"Radius" by BigBen2k

[Dix Dogfight]

I finally remembered what it was that this block reminded me of. The History Channel did a story on radar development. While the Brits came up with a way to make it work, they could not manufacture an intricate part (looking much like Ben's block) in mass quantities. A US engineer came up with the idea of stamping thin sections and bonding them together to build the required height.

That would be the Magnetron oscillator, and they sorta look like the attached picture. However the newer magnetrons that we have in our microwave ovens looke like a hybrid between the RADIUS and the magnetron in the picture.

cheers

[MadDogMe]

BigBen!, how are you planning on doing the outlet?, like a two in one pipe?(like phase change, inlet inside outlet). it'd be difficult but doable. I'd alter the size of the holes around the circle of my design to balance out the flow, to allow for the offcentre outlet. but I'd like to try to incorporate the 'two in one' idea into my design as well(no need to alter hole size for balancing flow ).

Or were you planning on multi outlets?...

*edit*

There would be a 1 7/8 inch copper tube soldered to the top, for the 1 7/8 ID vinyl tube. The center inlet is fed from a 1/2 inch tube that's inside the big tube.

Read it properly this time , I still can't visualise the 'cube res' part, are you gonna carry the two pipes all the way to the res/pump without seperating them?. you could make a seperator that fits to the block yes?, is that the 1 & 7/8 Cu tube thing?.

PS. I think I got the Res thing now, you don't have to submerse the pump either do you!!, just use the res to 'separate' the two tubes , Kudos!!...

It's very simular to what I had in mind, except you took it further with the two in one pipe . the second pic of the 'slots' in a circle, they're the outlets from the main block chamber yes?,are you gonna use Cu to go on top of that?, or Plexi?(you could use plexi tube to attatch the 1&7/8 tube to, on the top plate. and 1/2 tube attatched to the 'slot' plate) to keep the weight down, I'd use plexi for the plate with slots in as well, make for a cool visual experience as well ...

[bigben2k]

MadDog: glad you were able to figure it out!

I got tired of the usual inlet/outlet scheme. I needed something that would allow for an unencumbered outlet (read: less restrictive). I'm surprised no one used it before.

The cube can be just about anything. A plexi cube with two 1/2 barbs on it should do. Think about a cup, where the twin tube is attached to the top opening, and the 2 barbs are stuck at the bottom of the cup.

I've got some preliminary drawings from Utabintarbo.

I also got in touch with Fixittt, who tells me that a 1/8 inch endmill is about as small as he can go, so I'm running new numbers.

[Cathar]

1mm end-mills can and do exist. They don't really come any smaller than that though unless you start getting into the really expensive machinery.

[MadDogMe]

Have you done away with the slots?, I'm using eight 'holes'(non~symetrical) in my design, venting into a chamber rather than a large tube(pipe ). did you work out the outer tube size so you'll have the optimum area inside?(outside of the 1/2 inlet tube) so it equals the 1/2 tube inner?.
You'll want the res as close to the block as possible as well yes?, to minamalise time spent hot next to cold?. though I agree it's a small temp change only there for a very short time...

With the res, you're going to empty the block outlet into the res' but carry the block inlet tube straight through it to the pumps outlet?. that's the picture I've got. it should revolutionise multi outlet blocks!.

[bigben2k]

Originally posted by Cathar
1mm end-mills can and do exist. They don't really come any smaller than that though unless you start getting into the really expensive machinery.

Is that what was used for your block? Fixittt says 1/8 (3.2 mm), because 1/16 (1.6 mm) is too small for his CNC: Special high-speed CNC would be required.

MadDog: I haven't done away with them, this is Utabintarbo's interpretation. I haven't checked to see if this plexi configuration can withstand any stress (i.e. the big 2 inch hose clamp), but it looks good!

The ratio of the area between the 1/2 tube, and the 1 7/8 tube is pretty small. Yes I did check it, and it's not a problem. I don't care how fast the water exits, I just want a 1/2 inlet that's as unobstructed as possible. As for temperature differential, given the flow rate, the deltaT between inlet and outlet will be pretty small, so again, not a worry.

Yes' it's a new outlet scheme. I'm hoping it will catch on. That would be the least of my contribution to the watercooling world!

It's certainly an improvement over BladeRunner's video card cooler, with 3 tubes in, and 3 tubes out!

[Arcturius]

I realize this won't give precisely the same output that you're looking for, but what if you simply use a bandsaw to cut across the diameter of the block a bunch of times?

