Dear Hakun,
Your approach sounds correct. In addition to changing the tower mode shapes (which it sounds like you’re working on), the other input that needs to change is the time step (DT) in the primary FAST input file. The introduction of the CS foundation will introduce higher natural frequencies in the FAST model, which will require the use of a smaller time step. My guess is your results are “blowing up” because you didn’t reduce the time step. See my post dated Nov 4, 2010 in the forum topic found here for rules of thumb on how to choose proper time steps: Error in FAST, working in ADAMS.
Best regards,