Optimal-Air-trap-Design

[msv]

Originally posted by #Rotor

anyone care to enlighten me as to why an air-trap is required, if it's easily possible to purge all the air in a closed loop.

Rotor: I´m not the one to lecture You (even though the audience might find it amusing), so I won´t, BUT:
I bleed my system by tilting the case back and forth, turning it upside down etc. until all air has found its way to the T-connection for filling/bleeding.
Since my laziness by far outweights my physical strength my upcoming water cooling upgrade will include an airtrap instead of just a T-connection.
regards
Mikael S.

[V12|V12]

Lead Salad:

Hey, that idea looks very familiar! Like the one that I first
started using a while back. It's a solid and reliable design, BUT
the problems that I see with it are;

1) Not that I'm a stickler on bends, but that bend won't do a huge amount of restriction if you have a nice torquey pump. What it will do is make the water rapidly accelerate out of the tube and
straight into the top of the cap, which combined with the bend
WILL add up.

2) Having the incoming water entering at the top will give the air
a closer path to escaping...When I tried this, I found that it
works VERY fast/well. The problem is; The water will start a small
fountain stream, which will hit the top of the trap, which will
cause some of the water to flow around the top of the trap's rim,
causing a "rolling-effect" which sort of traps air (depending on
how close the water lvl is to the top) and mulches it up and when
the water falls back into the Trap it's mixed up with air-bubbles.

3) Another thing that I've noticed from my similar trap, is that
if you have the water level too low, the falling water from the
top of the trap will hit and splash, thus pushing air into the
water and eventually it will be sucked down into the inlet.
Again, I'm no expert, and since this isn't exactly like the one
that I made, it might not have any of these problems at all, I
hope it doesn't! Nothing is more frustrating than to spend hours
planning + building something that looks good on paper, only to
find out that once it's built, you've missed something or that it
just doesn't work the way you thought it might? I know all about
that! If I would change anything, I would take a look at Turbo's design, while it doesn't work super fast, it does work and won't create anymore bubbles than you started out with... Or you could try a "T" style at the bottom, so that the incoming water will still be exposed to the res water and won't fountain up? GOod Luck!

[bigben2k]

Yep, that's the essential flaw with that design. It has to be carefully tuned.

I know I couldn't do it: my flow rate is going to be too high.

[Althornin]

Originally posted by bigben2k
Yep, that's the essential flaw with that design. It has to be carefully tuned.

I know I couldn't do it: my flow rate is going to be too high.

Why dont you just add an extra bend to the top of the input tube?

[LeadSalad]

That's what I've ended up doing, adding a 90 deg copper pipe to the end of the pipe forcing the water against the wall of the trap. It seems to work well, but I can always just stick some pliers in there and remove the copper turn and flip the trap over to achieve what Virus was talking about.

[V12|V12]

I don't like "flaw" sounds negative... it's a GOOD design, you just have a "tuning" point that's not very wide, but once in that point, it's VERY effective and fast! Like my old one, it cleared bubbles in like 3min tops, VERY fast, but as long as the water level didn't drop below eh ~ 70% full and no higher than 90%, perfect. You could always try the "T" out on the bottom, say a nice 3/4-1+" would work wonders! Though it would take a little longer to filter, it would work well, and you'd have barely any in-Res turbulence. Time is not a huge factor to me, b/c hey, long as things are running and temps won't increase that much while things are filtering out, hell I'll go watch TV or play a game and poof, air is gone... That is of course with my other late attempts. This new res works fast and it's ultra quiet, plus TONS of lucite for Eye-candy... soon as I get some UV-LED's (expensive though) It'll be a BEASTLY looking wonder

Wait, one more thing you might wanna try, which would reduce turbulence and not hamper flow as much as bends is, what I call "flow-baffles." Just pieces of plastic etc, that prevent high flow areas, but are spaced apart enough to still allow an overall high amount of access to the outlet... I still believe that Turbo's+ a "T" would be the simplest and most effective, IF you want a true Resevior's function. Good luck!

