IEEE TRANSACTIONS ON COMPONENTS AND PACKAGING TECHNOLOGIES, VOL. 25, NO. 4, DECEMBER 2002 533

Enhanced Electronic System Reliability—Challengesfor Temperature Prediction

John D. Parry, Member, IEEE, Jukka Rantala, and Clemens J. M. Lasance, Associate Member, IEEE

Abstract—Using telecommunication as an example, it is arguedthat the electronics industry badly needs a change in attitudetoward reliability thinking. The role of thermal design and reli-ability qualification is discussed in context to current industrialneeds for short design cycles and rapid implementation of newtechnologies. Current and future practices are discussed in thecontext of newly-emerging reliability standards. Finally, twomulti-company projects targeting the improvement of reliabilitythrough better temperature-related information are described.

Index Terms—Physics of failure, prediction, reliability, stan-dards, telecommunication, thermal.

I. INTRODUCTION

T O INTRODUCE the reliability requirements we face forthe future, we have chosen to focus on telecommunica-

tion as an example. Personal telecommunication is becoming in-creasingly integrated into our daily activities. Broadband mobilenetworks promise high-speed web access from anywhere in theworld. Ubiquitous computing brings us connection to local net-works and in the future will connect various sensor systems inour homes. However, full advantage of such technology can notbe realized unless the telecommunication system is as depend-able as a car. Just as turning the ignition key should produce theright engine response first time, every time, so should the “con-nect” button on a mobile phone. If the connection is lost severaltimes a day, the system won’t be fun to use, and users won’tbe convinced that financial transactions are being processed se-curely. In the future, personal trusted devices (PTDs) combiningall the functions of a phone, organizer, secure web browser forshopping and personal finance, electronic cash, credit card, IDcard, driver’s license, and keys to car, home, and work place willbe technically possible. To gain widespread acceptance, how de-pendable does such a device and the infrastructure that supportsit need to be?

Telecommunication system integrators now have to push thelimits of the currently known technologies to create new prod-

Manuscript received September 27, 2001; revised February 19, 2002. Thiswork was supported in part by the EC under Contract IST-1.999-I2529 PROFIT.This work was presented in part at the Seventh Therminic Workshop, Paris,France, 2001. This work was recommended for publication by Guest EditorsC. Lasance and M. Rencz upon evaluation of the reviewers’ comments.

J. D. Parry is with Flomerics Ltd., Surrey KT8 9HH, UK (e-mail:john.parry@flomerics.co.uk).

J. Rantala is with Nokia Research Center, Nokia Group, Finland FIN-00045(e-mail: jukka.i.rantala@nokia.com).

C. J. M. Lasance is with Philips Research Laboratories, Eindhoven 5656AA,The Netherlands (e-mail: clemens.lasance@philips.com).

Digital Object Identifier 10.1109/TCAPT.2002.808001

ucts. To date the computer industry has been the first sector toutilize new technologies, but increasingly the telecommunica-tion industry is taking the lead despite the distinction becomingblurred. Heavy competition with short design cycles forces theuse of technologies before there is adequate experience of theirfield reliability. Increasing component power consumption andhigher data clock frequencies of digital circuits force the de-sign into smaller tolerances, and drive the demand for methodsto predict technology and system reliability through simulation,augmented by accelerated laboratory tests. Even in consumerelectronics where especially audio and video products have en-joyed relatively large design margins with respect to reliabilityand performance, products are being designed closer to theirlimits, forced by device miniaturization and reduction in systemvolume. Reliability prediction has to be linked to overall riskmanagement, providing estimates of how big a reliability riskis taken when a new technology, without previous field expe-rience, is used. From this, system integrators can compare themonetary benefits from increased sales and market share againstthe possible warranty and maintenance costs.

