A ?NEW? Waterblock Design

[Spartacus51]

I decided to sit down and think about the best possible waterblock design tonight,and I came up with some interesting ideas. This post is about the one I think may prove the most effective, and as far as I can tell has never been implemented (this may be a bad thing, because most the good inventions are already taken. I got so excited the first time I made the wheel, how was I to know it'd already be done?)

So anyway, the basic concept behind the design is concentric circles. Nearly all of the components of the design can be accomplished with concentric circles. I have in fact included a dot that is the center point of the circles on the diagram.

I liked the idea of a spiral waterblock, starting the cold water directly on top of the processor and moving it away from the center. The problem with the spiral (as best I can tell) is the no one is making true spirals. All the designs I've seen have been spirals whose radius doesn't increase proportionally to the angle. The problem with doing a non proportianal spiral is that it can't be programmed and milled directly, everyone has tried to drill individual holes and then dremel the inside smooth, but I have seen no successful outcomes of this. Rough edges = turbulence = lower flow = bad. The problem I see with a true spiral is that it creates numerous dead spots, and some of these are in important places.



So I like the nipple in the middle idea, and I like the arcs as compared to perpendicular intersections. It was in imagining this that I arrived at my new design.

Tips and Tricks:
I feel that the best way to put the exhaust nipple on this block is in the side. So when I mill the block I plan to mill the circles only and drill the hole for the nipple to meet the bottom of the water chamber of the block.

The rounds coming from out of the area surrounding the intake nipple should make flow in this area easier, but may not be necessary to the block's effectiveness.

Channel can be milled to whatever width is desired, I tried to indicate that my channel would be comparable in size to the nipple holes.

If anyone has already tried this, let me know your results. I'm in no way trying to steal anyone's invention, although I am certain that I arrived upon it independently of any block I've already seen.

Suggestions on how to improve it and/or reasons it will fail/succeed are welcome and requested.

[Ottoman]

assuming I understood u correctly...
water enteres the middle, splits in 2, and goes around the block and exits the bottom....


would the vertical suspension of the waterblock effect the way it flows, like mor out the bottom then the top, possibly cooling only lower half of the waterblock???

and u need mass to the block as well, so there's area so the block can pass the heat off to the water...

and if the pump were to fail u'd have some time before the cpu overheated...

[Spartacus51]

I suppose I should have labeled it better, that would be a top view of the water channel. There would still be some mass to the block, and I plan to piggyback a HSF onto just in case of the event of pump failure, I would certainly have time to turn off my computer and fix the pump. The hole out what you assumed to be the bottom (if I'm understanding you correctly) is actually the outlet hole out the side. water would be injected into the center directly above the core (ideally)

[Diggrr]

That's alot like the new IceFlow block from overclock watercool. Although their basic layout is more square, and yours is round. It's supposed to be beating their present blocks though, so the design is a good one.

[Sonny]

Build it & tell us how good it performs. Looks like a winner to me, that isnt much but what the hey;) You probably will need to do some tests regarding water pressure, high or low, to see what works with it.

One question; Will the inlet be in a perfect 90° angle?

[Christoph]

Looks like an interesting design, I disagree with you about spiral waterblocks. Spir@ls (http://dtekcustoms.safeshopper.com/7/31.htm?11), as far as I can tell, is a true spiral, but you won't be able to order them for a while.
I like your design, but I see three problems.
The first is that there is potential for some turbulence where the one channel splits into two. I'd propose something like this;
http://imagehost.auctionwatch.com/preview/id/ideamagnate/waterblock.png

The second problem I see is that the water will go much more slowly through the circle section than it will through the first section. To fix this, you could simply make the circular channels carry half the volume of the straight channel. (Remember that 1/2 the volume doesn't mean half the width.)

The third problem I see is that there's a lot of the block that won't have any water flowing through it (the red area in the center of the block especially). To remedy this, you could make several concentric circles connencted, kinda like this:
http://imagehost.auctionwatch.com/preview/id/ideamagnate/untitled.png

Just a couple ideas. Nice design though.

[Holst]

That looks a bit like my waterblock.

Im trying to keep the flow rate up without sacrificing surface area.

I should be making it during the week.

Its going to be very similar in design to the dangerden blocks.

