WaterBlock is all done!Check it

[a1cnolan]

Okay guys so here are all the links to the pics. oh and i know i need more lapping. i only went to 400 so far.

To fill you guys in. I decided to use press fits and layers to make this block.

The reasons why i chose to make press fits was threefold. 1 is that when you press something in with a hydraulic press you know exactly how much pressure it took to get it in there and thus you know how much it will take to get it out.

number two is that when it saved on manufacturing costs(less hardware to purchase)

and 3 is that i liked the idead of not having to use o-rings and the space i saved because of that and not running bolts through the block. This definitly allowed for more fins(because my .093" cutters all broke) and it hopefully lessened restricion

The block has been pressure tested to 25psi because Captain Slug said that 20 would be enough although it took about 3 tons of pressure to get everything inside so i don't think its going to break.

i haven't tested it yet because i don't have the rig i plan on running it on yet but as soon as i do ill let you know how it does.

The reason i used layers is that i didn't have copper thick enough.i was stuck with 1/8" so i had to adapt and overcome..

For the aesthetics i love the contrast on the bottom between the copper and aluminum because it almost looks like one piece of metal. The center hose barb is aluminum-bronze and the outers are made from what we call naval brass..(all it is is regular brass but it seems to have a different sheen).

Anyway hope you like the pics

http://www.imagedump.com/index.cgi?pick=get&tp=446245

http://www.imagedump.com/index.cgi?pick=get&tp=446247

http://www.imagedump.com/index.cgi?pick=get&tp=444072

[clocker2]

Looks very nicely made...with the exception of the barbs.
A bit of shaping would have gone a long way on those.

Have you tried it out yet?

[a1cnolan]

Well i liked the idea of smooth hose barbs. i think they look good and will work fine. Ive used straight barbs on other things before around the shop. fixing coolant systems and what not and found that they work just as well in most situations. If it does become an issue, they are easily replaced. they are not threaded in. just pressed to .005" which is .002" more than required for a water tight fit. the base plate is .062" thick. the finned plate is .126 thick and thus the fins are too. The top copper plate is .060 thick and the slot allows the water to get all the way out to the ends due to the 1/2 inlet against the thin slot. the hope is it will create a pressure when entering the fins like a small scale version of putting your thumb over your garden hose and then spraying it at the inside of a bottle cap. basically it is constantly flushing with new water at high velocity...

As soon as i can get my new rig i will get my temperatures posted. The things im trying to decide on now are the pump/radiator. My goal is to make a rig that is $75 or less that performs better than anything out there. I want to max out the efficiency of the block by maxing out the heat removal from the radiation. My eventual goal is to design a block that can be at room temperature at idle.

[caddy359]

looking mighty fine nice job - looking forward to some tests and rig pics

[learning curve]

How would you get it apart for cleaning when corrision begins? The copper and aluminum will react sooner or latter.

[Captain Slug]

How would you get it apart for cleaning when corrision begins? The copper and aluminum will react sooner or latter.
He alodined the aluminum, which is the same process that's used to protect Navy aircraft from salt water corrosion.
My eventual goal is to design a block that can be at room temperature at idle.
Not even remotely possible with straight water-cooling. Even those that are using geothermal or well water cooling are getting their idle temps to ambient. 100% thermal efficiency is impossible.

[a1cnolan]

okay well really close

[a1cnolan]

oh by the way cap....i have so many questions that i don't know how to word that its ridiculous. im realy interested in designing waterblocks. but im really into finding new ideas. not just reinventing the wheel. the problem is the more i read the less i can get a hold on a good direction to go with it..

the only problem is i just don't know how to ask my questions really...but thanks a ton with the help designing this...it was a blast to make.

[Captain Slug]

the only problem is i just don't know how to ask my questions really...
Well I don't really have any kind of scientific background, I've been teaching myself everything for quite a while so if you find your questions too long or cumbersome to post in a thread you can just PM them to me.

[damarble]

Well shoot, I wanted to check it out, but the site is blocked at work. So I'll have to make my comments when I get home.

Edit - Looks really good to me. Can't wait for some temp readings.

So the block is only about 1/4" thick without the barbs? Nice.

