SPRING 2021

electronics-cooling.com

FEATURED IN THIS EDITION

14 WHY CHEMICAL COMPATIBILITY

IS VITAL TO YOUR LIQUID
COOLING SYSTEM

20 BETTER COOLING BY REMOVING

MATERIAL, INSTEAD OF ADDING 6 CALCULATION CORNER
MATERIAL, TO UNLOCK THE FULL FAN COOLING PART 1: DETERMINING
POWER OF GaN ELECTRONICS FLOW RATE

24 THERMAL PACKAGING – 8 STATISTICS CORNER

FROM PROBLEM SOLVER TO CONFIDENCE INTERVALS
PERFORMANCE MULTIPLIER
© Copyright 2021 Electronics Cooling
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$) Location C-bore C-bore Base Width Base Length
Quantity Unit price (US$)
Diameter Effective
Length/ Fin Removal Area from Center Component side FIn side
1 122.73
ID THRU 115 The cut section interferes with fins. Suggested size 65.3mm or 68.8 mm
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
Genevieve Martin 1000 Germantown Pike, F-2
Plymouth Meeting, PA 19462 USA

4 COOLING EVENTS
Phone: +1 484-688-0300; Fax:+1 484-688-0303
info@lectrixgroup.com
News of Upcoming 2021 Thermal Management Events www.lectrixgroup.com

CHIEF EXECUTIVE OFFICER

6 CALCULATION CORNER Graham Kilshaw | Graham@lectrixgroup.com

Fan Cooling - Part 1: Determining Flow Rate VP OF MARKETING

Geoffrey Forman | Geoff@lectrixgroup.com
Ross Wilcoxon and Genevieve Martin
EDITORIAL DIRECTOR
10 STATISTICS CORNER James Marengo | James@lectrixgroup.com

Confidence Intervals CREATIVE DIRECTOR

Chris Bower | Chris@lectrixgroup.com
Ross Wilcoxon
BUSINESS DEVELOPMENT DIRECTOR
14 WHY CHEMICAL COMPATIBILITY IS VITAL TO Janet Ward | Jan@lectrixgroup.com

YOUR LIQUID COOLING SYSTEM PRODUCTION COORDINATOR

Jessica Stewart | Jessica@lectrixgroup.com
Elizabeth Langer
SENIOR GRAPHIC DESIGNER
19 TECHNICAL EDITORS SPOTLIGHT Kate Teti | Kate@lectrixgroup.com

CONTENT MARKETING MANAGER

Danielle Cantor | Danielle@lectrixgroup.com
20 BETTER COOLING BY REMOVING MATERIAL,
INSTEAD OF ADDING MATERIAL, TO UNLOCK ADMINISTRATIVE MANAGER
Eileen Ambler | Eileen@lectrixgroup.com
THE FULL POWER OF GaN ELECTRONICS
ACCOUNTING ASSISTANT
Remco van Erp
Susan Kavetski | Susan@lectrixgroup.com

24 THERMAL PACKAGING – FROM PROBLEM EDITORIAL BOARD

SOLVER TO PERFORMANCE MULTIPLIER
Ross Wilcoxon, Ph.D.
Avram Bar-Cohen Associate Director
Collins Aerospace

26 IN MEMORIAM OF AVRAM BAR-COHEN

ross.wilcoxon@collins.com

Genevieve Martin
32 INDEX OF ADVERTISERS R&D Manager, Thermal & Mechanics Competence
Signify
genevieve.martin@signify.com

Victor Chiriac, PhD, ASME Fellow

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

►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
For subscription changes email
retrieval system of any nature, without the prior written permission of the publishers (except in accordance with the Copyright Designs and Patents Act 1988). info@electronics-cooling.com
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 Reprints are available on a custom basis at
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. 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
Please call +1 484-688-0300.
in conjunction with the ElectronicsCooling trademark.

F R E E S U B S C R I P T I O N S
Lectrix®, Electronics Cooling®—The 2021 Spring 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
Genevieve Martin
Associate Technical Editor

An letter to our community and devoted readers: How can our community contribute in this pandemic time?

2020 has been a long and challenging year, both from professional and personal perspectives. We all know that
Covid is still around and 2021 will probably remain difficult.

The technical editorial board wishes each of you and your beloved ones, a happy, healthy, and prosperous new
year 2021. We are committing ourselves this year again to provide high quality issues that provide knowledge and
insight. This issue of Electronics Cooling magazine includes articles on Innovative cooling solution for GaN-
based components, discuss chemical material compatibility for your liquid cooling system, provide guidelines on
how to determine flow rate of a fan based system, continue our calculation corner with statistic column, and last
but not least deliver a tribute to Avi Bar-Cohen who left us October 10th, 2020 at the age of 74.

Let’s start the year with a positive touch. As technology enthusiasts, we are creative enough to find solutions to problems. Worldwide, initiatives have
emerged to help counter the virus; both by being innovative to react fast in the face of emergency and providing new product designs for healthcare
workers and by supporting the development of new medical treatments and vaccines.

Recent pandemic shows once more how collaboration, multi-disciplinary approach is key to success. In this editorial, I would like to stay close to
our community, which has a deep understanding of computer fluid dynamics simulations. Let’s see how our expertise is helping solve the problems
by working closely with other medical and lighting expertises.

Several studies are indicating that airborne transmission is a significant factor in the spread of the SARS-CoV-2 virus and of other viruses that
cause diseases like SARS (severe acute respiratory syndrome), MERS (middle east respiratory syndrome), and influenza [1][2][3]. Natural air flow
resulting from movement, temperature changes and recirculating air-conditioning in indoor spaces contributes to the rapid spread of viruses such
as SARS-CoV-2. This is an obvious challenge in battling the virus, as air cannot be easily contained; however, risks can be mitigated by applying ultra-
violet C (UV-C) light to reduce the virus concentration in the air while at the same time preventing human exposure to UV-C irradiation. Indeed,
both in-duct and upper-air disinfection systems leverage air flow models to provide the right UV-C intensities to achieve effective disinfection.

Recently, both the WHO and CDC recommended [4][5] the use of upper-room ultraviolet germicidal irradiation (UVGI) systems as a supplemental
air-cleaning measure to reduce the transmission of airborne bacterial and viral infections in public buildings, hospitals, military housings, and classrooms.

UV-C is an established measure for disinfection. UV-C light is a category of ultraviolet light with wavelengths between 100-280 nanometres (nm)
and is the most effective UV light for disinfection. It has been applied ever since it was discovered to be an effective tool in preventing the spread of
contagious diseases by disinfecting water, surfaces and air in very short time periods. UV-C light inactivates viruses and microorganisms such as
bacteria, moulds, spores, fungi and yeasts by destroying their DNA or RNA. It is generated by well-known lamp manufacturing technology and it
is sustainable and more environmentally friendly than several other disinfection methods.

Beginning in the mid-1800’s, researchers realized that microorganisms responded to light, leading to the use of ultraviolet light as a germicide.
Significant research over the past 150 years has led to numerous applications of UV-C light to disinfect air to reduce the spread of viruses.

The pandemic of 2020 has accelerated how our community has worked to reduce the spread of Covid through the combination of detailed
CFD simulations, advanced optical systems and knowledge of viruses. A future issue of Electronics Cooling Magazine will include an article
that describes the history of using light as a disinfectant, recent developments targeted at addressing the Covid-19 pandemic, and the technologies,
which are enabled by advances in electronics cooling, that are used to generate the disinfecting light. Stay tuned!

As always, I encourage you to contact us if you have special wishes or want to publish yourself an article in one of the upcoming issues.

– Genevieve Martin

[1] E.A.Nardell, R.R. Nathavitharana, “Airborne Spread of SARS-CoV-2 and a Potential Role for Air Disinfection”, JAMA. 2020;324(2):141-142
[2] L.Marr, S.Miller, K.Prather, C.Haas, W.Bahnfleth, R.Corsi, J.Tang, H.Herrmann, K.Pollitt and J.L.Jimenez, “FAQs on Protecting Yourself from COVID-19 Aerosol Transmission”,
https://tinyurl.com/FAQ-aerosols (2020)
[3] https://www.medrxiv.org/content/10.1101/2020.10.13.20212233v1
[4] WHO guidelines on tuberculosis infection prevention and control (2019 update)
[5] https://www.cdc.gov/infectioncontrol/pdf/guidelines/environmental-guidelines-P.pdf

2 Electronics COOLING | SPRING 2021

 OCTOBER 19-20 2021

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:

www.thermal.live

presented by
COOLING EVENTS
News of Upcoming 2021 Thermal Management Events

INTERNATIONAL CONFERENCE AND EXHIBITION ON THERMAL & POWER SOLUTIONS

Online Event
26-29
NEW THIS YEAR – A TECHNOLOGY CROSS-OVER EXTRAVAGANZA! CICMT, High Temperature,
and Thermal & Power Packaging come together for a great opportunity for you…One location | One
registration | Three times the content, networking, and learning! The Thermal event has also been
upgraded from a Workshop to a full Conference to allow for more attendees, exhibitors, speakers, and
networking!

Previously, this event has been organized annually by IMAPS (since 1992 in Workshop format) to
specifically address current market needs and corresponding technical developments for electronics
thermal management. Presentations on leading-edge developments in thermal management components,
materials, and systems solutions for effectively dissipating heat from microelectronic devices and systems
are sought from industry and academia. The Workshop emphasizes practical, high-performance solutions
that target current and evolving requirements in mobile, computing, telecom, power electronics, military,
and aerospace systems. Single-company product development concepts are acceptable subjects; however,
all abstracts will be judged on their novelty and innovative contributions to the industry knowledge.

Desc. source: electronics-cooling.com

► www.imapseurope.org/event/cicmt-2021/

MAY THERMAL MANAGEMENT INNOVATION WEST-COAST 2021

Online Event
10-14
The Annual BEV Thermal Management Innovation USA Congress, West-Coast 2021 is the number one
event to match buyer requirements with expert solutions during the Automotive sector’s battery evo-
lution. Following on from the tremendous success of the previous three events, Thermal Management
Innovation USA has firmly established itself as the automotive industries first-class event to gain a clear
and pragmatic view of the key challenges and current need to know learning objectives surrounding
advanced battery thermal management systems; to increase efficiency, range, battery health, and optimize
solutions for increasingly demanding advanced charging requirements.

The series analyzes innovative battery management solutions, explores the most crucial engineering and
material challenges, and benchmarks strategic imperatives for next-generation BEV advancement. We
welcome you to join the industry’s largest technical meeting for Thermal Management professionals and
foremost communication network for OEMs, technology and solutions providers, and leading Research
Institutes alike; where powertrain experts will engage during a series of case study presentations, interac-
tive panels and unparalleled networking opportunities.

