FALL 2020

electronics-cooling.com

25
TH

ANNIVERSARY

Volume 15, Number 1.

February 2009

cooling electronics with nanofluids:

laminar convective heat transfer
mini- and microchannels in thermal
interfaces: spatial, temporal, material,
and practical significance
when Moore is less: exploring the 3rd
dimension in ic packaging

WE'RE CELEBRATING 25 YEARS OF ELECTRONICS

COOLING WITH A LOOK BACK AT SOME OF YOUR
FAVORITE ARTICLES

ALSO FEATURING:

4 | A TRIBUTE TO BRUCE GUENIN

6 | TECHNICAL EDITOR SPOTLIGHT

8 | THERMAL LIVETM 2020 TECHNICAL PROGRAM

Subscriptions to Electronics Cooling are FREE

© Copyright 2020 Electronics Cooling

 Oct o b er 2 0 – 2 1 2 0 2 0

ONLINE EVENT
The Largest Single Thermal Management Event of The Year - Anywhere.

Thermal Live™ is a new concept in education and net­working in thermal management - a

FREE 2-day online event for electronics and mechanical engineers to learn the latest in
thermal management techniques and top­ics. Produced by Electronics Cooling® magazine,
and launched in October 2015 for the first time, Thermal Live™ features webinars,
roundtables, whitepapers, and videos... and there is no cost to attend.

For more information about Technical Programs,

Thermal Management Resources, Sponsors & Presenters

please visit:

thermal.live

Presented by
Alpha’s Online Heat Sink Customization
Quickly and Easily Create a Custom Heat Sink

Note: Note:

Estimated Price Estimated Price

(Without Tax): Add Hole [Edit/Delete] (Without Tax):
Change Base Size [Edit/Delete]
Quantity Unit price (US$) Location C-bore C-bore Base Width Base Length
Quantity Unit price (US$)
ID Diameter Effective Length/ Fin Removal Area from Center Component side FIn side
115 The cut section interferes with fins. Suggested size 65.3mm or 68.8 mm
1 122.73 THRU 1 162.35
X-axis Y-axis Diameter Depth Diameter Depth
5 29.29 5 43.56
1 4.5 Through Minimum fin removal 15 15 - - - - Add Hole [Edit/Delete]
10 10.01 10 28.71 Location C-bore C-bore
2 4.5 Through Minimum fin removal -15 -15 - - - -
50 5.56 50 14.60 ID Diameter Effective Length/ Fin Removal Area from Center Component side FIn side
100 4.31 3 4.5 Through Minimum fin removal -15 -15 - - - - 100 13.70 THRU X-axis Y-axis Diameter Depth Diameter Depth
200 RFQ 4 4.5 Through Minimum fin removal 15 15 - - - - 200 RFQ
1 6 Through 8 15 15 8.00 2.0 - -
Complete Imput Complete Imput
2 6 Through 8 -15 -15 8.00 2.0 - -

Add Tapped Hole

Proto Lead Time No MOQ, No Min.

1-2 Weeks No NRE Fee Order Value Quick and Easy
Typical production In most cases, a tooling Alpha can supply from a single A custom heat sink can
lead time is 2-3 weeks. fee will not be required. piece to production volumes. be designed in minutes.

ALPHA Co., Ltd. 256-1 Ueda, Numazu City, Japan 410-0316 ALPHA NOVATECH, INC. 473 Sapena Ct. #12, Santa Clara, CA 95054 USA
Head Office Tel: +81-55-966-0789 Fax: +81-55-966-9192 USA Subsidiary Tel: +1-408-567-8082 Fax: +1-408-567-8053
www.micforg.co.jp Email: alpha@micforg.co.jp www.alphanovatech.com Email: sales@alphanovatech.com

www.alphanovatech.com
CONTENTS www.electronics-cooling.com

2 EDITORIAL PUBLISHED BY
Lectrix
Bruce Guenin 1000 Germantown Pike, F-2
Plymouth Meeting, PA 19462 USA
Phone: +1 484-688-0300; Fax:+1 484-688-0303
4 TRIBUTE TO BRUCE GUENIN info@lectrixgroup.com
www.lectrixgroup.com
Past and Present Technical Editors Tribute to Bruce Guenin’s
22 Years of Service for Electronics Cooling Magazine CHIEF EXECUTIVE OFFICER
Graham Kilshaw | Graham@lectrixgroup.com

6 TECHNICAL EDITOR SPOTLIGHT VP OF MARKETING

Geoffrey Forman | Geoff@lectrixgroup.com

EDITORIAL DIRECTOR
8 THERMAL LIVETM 2020 TECHNICAL PROGRAM Jennifer Arroyo | Jennifer@lectrixgroup.com

CREATIVE DIRECTOR
Chris Bower | Chris@lectrixgroup.com
A Collection of Articles Featured by the Electronics Cooling Editorial Board
VP BUSINESS DEVELOPMENT TEAM LEADER
Bill Baumann | Bill@lectrixgroup.com
18 ADVANCES IN HIGH-PERFORMANCE BUSINESS DEVELOPMENT DIRECTOR
COOLING FOR ELECTRONICS Janet Ward | Jan@lectrixgroup.com
Clemens Lasance and Robert Simons
PRODUCTION COORDINATOR
Jessica Stewart | Jessica@lectrixgroup.com
30 ESTIMATING PARALLEL PLATE-FIN HEAT SINK
THERMAL RESISTANCE LEAD GRAPHIC DESIGNER
Kristen Tully | Kristen@lectrixgroup.com
Robert Simons
CONTENT MARKETING MANAGER
Danielle Cantor | Danielle@lectrixgroup.com
34 CALCULATING INTERFACE RESISTANCE
ADMINISTRATIVE MANAGER
M. M. Yovanovich, J. R. Culham and P. Teertstra Eileen Ambler | Eileen@lectrixgroup.com

38 CALCULATING SPREADING RESISTANCE IN ACCOUNTING ASSISTANT

Susan Kavetski | Susan@lectrixgroup.com
HEAT SINKS
Seri Lee EDITORIAL BOARD
Bruce Guenin, Ph.D.

42 THERMAL INTERFACE MATERIALS: A BRIEF

Consultant
San Diego, CA
REVIEW OF DESIGN CHARACTERISTICS AND sdengr-bguenin@usa.net

MATERIALS Ross Wilcoxon, Ph.D.

Ravi Mahajan, Chia-Pin Chiu, and Ravi Prasher Associate Director
Collins Aerospace
ross.wilcoxon@collins.com
48 HEATPIPES FOR ELECTRONICS COOLING
APPLICATIONS Genevieve Martin
R&D Manager, Thermal & Mechanics Competence
Scott D. Garner Signify
genevieve.martin@signify.com

45 ADVANCED COOLING IN POWER Victor Chiriac, PhD, ASME Fellow

ELECTRONICS Co-founder and Managing Partner
Global Cooling Technology Group
Sukhvinder S. Kang vchiriac@gctg-llc.com

62 INDEX OF ADVERTISERS ►SUBSCRIPTIONS ARE FREE

Subscribe online at
www.electronics-cooling.com
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, or stored in a
retrieval system of any nature, without the prior written permission of the publishers (except in accordance with the Copyright Designs and Patents Act 1988). For subscription changes email
The opinions expressed in the articles, letters and other contributions included in this publication are those of the authors and the publication of such articles, letters or other contributions
info@electronics-cooling.com
does not necessarily imply that such opinions are those of the publisher. In addition, the publishers cannot accept any responsibility for any legal or other consequences which may arise
directly or indirectly as a result of the use or adaptation of any of the material or information in this publication. Reprints are available on a custom basis at
reasonable prices in quantities of 500 or more.
ElectronicsCooling is a trademark of Mentor Graphics Corporation and its use is licensed to Lectrix. Lectrix is solely responsible for all content published, linked to, or otherwise presented
in conjunction with the ElectronicsCooling trademark.
Please call +1 484-688-0300.

F R E E S U B S C R I P T I O N S
Lectrix®, Electronics Cooling®—The 2020 Fall Edition is distributed annually at no charge to engineers and managers engaged in the application,
selection, design, test, specification or procurement of electronic components, systems, materials, equipment, facilities or related fabrication
services. Subscriptions are available through electronics-cooling.com.
EDITORIAL
Bruce Guenin
Associate Technical Editor

25 YEARS OF ELECTRONICS COOLING

As the “elder statesman” on the Editorial Board of Electronics Cooling, it is my sincere pleasure to have been given
the opportunity to write this milestone editorial. Electronics Cooling has always been of such a scale that its story
must be told in terms of the individuals who shaped it at any given time. Memory can be selective, so I apologize
in advance for any distortions or omissions in my retelling of this saga.

The first issue of Electronics Cooling was published in June 1995, with a publication schedule of three issues per
year. The fact that Electronics Cooling came into existence at all, is due to the vision and dynamism of a single
person. That person is Kaveh Azar, who, at the time, was at ATT Bell Labs in Massachusetts. Contemporaneously,
he had founded the company, Advanced Thermal Solutions, Inc. He was aware that in many small-to-midsize
electronics companies, there would be a thermal engineer or two and a test technician. He also knew that it was quite common for the thermal
engineer to get assigned to the development of product in its very last stages. This was due to the assumption of management that any shortfall in
thermal performance could simply be solved by specifying a more efficient heat sink or a more powerful fan. (Unfortunately, as we well know, life
is rarely so simple.)

Thermal engineers tended to feel isolated in their own organization and lacked a network of thermal engineers beyond their company to help them
expand their skills or knowledge base. He realized that the worldwide community of thermal engineers needed a centralized source of practical in-
formation and know-how to keep up with the rapidly increasing demands of the electronics industry. The core vision he had for Electronics Cooling
was that it would offer content such as technical briefs (describing test or simulation techniques), information on new products (including software,
hardware, and materials), news briefs (highlighting conferences, publications, patents, etc.) and in-depth technical articles (conveying sufficient
detail to the readers that they could apply the results to their own work). In summary, his vision was that Electronics Cooling would become a com-
prehensive source of current and practical thermal management information of archival value.

Due to his entrepreneurial spirit, he convinced others of the significance of this vision, and in short order, had received a commitment from Flo-
merics to support the publication and distribution of a new magazine—Electronics Cooling. He attracted two of the leading thermal engineers of the
day to join this enterprise as Associate Editors: Gordon Ellison, author of the first text book on heat transfer devoted entirely to electronic systems,
and Clemens Lasance, of Philips Research in Eindhoven, who had achieved a prominent profile in the international thermal community as he cam-
paigned for greater rigor in thermal engineering and in the calculation of component reliability, as influenced by their temperature.

Electronics Cooling very quickly became a forum that enabled the leading thermal engineers of the day to share their work with its diverse read-
ership. In fact, Kaveh’s vision was so compelling that, for those who either came on board as an editor or served as a frequent contributor, their
participation invariably became a labor of love. Kaveh had long wanted to have a regular column that he had named “Calculation Corner.” I first got
his attention as a result of my presenting an evening tutorial at SEMI-THERM® dealing with spreadsheet methods of thermal calculations. In 1998,
he approached me with the proposition that I author the column, and I agreed. The first installment of “Calculation Corner” was published in the
September 1998, issue. The column has been a regular feature in Electronics Cooling ever since.

Shortly thereafter we were joined by Bob Simons, who had had a distinguished career at IBM in the thermal engineering of mainframe computers,
and Jim Wilson, an engineering fellow at Texas Instruments, and an expert in the thermal engineering of defense electronic systems. With Kaveh
at the helm and with four associate editors on staff, in the year 2000, Electronics Cooling began publishing four issues per year. Bob and I shared the
“Calculation Corner” column and Clemens and Jim authored the “Technical Data” column. In time, as they were running out of relevant technical
data to publish, Clemens proposed a new column, “Thermal Facts and Fairy Tales,” devoted to exposing misconceptions and faulty practices in the
industry, that were accepted as valid by numerous engineers. Jim alternated with Clemens in authoring this column.

It was a good time for Electronics Cooling—the economy was good with strong advertising revenues and equally important, thermal engineers still
had enough discretionary time that they could commit to writing up their work for publication. The “lean and mean” era was still in the future. We
even had the luxury of having Tony Kordyban (author of the book, Hot Air Rises and Heat Sinks) as an occasional guest contributor, as he inserted
his dry and quirky humor into investigations of common yet questionable practices in the industry. This “smooth sailing” era came to an end in
2007 when Kaveh left Electronics Cooling to devote more time to his other enterprises. In 2007-2008, the financial crisis happened and Flomerics
was acquired by a larger company and was compelled to end their financial support for the magazine. More broadly, the crisis put financial stress
on all industries, including publishing.

2 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

After a period of uncertainty, Electronics Cooling was acquired by ITEM Media in 2010. The editors were very relieved, since there was a real risk that
the publication would go out of existence. In fact, in the Spring 2010, issue, our first issue after the regime change, Jim Wilson’s editorial was head-
lined, “We are back—and in print.” ITEM took on the challenge of creating a sustainable business model around Electronics Cooling, while at the
same time maintaining the editorial independence of the Assoc. Technical Editors, now referred to as the Editorial Board. Along the way Clemens
and Bob left the magazine. Madhu Iyengar, of IBM at the time, and Prof. Peter Rodgers, then at the Petroleum Institute, in the U.A.E, replaced them
on the magazine. Madhu began contributing to “Calculation Corner” column and Peter likewise to the “Thermal Facts and Fairy Tales” column.

Meanwhile, structural changes continued in the tech industry worldwide. The tech sector in the west saw a lot of mergers, acquisitions, restructuring
events, and layoffs. In this relentless economic environment, as would be expected, time pressures on engineers increased markedly. However, I’m
glad to say that, despite these challenges, the editors were able to solicit articles on a wide variety of topics that met the traditional quality standards
of the publication. However, it took a lot more effort than in the past to locate authors who could find the time to craft a quality article on a com-
pelling topic.

In 2014, Madhu joined one of the internet giant companies and had to leave the magazine. By 2016, Peter also left to focus all of his energies on
the demands of his academic career. In 2016, Jim resigned after being a valued contributor to Electronics Cooling for 18 years. Later in 2016, Ross
Wilcoxon, then a principal mechanical engineer at Rockwell Collins, and Victor Chiriac, then the thermal technologist at Qualcomm, joined the
magazine. Ross has contributed to the “Thermal Facts and Fairy Tales” and “Calculation Corner” columns and has recently started a new column,
“Statistics Corner.” Victor launched the column, “Technology Corner,” focusing on the megatrends in the tech sector with a focus on the semicon-
ductor industry. In 2017, the number of issues per year was reduced to three. In 2019, Genevieve Martin joined us. She is an R&D manager at Signify
(formerly a lighting division of Philips), in The Netherlands. She is active in several EU conferences and consortia dealing with LED test standards.
Once again, there was a full quota of technical editors, whose expertise covered a wide swath of relevant technologies.

We owe a debt of gratitude to the upper management of Lectrix (formerly ITEM Media) Graham Kilshaw (CEO) and Geoffrey Forman (VP of Mar-
keting) for their role in acquiring Electronics Cooling in 2010, and blending it into their print assets as well as giving it a much more robust online
presence, thereby integrating it into a sustainable business model. All this was done while retaining their commitment to maintaining the quality
standards that have distinguished Electronics Cooling over the decades. Without this intervention, Electronics Cooling would not now be celebrating
its 25th anniversary.

We also applaud their hiring for the first time in 2019, a full time Editorial Director, Jennifer Arroyo, to oversee all the print and online content. She
is a seasoned professional, having served as Editor-in-Chief of ECN (Electronic Component News).

Finally, I would like to express our gratitude to you, our loyal readers, for your support over the decades. Without your support, this adventure
would not have been possible.

The saga of the ups and downs and comings and goings of the folks associated with Electronics Cooling is a microcosm of the wider world around us.
However, despite all of these challenges we are still motivated by the original vision of Electronics Cooling, that at its very core was seen as a means
of bringing engineers today to share their expertise with each other to enable them to be more effective in their professions and, ultimately, to better
serve our larger society.

Speaking of goings, I’ve decided that the time has come for me for me to recede into the background, after spending 22 rewarding and challenging
years as an Assoc. Technical Editor of Electronics Cooling. I have felt both privileged and humbled by having the opportunity to serve you for so
many years. I still hope to provide behind-the-scenes assistance to Ross, Victor, and Genevieve, as they would wish and as I am able. Meanwhile, I
pass the torch to another generation of editors and wish them the best of success in these very challenging times.

In this editorial, I’ve spoken of the challenges faced by Electronics Cooling, that were mainly economic in origin. However, we are now facing the
most dire challenge in our lifetimes—the COVID-19 pandemic.

At this juncture, no one knows when the world will get this pandemic under control and what its ultimate impact will be on ourselves, our families,
and our society.

All I can do, in this final editorial of mine, is to convey my deepest regards to you and your families and my sincere wish that you all stay safe and
healthy.

Now, I’d like to add a few words about the content in this special 25th anniversary issue. In honor of this event we have deviated from our usual for-
mat and have reprinted articles that were individually selected by each of the Associate Technical Editors from the top 100 articles ranked according
to reader popularity. The relative popularity of all the articles at the Electronics Cooling website was determined by Google on the basis of the number
of unique page views of each article. We hope you enjoy them.

Electronics-COOLING.com 3
PAST AND PRESENT TECHNICAL EDITORS
TRIBUTE TO BRUCE GUENIN’S
22 YEARS OF SERVICE FOR
ELECTRONICS COOLING MAGAZINE

On behalf of the past and present members of the Technical Editorial Board of Electronics Cooling, it is a plea-
sure to write a few words of thanks to Bruce Guenin for his 22 years of service to the publication. As some of
you may know by now, Bruce has decided to step down from his current EC responsibilities in order to pursue
other exciting activities. We wish Bruce all the best in his future endeavors!

To date, Bruce is the longest serving Associate Technical Editor of Electronics Cooling—his commitment and contribution to the magazine dates
back to 1998. Over the decades, Bruce witnessed many changes in the editorial boards and the magazine’s ownership, helped overcome various
technical and editorial challenges, championed various initiatives, and maintained his outstanding service and loyalty to the thermal community
readership. Under his technical stewardship, the publication has grown and the article quality has always been top notch, as Bruce believes in
promoting and applying the highest standards for the technical review process prior to publication. It is the practical value to the reader that he
pioneered and professed since the early days of Electronics Cooling, and we celebrate Bruce for his long-term vision and dedication as he passes on
the baton to us, his editorial board peers.

Thank you Bruce, for your outstanding activity, creativity, inspiration, and commitment to high-quality articles and teachings, from the young to
the more seasoned practicing engineers, as well as for the overall (global) thermal community. We will continue to keep the EC tech tradition alive
and contribute to the publication's continued success.

Below are a few short stories and memories from Bruce and from past technical editors of Electronics Cooling. Enjoy and join us in thanking
Bruce for his outstanding service to our Electronics Cooling community! (compiled by Victor Chiriac, Assoc.Tech. Editor on behalf of the
current Assoc. Tech. Editorial Board members Genevieve Martin and Ross Wilcoxon)

Some personal notes on Bruce’s retirement from Electronics Cooling

(Clemens Lasance, EC Tech Editor, 1995 – 2013)

In 1995, Kaveh Azar, Gordon Ellison, and I started the magazine to fill the gap between academic and advertorial-driven journals. Its purpose from
the start was to provide the community of thermal engineers and designers with peer-reviewed articles in which mentioning the name of a company
was considered a mortal sin. When Gordon stepped down his place was filled by Bob Simons and Bruce, and when Bob retired Jim Wilson became
his successor. I have very fond memories of this period, and I learned a lot from my brother-editors. All articles and columns were peer-reviewed,
and I don’t recall a single situation where we didn’t reach consensus, often after some fierce discussions. Especially between Bruce and me, he the
wisest, me the more outspoken (being Dutch). From a scientific/technological point of view his contributions over the years (especially his “Calcu-
lation Corner”) were highly praised, and, in my humble view, should be archived in a volume to facilitate access. Apart from the magazine, we met
regularly at various thermal conferences and JEDEC meetings.

A long time ago Bruce and I found out that he played the guitar and me the piano. I suggested that we should perform as a duo at some SEMI-
THERM conference, and I sent him the score of a piece by Mauro Giuliani. It never happened (probably due to the lack of a piano in the conference
hall). In 2014, Bruce and his wife Anita visited us in our little village near Eindhoven (now world-wide known as the Van Gogh Village), and I
vividly recall our long discussions related to religion and philosophy.

I miss you, Bruce, your wisdom and your humor. I resigned after 18 years, around 2013, he beat me by four years; for sure he will be remembered as
the King of Electronics Cooling. (notes compiled by Clemens Lasance, at the invitation of Genevieve Martin, EC Assoc. Tech Editor)

Some Random Reminiscences from my 22 Years at Electronics Cooling

(Bruce Guenin, EC Tech Editor, 1998 – 2020)

I was asked by my fellow Assoc. Tech. Editors to share my reminiscences regarding my years at Electronics Cooling. Below are a few anecdotes that
bubbled up from my selective memory into my conscious mind. I hope you find them interesting.

4 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

By the time I was invited by Kaveh to join the editorial staff of Electronics Cooling in 1998, I had already been rather active in publishing and giving
presentations at various venues. But what was different here was that, as far as my “Calculation Corner” column was concerned, it was basically
my decision what topic would be explored. Thus, the column was a conduit through which I could communicate directly with tens of thousands of
readers of the publication—a novel experience for me at the time.

In the beginning, the titles of my “Calculation Corner” columns had bland, technical titles, such as “Determining the Junction Temperature in a
Semiconductor Package, Part III—The Use of the Junction-to-Board Thermal Characterization Parameter,” published in the May 2002, issue. Over
time, I chose titles that were a bit more playful, such as, “The 45° Heat Spreading Angle—An Urban Legend?” in the November 2003 issue, and
“Thermal Vias—A Packaging Engineer's Best Friend,” in August 2004. These titles were certainly more “catchy” but they at least revealed to the
reader what the content was.

However, this playfulness backfired when I explored an interesting phenomenon I discovered though some simulations. They showed that the junc-
tion-to-case thermal resistance of a package, that normally is seen as an intrinsic property of the package, could, in fact, be influenced by the heat
spreading properties of the test board. (The board’s heat spreading ability was related to the presence or absence of copper planes in its interior.)
This finding was a significant one and should have been of interest to the readership. However, in my efforts to be clever, I gave the article the title,
“A Funny Thing Happened on the Way to the Heatsink,” which gave the reader no idea at all what the article was about. To get the humor, one would
had to have heard of the Broadway play, A Funny Thing Happened on the Way to the Forum (the story was set in Roman times). Of course, few of
our readers had heard of the play and would, therefore, would have been able to get the joke. Thus, the ultimate impact of this title was to obfuscate
what the article was about and provide no motivation at all to the reader to investigate the article further. This taught me a lesson and, from that
point on, I made no more lame attempts at humor in my “Calculation Corner” titles.

