New guy with my waterblock idea

[Zymrgy]

I have been pimping this idea around a couple of other forums & BigBen2k suggested that I try here...so here it goes.

The top is solid, with the "barbs" machined right in. The only tough part is the center inlet, with it being round on top & the cross on the bottom, with the transition between the 2 shapes happening along the full length of the inlet. Thats what wire EMD's are for. It is planned to have a 1/2 inlet & 2 3/8 outlets. The top will be anaodized alum to save $$$, & the bottom out of copper. Base thickness of the bottom is planned to be .093 inch. The base will start out at 1/4", the the pocket machined down. The "fins" are right now are set .06 down from the top of the block...I couldn't decide weather to make them flush or not. The plan is for them to be .040 (1mm) wide & the channels being the same width. The actual heigth of the fins is my biggest thinking point at the moment, right now the are set at .100" or so. . The idea came alot from Cathar's white water block, the most major mod being the inlet. My thinking is that a taper will allow for a smoother flow & hopefully better results that a immediate transisition from round to oval ( or cross ). There lies the challenge, tho. The only machine that is capable of making a cut like this is a wire EDM, which are not the fastest machines in the world. A cut like this would take 45min/ 1 hour. However, I have another idea that might work as well but I could cut on a mill, & that would be the inlet being a cone shape on the top to a certain depth & then I could conjure up some of my programming skills to taper out the cross on the back side. Even simpler would be to run a drill point down & the go to the cross.

I think I will call it the "X-BLOCK"






Thanks

A bit about me....I am a have been working as a machinist/programmer/network admin for way too long now. I am fortunate enough to have a boss that does not mind me doing "G jobs" on my own time. My biggest problem is finding the time.

[edit] shit...looks like I already posted in the wrong forum...shit again. Mods move as you wish [/edit]

[Tar Man]

Very good design you've got there

Only got 2 things to point out that youve probably already thought of; im thinking that the water flow will be a bit restricted at the points below? 3 channels (only outlined 2 each side - oops) meeting at 1 point can't be good for flow. Perhaps sacificing some fin area for more space at the edges would be a good idea?

I also find the tapered inlet to be a sound theory, just as long as you taper for the entire length of the inlet, having the smallest part of the taper as the opening.

Other than that, it sure would be a good performer!

[bigben2k]

Thanks for stopping by, Zymrgy.

Tar man: I think you've got the flow direction reversed: it's a center inlet.

[Balinju]

How about taking an idea from Ben's Block ?

If I were you, i would not leave the spot under the intlet "empty". As I said, I would either do what Ben did (sorry not did, is doing) in his block, i.e. a + shaped fin in the middle or otherwise I would include some type of cone shaped fin.

Maybe Ben could explain to you better than I am.

Am I in the wrong track or not??

[Balinju]

this is what i meant. add the red fin. Sorry for my paint skills

[bigben2k]

What Balinju is trying to say is, use flow seperators. It'll help seperate the flow nicely, instead of letting the flow break up over the top of the fin.

Also, you might want to extend the fins into a point, because a liquid that traverses a sudden rise/drop in channel size, is going to be more restricted than in a smoother transition.

Someone pointed me to a HS that would be perfect, to make a block like this. Heck if I had the time, I'd give it a shot.

Nice work, very nice work.

[BrianW]

Don't wanna be a stickler, but this thread may get better attention if it were posted int hte correct section.

BrianW

[Fixittt]

that is a very very nice rendering. Nice cad work.

I will leave the base alone and focus on the top of the block.

you say you want the fittings machined into the top of the block. Do you have any idea how thick the top peice is going to have to be to make that? and with all the extra machine time involved to make it. That and you will need a really long endmill, which will give you deflection. But that is no problem.

I dont really think the intergrated hose barb idea is worth the extra machine time involved. When drilling and tapping takes only a fraction of the time.

[Zymrgy]

The time in making the top piece with the integrated barbs is insignificant. Really. The material would start out as 1" thick stock & the barbs are planned to be 3/4" tall. Now, a consertive estimate on removing the material start to finish would be 2 minutes, if that. Aluminium is fun to machine. "Typical" feedrates start at about 100 inches per minute, at least on the machines I have access to. I am still amazed to this day at how fast I can turn a block to a pile of chips on them. As far as deflection, a 3/4" length of cut is fine for a 3/8 carbide endmill, which would easily fit between the barbs. Funny thing is, it will take the finish tool longer to cut than the roughing tool. Most of the material would be removed with a 1/2" or 3/4" roughing tool, like this one...


