Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992643
Cheng Xu, Z. Zhong, W. Choi
Fan-out wafer level packaging technology becomes more and more popular and attractive of its flexibility for integration of diverse devices in a tiny form factor. The small and thin fan-out wafer level package has low package strength and often faces crack issues. In the previous study, the fan-out wafer level package strength increased significantly when the wafers were lithographed passivation layers. The extreme thin passivation layers should not impact the package strength seriously. Therefore, the only explanation was any package material strength changed during the lithographing process. In this study, the thermal properties of epoxy molding compound were evaluated to understand the effect of epoxy molding compound on the package strength and reliability. Both Vickers hardness test and three-point bending test were used to evaluate the changing of epoxy molding compound strength. The results showed that the post-mold curing process and any post-mold thermal process had a significant effect on the epoxy molding compound strength.
{"title":"Epoxy molding compound effect on fan-out wafer level package strength during post-mold thermal process","authors":"Cheng Xu, Z. Zhong, W. Choi","doi":"10.1109/ITHERM.2017.7992643","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992643","url":null,"abstract":"Fan-out wafer level packaging technology becomes more and more popular and attractive of its flexibility for integration of diverse devices in a tiny form factor. The small and thin fan-out wafer level package has low package strength and often faces crack issues. In the previous study, the fan-out wafer level package strength increased significantly when the wafers were lithographed passivation layers. The extreme thin passivation layers should not impact the package strength seriously. Therefore, the only explanation was any package material strength changed during the lithographing process. In this study, the thermal properties of epoxy molding compound were evaluated to understand the effect of epoxy molding compound on the package strength and reliability. Both Vickers hardness test and three-point bending test were used to evaluate the changing of epoxy molding compound strength. The results showed that the post-mold curing process and any post-mold thermal process had a significant effect on the epoxy molding compound strength.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125613007","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992599
A. Muslu, Burak Ozluk, Enes Tamdogan, M. Arik
Commercially available light emitting diodes (LEDs) that have high efficiencies and long lifetime are offered in advanced packaging technologies. Many cooling systems were developed for current LED systems that enable a better removal of heat than counterpart devices offered earlier this decade. On the other hand, these lighting systems are still producing a considerable amount of heat that is still not effectively removed. Especially, p-n junctions of LEDs are the most critical regions where a significant amount of heating occurs, and it is crucial to determine the temperature of this active region to meet the lumen extraction, color, light quality and lifetime goals. In literature, there are some proposed junction temperature measurement methods such as Peak Wavelength Shift, Thermal (Infrared) Imaging and Forward Voltage Change methods mostly focused on blue LEDs. In this study, we are studying three common types of LEDs (Red, Green, and Blue) and comparing their forward voltage drop (Vf) behaviors. A set of theoretical, computational and experimental studies have been performed. It is found that optical power change with temperature in red LEDs are much higher than blue and green chips. The green LED chip experienced the largest slope having the largest change in forward voltage compared to other LED chips.