Anyone know of a place that sell's CHEAP Clear Tubes, say around 3" diameter x "X" ft long? Oh and some CLEAR Epoxy that won't yellow or change colors when exposed to Anti-freeze? HAHA actually it's funny, my clear 5-min epoxy turns are sort of dark turquoise after being exposed to the water+anti-freeze, looks sorta cool though. Thanks!

[bigben2k]

A bit of progress:
http://forums.procooling.com/vbb/sho...&threadid=5613

This is not really an improvement on my design, it's more of an aesthetic re-work.

On the inside of the case, I'm going to have to put at least one elbow for the flowmeter, and for the inlet barb.

I picked up a 3" long extension to screw into the tee (inlet), so maybe it'll work a little better.:shrug:

[neoacid]

Well, I haven't had any experience with water-cooling, but I'm currently putting together my first, and I've done a lot of reading of posts because i want to have as few problems as possible. Right now, though, i will not be including an air-trap in my design, just a reservoir. Unfortunately, my case is packed and very space limited, so the reservoir is going to be at the bottom of the case, which probably doesn't make for a very good air-trap, since bubbles will rise to the highest point of the water circuit.

Back to the subject of trying to make an all-in-one air-trap though, I think you can minimize the production of bubbles from a standard reservoir that has top inlet and bottom outlet without having to keep the water level within a certain small range. You would do this by using a flat piece of material slanted at an angle underneath the top inlet that goes down to about 1/3" the total length of the res. to act as a sort of slide for the water to smoothly flow on and hit the surface of the water without splashing as much. I'm not sure, but i think microbubbles should also be able to escape easier because the water will be spread out into a thin film going down the slide. Also, the right side of the resevoir underneath the slide should be much calmer and allow additional bubbles to rise to the surface. Please excuse my very hasty drawing.

[V12|V12]

Okay, here is what I see, starting (thanks for posting!) with #1

#1 That incline angle is going to FORCE the incoming water+air (Lets call it "mulch") down into to the water. Now since the mulch is moving rapidly, it has to pass through air, which would lead one to think that the air in the H20 would be released... Nope, it's moving too fast. So now it passes through another barrier of resistence, the H20; which the red indicates a swirling that will happen b/c the surface of the still water is being penetrated by fast moving air/water, which will actually push down MORE air into the water.

#2 Now that the mulch has made it down further into the water, and has release some air, while pulling some down with it, now that incline (I assume it will be the same width of the tube diameter?) creates a flow bottle neck for the outlet going to X-location. When the mulch travels down the incline and past the "lip" of it, some of the water will again be moving very fast and some slower, which will create another looping, which again, will mulch up the remaining air that's along with it.

#3 Ah, the slower moving water+air will now be released on the other side of the incline? Argh... Since the incline is blocking half of the res, the suction side (outlet at buttom) will be pulling as much water down as possible, limited by the incline's distance from the tubing wall. So you will have a Suctioned side. Meaning that there is more restriction of flow on that side, so ANY water that comes past the incline will be sucked down FAST, now the air really has NO chance of escaping to the non inclined side.

#4 If you could change or play around with the incline's angle, that would reduce the swirling effect of the incoming water+air...


Lastly, if you could post some pics or an MS-Paint of your case and the obstructions that are in the way, I could get a better idea of what you could do to make an efficient res/trap.

[neoacid]

Thanks for the feedback. I think i didn't explain my concept too well, but that's okay. I also realize that my concept is far from the optimal solution, but I hope it would be somewhat functional.
V12|V12, I'm interested to know what you think about what i am about to say, so please post back.

quote:
[hr][b]
Originally posted by V12|V12:

#1 That incline angle is going to FORCE the incoming water+air (Lets call it "mulch") down into to the water. Now since the mulch is moving rapidly, it has to pass through air, which would lead one to think that the air in the H20 would be released... Nope, it's moving too fast. So now it passes through another barrier of resistence, the H20; which the red indicates a swirling that will happen b/c the surface of the still water is being penetrated by fast moving air/water, which will actually push down MORE air into the water.
[\B][hr]