Desc. source: electronics-cooling.com

► www.battery-thermal-management-usa.com/

4 Electronics COOLING | SPRING 2021

JUNE APEC 2021
Phoenix Arizona Convention Center
9-12
APEC is the leading conference for practicing power electronics professionals and addresses a broad range
of topics in the use, design, manufacture, and marketing of all kinds of power electronics equipment. Join
us June 9-12, 2021 in Phoenix, AZ.

The Applied Power Electronics Conference (APEC) focuses on the practical and applied aspects of the
power electronics business. This is not just a designer’s conference; APEC has something of interest for
anyone involved in power electronics:

Equipment OEMs that use power supplies and dc-dc converters in their equipment
Designers of power supplies, dc-dc converters, motor drives, uninterruptable power supplies, inverters,
and any other power electronic circuits, equipment, and systems
Manufacturers and suppliers of components and assemblies used in power electronics
Manufacturing, quality, and test engineers involved with power electronics equipment
Marketing, sales, and anyone involved in the business of power electronics
Compliance engineers testing and qualifying power electronics equipment or equipment that uses power
electronics

Desc. source: electronics-cooling.com

► www.apec-conf.org

OCT THERMAL LIVE™

Online Event
19-20
Electronics Cooling’s Thermal LIVE™, the world’s largest online thermal management event, returns for its
7th year this October. The two-day program features experts in the thermal management field, presenting
live educational sessions on industry challenges, trends, and products. Past topics have included advanced
thermal techniques in power electronics, design and manufacturing of blind mate couplings, selecting
TIMs for different applications, calculations and design elements for liquid cooling, and more.

If you’re an electronics or mechanical engineer who works with thermal management, this event is
a can’t-miss.

Desc. source: electronics-cooling.com

► https://thermal.live/

NOV
THERMAL MANAGEMENT INNOVATION USA
TCF Center | Detroit, Michigan
16
The Annual BEV Thermal Management Innovation USA Congress, Detroit is the number one event
to match buyer requirements with expert solutions during the Automotive sector’s battery evolution.
Following on from the tremendous success of the previous three events, Thermal Management
Innovation USA has firmly established itself as the automotive industries first-class event to gain a clear
and pragmatic view of the key challenges and current need to know learning objectives surrounding
advanced battery thermal management systems; to increase efficiency, range, battery health, and optimize
solutions for increasingly demanding advanced charging requirements.

Desc. source: electronics-cooling.com

► www.battery-thermal-management-usa.com

Electronics-COOLING.com 5
C A L C U L AT I O N C O R N E R

Fan Cooling - Part 1: Determining Flow Rate

Ross Wilcoxon1 and Genevieve Martin2

Associate Technical Editors for Electronics Cooling
1
Collins Aerospace
2
Signify

F
or a number of reasons, including cost, simplicity, thermal resistance, that can be approximated as 1/(2ṁcp)*.
power consumption, noise, etc., natural convection is
the preferred approach for cooling electronic systems. To illustrate how these two effects can influence the overall
However, it is often the case that natural convection is thermal resistance in a forced cooling application, the thermal
simply not sufficient to remove dissipated power while meeting resistances for parallel plate heat sinks with different flow rates
other system requirements such as size. Therefore, cooling fans were estimated using the approach described by Simons [1] for
are commonly used to increase cooling capacity to achieve an air cooling a heat sink dissipating 100W with the geometries
adequate design. This series of two articles provides an overview shown in Table 1. Thermal resistances for different heat sinks
of the basics of effectively integrating cooling fans into a system and different flow rates are shown in Figure 1.
and understanding other impacts of the use of fans. This first
article focuses on how the use of fans impacts the cooling system Table 1 Example heat sink with plate fins
design and how to determine the flow rate that a fan can produce Geometry Value Unit
in a specific application.
Base Width 12.7 cm
Before describing how to design a fan-cooled system for elec- Fin Length 12.7 cm
tronics, it is useful to first ask the question ‘why do fans improve Fin Thickness 0.12 cm
cooling?’. While the benefits of a fan to improve cooling may
Fin Height 5 cm
seem obvious, a clear understanding of the physical mechanisms
by which flow rate impacts thermal resistance is critical to devel- 1.42,
oping effective designs. Fin Pitch 2.83, fins/cm
4.25
To begin, convection heat transfer (Q), from a solid surface to
a fluid, is described with the equation Q = h*A*(Tfluid - Tsurface),
where T is temperature, A is the surface area in contact with the
fluid, and h is the convection coefficient. In laminar flow, the
coefficient is constant, regardless of flow rate. At higher veloci-
ties, the flow becomes turbulent and the heat transfer coefficient
increases with velocity. While the surface temperature of a heat
sink may be approximately uniform, the fluid temperature in-
creases as it absorbs energy, with the fluid temperature at any
point in the system defined as Tfluid = ṁ * cp / Q’ + Tinlet, where ṁ
is the mass flow rate of coolant, cp is the specific heat of the cool-
ant, Q’ is the heat absorbed by the coolant to that point in the
system, and Tinlet is the temperature of the coolant when it enters
the system. A larger flow rate can potentially affect heat transfer
in two different ways: 1) by increasing the convection coeffi-
cient, which decreases the convective thermal resistance 1/hA,
and 2) by reducing how much the fluid temperature increases as
it flows through the system. This effectively adds an additional Figure 1: Example of the effects of flow rate and fin geometry on the thermal resistance
thermal resistance, which may be referred to as the advective of a heat sink

*A future issue of Electronics Cooling Magazine will include a Calculation Corner that provides a more general discussion on how to account for
advective thermal resistance effects.

6 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

For this specific heat sink configuration, when the flowrate is less increasing pressure drop. The next section discusses the flow rate
than ~5-10 m3/hr, the thermal resistance is dominated by the flow and pressure head generated by a fan and how system and fan
rate, i.e., advection thermal resistance. In this flow rate regime, characteristics are combined to determine the resulting system
there will be little to no benefit of changing the fin design. At the flow rate.
other extreme, above ~30 m3/hr, the benefits of increasing flow
rate on the overall thermal resistance become quite small. Instead, The pressure head generated by a fan as a function of the flow
the thermal resistance is dominated by the convection thermal rate it generates are generally described in terms of fan curves; a
resistance, which is decreased by increasing the heat transfer area few examples of which are shown in Figure 3. The different fans
by adding more fins. As described later in this article, the design referenced in this plot all operate with the same 24VDC input
of an air-cooled system typically begins by determining what flow power, but their physical geometries, expected use cases, price,
rate corresponds to a target temperature rise for the cooling air noise, etc. are quite different.
and then determining fan and heat sink designs that are suitable
for that flow.

In order to improve a fan cooled system, it is important that the

designers know where they are on the flow rate / thermal resist-
ance curve. If a design is in the advection-dominated regime,
increasing the size or number of fins will have minimal bene-
fit; more flow is needed. In contrast, if a system is operating in
the convection-dominated regime, there is little to be gained by
increasing the flow rate. Instead, the heat sink resistance needs
to be reduced, typically by increasing the number of fins. Figure
2 illustrates a typical result of modifying a heat sink to reduce
its thermal resistance: the pressure drop through the heat sink
increases. The pressure drop was calculated for the same three
heat sink configurations shown in Table 1, using the approach
described in Ref. [2]. In this case, the larger number of fins in
the same space reduces the flow area, which directly increases the
flow velocity and subsequently increases pressure drop.

Figure 3: Fan performance curve for three commercial fans operating at 24VDC, as
reported by manufacturers

In addition to fan curves, manufacturers also generally provide

characteristic data to help users understand their performance.
Table 2 lists manufacturer-reported operating characteristics for
the three air movers associated with the curves shown in Figure 3.

Table 2: Example fans

Fan A Fan B Fan C

Centrifugal
Type of Air Mover Axial Fan Axial Fan
Blower
Typical Use Commercial Systems Aerospace
Figure 2: Pressure drop through different heat sinks as a function of flow rate Maximum Flow
146 44 170
(m3/hr)
Thus, both the thermal and pressure drop characteristics of a Maximum Pressure
heat sink can be influenced by the flow rate generated by a fan. 230 287 498
(Pa)
While the previous figures and discussion apply to the thermal/
fluid effects in a heat sink, the behaviors also apply to the overall Power Consumption
9 9.6 43
thermal design of a system, in that the design generally requires a (W)
balance between reducing thermal resistance without excessively Speed (rpm) 6,800 2,800 14,600

Electronics-COOLING.com 7
The quantified flow and pressure head parameters correspond to high pressure drop systems*. The block arrows in the plot indicate
the extreme conditions (maximum flow is for zero pressure head, predicted operating points for each combination of fan and sys-
maximum pressure head is for zero flow rate), rather than actual tem curve, with the numbers in the block arrows indicating the
use conditions in which the fan must generate both pressure head flow rate for each operating point.
as well as flow. Thus, the actual operating point for a fan must be
determined by comparing the fan curve to the system curve (also Table 3 summarizes the predicted operating points for the three
called system resistance curve). different fans in the two different systems.

The pressure drop that occurs when fluid flows through a heat Table 3: Predicted operating points of fans
sink, or any system, is a function of the flow velocity. If flow is
completely laminar, the pressure drop is proportional to the System/Fan Combinations Fan A Fan B Fan C
square of the flow velocity; in completely turbulent flow, the pres-
sure drop is proportional to the velocity. Actual flow through a Flow (m3/hr) 118 42 160
system, which may include duct flow, turns, expansions, contrac- System 1
tions, screens, etc., generally consists of regions of laminar and Pressure Drop (Pa) 66 13 108
turbulent flow. As long as the flow of air is incompressible, local
average velocities are proportional to the volumetric flow rate Flow (m3/hr) 421 27 59*
through a system and the pressure drop as a function of the vol- System 2
umetric flow rate, V, generally follows the equation ΔP = C*Vn, Pressure Drop (Pa) 168 83 308*
where C is a constant and the exponent, n, is between 1 and 2.
The specific values of C and n depend on the system. This rela-
tionship for pressure drop as a function of flow rate is referred In both Figure 4 and Table 3, the operating point for Fan C in
to as the system curve and must be determined through testing, System 2 is shown with an asterisk. This is to indicate that the fan
simulations, or flow resistance analysis. will not actually operate at this point in an actual system due to
the fact that the system curve intercepts an unstable region of the
The flowrate in a fan-driven system is determined by superim- fan curve. Under stable operation, a fan curve has a downward
posing the system curve (pressure drop as a function of flow slope in which an increase in flow rate leads to a reduction in the
through the system) and the fan curve (pressure head generated pressure head generated by the fan. Under this condition, any
by the fan as a function of the flow rate it generates). The intersec- small increase in the system pressure drop (imagine an insect fly-
tion of these two curves is known as the operating point. Figure ing past the system exhaust and temporarily creating additional
4 shows nominal operating points for the three fan curves shown pressure drop), the flow rate that the fan can generate is reduced.
previously for two system curves selected to represent low and This reduction in flow leads to a reduction in the system pressure
drop and once the perturbation has passed (the insect has flown
away), the fan returns to its original operating point. In contrast,
if the performance curve has a positive slope in which fan pres-
sure head increases with flow rate, such as with Fan C in the flow
range of ~40 and ~100 m3/hr, any perturbations will cause the
flow rate to increase, which increases the pressure head, which in
turn leads to higher flow rate, and so on. When a fan is operated
in an unstable region in which the fan curve has a positive slope,
the flow rate can oscillate between the flow rates where the fan
curve has a negative slope. These oscillations may generate sub-
stantial noise and significantly reduce the life of the fan.

