Modelling the dynamic interactions of rolling bearings

Berlin / Logos-Verl. (2008) [Dissertation / PhD Thesis]

Page(s): VI, 141 S. ; Ill., graph. Darst.


This work deals with the generation and propagation of structure-borne sound by the rolling contact in bearings. Rolling bearings are widely used in machines and, in early stages of operation, the interaction of the rough surfaces of races and rolling bodies produces an excitation that will propagate to the surrounding structure. The dissertation presents a model for the determination of the structure-borne excitation due to rolling, using the condition of the surfaces and the dynamics of the rolling movement as input. The model is able to simulate the time evolution of the excitation generated by the bearing that is imposed to the machine. Additionally, this work describes the experimental evaluation of the transfer functions of the test machine, i.e. the influence of the path between the point where the excitation is generated (in the bearing) to the point where it is usually measured (on the machine’s housing). For this purposes, special actuators made of one rolling element are built, calibrated and mounted inside the bearing. This allows a broad-band determination on the transfer function with the machine mounted (closer to the real situation). The key contribution of this work is a model of the noise generated in rolling bearings. With a description of the source and the influence of the path between source and sensor, it is now possible to simulate the vibration of the machine that would be capture in its housing. Measurements in operation conditions are made for comparisons with cylindrical and spherical rolling bearings. This approach shows that a deeper insight in the understanding of machine vibration can be gained when one tries to model sources and treats the system as a filter that has to be described in order to evaluate the influence of the transfer path. The adaptation of the model for lubrication is also implemented and described. A part of the inverse problem,i.e. to separate back the bearing signal from measured signals is also done. Future work remains in trying to associate this vibration to the condition of the bearing’s surfaces. For this, more measurements to create a bigger database of surfaces are necessary.



Guimarïaes, Joïao Henrique Diniz


Vorländer, Michael


  • ISBN: 978-3-8325-2010-6
  • URN: urn:nbn:de:hbz:82-opus-25290