Ultimate block? Theory.

[Limeygreg]

BillA
Looks like we were typing and posting at the same time.

Okay, I think I see.

If we are going to use an aquarium pump then it's got to be low resistance high flow with larger channels and physically induced turbulence - al la "turbulators" say.

If we use a small, low volume, high pressure pump then the block should be low flow small channel as in the cross-drilled block say. But it would probably only need say three layers of tubes though, or would it need more (as in the recent block under discussion)?

By the way Bill, were the TIM's of any use??


Can O' Beans and mr. phantastik

How about those self drilling self tapping sheet metal screws, I'd think they would cut a nice coarse thread into a hole in a copper block. Also, how about those concrete screws that will go directly into concrete, they're pretty damed tough and should be able to form a thread in copper easily.

[BillA]

haven't tried them yet, in the midst of a wb series
I'll let you know (and thanks)

if one layer is effective (small holes, close together), then 2 layers should be sufficient to preclude upward heat propagation
'look' at the gradient(s) through the bp - which is the whole wb as there is no waterbox

[Can O' Beans]

I have the tools/items to put screw threads into the copper, thats no problem.

I'm just wondering if it's too fine of a thread, and would the holes benefit from a longer thread, as in 4-10 "threads" per inch. Would turbulate & spiral. The regular screw threads would turbulate, but probably wouldn't spiral. I have one of these scerws, has about 3-4 threads per inch. Will have to test it out soon.


I have two designs right now, and I'll probably make both of them as I have multiple systems to WC. But that won't be until I finish modding the case for my current project, then get the motherboard & do some measuring. Trying to "perfect" my designs now ahead of time.

[BillA]

[mr. phantastik]

Limeygreg: I think a good self-tapping screw would do the job quite nicely, and the threads would be pretty jagged because self tapping screws aren't exactly meant to be taken out.

Now this turbulence adding is all fine and dandy for the water inlet/outlets on the sides (presuming we're using the cross rilled style talked about earlier), but what about adding the turbulence to the smaller holes? Would it be dangerous to use self-tappers for such small holes (breakage)? What about one cross drilled hole being the same diameter as the inlet/outlet channels, to allow for more flow over the core, like this: ...O... We could then tap that centre channel too.

The problem with cross drilled blocks is usually you don't have the water comming in right over the core. Only way to do this with a cross drilled I guess would be to have two outlets and one in, like in the image. Or would i be correct in assuming that most of the flow would go through the small channels in the centre instead of all around the block. If anyone can't tell i'm a big advocate of low-flow, only because of my rad ;]

oh yes, btw, ph33r my 1337 ansi skillz

[bigben2k]

Ya'll,

I realize I may have confused a few people with the links I posted...

The spring idea is independant from turbulators. I just posted the link because I though that it was of interest (it came up in my search), since the cooling problem that we have to solve is this boundary layer thing, where in a tube for example, the coolant in the middle of the tube flows faster than the perimeter. The spring, any helicoidal form would probably do, but its purpose is strictly to induce a little bit of turbulence to break up where the coolant sticks to the walls.

As for turbulators, I was inspired by the TC-4 block, Morphling pyramid base block (and everyone else who tried, and failed) and this latest thing that I found in a magazine called "Flow Control": a flow meter based on turbulence, for which I'll post a picture Monday. Anyways, the idea is to force the water into a spin before it enters the block. For lack of a picture, try to imagine a boat propeller shaped thingy, fit right into a center inlet.

Following my design#2, where a middle tube is surrounded by a large 2 inch tube, it would be perfect... at least as far as my little knowledge can take me.

For those who prefer a channel type design, then you can use the spring. Alternatively, I would even advance that the channels could be grooved, similar to a riffle barrel. It's not meant to induce a circular motion, it's just a little add-on to help break up the boundary layer.

[dream caster]

Originally posted by mota


http://www.google.com/search?q=coppe...8&start=0&sa=N

Here is another interesting variation of microchannel design by Thermotek that they are calling ThermalPlane. Hmm, looks like the page with the info on that block has been removed from thermoteks site.



Dual pass microchannel designs are also being developed.



You can forget about making a better waterblock with a mill, drills, saws etc.

Mota, please update us extensively; when you tried a maze2 against a Lytron microchannel plate, waterblock was better.

[BillA]

mota's latest

the Lytron plates (actually made by Thermalex) are not effective with a concentrated heat source

[bigben2k]

Originally posted by unregistered
haven't tried them yet, in the midst of a wb series
I'll let you know (and thanks)

if one layer is effective (small holes, close together), then 2 layers should be sufficient to preclude upward heat propagation
'look' at the gradient(s) through the bp - which is the whole wb as there is no waterbox

Thanks for answering the question I've been asking for such a long time!!! I didn't quite believe that more than a row or 2 would be effective.

As for the Lytron (Thermalex) cooling plates, I agree, the baseplate thickness isn't conducive to effectively cool a concentrated heat source. On the other hand, if the baseplate was in the 5mm order, it would probably perform as well as anything else (IMHO), but I can't help thinking about poor Kevin's water chiller experiment (failure), so I'd think twice before going that route.

I called in sick today, so the pic I promised is going to have to wait until tmo.

