I was answering a CAESAR II question.
I've seen assemblies of several Victaulic coupling one after the next. So arranged, these assemblies provide lateral flexibility. The pressure thrust will also pull these couplings away from each other and limit the flexibility they provide in the axial direction.
If you use an untied expansion joint or a series of, perhaps squirmy, Victaulic couplings you must provide some sort of pipe support to control axial pressure thrust. Usually, one side is anchored and the other side is highly guided and running direct to a nozzle.

You may not be required to investigate pressure thrust but you better accommodate it - for expansion joints either with hardware around the joint (tie rods or hinges) or added pipe restraint.

Going back to your original attachment - you show assemblies of Victaulic couplings that add system flexibility by twisting at the couplings. These are flexible Victaulic couplings. This is not the axial flexibility I was addressing above. These assemblies act like linkages rather than beams in bending. The stiffness method employed by pipe stress programs does not provide this sort of response. The stiffness method establishes straight line response, not arcs of response. You design the assembly with sufficient clearance so that twist can occur freely without binding the overall assembly. It's kind of like flexible metallic hose - you install a hose long enough to accept the required displacement rather than evaluate strength of the hose if it is too short. I believe you will find that the Victaulic people would use CAESAR II to evaluate load on their rigid couplings but simply assure clearance on their flexible couplings.