Large volume laminar flow machine

[karimwassef]

Great ideas, herring. I especially like these.

If you have the divider in the center of the tank, the paddle would only need to move 4 inches. If the divider is at the 1/3rd mark, then I think that the paddles would move 8 inches.

In either case, that is not too far. I also think that the more paddles that you have the less unwanted wave action you will get at those paddles.

Excellent and very clear descriptions and illustrations.

 

[karimwassef]

Putting the paddles and the separator inside the tank actually simplifies the mechanism. It doesn't consume 1/3 of the tank left to right.

It does consume the thickness, but that's actually more practical IMO.

 

[herring_fish]

I m on the east coast so I am off to bed but which aplication are you wanting to explore further?

 

[Gorgok]

Won't those inside spaces/paddles basically be wasted energy as its just moving a (semi-sealed) tank around... Only the paddles that touch the rest of the raceway affect the flow there really.

If you just moved one paddle from the left to the right the whole way you would be moving much more water on the front of the raceway than with the short stroke of the multi setup.

Taking this track and paddle idea further though you could incorporate a rotator into the single paddle. So it acts like an oar when rowing. Perpendicular on the stroke, parallel on the return or when idle (so as not to kill the flow already established).

Still more complicated than necessary i think, especially if using a narrow/wide raceway tank like the second rectangle shown. Having the returns blow one way in the narrow bit for a time, then the other should create the same effect. And it keeps that space of the tank less hostile to life i would think.

 

[karimwassef]

Gorgok- I think th difference between a paddle and a pump is loss and flow pattern. The pump would need to channel the flow through pipes, a pressurized chamber, and a circular output with a narrow round flow pattern. That's a lot of wasted energy and needs the pattern to be laminarized. Pumps are also single directional adding to the complexity in reversing the flow (valves, etc...). The exception would be the torpedo pump from a few pages back. It's a direct flow it's a 3" intake and outlet and can even be used as a powered (submersible). It's still one directional, but I can see running two in series in opposite direction with a shared chamber between them. It may take several (4 pointing left, 4 pointing right, a chamber between the left and right) to get there.

It can work but it's getting very expensive to duplicate the paddle with a pump.

Edit- the pumps do have one key benefit: flexibility. The paddles are limited in stroke length in the channel. This is a hard physical limit. Pumps can operate for minutes in the same direction (emulating an infinite channel). So the paddles work for the 4-15 seconds I am looking for, but we still haven't resolved the surface wave effect.

I think the multipaddle concept is based on the idea of using the wasted volume between the paddles to create momentum. I don't think the seals between the paddles and channel will be tight, so this reduces the stress per paddle and uses the volume between them to get the motion going one way with less energy once started. Needs some engineering.

 

[karimwassef]

So, thinking more about this. The reason I went round was to avoid turning the push into surface waves. Not sure I proved that it works yet, especially when reversing direction.

The channeled paddles idea has the same question with one more complication: the ends need to be shaped like circles to channel the flow into the laminal rectangle block in the front.

That radius and shape needs to work at the speeds we want to create laminar flow.

 

[karimwassef]

The outlets need to be channeled. This means that the top of the tank at the outlets needs to be sealed to avoid the surface wave effect.

 

[herring_fish]

The outlets need to be channeled. This means that the top of the tank at the outlets needs to be sealed to avoid the surface wave effect.

What do you mean? I just can't quite paint a picture of what you are saying. Can you rewrithe it in a different way?

Also:
So are you still looking at both the round and the rectangle designs?

 

[herring_fish]

In post 113, I talk about the problem of a single paddle and show the splashing. Since you are trying to accelerate all of the water in the display tank at the same time, using more paddles allows more places to get a grip on the mass with less wave action and splashing. Yes that is a waste in one respect but you have to move it all if you want laminar flow. That is what is happening in the scuba experience. It’s all moving one way and then the other.

This wave action needs to be minimized to keep the water as laminar and ripple free as possible, in the front of the divider. Of course that is just how it looks in my imagination and from what I have seen. Experimentation at that scale is the only way to know which of these several ideas will work best in this application.

I have also thought about making a box that is fully closed below the water line so that it displaces the water between the front and back paddle. This would get most of the same splashing of the one paddle system but there would be less total volume in the system for the paddle (box) to accelerate. Again, there would still need to have close gaps between the back wall of the display tank, bottom and center divider.

This would be true of the circular tank or the linear. Depending on the design, the total volume might be cut by 30 to 40 percent.

Rotating the paddle(s) is a technique that is used to allow the paddle to work in one direction only or one direction more than the other. From what I understand earlier, we want the water to go in one direction, stop and reverse so a straight paddle setup should address that goal.

Any kind of pump(s) that cannot propel the entire cross section of the tank at the required speed would cause non-laminar flow. If it can meet that requirement, you then need to use the straw box or something like it to reduce the chaos of the pump and piping.

Again, if I get around to it, I will try to achieve a less ambitious tube of comparatively laminar water flow going down the center of the tank cross section. Cowling and direction jetting can improve the flow of a large diameter pipe outlet by pulling in water that is near the outlet to amplify the total tube flow size but that is not what we are looking for here unless we can come up with a better set of ideas.

Finally, with a 4 second thrust in one direction, it would be a 9 or 10 second cycle. A 15 second thrust, which is your new outside limit, would be about 30 to 35 seconds. Am I reading you correctly? Of course 10 and 30 seconds is a big difference. It will be interesting to see which end of the range you will be able to hit, if at all.

Something talls me that you get what you want if you are willing to work for it. Good Luck

 

[ericarenee]

The outlets need to be channeled. This means that the top of the tank at the outlets needs to be sealed to avoid the surface wave effect.

