In this paper, a new method of finite element model updating using neural networks is presented. Many previous model updating techniques have exhibited inconsistent performance when subjected to noisy experimental data. From this background it is clear that a successful model updating method must be
FINITE ELEMENT MODEL UPDATING USING ANTIRESONANT FREQUENCIES
โ Scribed by K. JONES; J. TURCOTTE
- Publisher
- Elsevier Science
- Year
- 2002
- Tongue
- English
- Weight
- 204 KB
- Volume
- 252
- Category
- Article
- ISSN
- 0022-460X
No coin nor oath required. For personal study only.
โฆ Synopsis
This paper uses antiresonant frequencies in the "nite element model updating of an experimental 6-m aluminum truss and analyzes the physical correctness of the updated model by using it to detect damage. Rigid elements are used to simplify the modelling of welded joints, and their dimensions are used as parameters in an iterative update based on eigenvalue and antiresonance sensitivities. An update using both natural frequencies and antiresonant frequencies is shown to produce a 48% better correlation to experimental frequency response functions (FRFs) than an update that uses only natural frequencies. The antiresonant updated model is used to predict FRFs for the truss in 112 damaged con"gurations. Pattern classi"cation and curve-"t algorithms for damage detection are tested. The curve-"t method correctly identi"ed damage 92)6% of the time compared to 76)1% for the pattern classi"er. The high quality of the model is attributed to the use of rigid elements that are updated using antiresonant frequencies.
๐ SIMILAR VOLUMES
In this paper, the use of thermoelastic measurements to improve a "nite element model is investigated. The originality of the procedure lies in the use of this stress sum "eld measurement and in the new solution method of the modelling error location stage. Measurement inaccuracies and expansion err
This paper addresses the problem of selecting a side constraint and determining the regularisation parameter in model updating. The weight to be attached to the constraint is determined by the regularisation parameter. Methods based on singular value decomposition, cross-validation, and L-curves are
Updating procedures are concerned with the construction of mass, stiffness and (possibly) damping matrices, near to some base matrices, which yield some response data close to some experimental values. One of the fundamental questions in updating relates to the criteria for allowable mass and stiffn
In this paper, an updating technique that includes antiresonances in the de"nition of the output residual is considered. Antiresonances are not a global system property, but are typical of each frequency response function (FRF), thus allowing the residual vector to be enlarged with data identi"ed fr