Electronics may be required to operate in harsh environments in automotive, aerospace, and defense applications. Solder interconnects in harsh environments may be subjected to extremely-low and high temperatures in the range of −65C to +200C in conjunction with significant strain loads. Furthermore, electronic assemblies may be subjected to extended periods of non-climate-controlled storage prior to deployment. Prior studies have shown that lead-free solder materials continue to evolve under varied thermal loads, leading to deterioration in mechanical parameters such as Ultimate Tensile Strength and Elastic Modulus. The material characteristics for non-linear modeling and reliability prediction are required for risk minimization with the use of new alloy formulations in high-reliability applications. The current work fills this state-of-the-art gap by measuring the mechanical characteristics of undoped SAC105 and doped SAC-Q solder alloys at low operation temperatures (−65°C to 0°C) at high strain rate after varied thermal aging periods up to one year. In addition, the evolution of Anand parameters for SAC solder alloys after prolonged thermal aging has been studied. The Anand model’s reliability has been assessed by comparing experimentally observed data with forecasted data using determined model constants for both solder alloys. The Anand parameters were applied in a FE-framework to simulate drop events for a ball-grid array package on a printed circuit board assembly.
{"title":"Effect of Property Evolution of Doped and Undoped SnAgCu Solder Alloys Under Shock and Vibration","authors":"P. Lall, Vikas Yadav, J. Suhling, David Locker","doi":"10.1115/ipack2022-97452","DOIUrl":"https://doi.org/10.1115/ipack2022-97452","url":null,"abstract":"\u0000 Electronics may be required to operate in harsh environments in automotive, aerospace, and defense applications. Solder interconnects in harsh environments may be subjected to extremely-low and high temperatures in the range of −65C to +200C in conjunction with significant strain loads. Furthermore, electronic assemblies may be subjected to extended periods of non-climate-controlled storage prior to deployment. Prior studies have shown that lead-free solder materials continue to evolve under varied thermal loads, leading to deterioration in mechanical parameters such as Ultimate Tensile Strength and Elastic Modulus. The material characteristics for non-linear modeling and reliability prediction are required for risk minimization with the use of new alloy formulations in high-reliability applications. The current work fills this state-of-the-art gap by measuring the mechanical characteristics of undoped SAC105 and doped SAC-Q solder alloys at low operation temperatures (−65°C to 0°C) at high strain rate after varied thermal aging periods up to one year. In addition, the evolution of Anand parameters for SAC solder alloys after prolonged thermal aging has been studied. The Anand model’s reliability has been assessed by comparing experimentally observed data with forecasted data using determined model constants for both solder alloys. The Anand parameters were applied in a FE-framework to simulate drop events for a ball-grid array package on a printed circuit board assembly.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124762765","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
S. Karajgikar, Jasper Kidger, A. Shaw, N. Edmunds, V. Mulay
In this paper, a standard air-cooled high-density storage system is reengineered to demonstrate the use of single-phase immersion cooling. The storage system consists primarily of seventy-two hard drives, two single socket nodes, two SAS expander cards, NIC and a power distribution board in a 4OU form factor. It is successfully demonstrated that the storage systems can be designed to support single phase immersion cooling while supporting hot swap and cooling redundancy requirement like an air-cooled system. In an air-cooled system, temperature gradient between the drives was as high as 19°C. The drives placed at the front of the system received cooler air while the drives placed in the rear received preheated air thus resulting in a temperature gradient. The drives used for the study were 20GB Helium filled sealed drives. For immersion cooling, seventy-two drives were cooled in parallel with temperature variance of less than 3°C. The other system components such as CPU, DIMMs, SAS chip and NIC had sufficient thermal margin. It was demonstrated that the system can operate reliably for facility coolant supply temperature as high as 40°C. The resulting power consumption of the pump was less than five percent of the total IT power. In addition, the proposed cooling solution may help mitigate acoustic vibrational issues for drives often encountered in air-cooling solution. The solution is virtually silent in operation.
