pressure drop measurements in gas flows 1

[themroc]

We intend to measure the pressure drop over a length of 1m vertical pipe in a 44mm ID tube.
Air velocities are up to about 20m/sec.
Pressure is just below atmospheric pressure.

We have got the the devices sorted in order to measure the pressure drop. We are only interested in meassuring the static pressure in the tube. We cannot meassure the pressure using a whole in the tube wall because here a liquid film with 2mm film thickness is flowing downwards.
Therefore we intend to use a small tube (1mm ID)to point
perpendicular into the gas flow, in order to meassure the static pressure. (See scetch)
Does it matter how far I point into the gas flow?
Does the shape of the tip has an impact on the measurement?
Does the internal diameter of the tube has an impact?
I want to avoid a pitot tube effect where I meassure also a velocity component.

Sketch:
| |Tube with liquid film at the wall
| |
| |
| ______
| ______ 1mm tube to measure the static pressure
| |

 

[dcasto]

The system will only have about 15 mm H2O drop. I'd believe that making the insertion tube flush WITH the pressure cell higher than the tap would keep the thin film from interfeering with the pressure measurement.

 

[BigInch]

Might have to calibrate for some extra turbulence, but should work.

http://virtualpipeline.spaces.msn.com

 

[sailoday28]

Just curious- How have you determined the annular film thickness. Shouldn't the film thickness vary as it is falling?
Generally such a measurement would assume a negligible radial pressure gradient. Can the 1 mm measuring tube be inserted so that in does not protrude beyond the 2mm film? Measurement of the film pressure would then indicate the gas static pressure.

Regards

 

[Warpspeed]

Your best bet is to use a pair of Pitot static source. These measure true static air pressure in a fast moving air stream without any difficulty.

http://en.wikipedia.org/wiki/Pitot_tube

These very simple and easy to make devices, work well enough to be able to operate a pressure altimeter in a jet fighter aircraft.

 

[themroc]

Sailoday28, we thought of doing so, but we want to operate at quiet varying flow conditions for the film flow, this includes using quiet viscous liquids (up to 100cp) This causes variing film thickness's, there is than always the danger that the tube is not filled entirely with liquid and this would cause also problems

One other reason is that as a pressure sensor we use an air dp sensor from sontay which operates for pressures up to 100 Pa. It is calibrated for air, so we don't quiet know how it would react on liquid in the system.

 

[themroc]

Warpspeed, we looked into this but usually those devices are bigger and difficult to install in a tube (44mm ID).
When looking up those devices they also seamed to measure the pressure similar to our setup (see stativc pitot device in your link) the pressure openeing is located perpetingular to the flow. Do you think it makes a difference to meassure the static pressure with a single tube as in our proposal?

 

[Warpspeed]

That depends on if you have a true boundary layer inside your pipe. Any roughness or pipe bends up stream will cause flow separation, and you may not have a true boundary layer to tap into along the pipe wall.

Any flow velocity at all directly across the face of an open pipe end will create an unstable negative pressure.

A pitot static source can be made very small, perhaps with 1/8" od copper capillary tube. Block the open end with a smooth aerodynamic round dome of solder, drill some radial pin holes in the pipe at around ten pipe diameters back from the nose, and polish up the outside of the pipe with very fine emery paper to remove drilling burrs.

Very simple to make, easy to try, and it will work extremely well. This is what everyone uses, and it will give a rock steady static pressure reading up to extremely high airspeeds.

 

[themroc]

Thanks for the info this was very helpfull,
Do you know suppliers of such small tubes, or do you recommend to build them on your own?

Ideally I would only need a capillary tube which we bend 90° and than cover the inlet hole. In a next step we would need to drill holes into the tube wall and polish them, Is that right?

Only one Question, do you recommend to drill two holes opposite to each other as shown in the wikipedia article. What is the reason for this?
Do they need to have a certain minimum hole diameter?

 

[Warpspeed]

Very small diameter copper capillary tube can be purchased from any good refrigeration or air conditioning parts supply house to the industry. Just the first Google example that came up, pipe sizes from 1.6mm to 3mm od:

http://www.szhongtai.com/tc4.htm?gclid=CMC6xrePrI8CFRnwYAodKyBdMg

That is just an indication of the sorts of small sizes of pipe that are available. It is usually cut from a roll, and you can buy short lengths easily enough.

Block the end with soft solder and drill several radial pin holes. By using multiple holes the pressures detected in the boundary layer average out between all the holes. The holes can be as small as you like, as there is zero hole flow once the pressure has stabilized inside the pitot tube. Something like a number 80 drill perhaps ?

Just polish up the outside of the probe tube, so the boundary layer is not pierced with a rough projecting edge at the holes. If it feels smooth to the touch, it will work fine.

Bend the pitot tube into a right angle and solder it into a threaded male pipe fitting. Screw the whole thing into the wall of your process pipe, and you are in business.

Best to mark the threaded fitting, so as it is screwed in, you know when the projecting nose is pointing directly up stream into the flow. It can be pointing slightly out by ten to fifteen degrees without serious errors being introduced. One more reason for having multiple pin holes arranged around the probe.

 

[themroc]

Sounds perfect, I will update whether this was working well or not. Just one thought.

Does it actually matter which way the tube is facing
In the following drawings the gas flow is coming from the top. Is there a difference as long the tube is parallel to the flow? I presume assembly 2) is better because the tube bend does not effect the flow?


1)
| |
| |
| |
| --------
| || |
| || |
| || |
| |
| |


2)
| || |
| || |
| || |
| ---------
| |
| |

 

[Warpspeed]

Yes it is VERY important that the probe end points directly into the flow.

The idea is that the smooth blunt aerodynamic round nose of the probe directs the flow evenly around the outside of the probe tube with minimum induced turbulence. The air then forms a stationary boundary layer that sticks to this probe tube. The very small holes communicate this pressure into the inside of the probe tube.

If you do it the other way around, the supporting bend in the tube will disrupt the flow. Turbulence will then disturb or even sweep away the boundary layer. It is essential that the pitot probe points directly into the oncoming flow.

Your drawing 2 would be the correct one, if the direction of flow is moving downwards.

 

[sailoday28]

themroc (Chemical)Again, how did you determine the annular film thickness?
At the film boundary where air and liquid meet how was velocity of film determined-This also relates to film thickness.

Regards

 

[themroc]

In case you know viscosity and mass flow of liquid you can determine the film thickness by using the so called Nusselt equation.
This is used in order to calculate film thicknesses in falling film evaporators and vertical condensers. In general film thickness increasses with mass flow and viscosity

 

[sailoday28]

themroc (Chemical)Thickness should be weakly dependent upon vertical pressure gradient.

Regards