My theory of calculating necessary head in a loop

maxSaleen · 12-27-2004, 10:10 PM

Hopefully my image posted correctly. If not I'll try again. I recently configured and ordred a watercooling kit from various vendors. I posted a thread about my concern that the MCP350 wouldn't have enough power to maintain adequate flow rates through the Black Ice III. This was based on my faulty assumption that it was a four pass radiator. I calculated the total head required for the loop to be 52". Factor in restriction of the loop, any pressure drops, and I fiqured I would be getting flow rates as if it were at 100" (though this was just a guess). At such a high head the flow rate would be very low.

Assuming that my picture posted correctly, follow my logic as I walk you through the loop. From the pump the water flows to the CPU water block 10" above the pump. From there gravity pulls it downward through the VGA blcok and to the inlet of the radiator (this part of the loop has no effect on the pump's head). From there (assuming that the Pro III is a four pass radiator which it isn't) the water would have to travel vertically 14" twice. This would add 28" to the required head. From the outlet of the radiator the water would have to travel vertically to the reservior 14". My reasoning is that the pump, without any considerations for flow restriction or pressure drops, would require at least 52" or 4.3' of head. Anyone think I'm correct in my reasoning? Anyone think I'm full of junk? Let me know, and thanks for looking. [IMG]

[/IMG]

jaydee · 12-27-2004, 10:55 PM

Pressure drop of the CPU block?
Pressure drop of the GPU Block?
Pressure drop of the Radiator?
Pressure drop fo the Res?
Pressure drop of the hose?
Pressure drop of the barbs?

All those will cause pressure drops gravity or not.

Gravity probably shouldn't be in the equation in a closed loop system. :shrug:

SlaterSpeed · 12-28-2004, 05:40 AM

Yup gravity dosent make any differance. Think about it, It allways goes back to where it started. Gravity is against it when going up but when going down it helps it so it all cancels out.

slater..

maxSaleen · 12-28-2004, 08:22 AM

I didn't mean that gravity caused a pressure drop. When I mentioned pressure drops I was reffering to the fact that the volumes of certain devices within the loop are greater than 3/8" id tubing. An example would be the reservior or the radiator.

SlaterSpeed: What you say about gravity pulling the water down that is not to say that it cancels the effect on head. Think about this scenario: You have a huge full tower case where one part of the loop is four feet above the pump. If the pump only has 3.5' of head it won't be able to push water through the loop. See where I'm going with this? Just because of the fact that the water eventually "falls" 4' doesn't mean it will assit the pump in pushing the water "up".

Take another look at my example. What I'm trying to convey is that everytime the water makes a vertical (or near vertical) run it has an impact on the head of the pump. This would be something interesting to experiment with to see if it is remotely accurate.

Anyone else?....

SlaterSpeed · 12-28-2004, 08:51 AM

Im pretty sure it would be able to. Without the pump running gravity would push the water from the top down the tube to the pump and then back up the other side just like a manometer tube. When you switched the pump on it would only have to circulate the water and overcome the flow resistance of the tubing. It wouldent have to put any extra energy in to attualy rasing the water up to the top as that energy would come from gravity pulling the water down on the otherside of the loop.

Afterall the head pressure required to pump water up 3.5m would be the same pressure you would get at the bottom of a tube of water 3.5mm high would it not??

Your right if the pump was pumping from one tank to another buti think its differant in a closed loop tubing system.

Makes sence to me, hope im right

bobkoure · 12-28-2004, 08:55 AM

Here's a thought experiment.
Take a pump with 3.5' head. Connect the inlet to the outlet with a single length, say 8' of tubing, ID large enough not to be much of a factor. Garden hose should be fine.
With the tubing laying on the same surface as the pump, it should work fine, right? (water goes 'round).
Now lift the center of the loop so it's higher than 3.5' above the pump.
Does the flow stop? If you think it does, maybe you should actually try the experiment, rather then just think about it. If you don't think the flow stops, then how is a closed system different from a single length of hose?
[edit] BTW, mounting a rad like that, with inlet and outlet at the bottom is not un-usable but will present problems with air being captured in the top tank, at least at fill-time.[/edit]

redleader · 12-28-2004, 09:48 AM

I run a large, upside-down radiator. Its not a huge issue, the air is pretty quickly flushed out, particularly if you've got a good res. If you're using a bleed line, then it could take a very long time.

