Table of contents

In this paper, the relations expressing the effects of roughness orientations and molecular mean free path on the roughness-induced flow factors are derived. Such consideration of the effects of roughness orientation on bearing performance is necessary when rotary actuators are used in magnetic recording systems. Using the perturbation approach and the coordinate transformation, the flow factors are derived and expressed as functions of the roughness orientations, the Peklenik numbers, the standard deviations of roughness heights of each surfaces, the inverse Knudsen number and the film thickness ratio of the lubrication film, where h/ sigma is the ratio of average film thickness to standard deviation of the composite roughness. Two orientation coefficients are proposed to reveal the combined effects of Peklenik number and orientation angle and the change of flow factors is analysed as the orientation coefficients vary. Finally, the static performance of a magnetic recording head flying above a rotating disc under the ultra-thin-spacing condition is analysed with roughness and rarefaction effects taken into account. The effects of surface characteristics of the moving surface on load capacity are more significant than those of the stationary surface.

Domain wall depinning and propagation in a long wire of amorphous Fe_77.5Si_7.5B₁₅ alloy in the as-quenched state are studied during AC magnetization under a DC biasing field produced by a solenoid perpendicular to the wire. The active walls undergoing the depinning and propagation are classified into positive and negative ones. Consequently there are alternating and unidirectional magnetic reversal modes, both having two distinct types. The alternating mode can easily be induced by an external magnetic pole equidistant from the wire ends. The experimental data obtained in this work and published elsewhere are discussed in detail. It is concluded that the eventual depinning of the closure domain wall from the end is induced by its elongation and transverse movements. The latter process is essentially the same as domain wall propagation. Both are dynamic with a dominant eddy current damping and domain wall shape-dependent DC critical field.

It is well known that in an unbounded anisotropic homogeneous medium there are eigen plane wave solutions, and the fields inside a circle can be expressed as an integral of eigen plane waves with respect to the polar angle. In this paper the fields inside a circle are expanded as a finite summation of eigen plane waves with different polar angles. Each eigen plane wave is a wavefunction of the first kind, which can be represented by a series. From this series form of wavefunctions of the first kind for anisotropic media, series representation with explicit coefficients of the circular cylindrical wavefunctions of the second, third, fourth kind for homogeneous anisotropic media are obtained. This paper further deals with the numerical implementation of the above eigenfunction representation for homogeneous anisotropic media. Applications of the theory are made to the case of two-dimensional scattering of a plane wave by an anisotropically coated cylinder and comparisons are made with publishing data. The computational efficiency of the method expounded in this paper is compared with previous work.

A set of experimental back-scattering patterns produced by metallic surfaces consisting of spheres on a flat surface is presented. The study is made within the plane of incidence for copolarized light (polarization parallel or perpendicular to the plane of incidence), for two sphere diameters phi =1.1 and 3.0 mu m, and for an incident wavelength lambda =0.633 mu m. Also, a simple ray-tracing method is presented for the analysis of a one-dimensional surface model (cylinders on a flat substrate) whose results are compared with those obtained from the experiment. The effects of the surface particle density on back-scattering patterns are analysed both experimentally and through an incoherent geometrical addition of rays.

The contribution of radiative charge exchange and radiative association in symmetrical ion-atom collisions to continuous EM radiation from weakly ionized gaseous plasmas has been considered within the semiclassical adiabatic theory. The differential (in lambda ) cross sections for spontaneous photon emission and the general expressions for the spectral coefficients of emission and absorption are given. The hydrogen and helium plasmas that are representative of two different optical types of gaseous medium have been studied in a broad range of T and lambda , 4000<or=T (K)<or=20000 and 200<or= lambda (nm) <or=1000. The domain of T and lambda , where ion-atom collisions contribute significantly to continuous plasma spectra, has been established. The case of weakly ionized alkali metal plasmas of Li and Na has been studied in the same interval of lambda but at lower temperatures, 1500<or=T (K)<or=3500. The relevance of the results to studies of laboratory plasmas is discussed.

