Analysis and Parametric Identification of Air Foil Bearing Rotor System

Abstract

With the increasing demand for high speed turbomachinery, the dynamic analysis of rotor bearing system has become very important. Foil bearings with their exceptional feature of negligible wear has gained a very important place in the field of turbomachinery but the nonlinearities present in the bearing call for the dynamic analysis of these bearings. In the present work, a dynamic modeling approach and parametric identification scheme of a single disc rotor mounted on foil bearings is presented. Rotor is analyzed with conventional lumped parameter model. The time varying bearing forces are obtained by solving compressible Reynold’s equation using finite difference scheme and the convergence is checked using Gauss Sidle method. Effects of number of bump foils, foil thickness and bump pitch at different speeds of operation of rotor are studied on the dynamic response of rotor. For solving dynamic equations, Newmark method is employed using Matlab. A solid model of the rotor is prepared and then it is analyzed to obtain various critical speeds of the system and to obtain mode shapes at those speeds. A generalized model is developed based on radial basis neural networks to predict the foil thickness and bump pitch from a given frequency response of the rotor. Also a scaled model of shaft, disc and bearing system is fabricated for testing.