Hi,
(this applies to op. processes and op. activities, too.).
A sequence of exchanges in a functional chain (FC) may not visit the same function twice and thus may not contain any cycles. This is what you can read at some places in some places on the web and a similar, even more restrictive constraint is present in Capella help ( https://forum.mbse-capella.org/t/functional-chains/4190/4) . The context menu item to create a FC/OP in xAB diagram by selecting the exchanges only appears if this is fullfilled. What is the reason for that constraint? It seems to be arbitrary.
But in a OPD and dataflow diagrams it is still possible to add an already present activity or function to the op process/FC effectively counteracting this rule. Its is of course the right thing but why does it work only there and not in a xAB as described above?