Selecting a specific fan for a specific system requires sufficient

knowledge of both the fan and system to avoid design prob-
lems. This may require an iterative process of selecting a fan,
determining what its performance will be in a system, and then
potentially modifying the system to improve its thermal resist-
ance and/or pressure drop, operating the fan at a higher voltage/
speed, or selecting a different fan until performance specifica-
tions are met. Generally, one way to begin a design is to define
Figure 4: Predicted operating points for the three fans in two different systems an allowable coolant temperature rise for the system, i.e., the dif-

For reference, the system curves used in this plot were System 1: ΔP = 0.032*V1.6, System 2: ΔP = 0.25*V1.75 with pressure drop in
*

Pascals and flow rate in cubic meters per hour.

8 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

ference between the inlet and exit air temperatures, and deter- The next article in this series will discuss specific issues related to
mining the mass flow rate needed to achieve that condition. The integrating fans into systems. This includes how, when and why
appropriate target temperature increase depends on the specific to use multiple fans, type of fans, how fan orientation (positive
application and thermal budget available to the designer3. Sim- or negative pressure) can influence designs, issues related to fan
ulations or rough analysis, such as with a spreadsheet or a flow noise, efficiency, power, etc.
resistance tool, can then be used to estimate the pressure drop
through the system for that flow rate. An initial fan selection can REFERENCES
be made by identifying fans that have a maximum (zero pressure 1. R. Simons, “Calculation Corner: Estimating Parallel
drop) flow rate that is ~130% of that flow rate and a maximum Plate-fin Heat Sink Pressure Drop” Electronics Cooling
pressure head that is ~150% of the estimated system pressure Magazine, December 2015, https://www.electronics-coo-
drop. Once the candidate fan has been selected, a more detailed ling.com/2016/04/calculation-corner-estimating-pa-
comparison between the system curve and the fan curve can be rallel-plate-fin-heat-sink-pressure-drop/ (accessed 1/10/21)
used to identify the operating point for the fan and the resulting
thermal conditions. 2. R. Simons, “Calculation Corner: Estimating Parallel Plate-fin
Heat Sink Pressure Drop”, Electronics Cooling Magazine, April
3
For example, the flow rate of cooling air supplied to avionics 2016, https://www.electronics-cooling.com/2016/04/calcu-
commercial aircraft is set to establish a 15°C difference between lation-corner-estimating-parallel-plate-fin-heat-sink-pres-
exhaust and inlet temperatures. sure-drop/ (accessed 1/10/21)

Electronics-COOLING.com 9
S TAT I S T I C S C O R N E R

Confidence Intervals

Ross Wilcoxon
Associate Technical Editor for Electronics Cooling
Collins Aerospace

T
he previous articles in this series [1, 2] described how
the mean and standard deviations of a set of data are
calculated and how they can be used to estimate the char-
acteristics of a population using the normal distribution.
Since measured data typically represents only a subset of an entire
population, one should recognize that the estimated mean and
standard deviation values determined from a sample set are not
likely to be exactly equal to the true values of the entire population.
In other words, when we calculate the mean, i.e., average, of a data
set, we should recognize that there is actually a range of values in
which the true population mean lies – and the size of that range
depends on the confidence level that we assign to our estimate.
This article describes a process that can be used to determine that
range as a function of the number of data points and confidence
level. More generally, this outlines the overall process for linking
probability distributions with confidence levels that is the basis of
a variety of statistical analyses.

Figure 1 shows an example of a normal distribution in which the

mean value is 8 and the standard deviation is 0.75. Two lines are
included in this plot: the probability distribution corresponds to the Figure 1: Normal distribution for mean of 8 and standard deviation of 0.75
probability of any specific value occurring within the population
while the cumulative distribution indicates what portion of the As mentioned previously, we don’t generally know the true mean
population is less than or equal to a given value1. The cumula- and standard deviation of an entire population because we only
tive distribution curve is equal to the area under the curve of the make measurements of a subset of it. The larger the subset (the
probability distribution. more samples we use in making our estimates), the more accurate
our estimates of the true values should be. When we use a small
Z-value is defined as the difference of a value from the mean, nor- sample size in an analysis, we can account for the additional un-
malized by the standard deviation. For example, with the mean of certainty due to sample size by ‘flattening’ the bell curve of the
8 and standard deviation of 0.75, a measurement of 9 would have a standard normal distribution and increasing the size of the ‘tails’
Z-value of (9-8)/0.75 = 1.33. The cumulative normal distribution (the areas under the curve that are far from the midpoint) using
for this value is 90.9%2; in other words, 90.9% of the population the Student t-distribution.
with a normal distribution would have a Z-value that is less than
or equal to 1.33. This is illustrated in Figure 1 as the area under The t-distribution looks very similar to a normal distribution, but
the cumulative probability curve for values of 9 and below. its specific shape depends on the number of degrees of freedom.

1
To use Excel to calculate the values of these curves, use the function =norm.dist(x, mean, stdev, dist), where x is the x-axis value, mean
and stdev are the mean and standard deviation of the population respectively, and ‘dist’ is FALSE for the probability distribution and
TRUE for the cumulative distribution.
2
This value can be calculated with Excel in two different ways (at least). The simple approach is the function =norm.dist(9,8,0.75,TRUE).
Another approach is to use the Z-value with a standard normal distribution, which has a mean of 0 and standard deviation of 1. Thus,
the function would be =norm.dist(1.33,0,1,TRUE). Both of these functions will return the same value of 0.9088.

10 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

The degrees of freedom (DoF) for a sample set corresponds to the With this baseline information on distributions established, we
number of independent values used in the analysis. In this case, can now describe how they are used to define what range of val-
we can consider a data set of n data point to be comprised of n-1 ues the true mean of a population is based on a small number of
independent values; the difference between the sample size and measurements. The Central Limit Theorem is a critical compo-
DoF is due to the use of the data points to estimate the population nent in establishing this. The Central Limit Theorem states that a
mean [3]. Thus, one data point is not independent and the other population of terms (X-µ)/(σ/n1/2) tends towards being normally
data points can be used to assess to assess the uncertainty. Figure 2 distributed, where X is a value in the population, µ is the mean,
shows t-distributions for 2, 4 and 10 DoF and compares them to the σ is the standard deviation and n is the number of samples used
normal distribution, again for a mean of 8 and standard deviation to estimate the population characteristics. This applies not only
of 0.75. As the DoF increases, the t-distribution converges to the to the data, but also to the distribution that we use to assess our
normal distribution; above 30 DoF, the t-distribution is virtually confidence in the mean that is calculated from a sample set.
identical to the normal distribution.
For example, assume that 50 measurements of a heat sink show
that its thermal resistance is 8°C/W with a standard deviation of
0.75°C/W. How confident can we be that the actual population
mean is somewhere between 7.9 and 8.1°C/W? Using the Central
Limit Theorem, we calculate a test statistic as (7.9-8)/(0.75/501/2)
= -0.943. We can look this value up on a standardized normal
distribution table, which shows that 17.3% of a normal population
that has a mean of 0 and standard deviation of 1 will have a value
of -0.943 or less3. Since the normal distribution is symmetric, we
will also find that 17.3% of the population will have value of 0.943
or greater. Thus, 34.6% of the normal distribution is either less
than 7.9 or greater than 8.1 and there is a confidence band of ~65%
that the true mean is between 7.9 and 8.1.

Typically, we are more interested in conducting the reverse anal-

ysis – namely, what range of values corresponds to a prescribed
confidence band. Also, we may not have the luxury to have a suf-
ficient number of measurements (more than 30) to justify using a
standard normal distribution, rather than a t-distribution, in our
calculations. The steps for determining the range of mean values
correspond to a specified confidence band are shown in Table 1.
This table includes example calculations for testing on ten heat
sinks that again showed a mean thermal resistance of 8°C/W with
a standard deviation of 0.75°C/W. The goal of the analysis is to
Figure 2: Normal and t-distributions for mean of 80 and standard deviation of 0.75

# Step Example (with Excel functions)

1 Measure n samples and calculate the nominal mean, µ, and n = 10, µ = 8°C/W, σ = 0.75°C/W
standard deviation, σ
2 Define confidence interval, C.I. that defines the range of C.I. = 90%; two tails (one on each side of the distribution) =>
means, leading to the size of the tails outside the C.I., α/2 area in each tail is (1-0.9)/2 = α/2 = 0.05
3 Calculate the inverse t-distribution for the tail size and tα/2 = t.inv((1- α/2),(10-1)) = t.inv(0.95,9) = 1.833
degrees of freedom, tα/2 t.inv inputs: probability of being outside the tail (1-α/2), DoF
(sample size – 1)
4 Calculate the distance from the lower limit of the Δ = tα/2 * σ/ n1/2
confidence interval to the mean, Δ =1.833 * 0.75 / sqrt(10) = 0.43
5 Calculate the range that corresponds to the confidence Confidence interval: 8 - 0.43 to 8 + 0.43
interval = µ ± Δ C.I. = 7.57 to 8.43°C/W
Table 1

3
Or, we can use the Excel function =norm.dist(-0.943,0,1,true) if that seems easier…

Electronics-COOLING.com 11
determine the range of values that we can be 90% confident that
the true mean lies within.

Figure 3 shows the 90% confidence intervals calculated for the t-

and normal distributions for data with the same mean and standard
deviations, but different sample sizes. As the sample size increases,
the lines for the confidence intervals come converge. At small
sample sizes, the confidence bands that are calculated using the
more appropriate t-distribution are much wider than those calcu-
lated using the standard normal distribution. In general, it may be
questionable that an extremely small sample size of 2-3 samples
is necessarily representative of a population – and that small of a
sample size also leads to an uncomfortably large confidence inter-
val. But when dealing with typical sample populations with 5-20
measurements, the t-distribution provides a reasonable approach
for estimating the confidence interval and can be evaluated for
whether it is likely from a normal distribution (a discussion for a
future article). When the sample size is greater than 30, the normal
distribution can be used to assess the confidence interval. How-
ever, since the t-distribution converges to the normal distribution
at large sample sizes, one can continue to use the t-distribution
even with larger populations. So in general, if these equations are
incorporated into a tool such as Excel, it is appropriate to use the Figure 3: Confidence bands calculated for different sample sizes of heat sink resistance
t-distribution even for very large sample sizes.

