风冷式散热器设计与研究

华中科技大学硕士学位论文风冷式CPU散热片的热分析及其优化设计姓名：张远波申请学位级别：硕士专业：机械电子工程指导教师：熊蔡华20060508ICPU(CentralProcessingUnit)19902,300230,000,000CPUCPUCPUCPUCPUANSYSUIDL(User-InterfaceDesignLanguage)CPUANSYSCPUANSYSAPDL(ANSYSParametricDesignLanguage)ANSYSCPUAPDLUIDLIIAbstractWiththequicklydevelopmentofelectronicsindustry,thenumberoftransistorsintegratedinCPU(CentralProcessingUnit)growsrapidly(from2,300in1990to230,000,000nowadays),andthelinewidthofchipreducesrapidly.Asaresult,thepowerandtheheatinCPUgrowsharply,whichresultsinthatthefunctionofCPUisimpacted.Thus,theimprovementofcoolingperformanceofheatsinksbecomesoneofthekeyissuesinelectronicmanufacturefield.Inthisthesis,thecoolingrulesandstructureoptimizationofaircoolingsystemarestudieddeeply.AfterthestructurecharacteristicsoftheexistingCPUheatsinksareanalyzed,CPUheatsinksanalysissoftwareanduserinterfacearedevelopedusingUIDL(User-InterfaceDesignLanguage)ofANSYSandintegratedintoANSYS.Byusingthisprogram,userscananalyzetheinfluenceofvariousparameterstocoolingperformanceofCPUheatsinksconvenientlyandquickly.Theoptimizationprocessofheatsinksisacircularprocessofpre-solution,solvingandposttreatment.Onthepremiseofsatisfyingtheconstraintofcoolingspace,themaximumtemperatureofheatsinksisminimized.Thestructureparametersofheatsinksareoptimizedinordertoenhancethecoolingperformanceofheatsinks.Theconcreteprocessisasfollows:firstly,thedesignvariablesandobjectivefunctionaredefinedandtheoptimizationalgorithmsarechosen,secondly,theanalysisandoptimalcontrolprogramisdevelopedinANSYSusingADPL(ANSYSParametricDesignLanguage),lastlytheoptimizationmoduleofANSYSisusedtoimplementthestructureparametricoptimizationofheatsinks.Aseriesofexamplesofanalysisandstructureparametricoptimizationofheatsinksareshowedinordertovalidatetheanalysisandstructureparametricoptimizationmethodsproposedinthisthesis.Finally,theconclusionofthisthesisandtheadvicesforfutureresearcharegiven.Keywords:HeatSinks,OptimizationDesign,APDL,UIDL,ObjectiveFunction______111.1504360109732005CB7241002000271.22120042.65GDP13.9%20106.51.4GDP7%CPU1GHz4GHz18~24100W38°CBTX[1]CPU1°C2CPU0.05m0.0005m3CPU90°C40°CCPU200W250g38pa40cfm1-11.1CPU1.131.3CPUCPUCPUCPU(DARPA-DefenseAdvancedResearchProjectsAgency)HERE31C(HeatRemovalbyThermo—IntegratedCircuits)(ONR)(DOE)NSFNASANSAITHERM(InternationalWorkshopOilThermalInvestigationsofICsandSystems)SEMI-THERM(InternationalConferenceonThermal,MechanicsandThermomechanicalPhenomenainElectronicSystems)THERMINIC(SemiconductorThermalMeasurementandManagementSymposium)MMRCoolChipsCooligy[2]C.J.Shih[3]BejanMoregaMinakamiIwasaki[4]41.3.11[5]1CPUCPU2CPU342CPUCPUCPUCPUCPU3CPU[6]51.3.21CPUCPU10[7]CPUCPU2364CPUCPUCPUCPU[8]CPU[9]Intel4(3.06GHz)0.6IntelCPU[10,11]1.4UIDLAPDLCPUUIDLANSYSANSYS7APDLANSYS822.1CPUCPUCPUCPUANSYSANSYSANSYSANSYSANSYS[12]ANSYSAPDLUIDLANSYSCPU2.2ANSYS2.1[13]2.12.2.19ANSYS2.2.2[14]τρλλλ∂∂=+∂∂∂∂+∂∂∂∂+∂∂∂∂TcqzTzyTyxTxvvzyx(2.1)λTvqτvcλ=xλ=yλ=zλ=(2.1)τλρλ∂∂=+∂∂+∂∂+∂∂TcqzTyTxTvv/222222(2.2)VS(2.2)[15]0)})]()()({[(=⋅∂∂−+∂∂∂∂+∂∂∂∂+∂∂∂∂∫∫∫dxdydzTTcqzTzTyTyTxTxvVδτρλλλδ(2.3)fTαSαSS=)()(fwwTTnT−=∂∂−αλ(2.4)(2.4)(2.3)0))21(}][2{(2222222=−−∂∂+−∂∂+∂∂+∂∂∫∫∫∫∫dSTTTdxdydzTTcTqzTyTxTaSfvVατρλδ(2.5)0)(=TJδ(2.6)10)(TJEmmTTT,,,21⋅⋅⋅)(TJ)2,1,(0==∂∂mTJm(2.7))(TJeJ(2.7))2,1,(01==∂∂=∂∂∑=mTJTJEimem(2.8)meTJ∂∂0=∂∂mTJ2.2.30=∂∂mTJ(mT)[16]}{}{}{}]){[][]([}]{[gcfcdmQQQTKKKTC++=+++•(2.9)][CcdmKKK,,gcfQQQ,,(2.9)}{}]{[}{QTKT=+•λ(2.10)(2.10)112.3[17,18]3TSS=),,,(τzyxfTw=qSS=wwqnT=∂∂−)(λfTααSS=)()(fwwTTnT−=∂∂−αλ2.3.1ANSYS2.2(a)(b)(c)(d)2.22.3.2[19]12Solid90[20,21]2.3(a)(b)(c)(d)2.32.4UIDLANSYS13ANSYSGUIANSYSUIDLANSYSANSYS(GUI)ANSYS[22,23]UIDLUIDL.GRN[24]ANSYS2.4.1UIDLUIDLANSYSUIDLANSYSANSYSGUIUIDLANSYSGUI[25,26]menulist8.0.ansmenulist8.0UIFUNC2.GRNUIFUNC.GRNANSYS2.4.2ANSYS2.414ANSYSPREP7PREP7SOLUTIOPNSOLUTIOPNPOST1POST1ANSYS112.42.4.3[27,28]UIDLANSYSANSYS2-5ANSYS(MainMenu)(StraightSunflower)(PlateFinsModule)(PinFinsModule)(CurvingSunflower)OKANSYS15ANSYSMainMenu12.52.5PlateFinsModuleThermalAnalysisPlateFins(54321xxxxx)ok22.6162.632.72.71742.82.82.5UIDLANSYS1833.13.23.2.13.11x2x3x4x5x1x=2x=0.08m3.11913x=0.05m4x=0.0025m5x0.002m0.0025m0.003m0.0035m3.2(a)5x=0.002m(b)5x=0.0025m(c)5x=0.003m(d)5x=0.0035m3.23.33.3203.323x=0.05m5x=0.003m4x0.002m0.0025m0.003m0.0033m3.4(a)4x=0.002m(b)4x=0.0025m(c)4x=0.003m(d)4x=0.0033m3.43.53.5213.54x0.002m~0.0035m4x=0.0025m34x=0.0025m5x=0.003m3x0.04m0.045m0.04m0.055m3.6(a)3x=0.04m(b)3x=0.045m(c)3x=0.05m(d)3x=0.055m3.63.7223.73.743x=0.05m4x=0.0020m5x=0.003m161820223.8(a)16(b)18(c)20(d)223.8233.93.93.93.2.23.101x2x3x4x5x1x=2x=0.08m3.102413x=0.05m4x=0.0025m5x0.002m0.0025m0.003m0.0035m3.11(a)5x=0.002m(b)5x=0.0025m(c)5x=0.003m(d)5x=0.0035m3.113.123.12253.1223x=0.05m5x=0.003m4x0.002m0.0025m0.003m0.0033m3.13(a)4x=0.002m(b)4x=0.0025m(c)4x=0.003m(d)4x=0.0035m3.133.143.14263.144x0.002m~0.0035m4x=0.003m34x=0.0025m5x=0.003m3x0.04m0.045m0.04m0.055m3.15(a)3x=0.002m(b)3x=0.0025m(c)3x=0.003m(d)3x=0.0035m3.153.163.16273.1643x=0.05m4x=0.002m5x=0.003m16161818202022223.17(a)1616(b)1818(c)2020(d)22223.173.183.18283.183.2.