In this work, we investigate the active vibration isolation and damping of sensitive equipment. Several single-axis isolation techniques are analyzed and tested. A comparison between the sky-hook damper, integral force feedback, inertial velocity feedback and LagLead control techniques is conducted using several practical examples. The study of single-axis systems has been developed and used to build a six-axis isolator. A six degrees of freedom active isolator based on Stewart platform has been designed manufactured and tested for the purpose of active vibration isolation of sensitive payloads in space applications. This six-axis hexapod is designed according to the cubic configuration; it consists of two triangular parallel plates connected to each other by six active legs orthogonal to each other; each leg consists of a voice coil actuator, a force sensor and two flexible joints. Two different control techniques have been tested to control this isolator : integral force feedback and Lag-Lead compensator, the two techniques are based on force feedback and are applied in a decentralized manner. A micro-gravity parabolic flight test has been clone to test the isolator in micro-gravity environment. ln the context of this research, another hexapod has been produced ; a generic active damping and precision painting interface based on Stewart platform. This hexapod consists of two parallel plates connected to each other by six active legs configured according to the cubic architecture. Each leg consists of an amplified piezoelectric actuator, a force sensor and two flexible joints. This Stewart platform is addressed to space applications where it aims at controlling the vibrations of space structures while connecting them rigidly. The control technique used here is the decentralized integral force feedback.

The thesis concentrates on a methodological research on categorical structural optimization by means of manifold learning. The main difficulty of handling the categorical optimization problems lies in the description of the design variables: they are presented in a discrete manner and do not have any orders. Thus the treatment of the design space is a key issue. In this thesis, the non-ordinal categorical variables are treated as multi-dimensional discrete variables, thus the dimensionality of corresponding design space becomes high. In order to reduce the dimensionality, the manifold learning techniques are introduced to find the intrinsic dimensionality and map the original design space to a reduced-order space. The mechanisms of both linear and non-linear manifold learning techniques are firstly studied. Then numerical examples are tested to compare the performance of manifold learning techniques. It is found that Principal Component Analysis (PCA) and Multi-dimensional Scaling (MDS) can only deal with linear or globally approximately linear cases. Isomap preserves the geodesic distances for non-linear manifold, however, its time consuming is the most. Locally Linear Embedding (LLE) preserves the neighbour weights and can yield good results in a short time. Kernel Principal Component Analysis (KPCA) works as a non-linear classifier and we proves the reason why it cannot preserve distances or angles in some cases.Based on the reduced-order representation obtained by Isomap, the graph-based evolutionary crossover and mutation operators are proposed to deal with categorical structural optimization problems, including the design of dome, six-story rigid frame and dame-like structures. The results show that the proposed graph-based evolutionary approach constructed on the reduced-order space performs more efficiently than traditional methods including simplex approach or evolutionary approach without reduced-order space.The Locally Linear Embedding is applied to reduce the data dimensionality and a polynomial interpolation helps to construct the responding surface from lower dimensional representation to original data. Then the continuous search method of moving asymptotes is executed and yields a competitively good but inadmissible solution within only a few of iteration numbers. Then in the second stage, a discrete search strategy is proposed to find out better solutions based on a neighbour search. The ten-bar truss and dome structural design problems are tested to show the validity of the method. In the end, this method is compared to the Simulated Annealing algorithm and Covariance Matrix Adaptation Evolutionary Strategy, showing its better optimization efficiency.In order to deal with the case in which the categorical design instances are distributed on several manifolds, we propose a k-manifolds learning method based on the Weighted Principal Component Analysis. The obtained manifolds are integrated in the lower dimensional design space. Then the two-stage search method is applied to solve the ten-bar truss, the dome and the dam-like structural design problems. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished

