When considering unbalanced pressure thrust on a pump, for eg with an unrestrained bellow in the line, should one consider the resulting force as a force on the nozzle, or as a force resisted at the pump holding-down bolts, via the internals and the casing? In the latter case there would logically be some casing distortion and misalignment, as in the case of a direct nozzle load, but what criteria should one apply in evaluating the effect? I can't find a satisfactory answer in previous discussions.

To : Willem De Klerk
I think the simple answer to your question is that the unbalanced pressure thrust actually occurs on the pump casing at the side opposite the open area of the pump nozzle. A vertical discharge pump would have a vertical down thrust load to the casing supports. This is for a bellows expansion joint of same diameter as the pump nozzle. Since most bellows joints have a slightly larger effective diameter, there would be a smaller thrust load on pump nozzle. There would be an additional thrust load on the pump nozzle, when the piping has a reducer fitting between the bellows expansion joint and the pump nozzle. The thrust load at a reducer is the product of pressure multiplied by the annulus area of the reducer.
You may need to confirm with the CAESARII software consultants, that the pressure thrust that is calculated by CAESARII is applied at the ends of the bellows expansion joint element. It is not applied at the elbows or other fittings at the end of the line where the thrust load would actually occur in reality. This discrepancy is less of a problem with tied bellows expansion joints. I always try to use a tied universal expansion joint assembly, and work the lateral movement of the assembly for needed flexibility. This avoids the piping restraints for pressure thrust.
Regards, Richard Yee

Yes, the pressure thrust is applied at the ends of the expansion joint element, in the axial direction of the joint, so as to put the joint in tension.