The axial gradient of planar RF plasmas produced by an even surface wave travelling between a pair of semi-infinite dielectric regions is obtained theoretically for two kinds of gas discharges, namely the so-called diffusion and recombination regimes. In this theory, two first-order approximations with respect to an electron-neutral species collision frequency for momentum transfer and the wave attenuation are employed to solve the electrostatic surface wave. Some typical characteristics are disclosed as follows. (i) The produced plasma density is expressed by a simple characteristic function for the two gas discharge regimes. (ii) The axial gradient of the plasma density is much less than that in the case of standard cylindrical tubes. The utilizability of the planar RF plasmas for industrial uses such as material processing by plasmas is discussed.

An experimental and theoretical study of the drift oscillations in plasma of inert gases with a density gradient across the magnetic field has been carried out under conditions when the plasma is destabilized by the ion inertia. Experimental investigations of the unstable oscillations are performed at low gas pressures and high electron densities. For the theoretical description of the drift oscillations a set of differential equations for a cylindrical bounded plasma is used.

The influence of a transverse magnetic field on the weakly ionized plasma instability caused by competition between the diffusion and thermodiffusion processes is investigated. A criterion of this so-called thermocurrent instability for the plasma carrying a current in electric and magnetic fields is derived. It is shown that the transverse magnetic field is favourable to instability growth. This is associated with the disturbances propagating at an angle to the electric field and with a change in the electron transport parameters. From a calculation of the electron transport coefficients, boundaries of the instability region are obtained in He, N₂ and CO₂.

The excitation of travelling nonlinear Langmuir waves in a plasma by a relativistic electron bunch is considered. The nonlinear excitation is described as relativistic oscillations of the momentum of plasma electrons in effective potentials produced by the bunch and by the perturbation of the electron density of the plasma. The condition for the optimal excitation of the wave yields a relation between the density of the plasma, the density of the bunch and the length of the bunch. Numerical results support the theoretical analysis and demonstrate a deep nonlinear modulation of the electron plasma density and excitation of a large-amplitude longitudinal field at realistic plasma and electron pulse parameters. The back action of the wave field on the bunch is considered. It is shown that sharp peaks develop in its density distribution.

The effect of electrode processes on the spatial uniformity of a self-sustained XeCl laser discharge has been experimentally studied. It has been shown that the longitudinal discharge structure, much as the classical one, is formed towards the end of the pre-breakdown stage. The primary disturbances of the uniform discharge structure have been observed during the formation stage at a current density of about 1-10 A cm^-2. It was supposed that non-local effects play a part in discharge filamentation. The effect of the rate of increase of current on the development of the discharge instabilities has been revealed.

Plasma ion dynamics following the application of a negative voltage pulse to a two-dimensional trench geometry are computed using a two-fluid model. The potential structure in the sheath both focuses and bunches ions, creating an ion density pulse that travels down the sidewall of the trench. It may be possible to exploit this phenomenon to improve ion implantation in the sidewalls of trenches and holes when using plasma source ion implantation.

This work presents an experimental study of argon microwave discharges maintained by a surface wave at 2.45 GHz, at atmospheric pressure and under low flow, in a capillary tube. Emission spectroscopy techniques are used to characterize these discharges. Special attention is paid to the determination of the electron temperature by using the line-to-continuum intensity ratio method. We show that, even in situations very close to local thermal equilibrium, simple substitution of the excitation temperature T_exc by the electron temperature T_e in the equation governing the line radiation emitted by a plasma should not be done. Causes for such an assertion are analysed and a parametric study is conducted to determine the sensitivity of the proposed method. This method is then applied to provide T_e from experimental data corresponding to an atmospheric pressure argon plasma produced by a surface microwave.