Le sujet de ce mémoire est l'étude d'un interféromètre de Fizeau a décalage de phase pour la mesure de surfaces planes et sphériques a des précisions nanométriques. Nous avons recensé et étudié les sources de limitation et recherche les conditions ou les solutions rendant leur contribution inférieure a un millième de longueur d'onde en valeur crête à crête (la longueur d'onde est égale a 632,8 nanomètres). Six sources potentielles d'erreurs ont été examinées: la source laser, la détection, l'élément qui crée le déphasage, le diffuseur mobile du système d'imagerie, les aberrations géométriques des composants optiques et la nature même des interférences. Nous avons optimise la configuration interférométrique. De nouveaux algorithmes ont été développes. Ils sont basés sur des méthodes des moindres carres non linéaires. Ces algorithmes sont auto calibrés mais demandent beaucoup de temps de calcul. Pour obtenir des resultats rapidement, nous avons complete nos techniques de depouillement par une méthode basée sur une correction a posteriori de la formule de carre. La validation a montré les performances de l'instrument. Avec les méthodes basées sur les moindres carrés, les mesures de bruit ont montre que les fluctuations aléatoires sont de l'ordre de 0,21 nanomètres rms ; les erreurs systématiques ont une contribution inférieure a 1,8 nanomètres crête à crête. La cale piézo-électrique qui réalise le déphasage, et le bruit de la détection, sont en fait les principales sources de limitation. La correction des défauts de positionnement de la cale piézo-électrique a été essayée sur l'algorithme de carre et sa post correction ; les résultats indiquent une exactitude meilleure que 1 nanomètre crête à crête. L'interféromètre de Fizeau a décalage de phase a été applique a l'étalonnage absolu de surfaces planes. Une méthode basée sur des combinaisons de translations et de rotations d'une des surfaces par rapport a l'autre, a été présentée. Cette méthode a été utilisée pour la mesure d'une surface plane en silice.

L'étude ferroviaire présentée dans ce mémoire traite de trois types de dégradations de profils des roues et des rails en service. Comme tous les solides devant assurer une liaison par contact, les roues et les rails ferroviaires sont affectés par des dégradations telles que l'usure et la fatigue. Dans un premier temps, on s'intéresse à l'usure régulière des profils des roues et des rails. Ce type d'usure influence en particulier la stabilité des véhicules, donc le confort et la sécurité. Nous exposons notre modèle de simulation numérique de l'usure transverse des profils ; sa validation étant assurée par des applications à des profils mesurés sur le réseau S.N.C.F. Nous abordons ensuite le phénomène d'usure irrégulière longitudinale des roues à grande vitesse. Nous traitons cette partie avec de nouvelles hypothèses sur un mécanisme d'apparition d'ondulations. L'origine de ce phénomène est due au couplage, à travers le contact, de la dynamique transversale des bogies et de la torsion de l'axe des essieux. En nous appuyant sur deux modèles de simulations numériques, nous montrons la validité des hypothèses et nous mettons en évidence le mécanisme d'usure irrégulière longitudinale des roues. La dernière partie est consacrée à la fatigue des rails soumis au contact roulant, on y étudie une forme particulière appelée le Head Checking. Une première phase, basée sur l'étude des efforts de contact dus au comportement en courbes de certains véhicules ferroviaires, nous permet d'isoler les caractéristiques du système véhicule-voie les plus contraignantes. Une seconde étape consiste à étudier les conditions d'amorçage des fissures par deux approches : le critère de Dang Van et la mécanique de l'endommagement. Les résultats obtenus nous permettent de proposer des dispositions qui semblent appropriées à la lutte contre le Head Checking.

In this work, we investigate the active vibration isolation and damping of sensitive equipment. Several single-axis isolation techniques are analyzed and tested. A comparison between the sky-hook damper, integral force feedback, inertial velocity feedback and LagLead control techniques is conducted using several practical examples.<p><p>The study of single-axis systems has been developed and used to build a six-axis isolator. A six degrees of freedom active isolator based on Stewart platform has been designed manufactured and tested for the purpose of active vibration isolation of sensitive payloads in space applications. This six-axis hexapod is designed according to the cubic configuration; it consists of two triangular parallel plates connected to each other by six active legs orthogonal to each other; each leg consists of a voice coil actuator, a force sensor and two flexible joints. Two different control techniques have been tested to control this isolator :integral force feedback and Lag-Lead compensator, the two techniques are based on force feedback and are applied in a decentralized manner. A micro-gravity parabolic flight test has been clone to test the isolator in micro-gravity environment.<p><p>ln the context of this research, another hexapod has been produced ;a generic active damping and precision painting interface based on Stewart platform. This hexapod consists of two parallel plates connected to each other by six active legs configured according to the cubic architecture. Each leg consists of an amplified piezoelectric actuator, a force sensor and two flexible joints. This Stewart platform is addressed to space applications where it aims at controlling the vibrations of space structures while connecting them rigidly. The control technique used here is the decentralized integral force feedback.<p><p> / Doctorat en sciences appliquées / info:eu-repo/semantics/nonPublished