Often the use of new technologies forms a substantial con-tribution to the overall risk involved in the product creationprocess. The use of chip scale packages and flip chips in handheld devices such as mobile phones and palm top computers isdriven by the demand for miniaturization. This drives the in-dustry infrastructure changes needed to produce and assembleboards capable of supporting such high density interconnecttechnologies in high volume and at low cost. Where there isno field experience in the use of a new technology and shortdesign cycles prevent extensive testing, physics-of-failure reli-ability prediction gives estimations of the reliability.

Currently emerging virtual prototyping and qualificationtools simulate the effect of mechanical and thermomechanicalstresses on reliability. For a subsystem or a single part thereliability in a specified environment can be predicted ratheraccurately. However, the prediction accuracy decreases whenthe whole system is considered, or when the geometry, materialproperties, use profile or operating environment are not prop-erly known. More work needed in simulation tool development,tool integration, model improvements, and in capturing data onthe reliability loads the system will encounter [1]. This articlediscusses the “knowledge gap” between temperature/stressanalysis and system lifetime assessment.

The technical aspects are mainly based on work performedby the CALCE Electronic Products and Systems Center, Uni-versity of Maryland (e.g., [2]–[4]). This article discusses howthese principles can be applied in practice.

1521-3331/02$17.00 © 2002 IEEE

534 IEEE TRANSACTIONS ON COMPONENTS AND PACKAGING TECHNOLOGIES, VOL. 25, NO. 4, DECEMBER 2002

II. COOLER ISBETTER?

The effects of temperature on electronic device failure havebeen mainly obtained through accelerated testing, during whichthe temperature, and in some cases power, are substantiallyincreased to make the test duration manageable. This data isthen correlated with actual field failures. MIL-HDBK-217,“Reliability Prediction of Electronic Equipment,” which con-tains failure rate models for different electronic components isbased on this correlated data. The total reliability calculationsare then performed either by “parts count” or “part stress”analysis. It has been updated many times with the last version,MIL-HDBK-217F Notice 2, published in 1995. Although nowdefunct, its basic methodology is the foundation for manyin-house reliability programs still in use and has been adaptedby Bellcore for telecommunications applications.

The basis of the handbook is the assumption that many of thechip level failure mechanisms that occur under accelerated testconditions are diffusion-dominated physical or chemical pro-cesses, represented by an Arrhenius-like exponential equation.This relationship is then used to predict failures under opera-tional conditions. Doing so assumes that failure mechanisms ac-tive under test conditions are also active during operation, andthat the above relationship holds at lower temperatures, giving adirect relationship between steady-state temperature and relia-bility. This is substantially incorrect, since some failure mecha-nisms have a temperature threshold below which the mechanismis not active, whilst others are suppressed at elevated tempera-tures. In the temperature range55 C to 150 C, most ofthe reported failure mechanisms are not due to high steady-statetemperature. They either depend on temperature gradients, tem-perature cycle magnitude, or rate of change of temperature [5].Considerable care is needed to ensure that the test conditions ac-celerate the principal failure mechanisms expected to be presentduring use, without suppressing or introducing others to thepoint where the results of the test are invalid, and to ensure thatmaterial property limits are not exceeded.

Straightforward application of the Arrhenius model has ledto widespread misconceptions that are then followed blindly.An example is the “10 C rule,” being that the life of a com-ponent doubles, for every 10C the steady-state temperatureis dropped. Although this holds for some failure mechanisms,in reality the life of the part is more likely to depend on thenumber of power on/power off cycles it experiences. Even whenthe exact failure mechanism is known, the use of the Arrheniusmodel contains uncertainties because it is very sensitive to thevalue of the activation energy used in the exponential term. Therange of activation energy for the same failure mechanism canvary by more than a factor of two, depending on the part design,materials and fabrication processes. Due to the exponential re-lationship, the predicted mean-time-to-failure (MTTF) can varyby a factor of 20.

An additional fundamental difficulty in using MIL-HDBK-217 type models for new and emerging technologiesand components is the lack of a wide, environmentally relevantdatabase of test data and experience of field failures. Thereforethe MTTF calculations would be based on many assumptions,the validity of which is not known.