In summary, the goal of the first three articles in this series has REFERENCES
been to provide a sufficient background to allow readers to better 1. Ross Wilcoxon, “Statistics Corner—Probability”, Electronics
understand future articles aimed at providing practical statistical Cooling Magazine, Spring 2020
analysis approaches. Hopefully, the articles did not achieve a
‘worst of both worlds’ status in which they included more theory 2. Ross Wilcoxon, “Normal Distribution”, Electronics Cooling
than engineers might want to see and less theory than statisticians Magazine, Summer 2020
would expect. Regardless of whether they achieved that or not,
with a basic statistical foundation in place we can now move on in 3. https:// blog.minitab.com / blog /statistics-and-quality-data-anal-
future articles to describe tools and analysis methods for solving ysis/what-are-degrees-of-freedom-in-statistics
the types of statistical problems that engineers may encounter.

12 Electronics COOLING | SPRING 2021

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

Why Chemical Compatibility Is Vital To Your Liquid

Cooling System

Elizabeth Langer
Engineering Manager, Thermal Business Unit at CPC

P
ower densities in electronic subsystems continue to in- might have. Specifically, when assessing concerns related chemical
crease, driving demand for more extreme cooling pow- compatibility, potential permeation and diffusive losses, it is im-
er alternatives that increasingly include liquid cooling portant to identify critical points of connection – such as tubing
as a viable candidate. To optimize thermal management junctions, manifold ports, and quick disconnect fittings – and
efficiency, sustainability and reliability, designers of systems that evaluate each one for risks to reliability and performance.
use liquid cooling are exploring innovative combinations of
component materials, including advanced thermoplastics, spe-
cialized elastomers, metal alloys and engineered fluids.

Whether designing a closed-loop, single-phase immersion, two-

phase immersion or direct-to-chip cooling system, component
material compatibility is critical to performance. This article pro-
vides general guidance regarding selecting the right ingredients
for a liquid cooling solution.

THINK HOLISTICALLY WHEN

SELECTING COMPONENTS
A variety of subsystems and components make up the architec-
ture that is critical to the successful and reliable operation of any
cooling system.

Each system component has the potential to interact with other

component materials. Therefore, materials interactions and depen-
dencies warrant detailed analysis during design and specification. Figure 1

Coolants are of particular interest, not only as the primary con- Overall, when it comes to material selection, one must think
duit of thermal transfer, but because they are in contact with all holistically. This includes accounting for all system components
wetted materials within a particular cooling system, as indicated and considering the potential effects of the environment, work-
in Figure 1. Some fluids may promote corrosion or biofouling in ing fluid, temperature, pressure, and mechanical loading, which
the presence of certain materials, creating the potential for flow might adversely impact performance.
blockage or failure of the cooling system. It is essential to under-
stand what all the materials are and the interactions that they This article will discuss liquids commonly used in liquid cooling

Elizabeth Langer
Beth Langer leads a team of design engineers focused on quick disconnect coupling product development. She
earned her Bachelor and Master of Science degrees in Mechanical Engineering from the University of St. Thomas
in St. Paul, MN. Over the course of her career, she has been responsible for product design, reliability and techni-
cal innovation for industrial products. Besides leading design solutions that serve CPC customers, she is an active
consulting member on ASHRAE and OCP committees.

14 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

applications and present an overview of the materials of construc- TAPPING COOLANT ALTERNATIVES
tion. Finally, we’ll provide guidance regarding the potential com- Obviously, choosing a coolant is a focal point when designing a
patibility of these fluids and materials when used together. liquid cooling system. From a compatibility standpoint, it is im-
portant to recognize that the fluid links virtually every compo-
nent as it circulates through the liquid cooling system. Table 1
provides a brief overview of a number of fluids typically used in
cooling electronics.

The first step in selecting a coolant is to consider operating and

storage temperatures. Fluid properties must be appropriate to
the application environment, such as a boiling point that sat-
isfies the thermal load and thermal efficiency needed without
exceeding the critical heat flux. Fluids must also have suitable
low temperature characteristics during storage and shipping, en-
vironmental exposures, particularly engineered dielectrics such
as fluorochemicals, as well as refrigerants. It is often necessary
to understand the environmental impact of the fluid throughout
the life cycle – how it's manufactured, the potential impact of it
leaching into the facility or atmosphere during use, and end-of-
life fluid reclamation requirements.

When selecting fluids, the ozone depletion and global warming

potentials need to be considered, particularly for refrigerants and
dielectrics. Over the last decade or so, the World Health Organi-
zation guidelines have increased emphasis on these parameters,
prompting the development of greener alternatives, such as 3M
Novec™, HFE coolant, and more environmentally friendly, fourth
generation hydrofluoroolefin refrigerants like R-1234 or R-1336.

Table 1. Liquids Commonly used in Liquid Cooling Systems for Electronics Table 2. Construction materials that are commonly used in electronics cooling applications

Electronics-COOLING.com 15
 Engineered thermoplastics can be an excellent choice, especially
when considering effects of weight, chemical compatibility, and
price over metal counterparts. When specifying thermoplastic
materials look for mechanical strength, chemical compatibility,
and thermal stability characteristics.

Polymer Limitations
Polymers, in particular commodity plastics and some thermo-
plastics, may present issues in certain applications.

Given the emerging prevalence of warm-water cooling systems,

polymer resistance to hydrolysis has become an important factor.
Polymers with hydrolyzable links may be at risk for severe prop-
erty degradation in hot water environments. The same risk may
apply for fluorochemicals in contact with fluorinated polymers. As
we know, like dissolves like, and there could be a risk of solubility
of certain plasticizers or additives into the coolant fluid.

Copper Since flammability may also be a concern with some polymers,

designs should include inherently non-flammable materials, spe-
cifically non halogenated thermoplastics. Long-term exposure to
a wide range of temperatures is certainly a key consideration for
Highest in durability and stability. Lower material selection in cooling systems.
Stainless Steel thermal conductivity. Higher cost.
Passivation increases corrosion resistance.
Additional risks associated with thermoplastics include chemical
attacks and crazing, cracking, discoloration, and, as previously
Table 2. Continued mentioned, extraction or leaching into the coolant. Fluid absorp-
Construction materials that are commonly used in electronics cooling applications tion, swelling and certainly thermal aging and degradation effects
over time and also mechanical loading and an internal pressure
In addition to thermal stability and chemical compatibilities, stresses are potential threats to integrity as well.
properties such as coolant toxicity, flammability, cleanliness re-
quirements, environmental impact, and cost should be consid- Elastomers
ered. And of course, when comparing coolant types and options, Elastomers can be engineered to meet a wide range of perfor-
all materials that the fluid may come in contact with throughout mance requirements.
the system should be evaluated.
Elastomers are polymers that have the property of viscoelasticity
SIZING UP MATERIALS OF CONSTRUCTION – they are rubbery and flexible – and are primarily used in com-
Electronics cooling system components are, in general, comprised ponents for fluid transport, such as tubing and hose, as well as
of three types of polymers – commodity plastics, engineered sealing components such as O-rings and gaskets. To understand
thermoplastics, and elastomers – and four types of metal alloys how elastomers tend to behave, we can look at how they’re made.
– aluminum, brass, copper and stainless steel. Table 2 provides a
high-level comparison of liquid cooling system component mate- Vulcanization, or the process of curing, creates permanent cross-
rials of construction. links in long polymer chains in elastomers. These chains ensure
that when stresses are loaded and unloaded, the elastomeric
Polymer Strengths component will return to its original position. In the case of an
Polymer properties can vary widely based on processing, additives, O-ring and a quick disconnect, for example, an elastomer will
fillers, and where they are on the spectrum, from commodity to maintain its seal.
ultra-high-performance thermoplastics and elastomers.
At a high level, specifying elastomers for use in a liquid cool-
Polymers can replace metal in many areas, and often provide ing application requires detailed analysis and evaluation with
additional benefits. For example, engineered thermoplastics like the selected unique coolant to ensure compatibility and long-
PPSU and PEEK can meet higher thermal, chemical and me- term reliability.
chanical requirements compared to metals while providing the
added benefits of reduced weight and better corrosion resistance For discussion purposes, some common material compound
at a potentially lower cost. categories that might be seen in these applications can be iden-

16 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

tified as hydrogenated nitrile, ethylene-propylene or EPDM, and clude an elastomeric O-ring seal, a polysulfone thumb latch and
chloroprene. HNBR has great chemical resistance, excellent me- stainless steel springs. However, only the interior surface of the
chanical properties, including tensile strength, tear modulus to connector and the elastomer seal would be wetted in a closed-
the wide temperature range and can be compounded for excellent loop cooling system and those materials need to be considered
resistance for high pressure applications. EPDM has excellent hot for compatibility relative to the selected coolant.
water and steam resistance characteristics. However, because it
has lower resistance to hydrocarbon, it is not well suited for any
refrigerant type application.

Chloroprene, which is commonly known as neoprene, is very

resistant to many chlorofluorocarbons or CFCs that are used as
refrigerant. It has low cost but moderate chemical resistance and
limited temperature resistance.

Some additional things to consider when specifying elastomers

are to consider the hardness (the durometer), the thermal ro-
bustness under both continuous and intermittent exposures,
and certainly the compounding as it relates directly to chemical
compatibility.

Metal Alloys
Compared to commodity plastic polymers, metal components in
liquid cooling systems are generally more stable, more durable
and have a perception of longer-term reliability. Metal compo-
Figure 2
nents also tend to be heavier and can be more expensive. In many
applications, the enhanced performance characteristics of met-
als warrants the additional investment. In other applications, the At a high level, fluids can affect polymers in two different ways:
right polymer may actually provide the best solution. physically and chemically. The first is generally reversible while
the other is not. For example, an O-ring compound in a quick
When considering metal alloys for use in systems, one should disconnect might have an affinity for a certain coolant, causing
account for mechanical strength, surface treatment, and cleanli- the O-ring to swell, which creates connection and disconnection
ness. While many refrigerants and engineered fluids are low- to issues that potentially lead to leaks. Replacing the O-ring with an
non-corrosive to metals, designers still must consider the operat- alternative plastic or specifying a different fluid could correct the
ing environment with regard to corrosion. problem.

