Playing with new water block desing
 

?http://www2.arnes.si/~mlivak/new%20block.JPG

Hi again

I was playing with solid edge, and came up with this one, I think it should be great. The outlet is where the two channels connectc again at the top of the pic, I will also round up the exit edge to lower the water resistance, and the top of the block will be special too, the inlet will be like nozzle to maximize the speed of the water flowing to the center of the block.
What do you guys think?

the machining of a block like that would take days!! but if you made one and it came out it'd probably be excellent, surface area is prety good and in the right spot assuming the inlet is the center one. Otherwise nice model. To make is a bit more feasable perhaps some cilindrical pins would be tons easier and less time consuming than the conical ones.

--Matt

Could a series of holes be drilled in the desired locations and then short pieces of either #10 guage or #8 gauge solid copper wire be soldered into place?

Jim

damn that looks cool morphling, I'd like you to make it, and compare it to your spiral style block

Careful... OCH is liable to steal that design... and make it crappy....

the wires could be effective for heat transfer if you used silver-based soldier. and even if they don't help much with the heat they'd probably add some turbulence in there for you.

--Matt

that'd be human labour time consuming. With the original design it is possible just to set the mill on it's way and leave it for a day or two

very cool looking (and origianal to me; ive never seen a block like that B-4). but could you make the channels a bit bigger (you have some wasted space, why not use as much of it as you can?)

why does that remind me of something a dominatrix would use?

man that would take forever to machine!!!!!!!!!!!!!!!!

Whoo..... glad I aint machining it. i bet it would take a day just for the pins!!

Wow, that block looks so cool (and weapon-like)! Do you think there would be much of an impact on performance if there were only spikes over the general area of the CPU core? It would help to simplify the design for mililng if you wanted to create them quickly, I suppose.

(Is it just me, or does the outline look like an emperor penguin? Maybe I need to take a nap.)

Yeah, I know it would take forever to machine this, but I'm not thinking about machining it, well except for first model :) . As for the size of the channels, they are not the final one but I'm not going to make them a lot bigger, I'm focusing on where the cooling is needed the most.

You not thinking of a casting are you?
Jim

Hmm :D

Casting would be to expensive(way to much power to melt the stuff!!!), but so would the day to mill the thing.:D I like it!!! maybe make one without the pins first and if it shows promise try one with. But damn that is a lot of small milling to do.

My poor mill is getting it's @$$ kicked with the little Copper block I am making now. Something got thrown off balance and it is vibrating without even milling anything.:( Just about got the block done though.http://www.dorrellco.com/copperWB3/001.JPG

I was wondering if it would be practical to have the molds cast and then final finished on a mill. You have to give me some slack here; I am not a machinist or metal worker so I have no real world knowledge of the costs in these matters.

Actually I have wondered how much profit there is in the making of the blocks, since it seems extensive time is involved in milling them out. No, I am not interested in going into the business! But, I find the the business interesting.

You guys have most likely explored this idea already I'm sure.

Jim

go to my site at www.custom-cooling.com and there is a few articles Fixittt wrote up about the cost in the links on the left.

And molding will be very expensive. The cost to melt all that copper is very high for small amounts of blocks. If you where making 100,000 at a time and where able to sell that many in a few months it would be profitable. But 5-30 at a time would push it up to $200+ per block roughly.

I was waiting for someone to show this type of design. I know that IBM’s high-speed machines had a similar design, but I wonder if they ever patented it? (more than probably) I wanted to know if a hybrid type of design would fly. The milling of copper can be time consuming, especially with a difficult design. OTOH, if one could use a copper heatsink (i.e. Alpha or something like this), the milling of the pins could be avoided. I know that many of the heatsink manufacturers are going with fins similar to those on the radiatiors in some cases. Why not use their copper heatsink for the surface area and embed it into a copper box (block). One can get very creative with the fins on the copper heatsink to direct the flow of the water to reach all areas of the copper block with very little dead area. I know that the poly-blocks are the premiere things into this arena, however these things are limited to a small following and the water flows can be questionable.

Many designs have dwelled into the flow problems, but surface area can also make up for the heat transfer problems. Perhaps, I am looking at this from the wrong vantagepoint.

yeah, do it without pins at first.

we could even settle on one design, have fix, jay, morph do it, send them all to someone to test, and see who's interpretation came out best

I allready made similar block with Al, and I'm also going to make this one without the pins, as for casting I might get someone to do it for me and if we work everything out maybe there will be this kind of block out for sale and I'm talking cheap here, 30$-40$.
So prepare to want one :)

So you are considering casting. Interesting, I would like to see how this plays out.

Jim

How would you get it so cheap per block? I got quotes for one of my blocks and it was not nearly that cheap unless I had them make a couple thousand at a time. And then you would still have to mill it to touch it up. The pots they use are huge, so they really don't like doing small quantities, and all the electricity to do it costs a fortune.

If you have a friend with a small smelter that only does a small number at a time then that maybe the ticket!!! My engineer budy was thinking about building his own smelter as he used one at the school he went to make all kinds of stuff, and now that the job he is at now has all the big CNC equipment he could design and make the molds. But it just isn't practical enough he says. Copper melting point is 1981F or 1083C which takes a lot of power. Aluminum melting point is ALOT better at 1220F or 660C!!!

I looked into getting casting done and MORPH. let me give you some advise if yolu are planning the casting route ok.

First thing U would want to do is get a hunk of the material they use for casting (Because I can almost gatentee its not a true copper material) machine the peice to the final desing. And test it. For preformance and for water leakage. When copper is melted it is also (gonna use this word but its not all the way correct) churned. They keep it flowing. And this creates air bubbles. When I machined a hunk of material the casting company gave me, it leaked so bad on the bottom. because of the small air pockets. Also castings look oitted when done so machining will be needed on the surfaces to get a nice smooth finish, but will still be pitted.

Also getting the patterns made on a bard is pretty expensive. I used to work next to L&M pattern in Tulsa. They quoted me $1200 to get a board of 12 patterns made. The patterns are a reverse coppy of the block.

If you are going to do an all machining operation and want to distribute them. I can almost garentee that your cost will have to go up on the blocks. Hell, I dont make any money as it is on the Spir@ls.

I hope this really works out for you Morph, and if there is any info I can help ya with let me know.


JIM. Its a logical step to try and find ways to speed up the operation and reduce the labor involved. It was a great idea.

Like I said, my friend told me that he could do it for aprox. that kind of cash, we'll make models and everything, and then we'll see if we can make them that cheap. But I'll first have Cu version without the pins and optimal base thickness and if it will perform good we'll try and go casting. But like I wrote before the thing that will separate that block from others will be venturi nozzle like inlet.

kewl kewl!!!!!

Yeah, I hope it will work, and as for the price of the molds if you're capable of doing it, you wouldn't charge yourself, right. You just have to work a lot of hours for free, but we all enjoy doing that kind of stuff, otherwise we wouldn't be here, talking about it. As for the casting, it should be preasure casting, that way there shouldn't be lots of air bubbles, and the final touch with the mill will be needed, but more for the looks