-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
My first WCS---Storm with results
- Thread starter Wet Neophyte
- Start date
OP
- Joined
- Jul 5, 2005
- Thread Starter
- #22
I have my data in from my GPM test.......
The loop with this configuration holds a little over 2 ⅓ cups of liquid primed.
This loop is doing 1.25 gallons a minute ( I was hoping for more
)......the pump by itself drains 1 gallon in around 15sec.
Next time I have that loop outside I'll add a 6" tube in between. Not that I'll be able to use it because the door barley closes on my case.... If I do add I'll need to bend some tubing.
Velocity
I need a Pitot tube.
The loop with this configuration holds a little over 2 ⅓ cups of liquid primed.
This loop is doing 1.25 gallons a minute ( I was hoping for more
)......the pump by itself drains 1 gallon in around 15sec.Next time I have that loop outside I'll add a 6" tube in between. Not that I'll be able to use it because the door barley closes on my case.... If I do add I'll need to bend some tubing.
Velocity
I need a Pitot tube.
The velocity of water within a pipe is subjected to physical resistance due to friction and turbulance. This can easily be demonstrated with a long garden hosepipe. Although the water may gush rapidly out of the tap or standpipe, by the time it comes out of the hose at the other end the velocity will have significantly slowed. The same occurs in swimming pool pipework, and all pipe installations must have the pipe sizes calculated in order for the water volume delivered through the pipe is not too high for the pipe diameter.
The maximum velocity in any suction pipe must not exceed 5 feet per second (1.52 metres per second).
The maximum velocity in any pressure pipe must not exceed 9 feet per second (2.74 metres per second).
And to that extent I am looking at a graph that shows 1/2 inch plumbing reaching 5.64 feet per second @ 5GPM. Also for reference, 2GPM=2.26 feet per second, 1GPM=1.13 Feet per second.
The maximum velocity in any suction pipe must not exceed 5 feet per second (1.52 metres per second).
The maximum velocity in any pressure pipe must not exceed 9 feet per second (2.74 metres per second).
And to that extent I am looking at a graph that shows 1/2 inch plumbing reaching 5.64 feet per second @ 5GPM. Also for reference, 2GPM=2.26 feet per second, 1GPM=1.13 Feet per second.
OP
- Joined
- Jul 5, 2005
- Thread Starter
- #24
According to this water blocks facts it reacts better to more pressure......and according to you if I put a spacer of tubing in between the pump outlet and water block inlet....I will gain at the most 25% more pressure? Is that right?
So I should definitely see a temperature difference in doing so. I will try it out.
According to the math in your graph what's the suggested length of space out of the pump?
So I should definitely see a temperature difference in doing so. I will try it out.
According to the math in your graph what's the suggested length of space out of the pump?
Wet Neophyte said:
In theory that is correct.
I will try to find out if that is true for your situation. All my data is for 1.5-2" plumbing. I will try to find some data to correlate to our smaller sizes. I think I will start by looking at the irrigation industry.
Please wait to do anything unless you are really itching to do so. I would hate for you to go to all that effort, only to find out you gained 3% more pressure.
- Joined
- Jan 15, 2004
Wet Neophyte said:
I would wait if you read my post there would not be a difference since the space is so small. I would wait until Travis gets back with more data so that I can work out a model for our uses.
Travis: I can't wait for the numbers
Ok Sewer here goes, if you want/need more specific stuff, let me know. This is going to a random collection of interesting data as I find it relevant to various extents.
Multi-Pump Systems
Centrifugal pumps in parallel provide additive flow conditions and when mounted in series provide additive head conditions. Therefore three matched pumps mounted in series will provide the same flow at three times the head (aprox.) compared to one pump. Three matched pumps in parallel wil provide the same head at three times the flow compared to one pump. A very popular variation on this theme is to provide mult-stage pumps with up to twenty stages mounted on a common shaft to provide a high head from a single pump.
http://www.lmnoeng.com/index.shtml
Here is an interesting program that will formulate pump discharge pressure depending on a whole host of parameters-
http://www.uengineer.com/pumpexam111.htm
http://www.the-engineering-page.com/forms/pump/c_pump.html
Here is a friction loss chart form 3/8-6"-
http://www.engineeringtoolbox.com/pressure-loss-plastic-pipes-18_404.html
Ok that is all I can find on that stuff. Other interesting tidbits, The D4 can flow about 2.5x the volume and 50% more psi than the DDC pump.
Ok back onto my original concern, the whole theory of water flow being compromised due to an immediate restriction may be less than stated do to the following. Pool pumps spin at 3450rpms, the new D4 spins between 1000-3000 rpms. I could not find an rpm number for the DDC, but since the the D4 and the DDC use the same impeller but diff. magnet assemblies and the DDC pumps less water and less psi I am going to guess it is around 2000-2500 rpms.
