Slug Flow Dynamic Results Interpretation

[Crusader911]

I am analyzing a slug flow case. We observe that an elbow gets slammed every 3-5 seconds and displaces less than 1" in one direction, and even less in the other two. The estimated rise in pressure due to the slugging has been given to me by the process department. I multiplied this by the cross-sectional area of the pipe to get an applied force. I first did a modal analysis, and found that one of the natural frequencies Caesar calculated matches the movement of the pipe very well.

Now to calculate the time the force is applied (and thus the impulse on the elbow). I have read quite a few articles discussing the subject that suggest using the time the pressure wave takes to traverse the longest distance between elbows in the system. Since this line is coming out of a furnace, I first tried using the total length of the helical coil. That produced displacements that were much larger than observed. I next tried a 14' span between elbows that was in the same direction as the largest component of the observed deflection. The displacement results were consistent with the observed deflection.

The dynamic stress report shows that this line failing B31.3 code (the stress exceeds the allowable). I have been working feverishly on this for four days, and I may no longer be thinking clearly, but I interpret these results to mean:

1. Failure due to secondary stress (fatigue) is more likely because the system exceeds code stress. The fact that it has not failed in over 3 million cycles (6 months) is because the stress it is experiencing is somewhere in the safety factor of the code allowables.
2. Inherent in the analysis I've done is the assumption of low-cycle fatigue. In other words, there is a fatigue factor that is 1 for anything under 7000 cycles, but should be applied in this case. If I go back and do that, the results will be even worse due to attenuated allowable stress.
3. I should put a snubber on that elbow and see if I can resolve the problem.
4. I want to go home and play Guitar Hero with my kids until my fingers are as numb as my brain.

 

[richay]

Check "Ohliger's" suggestions to this post in the CAESAR II Forum (on the COADE web site).

Your item #3 above may resolve the issue. However, if altering the system stiffness resolves the problem, see if a small, light fixed restraint will do the same thing AND not cause a static problem. From a "maintenance / performance" point of view, you want to avoid a real snubber.

Richard Ay
COADE, Inc.

 

[Crusader911]

Thank you, Richard. The client has decided to revise the piping in order to prevent the formation of the vapor pockets responsible for this problem. I may not have any more to do on this one.