Like i said, i know my design isn't optimal, but I do believe the slide would reduce velocity of the water falling in the reservoir due to friction (yes, small though it may be and even though the angle is fairly steep). Also, due to the fact that the water will spread out along the surface of the slide, this should help to keep as much air from being mixed in where the falling water interfaces with the pool of water. There will still be swirling, but this should be a lot better than just letting the water free fall from the top of the reservoir and take a direct path (water current) to the outlet at the bottom.

quote:
[hr][b]
Originally posted by V12|V12:

#2 Now that the mulch has made it down further into the water, and has release some air, while pulling some down with it, now that incline (I assume it will be the same width of the tube diameter?) creates a flow bottle neck for the outlet going to X-location. When the mulch travels down the incline and past the "lip" of it, some of the water will again be moving very fast and some slower, which will create another looping, which again, will mulch up the remaining air that's along with it.
[hr][\b]

Well, yes it might look that way, especially because my drawing was very poor. I already had in mind that side of the wall and the lip of the slide should not be too close together or this would create resistance. As long as the total surface area between the side of the wall and the lip is adequate, this will be no more flow restrictive than your tubing. Ideally, we want to slow down the water to release the air, so the more space the better, even though i don't consider the water to be moving slow enough at this point to really do any good. It is after the water gets past the slide that the water should slow down even more because it will be entering a larger pool of water. This larger pool of water should be less turbulent and more bubbles should be released here. This is also why I consider this part to be the air-trap side of the device. The first part down the slide acts more like the reservoir.

quote:
[hr][b]
Originally posted by V12|V12:

#3 Ah, the slower moving water+air will now be released on the other side of the incline? Argh... Since the incline is blocking half of the res, the suction side (outlet at buttom) will be pulling as much water down as possible, limited by the incline's distance from the tubing wall. So you will have a Suctioned side. Meaning that there is more restriction of flow on that side, so ANY water that comes past the incline will be sucked down FAST, now the air really has NO chance of escaping to the non inclined side.
[hr][\B]

I dont think this will happen anymore than in any other device. Of course the water will be moving faster the closer it is to the outlet. That will be true for every device, but not all of the water will take a direct path to the outlet. Some of the water will go directly to the outlet, and some of it will be allowed to flow to the rest of the reservoir. I also think the suctioning can be minimized by how far the slide's lip is from the bottom of the reservoir. (see picture)

For #4, a shallower incline angle will necessarily make the reservoir have to be larger (not desirable) to keep #1 from happening.

[neoacid]

um, guess i dont know how y'all do the quoting thing. I assumed it was html coding for a horizontal rule <hr>. Also, it's been a while since i code, so i put the wrong slash in my closing tag.

Anyway, forgot to post the pic for my explanation, so here it is:

[Althornin]

everyone wants to make a nice vertical cylinder for an airtrap - problem is, it hard to make optimal (as has been discussed).
Why not lay cylinder on its side, input on one end, out on the other?

[joesgarage11]

Here's my res:
Click for a larger image.



It's an acrylic box 8"x8"x2-1/2". The inlets and outlets are near the bottom separated by a 3-1/2" tall divider. Very few bubbles make it from one side of the res to the other.

It's attached to a passive cooled radiator made of 3/4" copper tubing and copper fins. View the Radiator Post Here:

[Zhentar]

I've finally got my air trap working quite well. Its a little large, but its a great res and air trap. I used a 2" PVC wye. at the two tips of the y, I put in a 2" to 3/4" reducer, where I screwed in two 5/8" barbs. at the top I added on a bit of PVC pipe, and at the top a screw cap. (actually, I havent put in the screw cap yet) It clears all air very quickly, and doesn't even have to be filled to the top of the wye to work fine.

[bigben2k]

Originally posted by Zhentar
I've finally got my air trap working quite well. Its a little large, but its a great res and air trap. I used a 2" PVC wye. at the two tips of the y, I put in a 2" to 3/4" reducer, where I screwed in two 5/8" barbs. at the top I added on a bit of PVC pipe, and at the top a screw cap. (actually, I havent put in the screw cap yet) It clears all air very quickly, and doesn't even have to be filled to the top of the wye to work fine.

Sounds awfully familiar

The above design really isn't optimal: the flow ought to be reversed, and the inlet should be at the bottom, on the left. That way, the coolant can flow over the baffle, forcing the air out, without being sucked back down too fast.