Monte Carlo simulation has been used to investigate electron motion in magnetically confined CO gas laser discharges. The influences of a perpendicular magnetic field on electron transport coefficients have been studied. Data on transverse and perpendicular velocities, mean energy and numbers of several types of collisions are evaluated at various magnetic field types and strengths. The results show that the applied magnetic field decreases the transverse velocity, mean energy and ionization coefficient, and increases the perpendicular velocity. The high perpendicular velocity suppresses localized thermal electron constriction, giving a higher input power loading. The applied magnetic field also increases the number of collisions, resulting in increasing losses of the electron energy obtained from the electric field, especially the losses in vibrational levels. The magnetic field distribution and strength have great effects on the electron transport coefficients and the discharge properties.

A Monte Carlo technique has been used to simulate ionization in a magnetron-like model. Spatial distribution of ionization as well as electron loss under non-uniform magnetic fields of various strengths has been investigated. It is found that ionization increases with the magnetic field over a wide range ( approximately 0.15 T), but due to electron loss via cathode absorption it decreases at strong fields. With increasing field strength, ionization is found close to the cathode, concentrated at the place where the magnetic field is parallel to the cathode surface, and dominant at the E*B drift direction. It is also found that electron loss is decreased by the confinement given by the non-uniform magnetic field component parallel to the electric field, while, increased by the E*B drift in the model and increased by the cathode absorption when the magnetic field is made stronger. Analyses of the results are given; the results can be applied to magnetron discharges.

Measurements were made of current growth waveforms corresponding to the glow-to-arc transition in a 1 cm positive point-to-plane gap in conditions similar to those in pulsed corona devices used for flue gas cleaning. The glow-to-arc transition was significantly accelerated using a freshly polished unconditioned cathode surface. A duration of glow-to-arc transition as short as 25 ns was observed using a graphite-coated cathode. A possible explanation in terms of the cathode sheath break-ups is presented.

A sustained current conduction in the breaking chamber after a current interruption has been observed. This current, which we named 'Post-recovery (PR) current', is in phase with the applied voltage and its value ranges from several milliamps up to several amps. The present study shows that the probable origin of the PR current is an ionic conduction of SF₆ between the contacts. The estimated temperature of the conducting gas is about approximately 2000 K. The evolution of the temperature is in good accordance with a computational model based on the resolution of two-dimensional Navier-Stokes equations.

In the framework of a one-dimensional model for plasma flow in ablative capillary discharges, we use a consistent method for obtaining the appropriate boundary conditions, and a new method for describing the radiative energy flux from the plasma to the wall. Good agreement is reached between the calculated and the measured values for the temporal behaviour of the plasma's main parameters. The agreement is particularly satisfying for the plasma conductance (resistance), the voltage across the capillary, the total input energy, the maximum input power and the plasma pressure. This gives us confidence that the model also predicts correct values for the remaining plasma parameters that are difficult to measure accurately, such as the plasma density, the temperature and the velocity (inside the capillary as well as at its exit). The knowledge of these parameters is crucial in order to use the ablative capillary as an efficient plasma source for different applications.

Diamond films were synthesized by the hot-filament chemical vapour deposition (HFCVD) method on Si(100) substrates with surface biasing pre-treatment. Evidence has been obtained from high-resolution cross sectional electron microscopy (HREM) that the diamond films grown on Si wafers showed localized epitaxy. There is an angle of approximately 7.3 degrees between Si(100) and D(100) hetero-epitaxial crystalline planes. The same type of twins (coherent twin boundaries of type Sigma =3) exists on and near the interface. It is shown that the pre-treatment of the substrate is a key factor for the localized hetero-epitaxy of diamond films on Si wafers.

