Industrial Application of a Model Based Rotor Balancing Technique
Résumé
Unbalance is one of the most common problems found in rotating machines in the context of industrial plants. This phenomenon can generate high vibration amplitudes that can lead to fatigue of rotor elements. Various approaches have been proposed in order to solve the rotor balancing problem, as the so-called influence coefficients technique. This paper presents an alternative balancing methodology for rotating machines, aiming at overcoming the limitations faced by the influence coefficients method. This alternative technique first identifies the model of the machine, and then the unbalance is determined by solving a typical inverse problem through a pseudo-random optimization method, namely Differential Evolution. The proposed balancing methodology does not require trial weights for its implementation; however, a reliable mathematical model of the rotating machine is required. In this paper the balancing methodology is applied to two different cases, for illustration purposes. The first one is a numerical study of a rotor composed by a horizontal flexible shaft on fluid film bearings and the second is a rotating machine operating in an industrial environment. The results indicate the effectiveness of the proposed technique for both application cases.