PUTTING IT ALL TOGETHER: However, in a chemical interaction, in which a plasticizer is ex-

CHEMICAL COMPATIBILITY tracted from a component such as tubing, the effects of that dis-
With a foundational understanding of the fluids, plastics and metals solved plasticizer on the fluid's performance can be dramatic and
that might be employed in a given liquid cooling application, one are irreversible, which can be a critical issue in sensitive high-val-
can assess potential chemical compatibility of system components, ue applications.
based on their make-up, to ensure reliable, long term operation.
General guidance can provide a good starting point. Table 3 pro-
While polymers and metals can be effective in any combination vides an overview of relative compatibility between various ma-
when appropriately specified, it is critical to distinguish wetted terials and coolant options. Remember, a holistic view of the full
materials of construction from structural materials. Wetted ma- application details is the best way to ensure that the right materi-
terials include all components that are directly exposed to the als are specified. It is incumbent on system designers to test com-
coolant and therefore, are indirectly exposed to one another. ponents under expected operating extremes for their applications
Structural materials are not exposed to coolant during normal to assess fluid and material interactions at application-specific
operation. Creating a list of wetted and structural materials early temperatures, pressures, and other environmental conditions.
in the design cycle can help avoid complications down the road.
A given component might potentially be built of a combination A key to successful liquid cooling design is to engage with compo-
of polymers and metals. Thus, it is important to distinguish the nent suppliers early. This allows for the identification of any ma-
wetted materials from structural materials within a given com- terials that might be exposed to coolant, and any other variables
ponent. For example, a quick disconnect, such as those shown that might be present so that a design solution that is optimized
in Figure 2, may be constructed of nickel-plated brass and in- for the specific system requirements can be developed.

Electronics-COOLING.com 17
Table 3. Typical fluid and material compatibilities

Material and coolant compatibility

When considering wetted components in a liquid cooling system, the following combinations are:

A = Recommended. Little or no potential for chemical reaction or corrosion.

B = Good options. Minor potential for chemical reaction or corrosion, with limited effect on system performance.
F = Not recommended. Mild to severe chemical or corrosive reactions likely. May impede system performance.

Sometimes

BIGGER
is
BETTER.
Introducing our Everis™ LQ8
and Everis™ BLQ6 connectors.

At CPC, we continue to expand our portfolio to meet your liquid cooling needs.
Our newest additions include the Everis LQ8 and Everis BLQ6. The Everis LQ8
is a purpose-built connector for liquid cooling that offers high-flow capacity to
optimize system performance. The Everis BLQ6 blind mate connector allows you
to address greater cooling needs with a large panel-mount connector.

Learn more at cpcworldwide.com/coolinglarge

18 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

TECHNICAL EDITORS SPOTLIGHT

Meet the 2021 Electronics Cooling® Editorial Board

VICTOR CHIRIAC, PhD | 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 since 2919. 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 U.S. issued patents, 2 US Trade Secrets and 1 Defensive Publication (with Motorola), and has published
over 107 papers in scientific journals.
► vchiriac@gctg-llc.com

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 Semitherm conference
and is an active reviewer and technical committee in key conferences Semi-Therm®, Therminic, Eurosime.
► genevieve.martin@signify.com

ROSS WILCOXON | ROCKWELL COLLINS ADVANCED TECHNOLOGY CENTER

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 US 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

Electronics-COOLING.com 19
F E AT U R E D

Better Cooling by Removing Material,

Instead of Adding Material, to Unlock the
Full Power of GaN Electronics

Remco van Erp

A
s society is moving from fossil fuels to more sustain- thin, epitaxial layer of GaN on a low-cost silicon substrate, making
able electrical power, the systems for converting and the material cost effective in growth as well as in processing, since
delivering this electrical power are becoming increas- established silicon fabs can be utilized. This low-cost GaN-on-Si
ingly important. These power electronic systems are is the driving force behind the commercial adoption of GaN in
responsible for controlling and shaping the electricity between power electronics.
the sources, such as solar and wind, and the end-use destination,
such as driving a motor or charging the battery of your phone. Aside from the benefits of faster switching and lower losses, a
The transistor is the fundamental building block of these convert- unique property that distinguishes the GaN power transistor
ers. While silicon transistors thrive in computing applications, from its silicon counterpart is its lateral device structure. Transis-
they have limitations when it comes to handling high power and tors can be placed side-by-side on the same chip and assembled
high voltages. Many applications are very constrained in space together into a power integrated circuit (Power IC), creating the
and weight and the challenge is to get more power in a smaller potential of high-power converters integrated on a small chip. To
package. Gallium nitride (GaN) transistors are key in achieving fully appreciate this feature, we have to jump back to the early dis-
this goal, but its full potential remains untapped due to thermal covery of the transistor. Shortly after Shockley, Brattain and Bar-
limitations. The paradigm of cooling has long been about adding deen demonstrated the first transistor, then-vice-president of Bell
more material and going bigger: more metal, bigger heat sinks. Labs, Jack Morton, spoke about the tyranny of numbers. For each
Yet, the limitation on heat extraction often lies at the interfaces extra component soldered together on a circuit, the system be-
between these components. Instead, in this article we discuss the comes more complex and likely to fail. The same factors hold for
alternative path to improve cooling by not adding material, but power converters: more complex topologies with multiple devices
by removing it. We evaluate several approaches of microchannel can improve efficiency, add functionality and reduce reliance on
liquid cooling of GaN power devices and show the benefits of bulky passive components such as inductors. However, the higher
moving the cooling inside the chip. component count is usually faced with reliability concerns. Just
as the integrated circuit helped to manage the tyranny of num-
Gallium nitride (GaN), a wide-band-gap semiconductor, has fa- bers for processors, the GaN-based power IC may do the same
vorable material properties compared to silicon for power con- for power conversion. One could imagine a centralized chip that
version. It enables faster-switching, smaller converters with lower contains all active high-voltage power switching components, po-
losses. GaN technology, which originated from the Nobel-prize tentially co-packaged with passive components such as capacitors
winning research on blue-light LEDs, has steadily matured over [3]. A system like this could be a disruptive technology for power
the last years. One such development is the capability to grow a conversion, leading to more compact and powerful systems.

Remco van Erp

Remco van Erp received his B.S and M.S degree in mechanical engineering from Eindhoven University of Technology,
the Netherlands, in 2014 and 2017, resepectively. He is currently pursuing his PhD degree in microsystems and
microelectronics at École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland. His research interest
focuses on microfluidics, heat transfer in microstructures and liquid cooling for high power density electronics.

20 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

While this will be beneficial from a power density point of view, it

aggravates the thermal challenges involved in maintaining junc- COP = Qmax/PpumpRtotal = ΔTmax/fΔpRtotal
tion temperatures at acceptable levels. Power dissipation will be
concentrated in a single chip, so it cannot be spread to multiple Rtotal is a summation of the individual contributions of thermal
large heat sinks. To fully exploit the possibilities of GaN-based resistances between the power dissipating hot spot and the cool-
power ICs, new strategies for effectively extracting the heat are ant at the system inlet. As the flow rate increases, Rtotal decreases,
required. Since GaN is grown as a thin layer on a silicon substrate, but Ppump quadratically increases due to dependency on both flow
and this substrate has no functionality except for being an inex- rate (f) and pressure drop (Δp). Ignoring any entrance effects, the
pensive carrier, it is possible to tap into the developments of in sil- COP can be seen as independent of die size. For example, two
icon microchannel heat sinks from past decades [1]. Microscopic chips can be placed side-by-side, doubling both the maximum
cooling channels etched directly in the silicon substrate can func- heat load and the flow rate, resulting in an equal COP. Plotting the
tion as a high-performance heat sink. The large surface-to-volume COP against the maximum heat flux, qmax, provides a benchmark
ratio of such microchannels provides high heat extraction while of heat extraction efficiency versus heat extraction capability.
the absence of thermal interfaces ensures a low overall thermal
resistance. In other words, it can turn the silicon substrate from a qmax = Qmax/Adie
low-cost substrate into a high performance heat sink (Fig. 1).
Figure 2 shows an overview of three approaches we compared using
In this work we evaluate several embedded microchannel cooling this approach. The first, in Fig. 2a shows an indirect cooling ap-
approaches for GaN-on-Si power devices, and benchmark their proach in which a silicon microchannel heat sink is attached with
performance. A useful non-dimensional metric for such bench- an intermediate thermal interface material (TIM) to a commercial
marking of cooling performance is the ratio of extracted power packaged GaN transistor [4]. A one-dimensional thermal resis-
(Qmax = ΔTmaxRtotal)to pumping power (Ppump = fΔp) at a certain tance network can be used to model the heat transfer between the
flow rate. This is known as the coefficient of performance (COP) device’s junction and coolant inlet. The junction-to-case thermal
and is given by the following formula: resistance (Rj-c) is given by the device manufacturer and consists

(a) (b) (c)

Figure 1: (a) Bulk GaN, (b) GaN-on-Si reduces cost, (c) Embedded microchannels for improved cooling.

(a) (b) (c)

Figure 2: (a) Indirect cooling, a microchannel cold plate is attached to a packaged die with an intermediate thermal interface material. (b) Direct die embedded microchannel cooling,
cooling channels are etched inside the backside of the silicon substrate. (c) GaN-on-Si device with a monolithically integrated manifold microchannel inside the silicon substrate, co-
designed with the electronics. Cooling channels are positioned below and aligned with the pads of the electronic device.

Electronics-COOLING.com 21
of the various conductive elements between the hot-spot, die, and improve cooling is not by adding a better heatsink, but by remov-
packaging layers. ing the packaging and thermal interfaces. In the approach shown
in Fig. 2b, similar 50 µm microchannels are directly embedded
• RTIM describes the thermal resistance due to the TIM, into the silicon substrate in the same approach initially proposed
• Rcond accounts for the conduction in the silicon cold plate by Tuckerman [1]. The second column in Fig. 3a highlights the
• and Rconv accounts for the convective heat transfer between the effectiveness of this approach: the contributions of Rj-c and RTIM
solid silicon domain and the liquid coolant. are eliminated, resulting a substantial reduction in thermal resis-
• Finally, Rheat is the thermal resistance due to the heat capacity of tance. Since similar channel dimensions are used as in the indirect
the coolant. Rheat scales with the flow-rate of the coolant (Rheat = approach, the pressure drop between the indirect and direct ap-
1/ρcpf) and does not depend on the geometry of the chip. proach remain comparable (Fig. 3b). As a result, at a comparably
high COP of 104, this direct approach achieves a 10-fold increase
All the other thermal resistance components scale directly with in maximum heat flux up to 500 W/cm2 (Fig. 3c) [5]. Further-
the surface area of the chip, i.e., increasing chip size reduces these more, Fig. 3a reveals that the convective thermal resistance actually
thermal resistances. represents the limiting factor for heat extraction. We cannot scale
the heatsink since the dimensions are constrained by the chip size.
To compare multiple designs, we show the thermal resistance In order to reduce Rconv, smaller channel sizes can be utilized. How-
normalized by surface area in Fig. 3a. This shows that the major ever, this comes at a cost of increased pressure drop, which reduces
contributions in thermal resistance are related to the packaging COP and may add to system integration challenges.
and TIM, which account for more than 80% of the total. If the
pressure drop of Fig. 3b is accounted for, this indirect method One potential approach to decouple Rconv and pressure drip is by
enables the extraction of about 55 W/cm2 at a COP value ex- utilizing a third dimension. The manifold microchannel heat sink
ceeding 104. (MMC) is a hierarchical design that addresses these issues. In a
MMC, manifold channels distributes the liquid efficiently over
From this finding, it becomes clear that the most effective way to the chip, reducing pressure drop and increasing temperature uni-

(a) (b) (c)

2000

Figure 3: (a) Normalized thermal resistance components of indirect microchannel cooling, direct microchannel cooling and co-designed mMMC cooling. The components of R refer
to the thermal resistance network diagrams illustrated in Fig. 1. (b) Normalized pressure drop (pressure drop multiplied by chip aspect ratio W/L) versus flow rate for the three
configurations. (c) Benchmark of cooling efficiency (COP) versus maximum heat flux capability (q) for a maximum temperature rise of 60 K.