maxSaleen · 12-28-2004, 09:52 AM

bobkoure: I actually plan on mounting the rad horizontally during fill and bleed. Thx for the heads up though. I'll try that experiment with an old hydor pump if I get the time today.

killernoodle · 12-28-2004, 09:53 AM

The falling water will feed the pump inlet. Just think of siphoning- the way it works is water is drawn to a point lower than where it is fed- say one bucket 1m high and another one on the ground. As soon as the water level is lower than the 1m bucket, gravity will assist in drawing the water out of the 1m bucket. This works even if the tube extends 10m in the air out of the 1m bucket and then goes to the bucket on the ground.

So, the falling water draws up the other water. A water level works much in the same way.

When you calculate a pump's head height, you must consider the absolute height of what it will be pumping to with 0 pressure on the inlet. A pump (considering everything is strong enough to sustain the pressure) attached to a long tube filled with water that is draped over the empire state building could pump water like it was 0' head pressure if the tube end matched the elevation of the pump (not taking friction into account). However, if you take the end of the tube and lift it over the max head pressure of the pump, it will not be able to pump the water.

In that same scenario, lts say you have a bucket on the roof that has 2 fittings on it. The setup is the same with the pump on the bottom and the other tube exits at the same level as the pump. This setup wont work because the falling water is no longer a variable in drawing up the water. The water can fall independently from the pump's actions, so the pump must pump the 1,300 so odd feet to the top of the building instead of the 0' of the previous setup.

It is difficult to describe, but it is the truth.

maxSaleen · 12-28-2004, 09:58 AM

Something just came to me....

Bobkoure, about your experiment: Extrapolate your reasoning out to an extreme. Think if the hose was 100' in lenght and you raised the loop so that its apex is nearly 35' above the pump. Still think the water will flow? What about a loop that is 1,000' in lenght? I'm not trying to be an ass about this, so if I have offended you I apologize. You input is greatly appreciated. I'm just trying to keep my thoughts in line with the scientific method as I understand it.

HammerSandwich · 12-28-2004, 10:04 AM

Quote:

Originally Posted by maxSaleen

What I'm trying to convey is that everytime the water makes a vertical (or near vertical) run it has an impact on the head of the pump. This would be something interesting to experiment with to see if it is remotely accurate.

Go play with a siphon. A big rise makes it a PITA to get started but will not stop it from flowing.

jaydee · 12-28-2004, 10:15 AM

Quote:

Originally Posted by HammerSandwich

Go play with a siphon. A big rise makes it a PITA to get started but will not stop it from flowing.

If it is a closed loop is cannot siphon. Gravity doesn't have any effect in a closed loop. Take your siphon hose and put it back into the tank and see what happens...

It takes just as much power to get the liquid back up to the tank. Cancels each other out. :shrug:

HammerSandwich · 12-28-2004, 12:05 PM

That's what I meant. A siphon will continue to flow regardless of the vertical rise above the source, so it's a good example of the countering forces.

torin3 · 12-28-2004, 12:41 PM

Quote:

Originally Posted by HammerSandwich

That's what I meant. A siphon will continue to flow regardless of the vertical rise above the source, so it's a good example of the countering forces.

Uhhh...no.

http://www.science.edu.sg/ssc/detail...arent=5&cat=54

Brians256 · 12-28-2004, 12:51 PM

torin3, you are absolutely correct. However, the 10m maximum height for a siphon applicable only as a differential in pressures caused by the ability for water to freely move.

If you have 20 meters of rise above the pump (ridiculous for a contained unit, but not ridiculous for a system like Bladerunner's where he has the PC hooked up to a reservoir in the ground), you won't get a vacuum. If the system is open to atmospheric pressure, you would not "see" the force (imparted by the pump on the outlet) at the inlet side. If it is closed, the head that the pump produces is lost in 1) friction and 2) pressure differentials allowed by air bubbles and flexible tubing.

