fed back as inputs to the machine learning model such that the adjusted input values ​​can be used to estimate a new performance index. The method of blocks 803 through 807 can be repeated until a performance index within a desired range is obtained. The performance index within the desired range may be an optimal performance index, or a substantially optimal performance index, for the cooling system.

[0139] This method provides a fast and accurate way to design cooling systems with high efficiency for heat transfer. By using machine learning to estimate the parameters used, it can provide more accurate design simulations than traditional methods.

[0140] FIG. 8B illustrates one example of a method that can be used to control a cooling system in use.

[0141] At block 811, data streams are obtained from one or more sensors 223. The sensors 223 may be distributed throughout the cooling system as shown in FIG. 2 or in any other suitable arrangement. The sensors 223 may be configured to sense temperature, pressure, mass flow, or any other parameter that may be used to determine a performance indicator of the cooling system. Data from sensors 223 may provide input values ​​including physical parameters of the cooling system.

[0142] In block 813, data obtained from the sensors 223 is used to calculate a performance index of the cooling system. The performance index may be calculated by estimating a dimensionless ratio using data from the sensors 223 and then calculating the performance index of the cooling system using the dimensionless ratio. The dimensionless ratio can be estimated using machine learning or any other suitable process.

[0143] In the embodiment of Figure 8B, the calculated performance metric includes a power usage effectiveness (PUE) for the cooling system or a portion of the cooling system, although other performance metrics may be used in other embodiments of the present application.

[0144] If the performance index is outside a threshold range, the performance index is provided as an input to a machine learning model in block 815. The data obtained from the sensor 223 in block 811 may also be provided as an input to a machine learning model. The machine learning model may be configured to determine how parameters of the cooling system should be adjusted in order for the performance index to be within a threshold range. The machine learning model used may be a non-linear regression algorithm or any other suitable type of algorithm.

[0145] The controlled system parameter may be the flow rate of working fluid through the cooling system, the diameter of a portion of the cooling system, the height of working fluid in the cooling system, or any other suitable parameter. These parameters may be adjusted by opening and closing valves as shown in Figure 7, or by controlling pumps in an active system, or by any other suitable means.

[0146] At block 817, parameters of the system are adjusted according to the output of the machine learning algorithm. For example, one or more control signals can be provided to one or more valves to control the opening and closing of the valves as needed. Similarly, in an active cooling system, the control signals may be provided to a pump or any other suitable component.

[0147] The adjusted parameters of the cooling system can then be used to calculate a new performance index for the cooling system. The method can be repeated as necessary using the adjusted parameters of the cooling system and further data obtained from sensors to determine a new performance index for the cooling system.

[0148] An example of the method of Figure 8B provides a control system for a cooling system that, through the use of machine learning, is fast, highly accurate, and highly responsive, allowing the cooling system to operate in a highly efficient manner.

[0149] 9A and 9B show the results obtained according to the examples of the present application.

[0150] 9A shows the estimates of the Nusselt number and friction pressure gradient for the two-phase cooling system 101 obtained by the best existing methods from the literature. They show a mean absolute percentage error (MAPE) of 24.4% for the Nusselt number and a mean absolute percentage error (MAPE) of 21.5% for the friction pressure gradient.

[0151] In contrast, FIG. 9B shows the estimates of the Nusselt number and friction pressure gradient for the two-phase cooling system 101 obtained by the method of the present application. In this example, the Nusselt number and friction pressure gradient were estimated using machine learning. These show a MAPE of 10.1% for the Nusselt number and 4.5% for the friction pressure gradient. It is therefore clear that the new Nusselt number and friction pressure gradient estimation method is a significant improvement over existing methods.

[0152] 10A and 10B show error histograms comparing results obtained with existing methods to results obtained with examples of the present application.

[0153] 10A shows an error histogram for the estimates of the Nusselt number. Plot 1001 represents the results obtained by the examples of the present application, and plot 1003 represents the results obtained by the best existing method from the literature. The histogram shows that with the existing method, only 30% of the estimates of the Nusselt number have an error of less than 10%, while with the present method, 76% have an error of less than 10%.

[0154] 10B shows an error histogram for the pressure gradient estimates. Plot 1005 represents the results obtained by the present application example, and plot 1007 represents the results obtained by the best existing method from the literature. The histogram shows that with the existing method, only 32% of the pressure gradient estimates have an error of less than 10%, while with the present method, 92% have an error of less than 10%.

