I prefer to avoid using guides and stops with 0" gaps for purposes of getting the pipe or nozzle loads to pass for a few reasons:

1. Requiring them means that your model is relying on a stiffness value of 10^12 lbf/in in order to get your pipe or nozzle to pass. Your structure generally cannot meet that promise.

2. Implementing a 0 gap, especially bidirectionally, means that your system can't reliably move back into its original shape after a cycle. Metal heats and cools and will get stuck in place, and not slide, generally as your model promises it will.

3. It's impractical to build and put into place.

What is your opinion on the matter?

Michael,
Additionally when we consider guide or stops with 0 mm gap, one or two rotational movements also gets arrested (see sketch).While doing piping stress analysis there are many assumptions, depending upon criticality of the system we should try to build our theoretical model to simulate actual condition of the system.

For critical system such as compressor, turbine etc. where 0 mm gap is required to achieve nozzle loads, I prefer to use rigid struts. Advantage is that we can achieve near to 0mm gap with almost negligible friction. Rotation of strut needs to be taken into account.

As your pipe heats up, so will the shoe that's attached to it. Would you not consider it effectively fused to the guides at that point?

Best way is to match model as much as possible to reality, while maintaining healthy portion of conservatism and simplicity.
Use your engineering judgement.

Pipe configuration, equipment, pipe size... etc. Many factors will influence output.
There is no rule of thumb that can be applied without thinking.


Mathematically speaking, pipe is infinitely long stick with 0mm diameter and supports are point restraints.
Moving back to reality, pipes and restraints are still very close to above mentioned mathematical model.

While some rotational restraint effect exists it is not significant to
alter calculation results.

When we design zero gap on axial stop, we are actually increasing the loads on the supporting structure. Hence it will increase the cost of the supports. But there are points where we need to restrict motion in one direction. There we anyhow need to provide ZERO gaps.