After Kaveh left the publication in 2007, we four Assoc. Tech. Editors adopted the organizational model of a string quartet. We essentially made
decisions by consensus. One effect is that since we had four issues per year and four editors, we decided that each of us would be an Acting Edi-
tor-in-Chief for one of the issues and would have the privilege of writing the Editorial for that particular issue. This was something I enjoyed very
much. It was yet another way of communicating directly with our readers, but one in which I could share my point of view on a broader array of
issues that had an impact on our profession and, on occasion, on our society as a whole. As with “Calculation Corner” I also had some fun with the
titles of my editorials, by mimicking the titles of well-known works that were current at that time, such as:

• “The Audacity of Engineering,” February 2009, from: Barack Obama, The Audacity of Hope
• “The Joy of Engineering,” June 2015, from: The Joy of Cooking, a classic cookbook
• “It Takes a Village—To Raise an Engineer,” June 2016, from: Hillary Clinton, It Takes a Village: and Other Lessons Children Teach Us

I’m sure that there was a bit of wincing on the part of readers familiar with these references. However, these titles probably helped get their attention.
I also had fun with two provocative titles for two installments of the “Thermal Facts and Fairy Tales” column:

• “The Junction-to-Case Thermal Resistance: A One-Dimensional Underachiever in a Three-dimensional, Conjugate Heat Transfer World”,
Spring 2018
• “Whatever Happened to the Predicted Data Center Energy Consumption Apocalypse?” Spring 2019 (I couldn’t resist the opportunity to use
the word, “apocalypse”)

Although I enjoyed writing these editorials and sharing my point of view on some broader issues with the readership, I had no idea to what extent
they were actually read. However, I was very pleased to have been approached by an engineer at a conference who said that after reading “The Joy of
Engineering” his son, who was about to start college, decided to major in engineering. That one acknowledgement certainly made it all worthwhile
for me. (compiled by Bruce Guenin at the invitation of the EC Tech Editorial Board)

Electronics-COOLING.com 5
TECHNICAL EDITORS SPOTLIGHT
Meet the 2020 Electronics Cooling® Editorial Board

VICTOR CHIRIAC | GLOBAL COOLING TECHNOLOGY GROUP

Associate Technical Editor
A fellow of the American Society of Mechanical Engineers (ASME) since 2014, Dr. Victor Adrian Chiriac
is a co-founder and a managing partner with the Global Cooling Technology Group. He previously held
technology/engineering leadership roles with Motorola (1999-2010), Qualcomm (2010-2018) and Huawei
R&D USA (2018-2019). Dr. Chiriac was elected Chair of the ASME K-16 Electronics Cooling Committee
and was elected the Arizona and New Mexico IMAPS Chapter President. He is a leading member of the
organizing committees of ASME/InterPack, ASME/ IMECE and IEEE/CPMT ITherm Conferences. He holds
19 US issued patents, 1 US Trade Secret, 1 Defensive Publication, and has published over 109 papers in
scientific journals and at international conferences.
► vchiriac@gctg-llc.com

BRUCE GUENIN | CONSULTANT

Associate Technical Editor

Dr. Bruce Guenin has spent many years in the electronics and computer industries, which has given him a broad
perspective on macro trends in these fields. He has been an editor of Electronics Cooling® since 1997 and has
contributed, to date, 35 installments of the tutorial column, Calculation Corner. His previous affiliations include
Oracle, Sun Microsystems, and Amkor. He is a past chairman of the JEDEC JC-15 Thermal Standards Commit-
tee and the SEMI-THERM® Conference. His contributions to the thermal sciences have been recognized by
receiving the Harvey Rosten Award in 2004 and the Thermi Award in 2010 from the SEMI-THERM® Conference.
He received the B.S. degree in Physics from Loyola University, New Orleans, and the Ph.D. in Physics from the
University of Virginia. He has authored and co-authored over 80 papers and articles in the areas of thermal and
stress characterization of microelectronic packages, electrical connectors, solid state physics, and fluid dynam-
ics and has been awarded 18 patents in these areas.
► sdengr-bguenin@usa.net

GENEVIEVE MARTIN | SIGNIFY

Associate Technical Editor
Genevieve Martin (F) is R&D manager for thermal & mechanics competence at Signify (former Philips Lighting), The
Netherlands. She is working in the field of cooling of electronics and thermal management for over twenty years in
different application fields. From 2016 to 2019, she coordinates the European project Delphi4LED (3 years project)
dealing with multi-domain compact model of LEDs. She served as General chair of SEMI-THERM® conference and
is an active reviewer and technical committee in key conferences SEMI-THERM®, Therminic, Eurosime.
► genevieve.martin@signify.com

ROSS WILCOXON | COLLINS AEROSPACE

Associate Technical Editor
Dr. Ross Wilcoxon is an Associate Director in the Collins Aerospace Advanced Technology group. He con-
ducts research and supports product development related to component reliability, electronics packaging and
thermal management for communication, processing, displays and radars. He has more than 40 journal and
conference publications and is an inventor on 30 U.S. Patents. Prior to joining Rockwell Collins (Now Collins
Aerospace) in 1998, he was an assistant professor at South Dakota State University.
► ross.wilcoxon@collins.com

6 Electronics COOLING | FALL 2020

 Industrial

Networking / Telecom Consumer

Renewable
Energy Solutions Server / AI

Autonomous
Systems

Liquid Cooling Solutions

Professionally designed, built and tested to withstand harsh environments

High performance Coolant Distribution Unit (CDU)

Impinging cold plate design with lower pressure drop
Broad range of testing equipment to ensure product quality and performance
Micro-channel cold plate block designs
Liquid direct cooling for memory DIMMs
High-grade tubing with low permeation, high-pressure capacity and
small bending radii achievable
Hot-swappable quick-disconnect capability for dry disconnect

www.delta-fan.com | dcfansales.us@deltaww.com Smarter. Greener. Together.

THERMAL LIVETM | OCTOBER 20 – 21, 2020

ONLINE EVENT
TM

OCTOBER 20 — 21, 2020

PRESENTED BY ELECTRONICS COOLING ®

8 TO REGISTER FOR THIS EVENT, PLEASE VISIT WWW.THERMAL.LIVE

 WELCOME TO
THERMAL LIVEtm 2020

The Largest Single Thermal

Management Event of The
Year - Anywhere.

Thermal LIVE™ is the electronics

and mechanical engineer’s free,
online resource for education and
networking in thermal manage-
ment. Learn the latest techniques
and topics directly from thermal
management thought leaders wit-
hout leaving your seat.

Join us for two full days of interac-

tive webinars, product demonstra-
tions, roundtables, whitepapers,
and more.

Produced By:

9
 THERMAL LIVETM | OCTOBER 20 – 21, 2020

THERMAL LIVE TM 2020 TECHNICAL PROGRAM

OCTOBER 20 – 21, 2020

TUESDAY | OCTOBER 20, 2020

DEREJE AGONAFER, Ph.D., NAE | 9:00 am ET

Presidential Distinguished Professor in MAE, University of Texas at Arlington (UTA)
KEYNOTE - Single Phase Immersion Cooling of Data Centers: Opportunities and Challenges
Data center cooling has never been more challenging as it is today with the thermal design power (TDP) of
the components rising by almost 50% in the past decade. With advancements in 2.5D and 3D packaging
architectures for realizing smaller feature sizes in this decade, an increase in local power densities has
necessitated the requirement of more efficient cooling technologies for thermal management. This growth
can be primarily attributed to advancements in high power CPUs and GPUs to support high power computing
applications like bitcoin mining, AI, and machine learning algorithms development. Due to low thermal mass
limitations, the typical air-cooled data centers present a relatively less efficient and complex cooling option
owing to the requirements of raised floors, air handling/chiller piping, power routing, complex filtration and
tight environmental control involved, etc. Single-phase immersion provides simplicity in terms of thermal
infrastructure, PUEs as low as 1.03, and reduction in CAPEX equal to or greater than 50%. Single-phase
immersion can also provide benefits of direct to chip cold plate cooling by eliminating the complexity and
improved upfront affordability and ease of operation. Immersion cooling may also prove to be an efficient
solution for 3D stacked and heterogeneously packaged dies because of its inherent ability to provide even
temperature profiles. Immersion cooling also enhances system reliability by protecting the IT equipment
from the harsh environmental effects of high temperature, dust, vibrations, and corrosive gases. Lack of
published data on material compatibility, long term signal integrity, changes required in server architecture,
and maintainability are cited as some of the primary challenges in large scale implementation. Other major
issues associated currently with single-phase immersion are using completely sealed HDDs or using solid-state
drives. Additionally, the lack of manufacturer warranties provided by the OEMs is another issue that is slowing
the widespread adoption of single-phase immersion cooling. Enhanced equipment reliability and ease of
implementation, especially for edge data centers, also make single-phase immersion cooling an attractive
alternative for data centers to switch to immersion cooling.

TIM JENSEN | 10:00 am ET

Sr. Product Manager for Engineered Solder Materials, Indium Corporation
Novel TIMs to Enable Advanced Thermal System Design for High Performance and Quantum
Computing Applications
Thermal system design is a critical aspect in both high-performance and quantum computing applications.
Part of that system is the thermal interface material (TIM). This presentation will begin with an overview of the
thermal demands in these applications. From there, the focus will be on unique TIM technologies that can help
enable or enhance active and passive cooling designs to ensure maximum performance. Because the use of
metal-based TIMs is proving to be increasingly critical for success, several attributes of various metal TIMs will
be reviewed to show how they can ensure the best performance in these applications.

10 TO REGISTER FOR THIS EVENT, PLEASE VISIT WWW.THERMAL.LIVE

 WWW.THERMAL.LIVE

THERMAL LIVE TM 2020 TECHNICAL PROGRAM

OCTOBER 20 – 21, 2020

TUESDAY | OCTOBER 20, 2020 | CONTINUED

KIMBERLY FIKSE | 11:00 am ET

Lead Sales Engineer, ACT
Advanced Thermal Techniques Across Diverse Industries
In this webinar we will share real world examples where creative approaches in thermal engineering have
allowed design engineers to push their design concepts forward. Join us as we explore some advanced
thermal management technologies, including Heat Pipes and Phase Change Material heat sinks. We will
provide practical applications for each that apply to a range of markets from Spacecraft thermal control to
defense and industrial applications.

CHRISTIAN MAINEGRA | 12:00 pm ET

Applications Engineer, Fujipoly
Advantages of Putty Type: Thermal Interface Gap Filler Materials
Fujipoly will reveal many of the advantages of putty-type thermal interface materials offer compared to
standard gap filler pads. Key topics will include stress strain comparisons of putties and standard gap fillers,
advantages of putty over dispensable thermal interface materials, as well as putty handling and application
suggestions.

C. T. KAO | 1:00 pm ET
Product Management Director, Cadence Design Systems
Celsius Thermal Solver: The Thermal Solution for Intelligent System Design
As power densities on IC designs continue to increase, controlling temperature on the chip is becoming a
major challenge for IC designers. High temperature impacts both the reliability and electrical performance
of ICs, and designers need a way to accurately perform thermal analysis to model the thermal impact of
floorplan changes. Furthermore, heat would propagate from the chips through the packages and PCBs to
the environment, which clearly indicates that solving thermal issues in modern electronic systems will need
to take a global approach but cover the details in the design at the same time. Multi-physics analysis is also
essential for system-level thermal solution, including both steady and transient simulation, electrical-thermal
co-simulation, and Computational Fluid Dynamics (CFD). Today’s advanced applications whether they are in
automotive, datacenter, mobile, healthcare or high-performance computing, system-level thermal analysis,
which is typically not addressed by other tools, is a critical factor in understanding the overall thermal behavior
and managing the thermal impact on performance, reliability, and quality.

THERMAL LIVETM | OCTOBER 20 – 13, 2020 ELECTRONICS COOLING | FALL 2020 11

 THERMAL LIVETM | OCTOBER 20 – 21, 2020

THERMAL LIVE TM 2020 TECHNICAL PROGRAM

OCTOBER 20 – 21, 2020

TUESDAY | OCTOBER 20, 2020 | CONTINUED

JONATHAN TAYLOR | 1:45 pm ET

Product Manager, Neograf Solutions, LLC
NeoNxGen™ Thermal Management Solutions – Increase the Thermal Performance, Not the
Layers!
The latest generations of high-performance consumer electronics need thermal spreaders that are both very thin
and highly thermally conductive. One solution is to stack several layers of thin synthetic graphite using layers of
adhesive. This method produces spreaders that are thermally effective but are expensive and have poor yields.
Layering synthetic graphite creates a final part that is very stiff and can easily delaminate if repeatedly flexed.
NeoNxGenTM flexible graphite is a new type of monolayer high-performance graphite heat spreader that is
currently being manufactured in high volumes. It has the thermal performance of a multilayer synthetic solution
but maintains the flexibility and manufacturing cost advantages of a single layer. NeoNxGenTM flexible graphite
is tunable to meet the thickness and performance needs of specific devices, such as the foldable screen in a
notebook computer or to replace the copper vapor chamber in a mobile phone.

MICHAEL GONZALEZ | 2:45 pm ET

Sales Engineer, CDCP, CEJN North America
Blind Mate Couplings: Design and Manufacturing Aspects
This session will focus on the increasing liquid cooling demand with required quality, tolerances and technical
design considerations of a blind mate coupling, which is needed to ensure a safe and reliable cooling solution
for the end user. Effects on manufacturing, QC, and production are also factors that are continually evaluated
throughout the design process.

ANAND SAMANT | 3:45 pm ET

Global Market Director for Consumer Electronics, Materion
Advanced Materials Strategies for Cooling Modules in Electronic Systems
Higher data-transfer rates and power-usage requirements result in increasing heat-management challenges
for thermal engineers in the electronics industry. At the same time, there is growing demand for thinner and
lighter devices, resulting in less available volume for cooling modules. Mismanagement of heat can not only
negatively impact short-term performance, but can also lead to failure of electronic components. In this
webinar, we will discuss innovative strategies, such as clad systems combining multiple metals, metal matrix
composites combining metals and ceramics, and high conductivity monolithic alloys – all of which can help the
thermal engineer achieve the desired structural and thermal properties for cooling module materials.

12 TO REGISTER FOR THIS EVENT, PLEASE VISIT WWW.THERMAL.LIVE

 WWW.THERMAL.LIVE

THERMAL LIVE TM 2020 TECHNICAL PROGRAM

OCTOBER 20 – 21, 2020

WEDNESDAY | OCTOBER 21, 2020

BRAD WHITNEY | 10:00 am ET

Product Director/ VP of Two Phased Cooling, Boyd Corporation
Utilizing Ultra-Thin Two-Phase Cooling & Ultra-Thin Titanium Vapor Chambers to Improve
Performance
Ultra-thin two-phase technologies such as Vapor Chambers, Titanium Vapor Chambers, and Heat Pipes can be
a cost-efficient way to vastly improve the performance of your thermal solution while enabling thinner, lighter
cooling. These passive solutions also offer extremely high reliability and lifetimes while remaining silent with
no moving parts. These factors make them ideal for more powerful low-profile and next-generation devices.
Learn how new technical advancements have enabled improved design and manufacturing options for thinner,
higher performance vapor chamber solutions.

ANDRES ABRAHAM | 11:00 am ET

Applications Engineer, Thermal Business Unit, CPC
Liquid Cooling: Getting in the Know About Flow
To optimize performance and enhance sustainability, engineers need to consider how managing fluid flow
for cooling within an advanced electronics application differs from traditional air-flow cooling systems.
Understanding the heat transfer coefficient and how heat dissipates through liquids exponentially more efficiently
than through gases is at the root of efficiencies to be gained. However, understanding flow coefficients alone
is not enough. It is only the first step in how to compare liquid cooling component performance or ultimately
arrive at system capacity. For example, at the connector level, the Cv value of a given quick disconnect (QD)
can vary widely based on the cooling fluid used and system operating temperatures. Also to be considered
is volumetric flow rate -- the volume of fluid that passes through a point in a system per unit of time (which is
usually expressed in gallons per minute.) Additionally, specific gravity and how fluid properties impact system
component requirements can not be ignored.

In this session, learn more about these calculations along with the three design elements within QD design
that impact flow. Size, valves, and mounting options are key variable features of liquid cooling connectors, and
these components and their interplay or impact on flow is not always intuitively understood. As you will learn,
in liquid cooling, efficient flow is essential to effective thermal management.

THERMAL LIVETM | OCTOBER 20 – 13, 2020 ELECTRONICS COOLING | FALL 2020 13

 THERMAL LIVETM | OCTOBER 20 – 21, 2020

THERMAL LIVE TM 2020 TECHNICAL PROGRAM

OCTOBER 20 – 21, 2020

WEDNESDAY | OCTOBER 21, 2020 | CONTINUED

KEITH PERRIN | 12:00 pm ET

Global Director for Electronics, Hexagon Manufacturing
Advanced Micro-structural & thermal modeling for Energy Generating, Wearable, Electronics
At Hexagon we're dedicated to bringing the virtual and physical worlds together to enable our customers to
drive their ideas forward and stay ahead. Nowhere is this more important than in the design and manufacture
of industrial and consumer electronics.

In this session we'd like to explore some of the application and possibilities to be had from the application
of advanced microstructural & thermal modeling techniques to further the development and performance of
flexible electronics and even piezoelectric fabric. Not only are we looking at wearing electronics. We're looking
at having them power themselves.

NICOLAS MONNIER | 1:00 pm ET

Product & Market Specialist, Stäubli North America
PRODUCT DEMO - Quick Connectors for Two-Phase Fluids
Following the adoption of liquid cooling, two-phase fluids are a growing interest in the engineering community
due to many thermal advantages. As a leader in the IT Cooling industry, Stäubli will present the main technical
considerations for using quick disconnects with two-phase fluids. This product demonstration will cover
technical aspects of materials, pressure, tightness, and the Stäubli products designed for these applications.

METODI ZLATINOV1 & DENVER SCHAFFARZICK2 | 1:45 pm ET

Technical Lead, R&D1,Director of Engineering2, ERG Aerospace
Metal Foams Make Better Heat Exchangers
When faced with high heat fluxes and challenging packaging constraints, it is time to look beyond conventional
fins and consider high-performance metal foam heat sinks, cold plates, and heat exchangers. Due to their high
surface area, turbulence-enhancing pore structure, and excellent thermal conductivity, open-celled aluminum
and copper foams can deliver unrivaled enhancement in heat transfer. With a range of available base materials,
pore sizes, porosities, and the freedom to machine, compress, braze, and bond, it is possible to tailor both
performance and form factor to the application at hand. In this webinar, ERG Aerospace will dive into the
characteristics and properties of metal foams, introduce the principles behind designing effective metal foam
cooling systems, and highlight some real-world applications of this technology.

14 TO REGISTER FOR THIS EVENT, PLEASE VISIT WWW.THERMAL.LIVE

 WWW.THERMAL.LIVE

THERMAL LIVE TM 2020 TECHNICAL PROGRAM

OCTOBER 20 – 21, 2020

WEDNESDAY | OCTOBER 21, 2020 | CONTINUED

JOSEPH FORBES PETRI | 2:45 pm ET

Electronic Assembly Solutions Account Manager, 3M
Thermal Management for a Connected World
In the world of IoT, and an increasing number of connected devices, electronic devices and components are
expected to provide greater functionality to meet higher performance requirements. While the operating
frequency increases, the amount of power dissipates as heat increases. The buildup of excess heat in electronics
can cause 55% of electronic systems to fail. Design engineers are faced with a serious problem – how do we
maintain performance and have a more reliable device? The need for thermal management is across multiple
market segments. This webinar will focus on thermal management solutions in two key market segments;
automotive and aerospace & defense. Join this webinar to learn how 3M™ Thermal Interface Materials can
help keep your devices performing at their best.

JOHN WILSON | 3:45 pm ET

Electronics Product Specialist, Siemens Digital Industries Software
Thermal Analysis, Electronics Reliability and the Influences of Variability
The traditional use of electronics thermal analysis in its basic function has always been to assess whether a
design will meet requirements representing a worst-case scenario. Improving a design to meet a performance
goal using parametric studies, or optimization techniques is widely adopted. Even assessing designs against
product use mission profiles using transient simulation is now more widely used whether for power management
and efficiency studies or for generating temperature cycling and amplitude insights for lifetime prediction and
reliability purposes.

How many thermal simulation-based design processes today encompass the wide influence of variability? By
combining electronics cooling CFD simulation with statistical approaches, engineers can more fully explore
the design space, verify decisions made, and also evaluate matters such as manufacturing quality variation.
This presentation introduces an approach to simulation-based reliability assessment to reduce design process
aspects that impact field failure. Studies performed utilized Design of Experiments and Monte Carlo methods
and tools, including Simcenter Flotherm software and Simcenter HEEDS design exploration software.

THERMAL LIVETM | OCTOBER 20 – 13, 2020 ELECTRONICS COOLING | FALL 2020 15

 INTERESTED IN ATTENDING THERMAL LIVE TM 2020?
REGISTER TODAY FOR FREE AT

W W W.THER M AL .LI V E

INTERESTED IN PRESENTING AT THERMAL LIVE TM ?

For Marketing Opportunities, Contact Our Business Development Team:

GRAHAM KILSHAW JAN WARD

e: Graham@lectrixgroup.com e: Jan@lectrixgroup.com
p: 610.382.5885 p: 610.382.5891

THERMAL LIVETM | PRESENTED BY ELECTRONICS COOLING® | WWW.THERMAL.LIVE

Call for Authors
and Contributors!
Want to be a part of the next issue of Electronics
Cooling? Have an article or blog post you’d like
to write for Electronics-Cooling.com?