(a) (b)

Figure 4: GaN power device with microfluidically co-designed mMMC. (a) Schematic of the mMMC structure. (b) Picture of a fabricated GaN power device with mMMC cooling.

22 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

formity. However, design of such flow-hierarchy in MMC heat ded microchannel cooling. Here, the GaN chip contains multiple
sinks traditionally require multiple bonded layers together. This is power diodes, which are all cooled with water flowing inside the
not only a cumbersome fabrication step, but also raises reliability chip. Several challenges need to be addressed in terms of packag-
concerns under repeated thermal cycling. In the third approach in ing and reliability in order to see such technology materialize, but
Fig. 2c, we explored a new design approach, where the design of at least it’s sure that cooling doesn’t need to be the limiting factor
the cooling and electronics go hand-in-hand from the start of the in a future with more compact and efficient power conversion.
fabrication. We monolithically integrated an MMC structure in a  
GaN-on-Si device, which we call mMMC, using a new fabrication REFERENCES:
method that exploits the etching selectivity between silicon and [1] D. B. Tuckerman and R. F. W. Pease, “High-performance heat
GaN [6]. The typical source-drain spacing of GaN transistors and sinking for VLSI,” IEEE Electron Device Lett., vol. 2, no. 5, pp.
the optimal range of microchannel pith are both in the range of 126–129, May 1981.
15 to 25 microns. We exploited this fact by etching small trenches
through the GaN epilayer, realizing a cooling channel in the un- [2] D. B. Tuckerman et al., “Microchannel Heat Transfer: Early
derlying silicon substrate, and finally sealing the trenches in the History, Commercial Applications, and Emerging Opportunities,”
GaN epilayer during the metallization step. Using this approach, in ASME 2011 9th International Conference on Nanochannels,
each source and drain of the transistor is linked to an individual Microchannels, and Minichannels, Volume 2, 2011, pp. 739–756.
cooling channel (Fig. 4a). We refer to this approach as a micro-
fluidic-electronic co-design. The bonding interface between the [3] R. Reiner and B. Weiss, “PCB-Embedding for GaN-on-Si
manifold and the microchannels was eliminated and the entire Power Devices and ICs,” CIPS 2018; 10th Int. Conf. Integr. Power
3D cooling structure was realized in a single silicon crystal. This Electron. Syst., p. 6, 2018.
monolithic structure ensures structural integrity as well as excel-
lent heat transfer. Fig. 3a shows the substantial reduction in Rconv [4] R. Van Erp, G. Kampitsis, and E. Matioli, “Efficient Micro-
realized using this approach, as well as the reduction in pres- channel Cooling of Multiple Power Devices with Compact Flow
sure drop (Fig. 3b). Combining these properties, we can show an Distribution for High Power-Density Converters,” IEEE Trans.
additional increase in maximum heat flux of almost 3-fold at a Power Electron., vol. 35, no. 7, pp. 7235–7245, 2020.
COP of 104, relative to the microchannels etched in the back of
the substrate. [5] R. Van Erp, G. Kampitsis, L. Nela, R. S. Ardebili, and E. Ma-
tioli, “Embedded Microchannel Cooling for High Power-Density
GaN has the potential to reshape power electronic, and improved GaN-on-Si Power Integrated Circuits,” in InterSociety Confer-
cooling can be an enabling technology that facilitates this technol- ence on Thermal and Thermomechanical Phenomena in Elec-
ogy. The silicon substrate on which GaN is grown does not need tronic Systems, ITHERM, 2020, vol. 2020-July, pp. 53–59.
to merely be an economic choice; in addition, it can become an
efficient heat sink that pushes the boundaries of what is possible [6] R. van Erp, R. Soleimanzadeh, L. Nela, G. Kampitsis, and E.
in terms of integration and power density. As an example, Fig. Matioli, “Co-designing Electronics with Microfluidics for More
5 shows an integrated GaN-based power rectifier with embed- Sustainable Cooling,” Nature, vol. 586, no. 7824, 2020.

Figure 5: 1.2 kV/120 W GaN-based full-wave bridge rectifier (FWBR) power IC with embedded microchannel cooling. Schematic (left) and prototype (right)

Electronics-COOLING.com 23
F E AT U R E D

Thermal Packaging –
From Problem Solver to Performance Multiplier
Reprinted from Electronics Cooling website, December 5, 2013

Avram Bar-Cohen

T
he increased integration density of electronic compo- as thermoelectric, materials to link on-chip hot spots to micro-
nents and subsystems, including the nascent commer- fluidically-cooled microchannels. Such intra/inter chip enhanced
cialization of 3D chip stack technology, has exacerbated cooling approaches are required to be compatible with the mate-
the thermal management challenges facing electronic rials, fabrication procedures, and thermal management needs of
system developers. The sequential conductive and interfacial homogeneous and heterogeneous integration in 3D chip stacks,
thermal resistances associated with the prevailing “remote cool- 2.5D constructs, and planar arrays. A conceptual ICECool device
ing” paradigm in which heat must diffuse from the active regions is shown in Figure 1.
on the chip to the displaced coolant, have resulted in only limit-
ed improvements in the overall junction-to-ambient thermal re-
sistance of high-performance electronic systems during the past
decade. These limitations of Commercial Off-The-Shelf (COTS)
thermal packaging are undermining the cadence of Moore’s Law
and leading to a growing number of products that fail to realize
the inherent capability of their continuously improving materials
and architecture. Continued application of this “remote cooling”
paradigm has resulted in electronic systems in which the thermal
management hardware accounts for a large fraction of the sys- Figure 1: A Cross-Sectional Conceptual Schematic of an Embedded Cooling, Gen-3
tem volume, weight, and cost and undermines efforts to transfer (ICECool) device.
emerging components to small form-factor applications.
An intrachip approach would involve fabricating micropores and
To overcome these limitations and remove a significant barrier microchannels directly into the chip [3,4] while an interchip ap-
to continued Moore’s Law progression in electronic components proach would involve utilizing the microgap between chips in
and systems, it is essential to implement aggressive thermal man- three-dimensional stacks [5,6], as the cooling channel. In addition
agement techniques that directly cool the heat generation sites to the inclusion of an appropriate grid of passive and/or active ther-
in the chip, substrate, and/or package. The development and im- mal interconnects, it is expected that a combination of intrachip
plementation of such “Gen-3” embedded thermal management and interchip approaches, linked with thru-silicon and/or “blind”
technology, following on the Gen-1 air-conditioning approaches micropores, will confer added thermal management functionality.
of the early years and the decades-long commitment to the Gen-2 These microchannels and/or micropores will be integrated into a
“remote cooling” paradigm, is the focus of the current DARPA fluid distribution network, delivering chilled fluid to the chip or
Intra/Inter Chip Enhanced Cooling (ICECool) thermal packag- package and extracting a mixture of heated liquid and vapor to be
ing program. Launched in 2013, ICECool aims to develop and transported to the ambiently-cooled radiator.
demonstrate “embedded cooling” techniques capable of removing
kW/cm2 chip heat fluxes and kW/cm3 chip stack heat densities, Some 30 years of thermofluid and microfabrication R&D, driven
while suppressing the temperature rise of multi-kW/cm2 sub- initially by the publication of the Tuckerman & Pease micro-
mm hot spots [1,2]. The ICECool program is composed of two channel cooler paper in 1981 [1] and more recently by compact
thrusts: a 3-year ICECool Fundamentals effort, involving several heat exchanger and biofluidic applications [7,8], has created the
university teams which are developing embedded cooling build- scientific and engineering foundation for the aggressive imple-
ing blocks and modeling tools, and a 2.5-year ICECool Applica- mentation of the “embedded cooling” paradigm. Nevertheless,
tions effort, led by several aerospace performers and culminating substantial development and modeling challenges must be over-
in functional electronic demonstration modules. ICECool per- come if Gen-3 techniques are to supplant the current “remote
formers are pursuing the creation of a rich micro/nano grid of cooling” paradigm. Successful completion of the DARPA ICE-
thermal interconnects, using high thermal conductivity, as well Cool program requires overcoming multiple microfabrication,

24 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

thermofluid and design challenges, including: 2. ICECool Applications (ICECool Apps), DARPA BAA 13-21,
Posted 6 February 2013.
• Subtractive and additive microfabrication in silicon, silicon 3. D.B. Tuckerman and R.F.W. Pease. “High-performance heat
carbide, and synthetic diamond of high aspect ratio, thin- sinking for VLSI.” IEEE Electron Device Letters, vol. 2, no. 5, pp.
walled microchannels and high aspect ratio micropores; low 126–129, 1981.
thermal boundary resistance, high thermal conductivity ther- 4. L.P. Yarin, A. Mosyak and G. Hetsroni. Fluid Flow, Heat Trans-
mal interconnect grids; on-chip, high power factor, high COP fer, and Boiling in Micro-Channels. Springer, Berlin, 2009.
thin-film thermoelectric coolers; and hermetic attachment of 5. A. Bar-Cohen and K.J.L. Geisler. “Cooling the Electronic Brain.”
liquid supply and liquid/vapor removal tubes. Mechanical Engineering, pp. 38–41, April 2011
6. A. Bar-Cohen, J. Sheehan and E. Rahim. ”Two-Phase Thermal
• Convective and evaporative thermofluid transport in mi- Transport in Microgap Channels – Theory, Experimental Results,
crochannels and micropores– removal of 1 kW/cm2 chip heat and Predictive Relations.” Microgravity Science and Technology,
fluxes with 2-5 kW/cm2 sub-millimeter “hot spots”; low pump- pp. 1–15, Sept. 2011.
ing power liquid-vapor manifolds with Coefficients-of-Perfor- 7. S.Garimella, T. Harirchian, Microchannel Heat Sinks for Elec-
mance (CoP) between 20 and 30; high-exit-quality, greater than tronics Cooling, Vol 1, Encyclopedia of Thermal Packaging, 2013,
90%, evaporative flows without flow instabilities and/or local WSPC, SG
dryout; and high fidelity thermofluid models for single- and 8. L.P. Yarin, A. Mosyak, H. Hetsroni, Fluid Flow, Heat Transfer,
two-phase flow in microchannels, microgaps, and micropores. and Boiling in Micro-Channels, 2009, Spring-Verlag, Berlin

• Thermal/electrical co-design which moves progressively from Disclaimer: The views, opinions, and/or findings contained in this article/presenta-
tion are those of the author/presenter and should not be interpreted as representing
passive, thermally-informed designs, which recognize the im- the official views or policies, either expressed or implied, of the Defense Advanced
pact of temperature on functional performance, to active ther- Research Projects Agency or the Department of Defense. Distribution Statement A,
mal co-design which places functional paths and blocks in the Approved for Public Release, Distribution Unlimited.
most favorable locations on the chip, to fully-integrated co-de-
sign which deals with the impact of microfluidic channels and
thermal interconnects on the electrical design and placement
of electrical devices and cells, to mature designs that interac-
tively balance the use of resources to optimize layout for energy
consumption/functional performance.