{"title":"Impact of junction temperature over forward voltage drop for red, blue and green high power light emitting diode chips","authors":"A. Muslu, Burak Ozluk, Enes Tamdogan, M. Arik","doi":"10.1109/ITHERM.2017.7992599","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992599","url":null,"abstract":"Commercially available light emitting diodes (LEDs) that have high efficiencies and long lifetime are offered in advanced packaging technologies. Many cooling systems were developed for current LED systems that enable a better removal of heat than counterpart devices offered earlier this decade. On the other hand, these lighting systems are still producing a considerable amount of heat that is still not effectively removed. Especially, p-n junctions of LEDs are the most critical regions where a significant amount of heating occurs, and it is crucial to determine the temperature of this active region to meet the lumen extraction, color, light quality and lifetime goals. In literature, there are some proposed junction temperature measurement methods such as Peak Wavelength Shift, Thermal (Infrared) Imaging and Forward Voltage Change methods mostly focused on blue LEDs. In this study, we are studying three common types of LEDs (Red, Green, and Blue) and comparing their forward voltage drop (Vf) behaviors. A set of theoretical, computational and experimental studies have been performed. It is found that optical power change with temperature in red LEDs are much higher than blue and green chips. The green LED chip experienced the largest slope having the largest change in forward voltage compared to other LED chips.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130587029","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992492
Young-kuk Kim
Efficient removal of accumulated heat is essential to maintaining device efficiency and ensuring long lifetime span. Polymer composites composed of highly conductive ceramic fillers and flexible polymer matrix are popularly selected as thermal management materials owing to their moderate thermal transport properties, gap-filling capability and easy processing ability with low cost. High thermal conductivity of polymer composites is one of the primary requisites for thermal management materials of electronic devices. We report abnormal increase of thermal conduction of Al2O3-based silicone polymer composites by induced clustering of thermally conductive BN nanoplatelets. Initially, polymer composite containing various amount of spherical Al2O3 fillers have been prepared and measured thermal diffusivity along through-plane direction (dt) was 0.7 mm2/s at best. This value of DT is only 5% of known TDs for bulk polycrystalline Al2O3. The main cause of the low thermal transport properties is assumed to be discrete distribution of thermally conductive fillers. Then, small amount of nanocrystalline platelets with high thermal conductivity are incorporated to enhance the connectivity between fillers. As-received BN powders were surface-oxidized with high temperature treatment under ambient atmosphere and exfoliated to be nanoplatelets after further ball milling with water to ensure uniform dispersion in polymer matrix. Al2O3 containing composite showed dt of enhanced to 1.0 mm2/s after subsequent addition of surface modified BN nanoplatelets by 6 vol%. Further addition of BN nanoplatelets resulted in abrupt increase in DT up to 1.5 mm2/s which is almost twice as high as dt of a pristine Al2O3-containing polymer composite. The clustering of thermally conductive ceramic fillers enclosed with polymer matrix is shown to be induced by colloidal interaction between BN nanoplatelets in the gaps between macroscopic Al2O3 spheres. These can be attributed to be main cause of rapid improvement of thermal diffusivity in the composites. Here, the benefits of induced clustering for thermal management composite materials are illustrated.
{"title":"Induced clustering-enabled thermal transport enhancement in polymer composites for efficient thermal interface materials","authors":"Young-kuk Kim","doi":"10.1109/ITHERM.2017.7992492","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992492","url":null,"abstract":"Efficient removal of accumulated heat is essential to maintaining device efficiency and ensuring long lifetime span. Polymer composites composed of highly conductive ceramic fillers and flexible polymer matrix are popularly selected as thermal management materials owing to their moderate thermal transport properties, gap-filling capability and easy processing ability with low cost. High thermal conductivity of polymer composites is one of the primary requisites for thermal management materials of electronic devices. We report abnormal increase of thermal conduction of Al2O3-based silicone polymer composites by induced clustering of thermally conductive BN nanoplatelets. Initially, polymer composite containing various amount of spherical Al2O3 fillers have been prepared and measured thermal diffusivity along through-plane direction (dt) was 0.7 mm2/s at best. This value of DT is only 5% of known TDs for bulk polycrystalline Al2O3. The main cause of the low thermal transport properties