Thermal Gap Pad for shock mounting? 1

[gjacc]

My basic problem is to ruggedize (ie. shock mount) a heat-generating device (35 Watts) inside a sealed enclosure.
I am considering using thermal gap pad material (Chomerics or 3M), but it is still very stiff...does anyone have experience using this material? are there alternative materials/ideas for a thermally conductive and flexible(compliant) solution?

Thanks in advance,
Greg

 

[MikaelFranzen]

Hey,
Don't really have a lot of experience with that, but this site may give you some useful info:

http://www.stockwell.com/pages/materials_thermal.php

Best of luck,
Mikael

 

[ko99]

Thermal gap pads are fine for low power density (<<10W/in2) but have pretty poor thermal performance compared to, say, thermal grease.

What's the contact area of the device-to-pad? For example, assuming it's one sq.inch and using Mikael's link, a 0.10" thick Berquist 1500 pad has a resistance of about 2.5 C-in2/W. This results a theoretical loss of 70C through the pad alone. In practice, not all 35W will go through the pad, especially if it's that resistant, so the deltaT will be less, but still very high.

Note that 0.10" is the thickness after compressing - you may need a thicker pad to cover the tolerance range of your assembly. And finally, the compressed pad may not be as compliant as you need for structural isolation.

Ideally, it's best to rigidly mount the device to the enclosure, then structurally isolate the enclosure. If you can't do this, there are alternatives but I don't know enough about your design to venture a suggestion.

ko (

http://www.ecooling.biz)

 

[gjacc]

Thanks for your responses. More specifics on my problem:
The 35W heat is dissipating from a cylinder 2.5"dia x 5"long. The sealed enclosure is 6"wide x 10"long x 12"tall and contains other items inside. The internals need to be shock mounted internally. No external shock mounting is possible. The internal temp should not exceed 50degC with room temp. outside, other design specs are TBD.

-One idea was to have fins inside and outside of the housing w/ fans to improve heat transfer, but I've found this to be inadequate (internal temp is too high).

-What about leaf springs or coil springs to conduct to the outside housing and provide shock mounting?

Thanks again,
Greg

 

[ko99]

Greg, if the heat is ~uniform over the surface of the cylinder, the power density is much less than my guess, so a gap pad might work. Using the example pad on one end of the cylinder (5in2) the deltaT is 14C. Note you'll still have to determine if the example pad is sufficient for your tolerance and isolation needs.

Another possibility is to rigidly mount the heat source to a thick aluminum or copper block or bracket to spread the heat over a larger area. Then attach the bracket to the enclosure with an isolating-thermal pad between them. The larger contact area may allow you more leeway in selecting a softer pad.

Regarding the springs, you can estimate the conductive thermal resistance yourself with R = L/KA

R = resistance C/W
L = length heat travels along the conductive path, m
K = conductivity of spring material, W/mK
A = cross-sectional area parallel to heat travel, m2

To the above, add the contact resistance from the heat source to the spring, and the spring to the enclosure -- not a very good method of transferring 35W!




ko (

http://www.ecooling.biz)

 

[keepinitcool]

Is this heat source fragile? And does it need to be insulated electrically from other parts? Most of the thermal pads you are talking about are designed to provide electrical insulation between components and some thermal conduction. Because of that, the thermal conduction is typically limited. If you don't need electrically insulating material then you should look into silicon sheet (impregnated with material for better thermal conductivity). This is what some of the gap pads are that you are talking about are, but some of these sheets do not insulate electrically and are thicker.

Bergquist makes these pads also, so you might check them out. Although again, their parts are typically electrically insulating.

 

[gjacc]

Thanks for all your feedback.

After further research I've concluded that a gap pad won't provide adequate shock/vibration protection (yes, it's fragile).
A flexible copper strap(s) with foam is the direction I'm will likely take this design.
Regards to all,
Greg