You could then to touch-up work with a dremel or something.

[utabintarbo]

Originally posted by Arcturius
I realize this won't give precisely the same output that you're looking for, but what if you simply use a bandsaw to cut across the diameter of the block a bunch of times?

You could then to touch-up work with a dremel or something.

I was thinking along these same lines at a point. The only major issue would be a baseplate seperate from the casing. This shouldn't be a big hurdle, as no new ground is broken there (see Innovatek, and just about all German w/b designs lately). Might be kinda pretty!

Let me see what I can do!

Bob

[bigben2k]

Originally posted by Arcturius
I realize this won't give precisely the same output that you're looking for, but what if you simply use a bandsaw to cut across the diameter of the block a bunch of times?

You could then to touch-up work with a dremel or something.

That's what inspired me when I posted the Pro/Snowflake block (although it couldn't be done that way, it is far simpler).

It couldn't really be done that way with Radius, because the cuts would chop up the area over the core. I also thought about diamond shapes spread in a circular pattern, but again, because the pattern is circular, it wouldn't work.

Sticking with a straight pattern of diamond shapes, amounts to nothing else but a pin block (or even a la Swiftech), which is not what I'm shooting for.

[morphling1]

bb2k, I thought of this idea long time ago, when I first think od radial design block, but it's not that simple to implement, somewhere on that line you will have to separate those to lines and for that you will need special piece like a block where that hose and a hose within come and then separate, plus the solution on the block isn't too simple either. Good idea but hard to implement, especialy to look like an end product not some gheto mod.

[Fixittt]

Yes 1 mm endmills do exist, but for any type of fast machining, they just simply cannot be used. Unless you are proto typing in wax of something soft. Copper would snap those things quickly. I have tried carving in copper with 1/16th end mills, and the cuts have to be soo light. It would take forever to do. I have had good sucess and tool life with 1/8 th endmills. They are small, yet beefy enought to do some good depths in copper. That is why I suggested that size to Ben. Now if you had the time, and are using CNC you might get away with 1/16 with light cuts. But I wouldnt recomend it, cause the cost could start to stack up.

[bigben2k]

Morphling: I figured that you had already explored this idea, from seeing the work that you've presented in the past.

Fixittt: so what do I do? Let me run the math on the quantity of copper over the core; I might yet go with 1/8 (3.2mm) endmill.

Alternatively, I could finish the critical area (over the core) by hand, since it seems to be the most trouble.

dunno:shrug:

[Fixittt]

Ok, here is another solution. Rought out most of the block with the 1/8 endmill. As much as you can. And maybe, just maybe, you can step down an endmill size, say finish it with a 1/16th If you use high RMPS slow feedrates, and shallow depths, then maybe you can have it go into the tight spots and work them out.

catch waht I am saying? Also that V shape would look awsome IMO.

[bigben2k]

I believe that you mean this:
(the 1/16 might have to be shallower)

[bigben2k]

The problem is that it really throws off the aspect ratio, of copper to channel, which I'm trying to maintain between 1.5 and 0.75 . The fins could be fatter (overall wider), but I'd have to calculate the fin array effectiveness, in that configuration.

Maybe...

[utabintarbo]

What if they were tapered from top to bottom. The larger cutter will take most of the mat'l, and the smaller can finish it up.

Bob

[Fixittt]

BINGO!!!!!!!!!!!!!!

[Cathar]

Originally posted by Fixittt
Yes 1 mm endmills do exist, but for any type of fast machining, they just simply cannot be used. Unless you are proto typing in wax of something soft. Copper would snap those things quickly. I have tried carving in copper with 1/16th end mills, and the cuts have to be soo light. It would take forever to do. I have had good sucess and tool life with 1/8 th endmills. They are small, yet beefy enought to do some good depths in copper. That is why I suggested that size to Ben. Now if you had the time, and are using CNC you might get away with 1/16 with light cuts. But I wouldnt recomend it, cause the cost could start to stack up.

Hmmm, when all you have is a hammer, everything looks like a nail eh?

Am getting my batch of waterblocks made up with a CNC mill with a 1mm mill bit. Works just fine and not too slow at all.

Requires a very high cutting speed and the coolant injected at a high velocity at the cutter (or so I'm told).

