• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Swirl pot resevoir

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
ppe1700

ppe1700

Member
​
Joined
Jan 9, 2007
here we go, an idea just popped into my head :confused::beer::)

since im custom making a stainless steel resevoir, for some reason my street rod racing days came back to me from when i used to fabricate with stainless...

how many of you know what a swirl pot is? well, basically in racing they use these for the fuel system but more importantly the water system.
the reason they use it for the fuel is irrelevant here so i'll explain why they use it for the water system..
basically as the water flows round the engine it comes close to boiling point, when water nears boiling more and more air bubbles form in the water. if this is left then it inhibits the water cooling efficiency as the air bubbles prevent contact or the ammount of water surface area (think i explained that ok...)

so, they have a swirl pot. the water coming in is made to swilr round and create a votex like pattern, the air bubbles go to the middle and leave the water. then the water out is at the bottom of this and is continued through the system.

im thinking this would be a helpful add on to my wc system. and i would have thought a must for peltier systems?

to make this all i will have to do is angle the barbs as i weld them in to whatever way water swirls. think its anticlockwise in the northern hemisphere :santa:
 
While its a nice idea (and yes i too know what a swirl pot is), the water temperatures reached in computer water cooling systems is no where near boiling, so it would have no use as there are no air bubble. It may help with filling initially to get air bubbles out though.

However i don't think it is such a bad idea aesthetically. If you made the res out of acrylic so one could see that water within, it would be cool to see the perpetual whirlpool inside it.
 
Usually the water in the pc is only a 1-2 degrees hotter than room temp.
I doubt a swirl pot is needed because a water cooled pc will not really boil water if yours does you got problems and you better just stick to modding cars.
 
how could the water in the pc be only 1 or 2 degrees hotter than room temp when the cpu under load will be around 45c ? you must live in a very hot room. are you saying that your radiator feels cool to touch?
if it feels cool that means its temp is less than body temp. i doubt this very much.
if the water in the pc was only 1 or 2c above room temp, then the conductivity of that system must be poor.
the idea i have for my system is:
res > pump > rad > cpu block > gpu block > res
 
Currently:

  • Current room temp @ monitor desk height: 20.8C (69.5F)
  • Air temp @ floor height case fan intakes: 19.8C (67.64F)
  • Air temp @ radiator output: 22.5C (72.5F)
  • Idle CPU die temps: 22C/22C (71.6F/71.6F)
  • Idle GPU die temp: 38C (100.4F)
  • Water temp: approx. 25C (77F)

You can see that even though I have two components in my loop that give off plenty of heat, the air temps from the rad aren't that much higher then room temps. Water temp isnt that much higher either.

Oh, and I have the same loop layout as you plan, and all I'm running is distilled water and Water Wetter.
 
ORLY!?
thats a verbal slap in the face lol
ive been thinking over this water system for like 2 weeks now and have all these outcomes worked in my head.

wonder if cheap vodka only in the loop would be better than water? lol
 

Attachments

  • Seperator.jpg
    Seperator.jpg
    55.8 KB · Views: 527
i was thinking something like the height of the tower case, or at least half the size of tower with 2" OD stainless tube.

i wonder how my fast flowing 1200LPH water will go round though

cheers for the link.. the people of that forum really layed into him for not working out a propper plan and then posting it / basically posting up a few components as a finished design :|
 
My vortex res holds a gallon or so.
Bleeding air from the 3 gallons of coolant is about instant, at least within seconds of turning it on for the first time.
No benefit in temps, though it sure is purty eye candy.
3000 lph pump (quiet one 3000).
This is my second version. The first was mounted in the case instead of on the wall, and only held about a quart.

FurtherTwisterRes.jpg


I built my first one about 5 years ago or so, after seeing a museum exibit of a 250 gallon tank of similar proportion.
The inlet inside the tank has a 1" ID 90° elbow in the 1" line feeding it....yeah, it's a manly system, but the chicks dig it too. :D

BTW, if the top weren't air-tight, it will/would suck air all the way down into my truck radiator in the basement. It's not good for the pump to have air bubbles in it, just so you're warned.
 