• Physics of Failure models that address the failure mechanisms

and reliability of the Gen-3 thermal management components,
including erosion and corrosion in microchannels, microgaps,
and micropores; failure modes induced in the electrically ac-
tive areas of the chip and/or substrate; and the impact of micro-
fabrication and embedded cooling operation on the structural
integrity and stress profile of the microchanneled substrate (in-
trachip) and/or the chip-to-chip bonding (interchip).

Successful development and implementation of this Gen-3 ther-

mal packaging paradigm places thermal management on an equal
footing with functional design and power delivery, transform-
ing electronic system architecture and unleashing the power of
nanofeatured device technology, while overcoming the SWaP
(size, weight, and power consumption) bottleneck encountered
by many advanced electronic systems. After decades of mere
“problem solving” with Gen-1 (HVAC) and Gen-2 (spreaders,
heat sinks, and TIMs) thermal management technology, it is ex-
pected that widespread adoption of Gen-3 “embedded cooling”
techniques will provide a significant performance multiplier for
advanced electronic components.

REFERENCES
1. Intrachip/Interchip Enhanced Cooling Fundamentals (ICE-
Cool Fun), DARPA BAA 12-50, Posted 7 June 2012.

Electronics-COOLING.com 25
In Memoriam of Avram Bar-Cohen
– Victor Chiriac, Electronics Cooling Co-Editor, January 8, 2021

It is with sad feelings that I am compiling this remembrance article for Professor Avram Bar-Cohen, who departed this world too soon.
Avi inspired many scientists, educators and practicing engineers with his deep understanding of the thermal field and excitement in
seeking new technical breakthroughs. He mastered the secrets of combining the fundamentals of the academic field with the practicality
of novel technology implementation into the industrial products… I remember Avi’s infectious smile and enthusiasm when I first met
him during his visit to Motorola in 1997, while I was a summer student. Over my career I had the privilege to interact professionally
with Avi in my technical leadership roles at Motorola and later at Qualcomm. He was an inspiration to many of us who were peers,
colleagues and friends. There are many more things that could be said, but I would like to share here with our readership a collection
of memories from several distinguished educators and scientists who worked with and knew well Avi… God Bless and Rest in Peace
Avi. You are living through your Work and through the People who you helped shape into today’s thermal community leaders…

See below a collection of thoughts from some of Avi’s peers

A few personal notes by Clemens Lasance, Philips Research Emeritus

On October 11, I was shocked by being informed about prof. Avram Bar-Cohen’s sudden death, and I still am. Avi inspired everyone in
our thermal community, first of all by the impressive number and quality of his books, chapters and papers, but also for the people who
had the privilege to meet him in person by his enthusiasm and his presence. I will always remember his irresistible smile. Allow me a
few words on his achievements without even trying to be complete. He was a thermal engineering pioneer since the early seventies, from
both an academic and a practical point of view. He is worldwide recognized as a leader in thermal science and technology. He authored
and co-authored over 400 publications, delivered over 100 keynotes, and guided more than 70 masters and PhDs at the universities of
Maryland, Minnesota and Beer Sheva (Israel). He was honored by over 10 institutions (amongst with the THERMI Award in 1997) and
became recently a member of the EU Academy of Sciences. I agree totally with the words of Mike Pecht at the U. of Maryland: ‘’Avi’s
scientific accomplishments are prodigious, but it was his commitment to helping others that will be his enduring legacy.’’

I first met Avi at the IHTC in San Francisco in 1986, attending a course presented by him and Allan Kraus on heat sinks. Then again
at the IHTC in Jerusalem in 1990, and in 1993 for two weeks at the NATO Advanced Study Institute on Cooling of Electronic Systems
held in Cesme, Turkey, where we both lectured and participated in the closing panel session. I recall vividly not only our get togethers,
with a whole bunch of famous thermal experts, among others Arthur Bergles, Frank Incropera, Allan Kraus, Dick Chu, Al Ortega and
Wataru Nakayama, but especially our snorkeling adventures in the Mediterranean. When I together with the U. of Delft organized
the first thermal management conference in Europe in 1993, I invited Avi as a keynote speaker which he happily accepted. For sure he
brought the conference at a higher level with his active participation.

The most important result of the successful EU project DELPHI was the compact thermal modeling philosophy. Avi, who was involved
in this subject in those days, and me agreed to write a book on this topic together with three other authors, a serious start was made
but it never was finished. Then, some 10 years later, as the editor of the famous Encyclopedia of Thermal Packaging, Avi invited me to
contribute a volume on Compact Thermal Modeling, so I asked him what’s in for me, and he said: All volumes of the Encyclopedia.I
replied that I liked the proposition, the problem being that my bookshelves couldn’t carry their combined weight. He answered: No
problem, I am an excellent carpenter, when I am in the neighborhood, I give you a call. In 2011, I was very honored by being asked to write
a contribution to the Festschrift celebrating Avi’s 65th birthday. The subject I chose was about heat sinks, exactly the same subject that
started our relationship back in 1986. Avi, I will miss you as a scholar, friend and carpenter. I offer my deepest condolences to his wife,
children and family.

A few personal notes by Professor Michael Ohadi, U Maryland/ARPA-E

I first came to know Avi first in 1986, when I was a graduate student at the University of Minnesota and attended Avi’s newly offered
Electronics Cooling course there. Later Avi moved to the University of Maryland to chair the Department of Mechanical Engineering
at the University of Maryland (UMD) between 2001 and 2010. Avi brought much enthusiasm and momentum to UMD’s mechanical
engineering department. Among his many contributions, the Department’s national rankings vastly improved, and its research
expenditures increased by 63%.

26 Electronics COOLING | SPRING 2021

 Electronics COOLING | SPRING 2021

At the international level, under Avi’s leadership in 2002 UMD’s mechanical engineering participated in a sponsored international
research and educational collaboration with the United Arab Emirates (UAE). Through this multi-year collaboration, UMD’s College of
Engineering helped establish an engineering institution in the UAE, which included six engineering programs, a mechanical engineering
program which boasted the highest student enrollment of the six. The programs included both undergraduate and graduate programs.
In 2012, the six engineering programs were granted ABET accreditation and became the only institution in the UAE at the time with
such high recognition.

Avi’s administrative duties did not stop him from his active research, collaborative, and scholarly activities. He reached out and
collaborated with many of the faculty in mechanical engineering and across the campus. His scholarly work between 2001 to 2010
included more than 150 journal articles, peer-reviewed conference papers, and invited plenary/keynote/other major lectures. The
Mechanical Engineering faculty, staff, and students at UMD will miss Avi’s engaging and intellectual presence, momentum building
character, and his affable smile.

A few personal notes by Madhu Iyengar, Google

Professor Avram Bar-Cohen was a giant in the fields of electronics packaging and cooling, and a pioneer who has significantly influenced
academia, industry, and governmental research. Beyond his technical accomplishments and leadership, he was a fantastic mentor
for individuals and the community. He fostered a culture of innovation, collaboration, and high technical achievements for several
decades, and founded vibrant and healthy international conferences, such as the IEEE ITherm and the ASME InterPack. Avi's more
recent leadership as the leader of the DARPA IceCool program has led to breakthrough research on embedded cooling of ultra-high
performance chips, and I expect this achievement to be a beacon that shines light for future progress in the field.

For me personally, Avi was a teacher, mentor, and friend over the last 25 years, and whose passing I mourn and grieve deeply. Avi
inspired fearlessness and total commitment towards the progress of the field, with a cheerfulness and exuberance that was infectious.
He will be missed, but will continue to guide us through our memories.

A few personal notes by Bob Simons, IBM Emeritus

I knew Avram Bar-Cohen during most of his professional career, having first met him in 1976 when he was at IBM for a short while.
Since that time he became an internationally recognized leader in the electronics cooling and packaging community. Through his many
publications, lectures, short courses, and research, he contributed significantly to establishing a scientific foundation for the thermal
management of electronic components and systems. Those who had the opportunity to know him and interact with him professionally,
will remember him for his energy and passion for promoting both the art and science of cooling electronics.

A few personal notes by Professor YC Lee, U of Colorado at Boulder

Avi had been my mentor since the beginning of my professional career in Bell Labs. In 1984, I took my first short course taught by
him. In 1988, I attended the first ITherm organized by him. In 1995, Avi made a strategic move to hold the InterPACK in Maui, and
I supported this conference as a coordinator of international liaisons and a session chair. In 2010, Avi began his service as a DARPA
Program Manager, and I had a chance to work with him closely as the PI of a Thermal Ground Plane (TGP) project. In 2014, I started
serving as the Editor of the Journal of Electronic Packaging (JEP), and the first person I asked for help was Avi. Avi and his DARPA
team members contributed three reviews on TGP, Nanothermal Interface Materials, and Cooling for Wide Bandgap Power Amplifiers.
These articles made a major impact. I was blessed to have Avi as my mentor, and I was fortunate to witness and support Avi’s truly
outstanding service to the packaging community. Avi, we miss you!

A few personal notes by Jim Wilson, Principal Fellow, Raytheon Technologies

Avi’s contributions to the field of electronics cooling are numerous and found in many places including this magazine. His influence
extended beyond the many technical contributions to include the mentoring and collaboration and support he showed to his colleagues
and friends. I had the privilege of knowing Avi from professional society interactions, as an associate editor of Electronics Cooling, as a
coworker at Raytheon, and most importantly as a friend. His first exposure to the challenges of cooling electronics came when he was
a new mechanical engineer at Raytheon and this helped motivate further graduate studies and Raytheon made an excellent investment
helping support his graduate work.

Avi was a true trail blazer who enjoyed pushing for disruptive ideas that challenged the way things are done and was a fearless advocate
for the voice of thermal engineers all over the world. While he could be a challenging and persistent leader, he was a genuinely kind
person. He always found time to review a paper or discuss the latest technical topic (sometimes with emails that were sent in the
middle of the night). Avi found joy in selflessly helping others succeed and hopefully passed this characteristic on to others as part of
his legacy. He is deeply missed.