{"title":"Single-Phase Immersion Cooling Study of a High-Density Storage System","authors":"S. Karajgikar, Jasper Kidger, A. Shaw, N. Edmunds, V. Mulay","doi":"10.1115/ipack2022-97490","DOIUrl":"https://doi.org/10.1115/ipack2022-97490","url":null,"abstract":"\u0000 In this paper, a standard air-cooled high-density storage system is reengineered to demonstrate the use of single-phase immersion cooling. The storage system consists primarily of seventy-two hard drives, two single socket nodes, two SAS expander cards, NIC and a power distribution board in a 4OU form factor. It is successfully demonstrated that the storage systems can be designed to support single phase immersion cooling while supporting hot swap and cooling redundancy requirement like an air-cooled system. In an air-cooled system, temperature gradient between the drives was as high as 19°C. The drives placed at the front of the system received cooler air while the drives placed in the rear received preheated air thus resulting in a temperature gradient. The drives used for the study were 20GB Helium filled sealed drives. For immersion cooling, seventy-two drives were cooled in parallel with temperature variance of less than 3°C. The other system components such as CPU, DIMMs, SAS chip and NIC had sufficient thermal margin. It was demonstrated that the system can operate reliably for facility coolant supply temperature as high as 40°C. The resulting power consumption of the pump was less than five percent of the total IT power. In addition, the proposed cooling solution may help mitigate acoustic vibrational issues for drives often encountered in air-cooling solution. The solution is virtually silent in operation.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130557563","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
For many years, traditional rigid PCBs have been utilized in many applications and have been shown to be reliable in a number of applications. There hasn’t been much research on flexible electronics device attachment techniques and process reliability. There has been prior research done on copper sintering using various methodologies, but there is no prediction model to apply the underlying information directly for predicting the performance of the printed electronics. The sintering process determines the electrical performance of printed traces, and it is necessary to comprehend and estimate the resistivity of printed traces. This study developed a regression model based on an Artificial Neural Network (ANN) to predict resistivity. Because flexible substrates allow for more flexibility, it is critical to create a reliable way of attaching components to circuits that can endure various motions. Micro dispensing equipment was employed in this investigation to print conductive traces, an electrically conductive adhesive (ECA), and low-temperature solder (LTS) for component attachment pads. There is little understanding of SMD attachments’ behavior on additively printed flexible substrates, and we examined different aspects of their performance in this study.
{"title":"Influence of Cure-Reflow Profile and High-Temperature Operation of Additively Printed Conductive Circuits on Performance and Reliability","authors":"P. Lall, Jinesh Narangaparambil, C. Hill","doi":"10.1115/ipack2022-97457","DOIUrl":"https://doi.org/10.1115/ipack2022-97457","url":null,"abstract":"\u0000 For many years, traditional rigid PCBs have been utilized in many applications and have been shown to be reliable in a number of applications. There hasn’t been much research on flexible electronics device attachment techniques and process reliability. There has been prior research done on copper sintering using various methodologies, but there is no prediction model to apply the underlying information directly for predicting the performance of the printed electronics. The sintering process determines the electrical performance of printed traces, and it is necessary to comprehend and estimate the resistivity of printed traces. This study developed a regression model based on an Artificial Neural Network (ANN) to predict resistivity. Because flexible substrates allow for more flexibility, it is critical to create a reliable way of attaching components to circuits that can endure various motions. Micro dispensing equipment was employed in this investigation to print conductive traces, an electrically conductive adhesive (ECA), and low-temperature solder (LTS) for component attachment pads. There is little understanding of SMD attachments’ behavior on additively printed flexible substrates, and we examined different aspects of their performance in this study.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"29 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131348638","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Thermal predictions in data centers have been utilized to reduce the electric consumption of thermal equipment in data centers. While most of the optimization of data center temperature has been performed through the utilization of Computational Fluid Dynamics (CFD) and heuristic methods, data driven modeling techniques are now also being used to optimize the data center temperatures. Some data driven models have been used on a static data set to obtain the steady state temperature predictions for given input variables while other data driven models have been trained to provide temperature predictions at live time. This paper aims to investigate the transient temperature prediction capabilities of two data driven models — Long-Short Term Memory (LSTM) and Nonlinear Autoregressive Neural Network with External Input (NARX). While these two methods have been previously studied on data center applications, they have not been compared with each other for transient temperature predictions for normal operations. The study also utilizes ensembles to provide better temperature prediction accuracy for smaller data sets. The study compared these two models based on an experimentally obtained data set and found that NARX outperforms LSTM for normal operations and that the data driven models are able to provide relatively good predictions even if the input variables are slightly outside the training domain.
