I am trying to analyse pipe resting on parrot swing supports that are used underground in South African mines. This support system consists of a number of pipes resting on a common beam which is suspended on chains. These support are typically placed every 3 meters and run along the length of the haulages which are often straight and long with little room for flexibility. The pipes can be at different temperatures, pressures and diameters. The pipes are tightly bolted down to the beams with u-bolts. The beam can lift, arc and twist but the multiple pipes add stiffness to the system and the different temperatures oppose thermal growth.

This system is very difficult to analyse because of the length of the pipeline, the degrees of freedom and the inherent rigidity.

I believe that pipes supported on parrot swings could be over-constrained and subject to high internal stresses.

Mr Diehl recommended to use a +Y Rod cnoded to a +y support when looking at the analysis of 2 pipes resting on a parrot swing.

We have modeled a short section of pipe (54 m long) using the rod and +y support at each restraint point.

If we consider one pipe to be hot and one pipe to be at ambient temperature, I would expect the expansion of the hot pipe to be restricted by the pipe at ambient temperature.

Our model showed unrestrained growth of the hot pipe that extended the pipe at ambient temperature i.e. the model showed no restriction in the thermal growth of the hot pipe. This result is not what I would have expected. We will e mail the model tomorrow for you to have a look at it.

I have been asked if any anchorage is required to oppose dynamic loads from water hammer as any dynamic load is resisted by the significant inertia of all the suspended pipes. A concern is that as the wave travels between bend pairs, an individual support will be subjected to the full water hammer force.

You need to address the axial constraint between the hot & cold lines.

Please could you elaborate a little further - the pipes are bolted to the beam - should I add lateral restraint?
Q) why does the hot pipe shows unrestrained growth - clearly we are doing something wrong here.
Thanks

I believe up until this point you were ignoring the cold line in your model. To get this axial reaction, I think you'll have to build a little more.

Model your supporting steel as a pipe. Support this pipe with your swing models. Add nodes on this pipe where the hot & cold lines rest. If you U-bolts are tight, define X, Y, & Z restraints at these new nodes defining CNodes using the hot & cold line contact points. You might have to add a rotational restraint on one side to keep things stable.

It may sound like a lot of work but if you choose proper node numbers, the duplicate feature will make it easy. Build a couple spans and test them before running the entire system.