Everything You Know Is Wrong March 2003

Answers to those Doggone Thermal Design Questions

By Tony Kordyban

Copyright by Tony Kordyban 2003

Hey Thermal Question Dude,

Somebody asked a while back about finishes for heat sinks.  Your answer about black anodizing dealt only with how it changes the radiation properties of the surface.  But what about the convection?  When you paint a heat sink black, or anodize the surface, aren’t you in effect covering up the surface with a blanket?  I would think the paint or anodized metal are thermal insulators.  So you might improve the radiation by making the surface black, but don’t you mess up the convection?

Overheated in Bakersville


Dear Over,

I hate to be picky — well, that’s not true.  If I did not love to be picky, I wouldn’t be in the thermal article writing business.  By adding a layer of insulation to the surface of my heat sink, I would not be messing up the convection heat transfer at all.  I would be messing up the CONDUCTION heat transfer within the body of the heat sink.

A quick review of the basic equation for convection will illustrate my picky point.

Q convection = h A (T air – T surface)


Q convection is the heat carried away by convection
h is the convective heat transfer coefficient, which depends on the air velocity and the temperature gradient in the thermal boundary layer
A is the surface area in contact with the air
T air is the reference air temperature
T surface is the surface temperature

Anodizing a heat sink does create a layer on the surface that could act as a thermal insulator. But when you want something to be an insulator, doesn't it always need to be pretty thick?

No place in the equation does it ask for the conductivity of the solid.  The solid could be made completely out of Jello or anodized aluminum, and the convective heat transfer behavior would be exactly the same.

Well, OK, not exactly the same.  Your gut tells you that an aluminum heat sink has to work better than one made entirely of Jello, so the conductivity has to matter somehow.  It does matter, indirectly, by changing the surface temperature term of the equation.  If the heat sink were made of polystyrene instead of aluminum, then T surface would be different.

Can we estimate how much an insulating blanket would change the temperature of a heat sink? Easy.  It depends on the heat flux (the power per unit area) conducting through the heat sink, the conductivity of the blanket, and the thickness of the blanket.  Then we just use the basic equation for 1-dimensional heat conduction, which you should know by heart now, but I will provide anyway.

Q conduction = (k A )/t  (T aluminum – T surface)


Q conduction is the heat conducted through the thickness of the anodized layer
k is the thermal conductivity of the anodized layer
A is the cross-sectional area of the heat sink
t is the thickness of the anodized layer

The heat flux is just Q conduction / A, so if we rearrange the equation a little bit, we can see that

(T aluminum – T surface) =  heat flux x (t / k)

I have not been able to discover the thermal conductivity of the anodized layer.  (It might even be a pretty good conductor for all I know!)  It is an oxide of aluminum that is somewhat porous, with the pores sometimes filled with organic material, depending on the anodizing process.  Let’s be very pessimistic and assume that it has conductivity similar to an insulator like plastic, say k = 0.1 W/mK.

One thing I was able to find out by surfing the web,  was that the layer of anodized material is very thin, about 2 x 10 -5 meter (about 0.0007 inches).  You can specify it to be thicker, but it won’t make your radiation properties any better, so why would you?

The only other number we need is the heat flux.  Of course, that depends entirely on your application.  In my fairly ordinary heat sinks for digital electronics I encounter  fluxes in the range of  100 to 1,000 W/m2.  You can figure out yours by dividing the power dissipation of your component by the surface area of your heat sink.

Let’s take my heat sink dissipating a heat flux of 1,000 W/m2.  Given an anodized layer thickness of 0.00002 m and conductivity of 0.1 W/mK, I get a temperature difference across the anodized layer of 0.2 degrees C.  I consider that negligible.

But don’t take away from this that the thermal insulating effect of anodizing is always negligible.  For me, and for most electronic heat sink applications, it probably is.  But not always.  If your heat sink has a thicker coating,  of paint, for example, or a very high heat flux, such as from a motor control power supply, then maybe the temperature difference would be important.  This simple analysis is a tool you can use to decide if it will be important.  I live by that old thermal aphorism, “Give an engineer an answer, and you feed his/her curiosity for a day.  Give an engineer a tool for generating answers, and he/she stops nagging you endlessly.”

If you find out you DO have to worry about the anodizing, then you’d really better start to worry about all the dirt that is going to collect on that heat sink over years of operation.  That blanket of grease and dust and cat hair could easily get to be a lot thicker than the anodized layer.


