Electrochemical sensor for sensing two-phase cooling fluid contamination of immersion cooling system

Abstract

An immersion cooling system includes an immersion tank defining an immersion chamber therein, an immersion working fluid, a first electrode, and a second electrode. The immersion working fluid is at least partially positioned in the immersion chamber. The first electrode is electrically coupled to a power source, and the second electrode is positioned proximate to the first electrode and defines a sampling region therebetween. The first electrode is coupled to a sampling region in which a sample portion of the submerged working fluid is positioned, and the second electrode is coupled to a microcontroller configured to measure at least a current across the sampling region, the sampling region being between the first electrode and the second electrode.

Description

[0001] Background and Related Technology

[0002] Computing devices can generate significant amounts of heat during use. Immersion cooling allows the temperature of computing devices and components to be regulated by an immersion fluid. However, contaminants in the working fluid may change the boiling temperature of the immersion fluid, change the conductivity of the immersion fluid, cause ion movement in the immersion fluid, cause dendrite growth or deposition, and other adverse effects. Summary of the invention

[0003] In some embodiments, the immersion cooling system includes an immersion tank defining an immersion chamber therein, an immersion working fluid, a first electrode, and a second electrode. The immersion working fluid is at least partially positioned in the immersion chamber. The first electrode is electrically coupled to a power source, the second electrode is positioned proximate to the first electrode and defines a sampling region between the first electrode and the second electrode. A sample portion of the immersion working fluid is positioned in the sampling region, and the second electrode is coupled to a microcontroller, the microcontroller being configured to at least measure a current across the sampling region between the first electrode and the second electrode.

[0004] In some embodiments, a method of measuring a liquid working fluid in an immersion cooling system includes: circulating an immersion working fluid in the immersion cooling system; positioning a sample portion of the immersion working fluid in a sampling region between a first electrode and a second electrode; applying a stimulation signal having a static voltage across the sampling region using the first electrode and the second electrode; varying a frequency of the stimulation signal; measuring an output signal across the sampling region; and determining a first concentration of a contaminant based at least in part on the output signal.

[0005] In some embodiments, a method of measuring and remediating a liquid working fluid in an immersion cooling system includes: circulating an immersion working fluid in the immersion cooling system; positioning a sample portion of the immersion working fluid in a sampling region between a first electrode and a second electrode; applying a stimulation signal having a static voltage across the sampling region using the first electrode and the second electrode; varying a frequency of the stimulation signal; measuring an output signal across the sampling region; determining a concentration of a contaminant based at least in part on the output signal using a microcontroller in communication with the first electrode and the second electrode. And remediating the immersion working fluid based at least in part on the concentration of the contaminant using the microcontroller.

[0006] This Summary is provided to introduce some concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.

[0007] Additional features and advantages will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the teachings herein. The features and advantages of the present disclosure may be realized and obtained by the means and combinations particularly noted in the appended claims. The features of the present disclosure will become more fully apparent from the following description and the appended claims or may be learned by practicing the present disclosure as described below. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to describe the manner in which the above and other features of the present disclosure can be obtained, a more specific description will be presented by reference to specific embodiments shown in the accompanying drawings. For better understanding, the same elements are represented by the same reference numerals in the various drawings. Although some of the drawings may be schematic or exaggerated representations of concepts, at least some of the drawings may be drawn to scale. It is necessary to understand that the drawings depict some exemplary embodiments, and the embodiments will be described and explained with additional specificity and detail through the use of the drawings, in which:

[0009] Figure 1 is a schematic diagram of an immersion cooling system according to at least one embodiment of the present disclosure;

[0010] Figure 2 is a schematic diagram of an immersion cooling system with an external condenser according to at least one embodiment of the present disclosure;

[0011] Figure 3 is an electrical diagram of an electrochemical sensor according to at least one embodiment of the present disclosure;

[0012] Figure 4 According to at least one embodiment of the present disclosure, Figure 3 Example profile of electrochemical sensor measurements;

[0013] Figure 5 is a schematic diagram of a machine learning model according to at least one embodiment of the present disclosure;

[0014] Figure 6 is a schematic diagram of an immersion cooling system with multiple electrochemical sensors according to at least one embodiment of the present disclosure;

[0015] Figure 7 is a flow chart illustrating a method of measuring contaminants in an immersion cooling system according to at least one embodiment of the present disclosure; and

[0016] Figure 8 is a schematic diagram of an immersion cooling system with remedial features in accordance with at least one embodiment of the present disclosure. DETAILED DESCRIPTION

[0017] The present disclosure generally relates to systems and methods for thermal management of electronic devices or other heat generating components. More specifically, the present disclosure relates to systems and methods for detecting contaminants in immersion cooling systems. For example, immersion cooling systems use an immersion working fluid that receives heat from computing devices, electronic components, and other heat generating components. The immersion working fluid then transfers the heat to a heat exchanger to remove the heat from the immersion cooling system.

[0018] In some examples, contaminants in the immersion working fluid may adversely affect one or more properties of the immersion working fluid, reducing the cooling capacity of the system. In other examples, contaminants in the immersion working fluid may be transported by the immersion working fluid and deposited elsewhere in the immersion cooling system. In at least one example, contaminant ions or particles in the immersion working fluid are deposited on the surface of the heat generating component, which reduces the thermal conductivity from the heat generating component to the immersion working fluid. In at least another example, contaminant ions or particles in the immersion working fluid are deposited on the surface of the heat generating component, which causes dendrite growth on the surface of the heat generating component and may cause an electrical short circuit.

[0019] In some embodiments, the immersed working fluid is a two-phase working fluid that receives heat from the heat generating component by evaporation. Evaporation can further cause deposition or precipitation of contaminants. For example, an immersion chamber surrounds the heat generating component in a liquid working fluid that conducts heat from the heat generating component to cool the heat generating component. When the working fluid absorbs heat from the heat generating component, the temperature of the working fluid increases. In some embodiments, the hot working fluid can be circulated through a thermal management system to cool the working fluid and / or replace the working fluid with a cold working fluid. In some embodiments, the working fluid evaporates, introducing vapor into the liquid of the working fluid that rises out of the liquid phase, and taking away heat energy from the heat generating component in the gas phase via latent heat of boiling.

[0020] In large computing centers, such as cloud computing centers, data processing centers, data storage centers, or other computing facilities, immersion cooling systems provide an effective thermal management method for many computing components under various operating loads. In some embodiments, the immersion cooling system includes a working fluid in an immersion chamber and a heat exchanger for cooling the liquid phase and / or a condenser for extracting heat from the gas phase of the working fluid. The heat exchanger may include a condenser that condenses the gas phase of the working fluid into a liquid phase and returns the liquid working fluid to the immersion chamber. In some embodiments, the liquid working fluid absorbs heat from the heat generating components, and one or more fluid conduits direct the hot liquid working fluid outside the immersion chamber to a radiator, a heat exchanger, or a lower temperature area to cool the liquid working fluid.

[0021] In some embodiments, computing devices or systems assigned to immersion cooling systems and / or high computing applications executed thereon require extensive thermal management. The working fluid boils to absorb heat to overcome the latent heat of boiling. Thus, the phase change from liquid to vapor allows the working fluid to absorb a significant amount of heat with a small associated temperature increase. In addition, the lower density allows for efficient removal of vapor from the immersion bath to remove the associated heat from the system.

[0022] In some embodiments, the thermal management system includes an immersion tank in which a two-phase working fluid is positioned. The two-phase working fluid receives heat from a heat generating component immersed in a liquid working fluid, and the heat evaporates the working fluid, causing the working fluid to change from a liquid phase to a gas phase. The thermal management system includes a condenser such as described herein for condensing the vapor working fluid back into a liquid phase. In some embodiments, the condenser is in fluid communication with the immersion tank through one or more conduits. In some embodiments, the condenser is positioned inside the immersion tank.

[0023] Figure 1 The conventional immersion cooling system 100 shown includes an immersion tank 102 including an immersion chamber 104 and a condenser 106 in the immersion chamber 104. The immersion chamber 104 contains an immersion working fluid having a portion of a liquid working fluid 108 and a vapor working fluid 110. The liquid working fluid 108 forms an immersion bath 112 in which a plurality of heat generating components 114 are positioned to heat the liquid working fluid 108 on a support 116.

