Does air flowing through a tube cause a pressure change?

[TheRumpus]

I have been involved in a fairly furious debate about a fluid dynamics experiment regarding the pressure of air passing through a tube attached to a moving car. It is similar in concept to the issue of whether or not an opening in an aeroplane fuselage would suck people out, where the opening did not interfere with the airflow.

The situation is in the accompanying diagram.

So the controversial issue is whether or not the pressure monitor would show a pressure which varies with the velocity of the car and whether the pressure it reads would be ambient pressure.

This is more tricky problem than appears, and many many people who ought to get the answer right and are versed in physics seem to get it wrong initially. The issue to consider is whether Bernoulli's equation applies etc etc. I feel very strongly for one position but I thought I might ask for opinions on this first, hopefully which would confirm my position.

I have a friend who is proposing to test this in the field with a car and some plastic tubing of about 2cm in diameter. If anyone can suggest potential problems with the test ( such as the difficulty of aligning and maintaining alignment of the T piece ) I would appreciate it.

 

[anorlunda]

Are you familiar with pitot static systems used to measure airspeed of airplanes?

https://en.m.wikipedia.org/wiki/Pitot-static_system

 

[TheRumpus]

Thank you for the welcome :)

I am and pitot tubes are not relevant because they are blocked tubes, there is effectively no airflow through the tube as there is in this experiment.

 

[f95toli]

Another thing worth googling would be differential pressure devices. These are e.g. used to measure the flow rate in water pipes (this is why flow rates are sometimes given in the unit of "pd" )

 

[russ_watters]

Thank you for the welcome :)

I am and pitot tubes are not relevant because they are blocked tubes, there is no airflow through the tube as there is in this experiment.

He said "pitot static", not just pitot...

 

[TheRumpus]

Oooooooops! He did :D

I will investigate.

 

[TheRumpus]

And a pitot static system is based around a pitot tube, which is effectively a blocked tube, which is not relevant to my diagram.

 

[TheRumpus]

Another thing worth googling would be differential pressure devices. These are e.g. used to measure the flow rate in water pipes (this is why flow rates are sometimes given in the unit of "pd" )

Such devices are interesting, but they rely on the venturi effect, constricting the fluid flow, and that is not happening in my experiment.

But if you think it is relevant, please suggest an effect and a conclusion as to whether a pressure difference dependent on velocity would occur.

 

[f95toli]

Such devices are interesting, but they rely on the venturi effect, constricting the fluid flow, and that is not happening in my experiment.

Why not? If the tube is not in any way affecting the flow, what does it do?

 

[TheRumpus]

Why not? If the tube is not in any way affecting the flow, what does it do?

Well that is the question!

Do you believe from the diagram it is doing something or nothing?

If it is doing something, presumably this would manifest itself in a pressure change detectable by the pressure monitor, if it is doing nothing, then presumably the pressure monitor would not detect any change from ambient, or possibly there is another alternative.

 

[anorlunda]

And a pitot static system is based around a pitot tube, which is effectively a blocked tube, which is not relevant to my diagram.

No it is not necessarily blocked. Even if there was a hole in the diaphragm allowing a through flow, there would still be a pressure difference. Airspeed is proportional to the pressure difference between the two ports.

On your diagram, as long as there is a flow resistance in the pipe, there will be a pressure drop.

Another analogy, a ramjet engine.

 

[TheRumpus]

No it is not necessarily blocked. Even if there was a hole in the diaphragm allowing a through flow, there would still be a pressure difference. Airspeed is proportional to the pressure difference between the two ports.

On your diagram, as long as there is a flow resistance in the pipe, there will be a pressure drop.

Another analogy, a ramjet engine.

In your diagram the tube is effectively blocked. They have to be otherwise the pitot tube would not pick up all of the dynamic pressure.

Well I did say not to assume viscous forces in the diagram.

The diagram you offer shows a very severely constricted tube, which is quite different to my diagram.

I don't quite know how reasonable it is to assume a lack of viscous force, I was proposing 2cm tubing, and was hoping this would be enough to be able to discount the viscous forces.

 

[Chestermiller]

It is definitely not similar to people being sucked out of a hole in the fuselage. In a plane at cruising altitude, the atmospheric pressure outside the cabin is only a small fraction of the cabin pressure. This is not the case in the system you describe.

 

[TheRumpus]

I should have mentioned in describing my plane analogy that it was an unpressurised plane.

 

[Chestermiller]

I should have mentioned in describing my plane analogy that it was an unpressurised plane.

Even so, are you aware that the atmospheric pressure at cruising altitude is less than half of that at ground level?