{"title":"Data Driven Modeling Advancements for Thermal Predictions in Data Center Applications","authors":"D. Patel, Y. Joshi","doi":"10.1115/ipack2022-97478","DOIUrl":"https://doi.org/10.1115/ipack2022-97478","url":null,"abstract":"\u0000 Thermal predictions in data centers have been utilized to reduce the electric consumption of thermal equipment in data centers. While most of the optimization of data center temperature has been performed through the utilization of Computational Fluid Dynamics (CFD) and heuristic methods, data driven modeling techniques are now also being used to optimize the data center temperatures. Some data driven models have been used on a static data set to obtain the steady state temperature predictions for given input variables while other data driven models have been trained to provide temperature predictions at live time.\u0000 This paper aims to investigate the transient temperature prediction capabilities of two data driven models — Long-Short Term Memory (LSTM) and Nonlinear Autoregressive Neural Network with External Input (NARX). While these two methods have been previously studied on data center applications, they have not been compared with each other for transient temperature predictions for normal operations. The study also utilizes ensembles to provide better temperature prediction accuracy for smaller data sets. The study compared these two models based on an experimentally obtained data set and found that NARX outperforms LSTM for normal operations and that the data driven models are able to provide relatively good predictions even if the input variables are slightly outside the training domain.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"17 Suppl 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131091182","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Modern data centers, which consume roughly 3% of global electricity, continue to experience increased demand. Therefore, green data centers that consume less energy and have minimal environmental impact are desirable. This study examines the potential energy savings and environmental benefits of applying airside economization with evaporative cooling in air-cooled data centers in the continental U.S. A generic data center that employs a Computer Room Air Conditioning (CRAC)-based cooling system with a total IT load of 400 kW is modeled at 925 locations using the National Renewable Energy Laboratory’s (NREL’s) TMY3 database. The energy savings and environmental benefits are evaluated in terms of key data center performance metrics: Power Usage Effectiveness (PUE), Carbon Usage Effectiveness (CUE), and the recently proposed Water Scarcity Usage Effectiveness (WSUE) metric, which quantifies the holistic impact of water consumption on regional water availability. Results are aggregated and analyzed at the U.S. State level. It is found that airside economization implementation in the continental U.S. is feasible 6.57% and 21.5% of the year on average based on ASHRAE recommended and allowable envelopes, respectively. Furthermore, results indicate that carbon footprint and water scarcity footprint can be reduced by up to 16% when economization is implemented based on the ASHRAE allowable envelope.