[0155] These histograms clearly show that the present application provides a more accurate estimate of the dimensionless ratio than the current best existing methods from the literature. This improvement in accuracy improves the accuracy of the calculated performance indices and the accuracy of the parameters adjusted to control or design the cooling system.

[0156] FIG. 11 illustrates an example of an apparatus 1101 that can be used to implement embodiments of the present application.

[0157] The device 1101 includes a controller 1103. The implementation of the controller 1103 may be a controller circuit. The controller 1103 may be implemented only in hardware, may have certain parts in software including only firmware, or may be a combination of hardware and software (including firmware).

[0158] As shown in FIG. 11, the controller 1103 can be implemented using instructions that enable hardware functions, e.g., executable instructions of a computer program 1109 in a general-purpose or special-purpose processor 1105, which may be stored in a computer-readable storage medium (disk, memory, etc.) for execution by the processor 1105.

[0159] The processor 1105 is configured to read from and write to memory 1107. The processor 1105 may further include an output interface through which data and / or commands are output by the processor 1105, and an input interface through which data and / or commands are input to the processor 1105.

[0160] The memory 1107 stores a computer program 1109 including computer program instructions (computer program code 1111) that, when loaded into the processor 1105, controls the operation of the device 1101. The computer program instructions of the computer program 1109 provide logic and routines that enable the device 1101 to perform the methods illustrated in Figures 3-6, 8A and 8B. The processor 1105 is able to load and execute the computer program 1109 by reading the memory 1107.

[0161] Thus, the device 1101: at least one processor 1105 and at least one memory 1107 having computer program code 1111; The at least one memory 1107 and the computer program code 1111, together with the at least one processor 1105, cause the device 1101 to at least: receive 301 a number of input values ​​including at least some physical parameters of a cooling system; using the obtained input values ​​to estimate 303 a dimensionless ratio that provides an indication of at least one of: heat transfer within at least a portion of the cooling system; and pressure loss within at least a portion of the cooling system; determining 305 at least one performance index of at least a portion of a cooling system using the dimensionless ratio; and using 307 the at least one determined performance indicator to control one or more physical parameters of at least a portion of the cooling system.

[0162] As shown in Fig. 11, the computer program 1109 can reach the device 1101 via any suitable delivery mechanism 1113. The delivery mechanism 1113 can be, for example, a machine-readable medium, a computer-readable medium, a non-transitory computer-readable storage medium, a computer program product, a memory device, a storage medium such as a compact disc read-only memory (CD-ROM) or a digital versatile disc (DVD), or a solid-state memory or an article of manufacture that includes or embodies the computer program 1109. The delivery mechanism can also be a signal configured to reliably deliver the computer program 1109. The device 1101 can communicate or transmit the computer program 1109 as a computer data signal.

[0163] Thus, the computer program 1109 includes program instructions for causing the device 1101 to perform at least the following, or program instructions for causing the device 1101 to perform at least the following, namely: Receiving 301 a plurality of input values ​​including at least some physical parameters of a cooling system; using the obtained input values ​​to estimate 303 a dimensionless ratio that provides an indication of at least one of: heat transfer within at least a portion of the cooling system; and pressure loss within at least a portion of the cooling system; determining 305 at least one performance index of at least a portion of a cooling system using the dimensionless ratio; and using 307 the at least one determined performance indicator to control one or more physical parameters of at least a portion of the cooling system.

[0164] The computer program instructions may be included in a computer program 1109, a non-transitory computer readable medium, a computer program product, a machine readable medium, etc. In some embodiments, but not necessarily all embodiments, the computer program instructions may be distributed across multiple computer programs.

[0165] Although the memory 1107 is illustrated as a single component / circuit, it may be implemented as one or more separate components / circuits, some or all of which may be integrated / removable, and / or may provide permanent / semi-permanent / dynamic / cache storage.

[0166] Although the processor 1105 is illustrated as a single component / circuit, it may be implemented as one or more separate components / circuits, some or all of which may be integrated / removable. The processor 1105 may be a single-core or multi-core processor.

