Table of contents

Recent research results pertaining to InN, GaN and AlN are reviewed, focusing on the different growth techniques of Group III-nitride crystals and epitaxial films, heterostructures and devices. The chemical and thermal stability of epitaxial nitride films is discussed in relation to the problems of deposition processes and the advantages for applications in high-power and high-temperature devices. The development of growth methods like metalorganic chemical vapour deposition and plasma-induced molecular beam epitaxy has resulted in remarkable improvements in the structural, optical and electrical properties. New developments in precursor chemistry, plasma-based nitrogen sources, substrates, the growth of nucleation layers and selective growth are covered. Deposition conditions and methods used to grow alloys for optical bandgap and lattice engineering are introduced. The review is concluded with a description of recent Group III-nitride semiconductor devices such as bright blue and white light-emitting diodes, the first blue-emitting laser, high-power transistors, and a discussion of further applications in surface acoustic wave devices and sensors.

Single-crystalline films of meta-nitroaniline (mNA) and 2-cyclo-octylamino-5-nitropyridine (COANP) 5- thick were grown between two transparent fused quartz plates on which layers of indium tin oxide had been deposited. Both linear and quadratic electro-optical effects of the crystals were examined by an ac modulation method. The figures of merit for mNA and COANP are estimated to be and , respectively. The crystals were of high quality due to the excellent protection provided by the cells and the figures of merit of the crystals did not decrease with time. It is believed that the method adopted in the paper is applicable to fabrication of electro-optical cells with other compounds if growth of those crystals from the melt is feasible. It is suggested that the electro-optical cells have excellent potential use for device fabrication because of their durability and relatively easy manipulation during crystal growth.

The effect of a thin metal coating in near-field fluorescence imaging is studied by using a model based on the plane wave decomposition, that takes into account the dynamic nature of the molecule's electronic structure, namely the modification of the fluorescence lifetime due to the near-field environment (the metal layer and the probe's presence). The electromagnetic wave crossing the metal layer generates a plasmon excitation, the effect of which is the more important when the molecules, considered as dipoles, are polarized perpendicularly to the surface and when they are in a saturated excitation state. The consequence of the plasmon excitation is a decrease in fluorescence lifetime and, for the case of molecules polarized perpendicularly to the surface, a decrease in lateral resolution. When the molecules are polarized parallel to the surface, plasmon excitation is weaker and its effect on the resolution seems to be beneficial. The interaction with the probe is also studied; its effect on imaging of molecules is weak in the presence of the metal coating.

This paper presents multi-level, multi-grid solutions for isothermal elastohydrodynamic circular contacts under pure entraining motion. The solutions are valid for quasi-static conditions and Newtonian lubricants. Multi-grid solutions allow rapid evaluation of contact pressure distributions and corresponding elastic film shapes. This allows the deployment of a large number of elements, thus improving the accuracy of the numerical solution compared with that of conventional finite difference schemes. The numerical results are compared with experimental measurements and reasonable agreement has been obtained under a wide range of operating conditions.

Stable two-molecule clusters can densely be monodispersed in films up to 10 mol% of concentration, by the method of quench deposition and subsequent annealing of the films. The clusters are stoichiometrically embedded in the crystal matrix. Their absorption spectrum is characterized by an excitonic peak well fitted to an asymmetrical Lorentzian function. Its peak energy (2.503 eV), line broadening (206 meV) and oscillator strength per molecule (0.14) are strongly enhanced due to the effect on the exciton of confinement. The present experiment suggests that there is a possibility of a new method for achieving well-identified, two-dimensional clusters or microcrystals at high densities.

Nanocrystalline maghemite O was prepared by a wet chemical method. The thermal stability of this material has been investigated by using differential thermal analysis (DTA) measurements and x-ray diffraction analyses. One exothermic peak without weight loss appears irreversibly on the DTA curve. It was confirmed for the first time that this exothermic peak corresponds completely to the maghemite-to-hematite (-to-) structural phase transition. The nanocrystalline maghemite O particles grow very slowly when the temperature is lower than the range of the exothermic peak; whereas, during and after the transition, the nanocrystalline hematite O particles grow rapidly with increasing temperature. Moreover, the transition temperature of nanocrystalline maghemite O is higher than that of the coarse-grained counterpart. In this paper, the authors have also suggested some preliminary physical interpretations of the experimental results.

