electronic cooling simulation

Copyright © 2002 Fluent Inc. JA173• of 4

J O U R N A L A R T I C L E S B Y F L U E N T S O F T W A R E U S E R S

Electronics cooling simulation is helping a majoraerospace manufacturer reduce design time whileminimizing system weight. Thermal management isimportant for safe and reliable operation of allelectronic equipment, and is especially so forairborne systems. Yet it can take several months andcost on the order of $10,000 to build and test a singleprototype for certain equipment. For HamiltonSundstrand, a maker of power modules for aerospaceand marine applications, these obstacles have beenovercome by using computational fluid dynamics(CFD) software to evaluate the effectiveness ofdifferent designs. By using simulations to betterunderstand the airflow within each proposed device,engineers can determine the best methods for heatremoval before the first prototype isbuilt. In this manner, the companycan often find novel ways toimprove thermal management whilereducing component weight.

During the past several years in allsectors of the electronics business,system functionality has beensteadily increasing while system sizehas been steadily decreasing. Thecombination of these trends has ledto a steady increase in the amount ofheat generated per unit volume.Removing internally generated heatrequires an effective path along

which the heat can flow from the heat-dissipatingcomponents to the surroundings. To meet this need, avariety of cooling techniques is available to designengineers including, for example, conduction, naturalconvection, forced-air cooling, radiation cooling, andliquid cooling. Selecting and eventually optimizingthermal management has relied on, traditionally,building a series of physical prototypes usingdifferent cooling methods, and measuring theperformance of each.

Hardware vs. software prototypingUnfortunately, the traditional approach is time-consuming and it is difficult to build physicalprototypes during the early stages of the design,

Electronic Cooling Simulation HelpsReduce Design TimeBy Dr. Samir El-Khabiry

Hamilton Sundstrand

Rockford, Illinois

JA173

Figure 1: CFD model of motor control module

Copyright © 2002 Fluent Inc. JA1173 • of 4

when other issues are being tackled. The constructionand testing of many prototypes is often needed tomeet a stringent design requirement. This can turninto an expensive process with the potential to delaythe entire development cycle. Another problem is thatalthough building and testing prototypes can yieldaccurate thermal performance measurements, it shedslittle or no light on the internal flow and heat transferconditions that determine why the design does ordoes not work. As a result, engineers obtain very littleinformation from each test and have to proceedmainly on instinct.

In order to address these issues, Hamilton Sundstrandengineers have been using computer simulations for anumber of years to create virtual prototypes of theirconcept designs and evaluate thermal managementwithout the time and expense required for physicalprototyping. Computational fluid dynamics (CFD)software allows users to build models that simulatethe flow conditions within an enclosure, making itpossible to evaluate virtual prototypes on a computer.Virtual prototyping can be performed at a muchlower cost and in much less time than physicalprototyping, and has the additional advantage thatengineers can determine important flow variables,such as velocity, pressure, and temperature at anypoint in the design, making it easier to optimizethermal performance.

One question facing any engineer who wants tocreate a virtual prototype is whether to use a generalpurpose CFD software package, or to use softwaredesigned specifically for electronics cooling analysis.The advantage of using software designedspecifically for electronics cooling is that it reducesmodeling time by providing a library of commonelectronic components that can be used to streamlinethe modeling process. The advantage of generalpurpose CFD software is that it can handle a widerrange of geometries and physical models. AtHamilton Sundstrand, engineers use both methodsand select the best one for each individualapplication, based on its operating characteristics.

Optimizing the thermal performanceof a moduleIn one typical example, the objective was to optimizethe thermal performance of a motor drive module fora flap/slat control unit. This product is used to extendthe flaps and slats of a commercial aircraft duringtakeoff and landing, and retract them when they areno longer needed. The module fits inside the wings ofthe aircraft and includes a powerful motor, withcontrol circuitry that dissipates a large amount ofheat. The initial concept design of the module wasgenerated using Pro/ENGINEER computer aideddesign software. The complexity of the geometrymade it desirable to use a general purpose CFD code,so FLUENT, from Fluent Incorporated, Lebanon,

New Hampshire, was chosen toconstruct the virtual prototype.Engineers imported the geometryinto GAMBIT, FLUENT'spreprocessor, which automaticallygenerated the mesh. Only minoradjustments were required tocorrect mesh elements withexcessive skewness that mightimpact the accuracy of the results.

