|Title||Πειραματικές Μέθοδοι στις Ταλαντώσεις / EXPERIMENTAL METHODS IN VIBRATION|
|Cycle / Level||2nd / Postgraduate|
|Core||Elective Courses belonging to the selected specialization||1||1||5|
|Academic Year||2018 – 2019|
|Faculty Instructors|| |
Konstantinos Karatzas, 3hrs
|Class ID|| |
Basics of vibration theory, Matrix algebra, computer programming with matlab or equivalent comp. environment.
Students should be able to apply the methods and algorithms taught in the class in order to:
(a) learn from vibration data and thus extract knowledge
(b) reveal “hidden” relationships between heterogeneous parameters
(c) visualize and explain parameter relationships and dynamic (vibration) system behaviour, as well as
(d) simulate, model and provide forecasts for a studied dataset.
Analytical modal analysis (structural model equations, modal model equations, frequency response function, impulse response functions, viscous and structural damping, classically and non-classically damped systems). Signal processing theory (Fourier transform for periodic and non-periodic signals, sampled time and frequency transforms, aliasing, leakage, auto- and cross-correlation, auto- and cross-power spectra, frequency response functions and coherence). Modal parameter estimation (multiple-input multiple-output time and frequency domain methods, operational modal analysis methods). Testing techniques (forced and motion transducers, data acquisition and analysis system, test set-up, FRF measurements, excitation considerations). Modal model validation methods (modal assurance criterion). Finite element model validation and updating (least squares, maximum likelihood, Bayesian inference methods). Data oriented investigations, analysis and modelling: Arithmetic, logical and nominal parameters, construction of feature space. Data investigation and analysis methods (Principal Component Analysis, Self Organizing Maps, Clustering). System modelling with the aid of experimental data (Artificial Neural Networks, etc). Analysis and optimization of experimental procedures. System diagnostics possibilities on the basis of experimental data. Experimental design (information theory methods for optimal sensor locations and optimal selection of excitation characteristics). Applications from the area of machine dynamics and turboengines.
Learning material is provided via ICT. Computational exercises are completed with the aid of Matlab and WEKA. All course material, information, announcements and communications are done via the Internet and the course web site.
Lecture slides and lecture notes are made available