Pitot tube measurement at nozzle 2

[mgri]

Hi all and thanks in advance for the help. I have the following query regarding taking measurements from a discharge pipe using a pitot tube.

There is an pipe outlet at the bottom of a pool of water which is 12" deep. I am trying to establish the volume flow rate out of this outlet using a pitot tube. I understand that the pitot tube will measure total head (i.e. static and dynamic pressure = stagnation pressure). The plan is to take this stagnation pressure measurement, subtract the static head due to the depth of the pool, calculate a streamline velocity based on the resulting dynamic pressure and with knowldege of the diameter of the pipe calculate a flow rate(m^3/s).

Two questions:

- Where should I position the tip of the pitot tube? Flush with the outlet, into the outlet pipe a few mm? I wouldn't expect much of a difference between either measurement but I'd like to hear other people's thoughts on this.

- Are there any other contributions to "static" head I should consider other than the depth of the pool when subtracting from the stagnation pressure reading? Would there be any static pressure from the pump a few mm or cm into the pipe outlet that wouldn't be present if I took the reading flush with the outlet? Again, I wouldn't have thought so but there is some debate about this.

- Assuming the flow in the pipe is fully turbulent, how good an approximation would taking the max dynamic pressure of the fastest streamline provide for flow rate?

Finally, if anyone has any easy alternatives, advice or recommendations I've love to hear from you.

Regards,

mgri

 

[BigInch]

Get an ultrasonic strap-on flowmeter.

The pitot should be well upstream into the stream, not anywhere near the discharge where you will get a faster velocity, if the pipe is perhaps not totally full or there is some other streamline anomaly, vena contracta, etc.

To know the pressure near the outlet, you must subtract pipe flow, fitting and valve lost heads from the surface elevation. To get an indication of velocity, you must have an accurate known value for static head, or you won't be able to separate what part of the head is due to velocity and what part is due to static head, or you must take a separate pressure gage reading at the same point, which is why a pitot tube is normally of the kind that takes its static pressure off of a 90[°] tap on the pipe and the inlet protrudes into the center streamline.

If turbulent, the profile is very much blunt, but measure at the centerline of the discharge x-section anyway.

 

[mgri]

Thanks BigInch, some great advice there. I do have a ultrasonic flowmeter available to me so I can use that to verify whatever flow reading I get. For various reasons I must also perform testing at the discharge outlet. I have a few follow up questions for you if you wouldn't mind clarifying a thing or two...

My pitot tube reading at the outlet will be of stagnation pressure (PSI) at the centre of the outflow. From this I will need to subtract some value for static pressure (PSI) to get a measure of the dynamic pressure. You rightly said that I need to get a measure of this true static pressure at the discharge but unfortunately my pitot tube is not of the pitot-static variety and measures only total stagnation pressure. So, here's a question I'd love if you could answer authoritatively: At the discharge outlet, which is under the 12" of water, would it be a valid approximation to subtract a value of static pressure based only on static pressure at this depth (i.e. P = Rho*G*h)? i.e. can we expect the static pressure at the outlet to be equal to the static pressure at the bottom of the pool?

Regarding your advice: "you must subtract pipe flow, fitting and valve lost heads from the surface elevation". Could you clarify this for me? As I am measuring only at the discharge outlet I am unclear why the various lost heads upstream play a part. By "surface elevation" do you mean static pressure at a given depth? For clarity I have included a simple diagram illustrating the scenario I am measuring.

Regarding getting a faster velocity at the discharge, I believe you mean that this could happen if there is a constriction or possibly some effect associated with a partially full pipe. The pipe is vertical and of constant cross section for approximately 1 foot before the outlet. Given the constant flow rate, Q throughout the pipe system I don't expect there to be a large increase in flow velocity at the outlet Vs a few cm into the pipe.

Any further advice is more than welcome.

mgri

 

[BigInch]

It seems like the 12" head could be used as your static head, but if you're interested in high accuracy, I think the surface must not be moving to make that totally correct.

I mentioned the piping and fitting losses only to cover the case of if there was some significant amount of piping between your pool water surface and the outlet point you're trying to read.

Vena contractas, or at the least, local velocity perturbations, can happen within the outlet pipe, for instance if a short outlet tube went through a pool wall, did not have well rounded entrance conditions, or the pipe protruded into the pool, there's a number of different configuratins with different Cd values you might want to look at. Google tank discharge coefficients, pipe entrance coefficients, etc. It could be subject to an inlet Cd for water entering the tube from the pool and also to an outlet Cd from water exiting the tube. If the pipe was short, there may be some interference generated at your measurement point by one or both of those effects. The longer the tube, the less interference towards the center is more likely. I'd recommend you stay as far away as possible from either the inlet or outlet of the tube you're probing.

BTW, I think the turbulent flow profile in pipe is generally something like 80-85% of maximum centerline velocity occuring at the turbulent/stagnent wall boundary, if I remember correctly. It shouldn't be too hard to check.

 

[ccfowler]

mgri,

For your Pitot tube readings to be of any value, you need to consider how well-developed the flow is at the location of your measurements. Will there be at least 10 pipe diameters of straight pipe (more would be better) upstream of your test plane? If the upstream piping configuration is sufficiently messy, 20 or 30 pipe diameters may not be adequate for good measuring conditions.

Also, you would be wise to use a Pitot-Prandtl tube (both static and stagnation sensing in one probe). For either type of probe, precise alignment will be very important. In the case of a Pitot-Prandtl tube, by keeping the probe precisely aligned parallel to the pipe centerline, this type of probe has an inherent characteristic of reporting the flow velocity component parallel to the probe centerline very accurately for actual flow misalignments up to about 30 degrees. A simple Pitot tube will simply provide misleading data for misaligned flows unless you go to considerable bother to adjust the probe alignment to determine exactly how the flow is misaligned with respect to the pipe centerline.

If there is not a sufficient amount of straight pipe upstream, it will be necessary for you to take several readings in a carefully selected pattern to determine the actual flow velocity profile to determine the actual total flow rate. I assure you that this can be very tedious if substantial accuracy is required. If the pipe is large enough and the flow profile is sufficiently disturbed, you may even encounter regions where the actual flow direction is reversed due to eddy currents.

Having done much work with these types of probes, I recommend that the use of just a simple Pitot tube to determine the flow rate should almost always be a last choice rather than a first choice. The opportunity for errors (inherent and accidental) are fairly well maximized with the configuration that you seem to be prefering in your posting.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.