Hey guys,

Thanks John for the background info.

This subject is one that I believe is "implied" by many engineers. When I model a restraint, say an anchor at a nozzle or a vertical support in a rack, I know for a fact that it is not infinitely rigid. There will be some "flex" from the vessel wall or the beam. I think we can all agree on this. How many times do you actually model in the beam flexibility? Or that of the vessel wall? I'd venture to say less than 10% of the time.

However, when you give piping loads to a structural or vessel engineer for their analysis and/or approval, we typically assume these restraints to be "infinitely rigid", even though we know this is not the case. Typically, I only go to the trouble to model restraint flexibilities when the restraint loads are borderline using the infinitely rigid model, and then usually as a last resort.

The same goes for trunnions and dummy legs. I often model in the dummy leg diameter (if its pipe) and length (without temperature or pressure) to get an idea of the flexibility of the support. I've also been known to use the structural modeller to model beam trunnions, which can get tedious (watch the orientations). Local stress issues and SIFs still need to be considered.

After having said all that, (looks like one of Mr. Breen's postings) I guess the point is: there are many things we do in our analysis that are assumed or implied, but we need to be aware of the implications.