{"title":"Predictions of Airside Economization-Based Air-Cooled Data Center Environmental Burden Reduction","authors":"Li Chen, A. Wemhoff","doi":"10.1115/ipack2022-92005","DOIUrl":"https://doi.org/10.1115/ipack2022-92005","url":null,"abstract":"\u0000 Modern data centers, which consume roughly 3% of global electricity, continue to experience increased demand. Therefore, green data centers that consume less energy and have minimal environmental impact are desirable. This study examines the potential energy savings and environmental benefits of applying airside economization with evaporative cooling in air-cooled data centers in the continental U.S. A generic data center that employs a Computer Room Air Conditioning (CRAC)-based cooling system with a total IT load of 400 kW is modeled at 925 locations using the National Renewable Energy Laboratory’s (NREL’s) TMY3 database. The energy savings and environmental benefits are evaluated in terms of key data center performance metrics: Power Usage Effectiveness (PUE), Carbon Usage Effectiveness (CUE), and the recently proposed Water Scarcity Usage Effectiveness (WSUE) metric, which quantifies the holistic impact of water consumption on regional water availability. Results are aggregated and analyzed at the U.S. State level. It is found that airside economization implementation in the continental U.S. is feasible 6.57% and 21.5% of the year on average based on ASHRAE recommended and allowable envelopes, respectively. Furthermore, results indicate that carbon footprint and water scarcity footprint can be reduced by up to 16% when economization is implemented based on the ASHRAE allowable envelope.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127963637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Passive cooling techniques are widely sought-after solutions to thermal management issues in high power electronics due to increased energy dissipation in reduced areas. Phase change materials (PCMs) present a promising secondary passive thermal management opportunity by absorbing a large amount of energy as an isothermal process. This phenomenon can be utilized in various ways as a thermal management tool; including temperature spike alleviation, energy storage, and secondary passive cooling. Though PCMs have promising passive cooling ability, often it is difficult to select an appropriate or effective PCM for the specific application due to deficiencies in a particular material property. Previous studies have demonstrated the ability to alter PCM properties through the homogeneous inclusion of nanoparticles. Thermal conductivity is a particularly important metric for enhancement via nanoparticles due to the typically low conductivity of PCMs with high latent heats. Previous studies demonstrate the successful augmentation of this property. A large limiting factor to enhanced PCM passive cooling is related to the propagation of the melt front, representing the region of large energy absorption. In many cases, the melt front moves too slowly to effectively transfer energy away from the device. Slow material response time can also be problematic in the re-solidification process, limiting cyclability. Work has been conducted to monitor the melt front response to a thermal load. Early in the melting process, conduction dominates the heat transfer mechanism. This paper will examine the impact of nanoparticle inclusion as a means of controlling the melt front propagation. Using nanoparticles to control the composite thermal conductivity should lead to optimization ability of PCM melt characteristics to align with thermal management needs.
{"title":"Melt Front Enhancement of Phase Change Materials via Nanoparticle Inclusion for Improved Heat Transfer and Cyclability","authors":"Joshua Kasitz, L. Marshall, D. Huitink","doi":"10.1115/ipack2022-97399","DOIUrl":"https://doi.org/10.1115/ipack2022-97399","url":null,"abstract":"\u0000 Passive cooling techniques are widely sought-after solutions to thermal management issues in high power electronics due to increased energy dissipation in reduced areas. Phase change materials (PCMs) present a promising secondary passive thermal management opportunity by absorbing a large amount of energy as an isothermal process. This phenomenon can be utilized in various ways as a thermal management tool; including temperature spike alleviation, energy storage, and secondary passive cooling. Though PCMs have promising passive cooling ability, often it is difficult to select an appropriate or effective PCM for the specific application due to deficiencies in a particular material property. Previous studies have demonstrated the ability to alter PCM properties through the homogeneous inclusion of nanoparticles. Thermal conductivity is a particularly important metric for enhancement via nanoparticles due to the typically low conductivity of PCMs with high latent heats. Previous studies demonstrate the successful augmentation of this property. A large limiting factor to enhanced PCM passive cooling is related to the propagation of the melt front, representing the region of large energy absorption. In many cases, the melt front moves too slowly to effectively transfer energy away from the device. Slow material response time can also be problematic in the re-solidification process, limiting cyclability. Work has been conducted to monitor the melt front response to a thermal load. Early in the melting process, conduction dominates the heat transfer mechanism. This paper will examine the impact of nanoparticle inclusion as a means of controlling the melt front propagation. Using nanoparticles to control the composite thermal conductivity should lead to optimization ability of PCM melt characteristics to align with thermal management needs.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126719518","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Potting is one of the most effective techniques for safeguarding electronics assembly in challenging harsh conditions, including shock and vibration. Interconnect failures are often preceded by delamination at the PCB and epoxy interface. PCB-Epoxy interfaces have not been extensively researched for interfacial fracture resistance under high thermo-mechanical loading. In this study, bi-material PCB-epoxy samples are made and exposed to long-term high-temperature aging followed by monotonic four-point bend loading. The evolution of the interfacial integrity under sustained high-temperature exposure has been quantified. The study looks at five distinct types of potting materials with varying properties. The specimens are exposed to a high temperature of 100°C and 150°C for 30 days, 60 days, 90 days, 120 days, 180 days, 240 days, and 360 days. Steady-state strain energy release rate, mode-I (KI), and mode-II (KII) stress intensity factors are determined for the PCB-Epoxy interface. Cohesive zone parameters for each of the PCB-Epoxy interfaces have been determined and implemented into a predictive cohesive zone model (CZM). The PCB-Epoxy bi-material specimen has been modeled in ABAQUS with a cohesive zone at the interface and subjected to mode-I four-point bend loading. Damage is considered to occur at the interface where the cohesive zone has been modeled. For both pristine and aged tests, the damage accumulation is predicted using the interfacial fracture parameters from the experiment.
