Flow comparison in high restriction loop with various blocks, pumps and fittings

[luke997]

In my quest for a decent loop flow I've experimented a lot with various pumps (2xD5, 2xD5 Strong, 2xDDC 3.25, 3xDDC 3.25), fittings setups (from extreme BP fest to just straight compression and tubing), SLI configurations (parallel, serial, EK bridge, BP Crystal Link), blocks (full mobo to nothing, EK wave GPU design to new more restrictive) and this brought interesting results which I think are worth sharing.

I thought it would be good idea - perhaps helpful to some - to combine all this results in a single table so the information is easily accessible in one place - rather than scattered in my worklog (thanks to eXa for the suggestion).

All in all from this data you can get very rough idea what you might expect with change of pumps, SLI configuration, removing fittings or blocks as well as change in the EK GPU blocks design or how many fittings are needed for specific loops.

Note #1
Every single time the actual flow was different (read much worse) than the predictions in various flow estimators - so although many changes across each revision and the measurements being far from scientific reduces the value of data - the fact of being real life results increases it a bit.

Note #2
There was no flow degradation over time from version to version due to any sort of gunk in the blocks are algae.
Combination of distilled + PT Nuke + Acetal blocks + Black Primochill tubing + very little chance of reservoir exposure to the sun works very well in that aspect.

Note #3
Once again - this is far from any scientific measurements and testing, it's just sharing of results in my particular loop - so obviously your mileage may vary

Enough talk - here's the data:


Whenever change occurs between revisions - it's been marked in bold text.

[coolhandluke41]

Thanks Luke ,very well done

[Utnorris]

Nice. A lot of info here. I assume this is a single loop if it only shows flow for one pump? Also, what is the difference between the pump flow and flow result? In the dual loops, how did you have them separated? I.e loop 1 was just the cpu and a rad?

[luke997]

Nice. A lot of info here. I assume this is a single loop if it only shows flow for one pump? Also, what is the difference between the pump flow and flow result? In the dual loops, how did you have them separated? I.e loop 1 was just the cpu and a rad?

This is all for the single loop only.

"Flow result" is the flow in the loop with all the pumps on (whether it's 2 or 3)
2 pumps is the flow in loop with 2 pumps switched on (where the total was 3)
1 pump is the flow in loop with 1 pump switched on.

Hope it's clear now.

[eXa]

AWSOME!
I think the 4 last colums are the most interesting. Only thing that is changing is the fittings and paralell/serial GFX, yet the flow changes quite alot.

I dont know if i deserve any credit tho, all i said that it was kinda too bad that all this info was "hiding" in your build log rather than here

Cheers!

[Martinm210]

Great work!! Thanks for sharing!

[Utnorris]

This is all for the single loop only.

"Flow result" is the flow in the loop with all the pumps on (whether it's 2 or 3)
2 pumps is the flow in loop with 2 pumps switched on (where the total was 3)
1 pump is the flow in loop with 1 pump switched on.

Hope it's clear now.

Yeah, I think I have it. If I understand correctly, in column one, this was a single loop using two D5's with an EK dual top. The 199LPH is the flow with both D5's turned on and the 130LPH is the flow with one of the D5's turned off but is still in the loop. Is that correct?

Just curious as to why you used LPH versus LPM?

[luke997]

Yeah, I think I have it. If I understand correctly, in column one, this was a single loop using two D5's with an EK dual top. The 199LPH is the flow with both D5's turned on and the 130LPH is the flow with one of the D5's turned off but is still in the loop. Is that correct?

Just curious as to why you used LPH versus LPM?

Yes, that's correct.

I've used LPH since that's what I have on my Aquaero for the flow, but this can be easily converted.