This work has shown that isotropic single-phase materials can exhibit remanence enhancement due to exchange coupling between spins in grain boundary areas. Magnetic materials with remanence enhancement are required to have nanocrystalline structures with grain sizes comparable to the domain-wall thickness. The presence of non-magnetic phases may result in de-coupling of magnetic grains, therefore leading to an increase in coercivity but a decrease in remanence. The demagnetization processes of single-phase materials with enhanced remanence are different from those of nanocomposites consisting of hard and soft phases, in that no exchange-spring magnet behaviour was observed for single-phase ribbons of with a nanocrystalline structure. A negative deviation of the demagnetization remanence from the Wohlfarth model [15] was observed due to exchange coupling.

Spatial pattern formation in strongly driven ferromagnets has been predicted theoretically but not yet verified by experiment. We present an experimental set-up to excite the ferrimagnetic resonance of magnetic garnet films locally at high power levels. The spin precession is observed optically by means of Faraday rotation. Using special garnet films with very low saturation magnetization, we were able to observe the formation of spatial patterns.

For an annular-type electromagnetic pump of a liquid metal, the magnetic field is found in closed form by means of the Fourier transform method for two-dimensional field analysis based on an equivalent current sheet model, so that the terms contributing to the normal thrust and the end effect are analytically identified. Analytical solutions for the magnetic vector potentials are compared with numerical ones obtained from the finite difference equation, and they turn out to be in good agreement regardless of the assumption of infinite permeability used in the analytical approach. As a result of closed-form solutions, the electromagnetic body force exerted on a liquid metal flow and the efficiency of the pump are calculated without going through tedious time-consuming numerical work usually accompanied by troublesome convergence problems. The calculated results for estimating the influence of end effects on the performance of the electromagnetic pump clearly show that the end effects give rise to the obstructive force to the liquid metal flow at inlet and outlet ends of the pump. In addition, the efficiency is found to be lowered due to end effects, especially when the speed of the flow is high.

We analyse how the frequency dependence of the dielectric permittivity affects the electromagnetic field radiated by a point charge uniformly moving in a medium. It turns out that a moving charge radiates at every velocity. We study the space distribution of the electromagnetic field and show that its oscillations are due to the time-dependent polarization of the medium induced by the moving charge. Spectral distributions of the radiated energy and the photon number are given. Consequences arising from a choice of polarization different from the usual one are discussed. An analysis of the Kramers-Kronig dispersion relations for the treated problem is given. This consideration may be useful for the experiments recently discussed in the literature.

We report experiments to study the effects of the laser-pulse duration, angle of incidence and target material on x-ray generation. Metallic targets (Cu and Al) were illuminated with picosecond and sub-picosecond KrF excimer laser pulses focused to maximum intensities of with 5 ps and with 150 fs pulses. The resulting plasma produces soft x-rays in the 1 keV range. Copper targets illuminated at about with 5 ps pulses gave the highest x-ray conversion efficiency (1.6%). In addition to the soft x-rays, a hard x-ray signal ( keV) was also detected with 450 fs pulses. This is assumed to be due to the fast electrons generated by the interaction of the higher intensity sub-picosecond laser pulse with the solid target.

In this work a discharge model for the symmetrical double-sided, double-polarity, LC inversion, high-pressure (SDLC) laser discharge is reported. The model is compared with the experimental circuit. Also a new design which couples two different SDLC discharges driven by a single spark-gap is built up and experimentally compared with an extended discharge model. This new circuit design allows one to drive two different lasers, with a controllable no-jitter delay.

Pulsed gas was injected into a carbon doped helium plasma in TPD-II to investigate electron capture processes. Anomalous enhancement in C II line intensities was observed and considered to be due to the double electron capture in - collisions. The rate coefficient was estimated to be (, and a spectral distribution of C II lines, which relates to population densities of product ions, was also measured. Based on these results, C II spectral distributions in the JIPP T-IIU tokamak were observed to assess the influence of the double electron capture in collisions on the edge plasma. A new spectroscopic procedure showed that this process is one of the dominant processes in the edge plasma.

The electrode-melting rate and molten-droplet parameters during arc welding with pulsed current are considered. The discreteness of the droplet detachment is taken into account. Calculations agree well with the experimental fact that there is a range of operational parameters within which one droplet is transferred per current pulse. Variations of the droplet parameters and the melting rate within this range are calculated. It is shown that the high-frequency limit of this one-droplet-per-pulse range is characterized by the least thermal load on the weld, the least amount of fumes and the highest melting rate.

