Cooling Of Automotive Traction Motors: Schemes, Examples . PDF

16d ago
1.68 MB
12 Pages

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TIE.2018.2835397, IEEETransactions on Industrial ElectronicsIEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICSCooling of Automotive Traction Motors:Schemes, Examples and ComputationMethods– A ReviewYaohui Gai, Mohammad Kimiabeigi, Yew Chuan Chong, James D. Widmer, Xu Deng, Mircea Popescu,Fellow, IEEE, James Goss, Member, IEEE, Dave Staton, Member, IEEE, and Andrew Steven Abstract—This paper presents a comprehensiveoverview of the latest studies and analyses of the coolingtechnologies and computation methods for the automotivetraction motors. Various cooling methods, including thenatural, forced air, forced liquid and phase change types,are discussed with the pros and cons of each methodbeing compared. The key factors for optimizing the heattransfer efficiency of each cooling system are highlightedhere. Furthermore, the real life examples of thesemethods, applied in the latest automotive traction motorprototypes and products, have been set out andevaluated. Finally, the analytical and numerical techniquesdescribing the nature and performance of different coolingschemes have been explained and addressed. This paperprovides guidelines for selecting the appropriate coolingmethods and estimating the performance of them in theearly stages of their design.Index Terms—Automotive applications, cooling, tractionmotors, thermal analysis, numerical analysis.NOMENCLATURE𝐴𝐴𝑙 ,𝐴𝑖 ,𝐴𝑜𝑐𝑝𝐷𝑓𝑠 ,𝑓𝑟𝑔𝐺𝑟𝐻ℎℎ𝑙𝑘Cross section area of heat path (m2).Linear current density (kA/m).Inlet and outlet cross section areas (m2).Specific heat capacity (J/kg).Diameter (m).Friction loss factor (dimensionless).Gravitational attraction force (m/s2).Grashof number (dimensionless).Fin extension (m).Heat transfer coefficient (W/m2K).Latent heat (kJ/kg).Loss coefficient (dimensionless).Manuscript received December 20, 2017; revised February 05,2017 and March 31, 2018; accepted April 24, 2018. (Correspondingauthor: Yaohui Gai.)Y. Gai, M. Kimiabeigi, J. D. Widmer, X. Deng and A. Steven are withthe School of Engineering, Newcastle University, Newcastle [email protected]; [email protected] ).Y. C. Chong, M Popescu, J. Goss and D. Staton are with MotorDesign Ltd., Wrexham, LL13 7YT, UK (email: [email protected]; [email protected]; [email protected]; 𝑒𝑑 𝑝𝑃𝑟𝑅𝑅𝑒𝑅𝑒𝑟𝑇𝑤 , 𝑇𝑓 𝑇𝑆𝑉𝑉𝑟𝜇𝜆𝜌𝛽𝜎Current density (A/mm2)Length of the surface (m).Number of fins (dimensionless).Natural Nusselt number (dimensionless).Forced Nusselt number (dimensionless).Pressure drop (Pa).Prandtl number (dimensionless).Convection thermal resistance (K/W).Reynold number (dimensionless).Rotational Reynold number (dimensionless).Wall and fluid temperatures (K).Temperature difference (K).Fin pitch (m).Axial velocity (m/s).Tangential velocity (m/s).Dynamic viscosity (Pa s).Thermal conductivity (W/m K).Density (kg/m3).Coefficient of the expansion (dimensionless).Tangential stress (kPa).I. INTRODUCTIONWHILE operating an electric motor, heat is generated dueto the electromagnetic losses, mechanical power lossesand other stray losses that take place in various componentswithin an electric motor. Through conduction, convectionand/or radiation, the thermal energy is transferred to a coolingmedium [1] on the basis of a temperature difference betweenthe hot and cold bodies. However, a detailed thermalmanagement is essential during critical operating conditions,such as overload running, phase changing and/or asymmetricfaults, to avoid failures that are usually due to the local hotspot formation, and material degradation [2-5]. Furthermore,the topic of magnetic losses and heat generation governs theperformance of the electromagnetic efficiency and longer lifeexpectancy. Firstly, excessively high temperatures can causeaccelerated insulation aging [6] and deterioration within someessential components, such as winding conductors [7].Secondly, the remanence and coercivity of the rare earthmagnets are inversely proportional to the temperature. As aresult of which, partial or full demagnetization at highertemperatures may occur [8, 9]. In case of the ferrite andrecycled magnets [10, 11], the lower rotor temperatures may0278-0046 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TIE.2018.2835397, IEEETransactions on Industrial ElectronicsIEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICSsignificantly boost the torque density and / or efficiency of themotor, as for the remanence sensitivity of these magnet to thetemperature is 200%-300% higher than the conventional rareearth magnets. In addition, the electrical resistivity of thewinding conductors is in a direct proportion to thetemperature. This process can lead to a positive feedback inwhich an accelerated loss and temperature rises occur in thewindings [12]. Finally, the