INVESTIGADORES
ROSALES Marta Beatriz
congresos y reuniones científicas
Título:
Crack detection in a spinning beam
Autor/es:
ROSALES MB; P FILIPICH, C; A CONCA, M
Lugar:
San Carlos de Bariloche, Argentina
Reunión:
Congreso; XIV Congreso sobre métodos numéricos y sus aplicaciones; 2004
Institución organizadora:
AMCA, CNEA
Resumen:
This paper deals with the detection of a crack in a spinning beam(rotor) by means of the measured frequencies method.comm{It iswell known that a structural component shows changes in itsbehavior with the presence of a crack and several methods arebased on the analysis of various structural parameters (mass,stiffness, modulus) and modal parameters (frequencies, modeshapes, modal damping). The technique using the measuredfrequencies} This technique as a crack detection criterion hasbeen extensively applied in the last decade meanly due to the factthat frequencies are, among other dynamical parameters, easilymeasured. However the inverse problem of determination of thecrack parameters (location and depth) for a given set of measuredfrequencies is not simple. An efficient numerical technique has tobeen employed so as to obtain acceptable results. In this studythe effect of the crack is modeled through the introduction ofintermediate flexional springs in a spinning beam of circularcross section and rotating around its longitudinal axis withconstant angular velocity. The beam-springs analytical model isfirst stated and the power series method is employed to obtain thesolution for a given set of data, say the springs constants, thecrack location or the frequency. It should be noted that thesprings and the crack depth may be related by some expression fromFracture Mechanics. Here a systematization of the series  givesrise to an efficient numerical method.  An algorithm is thenwritten and prepared to solve the inverse problem. Thenexperimental frequencies are measured in a cracked spinning beam.At this stage, this experiment is performed numerically, with aspinning beam with a notch.  The flexural frequencies areobtained. These are the input for the previous numerical algorithmto find the solution of the inverse problem: i.e. predict thecrack depth and location resp., given the measured frequencies.Numerical examples are included with an evaluation of the errorsin the results. The methodology has been tested previously in annon spinning Euler-Bernoulli beam with very promising results.