Not that I have any real experience with CNC milling, but the guys I'm using are perhaps the best in the business in my state (Victoria, Australia) and they didn't seem to think it would be any major problem for them when I quizzed them on the difficulty of machining copper with 1mm bits.

[morphling1]

Yes no problem milling with 1mm bit if you have highspeed cnc mill (different price class then regular cnc mills). At 1mm, to have suggested cutting velocity on cutting edge of the bit, you need to have more then 10000 rpms and with those rpms good torque/power so rpms doesn't just stall.
So if you try something like that on lower rpm, and have feed rate too big that would mean, bit can't cut, but rather push the material, which it can't , so it breaks. So to compensate you need to go very slowly ->milling one block for hours->too expensive

[MadDogMe]

morphling1 bb2k, I thought of this idea long time ago, when I first think od radial design block, but it's not that simple to implement, somewhere on that line you will have to separate those to lines and for that you will need special piece like a block where that hose and a hose within come and then separate, plus the solution on the block isn't too simple either. Good idea but hard to implement, especialy to look like an end product not some gheto mod.

You could do it by having a reservoir with a 1&7/8 barb one side and a 1/2 bulkhead fitting the other(plus a normal 1/2 barb for the block outlet/res to vent from), the 1/2(block in) tube would carry through the 1&7/8 barb to the bulkhead fitting then on to the pump outlet, the 1&7/8 tube would empty into the reservoir. Ta~Da! ...

Was this what you had in mind BB2K?, or were you going to make a more exotic reservoir?. I'd be interested in other ideas you had, even scrapped ones.(how many people have scrapped a 2Tube aproach?, I did, passed over it in a few seconds as unworkable )
I did'nt realise you were going for plexi from the pic. (it takes me a while to realise whats what sometimes without pics and BIG LABELS :shrug I thought it was so you could see detail underneath ...

PS. do you think there's a way you could implement multi in/outlets like BRs with this aproach then?...

[morphling1]

Don't be to critical about my hand drawing skils, but I currently don't have any normal 3d graphic card, so drawing in ProE isn't realy comfortable for me

This is the way I would do it, to be as minimalistic as possible, still a lot more complicated then normal two separate hose conectors, but water flow through this block would be realy restriced only where it counts.

[bigben2k]

Actually, I really admire your hand drawings, I wish that I could draw like that!

That's very much what I had in mind, but I didn't care to make the outer hose any smaller. In your representation, there's actually two extra 90 degree bends (sort of), and that's exactly what I'm trying to avoid.

MadDog: see Morphling1's drawing. I actually had top outlets in mind, but I wouldn't want my hoses to kink against the side case, so I'll have to switch to 2 side barbs. It means a 90 degree turn for the 1 7/8 flow, into a 1/2 opening, but that shouldn't be so bad.

Some progress, next post.

[MadDogMe]

I was thinking of doing something like the drawing of splitting the pipes you've done(but it would have been integral to the block top), but I'd have had the inner carry through and the outer vent from a barb next to the inlet, I gave up cause it'd been to 'tall' once I'd allowed enough distance for the flow from the 8 outlets(4 for all intents, cause each two are joined) to 'equalise', because the outlet would still be offset...

BBs going to use the entire diametre of the block as his outlet 'barb' I think, or that was the impression I got...

I like the idea of doing it small as a fitting!, that's excellent , I'd have both barbs on the same side you have the 'outertube' outlet, to do away with any 90Deg bends.

*EDIT* You beat me to it! ... I'm going to have to work out how my scanner works!, I like that 'on the fly' look as well ...

[bigben2k]

Utabintarbo pointed out to me this morning that the 1/8 channel that the minimum endmill dictates is too large, and after recalculating (I made a sum error), it's apparent that there would be only 8 fins left (and 8 posts on the outer edges, maybe).

So the CNC option is out, unless someone's got a very high speed (10'000 rpm??) CNC setup that can handle a 1mm endmill.

In the mean time, I've been exploring the possibility of dremelling this thing by hand. I found plenty of bits at www.widgetsupply.com but I'll need to run this by you guys:
Here are the bits I picked out, and pics:

type: Diamond burr

Cylinder shaped diamond burr, 1/16" in diameter x 1/4" long, with 1/8" shank.


type: Carbide burr

0.8mm x 3.6mm Densply Maillefer Cylinder Carbide Burr, 1/16" shank. Approximately 3/4" overall. Made in Switzerland.


type: HSS

0.8mm x 4.1mm Densply Maillefer HSS Cylinder Burr. 1/16" shank. Approximately 3/4" overall. Made in Switzerland.