{"title":"Predictive Cohesive Zone Modeling for Delamination at PCB-Potting Material Interfaces Under Four-Point Bend Loading With Sustained High-Temperature Exposure","authors":"P. Lall, A. Pandurangan, K. Blecker","doi":"10.1115/ipack2022-97427","DOIUrl":"https://doi.org/10.1115/ipack2022-97427","url":null,"abstract":"\u0000 Potting is one of the most effective techniques for safeguarding electronics assembly in challenging harsh conditions, including shock and vibration. Interconnect failures are often preceded by delamination at the PCB and epoxy interface. PCB-Epoxy interfaces have not been extensively researched for interfacial fracture resistance under high thermo-mechanical loading. In this study, bi-material PCB-epoxy samples are made and exposed to long-term high-temperature aging followed by monotonic four-point bend loading. The evolution of the interfacial integrity under sustained high-temperature exposure has been quantified. The study looks at five distinct types of potting materials with varying properties. The specimens are exposed to a high temperature of 100°C and 150°C for 30 days, 60 days, 90 days, 120 days, 180 days, 240 days, and 360 days. Steady-state strain energy release rate, mode-I (KI), and mode-II (KII) stress intensity factors are determined for the PCB-Epoxy interface. Cohesive zone parameters for each of the PCB-Epoxy interfaces have been determined and implemented into a predictive cohesive zone model (CZM). The PCB-Epoxy bi-material specimen has been modeled in ABAQUS with a cohesive zone at the interface and subjected to mode-I four-point bend loading. Damage is considered to occur at the interface where the cohesive zone has been modeled. For both pristine and aged tests, the damage accumulation is predicted using the interfacial fracture parameters from the experiment.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116712993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xiaoqiang Xu, A. Mirza, Lingfeng Gao, F. Luo, Shikui Chen
This paper proposes a density-based topology optimization scheme to design a heat sink for the application of a 3D integrated SIC-based 75 kVA Intelligent Power Stage (IPS). The heat sink design considers the heat conduction and convection effects with forced air cooling. The objective function is to minimize the thermal compliance of the whole structure. A volume constraint is imposed to reduce the overall volume of the designed heat sink to make it conformal to the underlying power devices. Some numerical techniques like filtering and projection schemes are employed to render a crisp design. Some 2D benchmarks examples are first provided to demonstrate the effectiveness of the proposed method. Then a 3D heat sink, especially designed for the 3D IPS, is topologically optimized. The classic tree-like structure is reproduced to reinforce the convection effect. Some comparisons with the intuitive baseline designs are made through numerical simulation. The optimized heat sinks are shown to provide a more efficient cooling performance for the 3D integrated power converter assembly.