The 12 marks around the sides are for bolts for the TEC coldplate as im trying to get some mega pressure onto it.

http://www.genasis.freeserve.co.uk/off-site/dccuwv.jpg

[SpaceRangerJoe]

i added something else to that design that i thought could help. now please excuse my (extremely) bad artwork. i can only do so much in two minutes with a touch pad. but feel free to laugh. the two wavy red lines, one on either side of the block, in the water channel, are more walls. just keep an open mind here, i know the picture is bad. its like machining two coolant channels on either side, instead of just one. i know this would complicate the process, but the extra walls would greatly increase the interior surface area, hence, more heat transfer. now i didnt change the dimensions of the channels, which you might want to do to help the coolant flowand to decrease turbulence in the water through the channels, but i think you can get the general idea from this picture. let me know what you think.

[SpaceRangerJoe]

sorry, it didnt attach the picture

[Spartacus51]

Alright, Model II is here... or at least design II. I took your suggestions and drew up another version, very similiar to that of SpaceRangerJoe. I suppose at some point I'll have to sit down and figure this all out mathematically, how to maximize surface area and all that (ahh yes, calculus for, FUN? At least now I know there is real world application of that crap) But anyway, I've designed a model II and actually a model III although I'm not entirely sure that I like that one. The idea was to minimize the hot spot in the middle from the large chunk of copper. (Although I'm not sure if it's necessary, it's possible there's enough surface area there to cool the entire thing, and it's likely not over the top of the core anyway. Will have to play with different ways of milling all of this, but I'll say now that it won't be a flat bottom at the inlet, will be a round so that the water is directed where to go. My concern with model three is that the large inner area may allow water to sit and swirl in that area while idealy it is pulled quickly away from the core into the outer chamber... will have to see what testing yields. As of yet I don't have a pump or the materials for a bong or a radiator. Will have to get on that... this whole setup may be a project for christmas break... Further ideas/suggestions?

[Spartacus51]

And Model III

[Christoph]

Looks great! I've only got one suggestion, and this time the picture's pretty, since I felt bad about defacing your artwork last time.
http://imagehost.auctionwatch.com/preview/id/ideamagnate/.mids/waterblock3a.png
I suggest that you make the inner semi-circle's inner radius match the ID of your tubing and nipple (didn't mean to get personal!) to minimize turbulence while maximizing velocity.

[Gravity Man]

to avoid air pockets forming, I inverted it. I also enlarged the very thin walls because they did not have enough contact with the base to allow for better heat transfer. I don't like IdeaMagnate's last idea (no offense), because I think it will hinder the high flowing nature of such a water block while not adding much to its cooling eficiency.

[Gravity Man]

A picture would help

[Spartacus51]

Agreed on the extra inner chamber, think it would hinder flow. Perhaps if it was a block to be used with a radiator, but I intend to bong cool (cheaper, and as best I can tell works better) so higher flow is a good thing. Block doesn't need to be inverted though, as that is a top view, so really all that was switched was left to right.

Perhaps this is just me liking math a bit too much, but does anyone have any idea how a setup would look comparing mass, surface area and such to figure out the max. values? Just an idea.

I'm considering another different design, again based on concentric circles, but with a slight twist... will try to get that posted sometime soon. Also will try to get a better diagram of the one I've already designed...

[Christoph]

About figuring out the surface area, will the block channels be square shaped at the bottom or rounded out? If you're going to make them square-shaped, then it'll be a snap to figure out SA if you give us the depth of the channels and the appropriate dimentions for the block. If you're going to do rounded channels, I call dibs on figuring it out. (I love that kind of stuff!):eh?:
Just post those funky dimentions, and I'll be glad to take it from there.

[The Overclocker]

have the split holes half the size to keep the flow rate up

[Spartacus51]

Hmmm, starting to move into final planning stages. Think I'm liking the design, but need to finalize some stuff. Bottom will be rounded, so magnate, you can have all that fun stuff. before we can optimize that stuff though I need some help. As this is my first waterblock design, just how deep should it be? I don't want the block to allow less flow than will becoming into it, don't want that to bottleneck. Also agree with the earlier post that the rounds on either side should allow for half the volume of the inner chamber. How much stock do I want to use? How much copper do I want left on the bottom after milling? Does it really matter? Looking for the voice of experience here if I can get it. Once I've got suggestions I'll figure out all my dimensions and get em posted on here so Magnate can have some fun.

[The Overclocker]

i dont know how deep it should be the spir@l is almost an inch deap but my becooling block is only 1 cm

[Spartacus51]

Do they have similiar remaining depths? As in thickness between channel and bottom?

[Spartacus51]

Alright, production phase beginning monday, need to know how thick my base should be and if there are any other suggested mod's before I get this thing going