François Louf

Analytical Approach for Mechanical Resonance Frequencies of High-Speed Machines

This paper presents an analytical approach for the determination of the mechanical eigenfrequencies of an electrical machine stator. This model is based on the calculation and the minimization of Rayleighs quotient. It uses an energetic approach from beam theory applied to stators with a Timoshenko kinematic model. This model is very specially adapted for stators with a thick yoke as the high-speed machines. This approach is validated by finite-element simulations and experimental measurements.

Low Computational-cost determination of Vibrational Behavior: Application to Five-phase Flux-Switching Permanent-Magnet Motor

This paper presents a study of the vibrational behavior of a five-phase flux-switching permanent-magnet (FSPM) motor with a low computational-cost. Magnetic and mechanical models are sequentially considered to predict the acoustic noise generated by the structure. After the description of the motor, stress due to magnetic source is obtained by finite element simulations, particularly at the interface between the air-gap and the stator. Then, the most significant eigenmodes of the stator structure, obtained by finite element simulation and by experimental measurements, are presented to determine how the motor is then excited. Finally, from the knowledge of the magnetic excitations and the modal basis, an analytical model is proposed to determine the dynamic behavior of the stator outer surface, instead of a costly dynamic finite element analysis. This model is applied to the five-phase FSPM motor and validated by experimental measurements.

Modeling and Optimization of a Linear Actuator for a Two-Stage Valve Tappet in an Automotive Engine

This paper focuses on external volume and power consumption optimization of an electromagnetic actuator used to move pins in a two-stage variable valve lift component within an automotive engine. This system is designed to change the valve lift at each engine cycle. To achieve this optimization, three models were developed: a magnetic, an electric, and a dynamic model. They evaluate the performances of a given actuator, check if it fulfills its role, and define the springs needed for the displacement of the pins. The optimization is made using a particle swarm method and can reduce the volume up to 20% with the same electric consumption.

Analytical approach for magnetic and acoustic modeling of flux-switching permanent-magnet motors: Application to geometrical optimization

This paper deals with multi-physics analytical approach for magnetic, mechanical and acoustic modeling of Flux-Switching Permanent-Magnet (FSPM) motors. The major advantage of this model is the low-computation cost compared to finite element simulation. It contains: - Magnetic model obtained by formal resolution of magneto-static Maxwell equation by Fourier series decomposition. - Mechanical and stator eigenfrequencies model based on energy method (Hamilton principle) with computation and minimization of Rayleigh quotient. - Vibratory and acoustic model obtained on computating the stator deformation through dynamic equilibrium equation in modal base formed by stator eigenmodes. First, analytical model will be introduced then validated by finite element simulation and/or by experimental measurements. Second, geometrical optimization will be presented for 3-phase 12/10 and 5-phase 20/18 FSPM motors.

An updating method for structural dynamics models with unknown excitations

This paper presents an extension of the Constitutive Relation Error (CRE) updating method to complex industrial structures, such as space launchers, for which tests carried out in the functional context can provide significant amounts of information. Indeed, since several sources of excitation are involved simultaneously, a flight test can be viewed as a multiple test. However, there is a serious difficulty in that these sources of excitation are partially unknown. The CRE updating method enables one to obtain an estimate of these excitations. We present a first application of the method using a very simple finite element model of the Ariane V launcher along with measurements performed at the end of an atmospheric flight.

Un estimateur d’erreur en relation de comportement pour les problèmes d’impact

Cet article présente un estimateur d’erreur en vue de contrôler la qualité des simulations éléments finis pour les problèmes d’impact. Nous nous intéressons à un problème de contact avec frottement de Coulomb en élastodynamique sous l’hypothèse des petites déformations. La mesure d’erreur proposée est une mesure d’erreur en relation de comportement basée sur une extension de l’erreur de Drucker développée pour les calculs de dynamique rapide, dans laquelle les relations de contact-frottement sont introduites au moyen d’un bipotentiel. Nous présentons dans cet article le principe de construction de cette mesure d’erreur ainsi que sa mise en oeuvre sur des exemples simples.

Electrical machine optimization using a kriging predictor

This paper presents an optimization on a surface response created by kriging. It focuses on the predictor creation, its refinement and on the advantage of such a method over a direct optimization. In order to reduce the number of evaluations needed to construct the predictor, an adaptive method for the trial site selection is introduced as an improvement for the Latin Hypercubic Sample (LHS) algorithm. The method is applied to the optimization of a flux switching synchronous machine and the results are illustrated for two and four parameters.

Variability of a Bolted Assembly Through an Experimental Modal Analysis

Industrial structures are mainly assemblies with complex geometries and non-linear characteristics. Friction and joint preload added to fabrication imperfections lead to a substantial gap between numerical models and real structures. In order to develop accurate generic models, it is then necessary to quantify the behavior variability, especially the one related to the joint conditions. The first part of this paper describes the iterative sizing procedure of an academic assembly which characteristics may vary depending on several input variables (e.g. value of the bolt torque, number and position of preloaded bolts, etc.). The properties of the bolted joint were optimized in order to satisfy a set of conditions in terms of tangential slipping, normal displacement and maximum stress level. The second part concerns the experimental modal analysis of the assembly. The main purpose is to characterize the relationship that exists between the input variables and the measured eigenfrequencies and modal damping of the assembly.