We use both experimental diagnostics and computational modelling to characterize the electrical properties and vacuum ultraviolet emission of positive column discharges in both pure xenon and mixtures of xenon with other rare gases. Discharge conditions include xenon pressures of 10-100 mTorr at room temperature, discharge currents of 100-500 mA and positive column diameters of 1.1-5 cm. Our primary interest is the efficient generation of large quantities of ultraviolet radiation for use in fluorescent light sources. We therefore focus on the total output of atomic xenon resonance radiation at 147 nm, reported in units of watts of 147 nm radiation per unit length of positive column, and also the efficiency by which electrical input power is converted into 147 nm radiation. Over the range of discharge conditions considered, we observe outputs in excess of of discharge length and efficiencies in excess of 80%, but we also find that there is a strong tradeoff between efficiency and output, and that these two figures of merit cannot be simultaneously maximized. A change in the behaviour of the discharge is observed to occur near 25 mTorr in 2.2 cm diameter tubes and is attributed to the transition from a discharge sustained by multistep electron-impact ionization to one sustained by single-step electron-impact ionization.

The pressure dependence of the cathodic current density in the case of an electrolyte and a metal cathode glow discharge operating in air and helium atmospheres was investigated. In all cases, at pressures above 100-200 mbar, violation of the similarity law was observed, which was attributed to the occurrence of the dissociative recombination of molecular ions of the applied gases (O, and ). Taking into consideration this process, the relation was deducted, which was found to be in accordance with experimental data. In the case of an electrolyte cathode this relation was found to be valid also at low pressures.

The effects of the plasma's gas-flow rate on phenomena such as the negative anode fall and the heat flux to the anode in high-intensity arcs are discussed on the basis of a numerical model of the anode's boundary layer. The modelling system consists of a rotationally symmetrical argon plasma formed between the outlet of a constrictor tube and a water-cooled flat copper anode perpendicular to the axis of the plasma flow which is directed towards the anode. The arc is operated at atmospheric pressure and at a current level of 200 A. The boundary conditions for the electron temperature and the electron number density at the anode's surface are obtained by solving the electron-energy equation and the diffusion equation at the anode's surface simultaneously with all conservation equations for the calculation domain. A diffuse anode attachment is obtained for a mass flow rate exceeding 0.2 g and a constricted anode attachment is found for rates below 0.02 g . There is a (probably unstable) transition region between these flow rates. The anode boundary layer of the diffuse attachment is strongly affected by the mass-flow rate. Increasing the mass-flow rate shifts the location of the peak of the electrical potential towards the anode while its magnitude decreases and the total heat flux to the anode increases. This is a consequence of the effects of the mass-flow rate on the axial profiles of electron and heavy-particle temperatures and of the electron number density. In contrast, the anode boundary layer of the constricted attachment seems not to be affected by the mass-flow rate.

The spark decomposition of and 50% mixtures (100 kPa) in the presence of a solid insulator struck by an electric arc is studied. The sparks were generated in a cylindrical Monel 400 cell between a stainless steel point and a stainless steel plane (gap space 0.9 mm) either under a 50 Hz ac voltage (90 J per disconnection) or by discharging a capacitor (3.59 J per spark). The gaseous by-products , , , , , , and were assayed by gas chromatography and their yields studied by varying the nature of the insulator material (Megelit, Kel-F (polytrifluorochloroethylene), Teflon, polypropylene, polyethylene, nylon) and the concentration of two additives: (0 and 0.2%) and (0 and 0.2%).

As it becomes vaporized under the effect of the discharges, the insulator produces species that can trap the fluorine atoms released from decomposition and enhance the formation of by-products. This explains the main results of this work which can be summarized as follows.

(i) Enhancement, by an order of magnitude or more, of by the insulators.

(ii) A quantity of formed in the presence of the insulator proportional to the enhancement of . The ratio between the levels of these compounds proved to be the same regardless of the initial concentration of or added to the and the intensity of the sparks.

(iii) Enhancement of the quantities of (up to fivefold), , and formed in the presence of an insulator whatever the gas sample studied.