[0024] Reference now Figure 2 In some embodiments, the immersion cooling system 200 includes an immersion tank 202 defining an immersion chamber 204 in which an immersion working fluid is positioned. The immersion working fluid in the immersion tank 202 has a boiling temperature that is at least partially related to one or more operating characteristics of the immersion cooling system, electronic components and / or computing devices in the immersion tank 202, computing or workload of the electronic components and / or computing devices in the immersion tank 202, external and / or environmental conditions, or other characteristics that affect the operation of the immersion cooling system 200. When the operating conditions of the immersion cooling system 200 change, the immersion cooling system 200 can change the mixing ratio of the immersion working fluid to change at least one characteristic (e.g., boiling temperature) of the immersion working fluid.

[0025] In some embodiments, the immersion working fluid transitions between a liquid working fluid 208 phase and a vapor working fluid 210 phase to remove heat from a hot or heat generating component 214 in the immersion chamber 204. The liquid working fluid 208 more efficiently receives heat from the heat generating component 214, and upon transitioning to the vapor working fluid 210, the vapor working fluid 210 can be removed from the immersion tank 202, cooled and condensed by the condenser 206 (or other heat exchanger) to extract heat from the working fluid, and the liquid working fluid 208 can be returned to the liquid immersion bath 212.

[0026] In some embodiments, an immersion bath 212 of a liquid working fluid 208 has a plurality of heat generating components 214 positioned in the liquid working fluid 208. The liquid working fluid 208 surrounds the heat generating components 214 and at least a portion of other objects or components attached to the heat generating components 214. In some embodiments, the heat generating components 214 are positioned in the liquid working fluid 208 on one or more supports 216. The supports 216 can support the one or more heat generating components 214 in the liquid working fluid 208 and allow the working fluid to move around the heat generating components 214. In some embodiments, the supports 216 are thermally conductive to conduct heat from the heat generating components 214. The support(s) 216 can increase the effective surface area from which the liquid working fluid 208 can remove heat by convection cooling.

[0027] In some embodiments, the heat generating component 214 includes an electronic or computing component or a power supply. In some embodiments, the heat generating component 214 includes a computer device, such as an individual personal computer or a server blade computer. In some embodiments, one or more heat generating components 214 include a heat sink or other device attached to the heat generating component 214 to conduct away thermal energy and effectively increase the surface area of ​​the heat generating component 214. In some embodiments, the heat sink of the heat generating component 214 is a vapor chamber having one or more three-dimensional structures to increase the surface area.

[0028] As described, the conversion of the liquid working fluid 208 to the vapor phase requires input of thermal energy to overcome the latent heat of vaporization and can be an effective mechanism to increase the heat capacity of the working fluid and remove heat from the heat generating components 214. As the vapor working fluid 210 rises in the liquid working fluid 208, the vapor working fluid 210 can be extracted from the immersion chamber 204 in the upper vapor region of the chamber. The condenser 206 cools a portion of the vapor working fluid 210 back to the liquid working fluid 208, removing the thermal energy from the system and reintroducing the working fluid into the immersion bath 212 of the liquid working fluid 208. The condenser 206 radiates or otherwise dumps the thermal energy from the working fluid into the surrounding environment or into a conduit to carry the thermal energy away from the cooling system.

[0029] In some embodiments of the immersion cooling system, a liquid-cooled condenser is integrated into the immersion tank and / or chamber to effectively remove thermal energy from the working fluid. In some embodiments, the immersion cooling system 200 for thermal management of a computing device allows at least one immersion tank 202 and / or chamber 204 to be connected to and in fluid communication with an external condenser 206. In some embodiments, the immersion cooling system 200 includes a vapor return line 218 that connects the immersion tank 202 to the condenser 206 and allows the vapor working fluid 210 to enter the condenser 206 from the immersion tank 202 and / or chamber 204, and a liquid return conduit 220 that connects the immersion tank 202 to the condenser 206 and allows the liquid working fluid 208 to return to the immersion tank 202 and / or chamber 204.

[0030] The vapor return line 218 can be cooler than the boiling temperature of the working fluid. In some embodiments, a portion of the vapor working fluid 210 condenses in the vapor return line 218. In some embodiments, the vapor return line 218 can be oriented at an angle so that the vapor return line 218 is not perpendicular to the direction of gravity. The condensed working fluid can then be discharged back to the immersion tank 202 or forward to the condenser 206 depending on the direction in which the vapor return line 218 is tilted. In some embodiments, the vapor return line 218 includes a liquid collection line or valve, such as a drain valve, which allows the condensed working fluid to be collected and / or returned to the immersion tank 202 or the condenser 206.

[0031] In some examples, the immersion cooling system 200 includes an air-cooled condenser 206. The air-cooled condenser 206 may require a fan or pump to force ambient air through one or more heat pipes or heat sinks to conduct heat from the condenser to the air. In some embodiments, the circulation of the immersion working fluid through the immersion cooling system 200 causes the liquid working fluid 208 to flow through one or more heat generating components 214. In the example of a heat generating component 214 with a vapor chamber heat sink, the dynamics of the liquid working fluid 208 can be used to move the vapor chamber working fluid within the vapor chamber and / or the boiling of the immersion working fluid by the vapor chamber can drive the flow of the immersion working fluid.

[0032] In some embodiments, the liquid working fluid receives heat in the cooling volume of the working fluid directly surrounding the heat generating components. The cooling volume is the area of ​​the working fluid (including both the liquid and gas phases) that directly surrounds the heat generating components and is responsible for the convective cooling of the heat generating components. In some embodiments, the cooling volume is the volume of the working fluid within 5 millimeters (mm) of these heat generating components.

[0033] The immersion working fluid has a boiling temperature below a critical temperature at which the heat generating component undergoes thermal damage. The immersion working fluid can thereby receive heat from the heat generating component to cool the heat generating component before the heat generating component is damaged.

[0034] For example, the heat generating component may be a computing component that is subject to damage above 100°C. In some embodiments, the boiling temperature of the immersion working fluid is lower than the critical temperature of the heat generating component. In some embodiments, the boiling temperature of the immersion working fluid is less than about 90°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is less than about 80°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is less than about 70°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is less than about 60°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is at least about 35°C at 1 atmosphere. In some embodiments, the working fluid comprises water.

[0035] In some embodiments, the working fluid comprises ethylene glycol. In some embodiments, the working fluid comprises a combination of water and ethylene glycol. In some embodiments, the working fluid comprises an aqueous solution. In some embodiments, the working fluid comprises an electronic liquid, such as FC-72 available from 3M, or a similar non-conductive fluid. In some embodiments, a heat generating component, support, or other element of an immersion cooling system positioned in the working fluid has nucleation sites on its surface that promote nucleation of vapor bubbles of the working fluid at or below the boiling temperature of the working fluid.

[0036] In some embodiments, sensors according to the present invention detect and / or measure the presence of one or more contaminants in an immersion working fluid. In some embodiments, the immersion working fluid is a two-phase working fluid that boils during operation of the immersion cooling system. In some embodiments, the immersion working fluid is a single-phase working fluid that is maintained in a liquid phase by operation of the immersion cooling system. Detecting the presence of contaminants in a single-phase immersion working fluid is beneficial to the operation of the single-phase immersion cooling system, but two-phase working fluids experience the additional consideration of the effect on the boiling temperature of the immersion working fluid and the increased deposition rate of contaminants on components in the immersion chamber due to boiling of the working fluid.

[0037] Figure 33 is a schematic diagram of an electrochemical sensor 324 configured to measure ionic contaminants in a solution immersed in a working fluid according to at least some embodiments of the present disclosure. The sensor 324 includes a first electrode 326 and a second electrode 328 defining a sampling region 330 therebetween. For example, a first sampling surface 332 of the first electrode 326 and a second sampling surface 334 of the second electrode 328 may be substantially parallel to each other with a constant electrode gap 336. In other examples, at least a portion of the first sampling surface 332 of the first electrode 326 and the second sampling surface 334 of the second electrode 328 are not parallel to each other, and the electrode gap 336 is defined by a portion of the first sampling surface 332 that is closest to the second sampling surface 334.