In 1993, the U.S. Army Material Systems AcquisitionActivity and CALCE began working with the IEEE ReliabilitySociety to develop an IEEE Reliability Prediction Standardfor commercial and military use. The standard is based onphysics-of-failure approaches to reliability and life cycleprediction [2]. Space restrictions prevent a lengthy description,but Fig. 1 shows how the environmental stress on a systemand its operational performance are combined to providelifetime information using data acquired from a spectrum ofstakeholders.

In the physics-of-failure approach, user-defined load and en-vironmental conditions are used in combination with layout andother input data and physics-based failure mechanisms to resultin a ranking of most probable failures. While it is realized thatmany data are still lacking, the philosophy behind the approachhas a sound basis.

The above-mentioned work has culminated in two IEEEstandards [6]. The first, “IEEE Standard Reliability Programfor the Development and Production of Electronic Systems andEquipment” (IEEE Std. 1332-1998) was developed to ensurethat every activity in a product reliability program adds value.In brief, IEEE 1332 identifies three objectives for a reliabilityprogram, based on how the supplier shall determine, meet, andverify the customer’s requirements and product needs. Thesecond, “IEEE Standard Methodology for Reliability Predic-tion and Assessment for Electronic Systems and Equipment”(IEEE Std. 1413-1998) identifies the required elements for anunderstandable, credible reliability prediction, and to providesufficient information for the effective use of the results, thusrequiring thorough documentation of the reliability assessment.

The likely future implications of these standards for the elec-tronics industry are not entirely clear. However, one can imaginetheir impact will be similar to the introduction of the ISO9000Quality Management System standard. Customers asked theirsuppliers whether their internal quality management systemswere ISO9000 certified, and started to show a strong prefer-ence for certified suppliers. Indeed, the impact of IEEE 1332 andIEEE 1413 may be more profound, since they have a much moredirect impact on what the customer is interested in (the quality ofthe supplier’s product) than ISO9000 does. IEEE 1332 requiresthe supplier to work closely with the customer, so the supplier’scompliance with IEEE 1332 should be readily apparent. How-ever, customers need to ask whether the reliability assessmentof the product they are considering purchasing is IEEE 1413compliant, and if so ask for the supporting documentation. “No”is unlikely to remain an acceptable answer for very long, withcustomers favoring suppliers who can supply reliability assess-ments for the operation of their products and the data to supporttheir claims.

The physics-of-failure approach requires that the root causeof the failure is identified. The difficulties inherent in this can beillustrated by considering die attach, wire bond and solder jointfatigue, as examples of package and interconnect failure mech-anisms. Wire bond fatigue might be due to delamination at thedie surface, while misregistration of the part during placement,or inadequate solder paste could reduce the life of the packageinterconnect. “Latch up” is an operational malfunction that can

PARRY et al.: ENHANCED ELECTRONIC SYSTEM RELIABILITY 535

Fig. 1. Physics-of-failure approach as proposed by CALCE Research Center.

occur in CMOS technology as a result of high junction temper-ature. The root cause of latch up could be die attach delamina-tion, resulting in a significantly increased die junction temper-ature. Alternatively, it could be caused by electrical overstress.Unfortunately, many early failures with a new technology, forexample process-related manufacturing problems, are difficultto detect using the physics-of-failure approach. The above-men-tioned IEEE standards take this difficulty into consideration bydemanding the basic assumptions of root causes are documentedand brought to the attention of the customer, so that their rele-vance can be assessed.

To close this section, the reader should be aware of the dif-ference between electronic systems that depend on processorsand those that do not. The need to increase the performance ofcomputers has driven many novel cooling concepts, becauseof the inverse relationship between absolute temperatureand clock speed. Many audio and video consumer products,however, are dominated by reliability issues, not performance.For these products, currently the dominant reliability problemsare not related to components but to interconnects, especiallysolder joints. This problem is aggravated by the introductionof lead-free solders, mainly because of the lack of reliablethermal and mechanical data. Furthermore, ongoing researchreveals that the relevant mechanical properties change consid-erably over time. It is striking to observe that no internationalstandards exist to specify the maximum allowable values of

known stressors such as the creep strain energy dissipated peroperational cycle. In short, in many cases products are testedusing standards that don’t make sense, while no standardsaddress the real causes of many of the problems that occur.

