This is something that I believe would be of assistant to others out there in the pipe stressing world.

When incorporating slug loads into pipe stress analysis we traditionally incorporate a lot of conservatism in the slug loading to account for uncertainties in the fluid flow. Now with the assistance of CFD we can remove a lot of this conservatism.

Refer to the following link; http://www.engineeringcfd.com/examples_slug2.html

Note the force vs. time graph down the page.

In this case the CFD analysis showed that the liquid would not remain structured as a slug for these particular operating parameters. It could be determined that for the process conditions considered, the slug you can see in this example would never occur, hence we would not need to incorporate slug loads into the stress analysis.

Not only do you get the benefit of the reduced slug loads in the analysis, you can also use the force graph to reduce the dynamic amplification factor using ASME B31.1 Fig II-3-2.

I hope this is of some use to someone out there

Neat stuff, but the 2 part question I would need to answer for a specific design is what are you gaining by straying from "conservatism" and what are the consequences of the simulation being wrong?

caveat emptor

Robin

The use of CFD gives you a more realistic and potentially simpler design at a lot less cost.

In some cases there is no reason to stray from the traditional methods, however I have come across cases that are very costly to design and construct. One particular case is a long liquid slug empting into a separator vessel on an offshore oil platform. The process engineers calculated some very high velocities as the slug got shorter and shorter as it flowed into to the separator. As we could not design the piping to be strong enough for the slug loads we were forced to install a complex system of control valves to reduce the slug velocities.

As for the analysis being wrong. CFD is a proven technology and providing you use a competent company such as the above link to do the analysis and you cover all the process conditions then this is not a risk. This risk is no different to the use of Caesar, I have seen far too many Caesar analysis that are totally incorrect due to the incompetent users.

Nigel,

All points made are excellent, competence being the most pertinent.