thermal impact on the geometricaldimensions of the motor’s physical structure, such as anarrowing within the airgap, may alter the motor’s nominalperformance, or, in serious cases, result in faults and failures[13]. To tackle the thermal challenges in an electric motor,alongside the minimization of the magnetic losses as sourcesof heat, one needs to carefully address the heat dissipationmechanisms for a given design in order to obtain a balancedheat distribution across different components.Generally speaking, a thermal design uses a closed or anopen cooling circuit to achieve a critical temperature balancewithin an electric motor. Heat from the inner components isconducted to the outer surface of the motor and then issubjected to the convective cooling. The former process is akind of passive thermal design, which is affected by materialproperties, geometrical layout and contact interfaces. Thisprocess is considered economical and does not produce anyadditional parasitic effects such as acoustic noise. Analternative method is the so-called active cooling design inwhich an extra source of energy is applied to circulate a fluidwith a high heat capacity in order to exchange and extract theheat from the hot surfaces [14]. This active method of coolingapplies an external force created by a special device, e.g.pumps, fans to generate sufficient coolant flow to remove heatfrom the interior parts of a motor. This approach provides ahigh convection heat transfer capacity but the extra provisionsare required for diminishing not only friction losses, but alsorisks of short circuit faults and corrosion [15]. Table I lists thetypical values for the tangential stress 𝜎, linear current density𝐴𝑙 , current density 𝐽 and heat transfer coefficients ℎ ofdifferent cooling methods.A diversification of cooling approaches have been pursuedto meet the cooling demands placed on various applications.However, the rapid growth of aerospace and traction industrieshave brought about increased requirements for electric motorssuch as compactness, high speed and high power density. Thisleads to significant rises in temperature in cases whereminiaturized motors are involved, thus necessitating a moresophisticated and complicated cooling system to keep theworking temperature within a safe range [16, 17].In this paper, a detailed analysis of the active type cooling:the natural [21-28], forced air [29-42], forced liquid [43-63]and phase change types [18, 66-68], are reviewed in SectionII. On this basis, a comprehensive summary of the convectionmethods as applicable to the automotive traction motorscooling contexts have been provided with the advantages anddisadvantages of each method being compared. The essentialelements for optimizing the cooling performance of eachmethod together with the leading applications are specificallyhighlighted. In addition, the latest automotive traction motorprototypes and products [69-75] employing these methods,have been set out and evaluated in Section III. In Section IV,the use of the analytical lumped-circuit and the computationalfluid dynamics techniques [90-94] for calculating the coolingperformance are proposed and discussed. Section V sets outthe conclusion in which a number of recommendationsconcerning the future developments and applications isoffered. The aim of this paper is to familiarize and equip theelectrical machine designers with a concise knowledge of thethermal background to improve upon the overall performanceof their products.TABLE ITHE TYPICAL VALUES FOR DIFFERENT COOLING METHODS [18-20].Cooling method𝜎, kPa𝐴𝑙 , kA/m𝐽, A/mm2ℎ, W/m2 KNatural convection1.5-55-30Forced Air 15 805-1020-300gasHydrogen 2570-1107-12100–1000cooledForced 100-20010-30200-25000cooledPhase change500–50000II. COOLING METHODSA. Natural passive coolingNatural cooling uses the on-site energy, combined with theconfiguration of motor components to dissipate heat. Thehousing is the main path through which the heat is removedfrom inner components to the ambient environment. Thedesign of the housing needs to be optimized in order tomaximize the rate of convective heat dissipation.In practice, correlations of convective heat transfer (HTC) ℎhave been developed for natural cooling to show that theNusselt number 𝑁𝑢 mainly depends on the Grashof number𝐺𝑟 and the Prandtl number 𝑃𝑟 [21], defined as equations (14):𝑁𝑢 ℎ 𝐿/𝜆(1)𝑁𝑢𝑁𝑎𝑡𝑢𝑟𝑎𝑙 𝑓(𝐺𝑟, 𝑃𝑟)2𝐺𝑟 𝑔𝛽𝜌 (𝑇𝑤 𝑇𝑓 )𝐿𝜇2𝑃𝑟 𝜇 𝑐𝑝 /𝜆(2)3(3)(4)A suitably designed finned housing can improve the heattransfer coefficient value as compared to a non-finnedhousing. The cooling fins are normally placed on the surfaceof the housing, and are oriented in such a way as to not disturbthe natural airflow. There are two types of fins branching offin different directions relative to the motor shaft: one being aradial fin array [22], the other an axial fin array [23].The heat transfer rate from fins to the ambient environmentmay rise either by increasing the heat transfer coefficientand/or the fin surface area. However, the natural convectionheat transfer coefficient depends on the ambient conditions. Acommon practice for improving the natural convection heattransfer is to extend the fin area; however, this increases theresistance of the air flow which in turn diminishes the gain0278-0046 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See standards/publications/rights/index.html for more information.