Let us know at
editor@electronics-cooling.com

www.Electronics-Cooling.com
F E AT U R E D

Advances In High-Performance
Cooling For Electronics
Reprinted from the 10th Anniversary Issue, November 2005

Clemens J.M. Lasance Robert E. Simons

*Philips Research Laboratory *IBM Corp.
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
Dear Readers, it is my distinct pleasure to share with you this milestone article authored by two former Editors of Electronics
Cooling. Both Clemens Lasance and Bob Simons had very distinguished careers in their professional lives and during their years
as Associate Technical Editors at Electronics Cooling. Clemens worked for Philips Research in Eindhoven. Bob served in the IBM
Advanced Thermal Laboratory in Poughkeepsie, NY. Both were world-class technical organizations. Over their long careers, they
encountered nearly every type of cooling technology allowed by the laws of physics. In this article, originally published in the
10th Anniversary issue of Electronics Cooling in November 2005, they described the physical mechanisms operative in 18 classes
of cooling methods. Furthermore, they quantify the cooling efficiency of each and evaluate the possibility of further performance
improvement in each of them. As one reads this very clear and concisely written article, one can’t help but think of the dramatic
advances made in many of these fields over the last 15 years. However, it still remains a very useful reference since it catalogs the
range of possible cooling options and provides references that can be a starting point for one wishing to get a more in-depth
understanding of any of them. In my opinion, because of its encyclopedic breadth, this article represents the high water mark for
quality in the 25-year history of Electronics Cooling. (Bruce Guenin)

T
INTRODUCTION (including a possible area enlarging factor) of 20,000 W/m2K. From
he need for new cooling techniques is driven by the Figure 1 it can be concluded that there will be a need for liquid cool-
continuing increases in power dissipation of electronic ing in the future of thermal management. This article briefly dis-
parts and systems. In many instances, standard tech- cusses a number of promising thermal management technologies
niques cannot achieve the required cooling perfor- that are emerging for possible electronics applications.
mance due to physical limitations in heat transfer capabilities.
These limitations are principally related to the limited thermal CONDUCTION AND HEAT SPREADING
conductivity of air for convection and copper for conduction. In all cooling applications, heat from the device heat sources, must
first travel via thermal conduction to the surfaces exposed to the
cooling fluid before it can be rejected to the coolant. For example, as
shown in Figure 2, heat must be conducted from the chip to the lid
to the heat sink before it can be rejected to the flowing air. As can
be seen thermal interface materials (TIMs) may be used to facilitate
thermal conduction from the chip to the lid and from the lid to the
heat sink. In many cases heat spreaders in the form of a flat plate
with good thermal conductivity may be placed between the chip
and lid to facilitate spreading of the heat from the chip to the lid or
heat sink. Vapor chambers are also used to spread heat from a con-
centrated chip or module heat source to a larger heat sink.

Figure 1: Heat transfer coefficient attainable with natural convection, single-phase For high-power applications, the interface thermal resistance be-
liquid forced convection and boiling for different coolants [1]. comes an important issue. Direct soldering (e.g., reflow soldering)
is often difficult, certainly when copper is used because of the
Figure 1 shows a comparison of various cooling techniques as a large CTE mismatch between Cu and Si. However, a few promis-
function of the attainable heat transfer in terms of the heat transfer ing materials are entering the market.
coefficient. To accommodate a heat flux of 100 W/cm2 at a tempera-
ture difference of 50 K requires an effective heat transfer coefficient Diamond-filled greases have been tested to have an effective ther-

18 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

mal conductivity of over 20 W/mK; however, the vendor claimed

60 W/mK [2]. Even more interesting is a nanostructured foil,
which utilizes a very fast exothermic reaction to create a soldered
connection virtually at room temperature [3]. Extensive long-
term reliability studies are in progress [4].

Figure 3: Example of effect of thickness on heat spreading for various heat source
areas, material thermal conductivities, and heat transfer coefficients (A in cm2, k in
W/mK, h in W/m2K).
Figure 2: Chip package with thermal conduction path to heat sink via TIMs.
AIR COOLING
Heat spreading is a very effective way of mitigating the need for so- It is generally acknowledged that traditional air-cooling tech-
phisticated high-heat-flux cooling options. Of course, to be effec- niques are about to reach their limit for cooling of high-power
tive, the benefits of decreasing the heat flux density by increasing applications. With standard fans a maximum heat transfer coef-
the area should outweigh the penalty of adding another layer that ficient of maybe 150 W/m2K can be reached with acceptable noise
the heat must be conducted across. This is an optimization prob- levels, which is about 1 W/cm2 for a 60°C temperature difference.
lem as discussed below. The options for advanced heat spreading Using ‘macrojet’ impingement, theoretically we may reach 900
solutions are two-fold: W/m2K, but with unacceptable noise levels. Non-standard fans/
dedicated heat sink combinations for CPU cooling are expected to
• Novel materials such as carbonaceous materials, met- have a maximum of about 50 W/cm2, which is a factor of 10 higher
al-matrix composites, ceramic matrix composites (e.g., dia- than expected 15 years ago. However, some new initiatives have
mond-particle-reinforced silicon carbide), or ScD (Skeleton emerged to extend the useful range of air-cooling such as piezo
cemented Diamond), all of them with much higher thermal fans, ‘synthetic’ jet cooling and ‘nanolightning’.
conductivities than copper, much less weight and tunable
CTEs [5]. PIEZO FANS
• Novel heat spreader technologies such as Novel Concept’s Piezoelectric fans are low power, small, relatively low noise, sol-
Isoskin [6] and Enerdyne’s Polara [55] that claim effective id-state devices that recently emerged as viable thermal manage-
thermal conductivities that compete with diamond. ment solutions for a variety of portable electronics applications
including laptop computers and cellular phones. Piezoelectric
By applying heat spreaders cooling methods such as loop heat fans utilize piezoceramic patches bonded onto thin, low frequency
pipes and low-flow liquid cooling may be augmented to accommo- flexible blades to drive the fan at its resonance frequency. The reso-
date higher heat flux applications. Figure 3 provides a graph show- nating low frequency blade creates a streaming airflow directed at
ing heat spreading results for a 300 W heat source of 2 cm2 area as electronics components. A group at Purdue reports up to a 100%
a function of thermal conductivity, thickness, and cooling bound- enhancement over natural convection heat transfer [7].
ary condition (i.e., heat transfer coefficient). Looking at the results
it becomes obvious that heat spreading is a complex phenomenon. ‘SYNTHETIC’ JET COOLING
This is because the conduction and convection effects cannot be An approach using periodic microjets coined ‘synthetic jets’ has
separated and the two effects compete: increasing the thickness initially been studied by Georgia Institute of Technology and is
increases the through-plane resistance but decreases the in-plane being commercialized by Innovative Fluidics. Due to the pulsat-
thermal resistance. For example, comparing the two upper curves ing nature of the flow, synthetic jets introduce a stronger entrain-
with the two lower curves, their order is changed. The results also ment than conventional-steady jets of the same Reynolds number
show that it is very well possible to use heat spreaders to decrease and more vigorous mixing between the wall boundary layers and
the required fluid-side heat transfer coefficient to easily manage- the rest of the flow. One of the test set-ups is shown in Figure 4.
able values, below 5,000 W/m2K, which could be fairly easily re- A synthetic jet entrains cool air from ambient, impinges on the
alized with hydrofluoroether (HFE) cooling fluids. For example, top hot surface and circulates the heated air back to the ambient
using an 8 x 8 cm2 heat spreader of some advanced composite with through the edges of the plate. A radial counter-current flow is
a k of 800 W/mK and a thickness of 4 mm results in a temperature created in the gap between the plates with hot air dispersed along
rise of 40°C with a heat transfer coefficient of only 2,500 W/m2K. the top and ambient air entrained along the bottom surface. The

Electronics-COOLING.com 19
idea was further explored by the development of flow actuators ficient called ‘nanolightning’ is being pursued by researchers from
using MEMS technology [8]. Purdue. It is based on ‘micro-scale ion-driven airflow using very
high electric fields created by nanotubes. As shown in Figure 5, the
ionized air molecules are moved by another electric field, thereby
inducing secondary airflow [9]. Cooling a heat flux level of 40 W/
cm2 has been reported. The technology is being commercialized
through a start-up company (Thorrn).

LIQUID COOLING
The widely known heat transfer guru John Lienhard [10] once
raised the question: “How much heat could possibly be carried
away by boiling?” The answer is: 2,000 kW/cm2 (based on water
molecules turning into vapor without influencing each other). The
highest reported experimental value is over 200 kW/cm2, using
high velocities and high pressures. Some commercially available
microcoolers can handle about 1 kW/cm2 so there is some room
for improvement. Liquid cooling for application to electronics is
generally divided into the two main categories of indirect and
direct liquid cooling. Indirect liquid cooling is one in which the
liquid does not directly contact the components to be cooled. Di-
rect liquid cooling brings the liquid coolant into direct contact
with the components to be cooled. The following sections discuss
the categories of indirect liquid cooling in the form of heat pipes
and cold plates and direct liquid cooling in the form of immersion
cooling and jet impingement.

HEAT PIPES
Heat pipes provide an indirect and passive means of applying liq-
uid cooling. They are sealed and vacuum pumped vessels that are
partially filled with a liquid. The internal walls of the pipes are
lined with a porous medium (the wick) that acts as a passive capil-
lary pump. When heat is applied to one side of the pipe the liquid
starts evaporating. A pressure gradient exists causing the vapor
to flow to the cooler regions. The vapor condenses at the cooler
regions and is transported back by the wick structure, thereby
Figure 4: Flow dynamics of normal jet impingement with an oscillating diaphragm. closing the loop. Heat pipes provide an enhanced means of trans-
porting heat (e.g., under many circumstances much better than
copper) from a source to a heat sink where it can be rejected to
the cooling medium by natural or forced convection. The effec-
tive thermal conductivity of a heat pipe can range from 50,000 to
200,000 W/mK [11], but is often much lower in practice due to ad-
ditional interface resistances. The performance of heat pipes scales
from 10 W/cm2 to over 300 W/cm2. A simple water-copper heat
pipe will on average have a heat transfer capacity of 100 W/cm2.
An example of a typical application of a heat pipe for electronics
cooling is given in Figure 6.

Although there is virtually no limit to the size of a heat pipe, the

effectiveness of a heat pipe decreases with decreasing lengths. For
heat pipes with a length less than about 1 cm the performance of
a solid piece of metal (e.g., copper) is often comparable. They are
Figure 5: ‘Nanolightning’ sketch. very effective as efficient heat conductors to transport heat to lo-
cations were more area is available. 2D heat spreaders (otherwise
‘NANOLIGHTNING’ known as vapor chambers) based on the heat pipe principle can
An interesting new approach to increasing the heat transfer coef- achieve much higher effective thermal conductivities than copper.

20 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

For example, a thin planar heat spreader has been developed that is effort to significantly extend cooling capability, Tuckerman and
claimed to have a thermal performance greater than diamond [12]. Pease [13] demonstrated a liquid-cooled microchannel heat sink
that removed 790 W/cm2 with a temperature increase of 71°C for a
600 ml/min flow rate with a pressure drop of 207 kPa. As a result
of the continuing increases in heat flux at the chip level, micro-
channel cold plates are receiving renewed attention.

MICROCHANNELS AND MINICHANNELS

The term ‘micro’ is applied to devices having hydraulic diameters
of 10 to several hundred micrometers, while ‘mini’ refers to diam-
eters on the order of one to a few millimeters. In many practical
cases, the small flow rate within micro-channels produces lami-
nar flow resulting in a heat transfer coefficient inversely propor-
tional to the hydraulic diameter. In other words, the smaller the
channel, the higher the heat transfer coefficient. Unfortunately,
the pressure drop increases with the inverse of the second power
of the channel width, keeping the mass flow constant, and limiting
ongoing miniaturization in practice.

Figure 6: Examples of heat pipes used in a notebook application. Garimella and Sobhan [14] published a very good review of the mi-
crochannel literature up to the year, 2000. They concluded, among
Besides standard heat pipes, loop heat pipes (LHP) such as those others, that “Given the diversity in the results in the literature, a
shown in Figure 7 are attracting increased attention. LHPs have reliable prediction of the heat transfer rates and pressure drops
the advantage over conventional heat pipes that the vapor and liq- in microchannels is not currently possible for design applications
uid paths are separated enabling much better performance of the such as microchannel heat sinks.” Mudawar [15] reviewed high-
liquid return loop. For example, Kim et al. [12] showed the ability heat-flux thermal management schemes, including ultra-high heat
to accommodate a heat flux of 625 W/cm2. fluxes in the range of 1,000-100,000 W/cm2. A recent overview was
also provided by Mohapatra and Loikitis [16].

Lee and Vafai [17] compared jet impingement and microchannel

cooling for high heat flux applications. One of their conclusions
is that microchannel cooling is more effective for areas smaller
than 7 x 7 cm. Integrated single and two-phase micro heat sinks
are treated by Gillot et al. [18]. They were able to cool about 450
W/cm 2 using both single- and two-phase heat transfer. For two-
phase flow the pumping power is about ten times lower and the
required flow rate is considerably lower. Kandlikar and Upadhye
[19] showed enhanced microchannel cooling by using off-set
strip fins and a split-flow arrangement. Cooling of over 300 W/
cm 2 at 24 kPa is claimed with a flow of 1.5 l/min. A paper devot-
ed to pumping requirements has been written by Singhal et al.
[20]. Useful graphs compare the performance of a whole range of
pumps that could be considered for microchannel cooling. Col-
Figure 7: Examples of loop heat pipes. gan et al. [21] at IBM published a practical implementation of a
silicon microchannel cooler (as shown in Figure 8) for high-pow-
COLD PLATES er chips.
Liquid-cooled cold plates perform a function analogous to air-
cooled heat sinks by providing an effective means to transfer heat They argued that it is not practical to form the microchannels di-
from a component to a liquid coolant. Unlike heat pipes they may rectly on the chip given the high cost of high-performance proces-
be considered active devices in that liquid is usually forced through sor chips. Instead, a separate microchannel cold plate is bonded
them by the action of a pump. For many years water-cooled cold to the back of the chip. This requires a very low interface thermal
plates were used in mainframe computers to cool high-powered resistance. If the microcooler is based on silicon, a rigid bonding
multi-chip processor modules. Vacuum-brazed finstock coldplates means that silver-filled epoxies or solder should be used. Power
are standard practice in defense electronics, and copper-based su- densities in excess of 400 W/cm2 are reported, for a flow of 1.2 l/
peralloy structures are used in high-energy lasers. In 1981, in an min at 30 kPa.

Electronics-COOLING.com 21
 ic technique for fabricating liquid cooling channels onto the backs
of high-performance ICs. They also built a system that would al-
low the on-chip cooling system to be connected to embedded flu-
idic channels built into a printed circuit board.

Figure 8: Pictures from IBM paper showing high-performance liquid cooling technol-
ogy using microchannels [21].

It may be possible to push microchannel heat transfer even further

by utilizing boiling. In addition to offering higher heat transfer
coefficients, boiling convection in microchannels is promising
because it requires less pumping power than single-phase liquid
convection to achieve a given heat sink thermal resistance. For the
same heat flux the pressure drops by a factor of 20. A review has
been published by Bergles et al. [22]. A prototype of a 1,000 W/cm2
cooling system based on boiling heat transfer in microchannel
heat sinks using a flow rate of 500 ml/min has been described in
[23]. The main practical problem with two-phase flow is its unpre-
dictability. Local heat transfer coefficients may change appreciably
over time leading to local temperature changes of 10-15°C [24].
Also backflow of already heated flow due to expansion of bubbles
is observed.

Increasingly, microchannel-like cold plates are becoming com-

mercially available. Mikros [25] claims 1,000 W/cm2K, 14-21 kPa
and 0.05 K/W/cm2 for its patented technology in which the fluid
impinges on the surface to be cooled. ACT (Advanced Cooling Figure 9: Vendor’s heat spreader and test results.
Technologies) [26] offers pumped liquid (both single- and two-
phase) cooling technologies in addition to loop heat pipes for ELECTROHYDRODYNAMIC
space applications. Their single-phase solution (see Figure 9) in- AND ELECTROWETTING COOLING
corporates a ‘unique oscillating flow heat transfer’ mechanism, As an alternative to a continuous flow set into motion by either
capable of cooling over 1,300 W/cm2. temperature differences or by mechanical means, liquid could also
be formed and moved in droplets of nano-to-milliliter size (see
Another company, Cirrus, sells microchannel high heat flux heat for example a nice demonstration by Nanolytics [30]) by means of
sinks, claiming over 100 W/cm2 heat flux capability and reduced electric fields.
pressure drops. Recently another player, iCurie [27], entered the
market. They claim a pumpless microchannel cooling system
based on capillary pumping with separate liquid-vapor loops. The
layout can be very flexible, only 300 micrometer thick. At least 200
W/cm2 can be cooled, with the additional advantage of a negligi-
ble effect of gravity. It is not a heat pipe because the principle of
operation is based on flow boiling, not evaporation. Other com-
panies offering microchannel coolers include Lytron, Novel Con-
cepts, Curamik, Microcooling Concepts (MC2 ), and Koolance.
Figure 10: Water forms a nearly perfect ball, as shown in left photo, suspended on
A completely different way of making microchannels is by using the tips of tiny blades of nanograss.
metal foams or metal made porous otherwise. Engineers from
Thermacore have described a family of liquid-cooled heat sinks Electrowetting on a dielectric film, in which the surface property
using porous metal. They accommodated a heat flux of 500 W/cm2 of a dielectric film can be modified between hydrophobic and hy-
for a 50 K difference at a pressure drop of 115 kPa using water [28]. drophilic states using an electric field, can be used to provide the
basis for a direct micropumping system. Electrowetting involves
On June 21, 2005, Georgia Tech [29] announced a novel monolith- control of the surface tension of a liquid and can cause a droplet of

22 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

liquid to bead (as shown in Figure 10) or spread out on the surface
depending upon its surface state.

One of the possible applications is cooling on a micro scale. The

recently published theoretical work of Pamula [31] has shown
a possible configuration based on fast moving droplets under a
chip. They showed that with 0.4 ml/min it is theoretically possible
to cool 90 W/cm2. Recently, Leuven University in collaboration
with Philips Research published two papers on this subject [32,
33]. The Philips approach differs from the Duke approach in that
it concerns an oscillating flow. At Bell Labs researchers coupled
electrowetting with nanostructured superhydrophobic surfaces
(coined ‘nanograss’) to result in dynamically tunable surfaces [34].
One application is the movement of droplets to cool hot spots;
however, no further heat transfer data are given.

A recent publication discussed another promising development, Figure 12: Typical configuration of liquid metal cooling loop for mobile applications.
the application of electrowetting to liquid metals [35]. The main
advantage besides a better heat transfer capability is the much IMMERSION COOLING
lower voltage required (2 instead of 50 V). However, no experi- Direct liquid or immersion cooling is a well-established meth-
mental data have been presented. od for accommodating high heat flux backed by over 30 years of
university and industrial research. With natural convection two-
The interesting aspect in combining microfluidics with elec- phase flow, generally termed nucleate pool boiling, the critical
tric control is that when all sizes scale down to micro scale, the heat flux using FC-72 is in the range of 5 to 20 W/cm2. However,
electro/-kinetic/-wetting/-osmotic forces become comparable to much higher heat fluxes up to 100 W/cm2 can be accommodated
pressure drop forces and therefore control of the liquid motion with surface enhancement of the heat source. Figure 13 illustrates
becomes easier. Of course, active cooling of a hot surface is one a device submerged in a pool of dielectric liquid. The heat dis-
thing, to remove heat from the heated liquid in a closed loop re- sipated in the device produces vapor bubbles that are driven by
quires additional heat exchange area. buoyancy forces into the upper region of the container, where the
vapor condenses and drips back into the liquid pool. One of the
LIQUID METAL COOLING disadvantages of this technique is the need for a liquid compatible
Of special interest is the work ongoing in the field of liquid metal with the device. Most often, water cannot be used because of its
cooling. Apart from heat pipes based on liquid metals, mainly for chemical and electrical characteristics.
the high-temperature range, an increasing amount of research is
devoted to the use of Ga-Sn-In eutectics that remain liquid down
to -19°C. In [36, 37] high-performance liquid metal cooling loops
are described using magneto fluid dynamic pumps, claiming over
200 W/cm2 cooling capacity, using a flow of 0.3 l/min at 15 kPa.
Examples of liquid cooling loops for electronics cooling applica-
tion are shown in Figures 11 and 12. Another advantage of liquid
metal is its much lower CTE compared to water and the fact that
freezing introduces fewer problems. Developments to extend the
use in cold environments to -40°C are ongoing.

Figure 13: Example of pool boiling (thermosyphon) cooling.

LIQUID JET IMPINGEMENT

Wang et al. [38] claim a cooling of 90 W/cm2 with a 100°C tem-
perature rise using a flow rate of only 8 ml/min. Researchers at
Georgia Institute of Technology studied a closed loop impinge-
Figure 11: Sketch of liquid metal cooling loop. ment jet [39]. Cooling of almost 180 W/cm2 has been realized us-

Electronics-COOLING.com 23
ing water, using a flow of 0.3 l/min at 300 kPa. The micropump
used 7 W to drive it.

At this point it is difficult to say which one is better, microchan-

nels or microjets. Microchannels are easier to fabricate and imple-
ment but the temperature nonuniformity is larger and the nucle-
ation is more difficult to control. Microjets achieve better cooling
uniformity but more fabrication steps are required and an initial
pressure is necessary to form the jet. An example of a commercial
concept for impingement liquid cooling is shown in Figure 14 [40].

Figure 15: Illustration of spray and jet impingement cooling.

Figure 14: Commercially available multiple jet impingement liquid cooling. A final method to be mentioned is inkjet-assisted spray cooling.
This method uses existing thermal inkjet technology. A critical
SPRAY COOLING heat flux of 270 W/cm2 is reported [45] using only 3 ml/min, with
In recent years spray cooling has received increasing attention as a COP of six, meaning that the inkjet pumping power is a factor of
a means of supporting higher heat flux in electronic cooling appli- six lower than the heat removed.
cations. Spray cooling breaks up the liquid into fine droplets that
impinge individually on the heated wall. Cooling of the surface is SOLID-STATE COOLING
achieved through a combination of thermal conduction through A thermoelectric or a Peltier cooler (as shown in Figure 16) is a
the liquid in contact with the surface and evaporation at the liq- small electronic heat pump that has the advantage of no moving
uid-vapor interface. The droplet impingement both enhances the parts and silent operation. Thermoelectric cooling enables cooling
spatial uniformity of heat removal and delays the liquid separa- below ambient temperature. The coolers operate on direct current
tion on the wall during vigorous boiling. and may be used for heating or cooling by reversing the direction
of current flow.
Spray evaporative cooling with a Fluorinert™ coolant is used to
maintain junction temperatures of ASICs on MCMs in the CRAY
SV2 system between 70 and 85°C for heat fluxes from 15 to 55
W/cm2 [41]. In addition to the CRAY cooling application, spray
cooling has gained a foothold in the military sector providing for
improved thermal management, dense system packaging, and re-
duced weight [42]. A research group at UCLA discussed chip-level
spray cooling for an RF power amplifier and measured a maxi-
mum heat flux of over 160 W/cm2 [43]. Isothermal Systems Re-
search manufactures SprayCool products [44].

Spray cooling and jet impingement (as shown in Figure 15) are
often considered competing options for electronic cooling. In
general, sprays reduce flow rate requirements but require a higher
nozzle pressure drop. Figure 16: Schematic of simple Peltier cooler.

