La part des matériaux composites dans les structures aéronautiques devient de plus en plus importante et devrait se situer aux alentours de 50 % d'ici 2012. Dans la phase d'assemblage des tronçons d'avions, les centres de production aéronautiques actuels utilisent des Unités de Perçage Automatique (UPA) pneumatiques positionnés manuellement sur un outillage dédié pour le perçage axial des panneaux de structure. Les panneaux à percer sont des assemblages de matériaux composites, aluminium et titane. En particulier, le perçage de matériaux composites (hétérogène et anisotrope) peut engendrer l'apparition de défauts, tels que le délaminage ou l'écaillage, influençant directement les propriétés mécaniques. Les paramètres de coupe caractéristiques et l'usure des outils nécessitent d'être suivis et maîtrisés pour diminuer l'apparition de ces défauts macro et microgéométriques. Cet article présente donc la problématique du perçage de panneaux multi-matériaux, en rappelant les spécificités des matériaux usinés, les moyens de générations actuels ainsi que les modes de dégradations rencontrés. Un point sera fait sur l'influence des paramètres d'usinage sur la qualité géométrique et les défauts de surface. Les conclusions montreront les résultats obtenus et les développements à mener dans un futur proche.

The present work considers the turbulent air flow inside an annular high speed rotor-stator cavity opened to the atmosphere at the periphery, where the pre-swirl ratio of the fluid is low. The interdisk spacing is sufficiently large so that the boundary layers developed on each disk are separated and the flow belongs to the regime IV of Daily and Nece (ASME J. Basic Eng. 82 (1960) 217–232). In such a system, the solid body rotation of the core predicted by Batchelor (J. Mech. Appl. Math. 4 (1951) 29–41) in case of infinite disks is not always observed: the flow behaviour in the whole interdisk spacing is governed by the level of the pre-swirl velocity of the fluid which is closely linked to the peripheral geometry (Debuchy et al., Int. J. Rotating Machinery, (2007)). In the first part of the paper, experimental results performed by hot-wire probes introduced through the stator including mean radial and tangential velocity components, as well as three turbulent correlations, are presented for several peripheral boundary conditions leading to the same value of the pre-swirl ratio. In the second part, comparisons between experiments, numerical and analytical results are provided. The numerical approach is based on the Reynolds Stress Modeling (RSM) developed by Elena and Schiestel (Int. J. Heat Fluid Flow 17 (1996) 283–289). A good agreement between the different approaches is obtained for the mean and turbulent fields and especially for the distribution of the core swirl ratio.

La séparation de la phase huileuse à partir d'effluents huile-eau comportant des émulsions secondaires est très difficile par les procédés traditionnels et nécessite l'emploi de l'ultrafiltration. Des émulsions huile-eau contenant 2 et 4 % d'huile ont été ultrafiltrées à l'aide d'un prototype de filtration dynamique comportant un disque tournant à grande vitesse près d'une membrane de 50 kDa de 190 cm2 de surface. La turbidité de la solution initiale à 4 % (21 900 NTU) chute à 2 NTU dans le perméat après 15 min de filtration, ce qui représente une rétention de 99,5 % de l'huile. Les flux de perméat augmentent avec la vitesse de rotation du disque pour atteindre environ 90 l.h-1.m-2 à 2000 tr.min-1 à 24 °C. L'ajout d'ailettes de 6 mm de hauteur permet de doubler le flux de perméat à la même vitesse. Les microémulsions sont constituées de 4 à 20 % de cyclohexane, 0,5 à 16 % de 2-butanol et de 0,25 à 8 % de dodecyl sulfate de sodium (SDS) dans de l'eau pure. Les flux de perméat, très élevés (200 l.h-1.m-2) à 2000 tr.min-1 avec un disque lisse pour la solution la plus diluée chutent à 15 l.h-1.m-2 pour la plus concentrée. L'ajout d'ailettes permet de doubler ces flux.

La modélisation de l'usure superficielle nécessite la mise en place d'un outil numérique de calcul des puissances dissipées lors du roulement sous charge de galets sur des pistes de came [1]. Compte tenu des dispersions relatives aux défauts géométriques et aux paramètres mal connus comme le coefficient de frottement par exemple, l'étude s'oriente vers une approche probabiliste [2] où le nombr e de cas simulés devient très grand et hors d'atteinte en des temps de calcul CPU raisonnables par une approche exacte. L'utilisation d'une approche simplifiée est alors nécessaire pour le calcul de la puissance dissipée dans le cas non hertzien généré par l'usure. Pour quantifier la précision (ou l'incertitude) d'une telle approche, nous présentons et discutons les choix qui ont été réalisés ; en particulier, l'approche simplifiée proposée par Kalker [3] est comparée avec les solutions de l'approche exacte, puis nous proposons des améliorations aux approches semi-hertziennes déjà utilisées dans [4] ou [5].

