No, I accounted for those (that's what took me so darn long to calculate!)

A small update.

The hydraulic equivalent to a 1 by 3.25 mm rectangular opening is: 1.53 mm (diameter). The closest imperial measurement is 1/16. There are, of course, 4 of these.

Now I can drill 4 * 1/16 holes in a PVC cap, and run a flow test, but off-hand, it's gonna be awfully small.

In the meantime, here's a render of the latest revision of the top (thanks again to Utabintarbo).

You're wright, just realized that I considered the fillet radius at 1mm instead of 0.5mm

Anyway theres a very fast way of calculating circular areas; just consider them squares and multiply by Pi/4 (0.785)

Because of jet slightly loosing it's velocity after it exits the nozzle though it is a slight difference. If your nit-picking for every last performance gain then the smallest gain counts too. With a larger flow/pump, the further the distance between the base plate and nozzle can be without hurting performance greatly.


I think It'll work fine. More than 5mm with our type of pumps we use will just kill the effect we're looking for and that 5mm is right on the edge of loosing performance.

It will pick up any heat in the center fin and it should hit the base unless the pump is so weak it won't push it.

I'm lost there. Have no idea what you mean.

That's how lazy I can be. :D

I did this for a hydorL30 pump. About 300G/6Ft. head and one foot of hose If I remember correctly. Going with a smaller nozzle can actually decrease performance. It has to do with the velocity and size/shape of the impingement. The pump can only produce so much velocity/pressure with a given size nozzle. Once you reach the highest velocity you can with a nozzle, going smaller won't produce any more and will just effect the size of the impingement area. Of course the bigger the better the pump, the smaller you can go until you see negative returns providing the pump heat isn't a "big" concern. I can draw it on ms paint to better explain it.

My pump is fairly powerfull: Little Giant 2-MDQ-SC (see PQ curve below). Max flow around 525 gph, max head 14.6 feet.

I understand what you mean about the nozzles: I recently (tried to) run the calcs for LiquidRulez here.

What I meant about the flow being at 1mm is just that: if you look at the first graph I posted (from Roscal's analysis), it shows that out of all the channels, most of the flow is concentrated at the bottom 1mm: the rest is almost stagnant. I'm kinda hoping that the new inlet geometry will kinda help this too.

That's one nice pump. I seen one at Ace yesterday and debated to get it or not.

I see what you mean now. Your getting more flow/velocity at the base of the block and towards the top of your fins the flow/velocity is slower or almost stagnate. You can see the effect of it in the thermo graph you got from Roscal also. I know why your seeing that. It's just the nature of the impingement. The velocity/flow is faster at the base in a thin layer where the water hits it and is deflected off the surface of the base. Then there is eddies (water swirling up and away from the faster layer), that swirl up and almost back the opposite way slowing the water flow down. How to change that I don't know. Some how you would need to direct the some of the flow/velocity at the top of the fins also.

what's happening on a big scale.
http://www.comebackandplay.com:8080/~jay/imping.bmp

If the nozzle exit is made big enough, I would think about adding some kind of deflector that sits half way down inside your channels and extends just enough into the center of the nozzle grabbing some of that faster moving water. Use maybe, thin copper sheet?

Just remembered your channels are 1mm. What i just said is impossible to do. :shrug:

I'm glad to see someone understands what I'm babbling about!

I'm not too concerned about the flow sitting at the bottom, because that's where most of the heat appears to be, plus the top is polycarb, so there really wouldn't be much benefit in trying to equalize the flow within a channel.

I think that part of the problem is/was this:

a) the simulation didn't have a nozzle plug, it had a huge 5/8" opening. With a nozzle, the flow should strike the baseplate, then exit smoothly (more or less).

b) the outlet had a swirl going on, which may have had a certain effect: that's been fixed (curved outside wall is now square, and the outlets within the top have been made larger).


As for adding deflectors, I wouldn't do that near the jet: I want to leave that one alone so it can "be all that it can be":D

I'm gonna send this to Roscal, for analysis round 2, but first I have to make sure that the nozzle will allow my pump to work within an effective range.

That seems more likely. As for the deflector, that would be just a waste of time.

I cant wait until this design become reality and ready for some real testing! It's gonna be interesting!

A bit of non-progress.

I e-mailed Ken from www.lmnoeng.com about a consult, instead of paying for the nozzle calculation. He offered me a discount on the standalone version: $120 instead of $150. Optionally, I could pay $30 for a one week access to it.