I still believe the top water surface will have to have some kind of wave control..Being Sealed or maybe even a flat slotted top just below the water surface to break the waves... but allow some water to move up.....To help the momentum....


This applies to any of the design ideas.. I would like to see a actually test.
I could be wrong....... Hummmmmmm Erica Wrong.... :uhoh3:

 

[karimwassef]

I agree on a sealed top to channel the flow for some distance.
Circle or oblong rectangle? Not sure yet.
Multiple paddles vs one vs pump. Not sure yet.

10svs. 20sec... Yes. I think this changes with size. On a small volume, the inertia s different than on 1000gal. The time it takes to reverse without huge waves changes. That's the struggle.

 

[herring_fish]

I also agree with the top plate wave limiter idea.

Make sure that it is stout, like ¼ inch plastic or something. A 5 or 10 gallon wave might be too big or too small but almost any wave going that fast could be a problem.

I have used a 2 gallon dump bucket that has pushed flimsy plastic up and out of the way. .. the wave that comes from a dump bucket is nothing compared to this potential issue.

In a mockup though, you might just get a little water on the floor and can adjust afterwards.

 

[karimwassef]

For a mockup, it'll be small and thin acrylic is ok.

For a real setup, this is likely just as thick as the tank's walls.

 

[karimwassef]

Where did you find this? What's it called?

 

[herring_fish]

Just Google these and look under Images:

linear guide rail and carriage
linear guide rail and carriage
rail and carriage
carriage and rail
linear rail
circular rail
curved rail
curved rail and carriage

 

[karimwassef]

thanks. interesting options. This one shows some potential benefits in controlling motion around a corner

 

[karimwassef]

most are screw driven, so I don't know how fast these can be... or how much power they can deliver.

 

[herring_fish]

Yes there are a ton of options. Somebody makes a solution for whatever you come up with. I can give you some more information once you get closer to making a few decisions that will narrow things down.

Unfortunately, I have to keep being the Grinch here. I have to keep reminding us about the scale of what you are wanting to do. This would be a lot of force that will be exerted against the water. That reversing action is a hard nut to crack but do-able. The expense may not be prohibitive at all "¦.but I think you will have to think about beefy construction.

I keep guessing. How many gallons do you plan to have in your tank? If it is 2000 pounds, imagine moving your car back and forth with an electric motor.

Personally, I would like to steer you into straight linear or circular driving but not both. If you go with circular, you probably don't need rails and carriages. If you go with the raceway, I would just stick with linear only.

I really like the piston, rod and crank shaft style of getting reciprocal movement for either the linear or the circular system.

Hopefully, other people will have some input on what we are looking into or have an entirely different direction that will help us. Otherwise, a few phone calls or Skype group calls would shorten the design cycle.

Please help us. Just throw out your great or stupid ideas. Perhaps we can use it as is or fill in the week spots and fine tune it. There is strength in numbers! Come on please chime in.

 

[Gorgok]

most are screw driven, so I don't know how fast these can be... or how much power they can deliver.

Power isn't a problem with a screw, or shouldn't be unless its made of garbage... I see paddles or the joints to the carriage breaking way before the actual drive. Speed could be an issue, but higher speeds will make taller waves anyway.

On the multi-paddle thing, lets just go by a perfect seal first... If you move the thing 5" in a 5" wide channel 5" tall you have displaced ~1/2 a gallon of water from the front of the setup and drawn in the same at the back. The interior space just moved... It could be made of foam or air or water, it doesn't affect the rest of the system other than adding (live) mass to the carriage. If you move a single paddle the same 5" you displace the same 1/2 gallon with the same wave on the outside paddles.

If you move that carriage instead something like 25" since now you only have the one paddle so room isn't an issue you instead displace ~2.5 gallons.

Now it won't have a perfect seal, so the exterior paddles will bleed some when it starts to move and the interior ones will catch some of it, and when it stops the exterior paddles will block most of the flow the interior ones make but not all. This might raise the efficiency of a multi paddle setup over a single one a few percent, but 'total displacement' seen by the tank is still based on the outside loop. Being 80% efficient at moving 2.5 gallons will beat 90% efficient at 1/2 a gallon.

All this thinking about stroke stuff got me thinking of engines, so now we can think about the bore too. With a narrow channel you need to move farther to cause the same displacement as a wide one. And the faster you try to accelerate the water the taller a wave it will create at the paddle. So at the same speed, a wider channel should create a smaller wave with more displacement.

How much displacement will you need to actually see an effect on the live side of the loop? Thats probably a percentage of the volume on that side... maybe 25% to start? If you can shift 25% of the volume it should be pretty noticeable i think, at least on a 4+' tank that would be a pretty big deal... Calculations for the minimum size to hit that should be pretty simple in any given tank. Stroke is limited to total length minus a little bit. Height is just he water level. Then the only variable you really have to play with is the width. On a 75 thats maybe 44" of stroke and 20" tall. With 4" of channel you displace 15 gallons with roughly 58 gallons of display or 25%. At 5" its 19 and 54, or about 35%. At 3" its 11 and 62, or 18%. Even at 2", 7.5 gallons for 66 its about 12%.

None of those sound too bad, but it is still short lived, always limited to stroke unlike with a pump (at least until you rotate/raise the paddle out at the end of a stroke).

 

[karimwassef]

This is 8' x 3.5'

The display tank is 2'
The overflow is 0.5'
The channel is 1'

So the ratio is 1:2

If we can make a good seal, one paddle is fine. otherwise, more paddles should make up for the poor seals.

<a href="http://s1062.photobucket.com/user/karimwassef/media/1_zpsvshs5lvt.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/1_zpsvshs5lvt.jpg" border="0" alt=" photo 1_zpsvshs5lvt.jpg"/></a>

The display tank is too small ... 4.5' x 2' but it can be scaled. in length.