Experimental proof of Venturi Effect

[T C]

You can't just use any random fan for any random subsonic application.

Which kind of fan is most suitable?

The larger the ratio the higher the required pressure

Is there any way to calculate the necessary pressure?

 

[russ_watters]

Which kind of fan is most suitable?

It's more complicated than just saying "kind of fan". It's still an axial fan, but you can google for photos of them and see how they differ from yours. More, narrower blades, smaller gap, different pitch and rpm, etc. Optimizing the exact fan for the exact application is a fairly deep engineering issue. Your fan is explicitly/purposely designed to *not* generate static pressure.

Any quality fan that is meant to be ducted (or otherwise obstructed) and provide static pressure will also be provided with a fan curve by the manufacturer that tells you how its performance varies with back-pressure.

Is there any way to calculate the necessary pressure?

Yes, like I said, by applying Bernoulli's equation to the Venturi effect. By now you really should be able to calculate the component pressures at various points in the system using the Bernoulli equation in addition to using the Venturi short-cut for static pressure only.
https://en.wikipedia.org/wiki/Venturi_effect

[oops, simultaneously deleted duplicate posts. Fixed now.]

 

[T C]

The link you provided has no reference to how to stop spilling over and how much static pressure is necessary to push a specific amount of flow to pass through a convergent nozzle as long as the flow remains subsonic.
Maybe I am not that adept and efficient to calculate it, but I am sure there are people here in this forum who can do that.

 

[russ_watters]

I'm calculating that with a 50% loss coefficient through the nozzle and your specified 1.9 m/s inlet and 19:1 area ratio, you would need 1200 Pa of static pressure at the fan. That's a pretty substantial fan, comparable to a heating/air conditioning unit pushing air through heating and cooling coils, HEPA filters and a large ductwork system. It's probably 10x what your fan is capable of.

 

[russ_watters]

The link you provided has no reference to how to stop spilling over..

The gap. The gap is too big. Reduce the gap. You have to reduce the gap.

and how much static pressure is necessary to push a specific amount of flow to pass through a convergent nozzle as long as the flow remains subsonic.

Obviously, you have to calculate that given your chosen condition. It sounds like you need a walk-through of this. Let's start here:
The Venturi effect page has a basic equation for calculating pressure difference. Do you see it?

 

[T C]

With minimal gap, if the flow is first released into a short tunnel before entering the nozzle, do we get better result then?

 

[jack action]

One question suddenly struct my mind. In a wind tunnel, the flow will be created by a blower and in my experiment too, the flow has been created by a blower. Why there will be no spill over in wind tunnels?

You don't measure the inlet velocity of the convergent nozzle at the outlet of the wind tunnel fan. They are not necessarily the same.

You've been told to take care of spillage in your apparatus. But this is not really your problem. You have to measure the inlet and outlet velocities without changing your setup. Right now, you are measuring the outlet velocity of a fan without a nozzle in front of it (apparatus #1) and the outlet velocity of a fan with a convergent nozzle in front of it (apparatus #2). You cannot assume everything is the same in both apparatuses.

You must have two anemometers: one in front of your fan - with the convergent nozzle over it - and another one at the outlet of your convergent nozzle. Of course, the anemometer you have right now is too big compared to your nozzle and will obstruct the flow to the point that it will affect greatly the measurements. (i.e. they will not represents the values without the anemometers present.)

 

[T C]

In short, the problem is in the anemometer size and shape, right?

 

[jack action]

The problem is that you change your apparatus in between your measurements. And you do it because your anemometer is too big.

 

[T C]

At present, my biggest problem is to reduce the spilling over of the flow to the minimal and measure the inlet and exit flow after that. Just that and nothing else.

 

[russ_watters]

At present, my biggest problem is to reduce the spilling over of the flow to the minimal and measure the inlet and exit flow after that. Just that and nothing else.

Not really, no. The inappropriate fan and low minimum velocity of the anemometer are big issues too. Other than simply measuring the area/velocity relationship you haven't specified a goal, and any/all of these issues will prevent you from measuring the area/velocity relationship.

 

[russ_watters]

You must have two anemometers: one in front of your fan - with the convergent nozzle over it - and another one at the outlet of your convergent nozzle.

You can use the same anemometer in two different places as long as you don't change the system in between. I was going to suggest on the inlet of the fan. I know that's not as reliable as the outlet, but it would be something. But that doesn't even matter since the fan is stalled and the velocity is lower than the anemometer is capable of measuring. As far as I can tell, the current setup can't be cleaned-up to make it work. He needs to start over from scratch with a bigger fan at least.

