For a dynamic analysis the way a rest support may be treated/modelled can be different from a static analysis. This is what I want:

If the weight of the (horizontal) pipe is significantly higher than the axial excitation force then it may be assumed that pipe movement at the rest support is restrained in all three (XYZ directions), just due to the friction force. Of course many other effects have to be considered, but I want to keep the example simple.

I noticed that CAESAR allows me to model a support with all kinds of restraint types, even allow me to use/specify a MU factor which is intended for making use of that friction factor. So I made a little test model and tried if I could set up such a situation.

As far as I understand CAESAR (remember I am a new user) I think that the X, Y and Z restraint behaviour are treated separately in the solver. So I don't manage to couple the X and Z restraint behaviour to the Y restraint status. In simple words: I want the X and Z restraints turned off as soon as the pipe is lifted off the support. In my test model I did not succeed in creating this situation.

Just to be sure: it is sufficient to model the support as a lineair element. The behaviour described above should be checked during the static analysis and then set 'fixed'. So if there was sufficient friction the restraint mode should become XYZ, if there was unsufficient friction the restraint mode should become '+Y' if not yet lifted or '' when lifted.

The current solution (workaround) I've chosen is to make 2 models, one for static analysis and one for dynamic analysis. Maybe stress engineers think this is wrong or strange, but there is a significant difference in how to model pipe supports dynamically or statically. Dynamically spoken many supports can be considered as XYZ (fixed) points, whilst they are just rest supports (+Y) in the static analysis.

Has anyone experience in this specific problem?