[0038] In some embodiments, the electrode gap 336 is less than 100 microns. In some embodiments, the electrode gap 336 is less than 50 microns. In some embodiments, the electrode gap 336 is less than 20 microns. In some embodiments, the electrode gap 336 is less than 10 microns. Smaller electrode gaps 336 allow for lower voltages and / or amperages required across the sampling region 330, however, the immersion working fluid may include larger debris particles that may adversely affect the sensor 324 if the larger particles enter the sampling region.

[0039] Sensor 324 may also include a power source including a direct current (DC) source 338 and an alternating current (AC) stimulus 340 having variable amplitude and frequency. Sensor 324 applies a current and a voltage selected from DC source 338 and AC stimulus 340 to a sample portion of the immersed working fluid in sampling region 330.

[0040] A measurement system (configured to measure voltage and current) is connected across the sampling region 330. The measurement system includes a microcontroller 342 configured to receive voltage measurements from a voltmeter 344 and current measurements from an ammeter 346. The stimulation signal provided by the DC source 338 and the AC stimulus 340 that drives the electrochemical signal includes a static voltage coupled with a small amplitude swept sine wave. Different contaminants in the immersion working fluid will have different static voltages. By sweeping the stimulation signal provided by the DC source 338 and the AC stimulus 340 through a voltage range, the electrochemical response at and / or around the static voltage can be measured.

[0041] The voltage across the electrochemical cell and the current through it are measured during the frequency sweep. These signals are sampled by an analog-to-digital converter and stored in memory as complex numbers with measured signal amplitude and phase. The resultant complex voltage and current measurements at each frequency point are processed to produce a complex impedance value at that frequency. The resulting two-dimensional impedance data points are processed, for example, by a microcontroller 342 to convert the measured amplitude and phase into real and imaginary coordinates. These points form a contour, such as Figure 4shown.

[0042] In some embodiments, multiple profiles will be captured over time to track and / or correlate the evolution of the profile shape based on changes in concentration and / or stimulation signals. A machine learning (ML) model can be trained based on known concentrations of one or more contaminants to allow the ML model to identify or calculate the concentration of one or more contaminants and / or other contaminants in the immersion working fluid relative to the stimulation signal.

[0043] Figure 5 It is a schematic representation of an ML model that can be used with one or more embodiments of the systems and methods described herein. As used herein, a "machine learning model" refers to a computer algorithm or model (e.g., a classification model, a regression model, a language model, an object detection model) that can be adjusted (e.g., trained) to approximate an unknown function based on a training input. For example, an ML model may refer to a neural network or other machine learning algorithm or architecture that learns and approximates complex functions and generates outputs based on multiple inputs provided to the machine learning model. In some embodiments, the ML system, model, or neural network described herein is an artificial neural network. In some embodiments, the ML system, model, or neural network described herein is a convolutional neural network. In some embodiments, the ML system, model, or neural network described herein is a recursive neural network. In at least one embodiment, the ML system, model, or neural network described herein is a Bayesian classifier. As used herein, a "machine learning system" may refer to one or more ML models that collaboratively generate one or more outputs based on corresponding inputs. For example, an ML system may refer to any system architecture with multiple discrete ML components that consider different kinds of information or inputs.

[0044] As used herein, an "instance" refers to an input object that can be provided as an input to an ML system for use in generating an output, such as a type of contaminant present, a concentration of a contaminant present, or a rate of change of a contaminant present. For example, an instance can refer to any event in which a contaminant is present. For example, a first profile or other measurement received at a microcontroller can be associated with a first contaminant present in a sample portion of the immersion working fluid. In another example, a second profile or other measurement received at the microcontroller can be associated with a second contaminant present in the sample portion of the immersion working fluid. In another example, a change in a profile or other measurement received at the microcontroller can be associated with an increase in the concentration of a contaminant in the sample portion of the immersion working fluid.

[0045] In some embodiments, the machine learning system has multiple layers, with an input layer 548 configured to receive at least one input training data set 550 or input training instances 552, and an output layer 558, with multiple additional or hidden layers 556 therebetween. The training data set may be input into the ML system to train the ML system and identify individual and combinations of contaminants or other attributes of the training instances that allow the microcontroller to identify concentrations of known or unknown contaminants.

[0046] In some embodiments, the machine learning system can receive multiple training data sets simultaneously and learn from different training data simultaneously.

[0047] In some embodiments, the machine learning system includes multiple machine learning models operating together. Each of the machine learning models has multiple hidden layers between the input layer and the output layer. The hidden layer has multiple input nodes (e.g., node 560), each of which operates on the input received from the previous layer. In a specific example, the first hidden layer has multiple nodes, and each of the nodes performs an operation on each instance from the input layer. Each node of the first hidden layer provides a new input to each node of the second hidden layer, and each node of the second hidden layer in turn performs a new operation on each of these inputs. The nodes of the second hidden layer then pass the output (such as the identified cluster 562) to the output layer.

[0048] In some embodiments, each of the nodes 560 has a linear function and an activation function. The linear function may attempt to optimize or approximate a solution having a line of best fit, such as reduced power cost or reduced latency. The activation function operates as a test for verifying the validity of the linear function. In some embodiments, the activation function produces a binary output that determines whether the output of the linear function is passed to the next layer of the machine learning model. In this way, the machine learning system can limit and / or prevent poor fits from propagating to the data and / or non-convergent solutions.

[0049] The machine learning model includes an input layer that receives at least one training data set. In some embodiments, at least one machine learning model uses supervised training. In some embodiments, at least one machine learning model uses unsupervised training. Unsupervised training can be used to draw inferences and find patterns or associations from (multiple) training data sets without known outputs. In some embodiments, unsupervised learning can identify clusters of similar labels or features for various training instances and allow the machine learning system to extrapolate the performance of instances with similar features.

[0050] In some embodiments, semi-supervised learning can combine the benefits from supervised learning and unsupervised learning. As described herein, a machine learning system can identify associated labels (such as known pollutants or pollutant concentrations) or features between instances, which can allow for the fusion of a training data set with known outputs and a second training data set that includes more general input information. Unsupervised training can allow a machine learning system to cluster instances from a second training data set without known outputs and associate the clusters with known outputs from a first training data set.

[0051] Figure 6 6 is a system diagram of an immersion cooling system 600 having a plurality of electrochemical sensors 624-1, 624-2, 624-3, 624-4, 624-5, 624-6 positioned in contact with a liquid phase of an immersion working fluid (e.g., liquid working fluid 608). In some embodiments, the immersion cooling system 600 includes electrochemical sensors 624-1, 624-2, 624-3, 624-4, 624-5, 624-6 positioned at various locations in the immersion cooling system 600. For example, a first electrochemical sensor 624-1 can be positioned in a lower portion of an immersion chamber 604 having a heat generating component 614. The first electrochemical sensor 624-1 can be positioned to measure one or more contaminants present in the liquid working fluid 608 in contact with the heat generating component 614. As described herein, two-phase immersion working fluids present an increased risk of dendrite growth during evaporation of liquid working fluid 608 in contact with heat generating component 614. For at least this reason, it may be desirable to position first electrochemical sensor 624-1 to measure contaminants proximate heat generating component 614.

[0052] In another example, the second electrochemical sensor 624-2 can be positioned near the liquid-vapor boundary between the liquid working fluid 608 and the vapor working fluid 610. The second electrochemical sensor 624-2 can measure the concentration of contaminants at or near the top of the liquid working fluid 608 in the immersion chamber 604 to provide a relative concentration between the bottom and top near the first electrochemical sensor 624-1, for example, to help determine the source or source location of the contaminants in the liquid working fluid 608.

[0053] In yet another example, one or more third electrochemical sensors 624-3 can be positioned proximate to the heat generating component 614, such as within a cooling volume proximate to the heat generating component 614. The one or more third electrochemical sensors 624-3 can measure the presence of contaminants proximate to the heat source in the immersion chamber 604, and thus measure the amount of contaminants present in the boiling portion of the liquid working fluid 608. As described herein, boiling of the liquid working fluid 608 may increase the deposition rate of contaminant ions and / or the rate of dendrite growth on the heat generating component 614 (e.g., on a pin or interconnect of a computing device). Measuring contaminants proximate to the heat generating component 614 can provide information to the microcontroller 642 to predict the risk of damage or failure of the heat generating component 614.

