3 Attachment(s)
Instead of sucking a bunch of hot air, try this bong instead!
After reading a thread here on this forum, I decided to pursue alternative method of building a bong. The primary feature of this bong uses a vacuum to boil the liquid-coolant, to bring about our cooling effect.
My liquid-coolant chambers are nothing more than Pentek whole-house water filter housings. These can be readily obtained online in many locations. There are many options to choose from. Mine is the larger 20" tall model, with clear sumps.
I got my housings earlier this week. All I could do was grin as soon as I started looking at them. They're sooo BIG and absolutely BEAUTIFUL. They play the part of looking like plasma coolant chambers from some scifi show. Now let me state that these components are part of a much larger project, in which I'm building a cluster with five motherboards all inside of a Pelican 1780 Transport case. The idea is to create something that looks like it would have been used on Stargate SG-1. So I'm looking for a definite alien appearance here.
http://www.hardforum.com/showthread.php?t=1428264
Now the one constraint in this design, is that everything must exist inside that case. Only the condenser or radiator will be mounted to the exterior. That's obviously necessary, since we have to dump the heat of the system outside. I mentioned a condenser, because I haven't decided on whether I will also be cooling my fluid with a Phase-Change system or not. But in this thread, I'm going to concentrate on the efficiency of the liquid-cooling loop using vacuum-induced boiling to release vapor and cool the liquid.
Now here's where I should point something out. The chief characteristic of this system, is the boiling of the liquid-coolant under a vacuum. But the means to produce boiling isn't limited to using low pressures. Boiling can also be easily achieved with ultrasonics, in the same way that ultrasonic humidifiers work. The major point here, is to get the fluid to boil. How, is irrelevant.
Others have tried using coolant-soaked wicks to produce evaporation by moving air across their surface. That is a fundamentally inefficient way of getting liquid to vapor-phase. One of the main reasons for this inefficiency is the ambient humidity the system operates in. Because of that inefficiency, a much larger (if not colossal) system must be built to get a lesser effect than using a vacuum.
The common technique of building what is essentially a swamp-cooler (with a bong), requires that air move over the surface of a fluid (aeration). But the ability of the air to capture and carry vapor away from the liquid is an indirect process. By definition, the area subject to evaporation by aeration, is significantly smaller than the entire volume of a fluid being caused to boil. In boiling, we're using cavitation within the fluid. The only difference here being that we're not adding heat to our system. We're dropping pressure to lower the boiling point.
So now, I will introduce you to the components of this system.
Shingoshi
An explanation of the components...
The picture of the Pelican should be obvious. The case has the following internal dimensions:
42.00" x 22.00" x 15.10" (106.6 x 55.9 x 38.3 cm)
The Pentek 20" and 10" filter housings are pictured with their clear sumps and blue caps. Mine have black caps instead. The housings are actually sold by Watts. You can see the dual opposing ports here. They have a 1" NPT female thread. They have a fairly heavy construction. And since I'm using them as vacuum chambers, pressure won't be as much of an issue. I'll just have to be concerned whether these can implode by heavy vacuum.
I kind of got ahead of myself (I wrote the paragraphs after this one, before giving the essential background). I should have explained the purpose and function of the injectors. So I'll give you this link to Wikipedia to explain them: http://en.wikipedia.org/wiki/Injector
As you will see from reading this, injectors are capable of drawing an additional fluid into an existing stream. In the end, the two streams are combined. This happens because the injector is creating a vacuum on the second fluid.
Quote:
Now, here's the special treat. I found these injectors on the net. And the first time I saw them, I simply gasped. Not only are they designed as I need them to be, they're also damn gorgeous. But then, there's always problems with realizing one's expectations.
The Mazzei Injectors come in many sizes. But only the 2", 3" and 4" models have suction ports larger than 1" NPT. The model that first got my attention was the #3090. However, I couldn't tell from the pictures just how beastly that thing is. It has 3" male NPT intake and outlet ports. The suction ports are 1.5" male NPT. As you can see, there are two of them. That's unusual. In fact, they're patented.
The intent, before finding out just how large the 3090 is, was to mount it between the two filter housings having fluid entering from the other two ports. However, now that I know just how large they are, I'm inclined to use the smaller 2" models. In that case, I will mount the housings directly together in series, with the Mazzei Injector mounted on the end port of the pair.
But when I thought about it some more, I realized that I really want to draw vapor from BOTH housings in the same manner that I intended to do with the 3090. In this case, I need a tee connection between the housings, with the 2" injector mounted in the middle.
Now that I went back and corrected my omission, I'll continue with the outline of this project.
The purpose here is to pump liquid-coolant through the injector at high-pressure. That high-pressure flow will cause a vacuum to be created in the housings. The liquid in the housings will then boil, and in result lower the temperature of the liquid. It's with this fluid that the system will then be cooled.
Shingoshi
Just for the benefit of everyone...
It should be pointed out that injectors of smaller size can be used in any liquid-cooling system to increase the flow of liquid through that system. You can always find an injector small enough to benefit your cooling loop. So, they're not just for creating vacuums. Although, they do do a good job of that too!
Shingoshi
I think it's the same as yours...
Quote:
Originally Posted by
Naja002
What is the liquid/fluid going to be?
I'm using an antifreeze mixture. But I'm considering other fluids as well. In the course of my research, I came upon antifreeze that's formulated from soy. I guess they create an alcohol with it which is nontoxic. They use it in the food industry. Especially for the production of ice cream. It has the special characteristic of causing the formation of smaller ice crystals. That would be beneficial to me, since I'm thinking of running a slurry through my system. Slurry, as in kind of like the consistency of a milk shake. But in any case, I'm looking for fluids that will support much lower freezing points.
Shingoshi
I just called to verify delivery of my case...
After getting the email from Pelican Tuesday that the case was on the way, I called Optics Planet to see what was happening on their end. Hopefully, I'll my case by next Friday. Hopefully!
I really can't do anything else without having that case. I have to map out the placement of the components I want to use. I'm trying to build the system with as little tubing in it as possible. I can achieve that by directly connecting as many of the components together. Only using tubing if absolutely necessary.
And truthfully, I'm not rich. This is possibly an undertaking I would have been better off not trying. We'll just have to see what happens. If any unforeseen problems, especially any unrelated to this project occurs, I'll be set back for months. It's happened to me before. But I felt if I shared this, someone else might benefit from seeing how I developed the concept here.
Shingoshi
Operator, operator! A direct connection please!
Most of the components in the system use NPT fittings. So I'll be able to connect the following components without tubing between them:
1.) Watts/Pentek Filter Housings = Reservoirs
2.) Mazzei Injector #2081A/2083X = Vacuum Pump
3.) Danner Hy-Drive 4800 GPH = Liquid-Coolant pump.
With the use of NPT adapters, most of these will connect directly. If I remember correctly, you don't use solder, brazing or similar means to connect PVC!
If I don't go the sub-zero route as we've discussed before, the use of other connecting techniques won't be required. I don't know yet. I haven't made up my mind.
Shingoshi
I think you missed a few things...
Quote:
Originally Posted by
Shingoshi
With the use of NPT adapters, most of these will connect directly. If I remember correctly, you don't use solder, brazing or similar means to connect PVC!
If I DON'T go the sub-zero route as we've discussed before, the use of other connecting techniques won't be required. I don't know yet. I haven't made up my mind.
Shingoshi
I think of PVC welding as a form of gluing. But I may be wrong there too. The fact is, I want NO heat sources near this case. Especially not open flames.
You missed what I said here. I said I'm thinking of NOT trying to obtain sub-zero temperatures. Some of the stuff you've said elsewhere really hit home. I understood what you were saying, and took it into account. So I'm trying to be more pragmatic about the capabilities of the build. The realities of economics are substantially influential here. I get it.
No. I don't expect to get sub-zero temperatures from using low-pressure boiling. I know that's how it's achieved in refrigeration systems. But this is an entirely different beast.
As far as ultrasonics go, I'm not sure about that either. Not because it would be difficult, but because it brings a whole new level of cost to the project. Although, ultrasonics in the form of humidifiers can be brought cheaply on the internet. You really wouldn't need too many of them to contribute to the liquid's cavitation.
Shingoshi
What I can do, depends on what I can have...
Quote:
Ummmmm, No, that's not hard to understand at all. What's hard to understand is how any of that keeps you from providing info on materials usage, system expectations, etc.
I will be limited to the performance of the combined components I can fit in this case. I just don't want to say one thing, and then have to eat my words, so to speak. So again, I'm trying to be conservative about my expectations and goals.
Not only is this case not very deep, neither are my pockets, if you know what I mean.
Shingoshi
Maybe you can help me with this...
The NPT ports on my reservoirs are 1" females. Pentek makes a head that has 1.5" female NPT ports. But supposedly, it won't fit on my sumps. So last night, I got the idea of tapping the existing threads to 1 1/4". I think there's enough room to do that. But I want to know what problems could I create by doing so?
1.) Could I weaken the ports too much. I don't see how. Because the 1.5" heads don't look any different than the heads I have.
I looked for the prices of 1 1/4" NPT taps. And they're almost as much as buying another smaller 10" filter housing with the 1.5" ports as standard (http://filter.filtersfast.com/search...core&view=list).
But I'm not sure if I would have the room to add yet another reservoir, no matter how small it is. It would be advantageous. Because then I wouldn't have so much trouble figuring out how I'm going to mount my injector. But again, I have to wait to see if it's even possible. I almost want to return these housings. But that would really kill my spirit in doing this. Because the cosmetics are integral to the build itself. I want to see the liquid boiling in the reservoirs. Just like scenes from old TV shows. Frankenstein and the like. So this is also about emotion. I want to continue to be fascinated with this build.
That's why I don't want to give any specs. It would just be premature.
Shingoshi
A challenge to all liquid-cooling builders...
If you've read the thread well, you will see that Mazzei Injectors will increase the flow of liquid in your system. There's one model in particular which I think most builders can use without modification to their systems. That model is the #484. It has 1/2" MPT fittings on both the intake and outlet of the injector. The suction port is 1/4" MPT. With those fittings, just about anyone can use this.
So the challenge is to get one, and see how much your temperatures drop as a result of the increased flow through your system. It will increase your flow 5 times over what you have now!
Just search for "Mazzei Injector #484". You could even do a product search for quick results.
MAKE SURE TO USE 1/2" BRAIDED PVC TUBING, AND RADIATOR HOSE CLAMPS!!
Like the kind you would get from Lowe's.
Good Luck!
Shingoshi