1.0 and 2.0 MeV Ti ions have been implanted into silicon at different angles to study longitudinal and transverse distributions. The profiles of implanted Ti ions in Si at different angles were measured by secondary ion mass spectrometry (SIMS). The longitudinal and transverse moments are compared with the TRIM'90 Monte Carlo simulation. The results show that the longitudinal moments-projected range and range straggling-are consistent with calculated values within 13% and 15% respectively and the transverse moment is in agreement with the TRIM'90 simulation within 29%. The depth distribution obtained is also compared to Gaussian and Pearson functions.

Structural, electrical and optical properties of thin films of the Cu-In-Ga-Te system grown by thermal evaporation have been studied relating the measured film composition to these properties. Above a substrate temperature of 250 degrees C the chalcopyrite structure is obtained and films deficient in Te by up to 16% show this structure. The optical absorption coefficients of the films present two or three inflection regions corresponding to several energy gaps, two direct and one forbidden direct gap. The first direct gaps lie between 0.46 and 0.88 eV and the second ones in the 1.01-1.37 eV range. The forbidden direct gaps are in the 0.94-1.83 eV range. The first gap appears only in the samples with Cu/In>1.06. The values of the second energy gap and the cell parameters of tetragonal phase depend on composition, i.e. on Cu/Ga atomic ratios. The measured resistivities and mobilities of samples are ranged in the 0.001-1.63 Omega cm and 5-38 cm² V^-1 s^-1 ranges respectively. The samples were p-type or n-type; those with a defect in Te were n-type. Thermal treatments have been done in a hydrogen flux at 350 degrees C. This annealing varies notably the initial number of phases of the films.

Schottky barrier contact using Au and ohmic contact using Al were made on n-GaN grown by hydride vapour phase epitaxy. The diodes were characterized in the range 77-373 K. Under forward bias, the ideality parameter n=1.04 and the threshold voltage is 1.1 V. The reverse bias leak current is below 10^-9 A on a reverse bias of -10 V. The temperature-dependence of the I-V characteristics shows two regimes of forward current transport: one at low voltage governed by thermionic emission and the high-voltage regime due to spatial inhomogeneities at the metal-semiconductor interface. The barrier height phi _B and the electron affinity X_S were determined to be 0.91 and 4.19 eV, respectively, by I-V measurement; and 1.01+or-0.02 and 4.09 eV, respectively, by C-V measurement. The results have been discussed.

Solutions are given for the decay of the local current in both a finite and an infinite superconducting circuit consisting of a linear array of loops. These give some insight into the behaviour of persistent currents in superconducting dipole magnets. Generating functions are derived for the polynomials associated with the finite chain solutions. The solution for the infinite loop model are found from the Green's function for the free particle Schrodinger equation.

The prediction of the properties of polymers in contact with water is important for many applications. In this paper two kinds of materials are investigated: an epoxy resin and its glass fibre composite. The measured complex dielectric constant is analysed with different effective medium models, taking into account the variation of water uptake according to time. Since the mixing laws do not describe the experimental results accurately, three new models are proposed, based on the assumption of a time-varying dielectric constant of the matrix during water immersion. They all give a better agreement between measurement and theory. In order to check their validity, these models are also applied to other data sets, found in the literature, and good correlations are found. Finally, we discuss the conclusions that one may derive from this kind of modelling, since many factors can influence the permittivity. Nonetheless, the new models are able to predict the variation of the dielectric constant for water immersion ageing.

The dielectric breakdown model is generalized and adapted to describe electric discharge propagation in inhomogeneous insulators. The influences of the injected space charge, inclusions and barriers with different dielectric permittivities and conductivities are investigated. The growth of bush, dendrite and skeleton types of discharge structure as well as transformations from one type to another are simulated.

The HAUP method for determining the linear optical properties of crystals is critically reconsidered. The expressions that describe the method are re-derived rigorously and the fitting procedures that have been used by several authors are reviewed. The systematic errors, being crucial in HAUP measurements, are investigated by means of experiments on test samples. The understanding of both the origin and the behaviour of these errors leads to an improvement of the fitting procedures and of the interpretation of the data. In this way, the reliability and the sensitivity of the HAUP method have been increased, which should provide a means by which to reconsider contradictory results of previous experiments, excluding systematic errors as a source of the dispute.

