Table of contents

An 80 watt cw CO₂ laser has been used to evaporate thin films of Al₂O₃ on to p- and n-type silicon. Films in the range 50-500 nm have been deposited. Capacitance-voltage measurements have been carried out on MOS devices made on these films. Results indicate changes of resistivity of the silicon substrate which could be attributed to the diffusion of laser-induced Al ions into the silicon. Another aspect of the films is the change of stoichiometry of aluminium oxide as a result of annealing. A time-dependent hysteresis has been observed.

The usual method of output coupling for the intracavity-generated second harmonic of 1·06 μm leaves room for improvement of the efficiency by a factor of two. Methods of achieving this factor of two are proposed. The simplest one was tried and gave promising results.

A preionization for the theta-pinch in a very low-pressure region is attained by adding an auxiliary discharge circuit to the usual z-discharge circuit. By applying this method, the contraction time of a theta-pinch in a wide pressure region is measured. Its dependence on the gas pressure shows a certain anomaly below 0·5 mtorr.

Recent advances in the study of microwave ferrites and garnets have resulted in a better understanding of the relationship between magnetic properties and the fundamental crystal structure. This article presents a review of the way in which the crystal structure of microwave garnets may be designed in order to realize the many different combinations of magnetic properties required by the designer of non-reciprocal microwave ferrites. The discussion relates to rare-earth and yttrium-containing garnets of the form Y_3-αR_αFe₅O₁₂ (R = rare earth Sm-Lu) and considers the effect of substitutions of Al, Ga, In, Mn, Ce, Ca and V on magnetization, gyromagnetic resonance, magnetic loss, high-power instability threshold and magnetostriction effects.

Direct microscopic measurements are reported, on the deformation of commercial bonded fibrous networks. The samples include paper and non-woven textiles. Square grid patterns were copied on to the materials and these were photographed during tensile straining. The structures of the deformed patterns were evaluated statistically. In general there was a significant correlation of local deformation with the local areal density of mass. This gave linear regression equations with which the predictions of a statistical theory of elasticity were reasonably consistent. Only a summary of this theory is given here. Its principal concern is with deriving the average over finite regions, of the scalar bulk relation at points. The present experimental evaluation is particularly apt since the modulus at points in a flat, bonded fibrous network is proportional to the areal density of mass at points.

The problem of obtaining sufficient information from the profile of a rough surface, in order to predict the variation of the true contact area between it and a smooth surface as a function of load, is discussed. The effects of load and orientation on surface profiles were examined experimentally for anisotropic and several isotropic surfaces, all with randomly distributed surface heights, and the observations were analysed by computer. The results for the isotropic surfaces agree well with theory, while those for the anisotropic surface are consistent with proposed simple surface models. The variation of the number of contact spots per unit area with load was also measured directly. If an assumption is made concerning the original height distribution, the predictions for the number of contacts per unit area agree well with the experimental observations, while the discrepancy between theoretical predictions and experimental observations for the number of contacts per unit length is less than 4%.

It was shown in 1966 by MacDonald that the conventional relation Λ=πtan δ between spatial logarithmic decrement and loss tangent fails for high damping conditions and that the exact relation reads Λ = 2πtan δ/2. This means that if one calculates tan δ from the spatial logarithmic decrement and then uses the conventional relation, a value is found for tan δ which is too small. Moreover, in 1966 Parke showed that the same conventional relation Λ = πtan δ between the free-vibration logarithmic decrement Λ and the loss tangent, fails in the opposite direction, under conditions of high damping.

It is the aim of this paper to show that: (i) the relation deduced by MacDonald is also valid for free vibrations provided that the proper, i.e. complex, frequency is used; (ii) the relations given by MacDonald can be deduced in a very simple way by making use of the concept of characteristic impedance.

Moreover, our analysis of the energy dissipation in the material showed that in the relations for the components of the relevant complex elastic modulus, M₁ and M₂, where M^* = M₁+iM₂, in highly attenuating materials the velocity of energy transport c₁ is a more proper parameter than c the velocity of propagation.

A study has been made of the thermoluminescence sensitivity of lithium fluoride crystals as a function of both grain size and the surrounding atmosphere when given prolonged annealing at high temperatures. Optical grade lithium fluoride and dosimetry grade lithium fluoride (Harshaw TLD-100) do not show the same response to high-temperature annealing. For a given technique of quenching from high temperature, the thermoluminescence sensitivity of optical grade lithium fluoride increases with temperature when annealed in the absence of air, whereas the thermoluminescence sensitivity of TLD-100 stays roughly constant. When annealed in air the sensitivity of TLD-100 decreases sharply with increasing temperature. Optical grade crystals when annealed in air show an increase in optical absorption in the visible spectrum, and infra-red absorption bands are found due to the presence of OH anions which are a known thermoluminescence poison. Optical grade crystals show a generally decreasing sensitivity with increasing grain size.

