Des forts courants pulsés d'une durée comprise entre 1,25 et 20 ms ont été appliqués à T = 77 K à des conducteurs Bi:2223 texturés par forgeage à chaud ou frittés. Pour des pulses de courant ${I>I_{\rm c}}$, deux comportements sont observés : quand la température moyenne de l'échantillon est inférieur à ${T_{\rm c}}$, le matériau retourne dans l'état supraconducteur avant la fin du pulse de courant. Au dessus de ${T_{\rm c}}$, l'échantillon transite dans l'état normal après le pulse de courant à cause de l'élévation de température. Ceci induit une puissance instantanée importante appliquée au conducteur de 182 GW/m3. À I = 15Ic, des champs électriques de 400 à 1000 V/m sont obtenus sur des éléments frittés et texturés.

Une linéarisation par l'algèbre spectrale a été développée dans une méthode de modulation thermique pour mesurer la distribution des charges électriques dans les isolants solides. La modélisation présentée permet d'évaluer les performances et les limites tant numériques que physiques de la méthode de mesure.

This paper describes a numerical method to determine the losses in a solid superconductor plunged in transverse magnetic field or in transport current. The model which is based on the Bean critical state, shows that the hysteretic phenomena are taken in account.

Highly transparent undoped and indium doped ZnO thin films have been grown on glass substrates by using the spray pyrolysis process. Conditions of preparation have been optimized to get good quality and reproducible films with required properties. Polycrystalline films with an hexagonal Wurtzite-type structure were easily obtained under the optimum spraying conditions. Both of samples have shown high transmission coefficient in the visible and infrared wavelength range with sharp absorption edge around 380 nm which closely corresponds to the intrinsic band-gap of ZnO (3.2 eV). Orientation and crystallites size were remarkably modified by deposition temperature and indium doping.

This paper proposes a methodology for the sizing of static converters which the structure is given. The sizing problem is translated into a constraint optimisation problem, and the constrained model of the converter is formulated symbolically in order to proceed the optimisation of its parameters. In the paper, the necessary symbolical treatments are presented to obtain a generic formulation of the sizing constraints from the definition of the converter structure and an analysis of its operating cycle.

In indirect band gap ${\rm GaAs}_{1-x}{\rm P}_x$ alloys, the Nx levels of the excitons bound to nitrogen have been resonantly excited. The detection of the photoluminescence (PL) at energies greater than the excitation energy (Anti-Stokes Luminescence (ASL)) shows a new PL band which appears at temperatures lower than 40 K. Its mean energy 2.205 eV is nearly independent of the alloy composition. We investigate ASL as a function of excitation photon energy, excitation power and temperature. Moreover ASL has been time resolved.

NdBaCuO superconductors have been textured by the oxygen controlled melt growth method. The addition of platinum or Nd422 does not refine the 422 phase, but the latter improves the microstructural quality. High purity precursors are necessary to achieve high Tc and Jc (up to 56 000 A/cm2). The flux pinning mechanisms are discussed in terms of oxygen deficiency, Nd-Ba substitution and softening of the vortex lattice.

In this work the influence of solid particles concentration from suspension on the buildup evolution for a single wire from a magnetic filter working in axial configuration is presented. The buildup differential equations are solved and the surface contour of the buildup at different moments is obtained. It is evidenced the existence of a suspension concentration for which the radial deposit extension is maximum.