[0167] References herein to "computer-readable storage medium," "computer program product," "implementable computer program," etc., or "controller," "computer," "processor," etc., should be understood to include computers having different architectures, such as single / multi-processor architectures, Von Neumann / parallel architectures, as well as specialized circuitry, such as field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), signal processors, and other processing circuits. References to "computer program," "instructions," "code," etc., should be understood to include software for a programmable processor, or firmware, e.g., instructions for a processor, or the programmable content of a hardware device, such as the configuration settings of a fixed function device, gate array, programmable logic device, etc.

[0168] As used herein, the term "circuit" means (a) A hardware-only circuit implementation (e.g., an implementation using only analog and / or digital circuitry); (b) combinations of hardware circuitry and software, such as (where applicable) (i) combinations of analog and / or digital hardware circuitry and software / firmware, and (ii) any portion of hardware processor(s) with software (including digital signal processors), software, memory(s) that work together to cause a device, such as a mobile phone or a server, to perform various functions; (c) may refer to one or more or all of a hardware circuit(s) and / or processor(s), such as a microprocessor(s) or portion of a microprocessor(s), that requires software (e.g., firmware) for operation, and that is absent when not required for operation.

[0169] This definition of circuitry applies to all uses of the term in this application, including any claims. As another example, when used in this application, the term circuitry may also cover merely a hardware circuit or processor and its (or their) accompanying software and / or firmware implementations. The term circuitry may also cover, for example, a baseband integrated circuit for a mobile device, or a similar integrated circuit in a server, cellular network equipment, or other computing or network equipment, if applicable to the particular claim element.

[0170] The blocks illustrated in Figures 3-6 and 8A-8B may represent steps in a method and / or portions of code in the computer program 1109. The particular order illustrated in the blocks does not necessarily imply a required or preferred order for the blocks, and the order and arrangement of the blocks may be varied. Additionally, some blocks may be omitted.

[0171] As mentioned above, the system, apparatus, method and computer program may use machine learning, which may include statistical learning. Machine learning is a branch of computer science that gives computers the ability to learn without being explicitly programmed. The computer learns from experience E on a class of tasks T and a performance measure P where experience E improves performance on the tasks of T as measured by P. Machine learning includes fully or partially supervised learning and fully or partially unsupervised learning. The machine learning may allow for discrete outputs (e.g., classification, clustering) and continuous outputs (e.g., regression). Machine learning may be implemented using different methods, such as, for example, cost function minimization, artificial neural networks, support vector machines, Bayesian networks, etc. Cost function minimization may be used, for example, in linear regression, polynomial regression, K-means clustering. Artificial neural networks, for example, have one or more hidden layers and model complex relationships between input and output vectors. Support vector machines can be used in supervised learning. Bayesian networks are directed acyclic graphs that express the conditional independence of a number of random variables.

[0172] The term "comprising" as referred to in this application is intended to have an inclusive, rather than an exclusive, meaning that any reference to X containing Y means that X includes one or more Ys. Where an exclusive meaning of "comprising" is intended to be used, this will be made clear by reference to "including only one of..." or the context will make this clear by using "consisting of."

[0173] According to this specification, various embodiments may be referred to. A description of a feature or function with respect to an embodiment indicates that the feature or function is present in that embodiment. The term "embodiment," "for example," "may," or "may" used herein, whether or not explicitly stated, indicates that such feature or function is present in at least the described embodiment, whether or not stated in the embodiment, and may, but need not, be present in some or all of the other embodiments. Thus, the term "embodiment," "for example," "may," or "may" used herein indicates a particular embodiment in a class of embodiments. A property of an embodiment may be a property belonging only to that embodiment, a property of that class, or a property of a subclass of that class (including some, but not all, of the embodiments). Thus, a feature described with reference to one embodiment and not described with reference to another embodiment is implicitly indicated to the extent possible, but not necessarily to be used in other embodiments as part of a working combination.

[0174] Although the embodiments of the present application have been described in detail as in the above paragraphs, the present application is not limited to the above embodiments. It should be understood that any modifications, equivalent replacements, improvements, etc. made under the concept and principles of the present application are within the scope of protection of the present application.

[0175] Features set out in the above description may be used in combinations other than those explicitly set out above.

[0176] Although functions have been described with reference to particular functions, those functions may be performed by other functions, whether or not described.

[0177] Although features have been described with reference to particular embodiments, those features may also be present in other embodiments, whether or not they are described.