In the presence of an insulator, the - mixture led to results rather similar to those of concerning the levels of decomposition products formed. In all cases however, i.e. in the presence or absence of insulator or added or , the mixture led to decreased formation of the main products: and .

Interference was seen to occur between the action of the insulator and that of and ; this is ascribed to reactions taking place between these molecules or their dissociation fragments.

A comparative study of ferroelectric domain configurations and surface morphology of butterfly single crystals of by the techniques of atomic force microscopy, optical microscopy and scanning electron microscopy is described. An unusual array of defects on the surface of a highly stressed region of the crystal was investigated in detail. Surface microstructural features displaying complex and distinct morphology associated with barium-rich phases were also studied.

Preparation of a new -ion-conducting two-phase composite electrolyte system , (in molar weight fraction) and studies of its ionic transport properties are reported. Nano-size ( nm) particles of insulating and insoluble were used as second-phase dispersoid particles, while a new compound, `a quenched/annealed [] mixed-system/solid solution', recently investigated in the present laboratory, was used as the first-phase host matrix in place of the conventional host salt AgI. Various routes of sample preparation were adopted. The composition exhibited the highest enhancement in the room-temperature conductivity from that of the host. We referred to this as the `optimum conducting composition (OCC)' with room temperature conductivity about . The coexistence of constituent phases was confirmed by x-ray diffraction (XRD) and differential scanning calorimetry (DSC) studies. Various ionic transport parameters, namely the conductivity, mobility, mobile-ion concentration, transference number and drift velocity of the conducting ion, were measured experimentally as functions of the temperature. The results have been explained on the basis of theories proposed for two-phase composite electrolyte systems.

The displacement of a non-wetting fluid by a wetting one in a porous medium is influenced, among other things, by two competing mechanisms: the flow of the wetting phase along crevices, giving the possibility of snapping off in throats; and its advance through the centres of the pore space under various pore- and throat-filling conditions, leading to a cooperative filling. The percolation process associated with these two mechanisms on porous networks in two and three dimensions is well understood; it is a classic bond percolation problem competing with an invasion percolation one. We present a three-dimensional pore-level model that describes these effects on a site and bond network representing pores and throats, respectively. The network elements may have various degrees of correlation among their sizes. Site and bond-size distributions may be any kind of function representing a real pore space. In this work, we are able to predict various kinds of patterns that arise when the two aforementioned mechanisms compete and to study the effect of the correlation's strength on the onset of each pattern, revealing the strong influence of the topology of the network in determining which process will dominate. Buoyancy forces are not taken into account in the present work.

In this study, various two-stage deposition processes have been investigated in order to produce device-quality of chalcopyrite thin films. In principle, these techniques involved the preparation of various metallic precursors (by co-evaporation and sequential deposition) and the subsequent reaction of these precursors with a controlled -Ar atmosphere. In the first approach, conventional co-evaporation processes were used to prepare metallic Cu-In-Ga-Se precursors at substrate temperatures as low as C. This process produced uniform and dense films, but x-ray diffraction studies revealed the presence of broad x-ray peaks (indicative of poor crystallinity). The improvement in crystallinity brought about by -Ar treatment was critically influenced by the selenization parameters (especially the reaction temperature and gas concentrations). Virtually no improvement in the material's quality was observed at selenization temperatures below C. Optimum material properties (single-phase material of high crystallinity) were obtained when these co-evaporated precursors were exposed to 10 vol% at final selenization temperatures of C. In the second approach, sequentially deposited triple layers (Ga/Cu/In, Ga/In/Cu and Cu/Ga/In) and multilayers (Ga/Cu/In/Ga/Cu/In) were reacted with -Ar. In general, selenization of triple layers resulted in films with poor crystallinity (morphological irregularities and separated and phases were present). However, reaction of multilayers with 10% in Ar at final temperatures of C resulted in single-phase material with uniform and dense surface morphologies. Photoluminescence studies indicated, in all cases, the presence of one broad donor-acceptor-pair transition. However, in the cases of selenized co-evaporated Cu-In-Ga-Se alloys and sequentially evaporated multilayers, this emission line shifted to higher energies, which indicated that Ga is present in the near-surface region of these specific samples. The production of single-phase films at relatively low processing temperatures (C rather than the C used in conventional processes) and the control of the Ga concentration gradient through the samples are important technological advantages of these two-stage processing techniques.