Nous avons synthétisé par alternance de dépôts électrolytiques et de traitements thermiques des couches de Bi 2212 de $3,5~\mu\rm m$ d'épaisseur sur les deux faces de rubans d'argent de $\rm 50~\mu m$ d'épaisseur. Grâce à une succession de fusions partielles, la composition moyenne de nos couches s'homogénéise bien que les précurseurs soient déposés de façon séquentielle. Comme dans le procédé classique connu sous le vocable "partial melt growth", nous fondons nos précurseurs mais vue leur faible épaisseur au lieu de les refroidir très lentement, nous ne les maintenons que brièvement dans l'état fondu, puis nous les trempons jusqu'à la température ambiante. Nous synthétisons alors la phase Bi 2212 sans repasser au-dessus de son point de fusion. Afin d'optimiser le procédé, nous avons caractérisé la microstructure de nos échantillons à chacune des étapes de leur élaboration. Nous présentons dans cet article la comparaison entre le procédé classique de fusion partielle et le nôtre, ainsi que la corrélation entre les traitements thermiques et les propriétés électriques. Au stade actuel de l'optimisation, nous préparons de façon reproductible des échantillons présentant une densité de courant critique de transport d'environ 30 000 A/cm2 à 77 K et 300 000 A/cm2 à 4,2 K, ce qui correspond à des courants de transport de respectivement 6,3 et 63 Ampères pour des échantillons de 3 mm de large, avec un rapport substrat sur supra de 7. Nous avons de plus vérifié que le courant critique était proportionnel à la largeur de l'échantillon, conduisant à un courant critique de 42 A à 77 K et 100 A à 63 K pour les échantillons les plus larges que nous ayons testés (20 mm).

We present, in this paper, a study of the thermomechanical behaviour of a single crystal in transformation plasticity using Taylor's approach, and we neglect the internal stress associated to the interactions between martensitic variants. We represent those interactions by a matrix deduced from a thermodynamical potential's optimization associated to the transformation. We also point out the contribution of the variant's reorientation mechanism to the thermomechanical behaviour.

En s'appuyant sur les résultats analytiques obtenus dans le cas élastique, nous établissons les formules permettant de calculer le champ des contraintes thermiques induit par une irradiation en balayage par un faisceau laser ayant une répartition spatiale d'énergie quelconque. La surface du matériau est considérée semi infinie et la température induite quasi stationnaire. Nous introduisons ensuite une notion de source virtuelle pour que les résultats théoriques puissent simuler un cas pratique.

Thin films of ${\rm Cu_2Mo_6S_8}$ have been in-situ deposited by pulsed laser deposition on R-plane ${\rm Al_2O_3}$ on which an epitaxial growth has been achieved for the first time, with the $(100)_R$ orientation, as shown by X-ray diffraction. The superconducting transition temperature has been determined from DC resistive and AC susceptibility measurements. The critical temperature Tc is typically between 9.5 and 10.3 K and the transitions are narrow: the full width at half maximum of the inductive signal $\chi ''$ ranges between 0.1 and 0.5 K. The critical current densities have been obtained by standard four probe electrical measurements: in zero field a value larger than $10^8\; {\rm A}/{\rm m}^2$ has been measured at only 1 K from the end of the transition, and values typically larger than $10^7\;{\rm A}/{\rm m}^2$ at 4.2 K in an applied magnetic field of 7 T have been observed. The upper critical magnetic field, Bc2, has been determined from the critical current measurements performed at 4.2 K in an applied magnetic field, for two samples. The experimental values of 11 and 12.5 T at 4.2 K are close to those reported on bulk samples (12.5 T). For the first time to our knowledge a surface resistance $(R_{\rm s})$ measurement has been carried out on a Chevrel phase-type compound: a value of 4.5 mΩ at 10 GHz and 4.2 K has been obtained.

In this paper, the problem of electromagnetic scattering from a 3D system of spheres is considered and an iterative solution that accounts for multiple scattering is proposed. The Mie formalism used for a single sphere is extended to account for multiple scattered fields between several particles. The translational addition theorems for spherical wave functions are used to express the electromagnetic field scattered by a sphere Si in terms of an incident field for a sphere Sk in a spherical coordinates system attached to the sphere Sk. In this work, the numerical convergence of the method is discussed and associated computational times are given. Numerical computations including Radar Cross Section (RCS) and radiation patterns for various 3D configurations are presented. Some of them are compared with free-space measurements made in the 8 to 100 GHz frequency band using vectorial network analyzers.

The shape of an inaccessible boundary between two different media is identified by applying a dc field. The electric potential is described by Fredholm's integral equation of the first kind. Numerical processing leads to a system of nonlinear and ill-conditioned algebraic equations. Their solution causes numerical problems. A new algorithm based on the Brouwer's fixed point theorem is proposed as a solution.