The properties of the stannates and zirconates (M₂Sn₂O₇ and M₂Zr₂O₇) of yttrium, lanthanum, cerium and neodymium are discussed, and the materials are considered for possible application in magnetohydrodynamic (MHD) power generation. Measurements are discussed of their structural properties, thermal expansion, thermal diffusivity, electrical resistivity, resistance to oxidation, to chemical corrosion, and to electrolytic corrosion. Thermal shock resistance is also qualitatively assessed. For non-electrical applications, lanthanum stannate has shown the most promise, especially in the corrosion tests. All the materials are insulators below 1000 K, except cerium zirconate which has ample electrical conductivity (> 10 Ω⁻¹ m⁻¹ above 1400 K) for use as an MHD electrode. Cerium zirconate has complicated structural and electrical properties, and must be prepared in low oxygen partial pressures. However, it is easily oxidized and may disintegrate. The equimolar ceria-zirconia solid solution formed when the oxides are fired in air would be more useful as an MHD material.

Polycrystalline GaAs layers have been deposited on sapphire substrates by the AsCl₃-Ga-H₂ vapour transport method, and treated with caesium and oxygen. A layer 0·6 μm thick when operated as a photocathode has shown a luminous sensitivity of 70 μA lm⁻¹, in transmission. Thicker layers have shown lower sensitivities. Spectral response curves are given for layers 0·6 and 0·8 μm thick and show yields out to 1·4 ev.

An attempt has been made to relate the results to an elementary theory for transmission photocathodes.

The Kerr magneto-optical effect has been used to study the very low-frequency hysteresis loops of individual 180° domain walls in grain-oriented Si-Fe sheet. Measurements on a number of walls show that the loops vary widely in shape and show similar non-repeatable properties to those reported previously at higher frequencies. The peak movement appears to be independent of the coercive force of the walls, suggesting a model for wall motion in which the retarding forces on the wall due to a combination of magnetostatic fields and imperfections.

The Seebeck coefficient, thermal conductivity and electrical conductivity of crystals of Bi₉₅Sb₅ doped with various amounts of Te have been measured at 80 K, using samples that were oriented so as to yield the highest thermoelectric figure of merit. A dimensionless figure of merit of 0·58 has been observed for a crystal doped with nominally 0·002 at.% Te in a magnetic field of 0·4 T, this being significantly higher than the best value that was found for an intrinsic crystal.

From the change of thermal conductivity with magnetic field it has been possible to estimate a value of 2.8 W m⁻¹ K⁻¹ for the lattice conductivity. A rather low Lorenz number is attributed to inelastic scattering of the electrons.

It is shown theoretically that it is possible to reduce the effect of an applied magnetic field on a Mössbauer source by suitable choice of the source temperature. Application of this method to a ⁵⁷Co-in-Pd source yields a value of J=7·2±0·8 for the spin associated with dilute Fe in Pd.

Natural conduction in nitrobenzene has been measured in electric fields of 50-50 000 V cm⁻¹, over a temperature range of 0-200°C using brass and stainless-steel electrodes. The electrode materials, dissolved air and degree of purification are observed to have a pronounced influence on the conductivity. The cathode appears to be the most likely source of electrons and conduction occurs as a result of the hopping process. The thermal activation energy in nitrobenzene is found to be 0·53±0·05 ev.

The range of electrons as a function of primary energy below 2 keV, the escape depth of slow secondaries and the efficiency of back-scattered electrons in the production of slow secondaries have been determined in thin aluminium layers deposited on platinum, gold and silver substrates, using an automatic yield measuring device which enables the measurements to be made from contamination-free surfaces.

Shock pressure and impulse caused by the absorption of Q-switched laser light are calculated by an approximate method in which the temperature dependence of the penetration depth of the electromagnetic radiation is considered.

Numerous investigations have shown that the wave velocity of a self-sustaining detonation is less than the ideal Chapman-Jouguet (C-J) value and that the deficit increases with decreasing tube diameter. Moreover, in small diameter tubes, the pressure gradient in the expansion behind the wavefront is steeper than that predicted by isentropic relations. Estimates of these deviations from ideality by the nozzle model of Fay, based on a viscous boundary-layer displacement, only partly account for the observed values.

