How to measure/determine resulting relief valve chatter forces

[Water_Guy]

We are in the process of performing fatigue studies on branch connection (like a weld-o-let) that contain relief valves under a "Chatter" scenario in order to predict future failure. The question is: How do you measure, and record, the resulting forces at the branch during a chattering event so that the proper fatigue analysis can be performed?

We thought of placing accelerometers around the branch connection, but didn't know how to convert those readings into resulting forces. We are expecting the forces to be the result of the pressure wave back from the valve due to the sudden closure and the forces from the relief valve Disk closing against the nozzle due to the spring.

Or if the branch was completely restrained, would load cells work?

Many Thanks,

 

[GD2]

Water Guy,
What's your relief piping system - open vs closed? If it is open system, you will most likely need to find the reaction force/moment by the open vent. Usually, the relief piping will be supported on the vessel or the piping upstream of the discharge elbow to reduce the reaction moment effect on the WOL. I am not an expert in this area but my guess will be that you will need to find the time history of the relief valve chattering to determine the number of cycles the WOL will see in its design life. By installing an accelerometer, you might be able to determine the number of chatters (cycles) the valve will see in a relieving event.
If the relief is a closed system, a dynamic analysis is required to find the force/moment which is very complex.

Now, when you are talking about the pressure wave phenomenon, are you not talking about the 'water hammer' situation? Fatigue curves are generated for materials that goes through cyclic stresses. Is a 'water hammer' effect comparable in determining the fatigue life?
It's an interesting subject and I would like to see more discussion on the subject.

Ganga D. Deka, P. Eng
Canada