answers to those doggone thermal design questions
by tony kordyban

hi tony,

i thought your treatment of the fan laws was good except for the section about how flow rate (cfm) changes with density. actually, the volume flow rate (cfm) doesn't change with fan inlet density, but the air mass flow (lb/minute or kg/sec type units) changes. the air mass flow is proportional to density (assuming incompressible flow, constant density, and all that stuff.)

i hadn't seen the information about noise before and that looks like it could be good guidance because noise probably varies with fan configuration and design specifics. please keep up the good work.

doug werner, pe 
douglas engineering 

(p.s. sorry, i couldn't think up a clever pseudonym. i thought having pe after my name was amusing enough.)

dear doug,

oops, you are right. you spotted the big boo-boo in the fan laws table in last month's coolingzone newsletter. shame on me for not deriving them all from first principles, as any true engineer would, instead of just copying the equations out of a book. i actually did know that axial fans, at a given rpm and a given back pressure, deliver a constant volumetric output, regardless of the density. those rotating blades are sweeping out a fixed volume on each rotation. at high altitude, there are just fewer molecules of air in each of those sweeps. i was paying all my attention to the rpm side of the table and let this error slip by.

as you point out, doug, the mass flow rate goes down with reduced density. this is pretty important, because the heat transfer to the air is related to the mass flow rate, not the volume flow rate. at constant altitude we tend to think of volume flow rate and mass flow rate as being the same thing with different units, but it is not true at all.

so here is the fan laws table again, with the correction inserted for the volume flow/density relationship:

beware that i copied this mistake (and all the correct information) from an old catalog. the same table appears on the corresponding website. just a couple of paragraphs away is a statement that the volumetric flow rate of a fan does not change with density, in their explanation of how to estimate the operating point at high altitude. i'm not pointing that out to pass the buck, or to pick on that particular fan vendor. i just want you to know how i made my mistake, so maybe you don't do it yourself too often. even a generally reliable source can have erroneous information. just because it's in print or on the web doesn't make it so. it has to be checked out.

so shame on me, but shame on you loyal readers (except doug), too, who didn't notice this obvious goof and call me on it.

if you do spot something questionable (besides my writing style), keep me honest by letting me know. maybe i'll come up with a valuable prize or two for the best goof spotter.

dear temperature guy,

thanks for all your articles. they are really very useful.

i have a problem in temperature measurement. i am trying to measure the effectiveness of a particular thermal interface material. i intend to put it between two aluminum blocks, then measure the temperature difference between them while heat is flowing through the interface.

i drilled a 1/8th inch diameter blind hole in each block, the bottom of the hole being exactly at the centre. i want to messier the temperature of each block by putting a rigid thermocouple probe into each hole. the problem that i have is i am not sure if the probe touches the aluminum inside the hole. i do not know anything about thermal pastes, and if i apply them will i be able to take the probe off? i mean are they very sticky? can the thermal paste withstand a temperature of, say, 150 degrees c? i am planning to take the entire assembly (along with the thermal interface material) to a temperature of 150 degrees c and age it for about 1 week and again measure the temperature difference. could you please provide me any details about thermal paste?

e. poxy of sticky wicket 

dear sticky,

the experiment you are embarking on is very ambitious, and i don't envy you the problems you will discover along the way. thermal paste will seem like a tiny question when get to the mysteries of how to separate the contact resistance from the thermal resistance of the interface material itself, or how much heat is passing through the interface, and how much is leaking out in all other directions by conduction, convection and radiation. but you didn't ask how to do the experiment. you want to know the properties of thermal paste.

a quick search of the web and coolingzone's own supplier directory immediately turned up two thermal paste data sheets: aos heat sink compound, and thermoset tc-222. (this is not an endorsement of these particular products. they were just the first things i came across in my search. please check with your favorite thermal goop supplier.) the data sheets claim they work between -40 and 200 degrees c, and stay paste-like under those conditions for years. (as we found in the previous question, check out those claims before you bet the farm on them.) this seems to be typical for thermal greases and pastes, so it shouldn't be too hard to find one that does what you want. so you don't have to worry about the paste turning into a solid glue-like substance at high temperature, trapping your thermocouple in the hole.

