Dynamics of non-linear structures: modal interaction and non-linear identification

Despite the combined improvements in computing power and modelling capabilities
one major hindrance to optimal design is the present incapability to capture
experimentally (identify) nonlinear dynamic behaviour which characterise many
engineering structures. Currently, it is standard engineering practice to apply
techniques that can provide an answer to whether or not the system being tested is
nonlinear but do not help to identify the type and magnitude of nonlinearity. In this
paper it is shown where and why the usual methods cannot be used and to suggest new
technique to identify the source of the non-linearity and its effect on the behaviour of
the entire structures. A mathematical model and numerical simulations of a 3 degrees
of freedom systems are first presented. Subsequently, a test-rig with a known
nonlinearity (location, type and magnitude) is designed and built. The methodology
applied to numerical simulation is then applied to the experimental data for
comparison. The final goal is to transform a set of measured data into a set of modal
parameters which eventually will allow the (re)construction (or validation) of a model
which can then be used for prediction.