is assumed to be discrete distribution of thermally conductive fillers. Then, small amount of nanocrystalline platelets with high thermal conductivity are incorporated to enhance the connectivity between fillers. As-received BN powders were surface-oxidized with high temperature treatment under ambient atmosphere and exfoliated to be nanoplatelets after further ball milling with water to ensure uniform dispersion in polymer matrix. Al2O3 containing composite showed dt of enhanced to 1.0 mm2/s after subsequent addition of surface modified BN nanoplatelets by 6 vol%. Further addition of BN nanoplatelets resulted in abrupt increase in DT up to 1.5 mm2/s which is almost twice as high as dt of a pristine Al2O3-containing polymer composite. The clustering of thermally conductive ceramic fillers enclosed with polymer matrix is shown to be induced by colloidal interaction between BN nanoplatelets in the gaps between macroscopic Al2O3 spheres. These can be attributed to be main cause of rapid improvement of thermal diffusivity in the composites. Here, the benefits of induced clustering for thermal management composite materials are illustrated.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130386281","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992588
A. Díaz, G. Neves, Luis Silva-Llanca, M. del Valle, J. Cardemil
Data center energy consumption in Latin America has increased considerably during last years. According to Datacenter dynamics, energy requirements during 2016 were expected to be around 3.85 GW. In Chile, the data center industry grew 14% between 2009 and 2010, whereas energy consumption increased 21.4% between 2012 and 2013. For this reason, many data centers in the country have started to evaluate efficient alternatives to reduce energy consumption such as the use of air containment techniques, air-side and water-side cooling systems. To date, existing free-cooling maps do not provide information about available hours during the year for implementing either air-side or water-side cooling systems in data centers in South America. This paper presents a thermo-dynamic analysis aimed to evaluate the potential use of air-side free-cooling systems in the Chilean data center industry. First, temperature and Relative Humidity (RH) variations, during three years, were obtained at 29 different stations throughout the entire country. The objective was to identify regions in Chile that meet data center thermal requirements proposed by the ASHRAE. Fiber-optic availability was also considered during the analysis. The thermodynamic model considered a white room with a thermal load of 20 kW, for which an air treatment unit was incorporated with the objective of providing cold air at 18° and 60% RH. An air treatment system was calculated at three different locations in Chile. These locations were selected since they offer high availability of fiber-optic connections (Chacalluta, Arica y Parinacota Region), strategic position for companies (Quinta Normal, Metropolitan Region), and low temperatures through the year (Carlos Ibanez, Aysen Region). Preliminary results have demonstrated that Chile is a relatively humid country. For this reason, cooling air must be dehumidified most of the time. The results also showed that even when low temperatures can be found in Carlos Ibanez, both Chacalluta and Quinta Normal offer excellent possibilities for the data centers industry. These two last locations offer more fiber-optic connections and temperature variations that lay within the range established by the ASHRAE.
拉丁美洲的数据中心能耗在过去几年中大幅增加。根据数据中心动态，2016年的能源需求预计约为3.85吉瓦。在智利，数据中心行业在2009年至2010年间增长了14%，而能源消耗在2012年至2013年间增长了21.4%。因此，该国的许多数据中心已开始评估减少能源消耗的有效替代方案，例如使用空气密封技术、空气侧和水侧冷却系统。到目前为止，现有的自然冷却地图没有提供南美洲数据中心在一年中可用于实施空气侧或水侧冷却系统的可用时间的信息。本文提出了一项热力学分析，旨在评估智利数据中心工业中空气侧自然冷却系统的潜在用途。首先，获得了全国29个不同站点三年来的温度和相对湿度(RH)变化。目的是在智利确定符合ASHRAE提出的数据中心热要求的区域。在分析过程中还考虑了光纤的可用性。热力学模型考虑了一个热负荷为20 kW的白色房间，其中包含一个空气处理单元，目标是提供18°和60%相对湿度的冷空气。在智利的三个不同地点计算了空气处理系统。选择这些地点是因为它们提供高可用性的光纤连接(查卡卢塔，Arica y Parinacota地区)，公司的战略位置(Quinta Normal，大都市区)，以及全年的低温(Carlos Ibanez, Aysen地区)。初步结果表明，智利是一个相对潮湿的国家。因此，大多数时候必须对冷却空气进行除湿。结果还表明，即使在Carlos Ibanez可以找到低温，Chacalluta和Quinta Normal都为数据中心行业提供了极好的可能性。最后两个地点提供了更多的光纤连接和温度变化，这些都在ASHRAE确定的范围内。
{"title":"Meteorological assessment and implementation of an air-side free-cooling system for data centers in Chile","authors":"A. Díaz, G. Neves, Luis Silva-Llanca, M. del Valle, J. Cardemil","doi":"10.1109/ITHERM.2017.7992588","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992588","url":null,"abstract":"Data center energy consumption in Latin America has increased considerably during last years. According to Datacenter dynamics, energy requirements during 2016 were expected to be around 3.85 GW. In Chile, the data center industry grew 14% between 2009 and 2010, whereas energy consumption increased 21.4% between 2012 and 2013. For this reason, many data centers in the country have started to evaluate efficient alternatives to reduce energy consumption such as the use of air containment techniques, air-side and water-side cooling systems. To date, existing free-cooling maps do not provide information about available hours during the