New guy with my waterblock idea

[hara]

Originally posted by Zymrgy

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

a 2 inch round hose has an area of 2026mm^2

The cone in the centre will surely direct the flow but structural integrity is less than if the fins were connected and also connected (long) fins have a tendency to pick up more heat than discontinuous fins (or so i've been told). Keep that in mind. And also I agree with bigben, taper the very end. If you decide to keep the cone, converging the pins to a point would help IMO.

The design is similar to this which I had in mind earlier.

[Zymrgy]

hara, my mistake, though you still got me lost.

area of a 7/16" dia...the size I am planning on making the initial hole for the inlet = .150 sq in

area of a 45mm square = .319 sq in

or am I just missing something?

[hara]

area of a 7/16" dia = 97mm^2

area of a 45mm square != .319 sq in but 0.07 sq in

[Zymrgy]

^^^^^....its too damn early for me to be doing math,

[hara]

Actually it's quite late here. As for performance, I don't think it will perform better than a WW but quite good nonetheless.

[Zymrgy]

yet another revised pic....


That one is not as tough to make as it might look.

[hara]

Going good! Another problem you'd be facing is the outlets. A radius style approach is the best IMO. Otherwise 1 in 2 out like you were going to do earlier.

[Zymrgy]

^^^yes, that really does not show up in my first renderings well. I am planning on doing a 1/4" radius (approx 6mm ) feeding both the outlets.

I am starting to look at this like you would the intake/exhaust manifolds on a car....on the intake side you want a bit of turbulance to help mix things up. On the exhaust side, you want as smooth as flow as possable to clear out the exhaust as freely as possable.

[bigben2k]

Hot damn that looks mean!!!

I think you've got it tweaked. Do you know the actual area of an AMD core? Or is this for a Pentium?

The Barton core is 7.47 by 14mm, if memory serves correctly. The fins must run 2mm past, and you can go with a 1.0mm baseplate.

[Fixittt]

Zymrgy

What software do you do your cad work in?

I have to say you got some nice skill there.

Just remember (Im sure you dont need to know this, but some might) that just because you can draw it doesnt mean it can be made like that. Just by the last revision Zymrgy has increased machine time, and added a few more tools. But damn it looks good. Also, have you thought about how your going to deburr it when your done? My thoughts would be media blasting it.


I truley understand about the material situation. Besides I would make 2 or 3 out of aluminum or wood before you ever touch any copper or brass.

[hara]

The pentium 4 core is squareish (11x11?), so is the palominio. Tbred 'B' is ~7.5 by ~11.7
Keep also this in mind. The core of the Amd is offset by 1.25mm. Damn Inches!

[Zymrgy]

Ok, I am getting close to final on this & a few things bother me. Right now the "fins" are 3/32" ( lets call it 2.5mm ) tall. This is done out of the limitations of the tools that I have on hand. The base plate is also the same thickness, 3/32" or once again 2.5mm. One concern of mine was the material strength. With the fins being set down from the top of the block there I am concerned about going any thinner. I have come up with a solution & comprimise, though. One idea is to make the center X portion 4mm tall & have the baseplate 1.5mm thick. This would mean that the center X would actually touch the top plate therefore supporting it. All of the other fins I would make 3mm tall. Maybe I will model this up like that later.

Yet another option that I am strongly considering is going with a 1.5mm wide channel/wall. Advantage is we have endmills on hand that will do the job easily. Disadvantage is I loose alot of surface area on the fins.

I think I will keep the same dimensions for the "square" around all of the fins. on the last model I posted, the flat part was 14mm square, with the ends tapering off to a 21.5mm square. My thinking on this is I really don't want to make a "core specific" block. This one covers quite a bit of area & should work well in a variety of situations...at least in my mind.

I am thinking I might have to actually buy some endmills to get this done right....at least something like a 1mm dia with 4 mm lenght of cut. Not a problem....or I could modify some of our existing tooling to do the job. These are the things that I was trying to stay away from, though.