III. ROLE OF THERMAL DESIGN AND RELIABILITY

QUALIFICATION IN PRACTICE

Thermal design has grown in importance over the past fewyears, as the technical challenges and costs of cooling have in-creased. Central to thermal design is thermal analysis. Thermalanalysis allows designers to quickly examine a range of coolingscenarios, and can ensure that catastrophic design mistakes arenot made (e.g., placing a strip cable connector upstream of ahigh powered or thermally-sensitive component). Over 90% ofthe thermal analyzes performed on electronics products over thelast few years have been steady-state based. In part this has beendriven by the focus on steady-state temperature from a perfor-mance and reliability standpoint, and thermal design for contin-uous operation. It is also relatively quick, compared to a tran-sient analysis. Transient analyzes take more time, so are not per-formed unless the results can be used to further some businessobjective, such as reduced cost or higher reliability. Generally,this is not possible at present.

Thermal “uprating” is the assessment of a part to meetfunctional and performance requirements when used outsidethe manufacturer-specified temperature range [7]. Despite the

536 IEEE TRANSACTIONS ON COMPONENTS AND PACKAGING TECHNOLOGIES, VOL. 25, NO. 4, DECEMBER 2002

traditional practice of thermal derating (e.g., [8]), upratinghas recently become interesting as a means of achievingcompetitive advantage, despite the legal issues it raises relatingto warranty. This kind of design trend comes from the need touse more powerful and faster components in very demandingenvironments whilst avoiding over design. To estimate therisks being taken with a given design, the extent to which themanufacturer’s limits on temperature are exceeded must beaccurately known [7]. This presents a further challenge for thethermal analysis.

Despite huge progress, the accuracy of thermal analysis re-mains seriously limited by the following.

Environmental Uncertainties

1) Limited knowledge of the actual operating environmentof the equipment.

2) The actual system power on/power off cycle and loading.3) Information on the peak and normal component operating

powers, etc.Limited Design Data

1) Lack of availability of thermal models of componentssuitable for design calculations.

2) Unknown material properties, or material property dataof unknown accuracy.

3) Unknown model accuracy and sensitivity.4) Inherent errors associated with the use of CFD analysis

of complex geometries [9].

The above classification reflects a natural, if somewhat ideal-ized, division of responsibility. The system integrator is respon-sible for resolving environmental uncertainties through discus-sion with the customer/user of the equipment. The componentsupplier is responsible for providing behavioral models of thesupplied components to allow the equipment manufacturer tomodel their performance within the system.

When design cycles are fast and new technologies are contin-ually being adopted, a virtual qualification approach is needed,so that design weaknesses can be identified by predictionand removed before the prototype used in highly acceleratedlife testing (HALT) is created—the “design–fix–build–test”methodology described in Fig. 1. Although experimental testsare used to verify the design, their number can be greatlyreduced by numerical simulation. Again, this places greaterreliance on simulation, requiring better tool support andbetter thermal and thermomechanical data, better models andbetter material properties. At present, analyzes performed toinvestigate the impact of design changes on the thermal, elec-tromagnetic compatibility, and thermomechanical behavior ofthe system are either performed separately, or are not performedat all due to the time required to replicate design changes acrossa range of tools. The ability to perform such analyzes concur-rently would allow design tradeoffs to be investigated muchfaster and at minimum cost. The long-term aim is to identifyweaknesses during design and use the knowledge to improvedesign practices. Ultimately this will provide designers withan analysis environment in which physical designs of knownreliability can be created.

We are still far away from a method that enables designersto address reliability requirements in a logical and user-friendly

way. In order to realize this goal, progress is needed in the fol-lowing.