 

[boneh3ad]

At present, my biggest problem is to reduce the spilling over of the flow to the minimal and measure the inlet and exit flow after that. Just that and nothing else.

 

[boneh3ad]

You can use the same anemometer in two different places as long as you don't change the system in between. I was going to suggest on the inlet of the fan. I know that's not as reliable as the outlet, but it would be something. But that doesn't even matter since the fan is stalled and the velocity is lower than the anemometer is capable of measuring. As far as I can tell, the current setup can't be cleaned-up to make it work. He needs to start over from scratch with a bigger fan at least.

That anemometer is large enough (as a fraction of the duct size) that it almost certainly appreciably changes the flow.

 

[russ_watters]

That anemometer is large enough (as a fraction of the duct size) that it almost certainly appreciably changes the flow.

Agreed. And it could be increasing or decreasing the velocity depending where the fan is on the curve. One of the reasons I suggested a bigger fan is both the anemometer size issue and the gap/re-circulation issue are decreased just by getting a physically larger fan/system (assuming same gap size in cm).

 

[boneh3ad]

Another option is rigging the system so that the fan is actually downstream of the funnel drawing air through it rather than pushing air through it. This is how wind tunnels operate.

 

[paradisePhysicist]

Summary:: I want to know how much well experimented Venturi Effect is. If there are people here who experimented with Venturi Effect and can give references to experiments that can show the validity of Venturi Effect.

Venturi effect is known for centuries. And most probably that's why experimental proofs are rare because it's already accepted. But, I want to know how close real results are in case of experiments regarding Venturi Effect. I am especially interested in results of experiment regarding velocity of fluid through a convergent nozzle in case of subsonic flow. I want to know this just to see how other factors like viscosity can effect the outcome.

Cool video with cool music:
https://ytprivate.com/watch?v=UNBWI6MV_lY

I never heard of Venturi effect until recently, and I didn't watch the whole video yet, so I am not sure. But my opinion of how it works is the pressure makes the gas move, the moving gas has less pressure because it is not moving vertically anymore but horizontally. This is how it has less pressure even though also in a tighter space.

 

[russ_watters]

My first attempt at this is shown in the photos below. It went much worse than I expected. Here are the results:

As you can see in the results, the velocity ratio is nowhere close to the area ratio. This is likely due to poorly developed (non-uniform) and swirling flow, caused by the "tube" being installed on the outlet of the fan instead of the inlet (the large central obstruction on the fan is probably part of that). In other words; the measurements are likely unreliable. I'll see if I can re-do the experiment with the fan reversed.

The poor pressure performance and corresponding loss of airflow with the addition of the obstruction was not a surprise. Even with a relatively modest area ratio and associated static pressure requirement, the fan airflow dropped by half and registered essentially no static pressure (0.02" or 5 Pa) in the big box.

Below is a video I uploaded to Youtube showing the performance change as the variable geometry outlet nozzle is actuated closed. Note, the fan shutoff pressure is 0.07" (17 Pa) and the the fan audibly slows due to the added resistance. By comparison, here's a random residential toilet exhaust fan that's a centrifugal blower, which can achieve around 0.6" shutoff pressure (the curve is cropped at the top though):
https://www.broan-nutone.com/getmed...9c93-b67cad266740/Spec-Sheet-695.pdf?ext=.pdf

 

[boneh3ad]

 

[russ_watters]

Early returns on the draw-through design are promising, with results within about 10% of predicted. Details tomorrow.

 

[cjl]

Draw through is almost certainly the way to go. It basically eliminates swirl and also gives much more consistent flow across the cross section, especially if you use a flow straightener on the inlet - a cheap and easy way to do this that I've seen used on homemade wind tunnels before is actually to just buy several large containers of drinking straws and cut them (the entire container, straws and all) to a length of a few inches and fill the inlet with these. You'll also get much better results if you have a diffuser on the exit for pressure recovery - this will greatly reduce the static pressure requirement on your fan, and allow it to operate much closer to its open air performance.

Something like this would be an excellent guide to follow, though I think that guide would get better results with a slightly shallower angle and smoother nozzle transition at the front end, and a bit larger exit on the diffuser. You can even further reduce the flow restriction and get higher velocities still if you put the flow straightener before the constriction, but if you do that, the constriction of the nozzle needs to be smooth enough to keep the flow nicely behaved and not to introduce a bunch of turbulence or separation off the walls. You can see those characteristics on this wind tunnel for example, which is a really good example of sort of what you want to shoot for.

