Vertical Heat Sink Mounting

[Gyde]

Hello All,

This is my first post to this forum so be kind if I miss convention anywhere. I am having some difficulty devising a cooling method for a project I'm working on and am looking for some tips. I am designing metal shielding for a detector that may be running for anywhere up to days at a time. This detector when running for so long has a tendency to begin heating up and that is further compounded by being completely surrounded by metal and plastic. The detector itself looks like a rather thin all aluminum briefcase (14.5"x16.5") and I need to cool the electronics section (~5.5"x15")on the back side. So my plan was to attach a heat sink to the back of the briefcase and run low profile fans within the shield but I can't actually add mounting holes to the flat aluminum backing. The added wrinkle is that this detector will be vertically oriented during operation meaning the heatsink would need to be vertically mounted as well but again w/o mounting hardware.

Is there some sort of thermal adhesive that is strong enough to hold such a large heatsink to a flat surface vertically while still providing good heat transfer? Knowing the surface area I need to cool and the other restrictions I outlined is there another method I should investigate (e.g. water cooling?)

Cheers in advance for any advice you all have.

Gyde

 

[EdStainless]

There are thermally conductive adhesives, but they aren't very good at either part of their job based on my experience.
Can you make a bracket to clamp the heat sink to the case?
Do you have ambient air available for cooling?
A flat plate with heat pipes in it that lead to a heat sink with fans that is positioned in a better place might be an option.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[IRstuff]

What is the power requirement?

While you might not be able to put holes in the box, do you have access to all sides of it? You could strap or clamp the heat sink. However, if your box was not designed to conductively transfer heat, you might need to use a thermally conductive epoxy anyway. Depending on the heat sink design, there might not need to be a void filler that also serves as the adhesive.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[Gyde]

Hey Ed,

That's been my experience as well with thermal adhesives so I guess there might not be a magic bullet glue for me. For reference here is a pdf of the detector actually

https://www.varian.com/sites/default/files/resource_attachments/ps2530he_0.pdf

. There is a large square area where it detects that I cannot block on the front or the back. Potentially I could have something that wraps around and interacts with the screws on the front of the device but I'm not sure how to get good clamping contact across the heatsink. Especially considering most large surface area sinks I am finding online seem to recommend top mounting with adhesive which isn't really my use case. Do you have an example flat plate w/ heat pipes you could link?

 

[IRstuff]

Looks like there are plenty of existing screws; why not replace them with longer ones and use them to hold the heat sink?

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[chicopee]

One aspect that I am somewhat aware is laser beam welding as it produces a narrow weld with a small HAZ. This type of welding is used on thin section on dissimilar metals. Tacking the heat sink to the thin wall detector using laser beam welding may be your answer.

 

[LittleInch]

Why not look at this stuff.

http://m.globalsources.com/gsol/I/Air-conditioner/p/sm/1148688676.htm

Either attach with tape or glue then blow air or water through it?


Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MintJulep]

Vacuum brazing.

 

[chicopee]

Vacuum brazing mentioned by MintJulep is also a good option.

 

[lilliput1]

Are you sure you need cooling? The unit is only 16 watt and need to be either potable or fixed. if fixed it need to fit within a slot in the equipment. See unit operating manual.

 

[Gyde]

Hey Lilliput,

We ran some worst case scenario tests where the unit ran for 24 hours within a foam enclosure and saw temperatures up in the mid to high 50s Celsius. The operating range for this device is only up to 40C and when running within ambient conditions and no enclosure regularly posts 31-32C temp measurements. So cooling is in some ways a better safe than sorry operation since this device is quite expensive and may be used for days at a time within metal shielding.

In reference to some of the other suggestions, vacuum brazing is an excellent operation but management is nervous about the operation and I can't blame them considering the cost of this device. These parallel flow tubes are really interesting and I've been thinking of good ways to circulate air through them considering the tight space requirements I have within the shielding (everything must be self-contained).


Thanks for all the help so far though folks.

 

[lilliput1]

Putting the unit in a foam enclosure? Is this how the unit will be used? If so then your idea of adding a fan will make it worst. The fan will generate heat that would be trapped in the enclosure. There must be free air around the unit to carry the heat away. Consider providing fins on the shielding and using thermal paste so heat would be conducted to the shielding and dissipate heat to the air by the added fin surfaces. Mechanically secure the shielding to the unit.

 

[Gyde]

Hey lilliput,

The unit was put in the foam enclosure only to act as a worst case heat transfer scenario so we could see how hot it might get. In practice it will be used within the metal shielding that will have about 45mm of depth behind it.

 

[lilliput1]

That is not a true worst case scenario then. It is a case where nothing will work because you trapped all the heat in.

 

[Gyde]

Something I have been thinking about recently are thermoelectric coolers. Does anyone have experience with these?