Identification and Correction of Errors in Analytical Models Using Test Data - Theoretical and Practical Bounds

摘要

The paper begins with a review of procedures for identifying error lacations in analytical (finite element) models using static and/or dynamic test data. A classification of possible error sources in an analytical model is then presented and their influence is discussed w.r. to the mass and stiffness matrices, the modal data and especially the error indicators used in the different localisation procedures. Bounds for error localisation are presented in this section. In order to derive practical bounds the influence of a typical measurement environment as given by incomplete and noisy test data on the error indicators is investigated. Using a simple academic bar example it is shown how modeling and measurement errors are superimposed. The limits for successful error localisation and subsequent model updating in presence of realistic test data are discussed.