Pitot Tube will not measure stagnation pressure

[mgpaul]

I have made a pitot tube that I am using to measure the velocity of air in a pipe. First, I placed the probe into the pipe such that the hole in the probe was parallel to the flow to obtain static pressure. It did this successfully and I proved this using a piezometer. However, I then turned the tube such that the hole was normal to the flow to obtain stagnation pressure, but I am getting the same readings as I did for measuring static pressure. I know that since my flow is moving the stagnation pressure should be higher. Any suggestions as to why this is happening? I have attached a basic 3d model of what my probe looks like.

 

[boneh3ad]

Wait do you have a hole in the front and side of your tube leading to the same interior chamber?

 

[mgpaul]

No, sorry the view in the picture does not show it. The only way for the air to get into the tube is through the hole you can see, the other end of the tube is sealed.

 

[boneh3ad]

What are the two holes we can see in your drawing, then?

 

[russ_watters]

How much pressure are you reading and do you have an idea how much velocity to expect?

 

[mgpaul]

the hole near the origin is where the flow is coming into the tube, and the hole at the end is where my pressure gage is attached to.

I am reading about 22 psi with no velocity and approximately 9 psi when I am flowing air. However, no matter which way the hole is facing, it reads the same pressure, which I am sure cannot be right. And no, I am not sure how much velocity to expect but I am not to concerned with it. I just need the tube to measure stagnation pressure when the hole is normal to the flow, and static pressure when it is parallel to the flow.

 

[cjl]

Do you have any idea what your flow velocity is (even roughly)? Also, what is your accuracy and resolution of your pressure gauge?

 

[rcgldr]

It seems unlikely that the static pressure would drop to 9 psi. I'm not sure what's driving the air through the wind tunnel, but the static pressure at the entry end of the wind tunnel will be a bit above ambient, and dropping down to ambient as it flows through and exits the wind tunnel.

Your static port seems ok, but a stagnation port needs to be the open end of a tube (a diferent tube), not a hole in the side of a tube, since the air will tend to flow around the tube, including the boundary layer that covers the static port.

 

[RandomGuy88]

I believe he is saying that he is using the same tube, but separate experiments to measure total and static pressure. The tube shown is rotated about its axis in order to make the hole parallel to the flow and then rotated 90 deg to make the hole normal to the flow. This would be mean that the flow is always perpendicular to the axis of the tube? Is this correct?

A typical pitot probe has the hole at the end of the tube and the flow is parallel to the axis of the tube. This presents less of a disturbance to the flow.

How big is the tube? How big is the hole in the tube? How big is the pipe you are putting this in?

I am reading about 22 psi with no velocity and approximately 9 psi when I am flowing air.

What is your pressure transducer referenced too? Is it a differential or absolute transducer? Either way I think that measuring 22 psi with no flow is wrong because atmospheric pressure should only be about 14-15 psi. I agree that a change of 11 psi is probably wrong. It would help to know approximately what kind of velocity you are expecting. Is it a few feet per second, a few hundred feet per second, supersonic? It does matter.

What kind of transducer are you using? Is it electronic or is it something like a u-tube manometer?

Be sure to check carefully for leaks. You should also check that your transducer is at least operating correctly by blowing into one end of the transducer and see how the measured pressure changes. If there is no change it is possible that the transducer is broken. For example there could be a leak from one end of the transducer to the other.

 

[mgpaul]

My apologies, I have not been receiving emails for some reason and so I had no idea people were still responding. I will try to answer everyone's questions as best I can here.

First off, RandomGuy88 you are correct. That is what I am trying to do.

Second off, all pressure measurements are GAGE measurements (i.e. atmospheric pressure is read as 0 psig).

My test setup measures volumetric flow rate, which came out to about 60 scfh (standard cubic feet per hour) which I backcalculated to be about 50 ft/s I believe, so were shooting for around there

My test is setup in a way that I have a valve which I can open and close, and a pressure regulator to control my outlet pressure. When my valve is closed, My pressure is about 22 psig (i.e. Inlet pressure or tank pressure). When I fully open the valve, I am getting a reading from my pitot tube at about 9 psig, and this is an electronic recording with a calibrated device. I checked for leaks in my system using "snoop" (essentially a soap and water compound that bubbles if there is a leak) and I have gotten no leaks.

my pitot tube OD is 1/4" and the diameter of the pipe where I am taking a reading is about 3/8" ID. While I understand that this could create boundary layer issues, I did try the pitot tube in a 2" ID pipeline with a volumetric flow rate of about 50,000 SCFH (this is subsonic flow). Even in this larger test setup and faster airspeed, I still am not reading a pressure difference when I rotate the pitot tube from normal to flow, to parallel to flow.

Also keep in mind that this test is being run in a pipeline, not a wind tunnel (sorry I should have specified). I also have piezometers and pressure taps (all measuring static pressure) to verify my results. While my probe is within 1% accuracy for static pressure, it is not reading stagnation pressure when I turn it normal to the flow.