24 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

When a positive DC voltage is applied to the n-type thermoele- new fine-grain micro-alloyed bismuth telluride materials that do
ment, electrons pass from the p- to the n-type thermoelement and not suffer the element geometry limitation and can offer higher
the cold side temperature decreases as heat is absorbed. performance [47]. Another serious step forward has been realized
by Nanocoolers through a proprietary wafer-scale manufacturing
The heat absorption (cooling) is proportional to the current and process. It concerns a monolithic process with thicknesses about
the number of thermoelectric couples. The main disadvantage is 1-2 micrometers (see Figure 18). They claim a tunable performance
that the heat transferred to the hot side is greater than the amount of 10-1,000 W/cm2 with a single stage ΔT of 50-70 K [48]. Micro-
of heat pumped by a quantity equal to the Joule heating (i.e.: I2R Pelt, a spin-off company from Fraunhofer and Infineon also sells
loss) that takes place in the Peltier elements. promising thin film thermoelectrics (see Figure 19). For their
near-future products they claim cooling of 160 W/cm2.
The three most important thermoelectric effects are the Seebeck,
Peltier, and Thomson effects. For thermoelectric cooling the
Thomson effect can be neglected. The Peltier coefficient Π and
Seebeck coefficient S are related to each other through Π = S ● T.
Thermoelectric materials are usually characterized by their Figure
of merit ZT, defined by:

σS2T
ZT =
λ

Where σ = electrical conductivity

λ = thermal conductivity
T = temperature in absolute units

This equation shows why it is difficult to obtain good thermo-

electric materials. A good heat conduction between the hot and
cold side) and a high electrical conductivity (to minimize Joule
heating). Due to the dependence of σ and λ (Wiedemann-Franz
law), it is almost impossible to optimize this ratio for the electron
contribution to the thermal conductivity. Hence, the approach is
to reduce the phonon thermal conductivity without a degrada-
tion of the electrical conductivity. The best materials so far are
alloys of Bi2Te3 with Sb2Te3 and Bi2Te3 with Bi2Se3. ZT is of the
order of one at room temperature. This value gives a Coefficient
of Performance (COP, a measure for the efficiency) of about 1 (see
Figure 17a), which compared to household refrigerators and air
conditioners (COP from 2 to 4), makes thermoelectric cooling
generally not competitive. The same holds for power generation
(see Figure 17b).
Figure 17: Comparison of thermoelectric technology with other energy conversion
Despite the low efficiency, the application areas are increasing and methods for (a) cooling and (b) power generation [46].
include infrared detector cooling, charge coupled devices (CCDs),
microprocessors, blood analyzers, portable picnic coolers. Princi-
pal applications are still accurate control of temperature and cool-
ing below ambient temperature.

One of the problems with traditional Peltier elements is their lim-

ited capability of cooling heat fluxes over 5-10 W/cm2. Because
the cooling density of a Peltier cooler is inversely proportion-
al to its length, scaling to smaller size is desirable. The material
structure produced by conventional crystal growth techniques
for producing bismuth telluride thermoelectric materials impose
significant limitations on thermoelectric element dimensions due
to poor manufacturing yields. This prevents thermoelectric ele-
ments from being made very short. Marlow Industries reported Figure 18: Comparison of standard and thin-film Peltier element.

Electronics-COOLING.com 25
 Figure 20: Comparison of effective thermal conductance.

SUPERLATTICE AND HETEROSTRUCTURE COOLING

Figure 19: Thin film Peltier element. For a number of years now, the strategy to improve thermoelectric
cooling has taken a new turn giving hope for the future. On a nano
A number of other research projects directed at miniaturizing Pel- scale level, coherent and incoherent transport plays an important
tier elements are worth mentioning. In 2004, Biophan Technol- role in the electron and phonon diffusion. Extensive research is
ogies [49] signed an agreement with NASA for characterization going on in this field. For example, Venkatasubramanian [56]
and joint development of high-density nanoengineered thermo- at RTI reported ZT values between two and three at room tem-
electric materials for use with implantable devices. They anticipate perature obtained with Bi2Te3/Se2Te3 superlattices. Cooling power
a breakthrough in power generation systems. A French/U.S. con- density is estimated as high as 700 W/cm2 at 353 K compared to 1.9
sortium published a paper [50] on the fabrication and modeling of W/cm2 in the bulk material (see Figures 21 and 22).
an in-plane thermoelectric micro-generator. They concluded that
a heating power of about 100 mW may be enough to produce 1
mW of useful electrical power in vacuum, using thin film technol-
ogy. A compact thermoelectric device may be able to produce 60
microwatt with an output voltage of 1.5 V in air. Applied Digital
Solutions uses its ThermoLife thermoelectric power generator to
power its implantable chips [51]. It generates 3V, has a diameter of
9.3 mm and weighs 230 mg. DTS in Germany uses thin film tech-
nology for their Low Power Thermoelectric Generators (LPTGs)
[52] that produce a few tens of microwatts in the volt range for a
temperature difference of a few degrees C. The generators have a
mass of 390 mg. Also in Germany, researchers at Dresden Univer-
sity [53] have found a way to make tiny thermoelectric generators
using copper foil as a template. Antimony-Bismuth thermocou-
ples are electro-deposited and after adding an epoxy film the cop- Figure 21: Estimated power density for superlattice devices as a function of current
per is etched away. The result is a cheap, flexible and recyclable
generator that converts environmental heat into electricity. The
growth by pulsed laser deposition of high-quality thermoelectric
cobaltate thin films on silicon has been reported by Yu et al. [54].
In addition, TEM characterization revealed nearly perfect crys-
talline structures of the C a3C o4O9 film formed on top of an SiOx
amorphous layer, suggesting self-assembly might be a viable tech-
nique for cobalt oxide-based thermoelectrics.

If the application is limited to temperatures above ambient tem-

perature matters are quite different. Because the heat flow paths
of the cooling and the conduction have the same direction,
high-thermal conductivity materials are preferred. Enerdyne’s Po-
lara heat spreaders are based on this principle [55]. A factor of five
improvement compared to copper is claimed as shown in Figure Figure 22: Potential COP as a function of ZT with various technologies. THOT refers
20. However, no samples are available as of July 2005. to the heat sink temperature

26 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

Thin-film related work is also being conducted at the Universi- nificant improvements in lower work function materials have no
ty of California Santa Cruz, based on SiGe/Si. The most recent future.
paper [57] quotes a cooling power density of nearly 600 W/cm 2
for a temperature difference of 4K below ambient for a 40 x 40 PHASE CHANGE MATERIALS
micrometer size area. The superlattice efforts of RTI are being AND HEAT ACCUMULATORS
commercialized through a spin-off company called Nextreme. Phase change materials are successfully used as heat-storing
Recent information reveals that despite their claimed value of materials for air conditioning, cool boxes, efficient fire-retard-
ZT = 2.4, they are not able to manufacture production samples ing powders, as functional materials for self-heating insoles for
with a ZT larger than 1.4. The focus is to reduce the parasitics boots and many other industrial applications. Their use for elec-
and to reduce even further the current 100 micrometer thick- tronics thermal management is limited to applications where
ness. time-dependent phenomena play a role. For example, reference
[65] discusses the use of phase change materials as compared to
However, there is still hope for a serious breakthrough. Very re- copper for use in a power semiconductor unit.
cently, Humphrey and Linke [58] published a paper called “Re-
versible Thermoelectric Materials.” They argue that nanostruc- Chemical heat accumulators should also be mentioned. For
tured materials with sharply peaked electronic density of states example, the use of composite materials based on granulated
(such as quantum wires) may operate reversibly, challenging the open-porous matrix filled with a hygroscopic substance can be
view that thermoelectric devices are inherently irreversible heat seen as a new approach to accumulate heat [66]. The advantage
engines. In this case, ZT values could reach a value of 10 at room is a significant increase in the heat that can be stored as com-
temperature, much above the value of 5 that is required for eco- pared to sensible heat and latent heat. For example, for a 100°C
nomical adoption of thermoelectric technology for mainstream temperature rise copper absorbs 40 kJ/kg. Evaporation of water
refrigeration and power generation. is associated with an absorption of 2,260 kJ/kg. The enthalpy of
a reversible chemical reaction can reach a value of 7,000 kJ/kg.
THERMIONIC AND THERMOTUNNELING COOLING A principal advantage of reversible chemical reactions for heat
Thermionic cooling is based on the principle that a high-work- accumulation is their ability to store the accumulated energy for
function cathode preferentially emits hot electrons [46]. Materi- a long time, if the reaction is controlled by the presence of either
als available have a work function of 0.7 eV or higher, which lim- a catalyst or a reagent. Hence, the major applications are in the
its the use to the higher temperature ranges (>500 K). Vacuum field of summer-winter heat storage for buildings, etc. Chemical
thermionic devices based on resonant tunneling have been pro- heat accumulators could potentially be used for outdoor elec-
posed more recently [59]. Cooling capabilities of 20-30°C with tronic applications when a night-day rhythm is present.
kW/cm 2 cooling power density can be achieved. However, since
the operating currents for the device are as high as 105 A/cm 2, CONCLUSIONS
effects such as Joule heating at the metal-semiconductor contact A number of approaches show interesting industrial potential
resistance and reverse heat conduction have limited the experi- for the cooling of high-power electronics. This prospect is at-
mental cooling results to <1°C. tested to by the number of small companies that are entering the
market. For example, there are now companies engaged in the
Devices based on quantum tunneling through a small gap are development and commercialization of microchannels, spray
being commercialized [60]. The spacing between the cathode cooling, synthetic jets, thin film Peltier elements. For heat flux
and the anode should be of the order of 10 nm, providing quite densities up to and maybe even beyond 50 W/cm2 air-cooling
an engineering challenge. Much larger cooling power than ther- may remain the cooling option of choice. For heat fluxes over
moelectric superlattice coolers are predicted by Hishinuma et al. 100 W/cm 2, some form of liquid-cooling appears to be the most
[61] (e.g. 10 kW/cm 2 for 50 K cooling at room temperature). viable option. Several papers have demonstrated solutions that
may be industrially feasible for application in the range between
Recently a study has been devoted to their potential use as ener- 500 and 1,000 W/cm 2. Considering the range of efforts underway
gy scavenging or power conversion devices [62]. Unfortunately, to extend conventional cooling technologies, as well as develop
the conclusion is that a gap an order of magnitude lower must be new ones, the future seems bright for accommodating high-heat
achieved to be of interest for these application fields. Even more flux applications.
worrying is another recent study [63] showing that contrary to
the results of Hishinuma only about 16 W/cm2 can be reached (Note: We deemed it instructive to include examples of commer-
with a ΔT of 40°C, while the maximum COP is only 0.25. More cially-available thermal solutions. However, it should be clearly
or less the same conclusion can be drawn from a paper presented stated that we do not intend to promote any of the mentioned
at THERMINIC 2005, September 27-30 [64]. products. Finally, we have tried to cover the state-of-the-art as
known to us. However, given the broadness of the field, we may
Herein, some weaknesses in prior studies are discussed and it is have overlooked some important new developments for which
clear from the conclusions that nanogap solutions without sig- we apologize.)

Electronics-COOLING.com 27
REFERENCES SemiTherm Symposium, San Jose, CA, pp. 1-7, 2005.
1. Lasance, C., Technical Data column, ElectronicsCooling, Jan- 22. Bergles, A.E., et al., “Boiling and Evaporation in Small Diam-
uary 1997. eter Channels,” Heat Transfer Engineering, Vol. 24, pp. 18-40,
2. Rogers, P., CALCE, personal communication, April 2005. 2003.
3. http://www.reactivenanotechnologies.com/ . 23. Faulkner, D., Khotan, M., and Shekarriz, R., “Practical Design
4. Subramanian, J., et al. “Room Temperature Soldering of Micro- of 100 W/cm2 Cooling System,” Proceedings of 19th SemiTh-
electronic Components for Enhanced Thermal Performance,” erm Symposium, San Jose, CA, pp. 223-230, 2003.
EuroSimE 2005, pp. 681-686, 2005. 24. Vadakkan, U., Intel, personal communication, March 2005.
5. Zweben, C., “Ultra-High Thermal Conductivity,” Proceedings 25. http://www.mikrostechnologies.com .
of 21st SemiTherm Symposium, San Jose, CA, pp. 168-174, 26. http://www.1-act.com/ .
2005. 27. Lee, J., “A Macro Cooling Solution With Micro/Nano Technol-
6. http://novelconceptsinc.com/index.htm ogy,” Proceedings of Next- Generation Thermal Management
7. Acikalin, T., Wait, S., Garimella, S., Raman, A., “Experimen- Materials and Systems,” Phoenix, AZ, June 2005.
tal Investigation of the Thermal Performance of Piezoelectric 28. North, M., and Cho, W., “High Heat Flux Liquid-Cooled Po-
Fans,” HeatTransfer Eng., Vol. 25, pp. 4-14, 2004. rous Metal Heat Sink,” Proceedings of IPACK 2003, Maui, HI,
8. Beratlis, N., and Smith, M., “Optimization of Synthetic Jet Paper # 35320, 2003.
Cooling for Micro-electronics Applications,” Proceedings of 29. http://www.gtresearchnews.gatech.edu/newsrelease/cooling.
19th SemiTherm Symposium, San Jose, CA, pp. 66-73, 2003. htm.
9. Peterson, M., Fisher, T., Garimella, S., and Schlitz, D., “Exper- 30. http://www.nanolytics.com .
imental Characterization of Low Voltage Field Emission from 31. Pamula, V.K., and Chakrabarty, K., “Cooling of Integrated Cir-
Carbon-Based Cathodes in Atmospheric Air,” Proceedings of cuits Using Droplet-Based Microfluidics,” Proceedings ACM
IMECE’03, Paper #41775, 2003. Great Lakes Symposium on VLSI, pp. 84-87, 2003.
10. Lienhard, J., “High Heat Flux in Japan,” Engines of Our Inge- 32. Oprins, H., Vandevelde, B., Beyne, E., Borghs, G., and Bael-
nuity, No. 448, http://www.uh.edu/engines/epi448.htm . mans, M., “Selective Cooling of Microelectronics Using Elec-
11. Thyrum, G., “Critical Aspects of Modeling Heat Pipe Assisted trostatic Actuated Liquid Droplets-Modelling and Experi-
Heat Sinks,” http://www.thermacore.com/pdfs/critical.pdf . ments, 10th THERMINIC, pp. 207-212, 2004.
12. Kim, J., and Golliher, E., “Steady State Model of a Micro Loop 33. Oprins, H., Nicole, C.C.S., Baret, J.C., Van der Verken, G., La-
Heat Pipe,” Proceedings of 18th SemiTherm Symposium, San sance, C., and Baelmans, M., “On- Chip Liquid Cooling With
Jose, CA, pp. 137-144, 2002. Integrated Pump Technology,” Proceedings of 21st SemiTherm
13. Tuckerman, D.B., and Pease, R.F., “High Performance Heat Symposium, San Jose, CA, pp. 15-17, 2005.
Sinking for VLSI,” IEEE Electron Device Letters, EDL-2, 5, pp. 34. Krupenkin, T. et al., “From Rolling Ball to Complete Wetting:
126-129, 1981. the Dynamic Tuning of Liquids on Nanostructured Surfaces,
14. Garimella, S., and Sobhan, C., “Transport in Microchannels – Langmuir, 20, pp.3824-3827, 2004.
A Critical Review,” Ann. Rev. Heat Transfer, Vol. 13, pp. 1-50, 35. Mohseni, K., “Effective Cooling of Integrated Circuits Using
2003. Liquid Alloy Electrowetting,” Proceedings of 21st SemiTherm
15. Mudawar, I., “Assessment of High-Heat-Flux Thermal Man- Symposium, San Jose, CA, pp. 20-25, 2005.
agement Schemes,” IEEE CPT Trans., Vol. 24, pp. 122-141, 36. Miner, A., and Ghoshal, U., “Cooling of High-Power Density
2001. Microdevices Using Liquid Metal Coolants, Appl. Phys. Lett.,
16. Mohapatra, S., and Loikitis, D., “Advances in Liquid Coolant Vol.85, 506-508, 2004.
Technologies for Electronics Cooling,” Proceedings of 21st 37. Ghoshal, U., Grimm, D., Ibrani, S., Johnston, C., and Miner, A.,
SemiTherm Symposium, San Jose, CA, pp. 354-360, 2005. “High-performance Liquid Metal Cooling Loops, Proceedings
17. Lee, D.Y., and Vafai, K., “Comparative Analysis of Jet Impinge- of 21st SemiTherm Symposium, San Jose, CA, pp. 16-19, 2005.
ment and Microchannel Cooling for High Heat Flux Applica- 38. Wang, E., et al., Micromachined Jets for Liquid Impingement
tions,” Intl. Jour. of Heat and Mass Transfer, Vol. 42, pp. 1555- Cooling of VLSI Chips, J. Microelectromech. Syst., vol.13,
1568, 1999. pp.833-842, 2004.
18. Gillot, C., Meysenc, Schaeffer, and Bricard, A., IEEE CPT, Vol. 39. Bintoro, J., et al., “A Closed Loop Impinging Jet Cooling for
22, No. 3, pp. 384-389, 1999. Computer Chip,” IMECE2003, Washington D.C., Paper #
19. Kandilikar, S., and Upadhye, H., “Extending the Heat Flux 43865, 2003.
Limit With Enhanced Microchannels in Direct Single-Phase 40. http://siliconpower.danfoss.com/Products/PDF/Shower%20
Cooling of Computer Chips,” Proceedings of 21st SemiTherm Power.pdf .
Symposium, San Jose, CA, pp. 8-15, 2005. 41. Pautsch, G., and Bar-Cohen, A., “An fiverview on the System
20. Singhal, V., et al., “Analysis of Pumping Requirements for Mi- Packaging of the Cray SV2 Supercomputer,” Ipack 2001 Con-
crochannel Cooling Systems,”IPACK’03, Paper # 35237, 2003. ference, Kauai, HI, 2001.
21. Colgan, E., et al., “A Practical Implementation of Silicon Mi- 42. Cader, T., and Tilton, D., “Implementing Spray Cooling Ther-
crochannel Coolers for High Power Chips,” Proceedings of 21st mal Management in High Heat Flux Applications,” Proceed-

28 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

ings of 2004 InterSociety Conference on Thermal Phenomena, “Thin-film Thermoelectric Devices With High Room-Tem-
pp. 699-701, 2004. perature Figures of Merit, Nature 413, pp. 597-602, October
43. Cotler, A., et al., “”Chip-Level Spray Cooling of an LD-MOS- 2001.
FET RF Power AmpliOer,” IEEE CPT Trans., Vol. 27, pp. 411- 57. Zeng, G., et al., “Cooling Power Density of SiGe/Si Superlat-
416, 2004. tice Micro Refrigerators,” Mat. Res. Soc. Symp. Proc., Vol. 793,
44. http://www.spraycool.com/\ash/index.html#home . 2004.
45. Bash, C., et al., “Inkjet Assisted Spray Cooling of Electronics,” 58. Humphrey, T., and Linke, H., “Reversible Thermoelectric Ma-
IPACK2003, Maui, HI, Paper# 35058, 2003. terials,” Phys. Rev. Letters, Vol. 94, 096601, 2005.
46. Chen, G., and Shakouri, A., “Heat Transfer in Nanostructures 59. Korotkov, N., and Likharev, K.K., “Possible Cooling by Reso-
for Solid-State Energy Conversion, J. of Heat Transfer, Vol. 124, nant Fowler-Nordheim Emission, Appl. Phys. Lett., Vol.75, p.
p. 242, 2002. 2491, 1999.
47. Bierschenk, J., and Johnson, D.A., “Latest Developments in 60. http://www.powerchips.gi/index.shtml .
Thermoelectrically Enhanced Heat Sinks,” ElectronicsCool- 61. Hishinuma, Y., Geballe, T., Moyzhes, B., and Kenny, T., “Re-
ing, Vol. 11, No. 3, pp. 24-32, August 2005. frigeration by Combined Tunneling and Thermionic Emission
48. Ghoshal, U., “Advanced Thermoelectric Cooling Systems,” in Vacuum: Use of Nanometer Scale Design,” Appl. Phys. Lett.,
Proceedings of Next-Generation Thermal Management Mate- Vol. 78, No. 17, pp. 2572-2574, 2001.
rials and Systems,” Phoenix, AZ, June 2005. 62. Despesse, G., and Jager, T., “Exploitation of the Thermotunnel
49. http://www.biophan.com/pr/release_051404.html . Effect for Energy Scavenging,” Appl. Phys. Lett.. Vol. 96, No. 9,
50. Jacquot, A., Liu, W.L., Chen, G., Fleurial, J.-P., Dauscher, A., pp. 5026-5031, 2004.
and Lenoir, B., “Fabrication and Modeling of an In-Plane Ther- 63. Chua, H., et al., “Thermionic and Tunneling Cooling Thermo-
moelectric Microgenerator,” Proceedings of Twenty-First In- dynamics,” Appl. Phys. Lett., Vol. 84, pp. 3999-4001, 2004.
ternational Conference on Thermoelectrics, pp. 561-564, 2002. 64. Gerstenmaier, Y. and Wachutka, G., “Thermionic Refrigeration
51. http://www.adsx.com/content/index.htm . with Planar and Nonplanar Electrodes: Chances and Limits”,
52. http://www.dts-generator.com/main-e.htm . Proceedings 11th THERMINIC, September 27-30, Belgirate,
53. Qu, W., et al., “Microfabrication of Thermoelectric Generators Italy, 2005, pp. 270-277
on Flexible Foil Substrates as a Power Source for Autonomous 65. Krishnan, S., and Garimella, S., “Thermal Management of
Microsystems,” J. Micromech. Microeng., Vol. 11, pp. 146- 152, Transient Power Spikes in Electronics – Phase Change Energy
2001. Storage or Copper Heat Sinks?”, IPACK2003, Maui, HI, Ppaper
54. Yu, H., et al., “In Situ Growth of C-Axis-Oriented Ca3Co4O9 # 35169, 2003.
Thin Films on Si (100),” Appl. Phys. Lett., Vol. 86, 082103, 2005. 66. Levitskij, E.A., Aristov, Yu. I., Tokarev, V.N., and Parmon, V.N.,
55. http://www.enerdynesolutions.com/tech_polara.html “Chemical Heat Accumulator,” Solar Energy Material and So-
56. Venkatasubramanian, R., Silvola, E., Colpitts, T., O’quinn, B, lar Cells, 44, pp. 219-235, 1996.