[Can O' Beans]

It looks like they should perform better on a peltier setup where the heat is a bit more widely distributed. But I think a decent waterblock would outperform it anyway.

[bigben2k]

Ok, here's that pic I promised.

It's a flow meter that relies on the swirling effect (induced) to measure the flow. As you can see, there's also a flow straightener, but I don't really need that...

[bigben2k]

To complete my entry...

If you take my design#2 (see page 4 in this thread), add the swirler in the intake, and re-arrange the rest of the fins to accomodate the swirl, this block might perform very well, with a lesser flow.

I'll have to figure out how, but I'm ready to start experimenting with it. I just need a fellow CNC'er to help me out with it... (once I've tuned the swirler and fins).

[BillA]

you need to study up a lot more on vortex (shedding) water meters
try Yokogawa
lots of head loss, lots of velocity

not gonna happen with an aquarium pump

[bigben2k]

Originally posted by unregistered
you need to study up a lot more on vortex (shedding) water meters
try Yokogawa
lots of head loss, lots of velocity

not gonna happen with an aquarium pump

I know about vortex flow meters, but that's not one of them. I'm describing a swirl meter, made by ABB. It's used in a slightly different range of applications than a vortex meter.

In any case, I have no intention of measuring flow: all I want to do is induce a spin. I'm not going to nozzle the coolant input either, although I might experiment with it.

I realize that swirling will add to the flow restriction.

As an alternative, I MIGHT slap a triangular cross bar at the intake, and do the vortex thing, but that'd be for later. Right now, I'm going to concentrate on the swirler.

[Dix Dogfight]

BB2K
I think you actually might be on to something here. But I think that the flat bottom needs some thought.
If the water comes swirling down towards the bottom and then exits out the side. Wouldn't that create a dead spot in the middle like a small "tornado of water that only circles around and never leaves the center?
I'm guessing that a small pyramid at the bottom where thw fins start would reduce this or?

[bigben2k]

I see what you mean, but considering the velocity, I don't think it's going to be an issue: I plan to keep the inlet (see pic, imagine the area where the yellow bars extends) to the equivalent cross section as the 1/2 inch tube, so...

1/2 in diameter = 0.196 sq. in. cross section

Take that cross section, and apply that measurement in a cylindrical pattern.

.196 sq in surface of cylinder = 0.125 inch height. (1/8 of an inch).

I don't think that 1/8 of an inch would be an issue.

[bigben2k]

And here's the pic. (not to scale)

[Dix Dogfight]

Ok got it.
If i understand your concept correctly, you want to:
1* Have a thin bottom.
2* Add lots of turbulence to that area to minimize the boundry layer and so on.....

Then Iv'e got one more suggestion on the construction:

Why don't you make the turbulator go all the way down to the bottom. Or why not make a insert in the bottom that looks like the attached pic. If you make the center of the turbulator small then a little heat would travel up the turbulator but most of the heat would be transferred from the bottom around the center axis.

cheers

Edit It would loke something like the black INNOVATECH block

[bigben2k]

That's a good idea! I was thinking about what I'd use as a fin, over the core area, and I hadn't figured anything out yet. I was thinking about pins, fins, and divots, but I guess I could go for a pin, and use divots everywhere else (a la #rotor).

I could even thread a hole to screw in that pin: I have to consider how it could be manufactured, just in case...

Alternatively, I could put (as you suggested) a bottom plug that would screw in: the rest could be made of some transparent material... but could a transparent block handle a 2 inch clamp? Otherwise I'm back to the innovatech concept, with anodized Aluminium, and I don't really care to get into that, unless Jaydee can source one

Thinking last night... that turbulator could be made by simply taking a copper stock plate/bar and twisting it.

[hemophilic]

I was considering making these.
My thought on this design are...
-easily manufactured channel
-easily reoriented in/out holes in solid copper top(to accomodate different mounting systems)
-center lug bathed in water (similar to innovatek)
-solid copper
-wide channel to accept 1/2" flow

Basically a solid copper innovatek with 1/2"

[bigben2k]

I don't think it would perform very well: it's the equivalent of a very thick baseplate, so the heat would have to travel quite some distance before hitting a point where it'd be released from the block, which would allow the core to get very hot.

That's the reason why the Innovatech block's center mass is finned (horizontally).

[hemophilic]

If that's the case, why does the innovatek perform so well then? Even with the ribs for increased surface area, it's still a rather large hunk of copper. In the same vein, why do the best aircoolers have thick baseplates as well?
Also, there would be significantly more flow with the wider channels and 1/2" barbs.

maybe I should have posted this somewhere else since this is a theory thread. I'm going for a producible block, not a theoretical masterpiece which is impossible to construct. Oh well.

[utabintarbo]

Here's my take on one of Ben's ideas (the inlet inside the outlet one). I couldn't come up with any good way to make his manufacturable (several things scare me there), so I went with 2 outlets to one inlet with flow directed over the core, and the outward flow "directed" towards the barbs. I could probably go to 3 or 4 outlets, but this will still outflow about 95% of all blocks now extant.

Comments?

Bob

[utabintarbo]

Here is another view:

[bigben2k]

That looks rather nice, and it does address some issues. What size barbs did you have in mind? How exactly is this easier to manufacture? I mean, I always assumed that one starts from a rectangular block of copper...

How do you propose to mount it?