[0054] In another example, the fourth electrochemical sensor 624-4 can be positioned in or near the condenser 606. In some embodiments, the condenser 606 condenses the vapor working fluid 610 into a condensed liquid working fluid 608 in or in contact with the condenser 606. In at least one example, the liquid working fluid 608 is converted into the vapor working fluid 610 that does not contain contaminant ions when boiling, and the condensed liquid working fluid 608 in or on the condenser 606 does not contain contaminant ions unless at least some of the contaminant ions in the liquid working fluid 608 originate from the condenser 606. Measuring the contaminant concentration at or in the condenser 606 using the fourth electrochemical sensor 624-4 can provide information related to the source or source location of the contaminant ions in the liquid working fluid 608 to the microcontroller 642.

[0055] In yet another example, the fifth electrochemical sensor 624-5 is positioned at a first end of the return conduit 620 of the immersion cooling system 600, and / or the sixth electrochemical sensor 624-6 is positioned at a second end of the return conduit 620. In some embodiments, the immersion cooling system 600 delivers the condensed liquid working fluid 608 from the condenser 606 to the immersion chamber 604 through the return conduit 620. By detecting contaminants and / or measuring contaminant levels at one or more locations in the return conduit 620, the fifth electrochemical sensor 624-5 and / or the sixth electrochemical sensor 624-6 can provide information related to the source or source location of contaminant ions in the liquid working fluid 608 to the microcontroller 642.

[0056] In some embodiments, the microcontroller 642 calculates the concentration of contaminants in the liquid working fluid 608 at one or more locations in the immersion cooling system 600. The microcontroller 642 or other control plane of the immersion cooling system may recommend or automatically perform remediation of the liquid working fluid to remove contaminants from the liquid working fluid 608 and / or replace contaminated liquid working fluid 608 with new, uncontaminated working fluid.

[0057] Figure 7 7 is a flow chart illustrating an embodiment of a method 744 for measuring and remediating a liquid working fluid in an immersion cooling system. In some embodiments, the method 744 includes positioning a sample portion of an immersion working fluid in a sampling region between a first electrode and a second electrode at 746, and applying a stimulation signal having a static voltage across the sampling region using the first electrode and the second electrode at 748. The method 744 includes varying the frequency of the input signal at 750 and measuring an output signal across the sampling region at 752, as described with respect to Figure 3 and Figure 4 As described.

[0058] In some embodiments, method 744 is performed on-line while the immersion cooling system is operating. For example, the sample portion of the immersion working fluid at 746 may be a portion of the liquid working fluid in the immersion chamber, in the condenser, in the return conduit, or in other portions of the immersion cooling system. Sampling multiple times and / or at multiple locations in the immersion cooling system may allow the microcontroller to determine the concentration of the contaminant at 754 based at least in part on the measured output signal.

[0059] In some embodiments, method 744 optionally includes remediating a portion of the submerged working fluid at 756. For example, remediating a portion of the submerged working fluid may include changing one or more filters in the submerged cooling system, such as near an inlet or outlet of a condenser, near an inlet or outlet of an immersion tank or chamber, near an inlet or outlet of a return conduit, or at other locations in the submerged cooling system.

[0060] In another example, remediating a portion of the immersion working fluid may include adding uncontaminated immersion working fluid to the immersion cooling system, such as opening a valve to add the uncontaminated immersion working fluid to the immersion cooling system. Adding the uncontaminated immersion working fluid to the immersion cooling system may dilute contaminant ions in the existing immersion working fluid, thereby reducing the effect of the contaminant ions on one or more properties of the immersion working fluid, such as conductivity, boiling temperature, or other properties. In other examples, diluting the contaminant ions in the immersion working fluid may reduce deposition rates and reduce dendrite growth. In at least one embodiment, adding the uncontaminated immersion working fluid to the immersion cooling system includes opening a valve to allow the uncontaminated immersion working fluid to flow from the reservoir through the valve into the immersion cooling system. For example, the valve may be controlled by a microcontroller, and the microcontroller may open the valve based at least in part on the measured output signal(s).

[0061] In yet another example, remediating a portion of the immersion working fluid may include replacing the contaminated immersion working fluid with uncontaminated immersion working fluid. Replacing the contaminated immersion working fluid with uncontaminated immersion working fluid may include opening at least a first valve to drain a portion of the contaminated immersion working fluid from the immersion cooling system, and opening at least a second valve to allow uncontaminated immersion working fluid to flow from the reservoir through the valve into the immersion cooling system. In at least one embodiment, the valves may be controlled by the microcontroller, and the microcontroller may open the valves based at least in part on the measured output signal(s).

[0062] Figure 8 800 is a schematic diagram of an embodiment of an immersion cooling system 800 with remediation capabilities. For example, the immersion cooling system 800 may include a reservoir 858 containing uncontaminated immersion working fluid 860 and a collection tank 862 configured to capture at least a portion of the contaminated liquid working fluid 808. In some embodiments, the reservoir 858 is connected to the return conduit 820 or other inlet to the immersion chamber 804 to introduce the uncontaminated immersion working fluid 860 into the immersion chamber 804. In other examples, the reservoir 852 may be connected to the immersion chamber 804 directly or through a dedicated conduit, or the reservoir 852 may be connected to the condenser 806 to deliver the uncontaminated immersion working fluid 860 to the immersion cooling system 800 via the condenser 806.

[0063] In some embodiments, the first valve 864 controls the flow of uncontaminated immersion working fluid 860 into the immersion cooling system 800. The first valve 864 can be controlled by the microcontroller 842. In other embodiments, the first valve 864 can be controlled by another control plane that receives contaminant and / or contaminant concentration information from the microcontroller 842.

[0064] The collection tank 862 may be connected to the immersion cooling system 800 via a second valve 866. In some embodiments, the collection tank 862 is connected to the return conduit 820 or other conduit of the immersion chamber 804 to receive the contaminated liquid working fluid 808 from the immersion chamber 804. In other examples, the collection tank 862 may be connected to the immersion chamber 804 directly or via a dedicated conduit, or the collection tank 862 may be connected to the condenser 806 to receive the contaminated liquid working fluid 808 from the immersion cooling system 800 via the condenser 806.

[0065] In some embodiments, the second valve 866 controls the flow of the contaminated liquid working fluid 808 into the immersion cooling system 800. The second valve 866 can be controlled by the microcontroller 842. In other embodiments, the second valve 866 can be controlled by another control plane that receives contaminants and / or contaminant concentration information from the microcontroller 842. In at least one embodiment, the microcontroller 842 or other control plane at least partially opens the first valve 864 and the second valve 866 simultaneously (e.g., the time period when the first valve 864 is open at least partially overlaps the time period when the second valve 866 is open) to introduce uncontaminated immersion working fluid 860 from the reservoir 858 into the immersion cooling system 800 while removing at least a portion of the contaminated liquid working fluid 808 into the collection tank 862.

[0066] Industrial Applicability

[0067] The present disclosure generally relates to systems and methods for thermal management of electronic devices or other heat generating components. More specifically, the present disclosure relates to systems and methods for detecting contaminants in immersion cooling systems. For example, immersion cooling systems use an immersion working fluid that receives heat from computing devices, electronic components, and other heat generating components. The immersion working fluid then transfers the heat to a heat exchanger to remove the heat from the immersion cooling system.

[0068] In some examples, contaminants in the immersion working fluid may adversely affect one or more properties of the immersion working fluid, reducing the cooling capacity of the system. In other examples, contaminants in the immersion working fluid may be transported by the immersion working fluid and deposited elsewhere in the immersion cooling system. In at least one example, contaminant ions or particles in the immersion working fluid are deposited on the surface of the heat generating component, which reduces the thermal conductivity from the heat generating component to the immersion working fluid. In at least another example, contaminant ions or particles in the immersion working fluid are deposited on the surface of the heat generating component, which causes dendrite growth on the surface of the heat generating component and may cause an electrical short circuit.

