DIY pressure drop curves

[Martinm210]

Just in case you ever feel like doing a little pressure drop testing...I was doing two curves today and figured hey...why not share my redneck setup..

So, might as well share a little DIY on creating pressure drop curves. It's pretty easy really, and you don't need much more than the following:

• Variable pump or valve: Valve should be free flowing like a globe valve or gate valve, pump could be a D5 variable. You could also run off of tap water pressure, but if you have any variability in your source pressure, your flow rate will change, so preferrably use a pump. I experimented with this and I'm pretty sure there is too much variability in pressure..better use a pump if you want something very accurate.
  
• Tubing for water columns: Any size will do, just need two colums of water. It doesn't matter if the tubing bends a bit at the base, and it doesn't matter if you have a high/low inlet or outlet. They will change the pressure, but the difference between will be the same regardless. The key to pressure drop is you want to measure the "ELEVATION DIFFERENCE".
  
• T-fitting: You need two 5/8" T fittings. you could use 1/2", but the pressure drop includes everthing between the two fitting centers, so minimze the tubing and 5/8" fitting will be less restrictive. I found that 5/8" T fittings and 1' of tubing is such a small amount of restriction you can ignore it. If you're anal enough you could create an equation for that then subtract it from your test run but I think most people ignore this small amount.
  
• Measuring Bucket: Any old trash can or 5 gallon bucket will do, preferably something clear you can easily see water level within, or you can attach a barb like I have and graduate the tubing for more precise reading.
  
• Stop Watch: Just need something that can click On and Off at least to the nearest second.

This is actually a fairly accurate test method for creating these curves. I feel comfortable that I can measure the pressure drop column of water to within 1/4" which is equivalent to .009 PSI (.1&#37 (try to find a guage with PSI graduations to the nearest hundredth. And flow rate (if your source is constant) is at worst within 1 second of time, at 3 GPM that's 1/60th or 1.6%, if you wanted better than that...get a bigger bucket!

Anyhow, this is the basic setup:
This is one of two test series I ran today, this one being about 10' of 3/8" ID tubing:


And this is what pressure drop looks like at very very low flow rate (this happens to be one of my 1/2" test)

Yes... I prefer to setup my laboratory lounging in the shade outside


It's an interesting test and teaches you alot about pressure drop and those curves. If you decide to try it let me know, I can help through the excel graphing part too!

[santiagodraco]

Woa, very cool! It's always nice to learn something about how this stuff works. Once you see it like this it makes perfect sense. Nice going

[MaxxxRacer]

ghetto but very cool.

But I still like my digi manometer (not that i use it )

[Fairydust]

Nice work.

Afaik the pressure tubing should either be very thin or have some other form of restrictive element to minimize the influence of the flow's kinetic energy, so only the static potential is measured, correct me if i am wrong.

[hecktic]

sticky please

[Martinm210]

Nice work.

Afaik the pressure tubing should either be very thin or have some other form of restrictive element to minimize the influence of the flow's kinetic energy, so only the static potential is measured, correct me if i am wrong.

All I know is the few curves I've produced that have also been published have been very similar.

Here is my MCR320 with 5/8" barbs, it flows a little better than published at the right end which it should with the larger barbs:


My MCW60 curve was very very close:


My MCW30 test was a bit less restrictive than published. I did do a little cleaning up on the barbs, but close regardless:



The only curve that threw me off a bit is the D-Tek fuzion. As rough as the published curve is I believe mine more after seeing how close I was to all of the swiftech curves. My guess is the test was done with a pressure guage with few increments, again why I like water columns better..



So, I think it's fairly accurate. It's not nearly as fast as having the mechanical equipement (and with good tools you could repeat the test much more often), but this was recreation for me and I don't have to worry about calibration either..

[speed bump]

Nice manometer setup.

One thing that would make it more accurate and simpler to test with though. Use a scale and weigh the water after a certain time period.

[Fairydust]

Ya the difference should be that big, especially when both lines are nearly identical affecting both line similarly. I Just remember seeing a testbed being criticized for that reason before. Anyway, using smaller diameter tubing on the pressure line might make it even cheaper and minimize impact even further.

[CyberDruid]

Nice Work Martin

[sick_g4m3r]

this is awesome. we need this and your flow estimator to be combined into a sticky.

[Martinm210]

Ya the difference should be that big, especially when both lines are nearly identical affecting both line similarly. I Just remember seeing a testbed being criticized for that reason before. Anyway, using smaller diameter tubing on the pressure line might make it even cheaper and minimize impact even further.

Yeah, I understand what you are saying and I think you are correct. You do get a little error in pressure from water velocity catching that edge and pushing it's way up. I observed about 2-3 inches when I was doing a pump max flow rate test. But just like you mentioned in this case since my end measurement is the difference between the two points both of which have the same flow rate and fittings and error. I also have pressure added to both measurements by the elevation of my collecting reservoir as well as the friction loss in the following collection tubing, BUT..again that same pressure is added to both locations so the end pressure difference is still accurate.

