|M.Sc Student||Zivan Ayal|
|Subject||A Method for Optimal Design and Modeling of Dynamic|
Test under Given Constrains
|Department||Department of Mechanical Engineering||Supervisors||Professor Haim Abramovich|
|Dr. Arie Elka|
|Full Thesis text - in Hebrew|
The aim of the research thesis was to develop an algorithm that enables the planning and designing of dynamic test systems, such that responses measured in field conditions are replicated optimally in laboratory tests.
A basic requirement of any system is to function correctly under field conditions in which it is intended to work. In order to ensure that the system will operate as intended in the field throughout its life-cycle, it is required to test the system under expected field conditions. In order to deal with the limitations of field tests, laboratory tests are performed in order to simulate field conditions in a controlled, safe, and low cost environment.
Laboratory test should be designed such that the test conditions would comply with field conditions (load distribution, energy flow etc.) experienced by the system. Unfortunately such replication represents a major challenge that requires suitable laboratory infrastructure and knowledge, frequently preventing the achievement of the aims. One can attempt to reconstruct the system's measured field responses directly by controlling the exerted laboratory excitation by means of: number of shakers (one or more), locations, frequencies and intensity of each spectral component. Such a method for controlling the system's responses is presented in the present research in the form of an Optimal Test Design (OTD) algorithm.
Development, verification and preliminary validation of the OTD algorithm were performed using a finite element model of a given beam. The beam’s model was verified using analytical solutions of a one-dimensional structure.
The algorithm sequence consists of three main steps: excitation frequencies are calculated using the NLLS (Non-Linear Least Square) procedure, the system's steady-state deflection is then calculated based on the beam’s mode shapes, and the force excitation locations and intensities are calculated based on the system's force-deflection transfer function.
The OTD algorithm was validated through laboratory tests consisting of four vibration configurations of a beam under free-free boundary conditions. Calculated responses, as a result of optimal excitation, were compared to measured responses and deflection deviations were found to be between 3% to 10% (depending on vibration configuration).
In conclusion, the present research presents the development, verification and validation of the OTD algorithm aimed at effectively modeling and designing of dynamic laboratory tests.