Electronics-COOLING.com 29
F E AT U R E D

Estimating Parallel Plate-Fin

Heat Sink Thermal Resistance
Reprinted from February 2003
Robert E. Simons
*IBM Corp.
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
Over the last 20 years, I have collected many simple tools into a spreadsheet that I can use for making quick engineering esti-
mates. One prominent worksheet in that collection is a calculator that makes use of this “Calculation Corner” and its companion
article, “Estimating Parallel Plate-fin Heat Sink Pressure Drop”, to predict the thermal performance of plate fin heat sinks. This is
just one example of the many extremely valuable “Calculation Corner” columns that Bob Simons contributed to the entire elec-
tronics cooling industry through Electronics Cooling. (Ross Wilcoxon)

A
INTRODUCTION interface, Rint. The maximum allowable temperature at the heat
s noted previously in this column, the trend of increas- sink attachment surface, Tbase, is given by
ing electronic module power is making it more and
more difficult to cool electronic packages with air. As a (1)
result, there are an increasing number of applications
that require the use of forced convection air-cooled heat sinks to
control module temperature. An example of a widely used type The maximum allowable heat sink resistance, Rmax, is then given by
of heat sink is the parallel plate configuration shown in Figure 1.
(2)

where Tair,in, is the temperature of the cooling air at the inlet to the
heat sink passages. At this point many thermal engineers will start
looking at heat sink vendor catalogs (or more likely today start
searching vendors on the internet) to find a heat sink that will fit
in the allowable space and provide a heat sink thermal resistance,
Rhs, less than Rmax at some specified flow rate. In some cases, it
may be useful to do a sizing to estimate Rhs for various plate-fin
heat sink designs to determine if a feasible design configuration is
possible. The remainder of this article will provide the basic equa-
tions to do this. The thermal resistance of the heat sink is given by

(3)

where h is the convective heat transfer coefficient, Abase is the ex-

posed base surface area between fins, Nfin is the number of fins, ηfin
Figure 1: Parallel plate fin heat sink configuration. is the fin efficiency, and Afin is the surface area per fin taking into
account both sides of the fin.
In order to select the appropriate heat sink, the thermal designer
must first determine the maximum allowable heat sink thermal To proceed further it is necessary to establish the maximum al-
resistance. To do this it is necessary to know the maximum allow- lowable heat sink volume in terms of width, W, height, H, and
able module case temperature, Tcase, the module power dissipa- length in the flow direction, L. It is also necessary to specify a fin
tion, Pmod, and the thermal resistance at the module-to-heat sink thickness, tfin. Using these parameters, the gap, b, between the fins

30 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

may be determined from coefficient is given by

(11)
(4)

where kfluid is the thermal conductivity of the cooling fluid (i.e.
The exposed base surface area may then be determined from air). The efficiency of the fins may be calculated using

(5)
(12)

and the heat transfer area per fin from
where m is given by
(6)

(13)
At this point it is necessary to specify the air flow rate either in
terms of the average velocity, V, between the fins or a volumetric
flow rate, G. If a volumetric flow rate is used, the corresponding and kfin is the thermal conductivity of the fins.
air velocity between the fins is
Using these equations, it is possible to estimate heat sink thermal
(7) performance in terms of the thermal resistance from the tempera-
ture at the base of the fins to the temperature of the air entering
the fin passages. It may be noted that the relationship for Nusselt
To determine the heat transfer coefficient acting upon the fins, an number (8) includes the effect of the temperature rise in the air
equation developed by Teertstra et al. [1] relating Nusselt number, as it flows through the fin passages. To obtain the total thermal
Nub, to Reynolds number, Re, and Prandtl number, Pr, may be resistance, Rtot, to the base of the heat sink it is necessary to add in
employed. This equation is the thermal conduction resistance across the base of the heat sink.
For uniform heat flow into the base Rtot is given by

(8) (14)

and kbase is the thermal conductivity of the heat sink base.

The Prandtl number is

(9)

where μ is the dynamic viscosity of air, cp the specific heat of air

at constant pressure, and k is the thermal conductivity of air. The
Reynolds number used in (8) is a modified channel Reynolds
number defined as

(10)

where ρ is the density of air. Equation (8) is based upon a com-

posite model spanning the developing to fully developed laminar
flow regimes and was validated by the authors [1] by comparing
with numerical simulations over a broad range of the modified
channel Reynolds number (0.26 <Reb < 175) and with some ex-
perimental data as well.
Figure 2: Effect of fin height and number of fins on heat sink thermal resistance at
Using the Nusselt number obtained in (8) the heat transfer an air velocity of 2.5 m/s (492 fpm).

Electronics-COOLING.com 31
 For purposes of illustration these equations were used to estimate
heat sink thermal resistance for a 50 x 50 mm aluminum heat
sink. The effect of increasing the fin height and the number of fins
is shown in Figure 2 for a constant air velocity, and in Figure 3 for
a constant volumetric flow rate. In both cases it may be seen that
there are limits to how much heat sink thermal resistance may be
reduced by either increasing fin length or adding more fins. Of
course, to determine how a heat sink will actually perform in a
specific application it is necessary to determine the air velocity or
volumetric flow rate that can be delivered through the heat sink.
To do this, it is necessary to estimate the heat sink pressure drop
characteristics and match them to the fan or blower to be used.
This is a topic for consideration in a future article.

REFERENCES
1. Teertstra, P., Yovanovich, M.M., and Culham, J.R., “Analytical
Forced Convection Modeling of Plate Fin Heat Sinks”, Pro-
Figure 3: Effect of fin height and number of fins on heat sink thermal resistance at ceedings of 15th IEEE SEMI-THERM Symposium, pp. 34-
a volumetric air flow rate of 0.0024 m3/s (5 CFM). 41, 1999.

32 Electronics COOLING | FALL 2020

Accelerate package thermal
modeling in your design
Generate clean, accurate, CAD based detailed chip
package thermal models in minutes for use in in
your full 3D electronics cooling system simulation
with Simcenter Flotherm Package Creator.

Now available for Simcenter Flotherm XT, CAD

centric electronics cooling software, and Simcenter
FLOEFD, CAD embedded CFD software.

Find out more about more online and discover

Simcenter Portfolio of simulation and test solutions
in domains such as thermal, structural, vibration,
reliability, acoustics and electromagnetics.

siemens.com/simcenter
F E AT U R E D

Calculating Interface Resistance

Reprinted from May 1997

M. M. Yovanovich, J. R. Culham and P. Teertstra

*University of Waterloo
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
My spreadsheet of useful analysis tools also includes multiple contributions from Mike Yovanovich and his co-workers. This article
provides a comprehensive overview of what factors lead to interface resistance and a straightforward method for predicting the
effects of materials, surface conditions, pressure, the use of a thermal interface material, etc. I have probably used my calculator
that implements the analysis, which is shown in this article, hundreds of times during my career. (Ross Wilcoxon)

T
INTRODUCTION
he exposed surface area of many of today’s high-pow- The subject of thermal resistance at interfaces between aluminum
ered electronic packages is no longer sufficient for the heat sinks and ceramic packages has been discussed by Lee [1], de
removal of the heat generated during normal operation. Sorgo [2], Latham [3], and Early et al. [4]. These articles primarily
Heat sinks are a commonly-used, low cost means of in- report test results for joint resistance as a function of contact pres-
creasing the effective surface area for dissipating heat by means sure for various interface types. The interfaces examined in these
of convective air cooling. While the use of a heat sink lowers the works involve either bare surfaces (air filled) or joints where the
fluid-side thermal resistance, it also introduces an interface resis- interstitial gap is filled with a material layer containing dispersed
tance across the contact formed between itself and the package thermally conductive fillers. Interstitial material layers currently
case. Under some circumstances, this contact resistance can be used by the industry, as described by de Sorgo [2], include ther-
substantial, impeding heat flow and reducing the overall effective- mal greases, thermally conductive compounds, elastomers, and
ness of the heat sink. Figure 1 depicts an electronic package heat adhesive tapes.
sink assembly which would typically be joined by plastic or metal
spring clips around the perimeter of the assembly.

Figure 2: Contact Configurations

Figure 1: Ceramic Package—Aluminum Heat Sink Assembly The objective of this article is to illustrate how to calculate the ther-

34 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

mal joint resistance for the interface formed by two conforming, general complex because it depends on several geometric and
rough surface shown in Figure 2a, as a function of contact pressure physical parameters, such as the Vickers microhardness correla-
for the low pressure range, between 0.035 and 0.35 MPa (5 and tion coefficients.
50 psi), commonly encountered in microelectronic applications
(Latham [3]). Peterson and Fletcher [5] verified by experiments The surface asperity slope is frequently not given. In this case, the
in vacuum that the following models, which were originally de- mean absolute asperity slope can be approximated by the correla-
veloped for metal-to-metal contacts, give very good results when tion equation proposed by Antonetti et al. [10]:
used to predict the contact conductance at interfaces formed by
metals (invar, Kovar, and alloy 42) and mold compounds (Poly- m=0.125(σ*106 )0.402
set 410B and 410C, MG25F-LMP, and MG45F-04) at the interface
temperature range: 20°C to 70°C and the interface pressure range: which was developed for the surface roughness range:
0.5 to 5.0 MPa. This work will focus primarily on bare joints, al-
though an example where the interface material is treated as a liq- 0.216 µ m ≤ σ < 9.6 µ m
uid, such as in the case of a thermal grease, will also be considered.
Interfaces with thermal compounds or elastomeric sheet materials The gap conductance, hg, is given by the approximation of Yova-
will be shown to be very difficult to model and will be discussed in novich [6]:
general terms only. The non-conforming wavy, convex, or concave
interfaces depicted in Figures 2b, 2c, and 2d, respectively, are ex-
ceedingly complex to model and therefore will not be considered
here. Since radiation heat transfer at most interfaces is negligible or where kg is the thermal conductivity of the gap substance. The
non-existent, it will not be included in this analysis. effective gap thickness Y, shown in Figure 3, can be calculated ac-
curately by means of the simple power-law correlation equation
CONFORMING ROUGH SURFACE MODEL proposed by Antonetti and Yovanovich [7]:
The thermal joint conductance, hj, of the interface formed by two
conforming, rough surfaces is given by the following simple mod-
el proposed by Yovanovich [6] and further described and used by
Antonetti and Yovanovich [7], Yovanovich and Antonetti [8], and
Yovanovich [9]. for the relative contact pressure range:

hj=hc+hg 10-5 < P/Hc < 2 x 10-2

The contact conductance is given by:

where ks is the harmonic mean thermal conductivity of the inter-

face:

the effective mean absolute asperity slope of the interface m, as

shown in Figure 3, is given by:

and where σ, also shown in Figure 3, is the effective RMS surface

roughness of the contacting asperities: Figure 3: Conforming rough surfaces

The gas parameter M accounts for rarefaction effects at high tem-

peratures and low gas pressures. This gas-surface parameter de-
pends on the thermal accommodation coefficients, the ratio of
The contact pressure is P and Hc is the surface microhardness of specific heats, the Prandtl number, and the molecular mean free-
the softer of the two contacting solids. The microhardness is in path of the gas. Song and Yovanovich [11] present correlation

Electronics-COOLING.com 35
equations for the calculation of the accommodation coefficients where hj,1 and hj,2 refer to the joint conductance between each of
for several gases as a function of the gas temperature. This com- the contacting surfaces and the interfacial material and t and k
plex gas-surface parameter depends on gas pressure and tempera- are the average thickness and thermal conductivity of the layer.
ture according to the relationship: Completing this analysis requires characterization of the relevant
surface parameters, such as the slope, roughness and microhard-
ness, for the various interstitial materials. In addition, for elasto-
meric materials the layer thickness t is not constant but instead
depends on the contact pressure. Additional research needs to be
where M0 denotes the gas parameter value at the reference values done before a model can be developed to address this complex
of gas temperature and pressure, T0 and Pg,0, respectively. Refer- phenomenon.
ence values of the gas parameter for air and helium are presented
in Table 1. APPLICATION TO ALUMINUM HEAT SINK-CERAMIC
PACKAGE INTERFACE
Table 1
Thermal Conductivity Gas Parameter Table 2: Thermal and Surface Properties for Aluminum-Alumina
Gap Substance
(W/mK) M0 x 106, m Conforming Rough Surfaces
air 0.026 0.373**
Thermal Surface
helium 0.150 2.05** Microhardness
Material Conductivity Roughness
MPa
thermal grease 0.20 – 0.70* 0.0 (W/mK) μm
doped thermal grease 1.68 – 2.58* 0.0
A1 5052 [14] 140 745 6.9
** – T0 – 50°C, Pg,0 = 1 atm
* – AOS Technical Data Sheets, 1995 A1 6061 [14] 180 705 0.7
A1 6063-T5 201 1094 0.4 (flycut)
INTERSTITIAL MATERIAL LAYERS
Although the conforming rough surface model presented in the Aluminum Nitride [13] 160 10044 0.45
previous section was developed for bare surfaces, it can also be
Alumina (96% A12O3) 20.9 3100 1.3 (ground)
applied to interfaces with thermal grease. By assuming that the
grease behaves as a liquid and fills all gaps between the contacting Copper [13] 397 924.1 0.45 (milled)
asperities, the existing model can be used by substituting M = 0
and the thermal conductivity of the grease into the gap conduc-
The aforementioned models will be used to calculate the joint
tance relationship. However, when solid interstitial materials are
resistances for the interface formed by an aluminum 6063-T5
used, such as thermal compounds, elastomers or adhesive tapes,
aluminum heat sink and Al2O3alumina package. The thermal
the joint conductance problem becomes much more complicated.
conductivities of the heat sink and ceramic package are k1= 201
As shown in Figure 4, the use of a solid interstitial material intro-
W/mK and k2 = 20.9 W/mK respectively. The harmonic mean
duces an additional interface to the problem.
thermal conductivity of the interface is ks = 37.85 W/mK. Since
the microhardness of the aluminum alloy is 1094 MPa, which is
much less than that of the alumina, it will be used to compute the
contact parameters. Based on a surface roughness for flycut alu-
minum of σ1= 0.4 µm and a surface roughness for ground alumina
of σ2 = 1.3 µm, the effective surface roughness of the interface is
calculated as σ = 1.36 µm. Since the surface slopes are not given,
Eq. (6) will be used to calculate the following values: m1 = 0.139,
m2 = 0.0865, respectively. The effective surface slope of the inter-
face is therefore m = 0.164. The thermal and physical properties
of air, helium and grease presented in Table 1 will be used in the
Figure 4: Thermal conductance across an interface with and without an interstitial gap conductance model.
material
In Figure 5 the joint thermal resistances, whose units are cm2
Using thermal resistance concepts, the overall joint conductance °C/W are plotted against the nominal contact pressure over the
for this problem is determined by the series combination: pressure range: 0.007 ≤ P (MPa) ≤ 0.35 for several cases. The bare
joint resistances with air or helium present in the gap are shown.
The effect of a thermal grease of thermal conductivity kg = 0.20
W/mK is also shown in Figure 5.

36 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

REFERENCES
1. S. Lee, How to Select a Heat Sink, Electronics Cooling, Vol. 1,
No. 1, June 1995, pp. 10-14.
2. Miksa de Sorgo, Thermal Interface Materials, Electronics
Cooling, Vol. 2, No. 2, September 1996, pp. 12-15.
3. Carol A. Latham, Thermal Resistance of Interface Materials
as a Function of Pressure, Electronics Cooling, Vol. 2, No. 2,
September 1996, p. 35.
4. Malcolm Early, Seri Lee, and Mark Pellilo, Thermal Perfor-
mance of Interface Material in Microelectronics Packaging
Applications, Proceedings of the 1995 International Elec-
tronics Packaging Conference, September 1995, pp. 534-544.
5. G.P. Peterson and L.S. Fletcher, Evaluation of the Thermal
Contact Conductance Between Substrate and Mold Com-
pound Materials, ASMEHTD-Vol. 69, Fundamentals of Con-
duction and Recent Developments in Contact Resistance, ed-
ited by M. Imber, G.P. Peterson and M.M. Yovanovich, 1987,
Figure 5: Joint thermal resistance of an aluminum heat sink-ceramic package as- pp.99-105.
sembly for various contact pressures. 6. M.M. Yovanovich, New Contact and Gap Correlations for
Conforming Rough Surfaces, AIAA-81-1164, presented at
SUMMARY AND DISCUSSION AIAA 16thThermophysics Conference, Palo Alto, CA., June
Simple correlation equations are presented and used to calculate 1981.
thermal joint resistances for a typical aluminum-ceramic inter- 7. V.W. Antonetti and M.M. Yovanovich, Thermal Contact Re-
face found in microelectronics applications. Flycut and ground sistance in Microelectronic Equipment, Thermal Manage-
surfaces are considered. Joint resistances are calculated for con- ment Concepts in Microelectronic Packaging From Com-
tact pressures between 0.007 and 0.35 MPa, which includes the ponent to System, ISHM Technical Monograph Series 6984-
practical microelectronic pressure range of 0.07 and 0.17 MPa 003,1984, pp. 135-151.
(see Latham [3]). The greatest joint resistances are found when 8. M.M. Yovanovich and V.W. Antonetti, Application of Ther-
air is present in the interstitial gap. In the contact pressure range mal Contact Resistance Theory to Electronic Packages, Ad-
of 0.007 to 0.35 MPa, the air joint resistance goes from 2.665 to vances in Thermal Modeling of Electronic Components and
1.903 cm2°C/W. Systems, Vol. 1, Editors A.Bar-Cohen and A.D. Kraus, Hemi-
sphere Publishing Corporation, 1988, pp. 79-128.
When silicon grease is placed in the gap, the joint resistance is 9. M.M. Yovanovich, Theory and Applications of Constric-
much smaller than the bare interface. The calculated values of the tion and Spreading Resistance Concepts for Microelectronic
joint resistance lie in the range 0.335 to 0.213 cm2°C/W which are Thermal Management, Cooling Techniques for Computers,
an order of magnitude smaller than the joint resistances of a bare Editor Win Aung, Hemisphere Publishing Corporation,
joint. If greases with thermally conductive ceramics are used, the 1991, pp. 277-332.
joint resistance can be reduced to values below 0.065 cm2°C/W. 10. V.W. Antonetti, T.D. Whittle, and R.E. Simons, An Approxi-
mate Thermal Contact Conductance Correlation, HTD-Vol.
The correlation equations which have been used are based on 170, Experimental/Numerical Heat Transfer in Combustion
conforming rough surfaces with interstitial substances which and Phase Change, 1991, pp.35-42.
perfectly wet all portions of the surfaces which form the gap. 11. S. Song and M.M. Yovanovich, Correlation of Thermal Ac-
Any non-flatness will result in interfaces with larger gaps which commodation Coefficient for Engineering Surfaces, ASME
will have larger joint resistances. If the interstitial substance does HTD-Vol. 69, Fundamentals of Conduction and Recent De-
not perfectly wet the contacting surfaces, this will also produce a velopments in Contact Resistance, edited by M. Imber, G.P.
more thermally resistive interface. The proposed models and cor- Peterson and M.M. Yovanovich, 1987, pp. 107-116.
relation equations therefore correspond to the best thermal joints
which have the smallest joint resistances.

The use of other interstitial materials, such as thermal com-

pounds, elastomers or adhesive tapes, has been shown to increase
the complexity of the joint conductance problem significantly. In
order to successfully model this problem, extensive research into
characterizing the surface properties and layer thicknesses for the
various interfacial materials is required.

Electronics-COOLING.com 37
F E AT U R E D

Calculating Spreading Resistance in Heat Sinks

Reprinted from January 1998

Seri Lee
*Amkor Electronics
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
A classical study calculating the thermal spreading resistance, used over the years by many practicing engineers to solve the
chip to substrate to PCB heat spreading challenges. This valuable material transcends time and is an excellent reference for our
thermal community. (Victor Chiriac)

Accident? Consider the scenario where a designer wishes to incor- In cases where the footprint of a heat sink need not be much larg-
porate a newly developed device into a system and soon learns that er than the size of the heat source, the contribution of the spread-
a heat sink is needed to cool the device. The designer finds a rather ing resistance to the overall device temperature rise may be insig-
large heat sink in a catalog which marginally satisfies the required nificant and usually falls within the design margin. However, in
thermal criteria. Due to other considerations, such as fan noise and an attempt to remove more heat from today’s high performance
cost constraints, an attempt to use a smaller heat sink proved futile, devices, a larger heat sink is often used and, consequently, the im-
and so the larger heat sink was accepted into the design. A proto- pact of spreading resistance on the performance of a heat sink is
type was made which, unfortunately, burned-out during the initial becoming an important factor that must not be ignored in the
validation test, the product missed the narrow introduction time, design process. It is not uncommon to find in many high perfor-
and the project was canceled. What went wrong? mance, high power applications that more than half the total tem-
perature rise of a heat sink is attributed to the spreading resistance
The reasons could have been multi-fold. But, under this scenario, the in the base-plate.
main culprit could have been the spreading resistance that was over-
looked during the design process. It is very important for heat sink The objectives of this article are:
users to realize that, unless the heat sink is custom developed for a 1. To understand the physics and parameters associated with
specific application, thermal performance values provided in ven- spreading resistance
dor’s catalogs rarely account for the additional resistances coming 2. To provide a simple design correlation for accurate predic-
from the size and location considerations of a heat source. It is un- tion of the resistance
derstandable that the vendors themselves could not possibly know 3. To discuss and clarify the concept of spreading resistance
what kind of devices the users will be cooling with their products. with an emphasis on the practical use of the correlation in
heat sink applications

S
INTRODUCTION
preading or constriction resistances exist whenever heat The correlation provided herein was originally developed in Ref-
flows from one region to another in different cross section- erences [1] and [2]. This article is an extension of the earlier pre-
al area. In the case of heat sink applications, the spreading sentation.
resistance occurs in the base-plate when a heat source of
a smaller footprint area is mounted on a heat sink with a larger SPREADING RESISTANCE
base-plate area. This results in a higher local temperature at the Before we proceed with the analysis, let us attend to what the tem-
location where the heat source is placed. Figure 1 illustrates how perature distributions shown in Figure 1 are telling us. The first
the surface temperature of a heat sink base-plate would respond obvious one, as noted earlier, is that the maximum temperature at
as the size of the heat source is progressively reduced from left to the center increases as the heat source becomes smaller. Anoth-
right with all other conditions unchanged: the smaller the heat er important observation is that, as the temperature rises in the
source, the more spreading has to take place, resulting in a greater center, the temperatures along the edges of the heat sink decrease
temperature rise at the center. In this example, the effect of the simultaneously. It can be shown that this happens in such a way
edge surfaces of the heat sink is ignored and the heat source is that the area-averaged surface temperature of the heat sink base-
assumed to be generating uniform heat flux. plate has remained the same. In other words, the average heat sink

38 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

thermal performance is independent of the size of a heat source. dimensions of 100 x 100 x 1.3 mm thick. According to the catalog,
In fact, as will be seen later, it is also independent of the location the thermal resistance of this heat sink under a given set of condi-
of the heat source. tions is 1.0°C/W. Find the maximum resistance of the heat sink if
used to cool a 25 x 25 mm device.
The spreading resistance can be determined from the following
set of parameters: SOLUTIONS
• Footprint or contact area of the heat source, As With no other specific descriptions, it is assumed that the heat
• Footprint area of the heat sink base-plate, Ap source is centrally mounted, and the given thermal resistance of
• Thickness of the heat sink base-plate, t 1.0°C/W represents the average heat sink performance. From the
• Thermal conductivity of the heat sink base-plate, k problem statement, we summarize:
• Average heat sink thermal resistance, R0 • As = 0.025 x 0.025 = 0.000625 m2
• Ap = 0.1 x 0.1 = 0.01 m2
We will assume, for the time being, that the heat source is cen- • t = 0.0013 m
trally mounted on the base-plate, and the heat sink is cooled uni- • k = 200 W/mK
formly over the exposed finned surface. These two assumptions • R0 = 1.0°C/W
will be examined in further detail. Figure 2 shows a two-dimen-
sional side view of the heat sink with heat-flow lines schematical- Therefore,
ly drawn in the base-plate whose thickness is greatly exaggerat-
ed. At the top, the corresponding surface temperature variation λ= π3/2/√0.01+ 1/√0.000625 = 95.68 m-1
across the center line of the base-plate is shown by the solid line.
The dotted line represents the average temperature of the surface tanh(λt)= tanh(95.68 x 0.0013) = 0.124
which is, again, independent of the heat source size and can be
easily determined by multiplying R0 with the total amount of heat λkApR0 = 95.68 x 200 x 0.01 x 1.0 = 191.4
dissipation, denoted as Q.