that could be the end of the story. but i have a question for you: why don't you want your thermal paste to turn into a solid glue-like substance, trapping your thermocouple in the hole? i don't think thermal paste is right for your application. and i don't think a rigid thermocouple probe is right for your application, either.

thermal paste (or grease, as some know it), is meant to be used in a joint between two flat plates. the grease layer is intended to be extremely thin, only there to fill in the microscopic imperfections in the surfaces where the two plates don't touch each other directly. after all, the conductivity of even the best grease is only about 1 watt/meter/c. that is pretty low compared to the conductivity of aluminum (180 watt/meter/c). the grease is only a good conductor compared to the air it replaces. grease is used precisely because when you squeeze the plates together, it moves out of the way to allow the plates to touch each directly wherever they can (unlike a solid gap pad).

when you mention a rigid thermocouple probe, i imagine a pair of wires encased in a metal tube, with either a flat or hemispherical end at the place where the temperature is supposed to be sensed. at least that is the kind of pre-assembled probe i see in catalogs. your drilled hole probably does not have a flat bottom (or a spherical one), unless you had it specially machined that way. the bottom of a drilled hole is cone-shaped, like the point of the drill. so it is unlikely that you can get good metal-to-metal contact between the end of the probe and the bottom of the hole. filling the hole with grease is an improvement, but not much. the grease also won't act as an adhesive, so there is nothing holding your probe in place at the bottom of the hole, even if you go make good contact. it could move around during your experiment, giving you all kinds of errors.

here is my recommendation for measuring the bottom of a drilled hole in an aluminum block with a thermocouple. it is based on gabbing with some colleagues, and a book called "temperature measurement in engineering, volume 1", by h. dean baker, written about 50 years ago. (it is still in print and available from omega engineering. it has some very cool illustrations, such as how to install a thermocouple in a spark plug electrode. you can consider this and endorsement, because i have actually read this book.) i have updated the method a little to use materials i have at least heard of. 

make your own thermocouples from 30 gage wire. i believe type t and type j are good up to 150 degrees c, depending on the insulation. make a junction by stripping the wires, then twisting them together all the way back to the insulation. solder (or weld, as you prefer), and clip off all the extra twists to leave a small bead right next to the insulation. clean out the drilled hole with a solvent to get the metal chips out.

i would want to press the thermocouple bead into the bottom of the hole, and then glue the whole thing in place. to get some pressure on the bead, form the wire into a spring that will fit into the hole. shove the assembly down into the hole until you feel it bottom out. while maintaining pressure on the wire, pour glue into the hole to lock it into position. you can use thermal epoxy, silver-filled cement, or whatever you want (that will stand up to 150 degrees c.) the thermal conductivity is not important, if you have made contact between the bead and the aluminum first.

i know you won't be able to pull the probe out after that. but why do you want to? it is easy and cheap to make more probes once you buy a spool of thermocouple wire. just snip off the end of the wire that is stuck in the block when you are done. if you want to re-use the hole, just drill out the glue.

this may not guarantee that you have perfect contact with the aluminum, but by gluing your probe in place, you will at least have repeatable results, because it won't move around during the experiment.

the reason i want you to make the bead of the thermocouple so close to the insulation is so that you don't have long pieces of exposed wire that could short together when you jam the assembly into the hole. the temperature you read is the place in the thermocouple circuit where the wires separate last.

thermocouple attachment is an art in itself, many decades old, and i am not a virtuoso at it myself. and i don't mind a bit if you disagree with me. i'd like to learn some of your trade secrets, too.

about 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 20 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 b.s. in mechanical engineering from the university of detroit and a master’s in mechanical engineering from stanford. in those 20 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 at https://www.amazon.com/hot-air-rises-heat-sinks/dp/0791800741. this advice column is an extension of that educational effort.