The box containing the controlmodule in this example is cooledwith forced air driven by a fan.The characteristic fan curve wasentered into the CFD model andused as a boundary condition forthe analysis, along with the heatsources from the variousFigure 2: Temperature distribution inside the motor control module


components. Oncecompleted, the simulationpredicted air velocity andtemperature distributionswithin the enclosure. Theseresults were used to selectcomponents for the modulethat are able to withstand thethermal conditions predictedby the analysis. By usingCFD, it was possible toselect components that werewell suited for theapplication from a thermalstandpoint, while it couldhave taken months and tensof thousands of dollars ofhardware prototypes toachieve the same goal usingphysical testing.

CFD analysis was also performed on another motordrive module. While similar to the first, the secondmodule uses natural convection as a coolingmechanism rather than forced convection. Becausethe geometry was again very complex, FLUENT wasdeemed to be the right thermal analysis tool. Basedon the analysis, engineers decided not to reduce thesize of the heat sink in the unit, but to utilize the extramass as heat capacitance to support the naturalcooling mechanism.

Electrohydraulic drive unit thermalmanagementHamilton Sundstrand engineers also faced thechallenge of designing an electrohydraulic drive unit(EHDU) that includes an integrated 65-kilowattvariable speed permanent magnet electric motor and a35-gallon per minute (at 3000 psi) hydraulic pump.The highly efficient design of this unit is achievedthrough a variable speed/variable displacementcontrol strategy, with reliance on design principlesadapted from Hamilton Sundstrand's long experiencein power systems.

Because the geometry was less complex in thisexample, engineers used Icepak, another softwarepackage from Fluent Incorporated that is specificallydesigned for electronics cooling analysis. It differs

from general-purpose CFD codes in its use of"objects" such as blocks, plates, vents, openings,fans, resistances, and sources. Objects, which arepre-defined 3D solid models, make it easier togenerate system-level models. Users simply pick acomponent and enter its characteristics, such asthermal conductivity or specific heat, rather thanmodeling it from scratch, as would be done withgeneral-purpose CFD software.

Dissipating 3 kilowatts with forced aircoolingThe concept design for the EHDU used forced air-cooling instead of liquid cooling for a particularapplication in order to save weight, space, andpower. This approach can be challenging from athermal management standpoint and the initial designshowed that a very large heat sink would have beenrequired to maintain junction temperatures atacceptable levels. The Icepak results also showed theexistence of recirculation zones that restricted thetransfer of heat to the outside of the enclosure.Engineers rearranged the positions of some of thecomponents in order to eliminate these recirculationzones and improve the airflow inside the unit. Thechanges were so successful that it became possible touse a heat sink only one-third as large as thatrequired by the initial concept design. The CFDanalysis also helped engineers select and position thefans. The resulting design was able to dissipate over3 kilowatts of power without exceeding the

Figure 3: Temperature distributioninside the controller of the EHDU


maximum temperatures of the electronic componentsinside the unit.

Virtual evaluation of replacing electroniccomponents on a printed boardWhen it became necessary to replace severalcomponents on a printed circuit board of anothermotor control unit, Icepak was again used todetermine the impact of the change on the thermalconditions inside the enclosure. The analysis showedthat the additional heat generated by the newcomponents raised temperatures beyond acceptablelevels. Several alternative designs were evaluatedusing CFD, and an effective cooling mechanism wasidentified and applied to the actual board.

As these examples show, both general-purpose andelectronics cooling CFD software can help optimizethe design of electronic equipment without the needfor prototyping. A key advantage of both softwaretools is that their postprocessing capabilities make itpossible to visualize airflow and heat transfer insidethe enclosure. Understanding the behavior of anexisting design usually makes it possible to iteratequickly to an optimized design. The net result is asignificant reduction in the time and cost required toresolve thermal management issues and bring aboutperformance improvements.

Figure 4: Board temperature distribution in the EHDU

electronic cooling simulation

Documents