As indicated in Figure 2, the maximum constriction resistance

Rc, which accounts for the local temperature rise over the average
surface temperature, is the only additional quantity that is needed Hence, the maximum resistance, Rtotal, is:
for determining the maximum heat sink temperature. It can be
accurately determined from the following correlation. Rtotal = Ro + Rc = 1.0 + 0.66 = 1.66°C/W

EFFECT OF SOURCE LOCATION

(1) In the following two sections, we will limit our examination to

the current example problem. As we shall see, the result of this
where
limited case study will allow us to draw some general yet useful
conclusions. Suppose the same heat source in the above example
(2) was not centrally located, but mounted a distance away from the

center. Obviously, the maximum temperature would further rise
as compared to that found in the above example. Figure 3 shows
the local resistances corresponding to two such cases:

Figures 1, 2: Two dimensional schematic view of local resistance or temperature

variation of a heat sink shown with heat flow Lines

EXAMPLE PROBLEM
Consider an aluminum heat sink (k = 200 W/mK) with base-plate Figure 3: Heat sink local resistance showing the effect of source location

Electronics-COOLING.com 39
From L to R, heat source at (37.5,0) and (37.5,37.5) the first one plots of Cf obtained for many other cases exhibit essentially the
is for the case where the heat source is mounted midway along same profile as that shown in Figure 4, with magnitudes at the
the edge, and the other, where it is mounted on one corner of the corners determined from Equation (3), and the domain of the
heat sink. plot defined by the maximum displacement of the heat source.
Based on this observation, a general conclusion can be made: for
For these two special cases, the maximum spreading resistance all practical purposes, as long as the heat source is placed closer to
can be calculated by using Equation (1) for Rc with input parame- the center than to the edges of the heat sink, the correctional in-
ters t and R0 modified as shown below: crease in the spreading resistance may be ignored, and Cf =1 may
be used. As noted above, this would introduce a small error of no
Rc = C x Rc (Ap, As, k, t/C, Ro/C) (3) greater than 5-10% in the spreading resistance which, in turn, is a
fraction of the total resistance.
with C = 2^(1/2) for the first case, and C = 2 for the second case.
It is to be noted that this expression is independent of the source So far, we have assumed a uniform cooling over the entire
size. Numerically, for the current problem with a 25 x 25 mm heat finned-surface area of the base-plate. Although this is a useful
source, it results in the maximum spreading resistances of 1.29 and assumption, it is seldom realized in actual situations. It is well
2.38°C/W, or the total resistances of 2.29 and 3.38°C/W for the first known that, due to the thinner boundary layer and the less down-
and second cases, respectively. For both cases, it can be shown that stream heating effect, a device would be cooled more effectively if
the average surface resistance has not changed from unity. it is mounted toward the air inlet side. Again, a numerical simula-
tion is carried out using our example problem with the boundary
For other intermediate source locations, numerical simulations layer effect included.
were carried out and a plot is provided in Figure 4 for the correc-
tion factor Cf , which can be used to compute the total resistance as Figure 5 shows the resulting modified correction factor as a func-
tion of the distance from the center of the heat sink to the heat
Rtotal = Ro + Cf Rc (4) source placed along the center line at y = 0: x = -37.5 mm corre-
sponds to the front most leading edge location of the heat source
Where: Rc is determined from Equation (1), given for the case and x = 37.5 mm the rear most trailing edge placement.
with the heat source placed at the center.

Figure 4: Correction factor as a function of source location Figure 5: Correction factor modified for boundary layer effect at y=0

The coordinates in Figure 4 indicate the location of the center As can be seen from the figure, it is possible to realize a small
of the heat source measured from the center of the base-plate in improvement by placing the heat source forward of the center lo-
mm: the case with a centrally located heat source corresponds to cation where Cf < 1. However, it was experienced in practice that
(0,0), and the cases shown in Figure 3 correspond to (37.5,0) and accommodating a heat source away from the center and ensuring
(37.5,37.5) for the first and second cases, respectively. Only one its mounting orientation often cause additional problems during
quadrant is shown in Figure 4 as they would be, owing to the as- manufacturing and assembly processes.
sumption of uniform cooling, symmetrical about (0,0). As can be
seen from the figure, the correction factor increases from 1 as the SUMMARY AND DISCUSSION
heat source is placed away from the center. It is worthwhile not- A simple correlation equation is presented for determining
ing that the increase is, however, very minimal over a wide region spreading resistances in heat sink applications. A sample calcula-
near the center, and most increases occur closer to the edges. tion is carried out for a case with a heat source placed at the center
of the heat sink base-plate and a means to estimate the correction
Unlike C in the earlier expression, Cf is case dependent (i.e. it de- factor to account for the effect of changing the heat source lo-
pends on the heat-source size). However, it was found that the cation is provided. It is to be noted that the correlation provid-

40 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

ed herein is a general solution which reduces to the well-known as the fins in a planar heat sink. This additional spreading in the
Kennedy’s solution [3] when R0 approaches 0: the mathematical fins usually affects the spreading resistance in a similar way to a
equivalent of isothermal boundary condition. Kennedy’s solution thicker base-plate. The current author found that an increase of
is valid only when R0 is sufficiently small such that the fin-side of 20% in the base-plate thickness during the calculation roughly ac-
the heat sink base-plate is close to isothermal. Otherwise, Kenne- counts for the effect of this additional spreading in the fins of the
dy’s solution, representing the lower boundary of the spreading same material for most planar heat sinks under air cooling. No
resistance, may result in gross underestimation of the resistance. modification is required for pin-fin heat sinks.

The earlier study revealed that, depending on the relative magni-

tude of the average heat sink resistance, the spreading resistance REFERENCES
may either increase or decrease with the base-plate thickness. If 1. 1S. Lee, S. Song, V. Au, and K.P. Moran, Constriction/Spread-
the heat sink resistance is sufficiently small, as in liquid cooled ing Resistance Model for Electronic Packaging, Proceedings
heat sink applications, the spreading resistance always increases of the 4th ASME/JSME Thermal Engineering Joint Confer-
with the thickness, and an optimum thickness does not exist. On ence, Vol. 4, 1995, pp. 199-206.
the other hand, if the heat sink resistance is large, as experienced 2. S. Song, S. Lee, and V. Au, Closed Form Equation for Thermal
in most air-cooled applications, the spreading resistance decreas- Constriction/Spreading Resistances with Variable Resistance
es with the thickness and a finite optimum thickness exists. Boundary Condition, Proceedings of the 1994 IEPS Confer-
ence, 1994, pp. 111-121.
It is to be noted that the present correlation calculates the spread- 3. D. P. Kennedy, Spreading Resistance in Cylindrical Semicon-
ing resistance only in the base-plate and does not account for the ductor Devices, Journal of Applied Physics, Vol. 31, 1960, pp.
effect of additional spreading that may exist in other places, such 1490-1497.

We approach liquid
cooling like engineers,
because we
are engineers.

It’s crucial to pay attention to connectors when designing liquid cooling systems. As a
component that is often in close proximity to valuable electronics, reliability is key. CPC,
a leader in connection technology, is trusted by manufacturers to create robust, high-
performing couplings. Our expert engineers will work with you throughout the design
process to help identify products that best meet your needs so you can feel confident in
your liquid cooling system. Learn more at cpcworldwide.com/liquidcooling

Electronics-COOLING.com 41
F E AT U R E D

Thermal Interface Materials: A Brief Review of

Design Characteristics and Materials
Reprinted from February 2004

Ravi Mahajan, Chia-Pin Chiu, and Ravi Prasher

*Intel Corporation
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
This article provides a review of the thermal interface materials, a topic of significant interest to our thermal community. The
study includes a comprehensive study of the types and the main features and characteristics of the thermal interface materials:
the thermal conductivity, the bond line thickness (BLT), the bulk and the contact resistances, also various useful correlations and
comparisons to empirical models. A valuable study for the practicing thermal engineer. (Victor Chiriac)

I
INTRODUCTION role in connecting the different aspects of the solution (i.e., the
n the last few decades, as microprocessors have continued to heat source [the die]) with the heat-spreader and/or the heat sink,
evolve along Moore’s law, providing increased functionality ensuring efficient transfer of heat. The focus of this article is to
and performance, there has been an associated increase in discuss the desired characteristics of the TIM, the different classes
cooling demand driven both by the increase in raw power and of TIMs, and their advantages and limitations.
in local power densities on the die, commonly referred to as “hot
spots”[1, 2]. Considerable attention has therefore been given to the
development of thermal solutions and to the characterization of all
aspects of the thermal solution. The scope of the development has
been constrained significantly by cost pressures and the desire to
develop solutions that meet stringent form factor constraints as the
marketplace has seen an evolution of form factors increasing the
form, fit, and function demands on the thermal solutions. In par-
allel, the relentless pace of microprocessor evolution implies that
solution developers must also deal with limited time schedules.

In light of the environmental pressures described above, industrial

and academic researchers have partnered well to effectively bring Figure 1: Schematic illustration of the two thermal architectures:
thermal solutions to the marketplace, while, in parallel, systemat- (a) Architecture I is typically used in laptop applications.
ically investigating the fundamental aspects of thermal solution (b) Architecture II is typically used in desktop and server applications.
development. As discussed in [2, 3], a simple classification of ther- (Legend: I – Heat Sink, II – TIM, III – IHS, IV – TIM, V – Die, VI – Underfill, and
mal solutions may be made into two architectures, schematically VII – Package Substrate)
illustrated in Figure 1. A key assumption here is that the bulk of the
heat transfer occurs through the inactive side of the silicon and that DESIRED CHARACTERISTICS OF A TIM
thermal management addresses this primary heat transfer path. As discussed in the introduction, the TIM acts to connect the
different parts of a thermal solution. After inserting a TIM
In the first architecture, a heat sink is typically applied directly to between the solid surfaces, the effective thermal resistance, RTIM,
the back side of a silicon device through a thermal interface mate- at the interface will have two components, i.e., the bulk resistance,
rial (TIM). Such architectures are often used in space-constrained Rbulk, of the TIM arising from its finite thermal conductivity and
applications, such as laptops. The second architecture has a se- the contact resistance, Rc between the TIM and the adjoining so-
parate heat spreader interfacing directly with the inactive side lids, as shown in Figure 2. RTIM may be expressed as:
of the die, and a heat sink interfaces to the heat spreader. In this
architecture there are two separate applications of the TIM. It is
(1)
important to note that in both architectures, the TIM plays a key

42 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

where BLT is the bond-line thickness of the TIM, kTIM is the thermal includes the effect of the interface resistance between the filler and
conductivity of the TIM, and Rc1 and Rc2 are the contact resistances the matrix on the effective thermal conductivity of composites.
of the TIM with the two adjoining surfaces. Note that RTIM, Rc1, and The modified Bruggeman model for KTIM/ km >>1 is given by:
Rc2 are area-normalized thermal resistances (K-m2/W). One of the
goals of thermal design is to reduce RTIM. This can be accomplished
by reducing the BLT, increasing the thermal conductivity and re- (3)
ducing the contact resistances Rc1 and Rc2. Let us examine in detail
how each of these may be accomplished.
where α is the Biot number and is given as:

(4)

where Rb is the interface resistance between the filler and the par-
ticle and d is the diameter of the particle. Note that Rb is expressed
in units of area-normalized thermal resistance (Km2/W). The slid
line in Figure 3 represents the results of the modified Bruggeman
model for aluminum filler laden silicon-based grease for different
values of α. Thermal conductivity of silicone was assumed to be
0.2 W/mK. Figure 3 shows that model matches very well with 40%
and 50% volume fraction with α = 0.06 and for 60% volume frac-
tion with α = 0.15. This suggests that interface resistance for 60%
sample is higher than 40% and 50% samples. Rb could arise be-
Figure 2: Schematic of various resistance components of RTIM. cause of imperfect mixing of the particle with the polymer matrix,
or due to phonon acoustic mismatch, or due to a combination of
THERMAL CONDUCTIVITY the two phenomenon.
In typical applications [4, 5], the TIM serves to conduct heat
through its thickness. The thermal conductivity of a TIM is typi-
cally enhanced by loading a soft, sometimes liquid-like polymeric
material matrix with conducting solid particles, such as alumi-
num, alumina, and boron nitride. If the design requirement is
that the TIM should be thermally conducting but electrically in-
sulating then the ceramic-based filler particles are more typically
chosen. Figure 3 shows the variation of thermal conductivity of
silicone based thermal grease as a function of filler (aluminum)
particle volume fraction [6] where km is the thermal conductivity
of the silicone oil.

Thermal conductivity of particle laden TIMs can be functionally

expressed as:

(2)
Figure 3: Thermal conductivity vs. particle volume fraction for a silicone based ther-
mal grease with aluminum fillers.
Where kf is the filler thermal conductivity, Φ, filler volume frac-
tion, and Rb is the contact resistance between the fillers and the Prasher et al. [6] have shown that the contribution due to phonon
polymeric matrix. There are various analytical models available acoustic mismatch at room temperatures or higher is important
in the literature [6]. For spherical particles, one of the most pro- only when the thermal conductivity of both the materials is very
minent models is the Maxwell model. This model matches the high. In the present case, it is safe to assume the Rb is arising from
data for spherical particles with Φ up to 30%-35%, after which imperfect wetting or mixing of the filler particles with the silicon
percolation phenomenon takes over. Maxwell’s model cannot be oil, as the thermal conductivity of silicon oil is very low. Therefore
used to predict the thermal conductivity for higher volume frac- the phonon acoustic mismatch component is very low at room
tion due to the assumptions built into it. Prasher et al. in [6] used temperatures. The reason for higher Rb for 60% volume fraction
the modified Bruggeman model due its ability to predict thermal could be due to difficulty in wetting the surface of the particle
conductivity from low to high volume fraction and also because it with the silicone oil as the volume fraction is so high.

Electronics-COOLING.com 43
 to an optimal filler loading for the minimization of RTIM [6].

CONTACT RESISTANCE
Prasher [8] showed that sum of contact resistance of TIMs with
the two adjoining substrates can be written as:

(6)

where σ1 and σ2 are the surface roughness of the substrates, Anominal

is the nominal area of heat transfer, and Areal is the real area of
heat transfer. As shown in Figure 5, the real area of heat transfer
is smaller than the nominal area because of the trapped air in the
valleys of a microscopically rough surface of the substrates. In [7]
Areal was calculated by performing a force balance due to: 1) ap-
Figure 4: BLT vs. P/τy for various TIM materials. plied pressure, 2) capillary force due to the surface tension of the
TIM, and 3) back pressure due to the trapped air.

Prasher [7] compared the model for phase change and grease type
TIMs. Figure 6 shows a comparison between the model and ex-
perimental data on phase change material. Based on this, a few
general design guidelines were proposed to minimize contact re-
sistance; i.e., (1) increase pressure, (2) decrease surface roughness,
(3) increase thermal conductivity of the TIM, and (4) increase ca-
pillary force by changing the surface chemistry.

Figure 5: Mechanism of heat transfer near the TIM substrate interface.

BOND-LINE THICKNESS (BLT)

BLT reduction is often another parallel goal of thermal design.
BLT is a function of various parameters such as application pres-
sure (i.e., pressure applied in bringing the two contact surfaces
together) and particle volume fraction. Prasher et al. [6] have de-
veloped an empirical model for the BLT of particle laden polyme-
ric thermal interface materials. They conducted a study on eight
different formulations of particle laden polymeric TIMs, which
included thermal greases and phase change materials and pro-
posed the following correlation for BLT:

Figure 6: Comparison of the surface chemistry model with experimental results for
(5) phase change materials.

RELIABILITY OF TIMS IN TEMPERATURE CYCLING

where τy is the yield stress of the TIM and P is the applied pres- In packages using thermal grease as the conducting medium
sure. This correlation was validated in the pressure range of 25- between the die and the thermal solution, grease pump-out during
200 psi application pressures. Figure 4 shows their results. Since operation of the part is a known failure mechanism [8]. Traditio-
the yield stress of the TIM increases with increasing filler loading, nally the power cycle test is a direct method to examine grease
BLT is higher for higher volume fraction. Therefore there are two reliability. However, it is a time consuming process due to its long
competing effects with regard to filler loading for the thermal re- heating and cooling times. In order to screen numerous thermal
sistance of the TIM: kTIM increases and BLT also increases with grease materials during the initial design phase of a microproces-
increasing filler volume fraction at the same pressure, which leads sor package, it is advantageous to utilize a quick turn test.

44 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

As described in [8], an accelerated mechanical test was developed (b) Keep the ratio of G’ and G” greater than or equal to 1. Fi-
to evaluate interface degradation, due to pump-out. An MTS uni- gure 8 shows the results on the degradation of thermal resis-
versal testing machine was used to simulate the squeezing action tance per cycle from Prasher and Matayabas [9] on eight diffe-
on the grease, caused by die warpage change. rent samples of thermal greases.

Figures 7a and 7b show the typical grease pump-out pattern and

temperature trend observed in the quick test, respectively. By
using this accelerated testing method, the grease pump-out phe-
nomena can be predicted so that product design time can be si-
gnificantly reduced.

Figure 8: Effect of G’/G” on the degradation rate, as measured by thermal perfor-

mance of gel TIMs subjected to temperature cycling.

The labels present the value of G for each sample. The data la-
bels in Figure 8 are the values of G for different samples. So far
Figure 7a: Post-test thermal grease pattern for the case with grease pump-out. the discussion has focused on the desired design characteristics
of the TIM. These need to be characterized experimentally. Over
the last few years, a number of techniques have been elucidated
and the researcher today can rely on the hot and cold plate steady-
state method [10], a thermal test vehicle in steady-state test [11],
a transient test [12], and sandwich-type samples by the laser flash
method [13].

RE-WORKABILITY
Another typical requirement of TIMs is that of re-workability.
Since in a number of applications, the heat sink is attached to the
device by the OEM, re-workability is a requirement to avoid yield
loss due to heat sink attach. Re-workability implies that the heat
sink should be easily removed and that the TIM should be easily
cleanable so that the heat sink may be reattached if needed. This
requirement has led to certain classes of materials seeing conti-
Figure 7b: Temperature trend for a case with grease pump-out.
nued popularity; e.g., filled greases, filled phase change materials,
and certain gels.
To solve the pump-out problem gel TIMs were recently developed.
Gels are also thermal greases but they cure due crosslinking of the
CLASSES OF TIM MATERIALS
polymer at high temperatures. Recently Prasher and Matayabas
Table 1 shows the characteristics of some of the typical TIM mate-
[9] proposed the following design rules for the formulation of
rials and their advantages and disadvantages.
TIMs to avoid pump out problems and also to have a low thermal
resistance:
SUMMARY
In this paper, a general overview of the issues that are important
(a) Minimization of G, where G is given by
in TIM design and the desired characteristics of TIMs have been
presented. It was demonstrated that important attributes of a TIM
(7) are its thermal conductivity, and its interactions with the mating
surfaces; i.e., the resultant BLT and wetting characteristics. Issues,
where G’ and G” are the storage and loss modulus of the polyme- such as reworkability and reliability, also play important roles in
ric TIM, and the final selection of the TIM.