Now in the instance of 2 speed pool pumps, hi=3450rpm and low=1725 rpms the flow is dramatically reduced. As you said Sewer it is almost always quadratic. So a pump on hi speed that pumps 100gpm will drop to 25gpm at half the impellor speed. Ok I am rambling now, Sewer see what you can make of the mess I just created!
Multi-Pump Systems
Centrifugal pumps in parallel provide additive flow conditions and when mounted in series provide additive head conditions. Therefore three matched pumps mounted in series will provide the same flow at three times the head (aprox.) compared to one pump. Three matched pumps in parallel wil provide the same head at three times the flow compared to one pump. A very popular variation on this theme is to provide mult-stage pumps with up to twenty stages mounted on a common shaft to provide a high head from a single pump.
http://www.lmnoeng.com/index.shtml
Here is an interesting program that will formulate pump discharge pressure depending on a whole host of parameters-
http://www.uengineer.com/pumpexam111.htm
http://www.the-engineering-page.com/forms/pump/c_pump.html
Here is a friction loss chart form 3/8-6"-
http://www.engineeringtoolbox.com/pressure-loss-plastic-pipes-18_404.html
Ok that is all I can find on that stuff. Other interesting tidbits, The D4 can flow about 2.5x the volume and 50% more psi than the DDC pump.
Ok back onto my original concern, the whole theory of water flow being compromised due to an immediate restriction may be less than stated do to the following. Pool pumps spin at 3450rpms, the new D4 spins between 1000-3000 rpms. I could not find an rpm number for the DDC, but since the the D4 and the DDC use the same impeller but diff. magnet assemblies and the DDC pumps less water and less psi I am going to guess it is around 2000-2500 rpms.
Now in the instance of 2 speed pool pumps, hi=3450rpm and low=1725 rpms the flow is dramatically reduced. As you said Sewer it is almost always quadratic. So a pump on hi speed that pumps 100gpm will drop to 25gpm at half the impellor speed. Ok I am rambling now, Sewer see what you can make of the mess I just created!
- Joined
- Jan 15, 2004
Travis: Even though I have no numbers for this because I was unable to get length with each discharge pressure (check your pm box). First off on a 1/2" system the friction losses are many times higher than in a 1.5-2" system so any gain would be minimized due to the added friction of our systems. Now last time I checked the DDC spins at 3600rpm. However I don't think rpm is a valid number to use since it depends on the size, shape and probably other things I'm forgetting of the impellor. Right now I'm leaning towards there is no substantial gain and there might be a loss.
Ok the reason I mentioned RPMs is like this. First for comparison sake, lets lay some baselines. For this discusssion lets assume the nominal velocity thru Neophytes plumbing to be 5 feet/sec.
All centrifugal pumps use rotation to "sling" water from the middle of the impeller to the outside and thru the volute to the piping. There is two ways this can be accomplished, spped or torque. Just like in cars, you can use a high revving(lower torque) motor to get moving quickly or a high torque(lower revving) motor to move more.
Back to Neophyte, when that impeller "slings" the water it is moving faster than 5 feet/sec. for x amount distance before it "slows" back down to nominal. For arguements sake well say it takes a D4 3" of tubing to "slow" back down to nominal and if you add a restriction closer than 3" you are messing with that acceleration stage of the pumps flow. So lets say that for 3" the water is doing 7.5feet/sec. Now we add restriction that slows that down 6feet/sec. Well we now have water not getting out of the pump fast enough so the pump builds backpressure and flow decreases.
I can make another analogy if that doesn't make sense.
All centrifugal pumps use rotation to "sling" water from the middle of the impeller to the outside and thru the volute to the piping. There is two ways this can be accomplished, spped or torque. Just like in cars, you can use a high revving(lower torque) motor to get moving quickly or a high torque(lower revving) motor to move more.
Back to Neophyte, when that impeller "slings" the water it is moving faster than 5 feet/sec. for x amount distance before it "slows" back down to nominal. For arguements sake well say it takes a D4 3" of tubing to "slow" back down to nominal and if you add a restriction closer than 3" you are messing with that acceleration stage of the pumps flow. So lets say that for 3" the water is doing 7.5feet/sec. Now we add restriction that slows that down 6feet/sec. Well we now have water not getting out of the pump fast enough so the pump builds backpressure and flow decreases.
I can make another analogy if that doesn't make sense.
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
Travis, you are getting this all far more convoluted than it needs to be. What you are saying is essentially correct, but there's a far easier way to present the issue/data.
The system has a flow/(back)pressure curve, which is the sum of the flow resistances of the components in the loop.
The pump has it's own PQ (pressure/flow) curve.
Where the two curves intersect, that is the flow rate that will be seen.
e.g.
The pump just works against the back-pressure, and the flow stabilises at the back-pressure that the pump can push that amount of flow against. All this business about longer outlet tubings doesn't matter.
The system has a flow/(back)pressure curve, which is the sum of the flow resistances of the components in the loop.
The pump has it's own PQ (pressure/flow) curve.