A horizontal tube really is better, but usually a bit more work to fit. Once I've got the shop/garage up and running, I'll post pics and results of my latest creation (referred to earlier).

[Zhentar]

The design seems to work very well; maybe I just didn't explain it right.

http://timcluff.home.mindspring.com/wc1.jpg

You can see it there; the right tube leads to the pump inlet. None of that pipe on top is neccesary for operation; it doesn't draw in extra air without it.

[V12|V12]

Zhentar:

Now that's what I would call a BARE BONES system! Nice, large, but Optimal only means "optimal" for ea. individuals situation. Some don't have space, some have full-towers (evil people! lol) Good work, btw what is that wire going to the top of the tube?

[V12|V12]

NeoAcid:

(Sorry I didn't reply sooner, been offline for a while) Anyhow, that looks a lot better than the previous one... I mean it should work, but as I've experienced, you never really know until you try it out, unless it's a classic design that has been tried before and proven to work. As someone said before, that design looks more prone to having initial problems right from the start, IE; it's a complex way of dealing with air+water. Is there anyway that you could try another design, like what other's have posted as working before? Not that you should give up on your design, but it looks like it would be more hassle to work with, than a more simple design like BB's, Turbo's, mine's not too bad, and others etc... Good luck though...

[Zhentar]

its a temp sensor for the compunurse I got from directron.com. Its got 2 sensors, I've got one in the res for the water, and one at the air intake for ambient temp.

[Shreder-X]

Hey i am having an insane amount of bubbles in my system which i think is quite good besides the bubbles. I have this res. And i believe all the bubbles are created by the water bouncing off of the acrylic top and being suck back in before bubbles can dissipate. Since i have both in and out right next to each other like in the picture. I am using a via Aqua 1300 which also works great.
My idea is to put a piece of sponge into the side where water is hitting? i think this has been done before but it should reduce the water from bouncing off of the acrylic. My other problem is that the the res is on its side in my drive bays.
Any input if this had been done before would be helpful since i don't want to go getting a sponge stuck in my res.

[Boli]

I have that res and before I modded the thng it just happily foamed the water away becoming less like and air trap and more like a air... creator.

The mod... simple. A pen top with its end sawn off sliped into the intake and now:




Damn it clears the system up WITHIN 5min. That is from straight foam the system creates while the water is being forced round. to as clear. .

Hope this helps everyone it surprised me when it worked so well.

~ Boli


NOTE: When I didn't stagger my inlet/oulet things went as bit pear-shaped with whirlpools and foam galore

[Shreder-X]

i am a little confused since i cannot tell from your photo did you put a pen cap into the res floating at the top of the water. Or is it inserted into the intake valve hampering how much water can travel through it?

[JCYC5]

Would this work?

[Pritorian]

That would work just the same way as putting a wall inn the middel of the res. Should work quite well i would think.

Isn´t that how DD has made is last res?

[Boli]

Originally posted by Shreder-X
i am a little confused since i cannot tell from your photo did you put a pen cap into the res floating at the top of the water. Or is it inserted into the intake valve hampering how much water can travel through it?

Its a pen CASE for one. (Think on those see through Biro pen cases with just the end swan off) and it is just dropped into the intake value - restriting the flow I know - but it serves its main purpose of stoping the flow of water HIT the surface and makes sure the water comeing in goes DIRECTLY into the water in the res - no dropping. It initially was set without it and the impact of the water falling on the surface creates turbulence and thus bubbles.

I can't show you a better pic unfortunaly without taking my entire system apart .

JCYC5: that would work but you'll hit the same problem I faced with the drop in the water.

If I had the money and the capabilities I would create another 1/2" barb which its screw end extened well into the the res. This would help the restriction in flow.

In my res the water goes down (helped by gravity) to the pump and automatically primes the thing and the air trapped in the water goes upwards. So in my oppinion the thing may not be the perfect airtrap/res but it is damn good at both.

Like I said I was chuffed when this all worked so smoothly as I just put the things in for asthectics and worked from there .

~ Boli