The pink stuff is a lubricant.

My little Dremel is rated to 35'000 rpm, but with little power. My question is: can I do this, in 1 mm, with the right Dremel?

[MadDogMe]

Utabintarbo pointed out to me this morning that the 1/8 channel that the minimum endmill dictates is too large, and after recalculating (I made a sum error), it's apparent that there would be only 8 fins left (and 8 posts on the outer edges, maybe).

Really?!, my design is an 8fin one , would I be able to optimise for 8 fins, the ideal width of fin I mean?. or does it not work with a fixed amount?, I mean the amount is dictated by the optimisation?...
************************************
I think you could do it but not totaly 'free hand', if you made up a 'suport' for making straight lines maybe?, to keep the dremel upright and 'carry the line' so to speak...
You can get a gizmo that makes a router from a dremel can't you?, I wonder?...

Can you not afford to take this to a specialist like Cather did?...(I know I could'nt )

[N8]

Originally posted by bigben2k
Utabintarbo pointed out to me this morning that the 1/8 channel that the minimum endmill dictates is too large, and after recalculating (I made a sum error), it's apparent that there would be only 8 fins left (and 8 posts on the outer edges, maybe).

So the CNC option is out, unless someone's got a very high speed (10'000 rpm??) CNC setup that can handle a 1mm endmill.

In the mean time, I've been exploring the possibility of dremelling this thing by hand. I found plenty of bits at www.widgetsupply.com but I'll need to run this by you guys:
Here are the bits I picked out, and pics:

type: Diamond burr

Cylinder shaped diamond burr, 1/16" in diameter x 1/4" long, with 1/8" shank.


type: Carbide burr

0.8mm x 3.6mm Densply Maillefer Cylinder Carbide Burr, 1/16" shank. Approximately 3/4" overall. Made in Switzerland.


type: HSS

0.8mm x 4.1mm Densply Maillefer HSS Cylinder Burr. 1/16" shank. Approximately 3/4" overall. Made in Switzerland.

The pink stuff is a lubricant.

My little Dremel is rated to 35'000 rpm, but with little power. My question is: can I do this, in 1 mm, with the right Dremel?

We use .020" diamond burrs (0.50 mm) and larger ones, like the one you have pictured to drill through glass, sapphire, ruby, etc. They do break very easily and would be a total bitch to put a side load on. We go through a lot of them just drilling straight holes. I also have the 35,000 rpm dremel and about 10 years of dremel experience. You can do just about whatever you want with it, IF you have the time and very strong, steady hands. It takes a long time and is easy to screw things up with a slight slip. They do make a Dremel drill press that would be ideal for aligning and holding the Dremel while you feed the material very slowly. Don't forget that you can probably use the fiberglass reinforced cut-off wheels to slice straight lines into the block, instead of 'milling' with the Dremel. The only problem is that those are harder to control if they hang up at all. Maybe worth a try.

[bigben2k]

I happen to have the dril press attachment, as well as the router table...

Diamond bit's out. Thanks for the tip, N8!

[Fixittt]

Ben I am having my CNC machine retrofitted with a 1 hp 30,000 RPM black and decker router motor. HAving the motor mount made.

We can try the smaller bits while still roughing the design out with a larger one.

I also have a large RIO pump that I use for my flood coolant. And I have access to a few rather large and expensive HAAS cnc centers.

Get me a CAD drawing.

[bigben2k]

Hum...

With a 1/8 endmill, I'm down to 8 fins and 8 posts. It's a no-go.

The copper to channel ratio has to be 1.5:1, with a fin height of 5mm, unless I try to calculate another fin height which might allow a thinner fin, but I doubt it.

It's going to have to be 1mm, or I'll have to do it by hand. I looked at the machining parameters (here ), but I'm having a hard time with it. I could calculate the length of all the required cuts, if it's any help.

What's you ropinion on the viability of doing this by hand?

Here's Utabintarbo's interpretation. The fin pattern is different than mine, but let's see if we're talking the same language, so far!

[N8]

Originally posted by bigben2k
I happen to have the dril press attachment, as well as the router table...

Diamond bit's out. Thanks for the tip, N8!

The other disadvantage with the diamond coated bits are that they would 'gum up' with the copper pretty quickly and then do nothing but bind. They can be cleaned with SiC (silicon carbide) dressing sticks, but just another PITA.