In physical experiments the percolation threshold for conductivity of conductor-insulator mixtures can be achieved by addition of conductive inclusions to the insulating embedding matrix, keeping the volume of the cell constant, or by compression of a mixture containing an initially nonpercolating quantity of conducting inclusions. The latter case we define as a pressure-induced percolation transition (PIPT). Two limiting cases of PIPT may be envisioned, namely rigid conducting particles in a deformable insulator and deformable conducting particles in a less deformable insulator. In both cases the percolation state is achieved at some critical pressure (strain) which depends on the initial concentration of the conducting phase. The experimental results on conductivity of powder conductor-insulator mixtures show that pressure-driven percolation differs substantially from the usual percolation model: even in the simplest case of a rigid conductor in a deformable insulator the critical index of conductivity close to the critical strain is much less than the expected value . The value is achieved at larger strains. The low value of close to the percolation threshold can be explained by considering the competition of processes of destruction and creation of contacts between conductive inclusions and clusters under the applied stress. In the case of deformable conductive inclusions interpretation of the PIPT is rather complicated. The strain sensitivity of the conductivity is maximal close to the percolation threshold; we consider some practical implications of anomalous strain sensitivity.

Pure and chromium-substituted lithium ferrites have been made in air by self-propagating high-temperature synthesis (SHS), a combustion process involving the reaction of lithium peroxide, iron oxide, chromium oxide and iron or chromium metal powders. Three series of SHS samples were produced: by SHS only; by SHS followed by sintering at C for 2 h; and by SHS in a magnetic field of 1.1 T followed by sintering at C for 2 h. X-ray data showed that a cubic spinel ferrite was produced in all cases, with single-phase products achieved after sintering. Systematic changes in lattice parameter were seen both as a function of Cr content and, for the sintered samples, as a result of the field applied during SHS. Scanning electron microscopy showed that the crystallites had sizes of order . Electron microprobe analysis and EDAX indicated that the samples were homogeneous and had the expected Fe-to-Cr ratios. Mössbauer and magnetic hysteresis data showed that there were significant changes in sublattice occupancy and nett magnetization with Cr content. Coercive forces in the Cr-doped ferrites were larger than those in pure compositions. Some differences between Mössbauer and magnetization parameters of the sintered zero-field and applied-field SHS samples were observed.

Correlations between the ferroelectric fatigue and variations in mechanical properties have been investigated by simultaneous measurement of the fatigue and the piezoelectric resonance of bar 8/65/35 PLZT ceramics. It was found that the fatigue reflects itself only in dielectric and piezoelectric properties. Namely, the remanent polarization, the complex dielectric and piezoelectric constants and the electromechanical coupling factor all start to decrease at the same number of switching cycles. On the other hand, the complex elastic compliance remains almost unchanged during the fatigue process. In addition, the electrical, mechanical and electromechanical quality factors, determined at the piezoelectric resonance frequency, remain constant with the increasing number of switching cycles. This suggests that there is no correlation between the ferroelectric fatigue process and the mechanical degradation of 8/65/35 PLZT ceramics.

Thermally stimulated depolarization current thermograms were obtained from modelling experiments on polar systems, assuming that the polarization can be described by a weighted summation of elementary Debye-like processes. The analysed situations were studied both by global and by cleaned experiments and the features of the obtained spectra were discussed and compared with the characteristics of the systems and the experimental conditions. The polarization distribution through the time relaxation axis during a cleaned experiment for a case characterized by a distribution of activation energies was calculated for all experimental steps. It was shown that the activation energy obtained by the Bucci method must be lower than the mean energy of the isolated processes.

We describe an analysis of phototransferred thermoluminescence (PTTL) using simple energy band models and considering the dynamics of charge transfer between localized trapping states during irradiation, illumination and heating. The various behaviours which are possible are highlighted and described. Particular attention is focused upon the dependences of the PTTL signal upon the illumination time and upon the wavelength of excitation. Some comparisons with experimental PTTL signals from crystalline quartz are discussed.

Measurements of temperature and electron number density measurements have been carried out on a plasma plume under various pressure conditions (p = 1-0.3 bar) using emission spectroscopy. The anode nozzle used does not generate a shock wave or an expansion-compression zone. Ar- (1%) is used as the plasma gas. Temperature values are obtained from various lines of neutral argon. Electron number densities are obtained from the Stark-effect broadening of the 486.1 nm H line and the continuum measurements. Measurements of temperature and density profiles reveal the general features of a low-pressure plasma jet, namely relatively low temperature and long heating zone of the expanded jet. The LTE criterion is satisfied at higher pressures ( bar). Deviations from ionization equilibrium occur at lower pressure. The variation of excited neutral argon and states has been studied for lower pressures. tends to be overpopulated compared with .

