would you trust epoxy/ steel filled epoxy aka araldite?

[LopeDogg]

anyone made stuff with copper and epoxy?
for certain things it would be a lot easier to make than using solder...

[Moparchris]

i made a gpu block with a copper base and i epoxyed the plastic top onto it and it worked for about 6 months but then i took the block out cause it dosnt fit on my new video card and tryed to take it apart cause it was prettymuch rubbish and it was so hard to get the thing apart with a screwdriver i broke the plastic top instead of breaking it at the join... its pretty damn strong stuff

[LopeDogg]

*g*_______________

[JFettig]

I would use JB weld over epoxy any day for this kind of stuff, its not clear but holds better than regular epoxy as far as I have experienced.

Jon

[BillA]

araldite is a type of epoxy with good warm water resistance, I've not seen it filled but I suppose it could be
I do not know the type of epoxy used in JB Weld

in any case the surface prep is key, solvent clean and sand lightly; the bonding is mechanical

[LopeDogg]

the nice thing about araldite is it can fill a gap bigger than solder can...

im basically building this:
http://www.overclockers.com/tips857/

i really think its the ultimate passive radiator.

the reason i think this "flat box" radiator is so great, is it has minimum thermal resistance from water to air, it contains a minimal amount of water, uses a minimal amount of copper, keeps the weight down...

i read a heatsink design page once, and they said 1mm spacing between fins creates the most airflow between them (passively).
my plan is to build 8 of these, spacing them 1mm apart...
my biggest concern is getting both sides of the box to stick to the flow directors... doing this with solder wont be easy.

should be a breeze with araldite.

ive already started on one using solder... the flow directors are 1.5mm copper wire (low surface area, chances of the lid sticking to it are minimal)
but its just a lame experiment till i can buy more copper.

last time i tried working with thin copper plates(0.7mm) they warped like crazy, and became somewhat useless
now im working with 0.56mm copper. as i heated it up, it warped... and i decided to heat the whole plate as evenly as i could...
i then let it cool down naturally
low and behold it was perfectly flat afterwards...

i used to throw water on my copper stuff after i finished soldering it (impatient)
and i also had stuff clamped... i also didnt heat evenly... maybe its a combination of these three things, or only one of them... anyone know?

(the plan is to build the radiator with araldite when i can buy more copper)

[|kbn|]

I think youd be better using copper pipes. They are cheap, and easy to do what you want. It wouldnt look bad if pollished either.
I think maybe 10 meters of pipe woud be plenty, that would cost you about £10.


Read this: http://www.over-clock.com/ivb/index.php?showtopic=4135

[redleader]

Thats two sheets of copper ? Why not just use pipe? Its cheap and pretty easy to solder with a 15 dollar torch.

[LopeDogg]

in a passive radiator, surface area isnt the only thing you need, you also need to harness the power of convection. to do that best, you need vertical surfaces 1mm-3mm apart for best results.
the radiator i want to build is gonna be pretty small, it will attatch to the non window side of my case. ill get maximum surface area, maximum air convection, minimum thermal resistance from water to air... maximum passive cooling in smallest space...
well, thats the theory...

pity the copper place is closed

[lolito_fr]

1mm-3mm apart

Not for a passive radiator. Check this:
http://www.coolingzone.com/Content/D...las/fcalc3.htm

[LopeDogg]

thanks for your posts guys btw...

hmm, i had a look at the calculator...
keeping my fins the same width, and number, while adjusting the space between them, and obviously the total width of the "heatsink" i found that the bigger the space between the fins, the greater the heat transfer number became.

the calculator suggests infinite space between the fins is the way to go...
(while keeping the number of fins the same, and accomodating width of heatsink)
i dont really know what the problem is, because i dont understand how its calculating the optimum spacing... but i know this: the formula is calculating fin spacing (and count) based on a fixed width heatsink... in my problem the heatsink can be as wide as i want, and i can practically fit in about 8 fins. so its not suitable.

to blab a bit more...:
if the calculator is applied to my problem, the conclusion you would draw from its results dont make sense, because when you have vertical plates in close proximity, you create convection currents, as air touches some hot metal, it gets hotter, and starts rising, now when you have conductors in close proximity, the rising hot air, gets hotter as it continues to move past the hot metal, so its speed increases as it rises, it also pushes other air creating a nice current...
well, thats the way i understand it...

erm...
i think this calculator is more relevant to my problem.
it gives me hope that what i want to do will work.
http://www.coolingzone.com/Content/D...ot/vp-isot.htm

( i basically just looked at their default values and results to use as a guideline)

ive built a convection test radiator before... it was a piece of crap(it JUST did the job), but its given me some experience, so now ive got an idea of what to expect.

the size of my plates aka flat boxes is known, the unknown is the optimum spacing, which i'll find out with experimentation after i build it *g*

[lolito_fr]

With regards to the page I linked to:
1. Looks like the calculated Q is only good if the fin spacing S is chosen to be the same as the calculated Sopt...
(based on the formulas here: http://www.coolingzone.com/Content/D...Problem_2.html )

2. Infinite fin spacing = infinitely large rad = infinite dissipation
If you can accommodate that, go for it!!

I'm not too sure what your rad design is - if it's anything like the one on the OCers article you pointed to, then this calculator is useless as there are no 'fins'. Unless you're intending to add fins to it (a good idea) ?

... or when you say "spacing" are you talking about the thickness of the box?
Cos this will affect the water side, not the air side...

[LopeDogg]

i intend to basicaly make a few of what that guy built, 2mm wide approximately, make a few of them and put them in paralell.

when i talked about the spacing, i meant the gap between each radiator plate...

[superart]

I think I see hat your thinking of doing.

Make a "heatsink" that is hollow inside and run water through it.

Am I understanding you right?

[LopeDogg]

pretty much

[Les]

Originally Posted by lolito_fr

Not for a passive radiator. Check this:
http://www.coolingzone.com/Content/D...las/fcalc3.htm

Have played with and initially found to be a distraction.
However a 120x120x50mm heat-exchanger with a temp differential of ~25c gives a fin spacing ~6.5 mm and h(not Q) ~ 11W/m^2*K.
OK 25c is too large for dT(fin-air) but not too large for dT(air out -air in) IF the heat-exchanger is positioned beneath a household radiator(25++c on mine(thermometer only up to 50c).


Edit1 : Corrected "optimum fin spacing" to 6.5mm(from 4mm)
Edit2 : Corrected h to 11 from 30

[lolito_fr]

Was looking at 10-15mm spacing for a 400*400mm stacked plate rad, with an optimistic fin-air dT of 15°C in my earlier ramblings: http://forums.procooling.com/vbb/showthread.php?t=8704

"IF the heat-exchanger is positioned beneath a household radiator"
nifty idea - but what happens in summer!?

[Jonas]

not to piss in your cheerios here but, there are way to many 'i believe' and 'should' in that article. i would say that this has never been done, at all. If not then the author of that article should brush up on definitive statements such as: 'this is what happened for ME'.

if it had actually been done then where are my RL photos? Easy enough to tell others your theory on something but if you are not going to back it up woth a trial on YOUR expensive toy, then frankly bugger off.

[Butcher]

On the original question, I used epoxy to fill gaps in a water block I couldn't solder (the gaps were too large to get solder in). It worked just fine - no leaks at all, even after I did a bit of resoldering on it. I was using solder for mechanical strenth on that, with the epoxy just as filler/sealant.

[LopeDogg]

thanks for the input guys...
ive had quite a nice breakthrough, new design, etc.
simpler better, doesnt need araldite.
but ill probably use it at some point for something, cos its good stuff.

ill post pics when its done.

happy new year.