Why is the Pressure Drop Uneven in Two Equal Impedance Tubes?

[wombat]

Hi All,

I've come across something that surprised me but I'm sure has a logical explanation...

I have a two small ID tubes that have exactly equal impedance to air flow (for the purposes of the discussion the flow is 2ml/min). I connect them in series with one end of the system exposed to a constant vacuum of 10kPa below atmosphere and the other end open to atmosphere.

[Tube A] [Tube B]
[Atmosphere] ========= + ========= [-10kPa Vacuum]
^
[ Pressure Read Here]

I measure the pressure between the two tubes relative to atmosphere and I don't get -5kPa (ie half of -10kPa), instead I get something less, like -4.8kPa. Why is there an uneven drop in pressure through the two tubes. I assume it's to do with changing density but I can't put it into words.

Thanks in advance,
Wombat

 

[wombat]

It looks like my ascii diagram got butchered as the forum code removes multiple spaces.

Pressure is read in between the two tubes (where the '+' is) .

Wombat

 

[NoTime]

Two guesses.
There is a pinch effect going on, like three lanes of traffic being squeezed into two.
It's an artifact of your measurement setup.

 

[Cliff_J]

The tubes themselves may present the same impedance but the mouth of the first tube should add another impedance to that tube and cause your pressure reading to be low, not high.

What about if a chamber was added between the two tubes to help remove any velocity related measurement error possibilities? Also, now the second tube would have an identical mouth presented to the chamber so its impedance should match the first.

Cliff

 

[wombat]

Thanks for the replies so far. There is effectively a chamber between the two as the T-piece where the transducer nozzle connects is quite a large part relative to the size of the tubes. I'm doubtful of any effects due the mouth are dominating the readings as the flow's are very low and as I say, the mouths are basically the same.

The measurement setup has been done numerous time with different equipment and all with the same result. I've also tried different tube lengths and ID's (25uM -> 560uM). I also have high quality flow-meters, transducers etc.

Wombat

 

[NoTime]

Also, now the second tube would have an identical mouth presented to the chamber so its impedance should match the first.

Wouldn't the change in air pressure alter the impedance?

There is effectively a chamber between the two as the T-piece where the transducer nozzle connects is quite a large part relative to the size of the tubes.

Is there a temperature difference that might have to be accounted for?
If you can measure flow directly, does inserting the chamber alter the flow rate?

 

[Cliff_J]

Wouldn't the change in air pressure alter the impedance?

Yes will lower the effect but again it should still be lower for the second tube. At least how I'm visualizing it...

How about 3 tubes with similar connections to read the pressure at two points along a longer tube?

Cliff

 

[NoTime]

Yes will lower the effect but again it should still be lower for the second tube. At least how I'm visualizing it...

Isn't -4.8 less than the expected -5 kpa?
From your other post and this I get the impression you think he said the pressure is higher than expected?

Edit: nevermind looks like I was misreading it.

 

[ceptimus]

As the gas pressure lowers, the volume of the gas increases. As the same mass of gas is passing through both sections, this means it has to flow faster in the second part of the apparatus. Because of the faster flow, there is more resistance and so you drop only 4.8 across the first (slow flowing) part, and 5.2 across the second (faster flowing) part.

Actually it is more complicated than that, as the gas will also cool down as it expands.

Don't expect linear results in this sort of system.

 

[wombat]

Thank you for your reply Ceptimus.

Just so I'm clear... Pressure drop is equal to volume-flow x resistance. Therefore I imagine that resistance could be in units such as Pa.secs/m^3 ?

I don't think cooling effects will be an issue for our system but I take your point.