Any thoughts on the design of my reservoir?

M

Mysterae

Gawd
Joined
Sep 11, 2005
Messages
648
I updated my worklog regarding the final design of my reservoir, and as always I'd like to share it with you folk to get your opinion on it.

The image below shows the old design; the top item is the Collector, the bottom item the manifold.

stage_20f.jpg


The problem was for the original design I needed an acrylic block in the size of 100x100x200mm, and this proved expensive. So I redesigned the block to use cheaper, easier to get hold of sizes, and this is what I have come up with:

stage_87.jpg


stage_88.jpg


The price of all the parts (excluding fittings) is under £40 and a lot less wasted material. Made from two acrylic blocks, acrylic tubing, DD fill-port, threaded rod and 18 nuts. Add on the necessary hardware for connection and control of the circuit.

The ½” ball valve at the bottom is for the drain circuit, the 6 ¼” ball valves at the top are for the control of each loop. The DD fill-port is obviously there to fill and top up the circuit, and this can be done without opening the case. There are threaded holes at the top to mount it to the top of the case.

Just in case you don't know how it will be used:

parallel_hex_schematic_v5.jpg


Any thoughts? And yes, I am mad :p .
 
I like it! But wouldn't you want your distribution vavles mounted lower to avoid circulating any air trapped at the top? If they were mounted at the bottom, they'd always be under water and sucking liquid.
Change that, and I'd buy one off of you! :D
 
I like the second one better too theseeker.

I don't know exactly how much it will hold yet, as I'm not completely 100% sure on the length of the tubing. I'm using sketchup to see what can physically fit, but say the water level was 140mm, and the radius of the tube is 27mm, that's about 320ml, just under a can of coke.
 
Cheers xhx! The look of the Collector is a bit confusing - the water will be flowing into the reservoir from the valves. I'm using the valves to restrict the amount of water that each loop can put in to the reservoir, this promotes better flow in the other unrestricted circuits. Look at the schematic and you will see what I mean.

The only thing I have got to be careful of is bubbles created by the splashing of the water in the res. A plate or ping pong ball floating in the middle should settle the water underneath it.

I would consider making more to sell, but I can't see a market for 6 port reservoirs. It could be adapted (stop ends fitted) to be used on any number of circuits...watch this space!
 
I would recommend a second fillport on the top cap.
Having a dedicated airbleed port- separate from the port you fill with, will greatly reduce the likelihood of water backup (and spillage) due to air trying to escape the system while filling.
 
Sprocket, I wondered about the need for a bleed valve, and after consideration I think the bleed valve on the radiator should be enough. It's at the highest point of the circuit where the air will naturally collect, but I will need to take a panel off and be careful to catch any droplets of water when bleeding the system.

As long as the collector has enough water so it doesn't suck air back into the loops, all should be well.

Thewhiteguy, it would look mean with all 1/2" barbs! However, can you imagine the amount of tubing that would be sprawling all over the place! Using 6/8mm tubing and connectors will keep this to a minimum and not look too untidy.

My AC stuff was delivered today, but I wasn't in! I specifically asked them to deliver it to my work, which they failed to do :rolleyes:. I'll have to wait until tomorrow now!
 
That's going to be awesome. I can't wait to see it finished. I just have to wonder, however, how two pumps stacked so close to one-another will produce much more waterflow than one single pump... Blame it on my ignorance of such low flow systems, but I'd like to know if I'm wrong on this one :).

PS/ How and where are you going to mount this? I can see this sucker taking up alot of space, with barbs equidistant around it's circumference. Good call on the valves too, although I would have mounted them on the side of your case to change waterflow on the fly :).
 
Arcygenical said:
I just have to wonder, however, how two pumps stacked so close to one-another will produce much more waterflow than one single pump...
Click to expand...
Makes no difference where in the loop the pumps are, when run in series it will double the pressure of a single pump. Should get about 1/3 more flow from 2 vs 1.
 
nikhsub answered the question regarding the 2 v 1 pump, to be honest I wasn't too sure. I've done it that way after asking this forum, and it seems a good idea to have 2 pumps in this config for my circuit. The first pump takes the water from the reservoir (always a good idea) and the 2nd pump to make up for any drop in pressure from the rad. Plus I'll need all the pressure I can muster to ge the water flowing in this circuit, the manifold will kill some of that separating it into 6 lines. I'll have to do a lot of testing to balance both pumps, and remove the internal pots and use a double gang pot to control both pumps simultaneously.