 

[T C]

I have found a few videos (video 1, video 2, video 3) on youtube. In all these videos, the venturi is fitted after the blower i.e. downstream but performing pretty well and without any kind of flow straightener. IMO a cone shaped venturi can perform better than square shaped one created by russ_waters.
@cjl, if a fan/blower performs better in open air instead when attached to the throat i.e. draw through mode, what's the use of it?

 

[russ_watters]

if a fan/blower performs better in open air instead when attached to the throat i.e. draw through mode, what's the use of it?

What's the use of a duct? Lots of uses/reasons to keep air contained. Any time you want a coherent mass of air traveling from one place to another you need a duct. Or if you pressurize it or want to force it to go somewhere it doesn't want to go (like through a heating/cooling coil). You could just as easily ask why we put water in pipes!

 

[T C]

In case of a wind tunnel, IMO the main purpose is to create a flow at a specific speed. If a open air blower can generate more speed in comparison when fitted at the throat of a convergent nozzle shaped duct, why should one need that?

 

[russ_watters]

In case of a wind tunnel, IMO the main purpose is to create a flow at a specific speed. If a open air blower can generate more speed in comparison when fitted at the throat of a convergent nozzle shaped duct, why should one need that?

Because a non-ducted fan doesn't generally generate higher velocity it only generates higher flow volume (if you use a proper fan). And it isn't coherent/uniform. It is also a huge waste of energy; most real/large wind tunnels are closed circuit.

 

[T C]

As per the post from cji, a non-ducted fan can generate more flow than a ducted fan especially when fitted at the throat of a convergent nozzle shaped duct. You are saying the opposite. Which one should I consider to be correct?

 

[russ_watters]

As per the post from cji, a non-ducted fan can generate more flow than a ducted fan especially when fitted at the throat of a convergent nozzle shaped duct. You are saying the opposite. Which one should I consider to be correct?

I don't see where he said that, but regardless, "flow" usually means volumetric flow rate, not velocity.

You have a bad habit of trying to connect unconnected things.

 

[T C]

and allow it to operate much closer to its open air performance.

That clearly means its ducted performance is worse than open air performance.

 

[russ_watters]

That clearly means its ducted performance is worse than open air performance.

"Performance". That word evidently doesn't mean what you think it does.

 

[T C]

Then kindly explain it.

 

[russ_watters]

Then kindly explain it.

I did!

 

[boneh3ad]

In case of a wind tunnel, IMO the main purpose is to create a flow at a specific speed. If a open air blower can generate more speed in comparison when fitted at the throat of a convergent nozzle shaped duct, why should one need that?

For one, a wind tunnel has a lot of other considerations other than just a specific speed. You have to worry about flow uniformity, angularity, and free-stream fluctuations.

 

[T C]

I just want to know whether the velocity will be same or less in case of flow through design or open air design.

 

[russ_watters]

I just want to know whether the velocity will be same or less in case of flow through design or open air design.

It depends on the fan. If the fan is capable of generating substantial static pressure the velocity through a duct can be higher than without.

But the volumetric flow rate will never be higher with a duct (for an axial fan, with the possible exception of a shroud/inlet cone).

 

[russ_watters]

Here's that second set of results in the draw-through configuration:

Notes:

• I tried to read the velocity on the short side of the large box and got zero (which is really just <50 fpm). The transition is too steep between the small and large box on the short side, and the flow separates from the side of the box, leaving a dead spot up against the side. Shape of the tunnel matters a lot.
• Shutoff pressure was 0.06", about the same as in the blow-through configuration. I didn't attempt to limit re-circulation at the tips. I suspect this wouldn't help much due to the noticeable rpm noise drop when closing the variable geometry inlet. In other words, while the fan shape is bad, the torque requirement for the asynchronous motor is a bigger problem. Most real-world HVAC fans or others with big motors are constant RPM and the torque increases as load increases (or the motor dies trying).
• Notice that in this test I was able to achieve a higher velocity in Box 2 than with the bare fan. But just barely. With a better fan, capable of generating static pressure when restricted, the velocity could be much higher in a small duct than at the fan outlet for a bare fan. But airflow (volumetric) will pretty much always be lower in a duct than on a bare fan.