That error does however relate to something if using this same setup in a pump pressure drop test where you only have one tube and you are measuring pressure. If I saw 2-3 inches at the max flow rate test, you have that small error growing for 0 to 3 inches that either could be removed mathematically, or reduced by using a smaller inlet at the top of the T fittings. Thankfully 2-3 inches is a small amount, it's only .1psi.

I think the next time I do a pump test I'm going to try and find a 5/8" T with a smaller top barb, maybe 1/4" or somthing. That or I can create a pressure vs flow rate equation for that pressure error and subtract it out from the collected results.

[Martinm210]

Thanks all, hope it's useful. I know it's a bit redneck, but it works and it works pretty well. Alot of pressure guages and equipment is often calibrated by columns of water. The nice thing about it is gravity doesn't lie, but more importantly this is something anyone with the interest and a few bucks to put the basic parts together can do themselves.

You could put your whole loop together between the two points and create a very accurate system pressure drop curve that would account for all the different fittings and specifics.

The frustrating thing for me that I've noticed is how several manufactures don't post pressure drop curves and as you can see it doesn't take really expensive equipment to produce a curve. My guess is, if they aren't posting a curve it may not be a good marketing piece of information to sell the product. The other difficult part of the published pressure drop curves is the different units used, granted it may not be anyones intentions to muddy the water, but when you have one curve in GPM vs PSI in one hand and the other in meters of water vs liters per minute it makes it hard to tell for us consumers what the heck is going on.

I do have to commend swiftech for posting data on all their products though, they are very open about it, and from what I can tell very accurate as well.

My next venture is to take a look at reservoirs vs T-lines, I'm pretty sure alot of people will be a bit suprised to see just how restrictive those reservoirs really are. When is the last time you saw a pressure drop curve for a reservoir??

I guess I'm a bit inspired by "Mythbusters" and "It's all in the name of science!"..lol

[sick_g4m3r]

My next venture is to take a look at reservoirs vs T-lines, I'm pretty sure alot of people will be a bit suprised to see just how restrictive those reservoirs really are. When is the last time you saw a pressure drop curve for a reservoir??

I guess I'm a bit inspired by "Mythbusters" and "It's all in the name of science!"..lol

Cant wait for the res pressure drops, when do you think you will do them?

And mythbusters rules! you actually do kind of remind me of adam.

[Martinm210]

Nice manometer setup.

One thing that would make it more accurate and simpler to test with though. Use a scale and weigh the water after a certain time period.

Not sure what you mean by measuring the weight of the water.

Water weighs 62.4lbs per cubic foot and pressure is a force per unit area so Feet of water is a very standard unit of pressure measurement for hydraulics you don't need to weigh anything.

One of the most common arguments I've heard around pressure and colums of water is folks thinking that the tubing size is important. If you were measuring wieght, that would be correct, but you are not. It's a measurement of force per unit area.

I could have one of those tubes be a 6" diameter pipe and the other one 1/4" and they would still measure the same amount of pressure head (Feet of water). One side would have significantly more weight, but it's distributed over a much larger area...in the end the force per unit area is the same.

Anyhow, I may not be understanding what you meant..but this is my understanding.

[Martinm210]

Cant wait for the res pressure drops, when do you think you will do them?

And mythbusters rules! you actually do kind of remind me of adam.

Haha...I won't tell you about my little venture yesterday testing the max pressure capability of Tygon R3603 on 80 psi?(it's high watever it is) household pressure

It reminded me of making water weenies with surgical tubing many years ago.

The tubing started to bulge...the clamp quickly gave way...and I ended up with a face full of water. I better start wearing my safety goggle,s I could have gotton tygon shrapnel in my eye

[sick_g4m3r]

Haha...I won't tell you about my little venture yesterday testing the max pressure capability of Tygon R3603 on 80 psi?(it's high watever it is) household pressure

It reminded me of making water weenies with surgical tubing many years ago.

The tubing started to bulge...the clamp quickly gave way...and I ended up with a face full of water. I better start wearing my safety goggle,s I could have gotton tygon shrapnel in my eye

i wish i saw that! whered you get that many PSI? was it air?

[Martinm210]

i wish i saw that! whered you get that many PSI? was it air?

No it was from my household water plumbing, I'm probably way off on the 80 psi, it's probably half that I just know we have really high pressure compared to any other areas of town. I should pick up a pressure guage just to see what it really is, I could use that if I ever decide to redesign and redo my underground irrigation systems.

It was a funny sight I'm sure....I laughed my self until I saw my camera also fell victim to the water explosion..

[Martinm210]

Cant wait for the res pressure drops, when do you think you will do them?

I gave away my commercial coolingwork bay res to a guy on overclock.net, so at the moment all I could do is test a home made reservoir. It should give similar results, but it would be better if I had a couple of commerial reservoirs to test side by side. So..in the next couple of weeks I'll at least get some fitting tested and my custom reservoir vs T fitting tested.

My guess is a 5/8" T fitting has very very little restriction and a reservoir is many time more restrictive..probably worse than some water blocks.