[0069] In some embodiments, the immersed working fluid is a two-phase working fluid that receives heat from the heat generating component by evaporation. Evaporation can further cause deposition or precipitation of contaminants. For example, an immersion chamber surrounds the heat generating component in a liquid working fluid that conducts heat from the heat generating component to cool the heat generating component. When the working fluid absorbs heat from the heat generating component, the temperature of the working fluid increases. In some embodiments, the hot working fluid can be circulated through a thermal management system to cool the working fluid and / or replace the working fluid with a cold working fluid. In some embodiments, the working fluid evaporates, introducing vapor into the liquid of the working fluid that rises out of the liquid phase, and taking away heat energy from the heat generating component in the gas phase via latent heat of boiling.

[0070] In large computing centers, such as cloud computing centers, data processing centers, data storage centers, or other computing facilities, immersion cooling systems provide an effective thermal management method for many computing components under various operating loads. In some embodiments, the immersion cooling system includes a working fluid in an immersion chamber and a heat exchanger for cooling the liquid phase and / or a condenser for extracting heat from the gas phase of the working fluid. The heat exchanger may include a condenser that condenses the gas phase of the working fluid into a liquid phase and returns the liquid working fluid to the immersion chamber. In some embodiments, the liquid working fluid absorbs heat from the heat generating components, and one or more fluid conduits direct the hot liquid working fluid outside the immersion chamber to a radiator, a heat exchanger, or a lower temperature area to cool the liquid working fluid.

[0071] In some embodiments, computing devices or systems assigned to immersion cooling systems and / or high computing applications executed thereon require extensive thermal management. The working fluid boils to absorb heat to overcome the latent heat of boiling. Thus, the phase change from liquid to vapor allows the working fluid to absorb a significant amount of heat with a small associated temperature increase. In addition, the lower density allows for efficient removal of vapor from the immersion bath to remove the associated heat from the system.

[0072] In some embodiments, the thermal management system includes an immersion tank in which a two-phase working fluid is positioned. The two-phase working fluid receives heat from a heat generating component immersed in a liquid working fluid, and the heat evaporates the working fluid, causing the working fluid to change from a liquid phase to a gas phase. The thermal management system includes a condenser such as described herein for condensing the vapor working fluid back into a liquid phase. In some embodiments, the condenser is in fluid communication with the immersion tank through one or more conduits. In some embodiments, the condenser is positioned inside the immersion tank.

[0073] Some conventional immersion cooling systems include an immersion tank including an immersion chamber and a condenser in the immersion chamber. The immersion chamber contains an immersion working fluid having a liquid working fluid and a vapor working fluid portion. The liquid working fluid creates an immersion bath in which a plurality of heat generating components are positioned to heat the liquid working fluid on a support.

[0074] In some embodiments, the immersion cooling system includes an immersion tank defining an immersion chamber, an immersion working fluid positioned in the immersion chamber, the immersion working fluid in the immersion tank having a boiling temperature, the boiling temperature being at least partially related to one or more operating characteristics of the immersion cooling system, electronic components and / or computing devices in the immersion tank, computational or workload of the electronic components and / or computing devices in the immersion tank, external and / or environmental conditions, or other characteristics that affect operation of the immersion cooling system.

[0075] In some embodiments, the immersion working fluid transitions between a liquid working fluid phase and a vapor working fluid phase to remove heat from a hot or heat generating component in the immersion chamber. The liquid working fluid more efficiently receives heat from the heat generating component, and upon transitioning to a vapor working fluid, the vapor working fluid can be removed from the immersion tank, cooled and condensed by a condenser (or other heat exchanger) to extract heat from the working fluid, and the liquid working fluid can be returned to the liquid immersion bath.

[0076] In some embodiments, the immersion bath of the liquid working fluid has a plurality of heat generating components positioned in the liquid working fluid. The liquid working fluid surrounds the heat generating components and at least a portion of other objects or components attached to the heat generating components. In some embodiments, the heat generating components are positioned in the liquid working fluid on one or more supports. The support can support one or more heat generating components in the liquid working fluid and allow the working fluid to move around the heat generating components. In some embodiments, the support is thermally conductive to conduct heat from the heat generating components. (Multiple) supports can increase the effective surface area, and the liquid working fluid can remove heat from the effective surface area by convection cooling.

[0077] In some embodiments, the heat generating component comprises an electronic or computing component or a power supply. In some embodiments, the heat generating component comprises a computer device, such as an individual personal computer or a server blade computer. In some embodiments, one or more heat generating components comprise a heat sink or other device attached to the heat generating component to conduct away heat energy and effectively increase the surface area of ​​the heat generating component. In some embodiments, the heat sink of the heat generating component is a vapor chamber having one or more three-dimensional structures to increase the surface area.

[0078] As described, the conversion of the liquid working fluid to the vapor phase requires input of thermal energy to overcome the latent heat of evaporation and can be an effective mechanism to increase the heat capacity of the working fluid and remove heat from the heat generating components. Because the vapor working fluid rises in the liquid working fluid, the vapor working fluid can be extracted from the immersion chamber in the upper vapor region of the chamber. The condenser cools a portion of the vapor working fluid back to the liquid working fluid, removing thermal energy from the system and reintroducing the working fluid into the immersion bath of the liquid working fluid. The condenser radiates or otherwise dumps the thermal energy from the working fluid into the surrounding environment or into a conduit to carry the thermal energy away from the cooling system.

[0079] In some embodiments of the immersion cooling system, a liquid-cooled condenser is integrated into the immersion tank and / or chamber to effectively remove thermal energy from the working fluid. In some embodiments, the immersion cooling system for thermal management of a computing device allows at least one immersion tank and / or chamber to be connected to and in fluid communication with an external condenser. In some embodiments, the immersion cooling system includes a vapor return line connecting the immersion tank to the condenser and allowing vapor working fluid to enter the condenser from the immersion tank and / or chamber, and a liquid return conduit connecting the immersion tank to the condenser and allowing liquid working fluid to return to the immersion tank and / or chamber.

[0080] The vapor return line can be cooler than the boiling temperature of the working fluid. In some embodiments, a portion of the vapor working fluid condenses in the vapor return line. In some embodiments, the vapor return line can be oriented at an angle so that the vapor return line is not perpendicular to the direction of gravity. The condensed working fluid can then be discharged back to the immersion tank or forward to the condenser depending on the inclination direction of the vapor return line. In some embodiments, the vapor return line includes a liquid collection line or valve, such as a drain valve, which allows the condensed working fluid to be collected and / or returned to the immersion tank or condenser.

[0081] In some examples, the immersion cooling system includes an air-cooled condenser. The air-cooled condenser may require a fan or pump to force ambient air through one or more heat pipes or heat sinks to transfer heat from the condenser to the air. In some embodiments, the circulation of the immersion working fluid through the immersion cooling system causes the liquid working fluid to flow through one or more heat generating components. In the example of a heat generating component with a vapor chamber heat sink, the dynamics of the liquid working fluid can be used to move the vapor chamber working fluid within the vapor chamber and / or boiling of the immersion working fluid by the vapor chamber can drive the flow of the immersion working fluid.

[0082] In some embodiments, the liquid working fluid receives heat from the cooling volume of the working fluid directly surrounding the heat generating components. The cooling volume is the area of ​​the working fluid (including both liquid and gas phases) that directly surrounds the heat generating components and is responsible for the convective cooling of the heat generating components. In some embodiments, the cooling volume is the volume of the working fluid within 5 millimeters (mm) of these heat generating components.

[0083] The immersion working fluid has a boiling temperature below a critical temperature at which the heat generating component undergoes thermal damage. The immersion working fluid can thereby receive heat from the heat generating component to cool the heat generating component before the heat generating component is damaged.

[0084] For example, the heat generating component may be a computing component that is subject to damage above 100°C. In some embodiments, the boiling temperature of the immersion working fluid is lower than the critical temperature of the heat generating component. In some embodiments, the boiling temperature of the immersion working fluid is less than about 90°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is less than about 80°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is less than about 70°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is less than about 60°C at 1 atmosphere. In some embodiments, the boiling temperature of the immersion working fluid is at least about 35°C at 1 atmosphere. In some embodiments, the working fluid comprises water.