Back to googling.:shrug:

A bit of actual progress:

I ordered a foot long bar of copper, 1/4 by 2". Will be here next week.

so you're going ahead with the design?

Yes, the fin pattern is final, so the base is definitely ready for milling.

I'll try out different nozzles, until I get it right:p

Good Ben.
I was beginning to worry that you were giving up on such a promising idea/design.

Glad to see that you are not.;)

I also ordered some extra, so if the milling goes well, I might ask Fixittt to make an extra one for Utabintarbo, heck maybe even Roscal! (and one for Fixittt too, of course!)

Finding a 1mm endmill is much easier, now that the channel depth has been reduced to 4mm.

A tiny update:

I found the endmill!

www.moldshoptools.com

They carry p/n EMS12FtiN, which is described as:
I save 10% by ordering a qty of 5 or more. They come out to $39.33 for the pack, which comes out to $7.87 each.

There's also a ball end endmill available, same specs, but the base price is $11.03.

I have been googling for more than a day for this...:p

I got some pics back from WallWart:

Here's the block's o-ring, that just happens to fit perfectly in the PVC fitting.

...and here's the completed inner res component:

(ref 22067)

I found an even cheaper source of endmills:

www.drilltechnology.com

I e-mailed them, and got quoted these:

At $2.50, I can send one to JD for the top, and the rest to Fixittt.

What do you think guys?

Oh... the bit specs:
2 Flute Extended Flute EndMills -Square Cut

I still have not seen exactly what needs to be milled on the top or the depths or anything. Did I miss it? :shrug: I would like to see how much of this I can do on the laser before you decide to buy endmills for it.

Hum, actually I tossed out the mod to the top that I dreamed up. All I need on the top is the openings for the outlet, but I also need that nozzle, which is a 5/8" plug, 1/2" thick, where the opening is 1/4", but has a cross in the middle of it that's 1.5 mm wide.

That cross is going to have rounded corners, to match the cross in the copper base. By "rounded corners", I mean where the fins intersect, since the channel would be cut using the same endmill.

See what I mean?:p

[edit] In polycarbonate, of course, which I would send you.

Here's the top again:
(oh yeah, screw holes too!)

...and here's Nicozeg's draft of the rounded corners:
(never mind the extra stuff)

What that green thing is and what you speak of here doesn't match up. On the green pick there is nothing but a hole in the middle. Are you saying there is going to be a modification to that hole?

I am not so sure I would be any better to do this than Fixitt. He will have the actual block and be able to match things up. It is hard to make a top as perfect as it can get without having what it is going to. Might be worth asking Fixitt to do the top aswell. If for some reason the base needs minor adjusting then he can transfer that right over to the top without much hastle. Either way though I or Fixitt will need a dxf of the top and all the measurments and depths of everything. I will then have to redraw it to work with my mill and also make several different drawings for each step.

mkay...

...but I'll still need that nozzle.

I still don't have SW2k3 installed, and I already need a mod to it: drop the fin height to 4mm (from 5mm).

If I get you the SW2K3 file, can I get you to come up with a DXF that Fixittt can use? (PM me an addy!)

I'm also thinking about dropping the baseplate from 1.5mm to 1.0 mm, but I think I'll go with 1.5mm and if Fixittt is not having too hard of a time making this beast, I'll ask him to make another one that's got a 1.0 mm base.

Let me know.


Copper stock will be here today or tmo! Woot!

PS: Those endmills (if "Fixittt approved") are actually 0.0400 in, close "enough" to 1.0 mm (0.0398 in.) .

Here's a pic of the top plate only (below):
(left: view from top. Right: view from bottom)

It is now 1/2" thick (another mod to the SW2K3 file). There will be an inner tube that I will glue to the top of it, and it is 5/8" ID, 3/4" OD. Then, I'll glue the outer tube.

The nozzle will simply be dropped in.

What I need is the 1/4" nozzle shaped opening in this "nozzle plug", with the crossbars that should finish in a sharp edge, to seperate the water nicely.

Get me the files and I will see what I can do. If I canot do it I can bug my buddy to give it a shot. He loves this stuff. :D I will probably be seeing him this weekend anyway.

I cannot make heads or tails out of that nozzel from that pic. Can you get a drawing of just the nozzel in SW? or even a CAD drawing? Hell anything from a 3D perspective or all the sides drawn out.

Kewl!

Here's a *rough* sketch of the nozzle. I don't have any models for it.

(nozzle curvature details on next post)