year for implementing either air-side or water-side cooling systems in data centers in South America. This paper presents a thermo-dynamic analysis aimed to evaluate the potential use of air-side free-cooling systems in the Chilean data center industry. First, temperature and Relative Humidity (RH) variations, during three years, were obtained at 29 different stations throughout the entire country. The objective was to identify regions in Chile that meet data center thermal requirements proposed by the ASHRAE. Fiber-optic availability was also considered during the analysis. The thermodynamic model considered a white room with a thermal load of 20 kW, for which an air treatment unit was incorporated with the objective of providing cold air at 18° and 60% RH. An air treatment system was calculated at three different locations in Chile. These locations were selected since they offer high availability of fiber-optic connections (Chacalluta, Arica y Parinacota Region), strategic position for companies (Quinta Normal, Metropolitan Region), and low temperatures through the year (Carlos Ibanez, Aysen Region). Preliminary results have demonstrated that Chile is a relatively humid country. For this reason, cooling air must be dehumidified most of the time. The results also showed that even when low temperatures can be found in Carlos Ibanez, both Chacalluta and Quinta Normal offer excellent possibilities for the data centers industry. These two last locations offer more fiber-optic connections and temperature variations that lay within the range established by the ASHRAE.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"6 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120807283","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992499
J. Ruiz, Yash Ganatra, A. Bruce, J. Howarter, A. Marconnet
Passive thermal management with phase change materials (PCMs) has become the one of the most promising methods to cool cell phone processors due to the relatively simple implementation and profound impact on processor temperatures. Enhancing the thermal properties of conventional PCMs, mainly thermal conductivity and latent heat storage, allows for an overall improved thermal management system. This study aims to improve the thermal conductivity of paraffin wax (a typical commercial PCM) by the introduction of an expanded graphite (EG) filler to form a paraffin wax composite, and then infiltration of the EG/paraffin composite into an aluminum foam matrix. The thermal conductivity of the EG/paraffin composites increases respectively to the volume percent of expanded graphite. While the thermal conductivity increased, there is some negative impact on latent heat storage compared to pure paraffin wax. The pore size of the aluminum foam matrixes also has a profound impact on both thermal conductivity and latent heat storage of the overall system. These results will allow for improvements in cooling techniques incorporated within cell phones and other mobile devices, allowing for future development of their processors (higher computational power), prolonged reliability, and longer anticipated life cycles.
{"title":"Investigation of aluminum foams and graphite fillers for improving the thermal conductivity of paraffin wax-based phase change materials","authors":"J. Ruiz, Yash Ganatra, A. Bruce, J. Howarter, A. Marconnet","doi":"10.1109/ITHERM.2017.7992499","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992499","url":null,"abstract":"Passive thermal management with phase change materials (PCMs) has become the one of the most promising methods to cool cell phone processors due to the relatively simple implementation and profound impact on processor temperatures. Enhancing the thermal properties of conventional PCMs, mainly thermal conductivity and latent heat storage, allows for an overall improved thermal management system. This study aims to improve the thermal conductivity of paraffin wax (a typical commercial PCM) by the introduction of an expanded graphite (EG) filler to form a paraffin wax composite, and then infiltration of the EG/paraffin composite into an aluminum foam matrix. The thermal conductivity of the EG/paraffin composites increases respectively to the volume percent of expanded graphite. While the thermal conductivity increased, there is some negative impact on latent heat storage compared to pure paraffin wax. The pore size of the aluminum foam matrixes also has a profound impact on both thermal conductivity and latent heat storage of the overall system. These results will allow for improvements in cooling techniques incorporated within cell phones and other mobile devices, allowing for future development of their processors (higher computational power), prolonged reliability, and longer anticipated life cycles.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129676977","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992476
R. Warzoha, A. Smith, A. Bajwa, L. Boteler
