EN1591-1: Stresses for pressure test not plausible

[r2user]

Hi ROHR2 Support

I did a flange calculation PN320 with a pressure test at 500bar.
The utilization acc. to EN1591-1 that i got was 9%. This looks very optimistic to me.
Now I needed to increase the pressure test pressure to 600bar and the utilization factor in the load case pressure test decreased to 0.9%.
This is not plausible.

Can you help me to understand this?

Best regards

Rohr2 User

[rohr2support]

Dear ROHR2 User

Thank you for the request and the model.

The EN1591-1 calculates the utilization factors for the flanges and bolts by determining the equilibrium of forces and moments at the flange in all load conditions.
This is done by multiplying the forces from pressure, piping and gasket acting on the flange by the corresponding lever arms to determine the carding moment on the flange. The stresses are then calculated based on these carding moments.
In your case (high pressure flange with high thickness) the lever arm according to EN1591-1 2014 for the internal Pressure is almost equal to ZERO, it is actually even slightly negative (the internal pressure moment is a relief to the carding moment from the gasket).
Therefore the calculated stresses acc. to EN1591 decrease with the pressure!
Unfortunately THIS IS NOT CORRECT.

If we have a closer look at the lever arms over the thickness of the flange you can see:

Hebelarmfehler_dicke.png

Here hH is the lever arm for the pipe and hP is the correction for internal pressure.
The effective lever arm, that the code applies to the internal pressure forces is (hH-hP).
In this case hP increases so that at 32mm of flange thickness, the effective lever arm (hH-hP) becomes ZERO.
If we increase the flange thickness even further to 34mm, the lever arm becomes negative so that the other carding moments on the flange are fully compensated and the stress utilization becomes ZERO.

This is an error in EN1591-1. It concerns all versions, up to and including 2014 (the latest published version).
We have informed the normative commitee of this problem. In the meantime a draft proposal for a correction has been circulated.
The lever arm is corrected using an additional correction hQ.

Hebelarmkorrektur.png

This is not a mathematical clean solution, but at least it avoids the abberations of negative lever arms.
This correction is included in ROHR2-Flange as an option and also PROBAD includes this correction by default starting with version 5.05.

If you want to check the background information, here is a technical note that explains the problem

SIGMA-TN-P012-EN_r3.pdf

This corrections should be applied to all new flange calculations, even as the revised code has not been officialy accepted or published.

best regards

ROHR2 Support