[0178] The term "a" or "the" referred to in this application is inclusive, not exclusive. That is, "X includes Y / said Y" referred to in this application indicates that X may include only one Y or may include two or more Y, unless the context clearly indicates the opposite. If "a" or "the" is intended to have an exclusive meaning, it must be clearly stated in the context. In some cases, "at least one" or "one or more" can be used to emphasize an inclusive meaning, but the absence of these terms should not be considered as a presumption of an exclusive meaning.

[0179] A feature (or combination of features) in a claim refers to the feature or (combination of features) itself, as well as features capable of achieving substantially the same technical effect (equivalent features). Equivalent features include, for example, modified features that achieve substantially the same result in substantially the same way. Equivalent features include, for example, features that perform substantially the same function in substantially the same way to achieve substantially the same result.

[0180] Various embodiments may be referred to in this specification using adjectives or adjectival phrases to describe features of the embodiments. The description of a property with respect to an embodiment indicates that the property is, in some embodiments, specifically as described, and in other embodiments, substantially as described.

[0181] Although an effort has been made in the above specification to direct attention to what are believed to be the key features, applicants should understand that they may seek protection through the claims to protect any patentable feature or combination of features shown in the above drawings and / or drawings, whether or not emphasized.

Claims

1. Obtaining a plurality of inputs including at least some physical parameters of a cooling system; using the obtained input values ​​to estimate a dimensionless ratio providing an indication of at least one of heat transfer of at least a portion of the cooling system and pressure loss of at least a portion of the cooling system, the dimensionless ratio being derived from a friction pressure gradient, which is a friction pressure loss per unit length; determining at least one performance index of at least a portion of a cooling system using the dimensionless ratio; and using said at least one determined performance index to control one or more physical parameters of at least a portion of the cooling system.

2. The apparatus of claim 1 , wherein the cooling system comprises a two-phase cooling system.

3. 3. Apparatus according to claim 1 or 2, wherein the physical parameters are controlled for the entire cooling system or for parts of the cooling system.

4. The apparatus of any one of claims 1 to 3, wherein the dimensionless ratio is estimated using machine learning.

5. The apparatus of claim 4 , wherein the dimensionless ratio is estimated using a non-linear regression algorithm.

6. Apparatus according to any one of the preceding claims, wherein at least one determined performance index is used to control one or more physical parameters of the cooling system during use.

7. The apparatus according to any one of claims 1 to 6, wherein the at least one determined performance index is used to control at least one of the flow rate of a flow passage in the cooling system, the position of a liquid level in a downcomer of a passive two-phase cooling system, and the rotation speed of a pump in an active two-phase cooling system.

8. An apparatus according to any preceding claim, wherein the plurality of input values ​​comprises information obtained from one or more sensors within the cooling system.

9. The apparatus according to any one of claims 1 to 3, wherein at least one determined performance index is used for modifying one or more physical parameters during a design of at least a part of the cooling system.

10. 10. The apparatus of claim 1, wherein the physical parameters of the cooling system include one or more of the following parameters: a geometrical shape of the cooling system, a working fluid in the cooling system, a thermodynamic condition in the cooling system, a flow condition in the cooling system.

11. Apparatus according to any one of the preceding claims, wherein the dimensionless ratio used is estimated from physical measurements of the cooling system.

12. The apparatus according to any one of the preceding claims, wherein the dimensionless ratio used is selected based on at least one of the geometry of the cooling system and the operating conditions of the cooling system.

13. Obtaining a plurality of inputs including at least some physical parameters of a cooling system; using the obtained input values ​​to estimate a dimensionless ratio providing an indication of at least one of heat transfer of the cooling system and pressure loss of the cooling system, the dimensionless ratio being derived from a friction pressure gradient, which is a friction pressure loss per unit length; determining at least one performance index of at least a portion of the cooling system using the dimensionless ratio; and and controlling one or more physical parameters of at least a portion of the cooling system using at least one of the determined performance indicators.

14. When performed by a processing circuit, Obtaining a plurality of inputs including at least some physical parameters of a cooling system; using the obtained input values ​​to estimate a dimensionless ratio providing an indication of at least one of heat transfer of at least a portion of the cooling system and pressure loss of at least a portion of the cooling system, the dimensionless ratio being derived from a friction pressure gradient, which is a friction pressure loss per unit length; determining at least one performance index of at least a portion of the cooling system using the dimensionless ratio; and controlling one or more physical parameters of at least a portion of the cooling system using at least one of the determined performance indicators; a computer program comprising computer program instructions for carrying out the steps of:

Images