[0085] In some embodiments, the working fluid comprises ethylene glycol. In some embodiments, the working fluid comprises a combination of water and ethylene glycol. In some embodiments, the working fluid comprises an aqueous solution. In some embodiments, the working fluid comprises an electronic liquid, such as FC-72 available from 3M, or a similar non-conductive fluid. In some embodiments, a heat generating component, support, or other element of an immersion cooling system positioned in the working fluid has nucleation sites on its surface that promote nucleation of vapor bubbles of the working fluid at or below the boiling temperature of the working fluid.

[0086] In some embodiments, sensors according to the present invention detect and / or measure the presence of one or more contaminants in an immersion working fluid. In some embodiments, the immersion working fluid is a two-phase working fluid that boils during operation of the immersion cooling system. In some embodiments, the immersion working fluid is a single-phase working fluid that is maintained in a liquid phase by operation of the immersion cooling system. Detecting the presence of contaminants in a single-phase immersion working fluid is beneficial to the operation of the single-phase immersion cooling system, but two-phase working fluids experience the additional consideration of the effect on the boiling temperature of the immersion working fluid and the increased deposition rate of contaminants on components in the immersion chamber due to boiling of the working fluid.

[0087] In some embodiments, an electrochemical sensor according to at least some embodiments of the present disclosure is configured to measure ionic contaminants in a solution immersed in a working fluid. The sensor includes a first electrode and a second electrode defining a sampling region therebetween. For example, a first sampling surface of the first electrode and a second sampling surface of the second electrode may be substantially parallel to each other with a constant electrode gap. In other examples, at least a portion of the first sampling surface of the first electrode and the second sampling surface of the second electrode are not parallel to each other, and the electrode gap is defined by a portion of the first sampling surface closest to the second sampling surface.

[0088] In some embodiments, the electrode gap is less than 100 microns. In some embodiments, the electrode gap is less than 50 microns. In some embodiments, the electrode gap is less than 20 microns. In some embodiments, the electrode gap is less than 10 microns. Smaller electrode gaps allow for lower voltages and / or amperages required across the sampling region, however, the immersion working fluid may include larger debris particles that may adversely affect the sensor if the larger particles enter the sampling region.

[0089] The sensor may also include a power source including a direct current (DC) source and an alternating current (AC) stimulus having variable amplitude and frequency. The sensor applies a current and a voltage selected from the DC source and the AC stimulus to a sample portion of the immersed working fluid in the sampling region.

[0090] A measurement system (configured to measure voltage and current) is connected across the sampling area. The measurement system includes a microcontroller configured to receive voltage measurements from a voltmeter and current measurements from an ammeter. The stimulation signal provided by the DC source and the AC stimulation that drives the electrochemical signal includes a static voltage coupled with a small amplitude swept sine wave. Different contaminants in the immersed working fluid will have different static voltages. By sweeping the stimulation signal provided by the DC source and the AC stimulation through a voltage range, the electrochemical response at and / or around the static voltage can be measured.

[0091] The voltage across the electrochemical cell and the current through it are measured during a frequency sweep. These signals are sampled by an analog-to-digital converter and stored in a memory as complex numbers with measured signal amplitude and phase. The resultant complex voltage and current measurements at each frequency point are processed to produce a complex impedance value at that frequency. The resulting two-dimensional impedance data points are processed, for example, by a microcontroller to convert the measured amplitude and phase into real and imaginary coordinates. These points form a profile.

[0092] In some embodiments, multiple profiles will be captured over time to track and / or correlate the evolution of the profile shape based on changes in concentration and / or stimulation signals. A machine learning (ML) model can be trained based on known concentrations of one or more contaminants to allow the ML model to identify or calculate the concentration of one or more contaminants and / or other contaminants in the immersion working fluid relative to the stimulation signal.

[0093] As used herein, a "machine learning model" refers to a computer algorithm or model (e.g., a classification model, a regression model, a language model, an object detection model) that can be adjusted (e.g., trained) to approximate an unknown function based on training inputs. For example, an ML model may refer to a neural network or other machine learning algorithm or architecture that learns and approximates complex functions and generates outputs based on multiple inputs provided to the machine learning model. In some embodiments, the ML system, model, or neural network described herein is an artificial neural network. In some embodiments, the ML system, model, or neural network described herein is a convolutional neural network. In some embodiments, the ML system, model, or neural network described herein is a recursive neural network. In at least one embodiment, the ML system, model, or neural network described herein is a Bayesian classifier. As used herein, a "machine learning system" may refer to one or more ML models that collaboratively generate one or more outputs based on corresponding inputs. For example, an ML system may refer to any system architecture having multiple discrete ML components that consider different kinds of information or inputs.

[0094] As used herein, an "instance" refers to an input object that can be provided as an input to an ML system for use in generating an output, such as a type of contaminant present, a concentration of a contaminant present, or a rate of change of a contaminant present. For example, an instance can refer to any event in which a contaminant is present. For example, a first profile or other measurement received at a microcontroller can be associated with a first contaminant present in a sample portion of the immersion working fluid. In another example, a second profile or other measurement received at the microcontroller can be associated with a second contaminant present in the sample portion of the immersion working fluid. In another example, a change in a profile or other measurement received at the microcontroller can be associated with an increase in the concentration of a contaminant in the sample portion of the immersion working fluid.

[0095] In some embodiments, the machine learning system has multiple layers, with an input layer configured to receive at least one input training data set or input training instances, and an output layer with multiple additional or hidden layers therebetween. The training data set can be input into the ML system to train the ML system and identify individual and combinations of contaminants or other attributes of the training instances that allow the microcontroller to identify concentrations of known or unknown contaminants.

[0096] In some embodiments, the machine learning system can receive multiple training data sets simultaneously and learn from different training data simultaneously.

[0097] In some embodiments, the machine learning system includes multiple machine learning models operating together. Each of the machine learning models has multiple hidden layers between an input layer and an output layer. The hidden layer has multiple input nodes, where each of the nodes operates on the input received from the previous layer. In a specific example, the first hidden layer has multiple nodes, and each of the nodes performs an operation on each instance from the input layer. Each node of the first hidden layer provides a new input to each node of the second hidden layer, and each node of the second hidden layer in turn performs a new operation on each of these inputs. The nodes of the second hidden layer then pass the output (such as the identified clusters) to the output layer.

[0098] In some embodiments, each of the nodes has a linear function and an activation function. The linear function can attempt to optimize or approximate a solution with a best fit line, such as reduced power cost or reduced latency. The activation function operates as a test for the effectiveness of the linear function. In some embodiments, the activation function produces a binary output that determines whether the output of the linear function is passed to the next layer of the machine learning model. In this way, the machine learning system can limit and / or prevent poor fits from propagating to the data and / or non-convergent solutions.

[0099] The machine learning model includes an input layer that receives at least one training data set. In some embodiments, at least one machine learning model uses supervised training. In some embodiments, at least one machine learning model uses unsupervised training. Unsupervised training can be used to draw inferences and find patterns or associations from (multiple) training data sets without known outputs. In some embodiments, unsupervised learning can identify clusters of similar labels or features for various training instances and allow the machine learning system to extrapolate the performance of instances with similar features.

[0100] In some embodiments, semi-supervised learning can combine the benefits from supervised learning and unsupervised learning. As described herein, a machine learning system can identify associated labels (such as known pollutants or pollutant concentrations) or features between instances, which can allow for the fusion of a training data set with known outputs and a second training data set that includes more general input information. Unsupervised training can allow a machine learning system to cluster instances from a second training data set without known outputs and associate the clusters with known outputs from a first training data set.

[0101] In some embodiments, the immersion cooling system includes electrochemical sensors positioned at various locations in the immersion cooling system. For example, a first electrochemical sensor can be located at a lower portion of an immersion chamber having a heat generating component. The first electrochemical sensor can be positioned to measure one or more contaminants present in a liquid working fluid in contact with the heat generating component. As described herein, a two-phase immersion working fluid presents an increased risk of dendrite growth during evaporation of the liquid working fluid in contact with the heat generating component. For at least this reason, it may be desirable to position the first electrochemical sensor to measure contaminants close to the heat generating component.

