Please use this identifier to cite or link to this item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/326459
Type: Artigo
Title: Experimental Wear Parameters Identification In Hydrodynamic Bearings Via Model Based Methodology
Author: Mendes
Ricardo U.; Machado
Tiago H.; Cavalca
Katia L.
Abstract: When a rotor operates in very low rotating speeds, the fluid-film in hydrodynamic bearings is not completely formed, which allows direct metal to metal contact between the bearing and the journal. Numerous start/stop cycles also may lead to bearing wear. In this context, the present work aims to evaluate a model-based method for identification of hydrodynamic bearing wear parameters: more specifically, wear depth and angular position. The proposed method is based on minimizing an objective function, which compares the directional Frequency Response Function (dFRF) of the physical system with the response of the model developed here. It contemplates the numerical model of the rotor and the worn bearings, in this case, represented by linearized coefficients of damping and stiffness. Through experimental results, a deeper evaluation of the identification method is presented, such as the sensitivity to the number of points of the dFRF used in the objective function, different meshes of starting points, and the region of the dFRF used - around the resonance peak or around the oil-whirl peak. The analyses are conducted for different levels of bearing wear. The precision obtained in the results indicates that the presented method is a promising tool in the identification of bearing wear. (C) 2016 Elsevier B.V. All rights reserved.
Subject: Bearings Wear
Parameters Identification
Directional Frequency Response
Editor: Elsevier Science SA
Lausanne
Rights: fechado
Identifier DOI: 10.1016/j.wear.2016.12.002
Address: http://www-sciencedirect-com.ez88.periodicos.capes.gov.br/science/article/pii/S0043164816307359
Date Issue: 2017
Appears in Collections:Unicamp - Artigos e Outros Documentos

Files in This Item:
File SizeFormat 
000394400900011.pdf2.23 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.