PDA

View Full Version : Alternative leak testing technique


Peter_1
10-07-2011, 02:29 PM
Boyles law states that P * V = constant.
When performing a pressure test or a leaktest, we mostly use a nitrogen bottle on 200 bar (2,900 psi) with a pressure regulator on it.
Say you pressurize your system to 25 bar (360 psi) like me mostly do.
You stabilize the flow until you notice a complete standstill in flow.
You then can have 25 bar in you system, your manifold and in the small expansion chamber of the pressure regulator/reducer.
But you still have 200 bar at the inlet tube and a small part of the pressure reducer.
You then shut off the valve of the nitrogen bottle.
You then have a system with a big V and a small P (cooling system, manifold, and ½ of the pressure regulator) compared to the high P and small V at the inlet of the pressure regulator.

You wait some time and the smallest pressure drop that will happen in your system will result in a very fast pressure drop at the 200 bar side because you regulator tries to maintain 25 bar but there’s in that part only a very small volume available to let stabilize the lost pressure again.
Does this make sense? Sorry if I explained this already in the past.
install monkey
10-07-2011, 02:57 PM
your saying that the regulator is allowing flow into the system if there is a drop in pressure within the system. it wouldnt be a problem if after allowing for stabilisation of the system then the regulator is disconnected from the system and a valve cap is fitted to the manifold.
al
10-07-2011, 03:09 PM
Sounds like a good idea Peter, thank you, you rely on the regulator gauge to show the drop?

alec
Peter_1
10-07-2011, 03:35 PM
Sounds like a good idea Peter, thank you, you rely on the regulator gauge to show the drop?

alec

Indeed but on the HP side of the regulator. A pressure drop unnoticeable on you manifold - less than the thickness of the needle - will let drop the HP gauge.
But i not solely rely on this of course.
Peter_1
10-07-2011, 03:36 PM
your saying that the regulator is allowing flow into the system if there is a drop in pressure within the system. it wouldnt be a problem if after allowing for stabilisation of the system then the regulator is disconnected from the system and a valve cap is fitted to the manifold.

It's just the fact that the regulator must remain in the system to allow the fast pressure drop. It's the pressure regulator which will tell you that the systems leaks, not the manifold
Yuri B.
14-07-2011, 03:28 PM
Hello
My common sense tells me the point is the precision of the regulator. If its springs/ membranes' hysteresis is low, so reactance is quick , then "high" side would open when the P at the "low" side is, let's say 24.99 B, followed by pressures equalization. If the regulator is of a worse class, then the equalization may not take place till the "low" side is, for instance, 24.5 B (with "high" side showing 200B until the moment).
Should depend on the regulator.
Peter_1
14-07-2011, 05:50 PM
Tried already with 3 different brands with same results. But you have a point.
We're just back from a service call on a big automated refrigerant filling system for harvester machines. There they pressurize the system for 30 seconds to 5 bar, then vacuum for 9 minutes and do a vacuumtest for 16 seconds.
All not that long and not very high pressures IMHO to find out if system is leaktight.
desA
15-07-2011, 08:21 AM
A very cunning use of Boyle's Law. Excellent Peter, thank you. :)