Large volume laminar flow machine

[karimwassef]

the vertical sections are 15" high in a 16" high box (0.5" top and bottom).

<a href="http://s1062.photobucket.com/user/karimwassef/media/4_zpsbtumpq2a.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/4_zpsbtumpq2a.jpg" border="0" alt=" photo 4_zpsbtumpq2a.jpg"/></a>

and here's how it fits in the tank

<a href="http://s1062.photobucket.com/user/karimwassef/media/5_zpsnwoofvbl.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/5_zpsnwoofvbl.jpg" border="0" alt=" photo 5_zpsnwoofvbl.jpg"/></a>

 

[Floyd R Turbo]

Have you done any flow calculations? I think those would be pretty complex but if you're trying to push 40,000 GPH through this, I think you're going to run into some really high pressure (backpressure) as you increase flow due to localized friction.

Think about it in comparison to PVC pipe. If you use the same pump to push water through a 3" pipe vs a 1/2" pipe, you will get a significantly lower flow rate.

 

[herring_fish]

Floyd,

What are you doing here, slumming? Actually, I agree. I would think that the flow will be a bit anemic, especially when driven by propeller pumps.

 

[karimwassef]

this is 15" high.

The inlet is 2" wide, so that's 2" x 15" = 30 in2 cross section. That's the equivalent of a 6" diameter PVC pipe...

It gets better once it turns the corner, but then it gets tight again as it needs to split.

As it skinnies down, it basically splits the flow by half and then splits the cross-section by half, so the ratio is about the same...

Until the very end at the 1/4" openings. There's 64 of them, so effectively 16" at the exit. That's the equivalent of an 18" PVC pipe.

 

[theatrus]

I'm not well versed in CFD otherwise I'd suggest that route

 

[karimwassef]

Assuming 40,000 gph and 15" height throughout... here's what the geometry looks like.

width (in) ....count .....aperture (in2) ...~ PVC pipe (in) ...in / sec
2.00 ............... 1 ............... 30 ............... 6 ............... 85.6
1.25 ............... 2 ............... 38 ............... 7 ............... 68.4
0.75 ............... 4 ............... 45 ............... 8 ............... 57.0
0.50 ............... 8 ............... 60 ............... 9 ............... 42.8
0.25 ............... 16 ............. 60 ............... 9 ............... 42.8
0.25 ............... 32 ............. 120 ............. 12 ............. 21.4
0.25 ............... 64 ............. 240 ............. 17 ............. 10.7

 

[karimwassef]

No doubt it's a lot of flow, but if you can accept pushing 5000 gph through a 2" PVC pipe... that has an aperture area of 3.1 in2 and the equivalent flow is 100 in/sec

I may have made an error somewhere, but if I didn't then the highest flow rate here is equivalent to 86% of that...

 

[karimwassef]

The unknown is whether the propeller powerheads can handle being used under pressure.

 

[karimwassef]

The 15" height is really an arbitrary choice... I started with the full tank height of 24" with an overflow that's 2" lower, and then took 3" off the top and bottom to avoid surface waves and sand blowing action. That brought me down to 16". The plastic on the top and bottom of the box (using 0.5" thick acrylic) consumed another 1"... that's 15".

I could choose to raise the overflow to 1.5" and reduce the keepouts to 2". That would return 2.5" back to the box.

I don't think I need to though. I think this works.

 

[karimwassef]

back in post 351 (http://www.reefcentral.com/forums/showpost.php?p=23971063&postcount=351) I covered different pump options. The JGP 30000 does 7900gph but consumes a lot more power. Something like that (5x instead of 10x) could be an option too. Unfortunately, each consumes 660W... or 3.3KW of power use

If that's really the case, I might go back to the piston and paddle concept.

If only someone made a truly laminar uniform output propeller pump that's in a square stackable form.

 

[karimwassef]

Back to the pressure drop. We're pushing 40,000 gph through a 6" PVC pipe (equivalent) for about 10" of total length. These are all equivalents...

Using the Hazen-Williams equation on engineering toolbox,

http://www.engineeringtoolbox.com/hazen-williams-water-d_797.html

That results in a total pressure drop of 0.026 ft of head. Not much. I think it looks worse that it really is.