Hope this helps explain what I am doing, any further responses are greatly appreciated!

 

[RandomGuy88]

Ok so your probe is setup so that the axis of the probe is perpendicular to the flow right? Meaning it is setup like a cylinder in cross flow.

In this case when you rotate the probe so that the pressure tap is parallel with the flow you should NOT read static pressure in the pipe because the flow will accelerate over the cylinder surface. The static pressure on a cylinder in cross flow is much lower than the free stream static pressure.

Also what is the diameter of the pressure tap on your probe?

Now you have said that the static pressure measured by the probe matches the wall static pressure taps. Given that your probe will not actually measure free stream static pressure I have two ideas to explain the match between the probe and the wall tap.

First your transducer or probe are broken.

Second, if you are measuring static pressure with the wall pressure tap at the same axial location in the pipe at the same time you have your probe in the flow then you are getting incorrect static pressure measurements on the wall tap because of the very significant amount of blockage that you have. Your probe is much to big for this pipe. The probe really should be no more than a few percent of the total cross sectional area of the pipe. With the probe in the flow you will get a large amount of acceleration around the probe because the cross sectional area is constricted. This would causes wall pressure tap to measure an incorrect static pressure.

Also, what do you consider to be "no pressure difference". At this speed and assuming the density of air at approximately atmospheric conditions then the dynamic pressure (the difference between total and static pressure) is only about 0.02 psi. So make sure your device is sensitive enough to measure this.

 

[russ_watters]

My test setup measures volumetric flow rate, which came out to about 60 scfh (standard cubic feet per hour) which I backcalculated to be about 50 ft/s I believe, so were shooting for around there

50 ft/sec is really slow to be measuring the velocity with a pitot tube. That's near the bottom limit of pressure sensors that can measure thousandths of an inch of water gauge differential pressure. So there is almost no way a pressure gauge that measures in PSI will show any velocity pressure -- which appears to be what you are seeing.

 

[RandomGuy88]

50 ft/sec is really slow to be measuring the velocity with a pitot tube. That's near the bottom limit of pressure sensors that can measure thousandths of an inch of water gauge differential pressure. So there is almost no way a pressure gauge that measures in PSI will show any velocity pressure -- which appears to be what you are seeing.

That is not necessarily true. In the wind tunnel I work in we have no problem measuring pressures that low. In fact for some of my colleagues 50 ft/s is pretty fast for the tests they do in the tunnel.

It just depends on the particular transducer you are using.

 

[boneh3ad]

Also keep in mind here that your probe is taking up a substantial fraction of your pipe, so it is going to have a large effect on your flow field.

Would you mind perhaps drawing a picture of your setup just to make it more clear to everyone? It would probably prevent some of the confusion here.

 

[mgpaul]

Ok so I am in agreement here with everybody that using the probe in a 3/8" ID pipe is probably not best so I will no longer be using it in there, thanks for the advice. So yesterday I ran the tube in the 2" ID pipe yesterday (picture of setup attached) and I did see a very small difference using highly accurate pressure transducers (all calibrated). When the hole was normal to the flow, I read an average of 50.0975 psig. When parallel to the flow I read an average of 50.080 psig. However, using bernoulli's equation for incompressible flow (v = sqrt((2*(p0-p1))/rho) I still am only getting about 2.5 ft/s. So yes I am getting a slight difference, but I know I should be reading more. I did not attach a flow meter to my setup as another engineer was using it, so I do not have a volumetric flow rate to compare to, but I can guarantee the flow is moving MUCH faster than that (we had to wear ear plugs).

PS. in the picture, flow is fully developed with no bends or changes in the pipe upstream or downstream for at least 20 pipe diameters.

 

[boneh3ad]

So I do have a couple more comments. First, assuming that your probe protrudes 1.25" into the pipe (1" to get the hole into the center and another 0.25" since the hole can't be at the end of the tube), then the probe is still going to be blocking 10% of the area of the pipe, so it will definitely still cause the flow to accelerate just by being in there.

Second, let's assume for a second that your probe is just hanging out in a wide open area with no pipe to worry about. If the air is moving toward it at some velocity $U$, then the velocity parallel to the surface of your tube at the surface is going to theoretically be
v_{\theta} = -2U\sin \theta,
where $\theta$ is the angle around your probe (starting at zero on the right horizontal and measured counterclockwise). This is using potential flow theory, so it isn't perfect, but it's a decent estimate for the time being. In other words, over the top of your probe, the magnitude of the flow velocity is going to be roughly twice that of your incoming free stream. The flow should still stagnate at the front end, but the problem is that your hole that you are using as a pressure tap is not infinitely small, so it is measuring more than just the tiny bit of flow at the stagnation point.

