A benchmark is organised to quantify the variability relative to structure dynamicscomputations. The chosen demonstrator is a pump in service in thermal central units, whichis an engineered system with not well-known parameters, considered in its workenvironment. The blind modal characterisation of the separate pump components shows a5%–12% variability on eigenfrequency values and a less than 15% frequency error incomparison with experimental values. The numerical-experimental MAC numbers reach 0.7 atthe maximum, even after updating. An example of modal results on the pump assembly fixedis presented, which shows a larger discrepancy with measurement values, essentially due tothe modelling of the interfaces and boundary condition, and to the possible simplificationof the main components F.E. models to reduce their size. Though a significant frequencyerror, the first overall modes are correctly identified. If this tendency can be confirmedfrom all the participants’ results, the conclusion to be drawn is that, if the predictivecapability of F.E. models to represent the dynamical behaviour of sub-structures issatisfactory, the one relative to structures that are built-up of several components doesnot allow their confident use. Additional information issued from measurements is neededto improve their accuracy.

Cet article est dédié à l’étude des vibrations libres et forcées d’une structure àsymétrie cyclique, soumise à des non-linéarités géométriques, par la méthode de la balanceharmonique (HBM). Dans le but d’étudier l’influence des non-linéarités un modèle simplifiéa été développé. Après ajustement des paramètres du modèle, les équations du mouvement seprésentent sous la forme d’équations différentielles du second ordre, linéairementcouplées, où les non-linéarités se traduisent par des termes polynomiaux d’ordre deux ettrois. Les solutions périodiques de ces équations sont recherchées grâce à la méthode dela balance harmonique couplée avec une procédure de continuation. Dans le cas libre, enplus des modes non-linéaires similaires et non similaires, on met en évidence des modesnon-linéaires localisés. Dans le cas forcé, plusieurs types d’excitations sont considérées(excitation sur le premier mode propre linéaire et excitation detunée) et on étudieparticulièrement l’influence du niveau d’excitation sur la structure des réponsesdynamiques. Pour une excitation suffisamment perturbée, on montre que plusieurs solutionspeuvent coexister, certaines d’entre elles étant représentées par des courbes fermées dansle plan fréquence-amplitude.

Le Forage Grande Vitesse Vibratoire (FGVV) est un nouveau procédé de perçage qui permet de tripler la productivité des perçages profonds en supprimant les cycles d’évacuation des copeaux ainsi que la lubrification. Pour cela un porte-outil spécifique aété développé, qui entre en vibrations axiales auto-entretenues lors de la coupe et permet le fractionnement du copeau, facilitant ainsi sonévacuation. Pour anticiper uneéventuelle diminution de la durée de vie des roulements à billes de l’électrobroche résultant de ces vibrations, un modèle global du comportement dynamique du système tête vibratoire-électrobroche est présenté. Ce modèle est basé sur la modélisation du comportement des entités structurales par deséléments-finis de type « poutre-rotor» prenant en compte les effets dynamiques associés aux hautes fréquences de rotations de l’électrobroche, couplé à des modèles d’interfaces identifiés par la méthode de couplage de réceptance. Le modèle est validé à l’aide d’expérimentations qui montrent une bonne correspondance entre les résultats expérimentaux et numériques. Le modèle est ensuite utilisé pour calculer la durée de vie des roulements à billes de l’électrobroche lors de différentes opérations de perçage. L’influence de la fréquence de rotation et de l’avance sur la durée de vie des roulements estétudiée, permettant ainsi de faire ressortir des plages de fonctionnement optimal.

The paper presents the experimental work developed by the authors for the identification of rotordynamic coefficients of air bearings. Air bearings work at very high rotation speeds and are known to have nonlinear dynamic characteristics depending at least on the excitation frequency. The paper presents two very similar test rigs, the testing procedure and the algorithm for identifying the rotordynamic coefficients. The test rigs consist of a rigid rotor guided by fixed bearings and driven by a spindle. The dynamic loads are applied by two orthogonally mounted shakers applying two linear independent excitations. Two air bearings are analysed in the present paper. The test procedure is first developed for a simple circular bearing with easily predictable dynamic characteristics. Its rotordynamic coefficients are identified by using a least square procedure based on rational functions. The coefficients are compared to theoretical results in order to underline the limits of the identification algorithm. The procedure is next applied to a first generation foil bearing. Rotordynamic coefficients are presented for different working conditions (static loads and rotation speeds) and are discussed comparing them to circular air bearings.

