Dear List,

I have a question that maybe someone can help me with.

I have been asked to calculate a test result where a casting is pressurized and then a flow transducer output is monitored to determine
if there is flow through the casting above a certain level.

It looks basically like this:

15PSI >-----------[Flow =
Transducer]--------------------[Casting]----------------[leakage to
atmosphere......

If there is a leak the air will pass through the casting wall into atmospheric pressured air.

I have the ability to control the amount of orifice opening at the Flow Transducer via an analog driven valve. Range is 0 - 50 mL/minute.

I am looking for a leakage level of 15 mL/minute.

My question is how to calculate what that leak of 15 mL into atmospheric pressure will look like at the Flow Transducer with 15PSI behind it and
basically 15PSI on the other side. I could throttle back the flow rate through the transducer thus creating a pressure drop on the test side but I think the basically equal pressures will give me a different flow rate at the transducer than what I have flowing into static. Am I thinking right on this? I welcome your input on this.

Sid Roberts

 

I think the conventional method (simple version) for a leak test is to:

1) pressurise the test side
2) close the supply valve
3) take a pressure reading
4) wait x seconds
5) take a second pressure reading
6) calculate the differential

You mentioned that "I am looking for a leakage level of 15 mL/minute." I would have thought that at that rate it would be easier to
measure pressure rather than flow. However, if you measure pressure then the flow rate (which is what I assume the customer is specifying on their
product drawing) would have to be derived rather than being a direct measurement.

There are companies which specialise in leak testing (which is what your application sounds like). Depending upon how critical the application is, you may want to consider an off the shelf solution. Leak testing solutions can range from very crude to very sophisticated, depending upon the needs of the application.
You will also want to look very carefully at the mechanical installation. I had the misfortune to have to work with a rather crude
system once, and the biggest problem was not in the transducer or software, it was in the poorly designed pneumatic system (valves and piping) which consumed a good deal of the available system response.

One problem I would be concerned about with flow measurement would be in offset errors caused by flow restrictions in the pneumatic system between the transducer and the test piece. If the flow restriction is very little (i.e. a large diameter tube), then the flow rate may be so low as to be difficult to measure. If the flow restriction is high (i.e. a small diameter tube), then the offset error may consume a good deal of your resolution.
If you try to throttle back the flow rate with a valve, then you are reducing the pressure at the test piece. I would assume that the 15ml/min. leak rate is at a specified pressure differential across the casting. I think that this means you would have to either control the pressure at the test piece accurately, or perform a calculation (or table look up) to compensate for low pressure (which may be due to more leakage acting in concert with restrictions in your supply).
One more thing to keep in mind is that atmospheric pressure is not constant and your system may need to take this into account in some manner. If you instrument references ambient pressure in some way, then you are covered there. If not, then you may need to compensate in some way.



**********************
Michael Griffin
London, Ont. Canada
[email protected]
**********************