For the tapered inlet plan, I really don't think that the typical 1/4NPT barbs are an option, unless I want to make my own barbs

The machine I will be using has 30HP? ( not totally sure on that rating, but it is plenty ) & a 12,000 RPM spindle.

[Zymrgy]

& also, to Balinju & bigben2k, is this what you had in mind?


Now, all I did was switch the places from the "channel" & the "walls". I also added the tapered sides in, however, while not being impossable, they would be a pain in the ass to make that way, to say the least. Remember each channel is only .040 wide (1mm). Something more feasable might be to "step" the width down...this would allow me to make the fins taller as a bonus.

[bigben2k]

Nice to have such a machine handy!

No, I meant that you should taper the perimeter of the fin pattern, not the top of the fins.

Nice, very nice.

[Fixittt]

well, if you got that at hand and can play with it, then by all means go for it.

See the avitar for my name, that machine I have sitting in my garage. Its a maxnc15 Open loop system. While its alot slower and feeds are alot less, Its in my garage, so I can start it and walk away. Very handy to have. I also have a large manual mill too, for doing the easy stuff.


Also, if you have that machine handy, why not do it in copper?

[hara]

iMO, the circled fins are a waste of coolant as they are off the cpu.
Also, did you consider "squashing" the design? this is because AMD cpus are not square especially the Barton.

[hara]

And yes, much of the cooling will be done through the central fins.
This is a sketch of how the CPU is orientaited WRT the Holes (socket 462). This applies to Tbred A, B and Barton.

[hara]

This is what I meant:

Nice machines BTW

[Fixittt]

IMO

I dont really think that extra surface area is considered a waste of space. With a thinner base plate and thinner block.

[Sproket]

How the whitewater inspires us all.

[hara]

Originally posted by Fixittt
IMO

I dont really think that extra surface area is considered a waste of space. With a thinner base plate and thinner block.

Those little fins are 9mm away from the core* by my calculation. I could be off but by omiting them, he could reduce the nozzle further increasing the velocity where it matters. I could be wrong though.

@Zymrgy Keep your nozzle about ~45mm^2 of area. This will vary depending on the pump pressure.

*edit: centre of the core

[Balinju]

yes Zymrgy that is exactly what i meant

very very very ... nice machine btw, i would like to have one in my garage just as fixittt has

and also i agree with hara, make some calculations and the extra fins remove them because they could only be restring the flow of the water.

You need just a little bit more than 2mm of fins or 2mm of fins on each side of the core, ie, if the core is 10x10mm, then 14x14mm of fins would be enough

[Zymrgy]

Ok, I think I am getting the gist of what everyone is getting at. That & I am getting good at modifing my fins, lol. I decided to see what it would look like if I added a cone in the middle of the thing.



Looks pretty good....& should do something like direct the flow.

bigben2k...is that what you meant by adding taper to the fins?

originally posted by hara
Keep your nozzle about ~45mm^2 of area

am I reading that correctly...45 square mm? Wouldn't that be something like a 2 inch round hose coming in?

[Zymrgy]

Also, @ Fixittt, that would be a way sweet toy to have in a garage. nice. As far as making the top out of solid copper, well, we really don't work with much copper where I work, & none of the 1" thick stock hanging around. I am trying to make do with what we have....I can use material reminates for nothing. On alternative that I had not thought of before was making the top out of brass....but then brass is not as easy to machine as alum either.

The thing is, we send out parts for anodizing on a weekly basis. Most of those places charge a "lot" fee to do the work. So, if you want 1 part done, it costs $50. If you want 50 parts done, it costs $50. So, I could just slip my parts in for nothing.

[HMB]

That sure will be a neat WB, my suggestion, when u've settled on the findesign, make several bases with different thickness. That is if u have time to test them all

[Balinju]

My opinion on the second design with the cone.

I thin the cone is going to be difficult to mill, when our machinist was trying to mill the cone for our water block called squirrel, he had to stop in the middle of the process cause he thought that the pin was moving while being milled. But that would be a very nice thing to have inside a waterblock

another thing, just try to at least round the corners marked in red in the following picture. They cause cause some breakage of water flow.

[Balinju]

or otherwise you could also make them converge to a poing like a triangle if making them round is difficult

[bigben2k]

Actually, I meant that you should taper the very end of the fins.

Cathar calculated that the optimal fin-to-channel ratio should be between 1.5 and 0.75, so as long as you incorporate that over the core, and 2mm beyond, you're all set.

It looks very, very good.