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TIE.2018.2835397, IEEETransactions on Industrial ElectronicsIEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICSfactor. The optimization of the fin extension 𝐻, fin pitch 𝑆 andthe number of fins 𝑁 which are illustrated in Fig. 1, are themain ways to increase the natural cooling performance [2429]. The key design objective must be to maximize the rate ofthe heat dissipation, while minimizing the weight and volumeof the cooling fins.Fig. 1. Fin configuration geometry.The natural cooling approach, however, is appropriate onlyfor low or medium-power motors or large electric motors withsufficient heat transfer areas.B. Forced coolingForced cooling is a more popular approach than the passivetype discussed in Section A, as more power dense andcompact motors are being introduced to the market. Ascompared to the natural cooling, the forced cooling uses anexternal device and source of energy to create sufficientcoolant flow to exchange and extract the heat from the hottercomponents. The Reynolds Number 𝑅𝑒 is used to determinethe flow patterns, so-called flow regimes, under differentcooling media and architectures, and can be analyticallyestimated by (5). The heat transfer based on the forcedconvection method can be defined as a function of 𝑅𝑒 and 𝑃𝑟in accordance with (6) [21],𝑅𝑒 𝜌𝐷𝑉𝜇𝑁𝑢𝐹𝑜𝑟𝑐𝑒𝑑 𝑓(𝑅𝑒, 𝑃𝑟)mounted fan is limited by the speed of shaft. Hence anexternal fan or blower is employed to generate the optimallevel of air pressure independently of the shaft speed. For anEFC approach, the recirculated air is often cooled via theambient air through the external frames in case of the smallmotors or by an air-to-water heat exchanger in case of thelarge motors. The key benefit of the EFC scheme is that theinterior parts are better protected against pollutants which mayblock the ventilation ducts, with the risk of impeding theairflow. Furthermore, the cooling performance can beimproved by replacing the air with a suitable gas that hashigher heat conduction and higher specific heat capacity thanair, e.g. hydrogen [32]. This is owing to the fact that thesmaller and lighter gas molecules can result in a lowerwindage loss and better heat transfer than air.An OFC motor ventilation structure is illustrated in Fig. 3.The coolant air is continuously drawn from the ambientenvironment into the motor enclosure, and not re-circulated.Since, in this method, the motor is exposed to theenvironmental contaminants, provisions such as using filteringor employing indirect air channels need to be in place toprevent particles or moisture from entering the motor [33].Because of the accumulated pollutants, OFC motors areregularly dismantled for a clean-up operation once every twoor three years [34].Fig. 2. The ventilation structure of an EFC motor.(5)(6)1) Forced airIn a forced air cooling system, a fan or a blower isemployed to generate the continual passage of air through amotor or over its exterior. Depending on the enclosure of amotor, forced air can be divided into two different varieties: anenclosed fan cooled (EFC) motor and an open fan cooled(OFC) motor.An EFC motor consists of an inner and an outer flowcircuit. These are displayed in Fig. 2 [30, 31]. Therecirculating air from the inner circuit brings heat from theinner motor to the housing frame, where an outer flow circuitfunctions as a heat sink. The EFC configuration prevents afree exchange of air between the inside and the outside of themotor. An internal fan, either integral to the rotor or mountedon the shaft, circulates air inside the enclosure which promotesthe heat transfer to the frame. An exterior fan, makes thesurrounding air pass over the housing, thus removing heat tothe ambient environment. However, the efficiency of a shaft-Fig. 3. The ventilation structure of an OFC motor.Fig. 4. A typical fan characteristic curve [35].The cooling performanc

technologies and computation methods for the automotive traction motors. Various cooling methods, including the natural, forced air, forced liquid and phase change types, are discussed with the pros and cons of each method being compared. The key factors for optimizing the heat transfer efficiency of each cooling system are highlighted here. Furthermore, the real life examples of these methods ...