{"title":"Topology Optimization of Heat Sink for 3d Integrated Power Converters","authors":"Xiaoqiang Xu, A. Mirza, Lingfeng Gao, F. Luo, Shikui Chen","doi":"10.1115/ipack2022-98066","DOIUrl":"https://doi.org/10.1115/ipack2022-98066","url":null,"abstract":"\u0000 This paper proposes a density-based topology optimization scheme to design a heat sink for the application of a 3D integrated SIC-based 75 kVA Intelligent Power Stage (IPS). The heat sink design considers the heat conduction and convection effects with forced air cooling. The objective function is to minimize the thermal compliance of the whole structure. A volume constraint is imposed to reduce the overall volume of the designed heat sink to make it conformal to the underlying power devices. Some numerical techniques like filtering and projection schemes are employed to render a crisp design. Some 2D benchmarks examples are first provided to demonstrate the effectiveness of the proposed method. Then a 3D heat sink, especially designed for the 3D IPS, is topologically optimized. The classic tree-like structure is reproduced to reinforce the convection effect. Some comparisons with the intuitive baseline designs are made through numerical simulation. The optimized heat sinks are shown to provide a more efficient cooling performance for the 3D integrated power converter assembly.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124468261","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Heydari, Pardeep Shahi, Vahideh Radmard, Bahareh Eslami, Uschas Chowdhury, S. Saini, Pratik V. Bansode, Harold Miyamura, D. Agonafer, Jeremy Rodriguez
Removal of heat is becoming a major challenge in today’s data centers. Computing-intensive applications such as artificial intelligence and machine learning are pushing data center to compute intensive systems, such as GPU, CPU, and switches to their extreme limits. Racks of IT can approach up to 100kW of heat dissipation challenging traditional data center designs for enterprises and cloud service providers. Direct-to-chip liquid cooling utilizing cold plates is becoming a common method of removing heat from high heat density data center server racks. There are various methods of applying liquid cooling to data centers to address the high heat density components such as liquid to liquid (L2L), liquid to air (L2A), and liquid to single phase refrigerant (L2R). This study aims to investigate the thermo-hydraulic performance of the L2L cooling systems using cooling distribution units (CDUs). CDUs provide a cold secondary coolant (Propylene Glycol 25%) into the cooling loops of liquid-cooled server racks, with the CDUs providing liquid to liquid heat exchange between the primary facility water and secondary liquid used for cold plates. This study uses Thermal Test Vehicles (TTVs) which have been built to reproduce and simulate high heat density servers. Four different cooling loops are characterized experimentally, and detailed analytical and numerical simulations using CFD are developed for analyzing the cooling characteristics of the entire L2L cooling loop, including the CDU, for removing heat from the cold plates. Detailed Flow Network Modeling (FNM) has been performed to analyze precise hydraulic modeling of the secondary fluid flow, from the CDUs to the cooling loops, for predicting pressure drop and flow rate of the secondary coolant. A FNM properly sizes the pumping requirements of the L2L cooling system. Additionally, a system calculator has been created for quickly sizing all secondary loop piping for L2L heat exchanger deployments.
{"title":"Liquid to Liquid Cooling for High Heat Density Liquid Cooled Data Centers","authors":"A. Heydari, Pardeep Shahi, Vahideh Radmard, Bahareh Eslami, Uschas Chowdhury, S. Saini, Pratik V. Bansode, Harold Miyamura, D. Agonafer, Jeremy Rodriguez","doi":"10.1115/ipack2022-97416","DOIUrl":"https://doi.org/10.1115/ipack2022-97416","url":null,"abstract":"\u0000 Removal of heat is becoming a major challenge in today’s data centers. Computing-intensive applications such as artificial intelligence and machine learning are pushing data center to compute intensive systems, such as GPU, CPU, and switches to their extreme limits. Racks of IT can approach up to 100kW of heat dissipation challenging traditional data center designs for enterprises and cloud service providers. Direct-to-chip liquid cooling utilizing cold plates is becoming a common method of removing heat from high heat density data center server racks. There are various methods of applying liquid cooling to data centers to address the high heat density components such as liquid to liquid (L2L), liquid to air (L2A), and liquid to single phase refrigerant (L2R). This study aims to investigate the thermo-hydraulic performance of the L2L cooling systems using cooling distribution units (CDUs). CDUs provide a cold secondary coolant (Propylene