Please Help! Trying to decipher transient flow results

josh89

Hey guys,

I'm currently studying a Bachelor of Aerospace Engineering and doing my honours level research project in a field called Active Flow Control.

A brief intro to what the project involves:
The motivation of the project is to reduce the drag on trucks. Specifically I'll be putting a model into a wind tunnel and through two slots at the back there will be oscillating jets of air created using an internally mounted subwoofer.
The pulsed air will add momentum to the wake region, delaying the onset of shed vortices resulting from flow separation over the back of the model and consequently raising the base pressure (pressure on the back surface). Pressure drag is the dominant mode of drag for cruising trucks.

Where I'm at at the moment is I've built a prototype and I'm trying to find a way of accurately measuring the velocity of the air that's entering/leaving the slots. I've produced two 'rakes' of pitot static like tubes that have been attached to each slot.

Here are some photos so that you can get a better idea of the setup:
Prototype model with the back detached: http://tinypic.com/r/35l4ww3/7
Here you can clearly see the top slot: http://tinypic.com/r/2rhxx94/7
Inside of the pitot static 'rake': http://tinypic.com/r/vmru55/7
The complete set up: http://tinypic.com/r/2hnny8h/7

The rakes are mounted onto each slot and are connected via nylon tubing to a digital pressure measurement system. the metal tubes that face into the flow measure the total pressure and the ones that sit flush on the surface measure the static pressure. By subtracting the static pressure results from the total pressure results I hope to get the dynamic pressure, and hence the velocity.

So here's my problem: I'm getting static pressure that's bigger than the total pressure for every position along the both rakes (7 positions per rake, each position has a channel for total and a channel for static, so 28 channels in total).

As an example I've plotted the results of the middle position of the top slot here: http://tinypic.com/r/2jfetfa/7

As you can see the static pressure is greater than the total pressure and as a result the dynamic pressure (total-static) is inverted. I cannot understand how this is physically possible and so I've come to the conclusion that something's wrong with the experiment, but I can't work out what.

I've ruled out a few things so far. The hoses are definitely connected the the correct channels, the formula I've used in excel references the correct cells and all the channels were zeroed before every experiment.

If anyone has any ideas or suggestions I'd be very happy to hear them.

Thanks in advance!

josh89

P.s. If anyone knows of any other appropriate forums where I could post this thread that would be greatly appreciated as well.

RandomGuy88

Maybe the flow is not actually parallel to the surface the probes are mounted on. If the flow was at an angle relative to that surface than the flush mounted probe would read a higher pressure than static and the total pressure probe would read a lower pressure than the true total pressure.

Is it possible that when you bent the total pressure probes you kinked the lines?

What are all of your transducers referenced to?

I also have another comment. This setup won't be able to measure the velocity in both directions because the total pressure probes can only measure the total pressure from one direction. When the flow is in the opposite direction they total pressure probes will actually be measuring a static pressure and the static pressure taps will be measuring an incorrect static pressure because they are in the wake of the total pressure probes.

boneh3ad

The other thing to keep in mind is that the frequency response of a pressure transducer with a long tube connecting it to the flow is very low. Depending on your specific geometry, you may simply be cutting off the peaks because the long tubes you are using are damping out your response.

RandomGuy88

Great point boneh3ad.

When making unsteady pressure measurements you ideally want the transducer flush mounted to the surface. When this isn't an option then you want to minimize the length of the tubing and keep it as straight as possible. Its a very interesting problem though. Depending on the tube geometry and the frequency of the fluctuations, the magnitude of the measured pressure can be increased or decreased and the phase is shifted from the actual pressure. If the tubing is not to long it is possible to determine the transfer function of the tubing and then use the distorted pressure measurements and this transfer function to reconstruct the actual pressure at the probe. Determining the transfer function can be done experimentally or theoretically in some cases.