[0102] In another example, a second electrochemical sensor can be positioned near a liquid-vapor boundary between a liquid working fluid and a vapor working fluid. The second electrochemical sensor can measure the concentration of a contaminant at or near the top of the liquid working fluid in the immersion chamber to provide a relative concentration between the bottom and top near the first electrochemical sensor, for example, to help determine the source or source location of the contaminant in the liquid working fluid.

[0103] In yet another example, one or more third electrochemical sensors may be positioned proximate to a heat generating component, such as within a cooling volume proximate to heat generating component 614. One or more third electrochemical sensors may measure the presence of contaminants proximate to a heat source in the immersion chamber, and thereby measure the amount of contaminants present in a boiling portion of the liquid working fluid. As described herein, boiling of the liquid working fluid may increase the deposition rate of contaminant ions and / or the rate of dendrite growth on heat generating components (e.g., on pins or interconnects of a computing device). Measuring contaminants proximate to a heat generating component may provide information to a microcontroller to predict the risk of damage or failure of the heat generating component.

[0104] In another example, the fourth electrochemical sensor can be positioned in or near the condenser. In some embodiments, the condenser condenses the vapor working fluid into a condensed liquid working fluid in or in contact with the condenser. In at least one example, the liquid working fluid is converted into a vapor working fluid that does not contain pollutant ions when boiling, and the condensed liquid working fluid in or on the condenser does not contain pollutant ions unless at least some of the pollutant ions in the liquid working fluid originate from the condenser. Using the fourth electrochemical sensor to measure the pollutant concentration at or in the condenser can provide information related to the source or source location of the pollutant ions in the liquid working fluid to the microcontroller.

[0105] In yet another example, the fifth electrochemical sensor is positioned at a first end of a return conduit of the immersion cooling system, and / or the sixth electrochemical sensor is positioned at a second end of the return conduit. In some embodiments, the immersion cooling system delivers the condensed liquid working fluid from the condenser to the immersion chamber via the return conduit. By detecting contaminants and / or measuring contaminant levels at one or more locations in the return conduit, the fifth electrochemical sensor and / or the sixth electrochemical sensor can provide information related to the source or source location of contaminant ions in the liquid working fluid to the microcontroller.

[0106] In some embodiments, the microcontroller calculates the concentration of contaminants in the liquid working fluid at one or more locations in the immersion cooling system. The microcontroller or other control plane of the immersion cooling system can recommend or automatically perform remediation on the liquid working fluid to remove contaminants from the liquid working fluid and / or replace the contaminated liquid working fluid with new, uncontaminated working fluid.

[0107] In some embodiments, a method of measuring and remediating a liquid working fluid in an immersion cooling system includes positioning a sample portion of the immersion working fluid in a sampling region between a first electrode and a second electrode; and applying a stimulation signal having a static voltage across the sampling region using the first electrode and the second electrode. As described herein, the method includes varying the frequency of the input signal and measuring an output signal across the sampling region.

[0108] In some embodiments, the method is performed on-line while the immersion cooling system is operating. For example, the sample portion of the immersion working fluid may be a portion of the liquid working fluid in the immersion chamber, in the condenser, in the return conduit, or in other portions of the immersion cooling system. Sampling multiple times and / or at multiple locations in the immersion cooling system may allow the microcontroller to determine the concentration of the contaminant based at least in part on the measured output signal.

[0109] In some embodiments, the method optionally includes remediating a portion of the submerged working fluid. For example, remediating a portion of the submerged working fluid may include changing one or more filters in the submerged cooling system, such as near an inlet or outlet of a condenser, near an inlet or outlet of an immersion tank or chamber, near an inlet or outlet of a return conduit, or at other locations in the submerged cooling system.

[0110] In another example, remediating a portion of the immersion working fluid may include adding uncontaminated immersion working fluid to the immersion cooling system, such as opening a valve to add the uncontaminated immersion working fluid to the immersion cooling system. Adding the uncontaminated immersion working fluid to the immersion cooling system may dilute contaminant ions in the existing immersion working fluid, thereby reducing the effect of the contaminant ions on one or more properties of the immersion working fluid, such as conductivity, boiling temperature, or other properties. In other examples, diluting the contaminant ions in the immersion working fluid may reduce deposition rates and reduce dendrite growth. In at least one embodiment, adding the uncontaminated immersion working fluid to the immersion cooling system includes opening a valve to allow the uncontaminated immersion working fluid to flow from the reservoir through the valve into the immersion cooling system. For example, the valve may be controlled by a microcontroller, and the microcontroller may open the valve based at least in part on the measured output signal(s).

[0111] In yet another example, remediating a portion of the immersion working fluid may include replacing the contaminated immersion working fluid with uncontaminated immersion working fluid. Replacing the contaminated immersion working fluid with uncontaminated immersion working fluid may include opening at least a first valve to drain a portion of the contaminated immersion working fluid from the immersion cooling system, and opening at least a second valve to allow uncontaminated immersion working fluid to flow from the reservoir through the valve into the immersion cooling system. In at least one embodiment, the valves may be controlled by the microcontroller, and the microcontroller may open the valves based at least in part on the measured output signal(s).

[0112] In one example, the immersion cooling system may include a reservoir containing uncontaminated immersion working fluid and a collection tank configured to capture at least a portion of the contaminated liquid working fluid. In some embodiments, the reservoir is connected to a return conduit or other inlet to the immersion chamber to introduce the uncontaminated immersion working fluid into the immersion chamber. In other examples, the reservoir may be connected to the immersion chamber directly or through a dedicated conduit, or the reservoir may be connected to a condenser to deliver the uncontaminated immersion working fluid to the immersion cooling system via the condenser.

[0113] In some embodiments, the first valve controls the flow of uncontaminated immersion working fluid into the immersion cooling system. The first valve may be controlled by a microcontroller. In other embodiments, the first valve may be controlled by another control plane that receives contaminant and / or contaminant concentration information from the microcontroller.

[0114] The collection tank may be connected to the immersion cooling system via a second valve. In some embodiments, the collection tank is connected to a return conduit or other conduit of the immersion chamber to receive contaminated liquid working fluid 808 from the immersion chamber. In other examples, the collection tank may be connected to the immersion chamber directly or via a dedicated conduit, or the collection tank may be connected to a condenser to receive contaminated liquid working fluid from the immersion cooling system via the condenser.

[0115] In some embodiments, the second valve controls the flow of contaminated liquid working fluid into the immersion cooling system. The second valve can be controlled by a microcontroller. In other embodiments, the second valve can be controlled by another control plane that receives contaminant and / or contaminant concentration information from the microcontroller. In at least one embodiment, the microcontroller or other control plane at least partially opens the first valve and the second valve simultaneously (e.g., the time period when the first valve is open at least partially overlaps the time period when the second valve 866 is open) to introduce uncontaminated immersion working fluid from the reservoir into the immersion cooling system while removing at least a portion of the contaminated liquid working fluid into the collection tank.

[0116] The present disclosure relates to systems and methods for cooling heat generating components of a computer or computing device according to at least the examples provided in the following sections:

[0117] [A1] In some embodiments, an immersion cooling system includes an immersion tank defining an immersion chamber therein, an immersion working fluid, a first electrode, and a second electrode. The immersion working fluid is at least partially positioned in the immersion chamber. The first electrode is electrically coupled to a power source, the second electrode is positioned proximate to the first electrode and defines a sampling region therebetween. A sample portion of the immersion working fluid is positioned in the sampling region, and the second electrode is coupled to a microcontroller, the microcontroller being configured to at least measure a current across the sampling region between the first electrode and the second electrode.

[0118] [A2] In some embodiments, the power source of [A1] includes a direct current (DC) source and an alternating current (AC) stimulus.

[0119] [A3] In some embodiments, the power supply of [A1] or [A2] provides a stimulation signal to the first electrode, and the stimulation signal includes a static voltage.

[0120] [A4] In some embodiments, the stimulation signal of [A3] also includes a swept frequency sine wave.

