IDENTIFICATION OF DIFFERENCES BETWEEN FINITE ELEMENT ANALYSIS AND EXPERIMENTAL VIBRATION DATA

This paper presents a general procedure for identifying and quantifying the difference between finite element (fe) models and experimental data, which has been developed and demonstrated with simulated experimental data. The differences, which can be computed for linear, viscously damped structures, are presented in terms of mass, damping, and stiffness coefficients. This enables possible modelling problems to be identified in the fe or analytical model. From data generated for a damped cantilever beam and for a damped simply supported beam, it was found that the accuracy of the computed differences increased as the number of experimentally measured modes, included in the calculations, is increased. When the number of experimental modes is at least equal to the number of translational degrees of freedom, both the location and magnitude of the differences can be computed very accurately. When the number of modes is less than this quantity, the location of the differences may be determined, even though their magnitudes will be underestimated. When too few modes are available, neither the location nor the magnitudes of the differences can be identified. In practice, the experimental frequencies and mode shapes will need to be measured very accurately, before the differences can be attributed to deficiencies in the analytical model. If the experimental data is not precise, the computed differences can still provide considerable insight into the possible locations of deficiencies. (TRRL)

  • Availability:
  • Corporate Authors:

    The Modino Press Limited

    3 Greenway
    London,   England 
  • Authors:
    • LAWRENCE, C
  • Publication Date: 1988-6

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00492771
  • Record Type: Publication
  • Source Agency: Transport Research Laboratory
  • Files: ITRD, TRIS
  • Created Date: Mar 31 1990 12:00AM