Why does pressure gauge fluctuate during static test? 1

[USAeng]

This is on about 1/8 to 3/4" welded pipe. We installed this pressure gauge on a test piece for static pressure testing piping using hydraulic oil. We pressurize the piping to about 200 psi and hold for 10 minutes. First we pump the pipe full of oil. Next we take off a quick disconnect on the other side of the pipe which was letting the oil fill the pipe. Then we continue to pump until we get to the required pressure. Next we turn off the pump and disconnect the line going to the pipe from the pump via another quick disconnect. Therefore the pipe is isolated and there is a gage inline with some fittings on the end which lets up monitor the pressure. This gauge will actually rise during the test by 5-10psi which is what I cannot figure out. Then it might drop and rise from time to time as well. The temperature does not rise that much to make this happen. As a rule of thumb I was taught that for liquid filled gauge the pressure moves with temperature about 1% of the range of the gauge per about 40 degree F difference. I do not fully understand the operation of the internals of one of these gauges - here is the exact one

http://www.mcmaster.com/#4088k34/=c1k8as

but hopefully someone can give me an idea how the gauge could actually rise with time without any extra input. Thank you.

 

[MikeHalloran]

First, a caution: Many people fail to read the little slip of paper that comes in the box with liquid-filled gages, so they don't snip off or pierce the little rubber teat that vents the top of the gage housing to atmosphere. Or maybe they consciously decide not to do it, commonly for handheld test gages, because then a little glycerin or oil will run out when the gage is tipped.
That causes the gage reading to be both erratic and erroneous.

Once you've fixed that, you have two more uncertainties.

One, unless you use special gages, or evacuate before filling, you can't get all the air out of the system.

Two, the statement 'the temperature does not rise that much.." suggests that the temperature of some part of the system is changing, and that you are 'eyeballing' how much the pressure should change in response. If the temperature of any part of the system is changing _at_all_, you should expect the pressure to also change. ... and in the time it takes to properly calculate how much it should change, the system temperature will probably stabilize, so the time is better spent getting coffee.



Mike Halloran
Pembroke Pines, FL, USA

 

[zdas04]

USAEng,
Whoever told you that "gauge pressure changes 1% per 40F" is so full of crap that I don't know where to start. First, this is saying that the temperature change is a function of the calibrated range of the gauge instead of an intrinsic property of the fluid. That would say that a 0-100 psig gauge would change 1 psi for 40F change, but a 0-10,000 psig gauge would change 100 psig. What nonsense.

Second, the coefficient of thermal expansion is an intrinsic function of the fluid. For a water filled system, the number is about 100 psi/[°]F. Really. Different oils have different coefficients, but none of them is very far from the water number. So if pressure changes 10 psi, then temperature changed 0.1[°]F. Credible? Certainly. You will never have a total thermal equilibrium. The steel will conduct heat at a different rate than the air or the oil. If you are looking for an absolutely constant test, either use a gas (several orders of magnitude smaller coefficient of thermal expansion) or build an environment that is absolutely isothermal and allow enough time for the fluid and the steel to reach ambient temp (several hours after fill).

You might have the gauge problem the Mike mentioned, but you certainly have unreasonable expectations about the effect of temperature change on pressure in a closed, liquid-filled system.

David