 

[TheRumpus]

I am, for the cruising height of a jet aeroplane, the atmospheric pressure is indeed less than half that of ground level.

My experiment is concerned with pressure effects or lack of them due to relative air velocities. Which is also valid for aeroplane scenarios and openings in the fuselage, providing the aircraft are unpressurised.

 

[TheRumpus]

A slightly improved diagram of the scenario

 

[russ_watters]

And a pitot static system is based around a pitot tube, which is effectively a blocked tube...

Not really, no: the pitot port and static port are separate pressure ports on separate tubes that just happen to be concentric. You can measure both pressures together to get the difference or you can measure them separately. The behavior of the static port is exactly what you are asking about. And the answer is, as the name implies, it reads the static pressure.

However, while you have expressed that this is in the context of getting sucked out of an airplane, the scenario is not the same as your setup because of the influence of the fuselage.

 

[TheRumpus]

Yes it is.

The pitot tube measures the total pressure which is a combination of dynamic pressure and static pressure. It does so using what is effectively a blocked tube ( pressure chamber in the wikipedia diagram). The static pressure is also measured in this system using a SEPARATE port. The difference between the two gives the DYNAMIC pressure which is related to windspeed, then this is used to give indicated air speed.

(A pitot-static TUBE ( not system) which is another alternative incorporates the static intake from another SEPARATE ISOLATED part of the structure which holds both the static and the pitot elements. )

The scenario I have described is exactly the same as having an opening on the side of the fuselage ( where it is straight and away from wings etc which accelerate the air) of an unpressurised aeroplane. Where the cross section of the fuselage does not change with distance along the fuselage, the air is not accelerated and so the fuselage should not affect airflow or pressure. Where the fuselage does change cross section, e.g. near the front, near the wings, and near the end then airflow will be affected. My contention is that where the cross section of the fuselage is constant with distance along the length of the fuselage the same effect will be had on airflow as in the tube in the instrument in my experiment. Similarly a train where it has a constant cross section with distance along the train, will not cause airflow changes, except due to viscosity issues and boundary layer issues, which I have tried to ignore in this experiment.

 

[Chestermiller]

Yes it is.

The pitot tube measures the total pressure which is a combination of dynamic pressure and static pressure. It does so using what is effectively a blocked tube ( pressure chamber in the wikipedia diagram). The static pressure is also measured in this system using a SEPARATE port. The difference between the two gives the DYNAMIC pressure which is solely related to windspeed, then this is used to give indicated air speed.

The scenario I have described is exactly the same as having an opening on the side of the fuselage ( where it is straight and away from wings etc which accelerate the air) of an unpressurised aeroplane. Where the cross section of the fuselage does not change with distance along the fuselage, the air is not accelerated and so the fuselage should not affect airflow or pressure. Where the fuselage does change cross section, e.g. near the front, near the wings, and near the end then airflow will be affected. My contention is that where the cross section of the fuselage is constant with distance along the length of the fuselage the same effect will be had on airflow as in the tube in the instrument in my experiment. Similarly a train where it has a constant cross section with distance along the train, will not cause airflow changes, except due to viscosity issues and boundary layer issues, which I have tried to ignore in this experiment.

This is a pretty straightforward problem to model (analytically). Do you have any idea how to do it? Are you aware that the problem you have described is the same as the car and T being stationary, and the wind (in the far field) blowing past it at constant far field velocity v?

 

[TheRumpus]

I am aware of that completely.

Most of the people I am having a furious debate with realize that, but the scenario I describe in the diagram was the scenario that was agreed to consider.

 

[Chestermiller]

I am aware of that completely.

Most of the people I am having a furious debate with realize that, but the scenario I describe in the diagram was the scenario that was agreed to consider.

So what is the mathematical solution for the velocity and pressure distribution outside the car, neglecting the T?

 

[TheRumpus]

Well where the car is not changing cross section, the air is at something close to ambient pressure. Where the car is changing cross section air is accelerated as indicated in my diagram.

I don't quite understand why you are so focussed on what is happening near the car, my question is regards what is happening at the T

 

[Chestermiller]

Well where the car is not changing cross section, the air is at something close to ambient pressure. Where the car is changing cross section air is accelerated as indicated in my diagram.

I don't quite understand why you are so focussed on what is happening near the car, my question is regards what is happening at the T

I didn't ask for words. I asked for actual equations.

If you can't figure things out without the T present, you certainly won't be able to work it out with the T present.

 

[TheRumpus]

This is odd.

I asked the question in the OP as to what people think would be the relationship between the pressure recorded by the pressure monitor and the speed of the car, and now I am being asked to provide answers!