To Mr. Kordyban,

What about all these thermal conferences and symposiums?  I subscribed to a couple trade magazines on electronic packaging and before I could start reading them, my mailbox was flooded with brochures for various meetings called SemiTherm and I-Therm and InterPack and IMAPS, not to mention  the CoolingZone conference.  Like lots of packaging engineers, I am responsible for sheet metal, injection molded plastic parts, stress analysis, cable harnesses, EMI, safety, shock and vibration, and even shipping containers.  I do thermal analysis only part time. 

Do these conferences have anything for engineers like me?  Some of the topics sound interesting, but some seem really exotic.  Is it just a bunch of professors and PhD candidates showing off for each other?  Or can I learn something I can use on the job right away? 

Metal-Bender from Pittsburgh

Dear Bender,

These things are always a mixed bag.

My biggest gripe with technical conferences of every stripe is that there is often a disconnect between the label on the package and the product inside.  They should have warning on every paper like “Some settling of contents may occur during shipping.”  The abstracts and the bodies of the presented papers sometimes have little in common.  For instance,  you spot the title of a paper, “A simple method of determining the thermal conductivity of a printed circuit board,” which sounds like it could be very useful.  So you travel across four time zones to hear the talk, and it ends up being the derivation of an equation for heat conduction from a single point source through a composite plate.  It is almost like a printed circuit board, except the equation is only valid for circular sources on infinitely large plates, or some such extremely limited piece of the puzzle.  It’s not what you thought it would be when you read the abstract.

There are academic papers from universities, but there are also talks by people from industry who are working on real products, just like you.  You will find some duds and some winners in both categories.  If you attend conferences on a regular basis, what you will come to realize is that in thermal engineering, there are very, very few breaththroughs to report on.  The technology is mature, and progress is made in baby steps.  For a first-time attendee, that might be difficult to listen to.  It is like trying to learn all about the history and culture of a foreign country by reading random stories from their daily newspapers.

So maybe you shouldn’t count on gaining a lot of useful information from the presented papers at a conference.  There are still three things that might pay off for you.

1.  Short courses.  Almost every conference offers 1 or 2-day short courses.  These can be very useful, especially if you have not had a lot of formal education in thermal engineering and have had to pick it up as you go along.

2.  Vendor exhibits.  Here is your chance to pick up catalogs, samples, and nifty pens and flashlights.  Plus you can quiz the vendors for hours without feeling guilty about not giving them a purchase order at the end of the day.  This is a great opportunity to see lots of suppliers side-by-side, face-to-face, especially if your office is off the beaten track.  And you get to hear the questions your competitors ask them, too.

3.  Which leads into perhaps the most valuable feature of any conference:  you get to sit at big tables and eat lunch with a bunch of strangers.  After about 20 awkward minutes, you will begin to swap stories, maybe about your plane trip at first, but eventually you will start sharing stories about your problems at work.  And if you are lucky, or persistent, you will find somebody who has already solved a problem that has been bugging you for months.  I think there is more productive exchange of technical information among the attendees in the halls and lobbies of the hotel than there is in the meeting rooms.

So I recommend attending a conference once a year or so, if you can swing it.  If you happen to nod off during a presentation, no big loss.  Just make sure you’re awake during the coffee breaks!

I can’t really endorse one conference over another, since I haven’t attended all of them to compare fairly.  The only hint I can give you about guaranteed quality is if you see my name listed as one of the presenters.  (Judge for yourself whether that is a sign of high or low quality.)


Isn’t Everything He Knows Wrong, Too?

The straight dope on Tony Kordyban

Tony Kordyban has been an engineer in the field of electronics cooling for different telecom and power supply companies (who can keep track when they change names so frequently?) for the last twenty years.  Maybe that doesn’t make him an expert in heat transfer theory, but it has certainly gained him a lot of experience in the ways NOT to cool electronics.  He does have some book-learnin’, with a BS in Mechanical Engineering from the University of Detroit (motto:Detroit— no place for wimps) and a Masters in Mechanical Engineering from Stanford (motto: shouldn’t Nobels count more than Rose Bowls?)

In those twenty years Tony has come to the conclusion that a lot of the common practices of electronics cooling are full of baloney.  He has run into so much nonsense in the field that he has found it easier to just assume “everything you know is wrong” (from the comedy album by Firesign Theatre), and to question everything against the basic principles of heat transfer theory.

Tony has been collecting case studies of the wrong way to cool electronics, using them to educate the cooling masses, applying humor as the sugar to help the medicine go down.  These have been published recently by the ASME Press in a book called, “Hot Air Rises and Heat Sinks:  Everything You Know About Cooling Electronics Is Wrong.”  It is available direct from ASME Press at 1-800-843-2763 or at their web site at http://www.asme.org/pubs/asmepress,  Order Number 800741.