Electronics-COOLING.com 45
 Table 1: Summary of Characteristics of Some Typical TIM Materials

TIM Type General Characteristics Advantages Disadvantages

• High bulk thermal conductivity

Typically silicone based matrix
• Thin BLT with minimal attach pressure • Susceptible to grease pump-out and phase separation
loaded with particles (typically
Greases • Low Viscosity enables matrix material to easily fill surface crevices • Considered messy in a manufacturing environment due to a
AIN or ZnO) to enhance thermal
• TIM curing not required tendency to migrate
conductivity
• TIM delamination is not a concern

• Higher viscosity leads to increased stability and hence less

• Lower thermal conductivity than greases
Phase Polyolefin, epoxy, low molecular susceptible to pumpout
• Surface resistance can be greater than greases. Can be reduced
Change weight polyesters, acrylics • Application and handling is easier compared to greases
by thermal pre-treatment
Materials typically with BN or Al 2O3 filters • No cure required
• Requires attach pressure to increase thermal effectiveness
• Dealamination is not a concern

• Cure process needed

Al, Al 2O3 , Ag particles in silicone, • Conforms to surface irregularity before cure
Gels • Lower thermal conductivity than grease
olefin matrices that require curing • No pump out of migration concerns
• Lower adhesion than adhesives; delamination can be a concern

• Cure process needed

• Conform to surface irregularity before cure
Typically Ag particles in a cured • Delamination post reliability testing is a concern
Adhesives • No pump out
epoxy matrix • Since cured epoxies have high post cure modules, CTE
• No migration
mismatch induced stress is a concern

REFERENCES Model for Thermal Contact Resistance of Fluidic Interstitial

1. Mahajan, R., Nair, R., Wakharkar, V., Swan, J., Tang, J., and Thermal Interface Materials,” Journal of Heat Transfer, Vo-
Vandentop, G., “Emerging Directions for Packaging Tech- lume 123, 2001.
nologies,” Intel Technology Journal, Semiconductor Tech- 8. Chiu, C-P., Chandran, B., Mello, M., and Ken, K., “An Ac-
nology and Manufacturing, Volume 6, Issue 2, 2nd Quar- celerated Reliability Test Method to Predict Thermal Grease
ter, 2002, http://developer.intel.com/technology/itj/2002/ Pump-Out in Flip-Chip Applications,” Proceedings of 51th
volume06issue02/index.htm Electronic Components & Technology Conference, Orlan-
2. Mahajan, R., Brown, K., and Atluri, V., “The Evolution of do, FL, May/June 2001, pp. 91-97.
Microprocessor Packaging,” Intel Journal of Technology, 9. Prasher, R. and Matayabas, J., “Thermal Contact Resistance
3rd Quarter, 2000, http://developer.intel.com/technology/ of Cured Gel Polymeric Thermal Interface Material,” Sub-
itj/q32000.htm mitted to IEEE Transactions on Components and Packa-
3. Atluri, V., Mahajan, R., Patel, P., Mallik, D., Tang, J., ging Technology, 2003.
Wakharkar, V., Chrysler, G., Chiu, C-P., Choksi, G., 10. Aoki, R., and Chiu, C-P., “A Testing Apparatus for Ther-
Viswanath, R., “Critical Aspects of High-Performance Mi- mal Interface Materials,” Proceedings of 1998 International
croprocessor Packaging,” MRS Bulletin, January 2003, Vo- Symposium on Microelectronics, San Diego, CA, November
lume 28, No. 1, pp. 21-34. 1998, pp. 1036 – 1041.
4. Chiu, C-P., Solbreken, G., LeBonheur, V., and Xu Y., “Ap- 11. Solbrekken, G.; Chiu, C-P.; Byers, B.; and Reichenbacher,
plication of Phase Change Materials in Pentium III and D., “The Development of a Tool to Predict Package Level
Pentium III Xeon Processor Cartridges,” Proceedings of Thermal Interface Material Performance,” Thermal and
the International Symposium and Exhibition on Advanced Thermomechanical Phenomena in Electronic Systems –
Packaging Materials Processes, Properties and Interfaces, Proceedings of the Intersociety Conference, Las Vegas, Ne-
March, 2000, pp. 265-270. vada, May 2000, pp. 48-54.
5. Chiu Chia-pin, Viswanath Ram, and Solbreken Gary, 12. Chiu, C-P., and Solbrekken, G., “Characterization of Ther-
“Pentium II Processor Packaging-Thermal Management mal Interface Performance Using Transient Thermal Analy-
of SRAM’s in SEC Cartridge,” Proceedings of SPIE – The sis Technique,” Proceedings of 1999 International Systems
International Society for Optical Engineering, V 3582, No- Packaging Symposium, San Diego, California, January
vember 1998, pp. 833-838. 1999, pp. 227-232.
6. Prasher, R., Shipley, J., Prstic, S., Koning, P., and Wang, J-L, 13. Chiu, C-P., Maveety, J. and Varner, L., “Thermal Conduc-
“Thermal Resistance of Particle Laden Polymeric Thermal tivity and Resistance Measurements of Three-Layer Com-
Interface Materials,” Journal of Heat Transfer, Volume 125, posites with Solders,” Proceedings of 2000 International
December 2003. Symposium on Microelectronics, Boston, Massachusetts,
7. Prasher, R., “Surface Chemistry and Characteristics Based September 2000, pp. 846-851.

46 Electronics COOLING | FALL 2020

 CALL FOR PAPERS
37th Annual Semiconductor Thermal Measurement, Modeling and Management Symposium
March 22-26, 2021 at the DoubleTree by Hilton, San Jose, CA USA

About SEMI-THERM
SEMI-THERM is an international symposium dedicated to the thermal management and characterization of
electronic components and systems. Its goals are to:
• Provide knowledge covering all thermal length scales from integrated circuits to facility levels
• Foster discussions between thermal engineers, professionals and industry experts
• Encourage the exchange of information on academic and industrial advances in electronics cooling
Topics Include: Component/Board/System Thermal Design, Fluid Movers, Acoustics, Advanced Materials,
Measurement Methods, Modeling & Simulation, Additive Manufacturing, Reliability, etc.
Applications Include: Processors/ICs/Memory, 3-D packaging, Computing Systems, Data Centers,
Portable/Consumer/Wearable Electronics, Power Electronics, Harsh Environments, Defense/Aerospace
Systems; Solid-State Lighting & Cooling, Biomedical; Micro/Nano-scale Devices, etc.
Symposium Highlights

Technical Technical Vendor Vendor Product Tear Panel Poster/Dialog

Tutorials
Sessions Short Courses Exhibits Workshops Downs Discussions Session

Three options for participating in the technical program (additional details provided on FAQs):
Peer-reviewed paper: Submit a full manuscript for peer review in October. Authors notified of acceptance
in November. Reviewer comments provided to authors in December. Final manuscript due in January.
Manuscripts will be provided to conference attendees and uploaded to IEEE Xplore.
Non-peer-reviewed paper: Submit an extended abstract (2-5 pages) that describes the scope, contents,
key results, findings and conclusions. Authors notified of acceptance in November. Final manuscript due in
January. Manuscripts will be provided to conference attendees and uploaded to IEEE Xplore.
Presentation only: Submit an extended abstract (2 -5 pages) that describes the scope, contents, key
results, findings and conclusions. Authors notified of acceptance in November. Final presentation slides
are due in March. Presentations will be provided to conference attendees.
Note for authors of accepted 2020 SEMI-THERM ( 36) presentation/manuscripts: Please refer to the
FAQs for guidance on resubmission of manuscripts and presentation-only submissions:
https://semi-therm.org/presenters

Awards: All papers with manuscripts are eligible for the Best Paper Award. Student papers presented at the
conference are eligible for Student Scholarships. Presentation-only submissions are not eligible for awards.

Manuscripts and extended abstracts Date that authors are notified of Photo-ready full manuscript
submission deadline acceptance submission due date
Oct 4, 2020 Nov 15, 2020 Jan 17, 2021

Upload your paper electronically in RTF, DOC or PDF formats at www.semi-therm.org.

For further information please contact the Program Chair
Marcelo del Valle, E-mail: mvalle@infinera.com
Visit our website: http://www.semi-therm.org
** All authors qualify for reduced symposium rates **
F E AT U R E D

Heat Pipes for Electronics Cooling Applications

Reprinted from September 1996

Scott D. Garner
*Thermacore Inc.
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
This article is one of the most read articles by our community readers. It features a comprehensive overview of the operation, the
design, the limits, and the performance of heat pipes. This passive technology is the most efficient means of heat transportation
at relatively low cost. Over the last decades, heat pipe designs have evolved to be thinner (flat heat pipes) and to provide heat
spreading capabilities (vapor chamber). They are widely used in applications ranging from consumer to IGBT. (Genevieve Martin)

A
INTRODUCTION
ll electronic components, from microprocessors to high As heat is input at the evaporator, fluid is vaporized, creating a
end power converters, generate heat and rejection of this pressure gradient in the pipe. This pressure gradient forces the
heat is necessary for their optimum and reliable opera- vapor to flow along the pipe to a cooler section where it condenses
tion. As electronic design allows higher throughput in smaller giving up its latent heat of vaporization. The working fluid is then
packages, dissipating the heat load becomes a critical design fac- returned to the evaporator by the capillary forces developed in the
tor. Many of today’s electronic devices require cooling beyond wick structure.
the capability of standard metallic heat sinks. The heat pipe is
meeting this need and is rapidly becoming a mainstream ther-
mal management tool.

Heat pipes have been commercially available since the mid

1960’s. Only in the past few years, however, has the electronics
industry embraced heat pipes as reliable, cost-effective solutions
for high end cooling applications. The purpose of this article is
to explain basic heat pipe operation, review key heat pipe de-
sign issues, and to discuss current heat pipe electronic cooling
applications.

HEAT PIPE OPERATION

A heat pipe is essentially a passive heat transfer device with an
extremely high effective thermal conductivity. The two-phase
heat transfer mechanism results in heat transfer capabilities from
one hundred to several thousand times that of an equivalent
piece of copper.

As shown in Figure 1, the heat pipe in its simplest configuration is

a closed, evacuated cylindrical vessel with the internal walls lined
with a capillary structure or wick that is saturated with a working
fluid. Since the heat pipe is evacuated and then charged with the
working fluid prior to being sealed, the internal pressure is set by
the vapor pressure of the fluid. Figure 1: Heat pipe operation

48 Electronics COOLING | |SPRING 2020

FALL 2020
 Electronics COOLING | FALL 2020

Heat pipes can be designed to operate over a very broad range The most important heat pipe design consideration is the amount
of temperatures from cryogenic (< - 243°C) applications utilizing of power the heat pipe is capable of transferring. Heat pipes can
titanium alloy/nitrogen heat pipes, to high temperature applica- be designed to carry a few watts or several kilowatts, depending
tions (>2,000°C) using tungsten/silver heat pipes. In electronic on the application. Heat pipes can transfer much higher powers
cooling applications where it is desirable to maintain junction for a given temperature gradient than even the best metallic con-
temperatures below 125-150°C, copper/water heat pipes are typi- ductors. If driven beyond its capacity, however, the effective ther-
cally used. Copper/methanol heat pipes are used if the application mal conductivity of the heat pipe will be significantly reduced.
requires heat pipe operation below 0°C. Therefore, it is important to assure that the heat pipe is designed
to safely transport the required heat load.
HEAT PIPE DESIGN
There are many factors to consider when designing a heat pipe: The maximum heat transport capability of the heat pipe is gov-
compatibility of materials, operating temperature range, diameter, erned by several limiting factors which must be addressed when
power limitations, thermal resistances, and operating orientation. designing a heat pipe. There are five primary heat pipe heat trans-
However, the design issues are reduced to two major consider- port limitations. These heat transport limits, which are a function
ations by limiting the selection to copper/water heat pipes for cool- of the heat pipe operating temperature, include: viscous, sonic, cap-
ing electronics. These considerations are the amount of power the illary pumping, entrainment or flooding, and boiling. Figures 2 and
heat pipe is capable of carrying and its effective thermal resistance. 3 show graphs of the axial heat transport limits as a function of
These two major heat pipe design criteria are discussed below. operating temperature for typical powder metal and screen wicked
heat pipes. Each heat transport limitation is summarized in Table 1.

LIMITS TO HEAT TRANSPORT

Figure 2: Predicted heat pipe limitations

As shown in Figures 2 and 3, the capillary limit is usually the lim-

iting factor in a heat pipe design.

Table 1: Heat pipe heat transport limitations Figure 3: Predicted heat pipe limits

Electronics-COOLING.com 49
The capillary limit is set by the pumping capacity of the wick struc- and conduction losses back through the condenser section wick
ture. As shown in Figure 4, the capillary limit is a strong function and wall.
of the operating orientation and the type of wick structure.
Figure 6 shows a power versus ΔT curve for a typical copper/water
heat pipe.

Figure 5: Wick structures

Figure 4: Capillary limits vs. operating angle

The two most important properties of a wick are the pore radius
and the permeability. The pore radius determines the pumping
pressure the wick can develop. The permeability determines the
frictional losses of the fluid as it flows through the wick. There
are several types of wick structures available including: grooves,
screen, cables/fibers, and sintered powder metal. Figure 5 shows
several heat pipe wick structures.

It is important to select the proper wick structure for your appli-

cation. The above list is in order of decreasing permeability and
decreasing pore radius.

Grooved wicks have a large pore radius and a high permeability,

as a result the pressure losses are low but the pumping head is
also low. Grooved wicks can transfer high heat loads in a hori-
zontal or gravity aided position, but cannot transfer large loads
against gravity. The powder metal wicks on the opposite end of
the list have small pore radii and relatively low permeability. Pow-
der metal wicks are limited by pressure drops in the horizontal
position but can transfer large loads against gravity. Figure 6: Predicted heat pipe Delta-T

EFFECTIVE HEAT PIPE THERMAL RESISTANCE The detailed thermal analysis of heat pipes is rather complex.
The other primary heat pipe design consideration is the effective There are, however, a few rules of thumb that can be used for first
heat pipe thermal resistance or overall heat pipe ΔT at a given pass design considerations. A rough guide for a copper/water
design power. As the heat pipe is a two-phase heat transfer device, heat pipe with a powder metal wick structure is to use 0.2°C/W/
a constant effective thermal resistance value cannot be assigned. cm2 for thermal resistance at the evaporator and condenser, and
The effective thermal resistance is not constant but a function of a 0.02°C/W/cm2 for axial resistance.
large number of variables, such as heat pipe geometry, evaporator
length, condenser length, wick structure, and working fluid. The evaporator and condenser resistances are based on the outer
surface area of the heat pipe. The axial resistance is based on the
The total thermal resistance of a heat pipe is the sum of the resis- cross-sectional area of the vapor space. This design guide is only
tances due to conduction through the wall, conduction through useful for powers at or below the design power for the given
the wick, evaporation or boiling, axial vapor flow, condensation, heat pipe.

50 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

For example, to calculate the effective thermal resistance for a 1.27

cm diameter copper/water heat pipe 30.5 cm long with a 1 cm di-
ameter vapor space, the following assumptions are made. Assume
the heat pipe is dissipating 75 watts with a 5 cm evaporator and
a 5 cm condenser length. The evaporator heat flux (q) equals the
power divided by the heat input area (q = Q/Aevap; q= 3.8 W/cm2).
The axial heat flux equals the power divided by the cross sectional
area of the vapor space (q=Q/Avapor; q = 95.5 W/cm2).

The temperature gradient equals the heat flux times the thermal
resistance.

ΔT = qevap* Revap + qaxial * Raxial + qcond* Rcond

ΔT = 3.8 W/cm2 * 0.2°C/W/cm2 + 95.5 W/cm2 * 0.02°C/W/cm2 + 3.8 W/cm2 * 0.2°C/W/cm2

ΔT = 3.4°C

It is important to note that the equations given above for thermal

performance are only rule of thumb guidelines. These guidelines Figure 8: High-end CPU heat pipe heat sink
should only be used to help determine if heat pipes will meet your
cooling requirements, not as final design criteria. More detailed
information on power limitations and predicted heat pipe ther-
mal resistances are given in the heat pipe design books listed in
the reference section.

HEAT PIPE ELECTRONIC COOLING APPLICATIONS

Perhaps the best way to demonstrate the heat pipes application to
electronics cooling is to present a few of the more common ex-
amples. Currently, one of the highest volume applications for heat
pipes is cooling the Pentium processors in notebook computers.
Due to the limited space and power available in notebook comput-
ers, heat pipes are ideally suited for cooling the high-power chips.

Figure 9: High-power IGBT heat pipe heat sink

Fan assisted heat sinks require electrical power and reduce bat-
tery life. Standard metallic heat sinks capable of dissipating the
heat load are too large to be incorporated into the notebook pack-
age. Heat pipes, on the other hand, offer a high efficiency, passive,
compact heat transfer solution. Three- or four-millimeter diam-
eter heat pipes can effectively remove the high flux heat from the
processor. The heat pipe spreads the heat load over a relatively
large area heat sink, where the heat flux is so low that it can be ef-
fectively dissipated through the notebook case to ambient air. The
heat sink can be the existing components of the notebook, from
Figure 7: Typical notebook heat pipe heat sink Electromagnetic Interference (EMI) shielding under the key pad

Electronics-COOLING.com 51
to metal structural components. Various configurations of note- REFERENCES
book heat pipe heat sinks are shown in Figure 7. 1. Brennan, P.J. and Kroliczek, E.J., Heat Pipe Design Hand-
book, B&K Engineering, NASA Contract No. NAS5-23406,
In addition, other high-power electronics including silicon con- June 1979.
trolled rectifiers (SCRs), insulated gate bipolar transistors (IGBTs) 2. Chi, S.W., Heat Pipe Theory and Practice, Hemisphere Pub-
and Thyristors, often utilize heat pipe heat sinks. Heat pipe heat lishing Corporation, 1976.
sinks similar to the one shown in Figure 9, are capable of cooling 3. Dunn, P.D. and Reay, D.A., Heat Pipes, 3rd. Edition, Perma-
several devices with total heat loads up to 5 kW. These heat sinks gon Press,1982.
are also available in an electrically isolated versions where the fin 4. Eastman, G. Yale and Ernst D.M., Heat Transfer Technology
stack can be at ground potential with the evaporator operating at (Heat Pipe), Kirk-Othmer: Encyclopedia of Chemical Tech-
the device potentials of up to 10 kV. Typical thermal resistances nology, Volume 12, John Wiley and Sons, Inc., 1980.
for the high-power heat sinks range from 0.05 to 0.1°C/W. Again, 5. Peterson, G.P., An Introduction to Heat Pipes Modeling,
the resistance is predominately controlled by the available fin vol- Testing, and Applications, John Wiley and Sons, Inc., 1994.
ume and air flow.

52 Electronics COOLING | FALL 2020

 From the Creators of Interference Technology

I N T R O D U C I N G

THERE’S ONLY ONE SPOTLIGHT

ARE YOU IN IT?

Positioning Ambitious B2B Electronics Companies
for Aggressive Growth

www.lectrixgroup.com

PHILADELPHIA, PA | HONG KONG

F E AT U R E D

Advanced Cooling for Power Electronics

Reprinted from July 2017

Sukhvinder S. Kang
*Aavid
*Note: Affiliation as cited in the original article

EDITOR'S COMMENTS:
This special edition spans 25 years of Electronics Cooling. Power electronics could not be omitted from the list. Over the years,
power electronics became a hot topic (e.g. MOSFET, IGBT, data center, EV/HEV, etc.). This market is estimated at more than $10
billion with a compound annual growth rate of +4%. A diversity of cooling solutions can be used, but in the end, the final choice
depends strongly on the application requirements. (Genevieve Martin)

Power electronics devices such as MOSFETs, GTOs, IGBTs, trends in power semiconductors and devices is reviewed in [2, 3].
IGCTs etc. are now widely used to efficiently deliver electrical Silicon remains the workhorse material for power semiconduc-
power in home electronics, industrial drives, telecommunica- tors and to avoid device failure due to thermal runaway, effec-
tion, transport, electric grid, and numerous other applications. tive cooling is critical. Figure 1 shows the maximum safe junction
This paper discusses cooling technologies that have evolved in temperatures for silicon devices [2]. Wide-bandgap semiconduc-
step to remove increasing levels of heat dissipation and manage tors like SiC and GaN offer the advantage of high temperature
junction temperatures to achieve goals for efficiency, cost, and operation. However, available packaging technologies, passives
reliability. Cooling technologies rely on heat spreading and con- and peripheral components, solder materials, reliability consid-
vection. In applications that use natural or forced air cooling, wa- erations, and cost presently limit the junction temperatures to
ter heat pipes provide efficient heat spreading for size and weight ~175°C even though the semiconductor device can, in principle,
reduction. Previous concepts are reviewed and an improved heat operate at much higher junction temperatures [4]. The maximum
sink concept with staggered fin density is described for more iso- safe junction temperatures in SiC could exceed 300°C [5] so that
thermal cooling. Where gravity can drive liquid flow, thermo- even in high ambient temperatures, sufficient cooling may be pro-
siphons provide efficient heat transport to remote fin volumes vided by smaller and lower cost heat sinks resulting in improved
that can be oriented for natural and/or forced air cooling. Liquid volumetric power density.
cold plates (LCPs) offer the means to cool high heat loads and
heat fluxes including double sided cooling for the highest den-
sity packaging. LCPs can be used both in single phase cooling
systems with aqueous or oil-based coolants and in two-phase
cooling systems with dielectric fluids and refrigerants. Previous
concepts are reviewed and new concepts including an air-cooled
heat sink, a thermosiphon heat sink, a vortex flow LCP, and a
shear flow direct contact cooling concept are described.

P
INTRODUCTION
ower electronics devices and systems are vital in the efficient
generation, transmission and distribution, conversion, and
a huge variety of end uses of electric power. More and more
applications are adopting power electronics technologies to im-
prove energy efficiency, reliability, and control and it is anticipated
that in the future all electrical power will flow through a power Figure 1: Critical thermal runaway temperature and estimated maximum safe oper-
semiconductor device at least once [1]. The state of art and future ating temperature of Silicon devices [2]

54 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

A number of thermal management solutions in use for cooling for “double-sided” cooling. Both these modules are designed to be
power electronic modules in automotive applications are re- mechanically bolted to heat sinks or liquid cold plates.
viewed in [6, 7]. The coolant in such applications is available at
temperatures above 100°C so cooling must be accomplished with
a low temperature difference between the semiconductor and the
coolant. Many highly integrated cooling solutions are presented
in [6] focused on the thermal management challenges in this se-
vere application. Similar cooling concepts can be applied also to
other applications like industrial drives, wind power converters,
HVDC power transmission, etc. depending on the constraints
of each application. Various packaging designs and cooling solu- Figure 2: IGBT packages for single- and double-sided cooling [2]
tions for reducing the thermal resistance of high-power modules
are described and compared in [8]. With micro-channel liquid Interest in hybrid car and railway applications in recent years have
cold plates [9, 10] the cold plate thermal resistance can be reduced led to customized packaging solutions with double sided cooling
to such an extent that the internal thermal resistance of the elec- from Alstom [12], Denso [13] and others. These modules are
tronics package becomes the dominant thermal resistance. Phase shown in Figure 3. A number of other packaging concepts and
change cooling using forced convection boiling of refrigerant flu- related cooling solutions have previously been discussed in [6].
ids in cold plates is described in [11] including the more isother-
mal operation that comes from using the latent heat rather than
the sensible heat of the coolant to provide the cooling effect.

Cooling solutions in use today include:

• Natural and forced convection air-cooled heat sinks

• Single and double sided cooling with liquid cold plates
• Micro-channel liquid coolers built into power module base
plates or integrated with the DBC substrate
• Jet impingement and direct contact liquid cooling of module Figure 3: Custom power module designs for double sides cooling
base plates or DBC substrates
• Two-phase liquid cold plates with boiling of dielectric refrig- Figure 4 shows a schematic of a standard power semiconductor
erant coolant package mounted on a heat sink for cooling and the corresponding
thermal resistance network. The key parameter for device cooling
With the exception of underwater or marine applications, heat performance is the junction-to-air thermal resistance, Rja. This is
removed from electronics systems is ultimately dissipated to the the sum of the internal thermal resistance within the device pack-
air. Systems with air-cooled heat sinks dissipate heat directly into age, Rjc, the interface thermal resistance from the device to the heat
the air that flows through the system either by natural or forced sink, Rcs, and the heat sink to air thermal resistance, Rsa.
convection. Systems that use liquid cooling transfer heat from the
electronics into the liquid that in turn transfers the heat to air in
a liquid-to-air heat exchanger such as an automotive radiator. In
two-phase systems, heat from the power electronics transforms
liquid into vapor by boiling or evaporation and the vapor is then
returned to liquid phase when it transfers heat to the air in an air-
cooled condenser.

This paper reviews some of the existing cooling solutions and

presents new concepts for an air-cooled heat sink, loop thermo-
siphon heat sink, a liquid cold plate, and a direct liquid cooling
concept. Approximate performance levels are also presented.

POWER MODULE PACKAGES Figure 4: Schematic of standard IGBT package on a heat sink and key thermal re-
Figure 2 shows two common packages used for power modules. sistances
The IGBT module on the left is commonly used in applications
below 6.5 kV and exposes one flat surface for “single-sided” cool- The heat sink thermal resistance can be expressed through the
ing. The other module is well suited for stacking in series for high following equation commonly used in heat exchanger design lit-
voltage applications and provides both top and bottom surfaces erature [14].