Where's it going to go? Here:

stage_89.jpg


The top and two front panels are removed for clarity. There's a hole in the top lined up to the DD fill-port. It sure does take up a lot of space. It's 200mm in diameter and 280mm tall including connectors. I may make it even taller... :rolleyes:

I thought about the placement of the valves, and those that you can mount like you say are rare and expensive. After thinking about it, after all setup has been done and configured, I doubt I'd alter them anyway. But they come in handy for bleeding and filling the system. If they were computer controlled now that would be awesome!
 
I've updated my worklog with the finished article. So for those of you that don't venture to the Worklog section of the Forums, here's some images.

From conception to design to realisation:

collector_01a.jpg


collector_02a.jpg


collector_03a.jpg


collector_04a.jpg


collector_05a.jpg


Further info can be found in this post in the worklog.

Time to get it into circuit and run some tests!
 
Did you end up plumbing the pumps in parallel after all? I was looking at the fact you've got a bleed valve at the rad, you could put a removable piece of tubing on it to have it run any spilled fluids outside of the case and after it's all bled wrap a paper towel around the tubing at the valve and pull the tubing off. This would give you an easy way to route any spillage away from the internals and have a way to mop it up quickly when you pull the tubing off.

The res came out looking very sharp! One question about it though, what are you using to seal the ends? I don't see any o-rings.
 
I'll be doing a bit of testing tomorrow to find the optimal config of the pumps. I've found that the bleed screw on the rad is only good for letting air in, because of where it's positioned (top of the case). Plus the rad will be the first thing to drain, leaving all the components below it still containing coolant. The best drain point is at the most bottom, outside of the case to eliminate the chance of drips completely.

The Collector is sealed at both ends of the tube buy olive coloured o-rings, If you look closely at the first image you can see it in the bottom acrrylic block. This does look extreme for a reservoir, I love it!
 
Keep us updated on how the testing pans out, I'm very interested in it as I'm certain others are.
 
Looks pretty nice.

I have to say, however, that the fillports concern me. It seems like many people are experiencing humidity-related corrosion of both the galvanized plug and the anodized aluminum fillport bodies. I've been using the clear anodized version for ~6 months, and haven't seen any issues, but you should keep this in mind, and just make sure to keep an eye on it every now and then.

I just replaced mine with an acetal/nylon variant, though I don't imagine this is something you'd consider using (as it's all black).
 
How much are you going to sell them for? me likey!!
 
Initial testing has been good, with a few niggling issues. The pumps at present are in series as per all previous testing.

This isn't a big problem, but the tube has small air bubbles attached to the inner side of the tube. It looks like jacuzzi in the image below, but those bubles are static and stuck to the sides! It might be something to do with the acrylic polish I used, maybe not. I imagine they will go eventually, like they do when you have them in the stuck in the tubing lines. If anybody has any tips on this, let me know!

collector_06.jpg


The other niggle is the DD Fillport. I've found it's not the best because of the length of it and the required water line. To kill all water turbulence and bubbles, the water level has to be high, and the spout of the DD Fillport stops me from filling up completely. I'll remove the barb that's on the bottom of the fillport, that may help..

phide, thanks for the heads up on the fillport corrosion issue. It's odd that a company should mix the metals in something like that. Hopefully the coolant will take care of it, but I'll check it often. I'll see if I can find an alternative plug to use.

Bbq, I don't know how much I would sell them for, it would depend on the demand. Obvioulsy the more I get made, the cheaper it would be. If anyone is interested, shoot me a PM and I'll see what the numbers are. I can't imagine lots of people would want one, because of the application but you never know!

thewhiteguy, the outlets at the bottom are 1/2" BSPP and the inlets at the top (x6) are 1/4" BSPP. The top inlets could be made as 1/2" BSPP at manufacture no problem. I don't understand what you mean by:

thewhiteguy said:
Centrifugal pump inlet restriction FTL.
Click to expand...
but I'm open to be educated.

All in all I'm well pleased. It doesn't leak or creak and is silent when filled with enough liquid. The system has coped with everything I have thrown at it, even with a high ambient temperature as today.
 
I've found that using the handle of a screwdriver to tap on the sides of a res (with it standing on end) will convince the bubbles to gather elsewhere. Just make sure to be light handed with the love taps so you don't ding the plexi and tap on the plexi where the bubbles are, tapping those bolts won't help.
 
thewhiteguy said:
Centrifugal pump inlet restriction FTL.
Click to expand...
What he means is that restricting the INLET of any pump is a bad thing. I think he is confused though as 1/4" BSPP is not a small size, this is the size of my res thread and the same thread SIZE as swiftech's stuff, Storm, Apogee etc.
 
