SOLVED watercooling project

[William]

I am doing a water cooling project for school and want to make as failsafe of a solution as possible. To do so, I am using a small car radiotor to ensure heat dissapation even without a running fan. Leaks are my lest concern but I will go overboard with sealing them. My thought though is on the pumps. I read the post earlier on running in pumps in parallel vs. in series. I have some responses and questions to this. If I use two pumps that are quite more powerful than I really need, would a reduced water flow matter all that much if I used a Y adapter? Also, couldn't impeller pumps be run in series so even if one failed it would not have the probs of pulling the water through to pump. I am also attaching a video card cooler to the system. Were would the optimum placement for this be? In a Y with the CPU? Before or after the the water goes through the CPU? Or on another pump/small radiator assembly sharring the same water reserve? All responses would be greatly apreciated.

 

[Richard]

I honestly don't think you'll impede the waterflow too dramatically with a "Y" connector. Just make sure you don't use one that's too small.

If I don't get a response from Dangerden soon I'll have to implement one in my new project as well.

To maintain an even distribution of workload I decided that the "Y" connector would be best if placed within the reservoir. (This would also only necessitate a single outlet.)

Originally, I intended to reduce turbulence in my watercooling rig. However, due to the findings on this website. Turbulent flow can actually improve flow rate.

http://gcisolutions.com/GCINotes1298.html

 

[William]

Richard999 (Mar 01, 2001 07:45 p.m.):
I honestly don't think you'll impede the waterflow too dramatically with a "Y" connector. Just make sure you don't use one that's too small.

If I don't get a response from Dangerden soon I'll have to implement one in my new project as well.

To maintain an even distribution of workload I decided that the "Y" connector would be best if placed within the reservoir. (This would also only necessitate a single outlet.)

Originally, I intended to reduce turbulence in my watercooling rig. However, due to the findings on this website. Turbulent flow can actually improve flow rate.

http://gcisolutions.com/GCINotes1298.html

Not really sure how that link applied, but turbulence is not really bothering me. I just kinda was wondering if this was the best setup. I actually believe that a car radiator is the best way to go because it is designed to dissapate tons and tons of heat. It is the same concept as using a 500gph pump to ensure plenty of water flow. In addition, the cost is hardly anymore, we got the radiator for $50 which is a little high and with good enough shopping, you can probably get one for cheaper, but it is better than the ones places like dangerden sell, although they have a nice size advatage if you want to put yours inside the case.

 

[Richard]

I linked that site, because I think it is useful to understand the fluid dynamics.

Getting back to your question.

The car radiator is of course overkill. Even without a fan, a small radiator like the one on DangerDen's website will be sufficient to keep a CPU cool. (As long as the pumps are still running!)

A car radiator is large, because it is tackling a great deal more heat.

About the 500 GPH pump. Once again. That's overkill. As long as you have water moving through the block at a constant flowrate you're fine. In fact, the argument could be made that if a fan does go out on the radiator and you're using a huge pump you won't give ample time for the radiator to do its job.

dunno260 (Mar 01, 2001 10:08 p.m.):
Not really sure how that link applied, but turbulence is not really bothering me. I just kinda was wondering if this was the best setup. I actually believe that a car radiator is the best way to go because it is designed to dissapate tons and tons of heat. It is the same concept as using a 500gph pump to ensure plenty of water flow. In addition, the cost is hardly anymore, we got the radiator for $50 which is a little high and with good enough shopping, you can probably get one for cheaper, but it is better than the ones places like dangerden sell, although they have a nice size advatage if you want to put yours inside the case.

 

[Newbie_Doo]

Car radiators are designed to dissipate tons of heat...at speed. Actually, I think that you will find that your radiator (if it's a full-size one) is WAY TOO MUCH for the amount of heat you intend to load it with. You might actually have a problem with flow rate with a full size radiator. That said, I am using a Hayden Transmission cooler on my computer, but the Danger Den and Swiftech radiators are highly respected. Both are copper-cored and dissipate heat well.

 

[William]

yeah the radiator is tons of overkill, but it did not really cost too much more than the ones at like danger den and is a lot more total overkill. I know that the others can dissipate the heat from the CPu as it is not putting many Watts of heat. We actually did a very rough calculation and figured that the car radiator can dissipate about 7000 watts of heat. We are also using a vacuum pump to create a negative presure inside to system which will stop water from coming out in the event that a leak would form.

 

[Colin]

Richard999 (Mar 01, 2001 10:36 p.m.):
I linked that site, because I think it is useful to understand the fluid dynamics.

About the 500 GPH pump. Once again. That's overkill. As long as you have water moving through the block at a constant flowrate you're fine. In fact, the argument could be made that if a fan does go out on the radiator and you're using a huge pump you won't give ample time for the radiator to do its job.

I have a 550 GPH pump in my system. Restricting the flow increases the temps. When I switched from 136 GPH to 550, my temps dropped 3 C.

 

[DamnFast]

dunno260 (Mar 02, 2001 09:34 a.m.):
yeah the radiator is tons of overkill, but it did not really cost too much more than the ones at like danger den and is a lot more total overkill. I know that the others can dissipate the heat from the CPu as it is not putting many Watts of heat. We actually did a very rough calculation and figured that the car radiator can dissipate about 7000 watts of heat. We are also using a vacuum pump to create a negative presure inside to system which will stop water from coming out in the event that a leak would form.

I don't think that creating a negative pressure inside the case will slow the leakage in the event if it happens. It will INCREASE the chance of leakage. This is true because the pressure inside the hose is much higher than the pressure outside the hose, so if there is a leak, it will become greater and greater to fill that negative pressure in the case.
Having a possitive pressure will be much better solution for that, although I don't think air pressure in the case would matter by much anyways.

 

[William]

DamnFast (Mar 02, 2001 12:20 p.m.):

I don't think that creating a negative pressure inside the case will slow the leakage in the event if it happens. It will INCREASE the chance of leakage. This is true because the pressure inside the hose is much higher than the pressure outside the hose, so if there is a leak, it will become greater and greater to fill that negative pressure in the case.
Having a possitive pressure will be much better solution for that, although I don't think air pressure in the case would matter by much anyways.

Actually it will. Fluids always go from an area of high pressure to one with an area of low pressure. By lowering the pressure of the system to creating one that will create a constant pressure to one less than that of the room, air wil flow into the area where a leak would form. The only byproduct would be air bubbles but the vacuum pump would take them out as it would be set to maintain a constant pressure. But that is an interesting question on whether it would actually cause leaks to occur. Also i did a little calcuation on the radiator, it is from an 86 ford, small car forgot the name, it should be able to dissapate easily 20,000 watts of heat energy which is a whole smacking lot.

 

[William]

also, should 256 CFM in and 256 CFM out of air be good enough for the case or should i use more?