If this wasn't correct, we'd have a terrible time getting water delivered to our houses, and systems like a hot-water recirculating pump wouldn't work.

edit: after reading Jaydee's post, it seems I am just restating what he said but in longer and windier terms. Ooops!

bobkoure · 12-28-2004, 03:51 PM

Quote:

Originally Posted by maxSaleen

Extrapolate your reasoning out to an extreme. Think if the hose was 100' in lenght and you raised the loop so that its apex is nearly 35' above the pump. Still think the water will flow?

If the water is flowing without the center of the loop raised (so friction is being overcome), then yes. There are some frictional losses, even with very large IDs, so you may get no flow at all with 1000' of hose. Garden hose comes in 100' lengths, though - try with that.
Offended? Heck, no - I'd prefer you thought I was an idiot, so long as you went out and did the experiment to see how well your mental model of the world and the actual world match. People kept Aristotle's model of gravity for a very long time - IMHO because they either didn't bother to try for themselves or they tried it the same way he did (pebbles dropped through water), which doesn't match out-of-water very well.

maxSaleen · 12-28-2004, 05:54 PM

Well I'll throw in the towel on this one. I was wrong. Though this illlustrates the purpose of forums. I threw an idea out there, we discussed it, and we sort of have a conclusion. This explains why koolance's diminuitive pumps are able to force water through relatively large loops (think of their full tower with four blcoks in series). So far the consensus is that only pressure drops and friction would inhibit flow in a closed loop, not necessarily the height of certain components relative to the pump. Anyone else have something to add?

If friction is a concern in closed loops do I hear some votes for 45 degree inlets and outlets on water blocks? Always something I wanted to try but have never had the time....

Brians256 · 12-29-2004, 12:16 AM

Friction is a concern in our loops, but you do want the 90 degree input for reasons of turbulence, I'd think. The outlet(s) would probably do better with swept style bends in flow. Anyone up to having curved flow outlets for their blocks?

I don't think it'd be significant in an overall system though, with modern consumer pumps like the D4/MCP650 and the DDC/MCP350. It was probably more of a concern with pumps like the Eheim 1048 pumps or the Rio "pumps". That's my guess. Anyone know what the headloss of a 90 bend is vs a swept curve and if it would be significant ?

12-27-2004, 10:10 PM	#1
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	My theory of calculating necessary head in a loop Hopefully my image posted correctly. If not I'll try again. I recently configured and ordred a watercooling kit from various vendors. I posted a thread about my concern that the MCP350 wouldn't have enough power to maintain adequate flow rates through the Black Ice III. This was based on my faulty assumption that it was a four pass radiator. I calculated the total head required for the loop to be 52". Factor in restriction of the loop, any pressure drops, and I fiqured I would be getting flow rates as if it were at 100" (though this was just a guess). At such a high head the flow rate would be very low. Assuming that my picture posted correctly, follow my logic as I walk you through the loop. From the pump the water flows to the CPU water block 10" above the pump. From there gravity pulls it downward through the VGA blcok and to the inlet of the radiator (this part of the loop has no effect on the pump's head). From there (assuming that the Pro III is a four pass radiator which it isn't) the water would have to travel vertically 14" twice. This would add 28" to the required head. From the outlet of the radiator the water would have to travel vertically to the reservior 14". My reasoning is that the pump, without any considerations for flow restriction or pressure drops, would require at least 52" or 4.3' of head. Anyone think I'm correct in my reasoning? Anyone think I'm full of junk? Let me know, and thanks for looking. [IMG][/IMG]

12-28-2004, 05:40 AM	#3
SlaterSpeed Cooling Savant Join Date: Nov 2003 Location: Suffolk, UK Posts: 234	Yup gravity dosent make any differance. Think about it, It allways goes back to where it started. Gravity is against it when going up but when going down it helps it so it all cancels out. slater.. __________________ aka. slater3333uk - The self proclaimed 'Middle Player' 'Liquified'