[=ACID RAIN=]

He alodined the aluminum, which is the same process that's used to protect Navy aircraft from salt water corrosion.
Yeah, but if it was 100% effective, wouldn't several block manufactures be using it? I'd still prefer all copper/brass systems because with the above block, one scratch and your protection is gone. Sorry, I just will never have real faith in any coatings when it comes to this hobby.

Block looks great, but minus 3 out of 10 for the aluminum IMO.

[Captain Slug]

Alodining and anodizing can penetrate the surface as deep as 1mm so you'd have to put a serious gouge on it. And how you would manage to do that inside of the block is beyond me.
A few block manufacturers (like Viscool) DO in fact use anodized aluminum tops. Most do not because it adds cost to the block. Derlin or Acrylic are much cheaper in comparison and don't require any post-machining processes.

Alodining is not commonplace because it's nowhere near as cheap or easy to do as anodizing can be. He made the top out of aluminum because he gets it for free from his workplace, which also has the equipment to alodine it.

[a1cnolan]

Yeah I wasn't trying to design a waterblock for mass manufacture. I was trying to design one for maximum heat transfer given the materials i already had handy. I also wanted to look into press fits as an alternative to all the bolting and o-rings that are done currently in the field of waterblock machining. I think they look cleaner and are more pleasing to the eye.

The block is totally about .400" thick.

On a side note i was cleaning out the coolant for one of our machines in the shop the othrer day and it occurs to me now that NONE of those chip are corroded even slightly...We haven't changed the coolant in that machine since it arrive 5 years ago. Just figured id mention that a good anticorrosive can take away all your worries.

[hitbyaprkedcar7]

Looks good :) Cant wait to see some temp readings/pics of rig :)

[FooTemps]

I like the press fit idea. It looks awesome and the profile is low too. I think this would be perfect for sff systems.

[ziggo0]

I like the press fit idea. It looks awesome and the profile is low too. I think this would be perfect for sff systems.

...or that chipset on the DFI Ultra-D :)

[Just Some Dude]

Your aluminum to copper joint is seamless.

Excellent work.

[[O-CuK]Marci]

He alodined the aluminum, which is the same process that's used to protect Navy aircraft from salt water corrosion.

And Swiftech anodised the MCW462 and MCW5000 series blocks - however, they still ended up with MASSIVE amounts of corrosion, hence they moved to copper or powdercoated tops for the MCW6000 series. Same goes for DangerDen with the original Maze4GPU block - their tops were anodised, however, more or less all of them ended up a mess after a relatively short amount of time, and thus they shifted to Delrin.

Anodising isn't enough to protect unless it's done absolutely perfectly. Slightest imperfection in the anodising and the corrosion will eat it's way from the inside out so that the anodised layer eventually flakes off.

[damarble]

Marci']And Swiftech anodised the MCW462 and MCW5000 series blocks - however, they still ended up with MASSIVE amounts of corrosion,

His block was alodined, which was not a mispelling of anodized.

Alodined > Anodized

[WhiteFro1306]

very nice work

[a1cnolan]

Your aluminum to copper joint is seamless.

Excellent work.

Well the press didn't go as expected. Because of the fact that the copper was so thin i decided to go with a .010 press on all sides. This is quite litterally like trying to fit the square in the circle hole. I warmed the aluminum in the heat treating oven to 300 F and then i dropped the copper in the liquid nitrogen for about 10 minutes(more than necesary but i was smoking a cigarette and didn't want to throw it out just do get it done early). Then i pressed it and there was still some bending around the edges. I made a steel die that would fit the hole just right and pressed it until everything was flat. Then i would go to the next section. I pressed it so hard that when it was done the copper was thinner than it was supposed to be but i said no worries and locked the whole thing in a vice and milled it until it was flat. So that is why it looks seamless. i like the way it looks though.

As far as the anodizing. ive never anodized anything in my life. I have been doing aircraft maintenance on about 16 different types of airframes for the air force. As a machinist i do all the bushing work for every jet on my flightline. I would never put copper and aluminum together in a fashion through which it would corrode.

While corrosion is not my specialty it is something i deal with quite a bit.As far as i know it, the reason alodining is so effective is because aluminum is a porous metal. and because of this it can readily soak up certain liquids. When you alodine a piece of aluminum depending on the thickness it has to be dipped for different periods of time and that is why it is so effective.

