My Actual Flow Rate !OMG! HELP!

[arch5]

I posted the pics of my fridge mod last week and I decided to actually test my flow rate. I measured the flow rate with a gallon size container and timed it using my trusty Swatch Irony which has a stop watch.

My Result was an astounding 1Gallon in 1minute-8seconds! OUCH! This equals a whopping 53 Gallons per Hour!

• I'm using a Danner 500Gph pump and it has a really strong water flow compared to the pumps that came with my old "Blizzard Kit".
  
• I have about 18ft of tubing. 9ft out, 9ft in. Hey, I like to roll my case around on my desk OK.
  
• I'm using about 70% WindshieldWasher Fluid / 30% Water.
  
• And I have a DD Maze1C-1 WaterBlock

The test was done at the end of the loop, on the return line back to my Res. The Res temp is a stable 0*C.

Heres some pics of my setup. I'm no longer using the Radiator you see in the first pic.

Is this flow rate acceptable?

 

[Aesik]

I don't know that I'd call that flow rate acceptable, but I would call it predictable. Most people have no idea how much head loss their system is producing and how low their flow rate really is.

 

[arch5]

Aesik,
I looked at some of your other posts and I found the one where you were describing the equation dealing with flow rates and Heat output. I can't really tell if that will help me. It seems like if I were to ONLY increase water flow by switching to a more powerful pump I could either remove more heat or less heat, thus changing the temp of my reservoir.

I guess I'm a little stumped.
I suppose I should have spent more time in Physics class.

 

[Aesik]

Increasing the flow rate while keeping everything else constant will accomplish a few things.

But first, you have to understand one thing. Once at equilibrium, the amount of heat being 'removed' from the cpu/block is a constant NO MATTER WHAT.

By increasing the flow rate two things happen: first, a higher mass of water is moving through the block, thus allowing more particles of water to pick up energy from the block. It also increases the Reynolds number, thus increasing the turbulence and efficiency of energy removal. This results in a lowering of the surface temperature of the block, and thus a lowering of the CPU temperature.

So, at first it sounds like a wonderful thing, but unfortunately there is the fact that nothing comes for free. An increased flow rate through your system will also mean that the liquid in your reservoir will be cycled more quickly and not have as much time to dissipate all of it's energy. If the system is extremely efficient at dumping energy out of the closed loop (this goes for the fridge, or a radiator) than this isn't a problem. However, if it can't dump all the energy by the time it gets cycled through again, then the water entering the water block will be at a higher temperature, raising the block temperature, raising the CPU temperature... you get the picture.

What it comes down to is the best balance for the entire system. Every system is unique and every system will have its ideal balance point of maximum efficiency. From looking at your system and your approximate flow rate, off-the-cuff I'd immediately recommend trying a higher flow rate pump and/or reduce the head losses produced by your fluid path.

 

[arch5]

Thanks for the helpful reply. It's a good thing theres room for my computer AND a fridge on my desk! Looks like I'll be moving things around.

Thanks

 

[Patchmaster]

I think you're asking the wrong question. The right question is: Is my system being adequately cooled? If the answer to that is "yes", then don't worry about the flow rate. If the answer is "no", then flow rate might be something you'd want to look at improving.

 

[arch5]

I think you're asking the wrong question.

No, I was asking the question that Aesik was kind enough to try and answer. Possibly I can re-phrase my question to get a little more specific though.

Is a flow rate of 53 Gallons per hour acceptable from a 500gph Danner Mag Drive Pump? What are other's Actual flow rates?

Patchmaster,
This question would help me determine if my system is being adequately cooled. If I was satisfied with my cooling, you're right, I wouldn't be concerned with the flow rate.

 

[CrackedSKull]

one easy thing you could try is reducing the flowrate.
Theres all kinds of ways to that(kinking the hose somewhat, making the pump pump higher, using a smaller hose for one section)

If that makes temps worse, then increasing will more than likely help you reduce temps-- to a point.

 

[Owenator]

I agree with Aesik that the Reynolds number will increase with a higher flow rate but you also need to consider that the fluid 'particles' will not be in the block as long with a higher flow rate and not be able to absorb as much heat individually. This is usually a optimization type problem where you need to test different flows. But , with a fridge involved, like Aesik said, I bet higher flow would probably help.