12-28-2004, 08:51 AM	#5
SlaterSpeed Cooling Savant Join Date: Nov 2003 Location: Suffolk, UK Posts: 234	Im pretty sure it would be able to. Without the pump running gravity would push the water from the top down the tube to the pump and then back up the other side just like a manometer tube. When you switched the pump on it would only have to circulate the water and overcome the flow resistance of the tubing. It wouldent have to put any extra energy in to attualy rasing the water up to the top as that energy would come from gravity pulling the water down on the otherside of the loop. Afterall the head pressure required to pump water up 3.5m would be the same pressure you would get at the bottom of a tube of water 3.5mm high would it not?? Your right if the pump was pumping from one tank to another buti think its differant in a closed loop tubing system. Makes sence to me, hope im right __________________ aka. slater3333uk - The self proclaimed 'Middle Player' 'Liquified'

12-28-2004, 08:55 AM	#6
bobkoure Cooling Savant Join Date: Feb 2003 Location: USA - Boston area Posts: 798	Here's a thought experiment. Take a pump with 3.5' head. Connect the inlet to the outlet with a single length, say 8' of tubing, ID large enough not to be much of a factor. Garden hose should be fine. With the tubing laying on the same surface as the pump, it should work fine, right? (water goes 'round). Now lift the center of the loop so it's higher than 3.5' above the pump. Does the flow stop? If you think it does, maybe you should actually try the experiment, rather then just think about it. If you don't think the flow stops, then how is a closed system different from a single length of hose? [edit] BTW, mounting a rad like that, with inlet and outlet at the bottom is not un-usable but will present problems with air being captured in the top tank, at least at fill-time.[/edit] Last edited by bobkoure; 12-28-2004 at 09:00 AM.

12-28-2004, 09:53 AM	#9
killernoodle Thermophile Join Date: Sep 2002 Location: Charlotte, NC Posts: 1,014	The falling water will feed the pump inlet. Just think of siphoning- the way it works is water is drawn to a point lower than where it is fed- say one bucket 1m high and another one on the ground. As soon as the water level is lower than the 1m bucket, gravity will assist in drawing the water out of the 1m bucket. This works even if the tube extends 10m in the air out of the 1m bucket and then goes to the bucket on the ground. So, the falling water draws up the other water. A water level works much in the same way. When you calculate a pump's head height, you must consider the absolute height of what it will be pumping to with 0 pressure on the inlet. A pump (considering everything is strong enough to sustain the pressure) attached to a long tube filled with water that is draped over the empire state building could pump water like it was 0' head pressure if the tube end matched the elevation of the pump (not taking friction into account). However, if you take the end of the tube and lift it over the max head pressure of the pump, it will not be able to pump the water. In that same scenario, lts say you have a bucket on the roof that has 2 fittings on it. The setup is the same with the pump on the bottom and the other tube exits at the same level as the pump. This setup wont work because the falling water is no longer a variable in drawing up the water. The water can fall independently from the pump's actions, so the pump must pump the 1,300 so odd feet to the top of the building instead of the 0' of the previous setup. It is difficult to describe, but it is the truth. __________________ I have a nice computer.

12-27-2004, 10:55 PM	#2
jaydee Put up or Shut Up Join Date: Dec 2001 Location: Spokane WA Posts: 6,506	Pressure drop of the CPU block? Pressure drop of the GPU Block? Pressure drop of the Radiator? Pressure drop fo the Res? Pressure drop of the hose? Pressure drop of the barbs? All those will cause pressure drops gravity or not. Gravity probably shouldn't be in the equation in a closed loop system. :shrug:

12-28-2004, 08:22 AM	#4
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	I didn't mean that gravity caused a pressure drop. When I mentioned pressure drops I was reffering to the fact that the volumes of certain devices within the loop are greater than 3/8" id tubing. An example would be the reservior or the radiator. SlaterSpeed: What you say about gravity pulling the water down that is not to say that it cancels the effect on head. Think about this scenario: You have a huge full tower case where one part of the loop is four feet above the pump. If the pump only has 3.5' of head it won't be able to push water through the loop. See where I'm going with this? Just because of the fact that the water eventually "falls" 4' doesn't mean it will assit the pump in pushing the water "up". Take another look at my example. What I'm trying to convey is that everytime the water makes a vertical (or near vertical) run it has an impact on the head of the pump. This would be something interesting to experiment with to see if it is remotely accurate. Anyone else?....

