A new varistor system of SnO2-Bi2O3-Nb2O5 was reported in this paper. Theelectrical field-current density characteristics of this system were investigated by doping different amounts of Bi2O3 and sintering the samples at various temperatures. It is found thatadding 0.75 mol% Bi2O3 to Nb-doped SnO2 ceramic resulted in maximum nonlinear coefficientand breakdown voltage with α = 14 and E 0.5 = 19 525 V/cm. To improve the density as well asthe nonlinearity of this system, different amounts of Co2O3 were added. The optimal conditionsfor the best nonlinearity were 1300 °C with 0.03 mol% Co2O3 addition. Deviation fromthis doping content, toward either higher or lower Co2O3 content, causes the deterioration of I−Vcharacteristics. It can be concluded that the incorporation of cobalt oxides into SnO2-based varistors improves the nonlinearity in the low and intermediate current density regionsbecause of the increased barrier height $(\Phi_{\rm B})$ . The experimental results were explainedwith the defect barrier model for SnO2-based varistors.

The transmittance (T) of thin films of CdI2, prepared by thermal evaporation technique on quartz substrates,have been measured over the wavelength range 200−900 nm. Fromanalysis of the transmittance data, the optical constants, therefractive index (n) and the extinction coefficient (k), havebeen studied. Analysis of the refractive index (n) yields a lowfrequency dielectric constant, average oscillator strength and average oscillator energy. From analysis of the absorptioncoefficient (α), the fundamental absorption edge can bedetermined. Both allowed direct transitions and allowed indirecttransitions are observed. The composition of films is checkedusing energy dispersive X-ray (EDX) spectroscopy technique. X-raydiffraction (XRD) measurements showed that the CdI2 filmsevaporated at room temperature substrates were characterized by a polycrystalline form. At large thicknesses the films indicated the inhomogeneity. The effect of annealing temperature (up to 523 K) onthe film properties has been studied.

Thin amorphous and polycrystalline tungsten oxide films have been prepared by Chemical Vapor Deposition (CVD) from metallorganic precursor − tungsten hexacarbonyl − at atmospheric pressure. The dependence of the composition and the structure of tungsten oxide films on the technological conditions has been investigated by XPS, XRD, DTA-TGA and Raman spectroscopy. As a result it has been established that: at high values of the flow-rates of the reaction gases amorphous films of very low density have been obtained; in the XPS spectra of the understoichiometric WO3−y (0 < y < 0.3) films besides W6+, also W5+ and W4+ stateshave been observed. First to observe in the Raman spectra of amorphous CVD-WO3 films is the band at ~ 950 cm−1,characteristic for terminalW6+ = O bonds in result of the presence of structural water. Theexistence of structural water in the amorphous material has been established by thermal analyze, also.

In this paper, a claw pole permanent magnet machine is studied using a 3D scalar potential formulation. First, the discretisation of Maxwell's equations using Whitney's elements is presented. Especially, the method to take the winding current distribution into account by means of source fields is detailed. Then, a numerical model of the studied machine is presented. Simulation results (inductance, EMF, torque) are discussedby comparison with experimental ones.

This paper presents a modelling of a Transverse Vernier Hybrid Reluctance Motor (TVHRM). The geometry of the magnetic circuit of this latter is especially 3D. The use of the 3D finite element method to do the analysis and to predetermine the motor characteristics is more accurate, but with an important computational time which makes unpractical a parametric study. An analytical model which will allow a rapid investigation of the motor characteristics is a good compromise between accurate results and a weakly simulation time. This is done through two comparable models based both on the equivalent reluctance network method. These analytical models are compared to the 2D finite element one. Finally the predicted characteristics are estimated with the practical test results.

We study gallium arsenide (GaAs) high voltage photoconductive switches triggered by a 30 ps neodymium: yttrium aluminium garnet (Nd:YAG) laser, i.e. using photons of energy smaller than the band gap. We measure optical absorption at Brewster incidence under optical pulse excitation and determine the extrinsic one photon and intrinsic two photon absorption coefficients. Analyzing the photoconductive resistance under different power densities of laser radiation, and using the absorption data, we demonstrate that only about 20% of the photons absorbed by the extrinsic process are converted into free electrons. We conclude that high efficiencies can only be obtained by using two photon absorption, which is feasible with ultrafast lasers and focussed beams.

Within the framework of the study of the aircraft structural material lightning, we present a work concerning the heating of metal sheets under the action of a moving electric arc. A 2D and 3D modeling of thermal phenomena occurring in the heated electrodes are used in order to study the influence of the arc root velocity and of the power and the surface power density brought by the arc roots on the electrode heating.

The remote field eddy current technique is used fordimensioning the grooves that may occur in the ferromagnetic pipes. Wepropose a method to estimate the depth and the length of corrosiongrooves from measurement of a pick-up coil signal phase atdifferent positions close to the defect. Groove dimensioningrequiresthe knowledge of the physical relation between measurements anddefect dimensions; therefore finite-element calculations areperformed to design parametric algebraic functions for modeling thephysical phenomena. Different models are possible; the choiceof this algebraic function is discussed from identification criteria. By means of new measurement formalism and twopreviously defined measurement relations, estimates of groovesizes may be given. In the first approach, algebraic function parameters andgroove dimensions are linked through a polynomial function; this approach is proved to bebetter than a second one which tries to take advantage of more physical considerations.

Knowledge of the water content at the subcellular level is important to evaluate the intracellular concentration of either diffusible or non-diffusible elements in the physiological state measured by the electron microprobe methods. Water content variations in subcellular compartments are directly related to secretion phenomena and to transmembrane exchange processes, which could be attributed to pathophysiological states. In this paper we will describe in details and compare two local water measurement methods using analytical electron microscopy. The first one is based on darkfield imaging. It is applied on freeze-dried biological cryosections; it allows indirect measurement of the water content at the subcellular level from recorded maps of darkfield intensity. The second method uses electron energy loss spectroscopy. It is applied to hydrated biological cryosections. It is based on the differences that appear in the electron energy loss spectra of macromolecular assemblies and vitrified ice in the 0−30 eV range. By a multiple least squares (MLS) fit between an experimental energy loss spectrum and reference spectra of both frozen-hydrated ice and macromolecular assemblies we can deduce directly the local water concentration in biological cryosections at the subcellular level. These two methods are applied to two test specimens: human erythrocytes in plasma, and baker's yeast (Saccharomyses Cerevisiae) cryosections. We compare the water content measurements obtained by these two methods and discuss their advantages and drawbacks.

We present a simple bias reversal technique for single electron transistors (SET) to remove fluctuations of tunneling resistance from the read-out signal at low frequencies. The gain of the device is kept constant under bias reversal by using asymmetric junction capacitances. In our Al/AlOx/Al devices with 1.2 μm island size and 100 × 100 nm2 tunnel junctions, the noise at 10 Hz is $6 \times 10^{-4} e/\sqrt{\mathrm{Hz}}$, independent of the bias modulation.