Amélioration de méthodes de modification structurale – Applications

Structural dynamic modification methods improvement – Applications

¹ École centrale Paris, Grande voie des Vignes, 92295 Chatenay-Malabry, France
² EDF DR&D, 1 Av. du Général de Gaulle, 92141 Clamart Cedex, France

Auteur de correspondance : corus@mssmat.ecp.fr

Reçu : 23 Novembre 2004
Accepté : 31 Mars 2005

Résumé

Une méthode proposant d'estimer l'effet de modifications structurales distribuées est exposée ici. Une évolution des formulations classiques permettant de prendre en compte la non-coïncidence des points de mesure et des points de couplage est présentée. La mise en œuvre de la méthodologie à un exemple académique et à un cas d'application industriel permet de mettre en lumière les avantages de la méthodologie exposée.

Abstract

Structural modification methods are designed to estimate the dynamic behaviour of a structure after a modification. This estimation is obtained by coupling an experimental model of the non modified structure and a numerical model of the modification For short delays and little resources, these methods, as presented in [1] for example, can be efficiently implemented. Few authors dealt with the problem of distributed modifications (W. D'Ambrogio and A. Sestieri in [2] and [3], K. Elliot and L. Mitchell in [4], or B. Schwarz and M. Richardson in [5]). Nevertheless, all proposed methods impose a specific test configuration with several measurement points over the substructures interface. The improvements presented here allow the use of data that do not require to be measured on the substructures interface. The difficulties generated by the lack of displacement continuity at the interface as well as the different locations of measurement and coupling points have been overtaken by means of data expansion and model reduction techniques. The expansion process uses a reduced displacement basis for describing the interface. It is derived from a linear combination of the measurement point motions. To select the appropriate basis, two indicators based upon two displacement estimators on the interface are introduced. Two case study are presented and discussed, showing the ability of this approximate approach to accurately foresee structural modification effects.