Electric/Thermal resistance

Probably not. Your second-best bet would be to lash up a thermal test bench, and actually measure it; your best bet would be to get the finger stock completely out of the thermal path.

Good Luck! Rich

One method would be to estimate the contact area as (normal force)/(yield strength). (BeCu is pretty hard, so this will be a small number--indium is much better.)

The path length of the small contact spots would be about equal to the surface finish of the material. That ought to get you within a factor of 2 or 3, which is probably OK as long as it isn't a significant number. If it is, you know what to fix. ;)

The interfacial thermal resistance between metals isn't usually a big issue, but it's very poorly understood--it goes almost flat at low temperatures, whereas iirc the models predict that it should go as some power of T.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

h

ch,

Thanks Rich, You win! I just was reeading G.White's book "Experimental Techniques in Low Temperature Physics". I quote from section 4.7 (4th edition) "The measured thermal conductance is alwasy greater than that calculated by the Wiedemann-Franz-Lorentz law from the electrical conductance. This 'excess' was often a factor of 100 more"......"The conclusion is that the majority of heat is carried across the interface by thermal waves rather than electrons."

A thermal test is needed.. sigh. Measuring resistance is so much easier.

George H.

uch

ll

t

Thanks Phil, I was reading a paper by R. Berman "Some experiments on Thermal Contact at low temperatures" Jour. App. Phys. 27, page 318 (April 1956) Where he finds that the thermal conductance is proportional to the applied pressure and 'mostly' independent of the area. He finds a T^2 behavior at low temperatures, but mostly T independent above 77K (which is all I care about at the moment.)

I wonder how much force I get from one finger of BeCu?

George H.

Use thermally and electrically conductive Silver filled epoxy, like the ALL the chip bonding boys do.

Cold version of SMT. Only need 80C for 4 hours.

EPOTEK H20E

(Such

will

ll

r

net

I can tell you in chip design the thermal conductivity is highly correlated to the area of the bonding wire, so I'm a bit dubious about the area not being a factor.

Such

l

bench,

k

he

Thanks Glimmerman, These are pressed contacts that can be moved.

George H.

=A0(Such

y will

mall

e

tor

me

l.net

These were for pressed contacts and not anything that is bonded, soldered or otherwise electrically joined. The forces involved were pretty large. 50 lbs to 250 lbs (and up to 4700 lbs at 4.2 K!) Here's Berman's data for Cu rod on a Cu rod at 77 K Load (Lbs.) Conductance(W/K)

50 0.16 100 0.325 150 0.49 200 0.66

He changed the area of the contact by a factor of four, but that didn't change the conductance by very much. (There's only data shown for this at 4.2K and it's bit hard to make out the different data points on the pdf file I have.... why do people always make the graphs so small?)

I wonder how well the numbers scale to low forces? (a line drawn through the data goes nicely through zero.)

George H.

No surprise.

The contact occurs by smushing the surfaces together until the local stress is less than the yield stress, so the actual contact area is pretty close to (normal force)/(yield stress). That doesn't depend at all on the projected area of the pieces.

The reason this is so is that metal has roughly 4000 times the thermal conductivity of air, not counting the interfacial contribution at the air-metal boundary.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

bench,

Then leave the epoxy uncured. :-)

You'd have to clean and redress every so often, as it eventually polymerizes.

(Such

small

he

ctor

g

ome

al.net

on

t- Hide quoted text -

Yes, Berman said essentially the same thing in the paper. I need to follow up on the citations to thsi article. I've got this 'crazy' idea that we could use this as a varible thermal resistance. I can certainly make a little flexure type vice where I can apply 100lbs to some piece of copper. that would give me 3 degree/watt, then 10 lbs would be 30 degrees/ watt. etc. It looks perfect for making a variable thermal resistance.

George H.

Such

will

st bench,

tock

s e

Yuck, that sounds like a gooey(sp) mess. And it's not at all clear that it would improve the thermal conductivity.

George H.