This paper is devoted to the description of a new full MLS SPH modelisation of rupture of thin shell filled with fluid and the prediction on consecutive fluid loss though the fracture. The paper first presents an efficient and controlled model for elastoplastic rupture of thin shells by a single layer of SPH balls. The proposed model controls as well static and dynamic instabilities using two basic ingredients: additional stress points to control the hourglass like instabilities and extend the Monhagan viscosity control method to shear and bending components of the generalized efforts. The fluid is modeled using standard SPH fluid model. The interaction is modeled using pin-balls method which is very natural in this type of formulation. Application examples are presented.

The use of composite laminates is increasing in these days due to desired directional properties and low densities in comparison of metals. Delamination is a major source of failure in composite laminates where a crack like entity can initiate and propagate between different layers of composite laminates under given loading conditions. Damage mechanics based theories are employed to simulate the delamination phenomena between composite laminates. These damage models are inherently local and can cause the concentration of stresses around the crack tip. In the present study integral type non-local damage formulation is proposed to avoid the localization problem associated to damage formulation. A comprehensive study is carried out for the selection of different non-local variables. Finite element simulations based on proposed non-local damage models and classical local damage model are performed and results are compared with available experimental data for UD IMS/924 Carbon/fiber epoxy composite laminate.

An algorithm of operational modal analysis in time domain is presented for thediscrimination between structural modes, harmonic excitations and spurious modes. Aclassification factor is developed and based on the construction of the damped signal- to-noise ratio which separates the physical modes from the numerical ones. The harmonicexcitations are then recognized at very high power indexes and zero damping. Furthermore, after the modal discrimination, the harmonic spectrum and the estimated frequency responsefunction can be derived. The method can be updated in the time domain either by a shorttime sliding window or a recursive technique to provide the modal monitoring in bothstationary and non-stationary process. An industrial application on a hydro-electricalgenerator is discussed to identify the machine frequencies under operational workingconditions.

This study presents a new simple lumped model for analyzing the thermally induced seizure of fully lubricated eccentric circumferential groove journal bearings (CGJB). The model represents a significant upgrade of a previous seizure lumped model developed by Pascovici & Kucinschi, in 2002. Main upgrades consist of: eccentric operation, introduction of hydrodynamic lubricant flow rate component, and evaluation of friction power losses based on the short bearing theory with Barwell’s hypothesis on the divergent zone. The viscosity–temperature variation is replaced with Tipei’s viscosity-clearance relationship, as in the previous model. The decrease of viscosity, as a consequence of temperature increase, does not always limit clearance loss, and the seizure process ends with a concentric journal-bushing merged system. Although CGJB’s are less sensitive to this seizure mechanism, the threat still exists and must be avoided. As a result, bearing designers and users should check whether if several safe-operation criteria are met. A recent Institute Pprime experiment on CGJB’s provided a base set of operating parameters, from which numerical simulations have been performed to determine in what conditions the CGJB fails, under thermally induced seizure.

Dans le processus de fabrication par enlèvement de matière sur machines-outils, l'apparition des vibrations est inévitable. Dans les situations où l'amplitude dépasse les limites de la précision dimensionnelle, le phénomène vibratoire est préjudiciable. L'étude expérimentale présentée ici permet d'identifier les principaux paramètres relatifs au comportement dynamique du système usinant. La localisation des déplacements de la pointe de l'outil dans un plan est démontrée grâce aux résultats expérimentaux. L'existence de ce plan et les corrélations avec les caractéristiques élastiques du système usinant permettent de simplifier le modèle dynamique 3D. Avec l'approche expérimentale réalisée, un modèle dynamique est proposé en concordance avec les résultats expérimentaux.

In this paper, an analytical approach is proposed to model chip formation in the case ofturning process. Numerical simulations of chip genesis are performed in order to fitefficiently the proposed analy- tical model. In particular, cutting edge radius influence,temperature and internal stresses distribution are studied using finite element modelling.Numerical model setting is made with experimental and literature data using forces andchip thickness.

