Possible modified Cascade SS

[Aaron Burton]

Ok, so I have a tendency to go all-out when it comes to my computer. I'm currently on aircooling (slk-800 and a 92mm vantec tornado). I can no longer stand the noise, even all my case fans on 7v, so I'm moving into watercooling. My goal is for a completely silent machine, (at least not audible above some minimal ambient noise). I have these blocks of silver at home that my dad gets from X-ray techs when they come to recalibrate the machines, (it's 999+ silver) and I figured I'd find the top of the line water block, buy it, then through lost wax casting, recast the base in silver. Since then, I've become quite familiar with Cathar's work, including the Cascade SS. So here's what I'm going for-- It's very similar to the cascade, but I found that after Cathar got rid of the secondary outlet, he never redesigned the outlet resevoir in the polycarb sheet. That didn't really make sense to me, because on the side opposite the outlet nipple, there's a big pocket of stagnant water. I realize that circulating water as quickly as possible is kind of missing the point, but I don't see any purpose being served by this pocket of warm water. Using what I know of elipses, I cut off the pocket, and replaced it with a circle. Ellipses are very curious in that anything originating from one focus (in a straight line), will be deflected by the edge toward the other focus. So I put the inlet on one focus and the outlet on the other focus. The circle I used to cut off the heat-pocket is consentric with the focus of the original ellipse. This way, once the water is propelled into the cups, it will be more easily pushed toward the outlet. In addition, I rearranged the number of jets, and increased the number to 100. The size of the array is a *little* smaller than the width of Cathar's original array. I also fixed the plenum so it would be a little more appropriate to the new array shape.

I have a couple reservations- First of all, I'm not sure if I'm ripping off Cathar's (and everyone elses in the forum)'s ideas, or just enhancing them. I'm also not so sure that my methods are really sound. I feel like I have the ideas right, but fluid dynamics are a bit more complicated then I think I'm giving it credit for. Any suggestions? comments? Help would be much appreciated.

[bigben2k]

The statement about the ellipse is correct, but I don't believe that it applies to fluid dynamics, especially not in this particular case (but then what do I know? ).

As has been doscovered by many people now, The cascade is easy to imitate, but very difficult to reproduce: there's a fine balance between jets and cups that must be tested, to find an optimal configuration (again, if you have money/venture capital, *call me* !!! ).

Cathar has been trying to get more jets, but is hitting close to fabrication limits, if he hasn't hit them: 100 jet may be possible, but it's not going to be easy.

Good luck, and keep us posted!

[Aaron Burton]

Originally Posted by bigben2k

The statement about the ellipse is correct, but I don't believe that it applies to fluid dynamics, especially not in this particular case (but then what do I know? ).

As has been doscovered by many people now, The cascade is easy to imitate, but very difficult to reproduce: there's a fine balance between jets and cups that must be tested, to find an optimal configuration (again, if you have money/venture capital, *call me* !!! ).

Cathar has been trying to get more jets, but is hitting close to fabrication limits, if he hasn't hit them: 100 jet may be possible, but it's not going to be easy.

Good luck, and keep us posted!

Yeah, I kind of figured I didn't really know what I was talking about :-p But I'm aware of what Cathar's been up to, and he's been trying to increase the number of jets to ~150. I'm not trying for nearly as many. Even so, that's not really the only change I made. What do you think of the idea of essentially lopping off one end of the waterblock? Also, what do you mean by "again, if you have money/venture capital, *call me* !!! " ? And still, Am I ripping off yours and Cathar's designs?

-I almost forgot- The reason I figured that the ellipse technique would work was because the water is at such high velocity near the jets, that though the streams would be turbulent, they would still be *vaguely* laminar. That's why I said that the ellipse shape would "help" push it along. I don't expect it to do a ton of work, but anything helps, I suppose.

[pdf27]

Originally Posted by bigben2k

The statement about the ellipse is correct, but I don't believe that it applies to fluid dynamics, especially not in this particular case (but then what do I know? ).

Correct, but it probably doesn't make any difference in this case. Provided any stagnant water isn't affecting the flow over the area you want to cool then it's pretty harmless as far as the cooling goes.