Last edited:
how did you make that? is that glass? what did you use for that?
i was thinking more along the lines of having an inlet and outlet on the sides, rather than the bottom. then it would have the vortex going all the way down to whatever is before it.
i dont get this:
BTW, it the top weren't air-tight, it will/would suck air all the way down into my truck radiator in the basement. It's not good for the pump to have air bubbles in it, just so you're warned.
:D

also how does it sound? can you hear the vortex?
 
Because my system pumps so much water (truck rad, 1" tubing, etc..) it's easily capable of sucking that vortex of air all the way down to the basement (about 5 feet below the tank.
I fill it completely with water, and the domed lid holds enough air for the vortex to form, though it's a slight vacuum. That allows for very fast bleeding.
I believe that the friction of the water being swirled against the down-pipe allows the air in the middle to pass freely downstairs. If I run it with the top open, the air displaces the water in my system and it overflows.
The first one didn't do this, but that was 3/8" tubing and a Danner model 2 pump.

Yes, it makes the pumping less efficient, but I've got cooling capacity to spare.

The res is made from an acrylic container that I bought at Wal-mart filled with spaghetti noodles...very cheap. The fittings on both sides of the tank walls are bushing type to avoid threading (and the spider cracking that goes with it). I just squeezed them together with jbweld covering the 3/4" long tail of pipe...hard to explain, but I'd make a pic if you needed.

My outlet on the first version was on the side of the base too, so it would sit in the case. I made a plate of plexiglass between the inlet and outlet (situated horizontally) with a hole in the center so that the vortex stayed centered.
The inlet on that one was at the top, and instead of using an elbow to create tangential flow, I bored the hole through the plexi tube at a tangent.

Finally, I use a truck radiator in my basement for my cooling system. It's 16x20 with the IO's modded to 1" ID female threaded copper couplings, and 4 Sanyo Denki San Ace's.
Between the super radiator, and the year-round chilly temps down there, I probably could cool 3-4 computers pretty well. My CPU usually idles below ambient temp at my desk, because ambient at the radiator is ~15C.
I have a heatercore inside my computer case to chill case air an avoid condensation...yes it's that chilly.
 
As you can see the water is only 26 deg
Typical room temp in spring (Australia) is 24 deg
Thats only a 2 deg difference and I have 2 GPUs 1 NB and 1 CPU (all of them over clocked and generating heaps of heat).




If you want to bleed better you might want to put a "divider" into the res so that bubbles can surface instead of being sucked back into the pump and though the loop again. Its quite effective even people like swiftec put a "divider" into their micro res....
 
Last edited:
Diggrr said:
Because my system pumps so much water (truck rad, 1" tubing, etc..) it's easily capable of sucking that vortex of air all the way down to the basement (about 5 feet below the tank.
I fill it completely with water, and the domed lid holds enough air for the vortex to form, though it's a slight vacuum. That allows for very fast bleeding.
I believe that the friction of the water being swirled against the down-pipe allows the air in the middle to pass freely downstairs. If I run it with the top open, the air displaces the water in my system and it overflows.
The first one didn't do this, but that was 3/8" tubing and a Danner model 2 pump.

Yes, it makes the pumping less efficient, but I've got cooling capacity to spare.

The res is made from an acrylic container that I bought at Wal-mart filled with spaghetti noodles...very cheap. The fittings on both sides of the tank walls are bushing type to avoid threading (and the spider cracking that goes with it). I just squeezed them together with jbweld covering the 3/4" long tail of pipe...hard to explain, but I'd make a pic if you needed.

My outlet on the first version was on the side of the base too, so it would sit in the case. I made a plate of plexiglass between the inlet and outlet (situated horizontally) with a hole in the center so that the vortex stayed centered.
The inlet on that one was at the top, and instead of using an elbow to create tangential flow, I bored the hole through the plexi tube at a tangent.

