P. M. Souare, M. K. Touré, Stéphanie Allard, B. Borzou, J. Sylvestre, B. Foisy, É. Duchesne
{"title":"High Precision Numerical and Experimental Thermal Studies of Microelectronic Packages in Still Air Chamber Tests","authors":"P. M. Souare, M. K. Touré, Stéphanie Allard, B. Borzou, J. Sylvestre, B. Foisy, É. Duchesne","doi":"10.1109/ESTC.2018.8546346","DOIUrl":null,"url":null,"abstract":"In order to contribute to a better understanding of the relevant thermal phenomena, we have performed a high precision comparison of a microelectronic package cooled by natural convection with experimental data obtained in the Still Air JEDEC configuration. The heat transfer problem was solved by a bidirectional coupling method iterating between the solid domain and the fluid domain. At the first iteration, the conduction heat transfer in the solid domain was solved using convection coefficients estimated from empirical relations available from the literature, for each solid-fluid interface. The numerical temperature field so-obtained at the solid-fluid interface was then applied as a boundary condition for the numerical simulation of the of air flow around the package by computational fluid dynamics (CFD), to find the velocity, pressure and temperature fields in the fluid domain. The convection coefficient of each element at the solid-fluid interfaces was then computed from the heat flux in the CFD results. Finally, the convection coefficients obtained from the CFD analysis were applied back to the solid domain, and the same procedure was repeated iteratively until the solution converged to a stable temperature field. The comparison between the heat transfer coefficients obtained from the CFD method and those obtained from the empirical relations shows significant differences, thus validating the utility and effectiveness of the iterative approach to obtain precise thermal results. The numerical results for temperature were compared with measurements from a test vehicle. The experiments were carried out under natural convection according to the JEDEC standards in a still air chamber for the horizontal and vertical positioning of the test vehicle. The large difference in temperature between the walls of the still air chamber and the test vehicle leads to large errors in predicted temperatures if the radiation effects are not properly treated in the heat transfer problem. The junction temperatures obtained with the numerical simulations in the vertical and horizontal orientations were in excellent agreement with the experimental measurements, with an absolute difference of less than 1 °C. Such agreement demonstrates the accuracy of our methodology for the modeling of natural heat transfer for microelectronic packages mounted on printed circuit boards in the horizontal and vertical orientations.","PeriodicalId":198238,"journal":{"name":"2018 7th Electronic System-Integration Technology Conference (ESTC)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 7th Electronic System-Integration Technology Conference (ESTC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ESTC.2018.8546346","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
Abstract
In order to contribute to a better understanding of the relevant thermal phenomena, we have performed a high precision comparison of a microelectronic package cooled by natural convection with experimental data obtained in the Still Air JEDEC configuration. The heat transfer problem was solved by a bidirectional coupling method iterating between the solid domain and the fluid domain. At the first iteration, the conduction heat transfer in the solid domain was solved using convection coefficients estimated from empirical relations available from the literature, for each solid-fluid interface. The numerical temperature field so-obtained at the solid-fluid interface was then applied as a boundary condition for the numerical simulation of the of air flow around the package by computational fluid dynamics (CFD), to find the velocity, pressure and temperature fields in the fluid domain. The convection coefficient of each element at the solid-fluid interfaces was then computed from the heat flux in the CFD results. Finally, the convection coefficients obtained from the CFD analysis were applied back to the solid domain, and the same procedure was repeated iteratively until the solution converged to a stable temperature field. The comparison between the heat transfer coefficients obtained from the CFD method and those obtained from the empirical relations shows significant differences, thus validating the utility and effectiveness of the iterative approach to obtain precise thermal results. The numerical results for temperature were compared with measurements from a test vehicle. The experiments were carried out under natural convection according to the JEDEC standards in a still air chamber for the horizontal and vertical positioning of the test vehicle. The large difference in temperature between the walls of the still air chamber and the test vehicle leads to large errors in predicted temperatures if the radiation effects are not properly treated in the heat transfer problem. The junction temperatures obtained with the numerical simulations in the vertical and horizontal orientations were in excellent agreement with the experimental measurements, with an absolute difference of less than 1 °C. Such agreement demonstrates the accuracy of our methodology for the modeling of natural heat transfer for microelectronic packages mounted on printed circuit boards in the horizontal and vertical orientations.