A practical approach to parameter identification for a lightly damped, weakly nonlinear system

Abstract

Fluidelastic systems are often lightly damped and exhibit weak non-linear damping. The weak linear and non-linear damping forces have a significant effect on the long term behaviour of the system. However, parameter identification methods tend to concentrate on identifying the strongest forces. In this paper the applicability of two identification methods to a single degree of freedom fluidelastic system is investigated using experimental data. The system is weakly non-linear under the flow conditions and is prone to the fluidelastic instability (i.e. self excited limit cycle behaviour). The crucial effect of the weak damping forces on the system stability and post-stable behaviour has been demonstrated. The non-linearity detection and identification has been done using an analytic representation of the displacement signal (FREEVIB method) and subsequently a non-linear decrement method was used for comparison. The identified models were used to predict the global behaviour of the system in the form of limit cycle amplitudes and this has been used as an indication of accuracy. It was found that the non-linear decrement method yields superior predictions, but it suffers from the limitation that the functional form of the system must be known a priori. Therefore it is concluded that employing both methods together provides a more powerful approach for parameter estimation in a lightly damped system where weak non-linear damping plays an important role.