Finally, I use a truck radiator in my basement for my cooling system. It's 16x20 with the IO's modded to 1" ID female threaded copper couplings, and 4 Sanyo Denki San Ace's.
Between the super radiator, and the year-round chilly temps down there, I probably could cool 3-4 computers pretty well. My CPU usually idles below ambient temp at my desk, because ambient at the radiator is ~15C.
I have a heatercore inside my computer case to chill case air an avoid condensation...yes it's that chilly.
Click to expand...


i do like your design. its a lot better to look at than what im probably going to do. i was thinking about a plastic design, but just with regular tube like grey or white. but i dont know how i can get the metal barbs attached without leaking.
im guessing the vortex resevoir is at a neutral pressure because the same ammount of water coming is the same as going out... well with the vortex making a slight vacuum its not much of a vac is it?

wal-mart over here is called ASDA, so ill take a look at what spaggeti they have. did it come with the lid too?
 
Love that idea Myco, sure beats the iffy electronic temp readouts.

@ppe1700, yes the lid comes with it complete with a clear silicone ring gasket.
I doubt there's too much vacuum because it is a sealed system.
A smaller setup might have the drawback of sucking just enough air through the vortex to keep a steady supply of air bubbles running through your pump, so if that happens during your testing, I'd not run it that way for long. It can damage the impeller, making it crash against the inside wall of the pump volute (thinking of the floating impellers like the Liange pumps have) or wear against the axle.
You might end up with making more than one res, which can get expensive in materials. I got lucky because I'm no engineer by any means.

I'm intrigued by the idea of a spin pot (as billb showed us), the benefit of a res with the size not much bigger than a t-line. An excellent bit of kit to be sure, one that I might copy myself. Possibly able to be mounted sideways to make a neat window on the case front....something I'll have to tinker with...hmmm....
 
I've been considering this, too, mostly as eye-candy for the grandsons. ;) Seems to me the inlet and outlet should both be tangent - inlet near the top, outlet near the bottom. If the outlet is bottom center then the water pressure is less and there's that pesky chance of sucking air, too. A tangent outlet would have slightly higher pressure than a centered one and would reclaim some of your pressure loss from the pot.

At 1'+ tall your swirl-pot should be tall enough to avoid problems. Add an input line to the top center and I'd say you'd have a nice-looking set-up. You might consider acrylic or polycarbonate tubing (also available in 2") instead of SS. A little more expensive and tougher to work with but it would be kinda' cool to be see the water inside ...
 
That's sort of how the first one was made, except the outlet was not at a tangent.

http://home.wmis.net/~jberg/Vortex.jpg

The baffle centered the vortex. Without it, the vortex aims straight toward the outlet and flow in the res becomes off-centered. Then the end of the vortex would shear off creating bubbles everywhere.
The vortex would oscillate, forming on and off every 5 seconds or so as the flow was perturbed and reformed.
With the baffle, the vortex cannot form sideways to the circular flow, so what would occasionally reach below it shears and the bubbles would rejoin the vortex above the baffle.
No air reaches the pump then because the outlet is still too far below the bubbles to suck them in.
It took 3 tries to get that one right. Fortunately all with the same tube so it didn't cost a fortune.

Here's an old pic of it in my YY cube case. Notice it's my Abit KT7A-Raid board...ahh the days.
http://home.wmis.net/~jberg/Reservior.JPG
The green stripe isn't algae, it's antifreeze dye coloring the once clear epoxy used to hold the baffle. Lesson learned about materials there. ;)
 
You must log in or register to reply here.

Similar threads

R
Water cooling gloop.
Replies
6
Views
753
Rakunarr
R
Railgun
External twin MO-RA 360 planning and setup
Replies
1
Views
249
Railgun
Railgun
Al_bundy
Liquid Extasy Waterblocks
2
Replies
35
Views
2K
67Elco
67Elco
Railgun
AM5/Zen4 is pretty limited
Replies
11
Views
1K
freeagent
freeagent
B
New build advice (for a rookie water cooling builder)
Replies
4
Views
1K
Zerileous
Zerileous
Back