1) Accuracy of thermal analysis.2) Accuracy of the reliability assessment based on

thermal/thermomechanical input.3) Knowledge of all active failure mechanisms.4) System level thermal analysis to address applica-

tion-driven stressors.5) Reliable field data.6) Failures caused by design errors distinguished from those

related to reliability.7) Physical understanding of various notoriously difficult

failure mechanisms, e.g., solder joint reliability, wirebond lift-off, wire bond heel cracking etc.

8) Realistic user-defined environmental profiles.

This is a massive task, and can only be achieved through the co-operative efforts between experts in many different disciplines.Progress is often hampered by “over the wall” design cultures.Also, electronics designers are often held responsible for reli-ability calculations. While the electronic engineer is certainlyresponsible for the power dissipation, the mechanical engineershould be responsible for the resulting temperature rise, and ul-timately reliability.

Apart from correct input data, accurate reliability analysis de-pends on the accuracy of a whole range of separate tools thatneed to be combined are thermomechanical, EMC, vibration,humidity, and thermal. One example is shown in Fig. 2, showinga first-order thermomechanical analysis result of a package ona board at the system level.

IV. A BOUT THE PROFIT PROJECT—SHRINKING THE

KNOWLEDGE GAP

Recognizing the need to improve the data and methodsused in physical design, the European Community has fundedthe PROFIT project (Prediction of Temperature GradientsInfluencing the Quality of Electronic Products, [10]). It had anumber of activities directed towards the main aim: to providethe designers who are responsible for yield improvement,performance, reliability, and safety, with reliable and accuratetemperature-related information The main goal of the projectis to develop methodologies for the analysis of time-dependentdata. Much progress in the development of “design-centric”thermal models has been made in recent years. Most notablehas been the development of steady-state compact thermalmodels for chip packages and other electronic components.These simple models capture the thermal behavior of the partto a high degree of accuracy and offer tremendous potential forvirtual design qualification.

Major improvements have been made to the techniques usedto acquire the input data required for accurate numerical anal-ysis: interface resistances, emissivities, local boundary condi-tions and local board thermal conductivities. Experimental testset ups have been built to exploit the use of transient temper-ature measurements, and novel nonlinear parameter estimationmethods will be used to analyze the data. Optimization and otherstatistical challenges have been supported by CQM (NL). Novel

PARRY et al.: ENHANCED ELECTRONIC SYSTEM RELIABILITY 537

Fig. 2. Peak solder joint shear strain for a CBGA on a four-layer PCB under normal operating conditions.

test dies [TIMA (F) and Budapest University (HU)] have facil-itated the detailed time-dependent analyses of packages.

Existing thermal analysis software has been improved, andthermal and thermomechanical analysis software developedand integrated to facilitate the application of the project resultsin performance and reliability calculations. Thermomechanicalfailures, especially related to interconnects are of great interest.Nokia (FIN), Philips (NL), and Flomerics (UK) have investi-gated the benefits of using detailed system-level temperatureinformation predicted for the equipment during operation todrive time-dependent thermomechanical stress calculationsand hence predict system reliability. In doing so, the partnershave produced failure and lifetime predictions that are morerepresentative of the equipment’s field operation, taking intoaccount the normal use cycle and changing environmental loadsusing a demonstrator constructed by Nokia. This demonstratesthat it is possible to make this an integral part of the productdesign process, providing the potential to perform lifetimepredictions at each design iteration.

The participating semiconductor manufacturers [ST Micro-electronics (I), Infineon Technologies (D), and Philips Semi-conductors (NL)] will use the project results to better under-stand the actual reliability of their products in different environ-ments, and through modelling be able to improve the thermaland thermomechanical behavior of their components. This willresult on one hand to better usable thermal data for end users,and on the other hand to the yield improvement through betterdefined rejection criteria based on in-line quality testing usingtransient temperature measurements. The thermal software ven-dors [Flomerics and MicRed (HU)] are using the results to de-velop their tools to support their customer’s physical design ac-tivities so that a more holistic view of the performance of thesystem can be obtained. This involves combining the simula-tion of thermal performance with EMC and thermomechanicalstress. Web-centric tools are also being developed to support theprovision of data and models.

