I guess nobody wants to convince you "to use a snubber to arrest the vibration, i.e. increase the natural frequency of the system".

Snubbers are rather shock absorbers, which in such applications form restraints between piping and the structure. The displacements due to expansion do not encounter resistance by snubbers, in theory.

True, in calculation one effect is modifying the natural frequency. That's because a snubber is in calculation like a switch: for Non- dynamic case=0 (no Restrain), for Dynamic case=1 (Restrain). What I don't understand is which is your proposal to improve the calculation for shock dynamics cases, since "as it is" does not convince you?