Une double enceinte climatique conçue sur le principe de la méthode de la boîte chaudegardée a été mise au point afin d’étudier en laboratoire le comportement thermique ethygrique de différentes configurations de parois en bois. Le dispositif est dotéd’équipements de régulation lui permettant de reproduire des sollicitations climatiquesstationnaires ou dynamiques proches d’un climat réel et de suivre l’évolution des profilsde température et d’humidité dans les parois testées. Dans ce travail, le suivi ducomportement hygrothermique d’une paroi en ossature bois comportant des matériauxhygroscopiques et soumises à différentes conditions est rapporté. Les résultats ont montréque la capacité du dispositif à bien reproduire les conditions imposées a permis demesurer les propriétés importantes de parois en régime permanent (résistance thermique) ettransitoire (amortissement et déphasage thermique). Les essais réalisés ont également misen évidence le comportement hygroscopique des matériaux testés.

In recent decades, problems related to the squeeze of fluid films in the presence of aporous medium draw attention of researchers and are the subject of many applied studiesfor industry and biomechanics. Our concerns in this paper are the numerical simulation ofthe viscous shear stresses effects on the fluid film characteristics between two discswith one porous. This study is based on the coupling, at the fluid film-porous discinterface, of the Darcy-Brinkman equations in the porous medium and the modified Reynoldsequation describing the flow in the fluid film. The system of equations obtained isdiscretized by the means of finite differences method and solved numerically using thetechnique of Successive Over-Relaxation (SOR). The results show that the viscous sheareffects increase the radial and the axial fluid film velocities as well as the squeezefilm velocity but decrease the response time. Moreover, these effects are enlarged forsmaller viscous shear parameter and for smaller fluid film thickness.

Cavitation erosion is a current wear type in hydraulic turbines, on pump impellers, on ship propellers, valves, heat-exchanger tubes and other hydraulic structures in contact with high-velocity liquids subjected to pressure changes. Much effort has been done to optimise the design and operating parameters in order to avoid cavitation wear for the equipment. But, this type of wear has also been observed in mechanical devices which such as plain bearings, seals, orifices in which fluid goes through severe restrictions. Cavitation damage may occur when surfaces in contact with fluid are subjected to vibrations, e.g.: water-cooled Diesel-engine cylinder liner. Sometimes, cavitation can initiate other types of wear such as adhesion or abrasion because of wear particles produced at the friction interface, making the failure analysis more difficult.Cavitation wear mechanisms were showed through three industrial examples. The first one is a hydrodynamic plain bearing which was heavily damaged by cavitation wear due to flow instability. The second one concerns a low speed translating plain bearing. Firstly, abrasion traces were observed on the contact surface. But a deeper study showed that solid particles produced by cavitation wear had been the main cause of the abrasion wear. The last example presents several cases of cavitation damage in oil lubricated plain bearing observed in medium/slow speed diesel engine for marine or power station applications. Because of fluctuation of radial force from crankshaft and instability of lubricant flow, variation of oil pressure can be sufficient to produce bubble inception, collapse and microjet formation process.

La simulation numérique joue un rôle croissant dans la conception des navires à côté des ou associée aux approches réglementaires édictées par les sociétés de classification. Ce texte, non exhaustif, veut illustrer les tendances actuelles et les difficultés rencontrées dans l'approche simulation numérique. Celles-ci sont de plusieurs ordres : problèmes de modélisation liés à la taille des structures et à la nécessité d'une approche multi-échelle, problèmes de méthodologie liés à la complexité de l'environnement, problèmes de couplage fluides structures. Enfin ces études doivent se dérouler dans un temps extrêmement court, avec des informations qui se précisent au cours du déroulement de l'étude, informations entachées d'incertitudes et de dispersions. Les grands codes éléments finis généralistes disponibles sur le marché sont en général utilisés, associés à des outils et développements spécifiques si nécessaire, dans des chaînes de calcul plus ou moins intégrées. Les situations extrêmes ou accidentelles font l'objet d'études assez sophistiquées, nécessitant un travail d'interprétation important.

Impact wear can be defined as the wear of a solid surface due to percussion, which is arepetitive exposure to dynamic contact by another solid body. It has not been studied asextensive as other wear mechanisms and the information on causes and actual data is quitescarce. In the present work an experimental wear tester is designed and used to study theimpact wear phenomenon experimentally. It creates the consecutive impacts between ballsand flat plate. The different ball sizes (up to 30 mm), impact velocities (up to 6m.s-1) and impactangle (10°− 80°) can be tried by tester. The mass lossof plate is measured as the wear parameter. These measurements for different specimensrevealed some aspects of their wearing behaviors. The obtained results could be helpful inoptimal practical designs and explanation of some phenomena associated with impact wear.