If your tube OD is 0.25", let's assume the hole you drilled is 0.125". That means that your hole is going to take up 60° of the circle. Using that, we can average the velocity you are actually measuring. First, look at the "stagnation" condition. The relevant problem here is integrating the velocity $\pm \pi/6$ about the measurement point to find the average (this is just in one dimension, I haven't even considered the second dimension of the hole). So, getting the average velocity of the flow over your hole is just solving
v_{avg} = \dfrac{2U}{\frac{7\pi}{6} - \frac{5\pi}{6}}\int\limits_{5\pi/6}^{7\pi/6}|sin\theta|\;d\theta \approx 0.512 U,
where I've taken the absolute value since we are looking to average the magnitude here. You are therefore measuring nearly half of your free stream velocity there, not stagnation. Doing the same thing for your "static port" configuration is solving
v_{avg} = \dfrac{2U}{\frac{2\pi}{3} - \frac{\pi}{3}}\int\limits_{2\pi/3}^{\pi/3}|sin\theta|\;d\theta \approx 1.910 U,
so you are measuring nearly twice your free stream velocity there. In other words, you shouldn't expect these to measure what you are hoping they will measure. I don't particularly feel like working out the double integral to get a more accurate estimate based on two dimensions, but it will tend to make your static port worse and your stagnation port a little bit better. Either way, it's not a good design.

 

[RandomGuy88]

So Boneh3ad put some math to show you what I mentioned above you are not properly measuring static pressure because of flow over the cylinder. He showed that this is not a minor error that can be ignored and it is probably much worse than the simple potential flow theory would indicate because of the additional acceleration due to the size of your probe relative to your pipe.

Most static pressure probes have the axis of the probe aligned with the flow in order to reduce the effects of the flow acceleration over the probe. In addition this also reduces the amount of blockage that the probe generates. You really should change the design of your probe.

I also want to point out that your calculation of velocity from the pressure difference is probably incorrect. A dynamic pressure of nearly 0.02 psi is closer to 50 ft/s. My guess is that you are using density in slugs/ft^4 but pressure in lbs/in^3. You need to convert psi to psf or convert your density to slugs/in^3.

 

[mgpaul]

Thank you very much for the input and this all makes sense. I figured this was the problem from the beginning but due to the scenario I am trying to use this probe in, using your typical pitot static probe is not very feasible, which is why I tried this design in the first place.

So in conclusion, I am probably going to just make your typical pitot probe with the tube being axial to the flow rather than perpendicular. I do have a question for this as well though before I go ahead and try it. If I put a pitot probe axial to the flow, does the hole size that is measuring stagnation pressure matter? In other words, can the hole be large (say 1/8") or does it need to be as small as possible?

 

[RandomGuy88]

It really needs to be as small a possible. Boneh3ad made a good point about how you end up integrating over a large area on the surface but the other thing you have to consider is that the large the hole the more influence it has on the flow. The flow separates at the edge of the hole and creates an unsteady flow above the tap and this will effect the static pressure you measure.

Are you trying to measure dynamic pressure or just total pressure? You really should not need a static pressure probe because your wall static pressure measurements should be nearly correct for the static pressure near the center of the pipe.

 

[boneh3ad]

Also keep in mind that you don't necessarily need a Pitot static probe. You already have a static port in your pipe, so you don't need to double up that measurement. Just measure the total pressure with a Pitot tube and the static pressure with your static port and subtract to get your dynamic pressure and velocity.

Also keep in mind that the longer your tube is heading from your probe to the transducer, the longer residence time needed to let the pressure seen by the transducer reach that of the pressure port. It isn't instantaneous and could take a substantial number of second depending on the geometry.

 

[mgpaul]

So the problem with using the static pressure at the wall is that in the scenario I am going to be testing, there is a re circulation point at the wall which will give me an inaccurate reading, hence the need for this probe to measure both static and total pressure.

 

[RandomGuy88]

Why is there a re circulation region on a straight wall?

 

[mgpaul]

The test setup I have now (or in the drawing a few replies back that I posted) was only in place to ensure the integrity of my pitot probe. Basically, I was just trying to see if my pitot probe would work or not, and now I know it does not so I will be changing my design.

In the application I need to use the probe, the cross sectional area of the pipe is changing. It is essentially acting like a converging diverging nozzle and making the flow go supersonic, and I have attached a drawing of what is happening on a basic level.

I need to prove that the flow is going supersonic after the choke point, and after proving this, stop the flow from going supersonic. I have a design ready to implement to stop the supersonic flow, however proving the flow is supersonic is another matter. Normally this could be solved using a combination of wall static pressures and stagnation pressure from a pitot tube, but in this case, actually getting the probe into the setup with the tools I have at hand and without drastically altering the product I am trying to test is causing a huge problem.

Long story short, I need to prove flow is supersonic and getting a pitot probe inside the flow at any point in the setup is very difficult in the best case.

Also, in the drawing I provided, I cannot measure the critical area at the choke point either.