A fairly stupid question about solder and copper

[JonJ678]

I have one of these

.
Ignoring a steel plate in the centre which I've removed anyway, you can clearly see that it is sealed with O rings and held together with four M4 steel screws. Now, were I to buy this

and replace the clear plastic with it, I'd have a lump of copper sealed by 2mm O rings.

What I am wondering is if I can remove the O rings, put a ring of solder in their place (hopefully I can find 2mm OD solder somewhere), screw it together and heat it up to around 220 C or so, melting the solder, without permanently damaging the copper?

Points of concern are different coefficients of expansion of copper and steel (stripped threads/brass screws?), if the copper is lacquered this lacquer isn't going to cope, and that the copper itself might oxidise significantly at this temperature.

The hope, obviously enough, is that I'll end up with a waterproof copper waterblock, with no O rings, which I can if need be take apart by heating it up enough to melt the solder. Essentially I don't really trust O rings to deal with cold temperatures and am trying to solve this (admittedly imagined) issue.

Cheers

 

[surge]

What I am wondering is if I can remove the O rings, put a ring of solder in their place (hopefully I can find 2mm OD solder somewhere), screw it together and heat it up to around 220 C or so, melting the solder, without permanently damaging the copper?

Points of concern are different coefficients of expansion of copper and steel (stripped threads/brass screws?), if the copper is lacquered this lacquer isn't going to cope, and that the copper itself might oxidise significantly at this temperature.

The hope, obviously enough, is that I'll end up with a waterproof copper waterblock, with no O rings, which I can if need be take apart by heating it up enough to melt the solder. Essentially I don't really trust O rings to deal with cold temperatures and am trying to solve this (admittedly imagined) issue.

Cheers

How cold do you want to go? Silicone rubber O-rings are good down to about -60 C. Info here:
http://supaseal.co.uk/o-ringmaterials.htm

They say Buna-N, which is most probably what your existing O-ring is made of, is good down to -40, although I'm not sure I would go that low with it.

If you go the solder route, I don't think the steel screws should be a problem, but the block will almost certainly distort and require lapping afterwards. You might also want to worry about solder creeping by capillary action into the innards of the block...

 

[JonJ678]

I'd be lucky indeed to see -60, probably not -40. Just starting another round of drafts, worst case scenario is whatever a peltier heat exchanger achieves with the hot side around 0 degrees under a 200W load.

Capillary action doesn't worry me too much, I'd break it apart after the first join to have a look, and omit the inner O ring if it caused problems.

I suppose I can work out roughly how much the copper would expand, compare this to shear strength of the threads and estimate the shear forces acting in the copper. Doesn't sound like much fun though, I think I'll email ek to ask about their O rings.

Cheers man

edit: you mentioned in the Pneumonic's thread that you have more information on pelts, I would indeed be interested. Feel bad trashing his thread further though. Where should I look?

 

[surge]

edit: you mentioned in the Pneumonic's thread that you have more information on pelts, I would indeed be interested. Feel bad trashing his thread further though. Where should I look?

If you have'nt seen it already there's a guy on Overclock.net called Ultrasonic who has posted a lot of info and build logs on their forums- I put a link to one of his build logs (and my own) in Pneumonics thread. He has his own site here with a TEC calculator + other useful stuff
http://watercooling.co.nz/node/60

Good to see you and Pneumonic planning some serious TEC kit! Will be following with great interest. Which particular TECs are you going to be using?
BTW I have loads of O-rings at work of all sorts of sizes (certainly in Buna-N, but not many silicone) so if you get stuck let me know

 

[errata]

heating copper to high temp will anneal it ie soften it
200C will be fine but is probably 2 low to melt a good soft solder suitable for joining metals

the much stronger hard 'silver' solders used to join copper melt even higher

coppers rate of thermal expansion is high you will need a solder that covers a decent range of temperatures as well!

 

[Pneumonic]

Cheers man

edit: you mentioned in the Pneumonic's thread that you have more information on pelts, I would indeed be interested. Feel bad trashing his thread further though. Where should I look?

hehe constructive criticism / questions are always welcome.. I am guessing we have a similar understanding.. i.e. we have read the same threads at XS / documentation available.. so input is always welcome please feel free to bash my thread as much as you like

 

[rjk]

hmmm, i think you would be best leaving the block as it is designed, if a crack or something develops in the joins then you will only get problems

plus the oring on the ek blocks smooths out hte surface imprerfections, you would need to effectivley get both the contact plate and the blocks top level with eachother to a level of precision that would be impossible to get perfect

 

[JonJ678]

Some inarguable points raised here, cheers. Solder is probably not the way to go, though abandoning hope of taking it apart and using jbweld would probably work.

Cheers for the offer surge, good odds I'll want to take you up on that. How are O rings sized? Nominal diamater and length at a guess, but how much choice is there in length, and what sort of tolerences do they work to i.e. is closest mm enough or does it have to be spot on? If it's possible to mill an appropriate slot around the block and fit an O ring this is probably the best way to go.

 

[surge]

Some inarguable points raised here, cheers. Solder is probably not the way to go, though abandoning hope of taking it apart and using jbweld would probably work.

Cheers for the offer surge, good odds I'll want to take you up on that. How are O rings sized? Nominal diamater and length at a guess, but how much choice is there in length, and what sort of tolerences do they work to i.e. is closest mm enough or does it have to be spot on? If it's possible to mill an appropriate slot around the block and fit an O ring this is probably the best way to go.

O-rings are sized by cross section diameter (e.g 2mm) and internal diameter since they are commonly employed in circular grooves. If you pop the outer ring out (using something soft to avoid scratching the sealing surface at the bottom of the groove- the top of a BIC biro works quite well since its plastic but pointed) it will probably assume a circular shape then you can measure the ID. They will stretch somewhat- so ID accuracy is not critical.