In this paper we show that the high-accuracy universal polarimeter can be used to investigate the optical properties of (almost) any crystal through which light can be transmitted. The only extra condition is that the reciprocal linear birefringence (or the reciprocal linear dichroism) of the sample is large with respect to the other optical effects that are present. The sample is allowed to exhibit reciprocal linear birefringence and reciprocal circular birefringence. This is the situation for which the high-accuracy universal polarimeter was originally intended. We show, however, that reciprocal linear dichroism and reciprocal circular dichroism may also be present. The method can, therefore, be applied to (weakly) absorbing crystals. Furthermore, optical effects are taken into account that are related to the breaking of time-reversal symmetry. These are the non-reciprocal circular birefringence, non-reciprocal circular dichroism, non-reciprocal linear birefringence and non-reciprocal linear dichroism. This means that, under some conditions, each (weakly) absorbing, magnetic crystal can be investigated with the high-accuracy universal polarimeter. We derive a unified formula for the intensity change of the light that propagates through the polarimeter. This expression can be used to determine the (complex) eigenpolarizations of an arbitrary sample. Moreover, it is shown how this unified formula can be translated to a formula in terms of the different birefringent and dichroic optical effects. The relevant formula for a specific case can, therefore, be given directly. The method is demonstrated by means of measurements on samples of NiSO₄.6H₂O. This material shows reciprocal linear birefringence, reciprocal linear dichroism, reciprocal circular birefringence and reciprocal circular dichroism, simultaneously.

The interference of two opposing coherent waves in a thin layer was studied. Expressions for energy fluxes outside and inside the layer are obtained for normal incidence. It is shown that the interference flux, whose magnitude and direction vary markedly with layer parameters, amplitude and phase of incident waves may be derived together with incident, transmitted and reflected wave fluxes. The distribution of the total energy flux by thickness inside absorbing and amplifying layers was examined for single incident wave. The dependence of transmission and reflection energy coefficients, T and R, on the thickness of a layer with negative absorption has been studied numerically in detail for the case of normal incidence.

Space charges formed in low-density polyethylene (LDPE), to which high voltage is applied, are usually composed of injected homocharges from electrodes and ionized heterocharges from bulk impurities. In this paper, the heterocharges in LDPE were discriminated from homocharges using charge-injection suppression layers. Recent advanced techniques have made it possible to observe directly the space charge distributed in dielectrics. However, it is difficult to study the characteristics of charge injection from electrodes and ionization of impurities individually as the two processes may interfere with each other and become inseparable. Therefore, a layer that suppresses the charge injection from electrodes was used to discriminate between injected and ionized charges. To observe the space-charge distributions, the pulsed electroacoustic (PEA) method, one of the advanced techniques, was employed. In LDPE with suppression layers on both surfaces, hetero-space charges, which are attributed to ionization of impurities, were observed. On the other hand, in LDPE without the suppression layers, small amounts of homo-space charges, which are attributed to net charges from both the electrode injection and the bulk ionization, were observed. Amounts of charge supplied from electrodes were estimated by subtracting the amount of space charge in the sample with suppression layers from that in the sample without suppression layers.

A method for obtaining analytical expressions for the energy (intensity) optical coefficients and their envelopes of a film with a linear variation in thickness is proposed. The method is demonstrated for the transmittivity, front and back reflectivities, and front and back absorptivities of a semiconductor absorbing film on a transparent substrate.