A mathematical model to describe the globular transfer in gas metal arc welding is developed. This work is both theoretical and experimental. Using the volume-of-fluid (VOF) method, the fluid-flow and heat-transfer phenomena during the impingement of a droplet on a solid substrate, arc striking, the impingement of multiple droplets on the molten pool and finally the solidification after the arc extinguishes are dynamically studied. A He-Ne laser in conjunction with the shadow-graphing technique is used to observe the metal-transfer processes. Theoretical predictions and experimental results are shown to be in good agreement, suggesting that the theoretical treatment of the model is sound.

Results of experimental determinations of the gas temperature and plasma parameters in a microwave discharge not at thermal equilibrium are presented. The investigation is concerned with high-pressure hydrogen under conditions such that radiation is emitted both by atomic and by molecular components of the plasma are primarily governed by the interaction of the excited atoms and molecules with heavy particles. Because this takes place the approximate models which are successfully applied to analyse data from spectroscopic diagnostics cannot be used to describe the radiative properties of the discharge. One of the main objectives of research consists of making headway in spectral methods of gas-temperature determination in the high pressure range by invoking for interpretation of experimental data calculations performed in the framework of the kinetic scheme of thermal non-equilibrium discharge in hydrogen as well as general physical reasoning relating to the role of interactions between heavy particles in non-equilibrium plasma radiation. The following measurement techniques have been used: spectral measurements of Balmer-series radiation; spectral measurement of the gas temperature by recording the Doppler broadening of the ; and spectral measurements of by recording the Fulcher band. In addition the electron concentration has been measured with the help of a microwave interferometer and pyrometric measurements of the temperature of a test solid body placed in a microwave-discharge plasma have been performed. Measurements of the gas temperature agree satisfactorily under the assumption that collisions of excited particles with heavy components of a gas-discharge medium are of considerable importance. Conclusions regarding the ion composition and degree of dissociation of hydrogen molecules can be drawn.

A theory is developed to account for the influence of flocculation on the ultrasonic attenuation spectra of emulsions. The theory assumes that flocculated emulsions contain `particles' which have effective properties determined by the size and volume fraction of the individual droplets within the flocs. The theory predicts that the attenuation spectra of emulsions depends on the size of the droplets, the size of the flocs, the packing of the droplets within the flocs and the fraction of droplets which are flocculated.

Droplet flocculation in oil-in-water emulsions has been monitored by ultrasonic attenuation spectroscopy, a technique which is sensitive to changes in the spatial distribution of the droplets within an emulsion. The ultrasonic attenuation spectra (1-150 MHz) of a series of 5 wt% corn oil-in-water emulsions (m) were recorded at 25C. Depletion flocculation was induced in the emulsions by adding various concentrations (7-100 mM) of sodium dodecyl sulphate to the aqueous phase. At low frequencies, the attenuation coefficient of the emulsions decreased with flocculation due to overlap of the thermal waves generated by the droplets. At high frequencies, the attenuation coefficient increased with flocculation due to an increase in scattering by the flocs. The results could be explained using an effective medium theory recently developed to account for the influence of droplet flocculation on the ultrasonic properties of emulsions. The ultrasonic technique was also used to monitor the breakdown of flocs under shear flow. The dependence of the ultrasonic properties of emulsions on flocculation means that ultrasonic attenuation spectroscopy can be used to study interactions among droplets in concentrated emulsions.

This paper is concerned with an experimental investigation in which an electrorheological (ER) fluid is applied to control vibration. The fluid is sandwiched between two planar electrodes, the upper one fixed while the lower one is subjected to an oscillatory sinusoidal motion and is energized either by a constant voltage or by a constant field. In this dynamic squeeze flow it is seen that the force transmitted across the fluid is greater when the applied electrical field is maintained at a constant level, controlled with reference to the instantaneous gap between the electrodes. In addition, we compare the experimental results with those from an existing theoretical model in which the fluid is assumed to have a bi-viscous rheological characteristic.