Martinm210

So there was some interest in elbows and some theorys floating around. I had another D5 show up for my second loop, so before I bundled it in, I went ahead with some testing:

First my setup was a new MCP655-B or the non variable variety, about 16' of 1/2" tubing and a 5 gallon bucket that I graduated using a gallon measuring container. All my tests consisted of filling it up beyond the 4 gallon mark, flipping my power supply switch, then clicking the stopwatch to start at the 4 gallon mark and stopping it at the 1 gallon mark.



First I tested the pump stock as is with 16.5' of tubing, about 5 feet was at the inlet into my measuring bucket, and the rest at the outlet.

Then I cut the tubing at the outlet and slipped in one 1/2" elbow (meant for 1/2" ID tubing) and tested.
Then I cut in a second elbow with another 1/2" and tested
Then I swapped out the last elbow with a 5/8" elbow and tested.

Then I started going back to the one elbow outlet but started removing a piece of tubing at a time until I came back up to the original flow rate.

What I found was the theoretical one 1/2" elbow = 1.5" of tubing is not correct for a 1/2" hose fitting elbow, it's actually closer to 4 feet!

So the theoretical
http://www.engineeringtoolbox.com/pv...ngs-d_801.html
is probably about right if you had an elbow with a 1/2" ID and you also didn't have the fitting losses at the sharp edges where you transition from hose to fitting. But it doesn't apply to a 1/2" hose fitting elbow, because the ID is closer to 3/8":

I found these main points to share:

• 1/2" elbow hose fitting = 4' of 1/2" tubing
  
• 5/8" elbow hose fittings are less than half that bad.
  
• There is a pretty big loss just in the fitting edges. This is why I like 7/16" ID tubing, it better fits the ID of the fittings
  
• The fitting losses are dependant on flow rate, the slower the flow the lower the pressure drop, consistent with water blocks
  
• Elbows and fitting losses are significant, and tubing losses are small
  
• You're better off sweeping ALOT more tubing than you are introducing an elbow, long live the snake!

Now for reality.... is one elbow going to make your temps soar? No, you probably wouldn't even see the difference with one, but a system full of these nasty little guys just might.

Hope this helps..
Martin

Update 7-22-07: Ran a pressure drop curve for 3 elbows along with a curve for a reservoir and T-line fitting, while my original test is still valid, I am getting roughly an equivalent to 4' of tubing, tubing itself is not all that restrictive.

Anyhow here are the PD curves, one elbow is actually more restrictive than an entire reservoir, and several time more restrictive than a T-line fitting.


Now for the reality...how does that relate to a systems GPM.

Here is one example
D-tek Fuzion, MCW60, D5 pump, 7 feet of tubing, one T-line = 1.66 GPM
D-tek Fuzion, MCW60, D5 pump, 7 feet of tubing, one T-line plus one elbow = 1.61 GPM
D-tek Fuzion, MCW60, D5 pump, 7 feet of tubing, one T-line plus 2 elbows = 1.56 GPM
D-tek Fuzion, MCW60, D5 pump, 7 feet of tubing, one T-line plus 5 elbows = 1.44 GPM
D-tek Fuzion, MCW60, D5 pump, 7 feet of tubing, one T-line plus 10 elbows = 1.29 GPM

Or how about a comparison to adding an MCW30 NB block (which flows quite well)
D-tek Fuzion, MCW60, D5 pump, 7 feet of tubing, one T-line plus one MCW30 = 1.58 GPM

One MCW30 = Two 1/2" elbows

There you have it, they do have an effect, but if you really need them and you have a good pump like the DDC or D5, the loss in flow rate isn't going to add up to much in regards to temperature.

FYI, I've attached a beta version of my estimator 1.4beta with the elbows input and curves working if you want to try this. Havn't released 1.4 yet, so this is just a temporary version.

Attached Files

MartinsWaterCoolingFlowRateEstimatorv1_4beta.zip (220.9 KB, 9 views)

ThePope

Nice, i always wondered.

__________________

trendy

Very good test to do man! I've seen people have 2-3 elbows in their loops :-x

__________________

Burn

Stickied

Great work! I really appreciate you taking the time to do this

__________________
My CNC Mill Project
We cannot change the cards we are dealt, just how we play the hand.
--Randy Pausch
Current Best Benchmark

Martinm210

You bet.
It was in interesting test, glad it helps!

hvacgaspiping

You could get even more flow with a long radius copper elbow and sweat a couple of barbs onto the end of it.

__________________
Proud member of Team Folders, Inc.

ira-k

Good Job!!!....I always thought it was about equal to one ft of tubing and I thought that was bad enough to not use them...WOW.. ....


EDIT: Oh Yeah your in my book-marks now....

__________________
...***Beware all ye Men born of Woman, beware the Hemogoblins***...
**E8600 @ 5751Mhz ***OCN Water Cooling Club And Picture Gallery ***150 Opty @ 3242Mhz **

Apollo4g

very nice Martin!!

Martinm210

Thanks guys...can't get enough tinkering with this stuff.

ira-k

I wonder how the flow would of been affected in an actual loop? It didn't really drop that much with a second elbow that seems strange to me...You don't have an extra rad and block you could stick on it do you?

Anyway it goes to show why we suggest people don't use them they kill flow IMO....Just like adding unneeded blocks to a loop do....Thanks for taking the time to do it and record it....

__________________
...***Beware all ye Men born of Woman, beware the Hemogoblins***...
**E8600 @ 5751Mhz ***OCN Water Cooling Club And Picture Gallery ***150 Opty @ 3242Mhz **