Where the two curves intersect, that is the flow rate that will be seen.
e.g.
The pump just works against the back-pressure, and the flow stabilises at the back-pressure that the pump can push that amount of flow against. All this business about longer outlet tubings doesn't matter.
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
Additionally water is essentially incompressible. It is not elastic. It doesn't take time to "speed up" after it leaves the pump.
Cathar said:
I realize as a liquid it is incompressible. And the speed up is not after it leaves the pump. It is the instant it leaves the trailing edge of the impellor vane and and however much further distance until the effects of backpressure and friction loss come into play.
Don't you agree that the water will have a different velocity thru the radiator, cpu wb, pump? Your own design accelerates water thru "jets", the impellor is accelerating the water via "vanes". Either way you are building pressure. One form is electricall/mechanical generated and the other is mechanical.
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
Unsure what the water velocity is inside some component after the pump has to do with the pump performance, aside from merely being expressed in terms of the back-pressure that the pump has to work against in order to push the water to achieve that velocity.
The theory is undue restriction too close to pump discharge will hurt potential peak psi/flow. That is all. I can't find any current info to back up my claim. I read this in a pool pump installation manual about 6-7 years ago and do not have that manual availible for reference. I suppose that could have been exclusive to that 1 brand pump design, but at this point who knows. I plan on calling a few pool pump suppliers and seeing if this is still true.
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
If we plot the pressure drop throughout a loop though, the pressure is always at the highest immediately after the water is flung from the impeller vanes, and pressure drops from that point onwards as we pass through various points of restriction. Because water is incompressible it really doesn't matter if the total restriction exists as one tiny little point straight after the pump outlet, or if it exists as a complex series of restriction points as what occurs in a water-cooling loop.
While I don't doubt that your pool pump manual said this, I suspect that the reason for the meaning may have gotten a little lost. Perhaps they were referring to having no unneccesary points of restrictions after the pump outlet so that the pump was able to provide sufficient pressure to push adequate flow through the pool's filter system.
i.e. they're just talking about the normal practise of not having unneccesary points of restriction.
While I don't doubt that your pool pump manual said this, I suspect that the reason for the meaning may have gotten a little lost. Perhaps they were referring to having no unneccesary points of restrictions after the pump outlet so that the pump was able to provide sufficient pressure to push adequate flow through the pool's filter system.
i.e. they're just talking about the normal practise of not having unneccesary points of restriction.
Ok I have some updated info here. I contacted Sta-Rite a international company that makes arguably the finest pumps you can buy. Anyhow, the technical rep. that I spoke with stated there is a industry rule of thumb that is this. "For best flow, you want on both the inlet & outlet 4x the I.D. of the pipe." So for a 1/2" system, we would want a mininum 2" of pipe until our first fitting. He is going to be looking for the documentation to back this statement up and e-mail that info to me in a few days.
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
Travis, they're talking about a mostly open-flow application. Small DC water-cooling pumps are a very different scenario.
Actually, for the pump inlet it is important that this is as unrestricted as possible, but this is already commonly known and is why everyone always recommends that the reservoir goes right before the pump inlet.
Additionally, flow restriction from tubing is proportional to the tubing length. In a computer we're only talking about very short runs in comparison with a pool scenario.
Still Travis, what would be most helpful for you here would be for you to try it for yourself, with waterblocks installed, and compare even 1/2" ID to 3/4" ID, and let us know what you find. If it makes anything like a 5% difference in flow rates I'll happily eat my words.
No one is going to run 2" pipe in their computer case.
Actually, for the pump inlet it is important that this is as unrestricted as possible, but this is already commonly known and is why everyone always recommends that the reservoir goes right before the pump inlet.
Additionally, flow restriction from tubing is proportional to the tubing length. In a computer we're only talking about very short runs in comparison with a pool scenario.
Still Travis, what would be most helpful for you here would be for you to try it for yourself, with waterblocks installed, and compare even 1/2" ID to 3/4" ID, and let us know what you find. If it makes anything like a 5% difference in flow rates I'll happily eat my words.
No one is going to run 2" pipe in their computer case.
- Joined
- Jan 15, 2004
Cathar: 2 inches of pipe is not unreasonable. However the problem is how would you attach that to the pump without the barbs and tubing. Also about using regular tubing isn't that where the no gain comes from? (The following is probably not accurate) If a 2 inch piece of pipe produces said effect then can we say its the lower friction in the pipe vs the tubing (assuming there is lower friction).
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
SewerBeing, and what about the step-down transition costs from the 2" pipe into the waterblock's 1/2" barbed fitting?
TravisDawes's posts seem to suggest that these small water-cooling pump makers don't know what they're doing, especially companies like Eheim who stick on their ideal sized barbs onto their pumps and sell the tubing to match.
TravisDawes's posts seem to suggest that these small water-cooling pump makers don't know what they're doing, especially companies like Eheim who stick on their ideal sized barbs onto their pumps and sell the tubing to match.
Similar threads