An optical fibre sensor to measure weak partial pressures of carbon dioxide has been designed and realized. This sensor is based on direct molecular absorption in the near infrared at 4.3 μm. It is an extrinsic and amplitude modulation type sensor. In the linear region between 0 and 2000 μbar, the calibration curve that represents the transmitted power versus the carbon dioxide partial pressure in air shows a sensitivity of 20 nW/mbar with a minimum detectable pressure of 25 μbar. This article describes a temperature sensitivity analysis for all optoelectronic components and proposes a calculation method to distinguish two successive levels of partial pressure in a surrounding atmosphere where the temperature can vary. The experimental results show that without using a reference signal, the maximum variation of temperature is ±1.4 °C. The use of a reference signal combined with thermoelectric cooling of the photodetector tolerates a temperature variation of ±25 °C.

In this paper, an error estimator is used to control the quality of finite element solution in linear magnetostatic. This method, based on the works of P. Ladevèze, consists in constructing a complementary admissible field from the one calculated by the finite element method. The quality of finite element solution is measured by comparing the couple of admissible solutions. This method has been developed in 2D in the case of scalar and vector potential formulations. To show the effectiveness of this error estimator, we present three simple examples, one of them having a known exact solution.

Lors de la conception de convertisseurs statiques en électronique de puissance, le besoin se fait grandissant de caractériser les connexions. Cet article présente le modèle adopté pour le câblage permettant de prendre en considération ses caractéristiques électriques ainsi que les moyens mis en œuvre pour l'obtenir. Deux approches permettant de calculer des éléments de circuit équivalent sont comparées : la méthode des éléments finis, méthode numérique connue dans d'autres contextes pour sa précision et sa fiabilité, et la méthode PEEC (Partial Element Equivalent Circuit), qui est une méthode semi-analytique, particulièrement bien adaptée à ce genre de problème. Une confrontation de ces deux méthodes avec des mesures pour un élément de circuit simple démontre la pertinence de l'approche semi-analytique, et la possibilité d'utiliser une méthode éléments finis, qui demeure cependant coûteuse en temps de calcul. Le calcul de schémas électriques équivalents correspondant à un exemple de connectique plus réaliste est ensuite réalisé par la méthode semi-analytique. La comparaison avec des résultats expérimentaux est encourageante.

Organic field-effect transistors (OFETs) usually operate in the accumulation mode, where the biasing of the gate induces the charging of the insulator-semiconductor interface; the bias is negative in the most common case of p-type semiconductors. We show here that the application of a positive gate bias leads to the formation of a depletion layer, and a subsequent decrease of the drain current. We develop an analytical model for this depletion mode of the OFET. It is shown that measurements in the depletion mode give access to parameters such as the doping level and the density of trap levels. The model is applied to data obtained on sexithiophene (6T) film and single crystal OFETs. A substantial amount of traps is found in unsubstituted 6T, whereas the dihexyl substituted 6T is practically trap free. 6T single crystals are characterized by a very low doping level, which can be related to their very high purity. The possible use of the depletion mode to increase the on-off current ratio of OFETs is discussed.

This paper presents a non-linear model of the induction machine which takes in consideration the influence of the saturation on its behaviour. The approach, based on the Park's two-phase model, is established using a finite-element flux calculation program. This program allows to validate hypothesis formulated in order to simplify flux-current relationships. A comparison between the dynamic response of the model and the response obtained by a general model based on the coupled electro-magnetic equations permits to validate it.

We present experimental results on low frequency noise in amorphous PIN diodes under forward bias by pulses and by direct current. We have developed a new method using a differential amplifier to measure the noise in pulse regime. The study separates the white noise and the 1/f noise depending on current levels. The white noise can be evaluated by the shot noise through the various interfaces damped by their dynamic impedance. The excess noise shows a behaviour in f−1/2, due to defects in the intrinsic zone and a current dependence in I3/2 , current limited by the NN+ junction.