1)Bellows Cyclic Life
There are two different fatigue/cycle curve, that is, ASME B31.3 Appendix X and EJMA. 500 cycles per B31.3 will be equivalent to 3000 cycles in the EJMA. Also there are some different quality assurance requirements.
An overly conservative estimation of cycles can result in a greater number of convolutions and expansion joint which is more prone to instability of expansion joint can lead to catastrophic failure. That’s why 200 cycles per B31.3(approximately 1000 cycles per EJMA) are commonly asked for the design of general service expansion joint but it depends on client and project.
4” long-term settlement is quite big for the lateral movement of 24” metal expansion bellows. You need to check the instability of bellows(squirm) in the EJMA calculation sheet from vendor.
I would like to recommend to have 2” off-set installation and connect flange to the tank nozzle after hydro-test of tank. During hydro-test of tank, tank will be settled(ex.:50% of total settlement) and the amount of settlement will be subject to soil conditions. Then, you can consider the only 2” deflection for bellows lateral deflection.
Flexible joints(metal expansion bellows, rubber bellows, Dresser coupling, Victaulic coupling, etc) are commonly used for the piping system with large differential settlement. But they have been causing various failure problems also. Therefore; the use of these assemblies should be avoided or minimized. If there is no alternative choice, the proper selection of bellows material and type is very important.
2) It depends on system(type of service, fluid medium, temperature, and connected equipment). For 2” line, do you really need flexible type joint ? Please do not forget to put some restraint to control pressure thrust when you use flexible hose.
_________________________
Sun Wee