This paper studies the effect upon the thermoluminescence (TL) signal of four feldspar samples of exposing them to infrared stimulation, as occurs during infrared-stimulated luminescence (IRSL) measurement. Together with pulse annealing measurements these results are used to show which part of the TL signal is removed by exposure to IR and which part is directly related to the IRSL signal that is observed. When the samples are preheated prior to measurement in order to remove any low-temperature (<200 degrees C) TL signals, a linear relationship is observed between the amount of TL that is lost and the IRSL light sum that is produced. The IRSL light sum is consistently four times larger than the amount of TL that is lost. Three possible explanations are proposed for this, but no conclusive evidence could be obtained to support any of them. A close similarity is observed in the pulse annealing results and the loss of TL due to IR exposure of the four samples. This is taken as evidence that under in stimulation a similar trap, or trap complex, provides the charge donor in all of the feldspar samples studied here in spite of the variability in the TL signals observed.

Four groups of grain sizes of LiF:Mg,Cu,P material were chosen for the experiment. The influences of the grain size on the photon energy response and dose response have been examined in the interval of 28-1250 keV and the dose range 0.1-30 Gy of ⁶⁰Co gamma -rays. The experimental results show that both the energy response and the dose response are grain size dependent. A possible correlation seems to exist between the energy response eta (E) and the dose response f(D). A brief discussion is provided.

CaSO₄:Dy and CaSO₄:Tm are highly sensitive phosphors used in the thermoluminescence (TL) dosimetry of the ionizing radiations. It is shown that, by using P, Mo or SO₃ as co-dopants, the TL sensitivities can be increased by 40-100%. The increase in sensitivity originates in various processes, which are discussed. P acts as a charge compensator, allowing the incorporation of Dy in higher concentration without causing quenching. Mo increases the efficiency of trap filling. It is claimed that use of Mo as a co-dopant will improve the sensitivities without disturbing the other dosimetric characteristics.

Filament-type sources of Li⁺ ions, suitable for use in Li⁺-ion-attachment mass spectrometry, have been made using a Re ribbon filament coated with various Li-containing materials (Li₂O, Li₂SiO₃, Li₃PO₄, zeolite A and beta -eucryptite). These thermionic materials were studied and compared in terms of total emission, purity and stability over time, as a function of filament temperature (heating power). Based on the experimental results, it was concluded that beta -eucryptite is the best coating compound capable of producing copious and stable emission of Li⁺ at low heating power.

Hexagonal ZnO films were deposited on a glass substrate at room temperature by RF sputtering a ZnO target using a mixture of oxygen and argon. Polycrystalline ZnO films prepared in the presence of excess oxygen or at a high oxygen pressure tend to have crystallites with the orientation of the c-axis parallel to the substrate and a large and relatively fast photoresponse. The time constants of the photoresponse can be much reduced by doping the surface layer with nitrogen. This fast photoresponse of the resulting films is related to the changes of barrier heights of grain boundaries due to photodesorption of the chemisorbed oxygen on the grain boundaries when the films are irradiated with uv light.

Many of the components of particle detectors for the next generation of high energy experiments need to be extremely radiation-hard if they are to continue to function for more than a few months. For example, in the first few year's operation of the large hadron collider, we expect about 10¹³ neutrons cm^-2 year^-1 1 cm from the beam pipe, with other particles causing about ten times as much additional damage. Diamond is being considered as both a detector material and perhaps as a structural component. We report on tests of chemical vapour deposition diamond films irradiated by 6*10¹⁴ neutrons cm^-2 of energy below 10 keV, plus (2.0+or-0.5)*10¹⁴ neutrons cm^-2 of energy above 10 keV peaking at about 1 MeV. The diamonds were metallized with titanium-gold contacts-strips on the front and a continuous electrode on the back. The leakage current (already extremely small) decreased by factors of about 50 as a result of the irradiation, both measured between adjacent strips and through the thickness of the samples. The detection efficiency of charged particles improved by 30-40% after irradiation. In addition, natural diamonds irradiated by 4*10¹⁴ neutrons cm^-2 showed very little optical absorption in the GR1 and ND1 bands (neutral and negative vacancy) corresponding to a concentration of vacancies of 4(+or-2)*10¹⁴ cm^-3.