This work describes relevant design considerations for the fabrication of a miniaturized thermal interface material characterization instrument that is capable of resolving interfacial thermal resistances (Rt) below 1 mm2.K/W. Leveraging previous work (Warzoha et al., 2017, Smith et al. 2016), the authors propose a reduction in the length scale of the primary heat meter bars to below 4 mm in order to sufficiently increase the temperature difference across the interface, thereby reducing the measurement uncertainty of Rt across high-performance materials. The analytical uncertainty analysis takes advantage of an increase in the number of temperature measurements that can be made across the length of each bar via infrared microscopy. In a preliminary numerical analysis, we find that extreme care must be taken to apply and remove heat uniformly from the end points of each bar, particularly as the length of the bar is reduced below 4 mm. To do this, longitudinal fins are directly integrated into the bottom heat meter bar assembly and are immersed in a heat transfer fluid that is advected within a custom cold plate assembly. We conduct a parametric study to determine the linearity of the thermal gradient along the length of each heat meter bar, which in turn provides us with an upper limit for the number of temperature measurements that can be made via infrared microscopy and therefore the minimum achievable measurement of Rt. Finally, we use this information to design a more suitable lower heat meter bar cooling technique for measuring the thermal resistance across a sintered silver-copper interface with an expected value of Rt = 0.1 mm2.K/W. To do this, we find it necessary to transition from a heat sink cooling mechanism to the use of jet impingement for heat dissipation at the bottom of the lower heat meter bar.
{"title":"Design considerations for a miniaturized TIM tester with extremely high measurement resolution","authors":"R. Warzoha, A. Smith, A. Bajwa, L. Boteler","doi":"10.1109/ITHERM.2017.7992476","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992476","url":null,"abstract":"This work describes relevant design considerations for the fabrication of a miniaturized thermal interface material characterization instrument that is capable of resolving interfacial thermal resistances (Rt) below 1 mm2.K/W. Leveraging previous work (Warzoha et al., 2017, Smith et al. 2016), the authors propose a reduction in the length scale of the primary heat meter bars to below 4 mm in order to sufficiently increase the temperature difference across the interface, thereby reducing the measurement uncertainty of Rt across high-performance materials. The analytical uncertainty analysis takes advantage of an increase in the number of temperature measurements that can be made across the length of each bar via infrared microscopy. In a preliminary numerical analysis, we find that extreme care must be taken to apply and remove heat uniformly from the end points of each bar, particularly as the length of the bar is reduced below 4 mm. To do this, longitudinal fins are directly integrated into the bottom heat meter bar assembly and are immersed in a heat transfer fluid that is advected within a custom cold plate assembly. We conduct a parametric study to determine the linearity of the thermal gradient along the length of each heat meter bar, which in turn provides us with an upper limit for the number of temperature measurements that can be made via infrared microscopy and therefore the minimum achievable measurement of Rt. Finally, we use this information to design a more suitable lower heat meter bar cooling technique for measuring the thermal resistance across a sintered silver-copper interface with an expected value of Rt = 0.1 mm2.K/W. To do this, we find it necessary to transition from a heat sink cooling mechanism to the use of jet impingement for heat dissipation at the bottom of the lower heat meter bar.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133851406","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992486
L. Jani, A. Poppe
With the advance of the semiconductor technology power density and related thermal management issues became design bottlenecks. These physical limits require design engineers to make several thermal aware decisions during the design process: the earlier the better. Modern hardware description languages have extensions for simulation of mixed-signal circuits (e.g. SystemC-AMS, Verilog-AMS, VHDL-AMS) but none of these approaches support co-simulation of the effect of the foreseen thermal environment of the design with the logic behavior. A relatively new simulation paradigm called logi-thermal simulation is aimed to fill this gap in the available set of simulation tools. Our framework for co-simulation of logic and thermal behavior called LogiTherm contains generic interfaces towards usual logic and thermal simulation engines. In our present framework setup SystemC and Verilog is supported as hardware description languages and two thermal field solvers, SUNRED and 3D-ICE can be used as thermal simulation engines. In this paper we present the recent developments of the LogiTherm framework that enable logi-thermal simulation of mixed signal designs. We demonstrate the capability of our system by presenting simulation results of a test system, which contains a microprocessor and mixed signal components as well.