The single easiest way to reduce a centifugal pump's flow rate is to restrict its inlet. They aren't self-priming not because the manufacturer didn't want to make them that way, but because they suck and sucking.

Good to hear that you've got a decent sized fitting down there.
 
As much as we'd all like to buy that kickass res, let's look at reality:

"Will we really by running 6+ blocks in parallel?"
 
Just to add to the 2 vs 1 pumps debate, 2 pumps will give you redundancy, just in case one fails, theres one still circulating, and keeping the loop alive, so in the event of a failure, you have time to shutdown the system and address the problem, because if one pump is down and you havent noticed it for a while, temperatures wont skyrocket on you but will be noticeable so your parts don't die. Situations like this where heat is distributed all over the place can get nasty, especially where theres NO flow and heat is repeatedly being added to the water, which could kill every part in the system at once :eek: .
 
Emission said:
Just to add to the 2 vs 1 pumps debate, 2 pumps will give you redundancy, just in case one fails, theres one still circulating, and keeping the loop alive, so in the event of a failure, you have time to shutdown the system and address the problem, because if one pump is down and you havent noticed it for a while, temperatures wont skyrocket on you but will be noticeable so your parts don't die. Situations like this where heat is distributed all over the place can get nasty, especially where theres NO flow and heat is repeatedly being added to the water, which could kill every part in the system at once :eek: .
Click to expand...

I doubt it, processor would probably cause the whole system to shut down. My pump didn't start day before yesterday (I have since remidied this, by cleaning out everything with some vinegar), the storm got to hot, and poof, computer shuts down.
 
Hehe, another nice feature of modern day processors, thermal overload protection :cool:, but im just talking worst case scenario, with water, heat can be transferred from other parts even AFTER the system shut down.
 
What? Your post is a giant mess of commas.

The worst case scenario is that the motherboard shuts the machine down in the event the CPU reaches a set temperature. Video cards will also shut down the core if it should get too hot.
 
Pump redunancy is a good thing, and would work well with the pumps in series, except for the restriction a dead pump may bring, but the other pump could overcome it.

However, thinking about flows with the pumps in parallel, it could be a problem. If one pump dies, the flow may go through that dead pump rather than through the rest of the circuit - it all depends on what has the highest restriction.
Lets say the dead pump has a lower restriction than the rest of my circuit, the water would flow out the collector/reservoir to the working pump, through the equal T (that combines the flow of both pumps before the rad) and possibly through the dead pump back to the collector! Not good. I think I'll simulate a dead pump when I have them in parallel and see what happens.

A lot of above may not make a lot of sense, I'll post some pictures so you know what I mean. Anyone have any idea how restrictive a dead D5 is in the opposite direction?
 
Well, depending on the kind of openings inside the pump its probably not much to worry about, only think I would worry about is having them in parallel, because it could start a reverse flow.
 
Hmm.. If you have a one-way valve, per pump, and the pumps are in parallel, and when one pump stops, the fluid can't push the 'door' open, then it would block the flow to one pump, so all the flow goes through the working pump.
 
I have 2 DDC+'s running in series as shown below. One pump powers the system just fine...

pumps.jpg
 
Emission, I agree. I'll know soon if the extra flow would be worth the risk. Bbq has the solution to that risk:

Bbq said:
Hmm.. If you have a one-way valve, per pump, and the pumps are in parallel, and when one pump stops, the fluid can't push the 'door' open, then it would block the flow to one pump, so all the flow goes through the working pump.
Click to expand...

That would work a treat Bbq! I wonder if I can find any suitable one-way vavles.. I'll experiment to see if a one pump failure in a parallel pump arrangement will cause problems. I don't think all flow will go through the duff pump, but some will.

Nice set up nikhsub1, having the pumps in series is no problem at all, even if one pump fails. I tried it with one D5 at setting 1 and the other at setting 5 and all was fine. I wouldn't want to run them for long like that though, it's best if they are both at the same speed imo.
 
nikhsub1 said:
I have 2 DDC+'s running in series as shown below. One pump powers the system just fine...
Click to expand...
Why the heck would you run them in series? They have plenty of head by themselves.

One-way valves are restrictive.
 
thewhiteguy said:
Why the heck would you run them in series? They have plenty of head by themselves.

One-way valves are restrictive.
Click to expand...
Series yeilds better flow with the restrictive blocks I run.
 
You must log in or register to reply here.
Back
Top