Electronics-COOLING.com 55
 spreading. For the same fin configuration as Figure 5, a copper
base would enable an ~40% reduction in the spreading resis-
tance in Table 1 from 12°C/kW to 7°C/kW or a 10°C reduction
in junction temperature.

An examination of Equation (1) shows that the heat sink ther- Table 1 shows that the solder layer, copper base plate, and paste
mal resistance can be improved by increasing the mass flow rate comprise 50% of the thermal resistance not associated with the
m. of the fluid through the heat sink (e.g. forced convection rath- heat sink. This is even more significant in liquid cooling where
er than natural convection), the heat capacity cp of the fluid (e.g. the liquid cold plate resistance is ~0.1 to 0.2X the air-cooled heat
a liquid versus a gas), the heat transfer coefficient h on the heat sink.
sink surface (e.g. smaller channel dimensions like micro-chan-
nels, turbulent flow using turbulators and boundary layer inter-
ruption rather than laminar flow, two-phase mechanisms such
as evaporation and condensation rather than single phase con-
vection) and the effective heat transfer area A. The effective area
is the actual heat transfer area in contact with the cooling fluid
multiplied by the heat transfer efficiency of the surface (e.g. fin
efficiency). Area enhancement by using closely spaced fins and
small scale flow channels (e.g. micro-channels) is the focus of
many new developments in heat sink manufacturing technol-
ogies.

The term in the parenthesis in Equation (1) is the effectiveness of

the heat sink and is defined as follows

From Equation (1) we can see that improving the efficiency of

the heat transfer surface (making it more isothermal) will im-
prove the effective area A and therefore the effectiveness E of the
heat sink. This is also evident in Equation (2) by noting that a Table 1: Approximate thermal resistances in IGBT air-cooled assembly shown in
more isothermal surface implies a smaller value of Ts in the de- Figure 5.
nominator. The surface efficiency includes both the heat spread-
ing resistance in the base and the fin efficiency of the fins.

FORCED AIR-COOLED HEAT SINKS

Table 1 shows an example of thermal resistances in the path
from the IGBT dies to the air for the aluminum forced air-
cooled heat sink illustrated in Figure 5. The heat sink base is 236
x 230 mm in size and cools one econopack IGBT module (162 x
122 mm) with a DBC substrate soldered to a 3 mm thick copper
heat spreading base. Aluminum fins are pressed into grooves in
the aluminum base. The IGBT module is mounted on the heat
sink using thermal grease to reduce thermal interface resistance.
Air flow rate through the heat sink is 0.25 m3/s at 25°C and 1 Bar
and the IGBT dissipates a total of 2 kW of heat.

Data in Table 1 shows that almost 30% of the thermal resistance

is due to inefficient heat spreading in the base. With guidance
from Equations (1) and (2), the base can be made more isother-
mal to reduce the heat sink thermal resistance. Figure 6 shows Figure 5: Illustration of IGBT module attached to an aluminum press-fin forced air-
an example of a heat sink with a copper base to improve heat cooled heat sink

56 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

increasing in the flow direction as shown in Figure 8. The fin sur-

face area is lowest in the upstream region where the cold air enters
the heat sink and increases in a stepped fashion towards the down-
stream region as the air heats up. In this way, the base of the heat
sink can be kept more isothermal so that power modules mounted
at different distances from the edge where the air enters can be
cooled to similar temperatures. Furthermore, the total pressure
drop through the heat sink and the total mass is reduced because
of the lower fin density in the upstream region. This type of design
is manufactured by bonding or brazing the fins to the base.

Figure 6: Air-cooled heat sink with copper base

Figure 7 shows an example of a heat sink with heat pipes embed-

ded in the aluminum base. The copper water heat pipes use copper
powder wick and are embedded in the base using Aavid’s hi-con-
tact technology. The heat pipes are oriented along the fin direction
to even out the front to back temperature gradient due to heat up
of the air. An ~60% reduction in the spreading resistance to 5°C/ Figure 8: Advanced heat sink with increasing fin density along air flow direction
kW or a 14°C reduction in junction temperature can be achieved.
The better heat spreading with heat pipes is achieved at a smaller LOOP THERMOSIPHON AIR-COOLED HEAT SINKS
heat sink mass compared to using a copper base. Recent advances The heat sink concepts discussed thus far position the heat sink to
using carbon nanotubes grown on copper mesh and powder wick air heat transfer surface area in the physical space adjacent to the
enable heat fluxes in excess of 500 W/cm2 [15] and reduction in power module. This can limit the electrical design and packaging
thermal resistance. Such nanotechnology is expected to be avail- flexibility of the power electronics system. Heat pipes have been
able in commercial products based on market demands. used to extend the heat sink fins some distance from the heat sink
base but when gravity can be used advantageously, two phase ther-
mosiphons can provide a higher performance solution. The sim-
plest thermosiphons are just gravity aided heat pipes with a groove
type wick surface. In this type of design, the vapor flow in the tube
is in the opposite direction to the liquid flow along the wall of the
tube. This limits the maximum power that can be carried because
of liquid entrainment in the high velocity vapor flow [16]. A loop
thermosiphon avoids this limitation by transporting vapor in one
tube and returning the condensed liquid through another tube.

Figure 9 shows an air-cooled heat sink that uses a loop thermo-

siphon to transport heat away from the immediate space around
the power electronics package. An evaporator is provided with a
flat surface where the power module is mounted and an enhanced
Figure 7: Air-cooled heat sink with heat pipes embedded in an aluminum base boiling surface on the inside to enhance heat flux capability and
reduce thermal resistance. The air-cooled condenser is similar to an
A further advance in heat sink performance is achieved if the automotive radiator with vapor flow inside extruded flat aluminum
press-fin joint is replaced by a metallurgical joint made by sol- tubes and aluminum folded fins outside the tubes to provide a large
dering or brazing the fins to the base. This decreases the thermal cooling surface area on the air side. The liquid in the evaporator of
resistance by ~4°C/kW enabling an additional ~8°C reduction in the thermosiphon absorbs heat from the electronics module and
junction temperature. Both the heat sinks pictured in Figures 6 changes into its vapor phase. Vapor travels through the vapor tube
and 7 incorporate such an improved base to fin joint. to the condenser where it rejects heat to the air and condenses back
into liquid phase. The liquid condensed then returns to the evapo-
A recent innovation consists of a heat sink design with fin density rator through the liquid return tube to complete the cycle.

Electronics-COOLING.com 57
 The main limitation of the two-phase loop thermosiphon is that
it is not suitable for moving platforms such as automobiles where
external body forces other than gravity could potentially move the
liquid out of the evaporator and cause dryout.

Since the working fluid inside the thermosiphon is dielectric, in

principle the power semiconductor can be immersed inside the
fluid in the thermosiphon. Figure 11 shows a prototype thermo-
siphon heat sink based on this type of “immersion cooling” con-
cept [17]. Periodic two phase thermosiphons can transport heat
against gravity using vapor pressure to drive the flow [18].

Figure 9: Air-cooled heat sink using loop thermosiphon

Thermal resistance of the thermosiphon heat sink can be deter-

mined by using the heat transfer coefficient of the boiling sur-
face inside the evaporator per Figure 10 and a condensation heat
transfer coefficient of ~1 kW/m2-K inside the condenser tubing.
Assuming a 122 x 162 mm IGBT dissipating 2 kW of heat, the
heat flux on the evaporator wall is ~10 W/cm2 and the boiling
heat transfer coefficient is ~5.5 W/cm2-K. Including a 3 mm
thick copper wall thickness, the evaporator thermal resistance is
1.3°C/kW. Assuming a 230 x 230 mm by 30 mm thick air-cooled
condenser with 0.25 m3/s air flow rate (same as in Table 1), the
estimated condenser thermal resistance is 9.1°C/kW so that the
total thermal resistance Rsa is 10.4°C/kW, about a third of the
31°C/kW value for the press fin heat sink studied in Table 1.
This huge improvement in thermal performance comes with the
added benefits of lower mass, lower pressure drop, and an essen- Figure 11: Thermosiphon air-cooled system with power electronics immersed in di-
tially isothermal heat sink base for good static current balance electric fluid. IGBT is clamped between enhanced boiling plates.
between parallel IGBT dies.
LIQUID COOLED COLD PLATES
In liquid cooled systems, liquid cold plates provide localized cool-
ing of power electronics by transferring heat to liquid that then
flows to a remote heat exchanger and dissipates the heat to air
or to another liquid in a secondary cooling system. Liquid cold
plates provide more efficient cooling and enable greater levels
of integration and major reductions in the volume and weight
of power electronics systems. Existing and emerging liquid cold
plate solutions in 2008 were extensively reviewed in [6] including:

• Tube type and fin type liquid cold plates to cool packages
with DBC substrates with and without copper base plates.
• Liquid flow through fins formed directly on copper and Al-
SiC base plates.
• Single and double sided cooling using liquid jet impinge-
ment.
• Direct cooling of the base plate or DBC substrate using con-
cepts such as the Danfoss “shower power” design using jet
Figure 10: Heat transfer coefficient on thermosiphon evaporator boiling surface impingement or liquid flow through meandering channels.

58 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

• Direct double sided cooling of back to back modules with In the example considered to illustrate performance, an IGBT dis-
and without fins integrated with DBC substrates. sipating 1,600 W is attached to each side of the VLCP with thermal
• Micro-channel coolers built into base plates or integrated grease at the interface. The IGBT base plate size is 140 x 190 mm
with DBC substrate or into specially customized package and uniform heat flux is assumed on the VLCP over the IGBT con-
designs. tact area. Cooling is provided by a water-glycol coolant with 50%
• Stacked power modules and liquid cold plates using extrud- glycol by volume flowing at 11 liters per minute (LPM) at a tem-
ed channel type, folded fin type, and micro-channel type perature of 80°C at the cold plate inlet. Under these conditions, the
cold plates. maximum temperature on the cold plate surface is 91.5°C, yielding
a thermal resistance of Rsf = 3.6°C/kW. Using Equation (2), the ef-
The paper [6] showed the great potential for achieving very high fectiveness is 47%. By comparison, the thermal resistances of suit-
levels of cooling performance and system integration using liquid able tube type and offset folded fin type liquid cold plates for the
cold plates. same application conditions were 19°C/kW and 5.6°C/ kW respec-
tively. The pressure drop of the VLCP was 52 kPa versus 10 kPa for
As noted earlier, Equation (1) guides us towards higher heat trans- the tube type and offset folded fin type designs.
fer coefficients using turbulent flow or small channel dimensions
such as micro-channels to reduce thermal resistance. Two recent-
ly developed concepts that apply these principles effectively are
presented next.

In many applications where liquid filtration is not desirable, flow

channels are required to be of a sufficiently large size (~2-3 mm) to
avoid clogging by particles and precipitates in the cooling liquid.
High liquid velocities and turbulator inserts are commonly used
to enhance the heat transfer coefficient in such large size channels
including in tube type liquid cold plates. This type of approach
provides limited opportunity for heat transfer area enhancement.
An alternative approach developed recently [19] uses helicoidal
flow paths to create strong secondary flow with high vorticity to
achieve high heat transfer coefficients at the flow channel wall and Figure 13: Stack of press pack diodes cooled using Vortex Liquid Cold Plates
parallel flow paths to reduce liquid velocities and the correspond-
ing pressure gradient. Area enhancement is achieved by orienting
the helical path so that its axis is normal to the base. This design
concept, named the “vortex liquid cold plate” or VLCP, is shown
in Figure 12 in the double sided cooling version. This design is
ideal for cooling a stack of press pack IGBT’s, thyristors or diodes
in series as shown in Figure 13.

Figure 14: Liquid cooling of power electronics package using high shear direct con-
tact concept

For liquid cooling systems where liquid filtration is possible, high

heat transfer coefficients may be obtained using high shear flow
in narrow channels. Figure 14 shows a high shear direct contact
(HSDC) cooling approach using a cooling plate that enables direct
contact between the liquid and the surface to be cooled. An elas-
tomeric O-ring (not shown) is provided between the cooling plate
and the electronic package to seal the liquid flow volume. Liquid is
distributed through a highly parallel manifold system and made to
flow at a low velocity through narrow flow channels in direct con-
tact with the surface to be cooled as shown in Figure 15. Ribs (not
shown) are provided on the surface of the cooling plate that contact
Figure 12: Vortex Liquid Cold Plate in a double sides cooling version the cooled surface and set the height of the narrow flow channels.

Electronics-COOLING.com 59
The combination of low liquid velocity and small channel height whole cooling plate so that temperature gradients over the pow-
insures laminar flow in the channels. By adjusting the spacing be- er electronics package are considerably reduced. In its simplest
tween the liquid supply and return channels in the manifold, the implementation, there is no area enhancement on the base plate.
flow length of the cooling channels and liquid velocity can be set to However, the design can work well with enhanced surfaces as
achieve the desired pressure drop characteristics. well. Figure 16 shows the heat transfer coefficient on a flat base
plate depending on flow channel height assuming fully developed
laminar flow in the channels. The actual heat transfer coefficient
will be higher because of entrance length effects.

We will use the same application example as for the VLCP to il-
lustrate the performance of the HSDC liquid cooling concept in a
double-sided cooling configuration. The liquid contact area with
the IGBT base plates on each side is 140 x190 mm and uniform
heat flux is assumed over this area. Cooling is provided by a wa-
ter-glycol coolant with 50% glycol by volume flowing at 11 LPM at
a temperature of 80°C at the cooling plate inlet. We assume a flow
channel height of 0.15 mm so that the local heat transfer coeffi-
cient is ~9,800 W/m2K. From Equation (1), the thermal resistance
is ~2.8°C/kW so the maximum temperature rise on the cooled
surface relative to the liquid inlet temperature will be ~9°C. As-
suming each flow channel is 10 mm long, average liquid velocity
in the channels will be ~0.5 m/s and the pressure drop will be ~3-5
kPa. This level of performance is very competitive with the other
Figure 15: High Shear flow cooling channels and flow pattern in the HSDC liquid cold plate designs mentioned earlier but this concept does require
cooling plate coolant filtration to avoid clogging the very narrow channels. By
reducing the flow channel height, the thermal resistance can be
improved even further but at the expense of higher pressure drop.

CONCLUSIONS
Significant advances have been made in cooling technology for
power electronics. This paper has discussed improvements in
applications ranging from air-cooling to liquid cooling. A sim-
ple equation was described to helps guide design choices. Copper
bases and heat pipes embedded in bases can significantly improve
heat spreading in air-cooled heat sinks. Even more dramatic gains
can result from using loop thermosiphons to more efficiently pick
up heat over the full base area in contact with the power module
and passively transport the heat for dissipation in an air-cooled
condenser. New liquid cold plate concepts are discussed that are
well suited for single side cooled packages as well as double sided
cooling of stacked press-pack type modules. Performance esti-
mates are provided through application examples.

Figure 16: Local heat transfer coefficient on power module base plate at 80°C cool- Emerging nanotechnology is mentioned that has the potential to
ant temperature significantly improve thermal performance of evaporative and
boiling surfaces in future products. Although reliability of some
It is obvious that the HSDC cooling plate may be designed to technologies need to be proven, it seems that cooling technology
provide double sided cooling so that power electronics devices will keep up nicely with increasing power dissipation levels and
may contact the cooling plate on both the top and bottom surface. compactness of power electronics.
Furthermore, depending on the requirements of the application,
the cooling plate may be made from non-conducting plastic or ACKNOWLEDGMENTS
electrically conducting metal. The author would like to acknowledge the contribution of his col-
leagues Cesare Capriz, Jeff Kimball, and Marco Moruzzi for tech-
An attractive feature of the HSDC design is that the incoming nical discussions and for preparing many of the illustrations and
cold liquid is supplied uniformly to cooling channels over the pictures included in this paper.

60 Electronics COOLING | FALL 2020

 Electronics COOLING | FALL 2020

12. Mermet-Guyennet, M., 2006, “Double-sided liquid cooling,”

Nomenclature
High Temperature Electronics and Ther-mal Management,
Q Heat Load (W) ECPE Seminar, Nuremberg, Germany, Nov. 2006.
13. Denso News, 2007, “DENSO Develops High Out-put
Rja Junction to air inlet thermal resistance [= (T j- Ta)/Q (°C/W)]
Power Control Unit and Battery Cooling System for Hy-
Rsa Heat sink contact surface to air inlet thermal resistance [=(Ts-Ta)/Q (°C/W)] brid Vehicles”, May 2007, http://www.denso-ts.com/news/
D0705-DENSO_Develops_PCU_eng.pdf
Rsf Heat sink contact surface to fluid inlet thermal resistance [=(Ts-Tf)/Q (°C/W)]
14. Kays, W.M., London, A.L., Compact Heat Exchangers, Third
Rcs Thermal interface resistance [=(Tc-Ts)/Q (°C/W)] Edition, McGraw-Hill, Inc., New York, 1984.
15. Kim, S., Weibel, J. A., Fisher, T. S. and Garimella, S. V., 2010,
T Temperature (°C) “Thermal Performance of Carbon Nanotube Enhanced Va-
por Chamber Wicks,” 14th International Heat Transfer Con-
REFERENCES ference, Washington, D.C., USA, August 2010, Paper no.
1. EPE and ECPE “Position paper on Energy Efficiency – the IHTC14-22929, pp. 417-424.
role of Power Electronics,” Summary of results from Europe- 16. Dunn, P.D., Raey, D.A., Heat Pipes, Fourth Edition, Pergam-
an Workshop on Energy Efficiency – the role of Power Elec- on Press, Copyright 1994 Elsevier Science Ltd.
tronics, Brussels, Belgium, Feb 2007 17. Barnes, C., 2008, “Immersion Cooling of Power Electronics
2. Benda, V., “Power Semiconductors – State of the Art and in Segregated Hydrofluoroether Liquids,” ASME Heat Trans-
Future Trends,” Transactions on Machine Power Electronics fer Conference 2008, Jacksonville Florida, USA, August 2008
and Drives, Global Journal of Technology and Optimization, 18. Filippeschi, S., 2006, “On Periodic Two-phase Thermosy-
Vol. 2, 2011, Online Publication, http://www.engedu2.net/ phons Operating Against Gravity,” Inter-national Journal of
V2/PE-G11.pdf Thermal Sciences, 45 (2), pp. 124-137.
3. Vobecky, J., 2010, “Future Trends in High Power Devices,” 19. Aavid Vortex Liquid Cold Plate, Patent Pending, 2011.
Microelectronics Proceedings (MIEL), 2010 27th Interna-
tional Conference on, pp. 67-72, May 2010.
4. Biela, J., Schweizer, M., Waffler, S., Wrzecionko, B. and Ko-
lar, J.W., 2010, “SiC vs. Si Evaluation of Potentials for Per-
formance Improvement of Power Electronics Converter Sys-
tems by SiC Power Semiconductors,” 2010, Materials Science
Forum, Vols. 645648, pp. pp. 1101-1106.
5. Wrzecionko, B., Biela, J., and Kolar, J.W., 2009, “SiC power
semiconductors in HEVs: Influence of junction tempera-
ture on power density, chip utilization and efficiency,” in
35th Conference of the IEEE Industrial Electronics Society
(IECON). Porto, Portugal, Nov. 2009.
6. Schulz-Harder, J., 2008, “Review of Highly Integrated Solu-
tions for Power Electronic Devices,” Integrated Power Sys-
tems (CIPS), 2008 5th International Conference on, Nurem-
berg, Germany, March 2008
7. Saums, D., 2011, “Vehicle Electrification Thermal Manage-
ment Challenges and Solutions Overview”, MEPTEC Ther-
mal Management Workshop, San Jose CA USA, March 2011.
8. Leslie, S. G., 2006, “Cooling Options and Challenges of High
Power Semiconductors,” Electronics Cooling, V. 12, No. 4,
Nov 2006, pp. 20-27.
9. Stevanovic, 2006 Solovitz, S.A., Stevanovic, L.D. and Beaupre,
R.A., 2006, “Microchannels Take Heatsinks to the Next Lev-
el,” Power Electronics Technology, November 2006, pp.14-20.
10. Valenzuela, J., Jasinkski, T, and Sheikh, Z., 2005, “Liquid
Cooling for High-Power Electronics,” Power Electronics
Technology, February 2005, pp. 50-56.
11. Levettand, D. B., Howes, J. C. and Saums, D. L., 2008, “Cooling
of IGBT Modules with Vaporizable Dielectric Fluid (VDF),”
IMAPS France Advanced Technology Workshop on Thermal
Management 2008, La Rochelle, France, January 2008.

Electronics-COOLING.com 61
Index of
ADVERTISERS

Alpha Novatech, Inc.

473 Sapena Ct. #12,
MADE FOR Delta
CPC Worldwide
1001 Westgate Drive
Electronics Ltd.
46101 Fremont Blvd.
LECTRIX
1000 Germantown Pike,
Santa Clara, CA 95054
STAYING POWER.
St. Paul, MN 55114 Fremont, CA 94538 U.S.A. Plymouth Meeting, PA 19462

t: +1 (408) 567-8082 t: (651)applications. t: (866) 407-4278

Quality quick disconnects are critical in liquid cooling
645-0091 t: (484) 688-0300
That’s why CPC uses patent-pending
e: sales@alphanovatech.com w: www.cpcworldwide.com e: dcfansales.us@deltaww.com
valve technology – so the disconnects withstand e: info@lectrixgroup.com
w: www.alphanovatech.com page:long-term
41 use without drips. Regardless of w:
the www.delta-fan.com w: www.lectrixgroup.com
configuration, CPC fittings are designed to page: page: 53
page: Back Cover meet 7
the needs of your applications.

Learn more about what we’re made for at

cpcworldwide.com/LQ

©2020 Colder Products Company

Malico Inc. Master Bond, Inc Mentor Graphics SEMI-THERM 37

No. 5, Ming Lung Road, 154 Hobart Street 8005 SW Boeckman Road Double Tree by Hilton
Yangmei 32663, Taiwan Hackensack, NJ 07601 Wilsonville, OR 97070 San Jone, CA USA
March 22 – 26, 2021
t: 886-3-4728155 t: +1 (201) 343-8983 t: (800) 592-2210
e: inquiry@malico.com e: main@masterbond.com e: sales_info@mentor.com t: +1 (408) 840-2354
w: www.malico.com w: www.masterbond.com w: www.mentor.com e: mvalle@infinera.com
page: Inside Front Cover page: 61 page: 33 w: www.semi-therm.org
page: 47

THERMAL LIVETM 2020

Online Event
October 20-21, 2020

t: (484) 688-0300
e: info@electronics-cooling.com
w: www.thermal.live
page: Inside Back Cover

62 Electronics COOLING | FALL 2020