Designing a Closed Loop Water Tunnel for Fluid Dynamics Experiments

[dcpopham]

I am trying to design and build a small scale desktop water tunnel so I can observe the fluid dynamics of airplane models. I was thinking of a closed loop system using a long fish tank with a circular hole at both ends. PVC pipe will be attached at both openings. how does going from a round intake to a square fishtank create turbulence? How many linear feet does it take to resolve the turbulence and result in a laminar flow? I suppose since I don't have the specifics of the dimension, I will say I need a 2 mph flow rate. Can someone show me the math to figuring out the distance for resolution of the turbulence?

In addition, does this concept even seem reasonable. Look forward to hearing what you all have to say. Thanks to everyone.

David

 

[john.phillip]

You need a flow conditioner. Is this tunnel about the size you're looking for ?

http://www.rollinghillsresearch.com/Water_Tunnels/Model_0710.html

 

[dcpopham]

Thanks John,

Yes I would say it would be that size but different design. One long fish tank and PVC pipe coming out one end and circling around to the other.

So, what you seem to be saying is that i would need a honey comb baffle of sorts to condition this water that is in the shape of the fish tank. I mean it doesn't do any good to condition the flow in the PVC because once it gets to the fish tank you will create turbulence.

If this is all true... does anyone know what or where I can get this honeycomb flow conditioner?

Thanks all!

 

[john.phillip]

You can experiment and design one, or you might buy some flow conditioners for flow meters and try out the best arrangement to limit the turbulence intensity within the levels you are looking for.

http://www.flowmeterdirectory.com/flowmeter_flow_conditioners.html
http://www.emersonprocess.com/daniel/Products/Gas/accessor/Profiler/Productdetail.htm
http://www.vortab.com/

 

[Danger]

I'm very curious about one aspect of this. How on Earth do you keep the dye from eventually making the water too murky to see the flow through? Is it chemically neutralized after each test run?

 

[Borek]

I'm very curious about one aspect of this. How on Earth do you keep the dye from eventually making the water too murky to see the flow through? Is it chemically neutralized after each test run?

Exactly my thought.

Not that it is that difficult to do - you may add some bleach to water, if the concentration is right dyes will be killed fast enough to be not able to make a full circle, but slowly enough so that you can see water movement around the model.

I am not saying that's the way it is done, but I am sure it can be done this way.

 

[Studiot]

Is water or space at a premium with you ?

Why recirculate rather than a two tank setup?

With two tanks you may be able to avoid a pump, making it easier to control the flow regime.

 

[EnglishScient]

honeycomb flow conditioner ? Why not use a bundle of plastic drinking straws.

 

[dcpopham]

Straws... that seems like it would work the same to me. Thank you for the suggestion. Does anyone see a problem with this concept or anything I need to keep in mind when using the straws. Thats a lot of work so I only want to do it once.

 

[dcpopham]

studiot, actually I am eventually constrained with both but that is not the reason for the design. A closed loop seems the easiest design. Plus, I need to be able to withstand a flow of 2 mph for a considerable amount of time.

 

[john.phillip]

We worked with straws when we built a wind tunnel some time ago, it served well to our purpose. I'm not sure whether it will work for you, but you should try out.

I remember a tricky part was to build the transition between the blower round pipe and the square test section, we applied a cubic function to make it.

 

[Studiot]

You can still recirculate, and even add a dye cleaner system, with a two tank setup.

The point about pumps is that they are pulsed in some way. Gravity is constant.

 

[dr dodge]

if you come from a round tube with flow/velocity and go into a square tube/tank the transition into the square will induce turbulence and somewhat negate the straws. I would suggest a clear plexi tube and keep all the in/out diameters the same.

as far as upstream and downstream piping, this should help:
http://www.flowcontrolnetwork.com/issuearticle.asp?ArticleID=182

a simple orfice plate with a differential gauge would help to determine the exact flow.
couple that with a good variable flow pump and you're in business.
(added suggestion, rather than desktop, how about something the size of a 2 drawer file with wheels, being as that thing will be pretty heavy when filled with fluid)
Water will get "funky" in no time (green yuck) so a different fluid, or a mix of alcohol/water will help that

good luck, and if they aren't too pricey, I'll take one

dr

 

[RonL]

I am trying to design and build a small scale desktop water tunnel so I can observe the fluid dynamics of airplane models. I was thinking of a closed loop system using a long fish tank with a circular hole at both ends. PVC pipe will be attached at both openings. how does going from a round intake to a square fishtank create turbulence? How many linear feet does it take to resolve the turbulence and result in a laminar flow? I suppose since I don't have the specifics of the dimension, I will say I need a 2 mph flow rate. Can someone show me the math to figuring out the distance for resolution of the turbulence?

In addition, does this concept even seem reasonable. Look forward to hearing what you all have to say. Thanks to everyone.

David

I came up with too many options, but you might consider a vertical tube with a tank above and below, each having sufficient volume to minimise turbulence. A diaphragm pump to handle grit and not be affected too much.
The descent of water would be completely free of turbulence from what goes on in the square tanks as water is being moved from bottom to top (if the correct baffling is used).

To add to what dr dodge has said, consider using a colored media so fine as to show the same action as ink, different colors in something like glass or sand with just enough variation of size or mass to be filtered at different stages of velocity.

Just a start point.

Ron

PS I forgot to mention two diaphragms in opposed cycle to eliminate pulses.