Project: Hush! - updated 26/12/23

JonJ678 said:
I wouldn't know where to start calculating that, it's going to be flow dependent for one thing.
Click to expand...

That's true, though the flow rate (velocity) of the water in the 4.8mm ID tubes is about a sixth that of the water in the 1/2" inlet since the cross-sectional area of the 48 combined tubes is 868.6mm^2 vs 126.7mm^2 for 1/2" tube.

Assuming laminar flow?
Click to expand...

Yes, based on a quick and dirty pipe-flow online calculator. But it doesn't have the function of introducing bends, so it's based on straight pipe.

I had a brief play with numbers, and the closest I could come was the resistance of the 49 tubes is about half that of 1" inner diameter tubing based solely on ratio of cross sectional areas.
Click to expand...

The pressure drop of water in a straight circular tube depends upon the tube diameter, length, and volumetric flow as given by the Poiseuille equation (delta P = (8 x mu x length of tube x volumetric flow)/(pi x tube radius^4).
mu is the dynamic viscosity (Don't ask me what it is exactly!). So halving the tube radius increases the resistance 16-fold. Who'd have thunk that basic fluid dynamics we were taught for haemodynaics would come in handy?!:D


Flow down 4.8mm tubes is going to be pretty turbulent at the best of times, which is great for heat transfer but ruins me mathematically.
Click to expand...

It's actually the other way round - larger diameter tubes actually convert to turbulent flow at lower flow velocity, because there's a bigger difference in velocity between the 'laminae'. I'm not worried about turbulent flow - there may be some at the bends, but it'll soon re-establish laminar flow after the bend. What does concern me is the frictional pressure loss from the sheer surface area the water is in contact with, and the length of the tubes (around 120-130cm).

How big is the copper structure? It's very hard to judge from the pictures.
Click to expand...

39.5cm x 45.5 x ~45 atm (some tubes need trimming).

If possible, which it looks like it might be, Id suggest (after vinegar) putting a ring of solder around each and every joint, then sticking the entire structure into a household oven at 230 degrees. If it fits, all you have to do is check every couple of minutes and the entire thing will solder itself together for you. if it doesn't fit, befriending the local pottery group and borrowing their kiln would be an alternative.
Click to expand...

Good suggestion!:) It's too big for our oven, and besides, our oven's dodgy and takes ages to cook - due to a furred up gasline I think, since the last one had the same problem. I doubt it could get up to 230C. I think other home ovens would have the same problem - I'd planned on doing it that way but from my calculations it would take a kilowatt oven about half an hour to get up to temperature, assuming no heat loss, and was told solder flux should get to heating temperature within 6 minutes.

A kiln would no doubt have no problem though, since I'd think they use a heftier heating element, though I could be wrong.

I had thought about getting it put into an oxygen free oven, but I can't find anywhere vaguely local, and it would probably cost a fair bit to get them to stick it in the oven..

Heating it with a blowtorch will require a massive blowtorch or considerable ingenuity. Soldering a heatsink is, as you've noticed, not so easy. How does sitting it on an electric hob sound? Keep all the copper reasonably hot, turn a blowtorch on the part you're soldering. Heat wont dissipate anything like as fast if the temperature gradient is less
Click to expand...

Don't have an electric hob unfortunately. I'll try the blowtorch and let you know how I get on.
 
Really good to see photo's after all the threads on this epic during the design stages.
And nice work so far, be good to see how this progresses and looks in its final form.
+1 sub

Makes my custom rad plans look very simple in comparison :o
 
Last edited:
Glad to see you ditch the idea of trying to get all the small bore tubes to meet up with a large bore tube drilled with umpteen holes whilst also having to fiddle about bending the ends to suit, that would have been an absolute nightmare. By plenum I take it you mean a rectangular manifold?

Also, you say this is also going to form the case of your pc, where is the mobo going to lay,? Is your first pic showing how it will sit, i.e. the mobo tray horizontal above the bottom fin array?
 
Didn't have time to read all the posts, but if you are still using the 8-9mm copper tube and you used a pipe cutter on it, it's probabbly not a good idea on such small bore pipe.

The pipe cutter will push in the end of the pipe and possibly reduce the diameter to 6mm or less on the ends.

If you do use a pipe cutter on this size of pipe, try to ream out the ends well, although I would strongly suggest using a junior hacksaw to cut it and using a fine file to clean up the ends, both internally and externally.
 
Last edited:
Bubo said:
Glad to see you ditch the idea of trying to get all the small bore tubes to meet up with a large bore tube drilled with umpteen holes whilst also having to fiddle about bending the ends to suit, that would have been an absolute nightmare. By plenum I take it you mean a rectangular manifold?

Also, you say this is also going to form the case of your pc, where is the mobo going to lay,? Is your first pic showing how it will sit, i.e. the mobo tray horizontal above the bottom fin array?
Click to expand...

The motherboard tray will be vertical; I'll post a photo up later with a rough mock-up how the motherboard tray will sit.

Cambofrog said:
Didn't have time to read all the posts, but if you are still using the 8-9mm copper tube and you used a pipe cutter on it, it's probabbly not a good idea on such small bore pipe.

The pipe cutter will push in the end of the pipe and possibly reduce the diameter to 6mm or less on the ends.

If you do use a pipe cutter on this size of pipe, try to ream out the ends well, although I would strongly suggest using a junior hacksaw to cut it and using a fine file to clean up the ends, both internally and externally.
Click to expand...

Thanks for the suggestion. :) The pipe cutter does leave a big ream on the ends (though it doesn't deform them). I need to trim the tubes down later anyhow (possibly with a dremel as the mini pipe cutter won't fit as the surrounding pipes are in the way), so I'll take a small spindly file to them when they're at the final stage.
 
I remember you talking about this a couple of months ago. The plan sounds good, the work looks excellent. I await the next instalment :)
 
/Thom\ said:
How much planning was put into this project out of interest cause it seems like a lot :p
Click to expand...

Quite a bit - lots of reading up on stuff and trying to find parts on the web mainly. The actual design changed around a few times for what sides of the case would be heatsink, and motherboard orientation.

Mr-White said:
looks huge :D

How much did all that copper set you back ? :eek:

MW
Click to expand...

Around £160, though on the plus side it's probably worth about that as scrap copper. Copper's expensive stuff, and I'd planned on using thinner sheet due to price - for the amount I needed it was looking like around £300-400 for thick copper sheet so it didn't figure in my plans. But then I found a really nice guy on ebay selling several large sheets and who was willing to cut the sheets to size. He offered me commercial funding:eek::); I think he thought I was making a prototype to go into production!
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom