Oscillating airfoils: system identification

Iain

Hi guys, I'm working on a simple aeroelastic system and have a 2 degree of freedom system (pitch & plunge) in unsteady flow. The technique is working well but I'm having problems in analysing the oscillations.

The code takes a long time to run and outputs the loads & pitch and plunge displacements. I can tell by eye whether these are growing (flutter) but am having problems getting a system identification algorithm to work.

I'm currently trying Prony's method as the results should, ideally, be able to be modelled as the sum of complex exponentials (i.e. the sum of damped cosine/sine signals). Unfortunately the efficacy of this approach varies depending on which airfoil/flow parameters I use. Sometimes my approximations are very good - sometimes they're not even close. There doesn't seem to be any way of knowing beforehand whether the Prony method is likely to produce good results. I should mention that my oscillation plots are smooth (no noise so Prony should be ok).

Could someone please suggest a simple method of system identification that I use for my model? If you could point me to a reference or two I'd be a happy bunny ) Iain