Experimental Validation of Inverse Techniques for the Remote Identification of Impact Forces in Gap-Supported Systems Subjected to Local and Flow Turbulence Excitations - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue (Article De Synthèse) Journal of Pressure Vessel Technology Année : 2011

Experimental Validation of Inverse Techniques for the Remote Identification of Impact Forces in Gap-Supported Systems Subjected to Local and Flow Turbulence Excitations

Résumé

Predictive computations of the nonlinear dynamical responses of gap-supported tubes subjected to flow excitation have been the subject of active research. Nevertheless, experimental results are still necessary for validation of the theoretical predictions as well as for asserting the integrity of field components. Because carefully instrumented test tubes and tube-supports are seldom possible, due to space limitations and to the severe environment conditions, there is a need for robust techniques capable of extracting relevant information from the actual vibratory response data. Although at the present time such analysis is overambitious, as far as the multisupported tube bundles of real-life components are concerned, the same instrumentation difficulties frequently apply in the case of laboratory test rigs. Therefore, the subject of this paper is of practical significance even in the more modest realm of laboratory experiments. The knowledge of the dynamical contact/impact (vibro-impact) forces is of paramount significance, and also the tube/support gaps. Following our previous studies in this area using wave-propagation techniques (De Araújo et al., 1998; Antunes et al., 1998; Paulino et al., 1999), we recently applied modal methods for extracting such information. Based on numerically simulated time-domain vibro-impact responses, the dynamical support forces, as well as the vibratory responses at the support locations, were identified from one or several vibratory responses at remote locations, from which the support gaps could also be inferred (Delaune et al., 2010). Also recently, for the related problem of friction force identification on bowed strings, preliminary experiments have shown the feasibility of these identification techniques (Debut et al., 2010). In the present paper, the modal identification techniques developed by Delaune et al. (2010) and Debut et al. (2010) are tested using an experimental rig built at Commissariat à l’Energie Atomique et aux Energies Alternatives (CEA/Saclay), consisting of a randomly excited clamped-free beam which impacts on an intermediate gap-support. Identification of the impact force, as well as of the beam motion at the gap-support, is achieved based on remote measurements of the beam response provided by two accelerometers. A significant feature of the experimental identifications presented in this paper is that, beyond the results obtained under a point-force shaker excitation, we test here an original technique to identify the gap-supported reactions in flow-excited systems, which was recently introduced by Delaune et al. (2010). As for most inverse problems, the identification results may prove sensitive to both noise and modeling errors. Therefore, regularization techniques discussed by Delaune et al. (2010) are used to mitigate the effects of unmeasured noise perturbations. Overall, the experimentally identified results compare reasonably well with the measured contact forces and motions at the gap-supports. Actually, even if our identifications are not immaculate at the present time, they remain nevertheless quite usable.
Fichier non déposé

Dates et versions

hal-04501759 , version 1 (12-03-2024)

Identifiants

Citer

Xavier Delaune, Philippe Piteau, Vincent Debut, Jose Antunes. Experimental Validation of Inverse Techniques for the Remote Identification of Impact Forces in Gap-Supported Systems Subjected to Local and Flow Turbulence Excitations. Journal of Pressure Vessel Technology, 2011, 133 (5), ⟨10.1115/1.4002926⟩. ⟨hal-04501759⟩

Collections

CEA DEN ISAS
3 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More