Measurement - Pressure Drop vs. Leak Rate

GURU SDI

Involved In Discussions
Hi all,

I want to measure a leak rate (air) for my assembly. In my technical drawing, it is specified as "X ccm" at testing pressure (P) for a testing time "t" secs.

But My supplier having an instrument which measures leak rate as pressure drop instead of actual leak rate. now i want to give allowable pressure drop for the leak testing process.

thank you

Bev D

Heretical Statistician
Staff member
Super Moderator
Alternativley you can develop the formula for yourself by performing a straight forward correlation study.

first select some parts (30 if possible but no fewer than 10 if they are rare)
have yourself and the supplier measure each unit twice.

determine the repeatability of your system and the supplier's system.
if both systems are sufficiently repeatable perform a regression on the first measurement made by yourself and the supplier - this wil give you the conversion formula based on the exact specifics (uniqueness) of each system.

GURU SDI

Involved In Discussions
Re: Pressure Drop Vs Leak Rate

Thanks Rob...

But this equation is arrived based on the perfect gas equation, which assumes the process takes place under constant temperature (practically temp is not constant during my test) and moreover i am having difficulty in measuring the exact volume of my test piece.

so i planned to take the readings in both pressure drop and leak rate for 50 samples.. & there should be linear relationship between these two variables...so I can arrive best line of fit between leak rate and pressure drop...

From the straight line equation, i my get pressure drop...

Any other views to measure...thanks in advance

GURU SDI

Involved In Discussions
Thanks Bev

Please see my post, will my study give good result..

gpainter

Quite Involved in Discussions
Compressed air should be looked at as a system. Use an ultrsonic leak detector to find leaks. Pressure drops can be a result of many different things, distance,couplings, etc. Develope a leak detection system. A single 1/16 leak can run about \$732 an year based on \$0.07kWH and up to 30% of compressed air is consumed by leaks. Compressed air is also not an energy efficient power source(5-10% at most)and costs 6-8 times the cost of electric.

Last edited:

dgriffith

Quite Involved in Discussions
If I have a closed system with 10 psi in it, and I measure the pressure again after 5 minutes and I have 8 psi in it, what is my average leak rate? I think it is 2 psi/5 minutes which is 0.4 psi/min.

When I connect an Air Data Test Set up to an aircraft and perform a leak check on the pitot-static system, I'm going to get pressure per unit time (or altitude or airspeed/time), and I'm not going to care about the volume of the system--Piper Cub or 787. The test set doesn't care.

What is of concern are the variables--most important being adiabatics. Unless the system is temperature stabilized, gas temperature changes will cause pressure changes--for a given volume--that will mimic a leak. Leak checks performed immediately after evacuation or pressurization will run that risk.

So, as a baseline, I would pressurize the system, wait 60 seconds, and measure/record the pressure every minute for 5 minutes. I would plot the values and see how curvy it was. Too curvy at the top means wait longer before starting to record. If it's not too curvy, then establish a 1st order slope (regression line). You could then predict a pressure leak anywhere along the line for a given time.
In other words, for a 2 minutes leak check x, after waiting to stabilize, you should get a pressure value y ?some error.

dgriffith

Quite Involved in Discussions
eguevara,
You may have to rephrase the question and be more specific.
One the surface, you measure the value of a pressure drop with a pressure gauge or pressure transducer. Also, the difference/change between where it was then and where it is now is the pressure drop.