AbstractThe objective of this thesis is motivated by the industrial needs to progress the development of integrating experimental test and simulation methods for nonlinear vibrating structures. The demonstration of the test and simulation integration approach adopted in this thesis is founded on three different complementary model types, namely, white, black and grey box models. The first part of the thesis presents a black-box data driven approach for modelling local non-linearities. In this case, the entire non-linear identification and simulation is derived from experimental data with no prior knowledge or assumption of the non-linear characteristics with demonstration done on measured data obtained from the experimental campaign of an aero-engine casing structure.
The second part of this thesis is devoted to the requirement for quantifying and associating physics-based parameters to each non-linear characteristic observed in an engineering structure. A novel method for non-linear system identification based on revising an existing black-box oriented state space model technique is presented. The proposed time-domain method is formulated based on converting a black-box oriented state space model algorithm for a nonlinear system to a grey-box state space model. The advantage of this method is the ability to extract a lower number of parameters with physics based interpretation. This method is tested and validated on simulated data and experimental data obtained from an aerospace structure. The final part of this thesis addresses the current challenge of predicting the vibration response and validating simulated models of assembled engineering structures with identified non-linearities. This challenge is investigated through a framework strategy that permits simulation model parameters to be updated and upgraded based on experimental observations, to constitute a white-box model approach. This is achieved by embedding the current linear modal analysis technique into a test-simulation integration framework for non-linear systems.
|Date of Award||28 Nov 2019|
|Supervisor||Branislav Titurus (Supervisor), Dario DiMaio (Supervisor) & David J Ewins (Supervisor)|