{"title":"Extending a multi-level logi-thermal simulation framework to a mixed signal thermal aware simulation environment using SystemC-AMS","authors":"L. Jani, A. Poppe","doi":"10.1109/ITHERM.2017.7992486","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992486","url":null,"abstract":"With the advance of the semiconductor technology power density and related thermal management issues became design bottlenecks. These physical limits require design engineers to make several thermal aware decisions during the design process: the earlier the better. Modern hardware description languages have extensions for simulation of mixed-signal circuits (e.g. SystemC-AMS, Verilog-AMS, VHDL-AMS) but none of these approaches support co-simulation of the effect of the foreseen thermal environment of the design with the logic behavior. A relatively new simulation paradigm called logi-thermal simulation is aimed to fill this gap in the available set of simulation tools. Our framework for co-simulation of logic and thermal behavior called LogiTherm contains generic interfaces towards usual logic and thermal simulation engines. In our present framework setup SystemC and Verilog is supported as hardware description languages and two thermal field solvers, SUNRED and 3D-ICE can be used as thermal simulation engines. In this paper we present the recent developments of the LogiTherm framework that enable logi-thermal simulation of mixed signal designs. We demonstrate the capability of our system by presenting simulation results of a test system, which contains a microprocessor and mixed signal components as well.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132849209","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992578
M. del Valle, Carol Caceres, A. Ortega, K. Nemati, B. Sammakia
The transient modeling of cross flow heat exchangers requires the use of complicated analytical solutions or complex numerical schemes to solve the set of coupled partial differential equations representing the air, coolant and wall temperatures. Since most of the dynamic mass flow or temperature perturbations encountered in data center environment are slow compared with the timescale of the heat exchanger, using a simpler quasi-steady model instead of a complex transient solution represents a good alternative. The present work shows the use of a quasi-steady heat exchanger model to recreate the transient response of a 12×12 in heat exchanger core under ramp and sinusoidal perturbations. A criteria to identify the regimes where the quasi-steady model is valid is developed. Finally the model is used to model a rear door heat exchanger and the results compared with experimental results.
{"title":"Quasi-steady modeling of data center heat exchanger under dynamic conditions","authors":"M. del Valle, Carol Caceres, A. Ortega, K. Nemati, B. Sammakia","doi":"10.1109/ITHERM.2017.7992578","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992578","url":null,"abstract":"The transient modeling of cross flow heat exchangers requires the use of complicated analytical solutions or complex numerical schemes to solve the set of coupled partial differential equations representing the air, coolant and wall temperatures. Since most of the dynamic mass flow or temperature perturbations encountered in data center environment are slow compared with the timescale of the heat exchanger, using a simpler quasi-steady model instead of a complex transient solution represents a good alternative. The present work shows the use of a quasi-steady heat exchanger model to recreate the transient response of a 12×12 in heat exchanger core under ramp and sinusoidal perturbations. A criteria to identify the regimes where the quasi-steady model is valid is developed. Finally the model is used to model a rear door heat exchanger and the results compared with experimental results.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"255 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132858322","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992525
J. Jörg, S. Taraborrelli, E. Sabelberg, R. Kneer, R. D. De Doncker, W. Rohlfs
This study examines the potential performance of submerged single phase direct impinging jet cooling and compares this cooling concept to a state of the art pin fin cooling systems. Submerged single impinging jets and arrays of impinging jets are characterized by high heat transfer coefficients. Especially in the stagnation region, in which the jet provides fresh cooling liquid to the surface, high heat removal rates occur. Thus, impinging jet arrays are a promising approach for hot spot removal as well as for thermal uniformity in large areas. In the present cooling concept, micro jets of approximately 1 mm diameter directly impinge onto the backside of a IGBT-semiconductor. In contrast to pin fin cooling, direct jet impingement cooling dispenses with the need for any kind of thermal interface materials (TIM) or heat spreader, and thus, minimizes the thermal resistance of the heat sink. A further advantage of submerged direct impinging jet cooling is hot spot removal. Once the exact location of heat generation is determined a small impinging jet can be directed at these hot spots. For this, local heat generation of the IGBT-semiconductor is first investigated by measuring the surface temperature on the top and bottom side by IR-thermometry. To observe the local heat generation the IGBT is operated for a very short time without any heat sink. This information is used in the development and design of the cooling chamber and the jet positioning. Finally, the cooling system is analyzed and compared to other common cooling systems. Both liquid cooling concepts are experimentally investigated with respect to heat transfer, pressure drop, and pumping power. A comparison and evaluation is provided with a special focus on applications in the automotive and electro-mobility sector.