I will however provide what I believe to be an answer to your question. Around the car where the cross section is constant along its length ( which is the parallel to the direction of travel ) the pressure should be ( at least beyond the boundary layer ) at ambient pressure because the car has not accelerated the air. Bernoulli's equation shows that for air that is no accelerated, i.e. no velocity change, there should be no pressure change.

In my diagram I show areas where the air is accelerated and Bernoulli's equation would suggest that where air is compressed by being displaced, the air pressure would increase, and where it is accelerated it would decrease. But again, these issues with the car are almost irrelevant to the T because the T is some distance from the car. This focussing on the car is very odd.

 

[Chestermiller]

This is odd.

I asked the question in the OP as to what people think would be the relationship between the pressure recorded by the pressure monitor and the speed of the car, and now I am being asked to provide answers!

I will however provide what I believe to be an answer to your question. Around the car where the cross section is constant along its length ( which is the parallel to the direction of travel ) the pressure should be ( at least beyond the boundary layer ) at ambient pressure because the car has not accelerated the air. Bernoulli's equation shows that for air that is no accelerated, i.e. no velocity change, there should be no pressure change.

This is correct, although, certainly, the air that reaches downstream (where the cross section is constant) has been accelerated at the nose of the car. However, by the time the air reaches the downstream locations, the pressure has returned to ambient. (Of course, there is no boundary layer if the fluid is inviscid).

Now, if the T is small in diameter and located in the constant-cross section region of the car, why would you think that there would be any significant interaction between the flow through the T and the air flow that is now parallel to the car? My intuition would say that the flow velocity through the T would be v. And, if this were the case, the pressure within the T would be ambient too, would it not?

 

[TheRumpus]

I agree completely, I think, I certainly agree the airflow through the T would be v, ignoring viscous effects, but a lot of people think that because there is a relative velocity of air in the T to the air in the stem of the T, there should be an induced pressure difference. You didn't quite make clear if you thought this would be the case or not, you seemed to be implying that there would be no pressure difference and that the pressure monitor would record ambient pressure and that there is no link to the velocity of the car and the pressure recorded by the pressure monitor . Is that correct?

 

[Chestermiller]

I agree completely, I think, I certainly agree the airflow through the T would be v, ignoring viscous effects, but a lot of people think that because there is a relative velocity of air in the T to the air in the stem of the T, there should be an induced pressure difference. You didn't quite make clear if you thought this would be the case or not, you seemed to be implying that there would be no pressure difference and that the pressure monitor would record ambient pressure and that there is no link to the velocity of the car and the pressure recorded by the pressure monitor . Is that correct?

Yes.

 

[TheRumpus]

Excellent, we are in complete agreement.

No one else so far besides yourself has expressed an opinion on the question I had asked, a bit timid perhaps? As the result is counter intuitive for many people and they are inclined to think that the flow of air would cause a pressure difference.

 

[russ_watters]

Yes it is.

The pitot tube...

You have gotten stuck on the "pitot" part and never let it go. The static port is what you should be focusing on. To help, here's a picture of a static port that is totally independent of a pitot tube:

The scenario I have described is exactly the same as having an opening on the side of the fuselage ( where it is straight and away from wings etc which accelerate the air) of an unpressurised aeroplane.

The photo above of static ports is exactly that. A few notes on your scenario:

1. This measures static pressure of the airstream moving past the plane...or car.
2. In a slight caveat to the above; in situations where the port is mounted somewhere where airflow has been deflected going around an object, the pressure measured may not exactly be the static pressure; it may be lower.
3. In your device, the tube for the airflow is superfluous. You've essentially defined it to not even be there.
4. It doesn't serve much of a purpose to mount it far away from the car except maybe to reduce the turbulence around it and avoid #2.

Where the cross section of the fuselage does not change with distance along the fuselage, the air is not accelerated and so the fuselage should not affect airflow or pressure. Where the fuselage does change cross section, e.g. near the front, near the wings, and near the end then airflow will be affected. My contention is that where the cross section of the fuselage is constant with distance along the length of the fuselage the same effect will be had on airflow as in the tube in the instrument in my experiment. Similarly a train where it has a constant cross section with distance along the train, will not cause airflow changes, except due to viscosity issues and boundary layer issues, which I have tried to ignore in this experiment.

This is all correct. It seems you do actually understand it...

No one else so far besides yourself has expressed an opinion on the question I had asked, a bit timid perhaps? As the result is counter intuitive for many people and they are inclined to think that the flow of air would cause a pressure difference.

Well;
1. We try to push people toward thinking through to the answers themselves. It's a better way to learn than just to spoon-feed people answers.
2. Intentional or not, your attitude was a bit off-putting and lowered responding to this thread on my priority list.