One note of caution is that the higher flow is acomplished by the pump putting out a higher pressure. This higher pressure overcomes the losses to get higher flow but in may also cause leaks. I would test it with caution. I used a high flow pump once in a college lab with the outlet valve closed and the pump built up enough pressure to pop off the tubing. Good thing I was only flow testing some flow holes with clean water.

O

 

[Lt. Max]

whats the black isolation/rubber? around ur tubes, or are they tubes?

 

[arch5]

whats the black isolation/rubber? around ur tubes, or are they tubes?

The black tubes are Rubatex Insulation Tubes. You can get them at your local hardware store or at least a similar brand item.
Give me a sec and I'll take a pic of one that wasn't used.

I've just throttled the intake to my pump by partially closing the valve on my intake tube. I have shutoff valve at both ends of my tubes so that I can close the lines from either side if I need to do any maintenance. They come in quite handy when you want to empty your res but don't want to get water everywhere.

 

[Lt. Max]

so why do u use so big tubes?

 

[arch5]

so why do u use so big tubes?

I'm not sure if I understand the point of your question. I didn't want to spend the time wrapping them with insulation tape though. Plus the tubes have a 1/2" Inside Diameter which is the Outside Diameter of my water tubes. If you're asking why they are so long it's because I like to be able to move my pc on my desk. The extra length gives me room to roll around. If you mean "Why am I using Insulation Tubes?" then I'll say that they have considerably helped to keep the water cooler because they aren't letting the cold get out. Here's a pic.

 

[Lt. Max]

ok that answers it.. i was wondering why u got so fat tubes.. i see it was more convenient.

 

[arch5]

Okay here's an update so far. I've reduced the size of my water intake by half by closing the ball valve halfway. My Res. temp has dropped back down to what it was before I reduced the intake.
HMMMM.

I'll try closing the intake valve even more. Update to follow.

Maybe my pump doesn't like the water so cold. Has anyone experienced this?

 

[Aesik]

Owenator, time to crack open your text books again If you do the calcs, faster flow through the BLOCK will always result in a lower block temperature. It's the flow through the cooling part of the loop that needs to dump the heat that can benefit from a slower flow.

 

[Lt. Max]

so anyone done tests to find the perfect flow rate?

 

[arch5]

OK, I'm not sure what's going on here but since I throttled the intake on my pump using a ball valve my reservoir temps have actually gone down a little (0.3*C). I haven't changed anything since my last post. My CPU temps are reading the same.

What's going on?

 

[Barryng]

It is very misleading to state that a pump has a certain rated flowrate unless one is talking about a positive displacement (e.g. piston, gear. etc.) pump with a fixed speed motor. I do not believe the type of pumps being discussed here are PD pumps but are much more akin to centrifigal pumps.

The output of a centrifigal pump is variable and is very dependent on the flow resistance of the system it is discharging into. Maximum flow (called "runout") is achieved when the pump is discharging into open air with no pipe or line connected to its discharge fitting. This is not only a useless configuration but pumps are not designed to be in runout and will be eventually damaged from vibration and excessive current flow (at least larger ones will be damaged).

The opposite extreme is when the pump is deadheaded and the flow is zero. Pumps are also not typically designed to be operated in this condition and damage will also occur from excessive heating within the pump casing.

Since a real application falls somewhere between these two extremes, the flow rate will vary between zero and runout flow depending on the system resistance. Most manufactures can provide a pump curve from which the actual flow for a given system characteristic can be determined. However, this would be a very impractical tool for the type of applications being discussed here.

If a manufacturer claims a given model is rated at 500 gph, he must also state the associated discharge head for this to be meaningful. Even then, every installation is different with a different flow resistance depending on tubing size, tubing material, tubing length, and the flow characteristics of the water blocks. As you can see, it is very unrealistic to expect the pump to live up to the advertised flow claim, especially when the conditions for this claim are neither specified or, in reality, known. For the smaller pumps we are discussing, I would not be suprised if the advertised claims are at runout when the flow would be maximum. If this is the case, it would be totally insane to expect anything close to this advertised flow.

 

[Neco]

ever stop to think the water is so cold it's becomming harder to pump ?


sort of like a slurpee or a milkshake.. as it gets colder it's harder to suck through a straw, because it's starting to freeze