Standardization was considered an important project deliv-erable to ensure that the results are ultimately suited for imple-mentation in emerging virtual prototyping methods and physics-based reliability analysis software. To this end the project con-sortium has worked with the IEEE JEDEC JC15.1 Committee toformulate proposals for the standardization of dynamic compactthermal models, which has resulted in two guideline proposalsfor steady-state compact thermal models.

V. RELIQUI PROJECT

Parallel to PROFIT, a separate project, in which Philips,Nokia and CALCE participate, has been started to address anumber of reliability issues. The project is called “RELIQUI,”an acronym of “Reliability and Quality Integration.” Theultimate goal is to provide a significant contribution to betterproducts at a lower cost, and in less time. The objective willbe realized through the development of virtual qualificationand prototyping tools, enabling the assessment of yield im-provement, performance, reliability and safety (collectivelyreferred to hereafter as “quality”) in every stage of the designprocess. Because quality prediction comprises many aspectsthat are related to general topics such as the development ofsoftware tools, physics-of-failure and standardization, it makessense to cooperate with other industries to share the costs andknowledge.

More specifically, the following sub-objectives can be distin-guished.

1) Exploring methods to enable Design-to-Limits.2) Virtual qualification enabling earlier choice between al-

ternative designs.3) Improving physics-based understanding of final product

reliability.4) Realizing faster test methods.5) Understanding the relationship between accelerated tests

on parts and system performance.6) The search for realistic temperature specifications.

538 IEEE TRANSACTIONS ON COMPONENTS AND PACKAGING TECHNOLOGIES, VOL. 25, NO. 4, DECEMBER 2002

The project focuses on the quality assessment of complete prod-ucts, which distinguishes the project from many other quality-driven projects that focus on some part of the system only (com-ponent, interconnect, board). The proposed project differs alsoin the use of operational conditions rather than standardized testconditions.

Conventionally the reliability testing has been based on ac-celerated testing by thermal cycling. At the moment more elab-orate highly accelerated life testing (HALT) and highly acceler-ated stress screening (HASS) procedures are gaining more im-portance in this. HALT is a procedure used to expose a board,sub-system or system to temperature, temperature change andmulti-axis vibration across a wide frequency range to exposedesign weaknesses. Once found, the cause of failures are re-moved through redesign. The redesigned equipment is re-testedand the process repeated until there is sufficient confidence inthe product design’s life-cycle reliability. HASS is performedafter HALT. It involves, for example, accelerated cycling of aproduct during the production phase to assure the productionprocess and confirm the final design.

HASS and HALT have proved to be effective tools for qual-ifying and improving the reliability of a product. However, de-spite their success, these techniques test the product under con-ditions that are very different from those the product will expe-rience in the field, so that care is needed in the interpretation ofthe results. If not done well HASS can indicate a much-reducedoperational lifetime.

Another prominent feature of the project is the possibilityto test many products, enabling statistically significant conclu-sions. Finally, the project will investigate coupling the results ofsystem-level thermal analysis software to reliability software,thereby including the various results from the PROFIT project.

VI. CONCLUSION

The electronics industry is facing a paradigm shift in theapproach to product quality. The integration of design-centricanalysis technologies is required to facilitate virtual qualifica-tion using the physics-of-failure based “design–fix–build–test”methodology for reliability prediction. In this context thermalmanagement is expected to continue to play a key role.

REFERENCES

[1] M. Lindell, P. Stoaks, D. Carey, and P. Sandborn, “The role of phys-ical implementation in virtual prototyping of electronic systems,”IEEETrans. Comp., Packag., Manufact. Technol. A, vol. 21, pp. 611–616, Dec.1998.