{"title":"Hot spot removal in power electronics by means of direct liquid jet cooling","authors":"J. Jörg, S. Taraborrelli, E. Sabelberg, R. Kneer, R. D. De Doncker, W. Rohlfs","doi":"10.1109/ITHERM.2017.7992525","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992525","url":null,"abstract":"This study examines the potential performance of submerged single phase direct impinging jet cooling and compares this cooling concept to a state of the art pin fin cooling systems. Submerged single impinging jets and arrays of impinging jets are characterized by high heat transfer coefficients. Especially in the stagnation region, in which the jet provides fresh cooling liquid to the surface, high heat removal rates occur. Thus, impinging jet arrays are a promising approach for hot spot removal as well as for thermal uniformity in large areas. In the present cooling concept, micro jets of approximately 1 mm diameter directly impinge onto the backside of a IGBT-semiconductor. In contrast to pin fin cooling, direct jet impingement cooling dispenses with the need for any kind of thermal interface materials (TIM) or heat spreader, and thus, minimizes the thermal resistance of the heat sink. A further advantage of submerged direct impinging jet cooling is hot spot removal. Once the exact location of heat generation is determined a small impinging jet can be directed at these hot spots. For this, local heat generation of the IGBT-semiconductor is first investigated by measuring the surface temperature on the top and bottom side by IR-thermometry. To observe the local heat generation the IGBT is operated for a very short time without any heat sink. This information is used in the development and design of the cooling chamber and the jet positioning. Finally, the cooling system is analyzed and compared to other common cooling systems. Both liquid cooling concepts are experimentally investigated with respect to heat transfer, pressure drop, and pumping power. A comparison and evaluation is provided with a special focus on applications in the automotive and electro-mobility sector.","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129443725","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}
Pub Date : 2017-05-01DOI: 10.1109/ITHERM.2017.7992594
Michael Ot, T. Wilde, Herbert Ruber
Since mechanical chillers account for a large fraction of electricity spent on cooling, data centers are looking for ways to reduce their usage. In High Performance Computing, a popular approach is to use readily available high temperature direct liquid cooling (HT-DLC) that allows for mechanical chiller free cooling of compute components. Additionally, it provides a means to re-use their waste heat. A potential application is to use the waste heat to produce still needed cold water via adsorption refrigeration. This paper analyses the first production level installation of adsorption technology in a data center in terms of energy flows, Return of Investment (ROI), and Total Cost of Ownership (TCO).
{"title":"ROI and TCO analysis of the first production level installation of adsorption chillers in a data center","authors":"Michael Ot, T. Wilde, Herbert Ruber","doi":"10.1109/ITHERM.2017.7992594","DOIUrl":"https://doi.org/10.1109/ITHERM.2017.7992594","url":null,"abstract":"Since mechanical chillers account for a large fraction of electricity spent on cooling, data centers are looking for ways to reduce their usage. In High Performance Computing, a popular approach is to use readily available high temperature direct liquid cooling (HT-DLC) that allows for mechanical chiller free cooling of compute components. Additionally, it provides a means to re-use their waste heat. A potential application is to use the waste heat to produce still needed cold water via adsorption refrigeration. This paper analyses the first production level installation of adsorption technology in a data center in terms of energy flows, Return of Investment (ROI), and Total Cost of Ownership (TCO).","PeriodicalId":387542,"journal":{"name":"2017 16th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129783984","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：