tell tale holes

[1169]

Tell tell holes in vessel flange reinforcements are usually 1/4" threaded. What is maximum pneumatic pressure that can be applied?

 

[jte]

1169-

I presume you are referring to reinforcing pads on shells at nozzles. A couple of psi should be enough for a bubble test. The pad is not intended to be a pressure retaining part; Don't overdo the test. The pad exists to provide local strength and rigidity.

See also thread794-25516 and thread292-97314

jt

 

[JStephen]

You can get an idea of the allowable pressure using the flat plate formulas in "Formulas for Stress and Strain".

For small nozzles and heavy wall thicknesses, allowable pressure will usually be higher than you'd want to test at anyway. But for an atmospheric tank, you could have a 36" manway in a 3/16" shell, and pressure there is limited by strength of the plates.

Usually no more than 2-5 PSI should suffice. No reason to use any more pressure there than the vacuum you use in a vacuum box.

 

[jte]

JStephen-

In my experience, you'll frequently be limited not by the plate thickness (unless you have narrow repads) but by the fillet welds at the outer edge. These welds by definition cannot be as strong as the plate, and with the pressure induced deformation of the plate will see not only shear but also bending.

jt

 

[VeryPicky]

I don't remember where is that requirement specified for vessels but for tanks API 650 paragraph 5.2.4.2 gives you the pressure of 103kPa (15 psi) for air soap test of repads (I assume this is what you are asking for).
For pressure vessels the same pressure is usually applied to the repad.

And: don't plug the holes after the test.

Putting Human Factor Back in Engineering

 

[JStephen]

Note that API-650 specifically says "up to 15 PSI". However, as noted above, this may overpressure large thin repads, even though they meet API-650.

 

[VeryPicky]

We applied 100kPa to all the repads on our pressure vessels and there was never a problem with bulging of repads for over 10 years.
But it never hurts to be vigilant when testing flimsy equipment.

Putting Human Factor Back in Engineering

 

[JStephen]

Read the recent thread on failures in double bottoms. The problem isn't flimsiness, it's simply a matter of using the appropriate pressure for the structure.

 

[JStephen]

I was just running a spreadsheet on the different sizes, maybe this will help. Treating repad (and shell) as fixed-fixed flat annular plate, stress of 20,000 PSI, using API-650 round repad sizes, testing for adequacy at 15 PSI:

3/16" shell- good up through 14" nozzle
1/4 shell- good up through 20" nozzle
5/16" shell- good up through 24" nozzle
3/8" shell- good up through 30" nozzle
1/2" shell- good up through 42" nozzle
5/8" shell- good up through 48" nozzle

If the tank diameter is fairly small, this will be overly conservative, but I know of no better analysis. For larger diameters (ie, field erected tanks), these numbers should be reasonable estimates. Normally, you don't have huge nozzles on thin walls, so the most likely problem area is shell manways.

 

[VeryPicky]

Thank you for the comment JStephen. The question was about pressure vessels and that is why I referred to "flimsy" structures. Pressure vessels rarely create the same concerns regarding wall thickness and subsequently repad thickness failure during 100kPa soap test.
The only reason the API 650 was brought up as a reference was because I could not find the requirement in pressure vessel code.

What are we talking about? Atmosheric pressure not more.


Putting Human Factor Back in Engineering