[2] M. Pecht, “Why the traditional reliability prediction models do notwork—Is there an alternative?,”Electron. Cooling, vol. 2, no. 1, pp.10–12, 1996.

[3] D. Das, “Use of thermal analysis information in avionics equipment de-velopment,”Electron. Cooling, vol. 5, no. 3, pp. 28–34, 1999.

[4] M. Osterman, “We still have a headache with Arrhenius,”Electron.Cooling, vol. 7, no. 1, pp. 53–54, 2001.

[5] P. Lall, M. Pecht, and E. Hakim,Influence of Temperature on Microelec-tronics and System Reliability. Orlando, FL: CRC Press, 1997.

[6] [Online]. Available: http://standards.ieee.org/catalog/reliability.html.[7] D. Humphrey, L. Condra, N. Pendsé, D. Das, C. Wilkinson, and M.

Pecht, “An avionics guide to uprating of electronic parts,”IEEE Trans.Comp., Packag., Manufact. Technol. A, vol. 23, pp. 595–599, Sept. 2000.

[8] M. Jackson, P. Lall, and D. Das, “Thermal derating—A factor of safetyor ignorance,”IEEE Trans. Comp., Packag., Manufact. Technol. A, vol.20, pp. 83–85, Mar. 1997.

[9] C. Lasance, “The conceivable accuracy of experimental and numericalthermal analyzes of electronic systems,”IEEE Trans. Comp. Packag.Technol., vol. 25, pp. 366–382, Sept. 2002.

[10] , “The European Project PROFIT: Prediction of temperature gradi-ents influencing the quality of electronic products,” inProc. 17th SEMI-THERM Symp., 2001, pp. 120–125.

John D. Parry (M’90) received the B.Sc. degree inchemical engineering from Leeds University, UK,and the Ph.D. degree from Birmingham University,UK, in 1985.

He is a Research Manager with Flomerics, Ltd.,Surrey, UK. He has been with Flomerics since itsinception in 1989, initially in the role of CustomerServices Manager. Since 1993, he has been involvedwith the company’s research activities and was theProject Coordinator for the European-funded Delphiand Seed Projects. He is currently involved in the

European-funded PROFIT Project and a number of nationally funded projects,leading to the creation of FLO/STRESS, a stress analysis module for use withFlotherm. His technical contributions include the development and applicationof thermal compact models, and has published several papers in this area. Priorto joining Flomerics, he was with Cham, Ltd., initially as a Project Engineer,and later as Technical Support Manager.

Dr. Parry is a Chartered Engineer.

Jukka Rantala received the M.Sc. and Ph.D. degreesfrom the University of Helsinki, Finland, in 1988 and1993, respectively.

He is a Senior Research Manager with theNokia Research Center, Helsinki, responsible forhardware integration of electronics, and also asProgram Director for Telectronics (telecommuni-cation electronics research program), Academy ofFinland. Previously, he was with the University ofHelsinki, Neste, Ltd., Institute for ManufacturingResearch, Detroit, MI, Institute for Polymer Re-

search, Stuttgart, Germany, and the Academy of Finland. He has authored orcoauthored more than 50 refereed articles and book chapters, most of themconcentrating on thermal phenomena and thermal characterization.

Clemens J. M. Lasance(A’94) received the Ph.D.degree in physics from Eindhoven Technical Univer-sity, Eindhoven, The Netherlands, in 1969.

He is a Principal Scientist at Philips Research,Eindhoven, where he has been on staff since 1969.In 1980, he took up a post within the Heat TransferGroup, CFT. From 1984 onwards, his main focushas been the thermal management of electronicsystems. In 1996, he moved to research, engagedwith a long-term research program in the field offluid dynamics and heat transfer with a special

focus on electronic parts and systems. He is an active participant in numerousprofessional and global industrial associations and an Associate Editor ofElectronics Cooling Magazine. Currently, he leads a ten-partner consortium inthe EU-funded projects called PROFIT.

Dr. Lasance received the SEMITHERM Significant Contributor Award in2001.