[Zymrgy]

@Fixittt

I use CadKey for all of my modeling work. It is my CAD program of preference....but then I know how to use it pretty well. One could get the same results with Solidworks, ProE, ect.

For my toolpaths I import the model into Mastercam.

As far as deburring the thing, yes, Glass beading was exactally what I was thinking. I don't have the patience to do it any other way.

[hara]

Given your tooling limitations you will have to redisgn it for average performance. A 1mm base with 4/4.5mm fins is ideal.

[Balinju]

imo 2mm of fin height is too little, if i were you, i would go for at least 4mm

[bigben2k]

Ah! Let me introduce you to the 400XL

In the Radius thread, you'll find a link for 1mm (actually, they're 0.0400") endmills.

I'm about to buy 10, for $2.50 a piece. If you want in, let me know.

1/4" depth of cut ( >6mm), square cut, center cutting, 2 flute, solid micro grain carbide, without coating (not needed).

[Fixittt]

here is something that I dont think anyone has thought of before.

Why not leave 4 or 6 small pillars around the inter wall set that will come in dirrect contact with the top of the block. Think of it as pillar supports for a house on the beach if you will. Will add support with very minamal flow restrictions.


Ahhh Mastercam, I should be getting that this week. I use solidworks 2001

[Zymrgy]

@bigben2k...

yes, those cutters are the ticket. However, I would recommend going with the 4 flute rather than the 2 flute. At these extreme lengths, the 2 extra flutes add alot of rigidity to the cutter. Seems to be the exact same price on the webpage also. I would definatly kick in my $10 worth also. Let me know ASAP.

[Zymrgy]

One more pic....I think I might stay with this one, it just plain looks good.


The top is pretty much the same. The final depth of the channel is 4mm. Baseplate is 1.5mm. The center X portion of the channels is 4mm tall, ie flush with the top of the block, done for support. All the other fins are 3mm in height. This is done as a comprimise for water flow. The inlet will have the same + shape as posted in my earlier pics. The center square part of the fins is 14mm & they taper down from there. Looks to be tons of surface area to me.

[Balinju]

you know ben you will be getting those end mills??

[hara]

Originally posted by bigben2k
Ah! Let me introduce you to the 400XL

In the Radius thread, you'll find a link for 1mm (actually, they're 0.0400") endmills.

I'm about to buy 10, for $2.50 a piece. If you want in, let me know.

1/4" depth of cut ( >6mm), square cut, center cutting, 2 flute, solid micro grain carbide, without coating (not needed).

Isn't the shortest possible endmill the best? I mean the 6mm has a better chance of snapping than a 4mm. What does the tin coating enhance in endmills?

[Balinju]

TA (TiAIN) - Sole choice for dry, semi-dry & HSM machining requirements. High hardness and heat resistance for machining toughest materials.

TC (TiCN) - Very high hardness qualities with excellent lubricity. Multi-layered for higher heat resistance and toughness. Excellent for all kinds of steel.

TN (TiN) - Excellent general purpose coating for all ferrous materials. Good combination of lubricity, heat resistance, and wear characteristics.

[Balinju]

if you need more i can also tell you the highest working temp of each and the colour

[hara]

Yeah, but what. You can have a higher feed rate, a more deep pass or just make the endmill last longer with titanium nitrate?

[Balinju]

1. There are not all titanium nitrate but only the 3rd one is titanium nitrate coat.
2. these make the end mills more hard and resiste to more heat.

an example, if you are cutting in very very hard metal, the mill tends to get hot (cause of friction), and if not enough coolant is supplied then the mill can bend or break.

it could be also possible that the mill gets hot enough even with a good amount of coolant, the coating makes the mills resiste more heat before bending or breaking, for example TiAIN coating makes the mills resist about 850 degress celcius while TiN coating only 450 degrees celcius.

the coating is different for every purpose