[0121] [A5] In some embodiments, the microcontroller of [A4] records multiple impedance data points

[0122] [A6] In some embodiments, the first electrode and the second electrode of any one of [A1] to [A5] are located in an immersion chamber.

[0123] [A7] In some embodiments, the first electrode and the second electrode of any one of [A1] to [A5] are located in a capacitor.

[0124] [A8] In some embodiments, the first electrode and the second electrode of any one of [A1] to [A5] are located in a return conduit connected to the immersion chamber.

[0125] [A9] In some embodiments, the first electrode and the second electrode of any one of [A1] to [A5] are located in a cooling volume proximate to the heat generating component.

[0126] [A10] In some embodiments, the first electrode and the second electrode of any one of [A1] to [A9] form a first electrochemical sensor, and the system also includes a second electrochemical sensor positioned in the immersion chamber and in communication with the microcontroller.

[0127] [B1] In some embodiments, a method for measuring a liquid working fluid in an immersion cooling system includes: circulating an immersion working fluid in the immersion cooling system; positioning a sample portion of the immersion working fluid in a sampling region between a first electrode and a second electrode; applying a stimulation signal having a static voltage across the sampling region using the first electrode and the second electrode; changing a frequency of the stimulation signal; measuring an output signal across the sampling region; and determining a first concentration of a contaminant based at least in part on the output signal.

[0128] [B2] In some embodiments, determining the concentration of the pollutant of [B1] includes inputting the measured output signal into a machine learning (ML) model.

[0129] [B3] In some embodiments, changing the frequency of the input signal of [B1] or [B2] includes changing a DC signal with an AC stimulus.

[0130] [B4] In some embodiments, positioning the sample portion of the immersion working fluid in the sampling region between the first electrode and the second electrode of any one of [B1] to [B3] includes flowing the sample portion of the immersion working fluid through a return conduit.

[0131] [B5] In some embodiments, circulating the immersion working fluid in the immersion cooling system of any one of [B1] to [B4] includes evaporating at least a portion of the immersion working fluid.

[0132] [B6] In some embodiments, the first electrode and the second electrode of any one of [B1] to [B5] are a first electrochemical sensor at a first position in the immersion cooling system, and the method also includes measuring a second output signal with a second electrochemical sensor at a second position in the immersion cooling system.

[0133] [B7] In some embodiments, the method of [B6] also includes determining a second concentration of the pollutant based at least in part on the second output signal.

[0134] [C1] In some embodiments, a method for measuring and remediating a liquid working fluid in an immersion cooling system includes: circulating an immersion working fluid in the immersion cooling system; positioning a sample portion of the immersion working fluid in a sampling region between a first electrode and a second electrode; applying a stimulation signal having a static voltage across the sampling region using the first electrode and the second electrode; varying a frequency of the stimulation signal; measuring an output signal across the sampling region; determining a concentration of a contaminant based at least in part on the output signal using a microcontroller in communication with the first electrode and the second electrode; and remediating the immersion working fluid based at least in part on the concentration of the contaminant using the microcontroller.

[0135] [C2] In some embodiments, remediating the immersion working fluid of [C1] includes opening a valve to add uncontaminated immersion working fluid to the immersion cooling system.

[0136] [C3] In some embodiments, remediating the submerged working fluid of [C1] or [C2] includes opening a valve to remove at least a portion of the submerged working fluid from the submerged cooling system.

[0137] The articles "a" and "an", "the" are intended to indicate the presence of one or more elements in the preceding description. The terms "comprising", "including" and "having" are intended to be inclusive, and mean that there may be additional elements in addition to the listed elements. Additionally, it should be understood that reference to "one embodiment" or "embodiment" of the present disclosure is not intended to be interpreted as excluding the presence of additional embodiments that also incorporate the features. For example, any element described with respect to the embodiments herein may be combined with any element of any other embodiment described herein. The numbers, percentages, ratios or other values ​​described herein are intended to include the value, as well as other values ​​of "approximately", "substantially" or "approximately" the value, as will be understood by those of ordinary skill in the art covered by the embodiments of the present disclosure. Therefore, the value should be interpreted broadly enough to include values ​​that are at least close enough to the value to perform the desired function or achieve the desired result. The value includes at least the expected variation in a suitable manufacturing or production process, and may include values ​​within 5%, 1%, 0.1% or 0.01% of the value.

[0138] In view of the present disclosure, those of ordinary skill in the art should recognize that equivalent constructions do not depart from the spirit and scope of the present disclosure, and that various changes, substitutions, and modifications may be made to the embodiments disclosed herein without departing from the spirit and scope of the present disclosure. Equivalent constructions, including functional "means plus function" clauses, are intended to cover structures described herein that perform the functions described, including structural equivalents that operate in the same manner and equivalent structures that provide the same functions. It is the express intent of the applicant not to invoke means-plus-function or other functional claims for any claim, except those in which the words "means for..." appear together with the associated function. Each addition, deletion, and modification to the embodiments that falls within the meaning and scope of the claims will be encompassed by the claims.

[0139] It should be understood that any directions or reference frames in the foregoing description are only relative directions or movements. For example, any reference to "front" and "back", or "top" and "bottom", or "left" and "right" only describes the relative positions or movements of the relevant elements.

[0140] The present disclosure may be implemented in other specific forms without departing from its spirit or features. The described embodiments should be considered illustrative rather than restrictive. Therefore, the scope of the present disclosure is indicated by the appended claims rather than by the preceding description. Changes within the meaning and scope of the equivalents of the claims will be included within their scope.

Claims

1. An immersion cooling system comprising: an immersion tank defining an immersion chamber therein; an immersion working fluid positioned at least partially within the immersion chamber; a first electrode electrically coupled to a power source; as well as a second electrode positioned proximate to the first electrode and defining a sampling region between the second electrode and the first electrode, wherein: The sample portion of the immersion working fluid is positioned in the sampling region, and The second electrode is coupled to a microcontroller configured to at least measure a current across the sampling region between the first electrode and the second electrode.

2. The immersion cooling system of claim 1, wherein the power source comprises a direct current (DC) source and an alternating current (AC) stimulus.

3. An immersion cooling system according to claim 1 or 2, wherein the power supply provides a stimulation signal to the first electrode, and the stimulation signal comprises a static voltage.

4. The immersion cooling system of claim 3, wherein the stimulation signal further comprises a swept frequency sine wave.

5. The immersion cooling system of claim 4, wherein the microcontroller records a plurality of impedance data points.

6. An immersion cooling system according to any one of the preceding claims, wherein the first electrode and the second electrode are located in the immersion chamber.

7. The immersion cooling system of any one of claims 1 to 5, wherein the first electrode and the second electrode are located in a condenser.

8. An immersion cooling system according to any one of claims 1 to 5, wherein the first electrode and the second electrode are located in a return conduit connected to the immersion chamber.

9. The immersion cooling system of any one of claims 1 to 5, wherein the first electrode and the second electrode are located in a cooling volume proximate to a heat generating component.

10. An immersion cooling system according to any one of the preceding claims, wherein the first electrode and the second electrode form a first electrochemical sensor; and The immersion cooling system also includes a second electrochemical sensor positioned in the immersion chamber and in communication with the microcontroller.

11. A method of measuring a liquid working fluid in an immersion cooling system, the method comprising: circulating an immersion working fluid in the immersion cooling system; positioning a sample portion immersed in the working fluid in a sampling region between the first electrode and the second electrode; applying a stimulation signal having a static voltage across a sampling region using the first electrode and the second electrode; changing the frequency of the stimulation signal; measuring an output signal across the sampling region; as well as A first concentration of a contaminant is determined based at least in part on the output signal.

12. The method of claim 11, wherein determining the concentration of the pollutant comprises inputting the measured output signal into a machine learning (ML) model.

13. The method of claim 11 or 12, wherein varying the frequency of the input signal comprises varying a DC signal with an AC stimulus.

14. The method of any one of claims 11 to 13, wherein positioning a sample portion of the immersion working fluid in a sampling region between a first electrode and a second electrode comprises flowing the sample portion of the immersion working fluid through a return conduit.

15. The method of any one of claims 11 to 14, wherein circulating the immersion working fluid in the immersion cooling system comprises evaporating at least a portion of the immersion working fluid.

Images