Electronics-COOLING.com 27
 37
Please Join Us At
SEMI-THERM 37
Virtual Symposium March 22-26, 2021
All Times PDT

www.semi-therm.org

Hear from Experts and Learn about Industry Innovations

Your choice of 4 short courses - Two courses are included with your registration

Liquid Cooling Panel Discussion

Keynote Presentation:
"Opening Opportunities for Thermal Design through Innovations in CFD"

Embedded Tutorial:
"Realistic Thermal Model for Human Skin in Contact with a Wearable Electronic Device"

8 Technical Sessions
• Consumer Electronics • Data Centers
• Automotive/Aerospace/Outdoor • Liquid Cooling
• Two-Phase Cooling I • Measurement Techniques / CFD
• Two-Phase Cooling II • Thermal Interface Materials

THERMI Award Presentation:

"Apollo - The Dawn of Semiconductor Thermal Management"

See the Latest Products and Solutions

A Virtual Exhibition is Accessible Throughout the Symposium
Eight Focused Vendor Workshops are Being Conducted Over Three Days

Attendees Exhibitors
If you are interested in thermal design, We are offering new options for exhibitors to
management or characterization of electronic maximize exposure to attendees, including a
systems and components, plan to attend. limited amount of speaking opportunities.
All programming subject to change
1
SEMI-THERM 37
Keynote Speaker
Tuesday March 23, 9:00 a.m.
Opening Opportunities for Thermal Design through Innovations in CFD

Presenter: Lieven Vervecken

Diabatix
Over the past decades, computational fluid dynamics (CFD) has evolved from a purely research discipline to a
reliable engineering practice. This evolution was driven by continued innovation in multiple domains, ranging
from hardcore mathematics to HPC architecture development. Concurrently with this evolution, CFD has
gradually taken on great importance in the thermal design process which has resulted in a countless number
of products that could not have been realized otherwise. Yet, the potential for discovering new possibilities in
thermal design through innovations in CFD remains enormous. This talk touches on a number of recent and
upcoming innovations in CFD which have the potential to set this in motion.
Lieven Vervecken is co-founder and CEO of Diabatix nv where he is responsible for the general management
development of the company. Diabatix is a Belgian technology scale-up specialized in generative design for
cooling components that helps multinationals all over the world to push the boundaries in thermal design.

Short Courses Monday, March 22

These two-hour classes provide practical, interactive training on a variety of specific skills on topics ranging from
thermal design & modeling to system level validation testing. Attendees can choose one Short Course in the morning
and one in the afternoon. All times PDT.

7:00 a.m. Short Course 1

Let’s Work Together: How Co-Design Leads to Better Solutions in Thermal Management
Presented by: Lauren Boteler, Army Research

7:00 a.m. Short Course 2

Design and Optimization of Heat Sinks
Presented by: Marc Hodes and Georgios Karamanis, Transport Phenomena Technologies, LLC

9:20 a.m. Short Course 3

Micro-Two-Phase Electronics Cooling…Getting it on its Way
Presented by: John R. Thome, J.J. Cooling Innovation

9:20 a.m. Short Course 4

Introduction to Electronics Cooling
Presented by: Patrick Loney, Northrop Grumman Mission Systems
2
SEMI-THERM 37
Liquid Cooling Panel Monday, March 22 11:40 a.m.
Join us Monday afternoon for the Liquid Cooling Panel. This panel will provide a broad perspective on the current usage of liquid
cooling across a range of industries, including data centers, ground vehicles and aerospace. We will also look forward to the near-term
trends according to these experts, then looking further out, with each panelist given the opportunity to describe the technological
advances in liquid cooling that they would most like to see.
Panelists:
Moderator: Suresh Pichai
Tim Shedd, Ph.D. Director, Innovation and Development
Research Division Oak Ridge National Laboratory
Director of R&D, Motivair Corporation Equinix Data Centers
Emre Gurpinar, Ph.D. Debabrata Pal, Ph.D.
R&D Staff, Electrical and Electronics Systems Technical Fellow
Collins Aerospace
Alfonso Ortega Ph.D.
James R. Birle Endowed Chair Professor of Energy Technology, Bapi Surampudi, Ph.D.
Villanova University Director, Villanova site of the NSF Center for Staff Engineer, Electric Powertrain
Energy Smart Electronic Systems Southwest Research Institute

Technical Sessions
Tuesday March 23 Sensitivity Analysis of a Calibrated Data Center Models to
Technical Session - Track 1: Consumer Electronics Minimize the Site Survey Effort
Increased System Performance and Reduced Surface Touch Saurabh Singh, University of Texas, Austin
(Skin) Temperature in Mobile Electronics Utilizing Composites
of Graphite with Ultra-High Spreading Capacity and Insulation
with Ultra-Low Thermal Conductivity Technical Session - Track 5: Liquid Cooling
Mitchell Warren, W. L. Gore & Associates Unified Method to Model Closed-Loop Liquid Cooling
Albert Chan, Cisco Systems, Inc.
An Analysis of Temperature Variation Effect on Response and
Performance of Capacitive Microaccelerometer Inertial Sensors Thermohydraulic Performance of Heat Sink with Sinusoidal
Jacek Nazdrowicz, Lodz University of Technology Microchannels Embedded with Pin-Fins for Liquid Cooling of
Microelectronic Chips
Thermal Acceptability Limits for Wearable Electronic Devices Bobby Mathew, United Arab Emirates University
Mark Andrew Hepokoski, ThermoAnalytics
Technical Session - Track 6: Measurement Techniques/CFD
Technical Session - Track 2: Automotive/Aerospace/Outdoor Cross Correlation Method for Images Alignment: Application to
Effects of Solder Voiding on the Reliability and Thermal 4 Buckets Calculation in Thermoreflectance
Characteristics of Quad Flatpack No-lead (QFN) Components Metayrek Youssef, IFSTTAR
Ross Wilcoxon, Collins Aerospace
Including Electrothermal Effects in Electronics Design with
Characteristics of Practical CTE-Matched Composites for Connected FANTASTIC BCI-ROMs
Electronics Thermal Management: Comparative Study Byron Blackmore, Mentor, A Siemens Business
Dave Saums, DS&A LLC
Thursday March 25
Technical Session - Track 3: Two-Phase Cooling I Technical Session - Track 7: Thermal Interface Materials
Numerical Investigation of Coolants for Chip-embedded CVD Polycrystalline Diamond for Laser Diode Applications
Two-Phase Cooling Firooz Faili, Element Six Technologies
Pritish R Parida, IBM T.J. Watson Research Center
Experimental Investigation of the Impact of Squeezing Process
Numerical Investigation of Thermal Spreading Resistance of on the Microstructure and Performance of Thermal Interface
Vapor Chambers Materials (TIMs)
Farzan Kazemifar, San Jose State University Rajath Kantharaj, Purdue University
Wednesday March 24 Metallic TIMs for Liquid Immersion Cooling and Cryogenic
Technical Session - Track 4: Data Center Temperatures for Quantum Computing
Determination of Cost Savings Using Variable Speed Fans for Dave Saums, DS& A LLC
Cooling Servers
Nicole Okamoto, San Jose State University Technical Session - Track 8: Two-Phase Cooling II
Design and Optimization Array of Micropillar Structures for
Effects of Different Coolants on the Cooling Performance of an Enhanced Evaporative Cooling of High-Powered Electronics
Impingement Microchannel Cold Plate Kidus Guye, Washington State University, St. Louis
Cong Hiep Hoang, Binghamton University - SUNY
Actively Cooled Two-phase Cold Plate for High Heat Flux Electronics
Michael C. Ellis, Advanced Cooling Technologies, Inc.
3
SEMI-THERM 37
THERMI Award Presentation, Wednesday March 24, 9:00 a.m.
Each year, SEMI-THERM honors a person as a Significant Contributor to the field of semiconductor thermal management. The
THERMI award is intended to recognize a recipient’s history of contributions to crucial thermal issues affecting the performance of
semiconductor devices and systems. The 2020 THERMI award is proudly presented to:

Ross Wilcoxon, Ph. D.

Collins Aerospace
Join us Wednesday morning for Dr. Wilcoxon’s presentation.
Apollo – The Dawn of Semiconductor Thermal Management

Hall of Fame Award Presentation, Thursday March 25, 11:50 a.m.

Dr. Dereje Agonafer

University of Texas at Arlington
This year’s recipient is Dr. Dereje Agonafer.
Join us Thursday afternoon for the presentation by this years Hall of Fame award recipient.
Each year the SEMI-THERM Educational Foundation Thermal presents the Hall of Fame Lifetime Achievement Award in recognition of
significant contributions to the field of electronics thermal management.

Embedded Tutorial, Thursday March 25, 9:00 a.m.

Bruce Guenin, Ph.D.

Join us Thursday morning for the Embedded Tutorial, presented by Bruce Guenin, Ph.D.
Realistic Thermal Model for Human Skin in Contact with a Wearable Electronic Device

REGISTER NOW www.semi-therm.org Virtual Symposium March 22-26, 2021

4
Index of
ADVERTISERS

Alpha Novatech, Inc. CPC Worldwide Ellsworth Adhesives LECTRIX

473 Sapena Ct. #12, 1001 Westgate Drive W129 N10825 Washington Drive 1000 Germantown Pike,
Santa Clara, CA 95054 St. Paul, MN 55114 Germantown, WI 53022 Plymouth Meeting, PA 19462

t: +1 (408) 567-8082 t: (651) 645-0091 t: (877) 454-9224 t: (484) 688-0300

e: sales@alphanovatech.com w: www.cpcworldwide.com e: info@ellsworth.com e: info@lectrixgroup.com
w: www.alphanovatech.com page: 18 w: www.ellsworth.com w: www.lectrixgroup.com
page: C2 page: 25 page: C4

Mentor Graphics SEMI-THERM 36 THERMAL LIVE

8005 SW Boeckman Rd 3287 Kifer Road, Online Event
Wilsonville, OR 97070 Santa Clara, CA 95051, USA October 19-20, 2021
March 16th – 20th, 2020
t: (800) 592-2210 t: (484) 688-0300
e: sales_info@mentor.com t: +1 (408) 840-2354 e: info@electronics-cooling.com
w: www.mentor.com e: drael@semi-therm.org w: www.thermal.live
page: C3 w: www.semi-therm.org page: 3
page: 28-31

32 Electronics COOLING | SPRING 2021

Accelerate package thermal
modeling in your design
Generate clean, accurate, CAD based detailed chip
package thermal models in minutes for use 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
 THERE’S ONLY ONE SPOTLIGHT

ARE YOU IN IT?

www.lectrixgroup.com

Problem first. Product last.

PHILADELPHIA | HONG KONG