Do you have a physical example of this ? If you stop the flow, and the vibration ceases, then you have your answer!
I would say that, if the vibration is caused by the fluid acting directly on the bellows, it would be a very difficult thing to calculate, especially if it involves turbulence. If you can identify the major mode shape using high speed photography, and its a relatively simple mode, the natural frequency might very roughly correspond to that of a beam with a distributed mass (the fuel fluid), where the axial and/or bending stiffness of the beam is equated to the matching stiffness of the empty bellows - but its a long shot.