Originally Posted by Aaron Burton

I feel like I have the ideas right, but fluid dynamics are a bit more complicated then I think I'm giving it credit for. Any suggestions? comments? Help would be much appreciated.

You probably don't need to worry too much about the fluid dynamics of the bulk flow. The critical bit is with the jets, and there you're just extending Cathar's principle so it shouldn't be a problem. With the bit you're changing just think of the water being totally incompressible and design it by eye - where does it look like the water is going. This will generally be good enough for your purposes. If you've got a clear top you could maybe add a few air bubbles as flow visualisation to check it once built, but TBH it really shouldn't matter.

Originally Posted by Aaron Burton

And still, Am I ripping off yours and Cathar's designs?

It really isn't likely to be a problem if you're only making it for yourself - in fact Cathar may well turn up and make helpful comments about your ideas (he has before). Just make sure you post pics when you're finished

[Aaron Burton]

Originally Posted by pdf27

Correct, but it probably doesn't make any difference in this case. Provided any stagnant water isn't affecting the flow over the area you want to cool then it's pretty harmless as far as the cooling goes.

But doesn't that mean that the heat from the stagnant water is being somewhat transferred to the bulk flow?

[|kbn|]

I was following the cascade XXX thread (the cascade with 149 jets?) and from what Ive read it seems more jets doesnt help that much, but improves the overall smoothness of heat over the core, because there are less hotspots between the jets.
One question I still have is does jet/cup size (with the right ratio) improve heat transfer much in terms of how it affects the boundry layer.

Also I dont know how easy lost wax casting is in silver, but it seems very hard to do in copper because of air bubbles etc.. I have enough silver for a few blocks but I couldnt find out enough about casting silver for me to try it. Have you done it before and what equipment did/will you use to do it?

[Aaron Burton]

I've done lost wax casting before, and believe me- it's not going to be worth your while. Once you buy all your materials, (wax, plaster, investment, investing table, crucibles, oven, etc. ) You'll be waayyy over the cost of just going and having it CNC milled. I once thought I could cast my SLK-800 in wax, and boy was I wrong. It's such a pain with something you need this precise, that it's just not worth it. RAM sinks and perhaps even a passive air NB cooler would work, but unless you already have a place to do it, it's not going to be worth your while.

About the water carrying heat but not conducting-- I'm not exactly sure what you mean by that. Since heat is just internal kinetic energy, in order for a particle to be "carrying" it, it needs to be conducting it. Perhaps I'm just misunderstanding you? Could you elaborate?

[|kbn|]

For the casting I was thining of different ways to do it. It would seem silver and copper could be cast in a steel mold, which could easly be made (if layered). There are plenty of DIY non expensive ways to melt metals too. The problem is with copper keeping oxygen away from it, I dont know if silver has the same problem. Thinking aobut the basics of casting silver - needs to get the silver past melting point, into the shape you want it, and then let it cool down, as well as easy to remove from the mold, anything Ive forgot?
Infact couldnt ceramic/clay be used as a mold? assuming it wouldnt crack at that temp..

Im not sure what I meant either..

[pdf27]

Originally Posted by Aaron Burton

About the water carrying heat but not conducting-- I'm not exactly sure what you mean by that. Since heat is just internal kinetic energy, in order for a particle to be "carrying" it, it needs to be conducting it. Perhaps I'm just misunderstanding you? Could you elaborate?

What is happening is that the water posesses internal kinetic energy (i.e. is hot). There are three methods of heat transfer - conduction, convection and radiation. In the case of a fluid at this sort of temperature, convection is the most significant. In a water block you effectively have forced convection - the working fluid (water) is being forced along. This overwhelms all the other heat transfer mechanisms within the water at this sort of temperature/water velocity.

Originally Posted by Aaron Burton

But doesn't that mean that the heat from the stagnant water is being somewhat transferred to the bulk flow?

The water is unlikely to be totally stagnant - there will be some mixing with the bulk flow. However, if it was totally stagnant what would happen is that the water would just sit there at the same temperature as the metal around it, having effectively no cooling effect.

I'm not quite sure if that's what you wanted to know. If it isn't, get back to me.