The error message signals that floating point parameter is not correct. So there are different possibilites for it. Above I mentioned some reasons, but there could be more, e.g. division through zero. So in my opionion you should check the real inputs of the function block in the error case. One input is in my opinion not correct. There exist also a open function block with which you can check the values before you assign them to the function block.
... View more