In this paper, we study the phenomena of thermo-chemical imbalance in a reactive mono-dimensional flow composed of a mixture of (79% nitrogen N2 and 21% oxygen O2). We are interested in modeling the physicochemical process that may be encountered in hypersonic flows, as vibrational excitation, dissociation and ionization, also the formation of chemical species to higher temperatures behind a detached strong shock. We put a special emphasis on vibrational relaxation model of CVD coupling. At these high temperatures, collisions electrons-atoms become very effective, taking in account the radiation that requires knowledge and modeling of all physicochemical processes (collisional and radiative). The mathematical model of flows at the atmospheric reentry is governed by Euler stationary equations coupled with the chemical kinetics and the radiative transfer equations. Our computational code is based on the finite differences method that used to discretize and resolve the obtained numerical model, where an appropriate mesh is selected in the relaxation zone in order to determine the flow parameters at each grid position

Cleavage fracture anisotropy has been studied on a ferrite-bainite low alloy pipeline steel plate. Mechanical tests have been performed on smooth and notched bars taken along the three principal directions of the plate at temperatures between 20 °C and −196 °C, and analyzed by finite element calculations with an anisotropic plasticity model. The fracture mode of specimens tested along the normal direction differs from that of specimens taken along the two other directions, together with a lower critical cleavage stress. This difference seems to be related to microtexture anisotropy and might explain the sensitivity of this steel to delamination at low temperatures.

Shrink fits are low-price connections which are widely used in industry and industrialconnections. In designing shrink fits it is important to consider radial interface foroptimized performance and also to choose an accurate method of fabrication and assembling.Parts which have to be assembled are usually exposed to thermo-mechanical loads. Mode andtime duration of heat transfer have a significant effect on required hydraulic force,stress time rate of creation in parts and joint ability to withstand against externalloads. Therefore, planning a set of appropriate thermal and structural procedures hassignificant role in reducing energy consumption, optimized performance and promoting thespeed of parts assembly. Despite of the fact, few researches have been done on shrink fitoperation and design, rather than dimensional design. In this study, shrink fits arestudied in two main processes: first heating and mounting process and then backing to theambient condition. A 3D coupled thermal and structural simulation based on FEM is done oneach process through well-known Solidworks Premium. To evaluate the accuracy, exactanalytical solution of two steel rings shrink fit is compared with the approach outcomes.Results of validated method are used for choosing the most optimum sub processes of shrinkfit fabrication.

The long-period components in earthquake ground motions, which attenuate gradually withdistance, can induce sloshing waves in the liquid containment tanks although they arelocated far away from the seismic source. The resulting sloshing waves generate additionalforces impacting the wall and roof of the tanks and may cause extensive damage on the tankstructure. Numerous examples of tank damages due to sloshing of fluid have been observedduring many earthquakes. Nevertheless, the effect of sloshing is usually primitivelyconsidered in most of the seismic design codes of tanks. On the other hand, the derivationof an analytical solution for the sloshing response of a liquid storage tank subjected toharmonic excitation includes many assumptions and simplifications. Most of the analyticalsolutions in the recent literature assumed the containing liquid to be invicid,incompressible and irrotational, and the tank structure to be an isotropic elastic platewith uniform stiffness, mass and thickness. Even though, experimental works are necessaryto study the actual behavior of the system, they are time consuming, very costly andperformed only for specific boundary and excitation conditions. However, appropriatenumerical simulation using fluid structure interaction techniques can be used to predictthe hydrodynamic forces due to the high-speed impacts of sloshing liquid on a tank walland roof. These simulations can reduce the number of experimental tests. The nonlinearfinite element techniques with either Lagrangian and/or Eulerian formulations may beemployed as a numerical method to model sloshing problems. But, most of the Lagrangianformulations used to solve such problems have failed due to high mesh distortion of thefluid. The arbitrary Lagrangian Eulerian techniques are capable of keeping mesh integrityduring the motion of the tank. In this study, an explicit nonlinear finite elementanalysis method with ALE algorithm is developed and sloshing phenomenon is analyzed. Theanalysis capabilities of the method are explained on a technical level. Although, thedeveloped numerical procedure is applicable to deformable structures, the accuracy of themethod is validated with the existing analytical formulation derived from potential flowtheory as well as the experimental data carried out on rigid tanks when subjected toharmonic and earthquake ground motions. High consistency between numerical andexperimental results in terms of peak level timing, shape and amplitude of sloshing wavesis obtained not only for non-resonant excitation but also for resonant frequencymotion.