I'm not sure if this block will be as good or better than anything out there. But I have been looking at the trend that the pc business is going through. Whether its vid cards or processors or ram or harddrives the way it seems every manufacturer is getting around the walls they hit is by doubling what you have already. For my next block i am thinking about dual inlets and quad outlets with possibly one really powerful pump or two separate pumps.

The idea right now is to separate the center with a 1mm wall as this one is and slamming basically the same arrangement i have now with twice the power...i would obviously mess with the fin thicknesses to accomodate but i think that is the only way i will be able to make a better block without upping the cost of manufacture because cutters are expensive as is and the only way i could ever do anything with smaller fins would be a laser engraver or a waterjet(which we have one on order but probably won't arrive until after i leave) because the if you put it on a graph and compared cutter life to diameter the cost to benifit ratio plummets exponentially..

p.s. thanks for all the input guys and thanks for the compliments. I am particularly proud of this one

[Captain Slug]

I forgot to say: Very nice work on the block.
"naval brass" is a brass composition that is 40% zinc and 1% tin. The normal Yellow Brass has 33% zinc and no tin.

And that's a very interesting idea. Nobody has contemplated or theorized a two inlet design yet because space is so limited. in the socket area. I have seen a few larger-bore center inlet blocks with four outlets though.

[a1cnolan]

I forgot to say: Very nice work on the block.

And that's a very interesting idea. Nobody has contemplated or theorized a two inlet design yet because space is so limited. in the socket area. I have seen a few larger-bore center inlet blocks with four outlets though.

Well firstly thank you very much. Secondly about the two inlet design. There is so much space saved by the press fit hose barbs rather than the screw in type that i am fairly positive i could fit two inlets..Another thing about it is that it allows you to make a hose barb as long as your drill bits are.so casically you could make them all different lengths which leaves a lot more room to deal with hose clamps and such. And it also allows for some other ideas i thought were quite interesting.But im not ready to present those ideas yet because i haven't formulated them in my mind enough to make them sound plausible. The basic idea is since you have as long a hose barb as you want you could build a case around the waterblock. basically attaching the board to the block and having all of the motherboard supported by it. Anyway im still working on the mechanics of it but once i have it in my head ill make sure to post it for ideas from people.

[razetek]

Nice work, I am in the liquid cooling business, I have seen a ton of blocks, and yours in quite stunning.

How much do you think the mass manufacturing machining cost would be?

[a1cnolan]

Mass manufacturing? Well i dunno. I think if you owned the machine it would be exceedingly cheap. If i owned the machines i think i could get it down to a 5-15 dollars a block. but that doesn't mean to say if you paid a machine shop to make the parts you wouldn't spend more.

I honestly made a lot of mistakes on this block and would be able to make it a lot more efficiently now...knowing what i did wrong. but thanks for the compliment. i also would have made the whole thing out of copper if i had the materials.

anyway i can't wait to get out of the military so i can get my own machines and have fun with them

[Joeteck]

How does one make something like this? What do you need? This is very interesting...

[Captain Slug]

How does one make something like this? What do you need? This is very interesting...
A basic understanding of what makes waterblock effective, CAD software to design your parts, the tools needed to do the work, and materials to make the parts from.

Your design should first be dictated by the tools you have access to otherwise you will run into great difficulty and cost trying to get your design made. Most current block are made using CNC milling equipment due to the complexities of the designs and the materials involved. CNC can lower the cost of mass production, but for single units can be quite expensive unless you have access to it to make it yourself. CNC work can cost anywhere from $60 to $150 per hour depending on the equipment, expertise, and materials.

There are very very few designs at present that the average person can make with more affordable tooling options. Cross-drilled blocks, while not being optimally efficient but still effective, can be made with a drillpress and a set of taps. The #rotor waterblock design (http://www.madshrimps.be/gotohowto.php?howtoID=34) can be made using a drillpress, rotary tool, and tapping bits (plus alot of time). It's a pretty decent block in terms of performance.
I'm working on a sligthly more modern block design (http://www.ocforums.com/showthread.php?t=471737) that will require only a drillpress, bandsaw (or hacksaw and mitre box), scrollsaw, and taping set that hopefully will perform similar to the newer revision of the White Water.

If you have a specific questions feel free to PM me. For background information on waterblock designing and a wide variety of design research I recommend checking out the waterblock design forum over at procooling.com (http://forums.procooling.com/vbb/forumdisplay.php?f=37).