Glycol 25%) into the cooling loops of liquid-cooled server racks, with the CDUs providing liquid to liquid heat exchange between the primary facility water and secondary liquid used for cold plates. This study uses Thermal Test Vehicles (TTVs) which have been built to reproduce and simulate high heat density servers. Four different cooling loops are characterized experimentally, and detailed analytical and numerical simulations using CFD are developed for analyzing the cooling characteristics of the entire L2L cooling loop, including the CDU, for removing heat from the cold plates. Detailed Flow Network Modeling (FNM) has been performed to analyze precise hydraulic modeling of the secondary fluid flow, from the CDUs to the cooling loops, for predicting pressure drop and flow rate of the secondary coolant. A FNM properly sizes the pumping requirements of the L2L cooling system. Additionally, a system calculator has been created for quickly sizing all secondary loop piping for L2L heat exchanger deployments.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124556772","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Himanshu Modi, Pardeep Shahi, Akiilessh Sivakumar, S. Saini, Pratik V. Bansode, Vibin Shalom, Amruthavalli Rachakonda, Gautam Gupta, D. Agonafer
Direct to chip liquid-cooling technique has been widely implemented for the cooling of processors with high thermal design power. To further enhance the efficiency of liquid cooling, ongoing research focuses on the optimizations at the cold plate level or by changing the flow configurations. But in all cases, the coolant which is pumped across the rack is pumped at a constant flow rate irrespective of the workload utilization at the individual server, resulting in excess pumping of coolant. A practical approach is to dynamically vary the flow rate to each server as per server workload utilization. In this study, transient analysis is performed by varying the flow rate across individual servers at rack level using CFD. A CFD model mimicking four servers placed at different heights on a standard 42U rack is developed. The flow variation through each of the servers is done using a damper arrangement representing a flow control device. A controller is integrated to automate the process of opening and closing of FCD to vary the flow based on the average outlet temperature from each server. A baseline simulation with all servers running at maximum power dissipation with a constant coolant flow rate is compared with cases where the coolant flow rate is controlled dynamically for varying thermal design power (TDP). The results shown analyze the impact of the dynamic response of the flow control device on transient thermal and hydraulic characteristics across the rack is done.
{"title":"Transient CFD Analysis of Dynamic Liquid-Cooling Implementation at Rack Level","authors":"Himanshu Modi, Pardeep Shahi, Akiilessh Sivakumar, S. Saini, Pratik V. Bansode, Vibin Shalom, Amruthavalli Rachakonda, Gautam Gupta, D. Agonafer","doi":"10.1115/ipack2022-97443","DOIUrl":"https://doi.org/10.1115/ipack2022-97443","url":null,"abstract":"\u0000 Direct to chip liquid-cooling technique has been widely implemented for the cooling of processors with high thermal design power. To further enhance the efficiency of liquid cooling, ongoing research focuses on the optimizations at the cold plate level or by changing the flow configurations. But in all cases, the coolant which is pumped across the rack is pumped at a constant flow rate irrespective of the workload utilization at the individual server, resulting in excess pumping of coolant. A practical approach is to dynamically vary the flow rate to each server as per server workload utilization. In this study, transient analysis is performed by varying the flow rate across individual servers at rack level using CFD. A CFD model mimicking four servers placed at different heights on a standard 42U rack is developed. The flow variation through each of the servers is done using a damper arrangement representing a flow control device. A controller is integrated to automate the process of opening and closing of FCD to vary the flow based on the average outlet temperature from each server. A baseline simulation with all servers running at maximum power dissipation with a constant coolant flow rate is compared with cases where the coolant flow rate is controlled dynamically for varying thermal design power (TDP). The results shown analyze the impact of the dynamic response of the flow control device on transient thermal and hydraulic characteristics across the rack is done.","PeriodicalId":117260,"journal":{"name":"ASME 2022 International Technical Conference and Exhibition on Packaging and Integration of Electronic and Photonic Microsystems","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123021116","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：