This thing is huge - it's 15" high by 32" wide and only runs for 10"... so it's got a very large surface area and very short travel distance.

That's the zero order (back of the envelope) result ignoring the impact of the corners, but I'll do a more detailed workup.

 

[herring_fish]

I get payed to be an engineer but I think that all that math stuff is just a bunch of witchcraft.

 

[karimwassef]

well... first order calculation is a lot higher than zero order.

The head of pressure is more like 11ft. There are three bottlenecks - I hadn't predicted any of them, but 20-20 hindsight... it's clear now.

The first is the actually run to the entry of the diverter. It's not the cross-section, it's the 2 ft run through it...

The second is the 6" run after the first tee. Again, not so much the cross-section as it is the run.

The third was in the third finest separators. I have little raisers that increase the neck length from 0.25 to 0.5"... that was just as painful as the other two bottlenecks. Each added about 1.5" of head !

I'll go back and fix those.

The funny thing is that the final comb structure and some of the finer features.. they weren't as big of a problem...

 

[salty joe]

When you compare the baffle to PVC pipe, are you also considering surface area (friction)? Looks like an awful lot of surface area.
If you spent the money for all those pumps and it did not quite work, would you end up using the pumps for your tanks anyway?

FWIW, I have a pair of prop powerheads driving water into my ~24"x40" filter with a 3" drain. It raises the water in the filter about 1". Tank and filter are on the same level.

Just saw the 11' head. I see no way prop pumps working, if that's the case.

 

[Michael Hoaster]

This thread hurts my head, but I'm following anyway…

I'm rootin' for ya, Karim!

 

[karimwassef]

This was a very useful exercise... thanks guys.

I pushed the pressure down to 4 ft of head, but it took a lot of going back and simplification. I'm still concerned about the difficulty of flow redirection after a Tee, but then again, with that much pressure, maybe it'll work as is.

<a href="http://s1062.photobucket.com/user/karimwassef/media/0_zpsfgokdzei.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/0_zpsfgokdzei.jpg" border="0" alt=" photo 0_zpsfgokdzei.jpg"/></a>

 

[Gorgok]

11 feet? Ouch. Though i have never really thought about how well a powerhead would handle head that seems like a lot.

There are prop pumps built that can handle more than that in head, but generally they seem to be the same size as the pipe.


Now that im on a deeper pass of this idea i actually wonder how well props multiply outputs... They have a fixed maximum velocity they can impart on the water, based on the pitch and rpm. If all the water in the channel is at that velocity any more simply won't help (basically be free wheeling). It is a big channel, and lots of restriction so you probably aren't hitting it, but its something to keep in mind.

 

[karimwassef]

Thanks. I'm down to 4 ft of head now.

There's a few pinch points but the breakdown is pretty uniform between the elements. So while I think I can improve it a little - maybe down to 3ft, there will always be an element of pushback.

When I mock up the Lego prototype, I should see someone this in action.

 

[karimwassef]

This thread hurts my head, but I'm following anyway"¦

I'm rootin' for ya, Karim!

Welcome to the club. This has been hurting my head for weeks. I'm stubborn / um tenacious / and when I have a difficult goal, I tend to dig in until it's done.

This has been frustrating - every time I think we're there, something is missing.

 

[karimwassef]

When you compare the baffle to PVC pipe, are you also considering surface area (friction)? Looks like an awful lot of surface area.
If you spent the money for all those pumps and it did not quite work, would you end up using the pumps for your tanks anyway?

FWIW, I have a pair of prop powerheads driving water into my ~24"x40" filter with a 3" drain. It raises the water in the filter about 1". Tank and filter are on the same level.

Just saw the 11' head. I see no way prop pumps working, if that's the case.

That was the biggest difference between my zero order and first order calcs. I just used the effective pipe the first time, but when I went back and calculated the hydraulic diameter of a duct - it hit me back hard.

For a wide and high duct, the equivalent pipe is fine. For a high aspect ratio (like 15" x 2" .... Or smaller), the hydraulic diameter is 2x separation. So, if I have a mile deep duct, that's only 1" wide, the hydraulic diameter is 2", regardless of the cross sectional area... That really messed things up and it's why I went from 0head to 11head.