Rolling bearing is probably the most widely used component in rotating mechanicalequipments and its condition monitoring and fault diagnosis to prevent the occurrence ofbreakdown is growing in interest since many years. Vibration signal based methods are themost popular and have been adopted in many kinds of condition monitoring systems. Startingin the early 60, an immense range of different methods has been proposed on this basis, toperform diagnosis, fault identification and classification of bearing faults. Among theothers, one typical approach consists in deep analysis of the most informative frequencyrange output of the system under test; the identification of this band is notstraightforward because the fundamental task consists in finding out the band which is themost informative in contents which, in turn, might not be corresponding to that one of themaximum response, as claimed by some authors. In this paper, Spectral Kurtosis and SupportVector Machine are analysed and compared and it is shown that they typically reach similarresults, in spite of their totally different approach. A brief description of both methodsis given and laboratory data are analysed from a lab rig which uses spare parts of a fullsize power transmission gearbox, designed by AVIO. By taking advantage of thesecomparisons, the analyses are conducted using classical indicators applied to the specificbands suggested by previous analysis such as the RMS and other statistical quantities.Multi dimensional graphs are reported to show the reliability of the obtained results.

An emerging research topic in civil engineering is the dynamic interaction between crowdsand structures. Structures such as footbridges, which oscillate due to the crossing of agroup of pedestrians, or stands within stadia or concert halls, which vibrate due to therythmic movement of the audience are of particular interest. The objective of this studyis twofold: modelling the movement of pedestrians with consideration ofpedestrian-pedestrian, and pedestrian-obstacle interactions, and the incorporation of apedestrian-structure coupling in the previous model. Frémond’s model, which allows us tosimulate the movement of an assembly of particles and accounts for collisions amongconsidered rigid particles, is presented and adapted to the crowd by giving a willingnessto the circular particles, which allows each pedestrian to move according to a giventarget. To handle the crowd-structure interaction in the case of lateral oscillations offootbridges, the Kuramoto differential equation governing the time evolution of thelateral motion of each pedestrian is implemented in the previous model. Preliminaryresults obtained from numerical simulations are presented and discussed.

This paper approaches a novel technique to estimate cable tension simply based on itsvibration response. The vibration response has been quite extensively adopted in the pastdue to its simplicity and, mainly, because the inverse approach allows the tensionestimation with the cable in its original site. A first tentative approach consists inusing a certain number of experimentally measured natural frequencies to be introduced inthe theoretical vibration formula; this formula, however, involves also the cable length,the cable mass per unit length and its flexural rigidity. Unfortunately, some problemsarise in its application to real structures, such as the case of suspended andcable-stayed bridges, because the exact cable length cannot be measured (it appears at thefourth exponent in the vibration formula); moreover section and weight can be estimatedwithin a certain degree of accuracy, whilst the boundary conditions are often defined withdifficulty. A novel extension of the method is here proposed, which takes advantage from amoving mass travelling on the cable. This is the case occurring when cables are verifiedwith magnetic-based technology to detect rope faults and cross section reduction. In thisway, the extracted natural frequencies are varying with time due to the moving load, andhence they have to be extracted adopting a time-varying approach. Although someapproximation linked to the shape modification must be introduced, a simple iterativeprocedure can be settled, by considering the cable length as an unknown. An estimation ofthe equivalent length is given, and successively this value is used to obtain anestimation of the cable tension.

For a controlled structure, the changes of physical characteristics like naturalfrequencies, modal damping coefficients, and what’s more on mode shapes will destabilizethe controlled system. To keep the control performance and guarantee its robustness, a newmodal control method using real-time identification is proposed to overcome theinstability of controlled structure, especially induced by the inversion of mode shape. Inthis paper, a modal model of time-varying mechanical structure is supplied by anidentifier. The choice of the kind of control permits to use this time-varying model foroptimizing and self-adapting the controller. In order to supply the controller, a modalobserver is used and also updated. For illustrative purpose, this proposed methodology iscarried out on a simple but representative time-varying mechanical structure. On thisexample, an inertia modification leads not only to low modal frequency shifts but also toinversion of a mode shape. This mode shape inversion is shown to destabilize thecontrolled structure when the control system is not updated. The overall procedure will bedescribed through simulations and performed for different operating conditions, which willprove that mode shapes have to be precisely determined and updated in the controller andobserver to guarantee a robust modal control with high performances in spite of thechanges of structure.

In order to overcome the loss of performances issue when scaling resonant sensors down toNEMS, it proves extremely useful to study the behavior of resonators up to largedisplacements and hence high nonlinearities. A comprehensive nonlinear multiphysics modelbased on the Euler-Bernoulli equation which includes both mechanical and electrostaticnonlinearities in the case of a capacitive doubly clamped beam is presented. This purelyanalytical model captures all the nonlinear phenomena present in NEMS resonatorselectrostatically actuated including bistability, multistability which can lead to severalphysical limitations such as noise mixing, frequency stability deterioration as well asdynamic pull-in. Moreover, close-form expressions of the critical amplitudes and pull-indomain initiation amplitude are provided which can potentially serve for NEMS designers asquick design rules.