12-28-2004, 09:48 AM	#7
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	I run a large, upside-down radiator. Its not a huge issue, the air is pretty quickly flushed out, particularly if you've got a good res. If you're using a bleed line, then it could take a very long time.

12-28-2004, 09:52 AM	#8
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	bobkoure: I actually plan on mounting the rad horizontally during fill and bleed. Thx for the heads up though. I'll try that experiment with an old hydor pump if I get the time today.

12-28-2004, 09:58 AM	#10
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	Something just came to me.... Bobkoure, about your experiment: Extrapolate your reasoning out to an extreme. Think if the hose was 100' in lenght and you raised the loop so that its apex is nearly 35' above the pump. Still think the water will flow? What about a loop that is 1,000' in lenght? I'm not trying to be an ass about this, so if I have offended you I apologize. You input is greatly appreciated. I'm just trying to keep my thoughts in line with the scientific method as I understand it.

12-28-2004, 12:05 PM	#13
HammerSandwich Cooling Savant Join Date: Oct 2003 Location: 15143 Posts: 358	That's what I meant. A siphon will continue to flow regardless of the vertical rise above the source, so it's a good example of the countering forces. __________________ www.procooling.com: It's true we are often a bunch of assholes

12-28-2004, 12:51 PM	#15
Brians256 Pro/Staff Join Date: Oct 2001 Location: Klamath Falls, OR Posts: 1,439	torin3, you are absolutely correct. However, the 10m maximum height for a siphon applicable only as a differential in pressures caused by the ability for water to freely move. If you have 20 meters of rise above the pump (ridiculous for a contained unit, but not ridiculous for a system like Bladerunner's where he has the PC hooked up to a reservoir in the ground), you won't get a vacuum. If the system is open to atmospheric pressure, you would not "see" the force (imparted by the pump on the outlet) at the inlet side. If it is closed, the head that the pump produces is lost in 1) friction and 2) pressure differentials allowed by air bubbles and flexible tubing. If this wasn't correct, we'd have a terrible time getting water delivered to our houses, and systems like a hot-water recirculating pump wouldn't work. edit: after reading Jaydee's post, it seems I am just restating what he said but in longer and windier terms. Ooops!

12-28-2004, 05:54 PM	#17
maxSaleen Cooling Savant Join Date: Dec 2004 Location: Orlando, Florida Posts: 383	Well I'll throw in the towel on this one. I was wrong. Though this illlustrates the purpose of forums. I threw an idea out there, we discussed it, and we sort of have a conclusion. This explains why koolance's diminuitive pumps are able to force water through relatively large loops (think of their full tower with four blcoks in series). So far the consensus is that only pressure drops and friction would inhibit flow in a closed loop, not necessarily the height of certain components relative to the pump. Anyone else have something to add? If friction is a concern in closed loops do I hear some votes for 45 degree inlets and outlets on water blocks? Always something I wanted to try but have never had the time....

12-29-2004, 12:16 AM	#18
Brians256 Pro/Staff Join Date: Oct 2001 Location: Klamath Falls, OR Posts: 1,439	Friction is a concern in our loops, but you do want the 90 degree input for reasons of turbulence, I'd think. The outlet(s) would probably do better with swept style bends in flow. Anyone up to having curved flow outlets for their blocks? I don't think it'd be significant in an overall system though, with modern consumer pumps like the D4/MCP650 and the DDC/MCP350. It was probably more of a concern with pumps like the Eheim 1048 pumps or the Rio "pumps". That's my guess. Anyone know what the headloss of a 90 bend is vs a swept curve and if it would be significant ?

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)