Numerical integration by Skinner of the non-steady equations of motion of the burned gases, assuming the Reynolds' analogy to hold between friction and heat transfer, shows that a growing region behind the front becomes time-steady as the wave recedes from its plane of origin. This prediction is confirmed in the present work by gas velocity measurements. An approximate solution of the non-steady equation is derived that is valid near the front and which allows a simple criterion to be stated for the attainment of a steady profile.

Wall heat transfer measurements are described using platinum resistance gauges coated with silicon monoxide to suppress the effect of ionization. Transfer rate measurements near the wavefront of oxy-hydrogen waves agree well with the calculations of Sichel and David for the situation obtaining near the C-J plane. The value of the friction coefficient derived from the heat transfer data is used to compute the pressure and gas velocity profiles; these profiles are found to be in satisfactory accord with the observational values.

Using the expressions for the second and third harmonic components of the current density in a plasma, subjected to an external alternating field and a d.c. electric field, the authors have investigated the conditions for maximum efficiency of harmonic generation in a plasma. It is concluded that (i) the intensity of harmonics is higher in the transmitted component than in the reflected component from a plasma slab, and that (ii) the efficiency of second-harmonic conversion is maximum when 0·80 <(ω_p/ω)² < 0·90 and dimensionless plasma thickness (ξ₀=ωz/c) in the range 3·20 <ξ₀ <3·50 and the efficiency of the third-harmonic generation is a maximum when 0·80<(ω_p/ω)²<0·95 and 2·50<ξ₀<3·40. The variation of the efficiencies in these limits is not significant.

A new geometry called the inverse thetatron is described in detail. The dynamics of the plasma motion are calculated using a simple momentum balance theory and measured experimentally. Preliminary results show that good symmetry and useful plasma speeds and displacement may be attained with this geometry thus alleviating the electrode problems common to other hard-core pinch geometries.

An account is given of the development, and verification by experiment, of a Monte Carlo method for obtaining the flux of radiation around a point source in an atmosphere of scattering particles, having an absorbing boundary. It is shown that the method gives satisfactory solutions in regions close to the boundary, where an earlier diffusion method became inaccurate.

The two methods have complementary features which make it profitable for them to be used in conjunction, each in the domain in which it is most accurate and convenient.

An account is given of measurements of the diffuse transmission of light from point sources over horizontal paths in the lower atmosphere. Observations were made at ranges up to 27 km in a variety of meteorological conditions.

For certain situations, it is shown that the transmission is a function of range expressed in units of the transport mean free path. Many of the results show that the transmission at long ranges decreases less rapidly with distance than at short ranges, a feature not reported hitherto.

The response of the Fabry-Pérot etalon to changes in optical length is analysed. The optical length is initially fixed and then allowed to change at a constant rate. The analysis is therefore applicable to experimental situations such as plasma refractive index measurements or spectral line profile investigations. Two time constants of interest, t_d and t_c, are obtained. t_d is the time lag between changes in the optical length and subsequent changes in the output intensity; it is equal to either the single or double transit time within the cavity, depending on the method by which the optical length is changed. t_c is the time during which transients occur and is related to the energy decay time within the cavity. Under certain conditions at high frequencies, the output exhibits harmonics and the necessary and sufficient conditions for this to occur are established. A physical explanation for the generation of these harmonics is given and the relevance of these harmonics to the experimental use of the Fabry-Pérot etalon at high frequencies is discussed.

The systematic variation of neutron flux distribution and buckling measurements as a function of fuel loading in a subcritical assembly (graphite, natural uranium) for two different coolant channel diameters have been investigated. The void ratio (volume of voids/lattice volume) in the two cases was 19% and 24%. The experimental results have been compared with theoretical predictions based on diffusion theory. For full assembly cases the results are in good agreement (as might be expected) with the predictions of Syrett's model for graphite moderated reactors. The correlation of 10 partially filled assembly cases was carried out in three ways:

1. Syrett method lattice parameters with reflector savings calculated from one- and two-group theory for an infinite-slab system.

2. Numerical solution of the two-group, two-region diffusion equations.

3. Two-group heterogeneous (source sink) theory for finite systems.

The results show that method (i) breaks down for all cases except those with the full assembly, method (ii) breaks down in nearly all cases and method (iii) holds well down to 36 fuel channels. The breakdown in method (i) and (ii) is most probably due to inaccurate streaming corrections to reactor systems of smaller size. In method (ii), excessive computer time is an additional complicating factor. If the experiments are analysed by heterogeneous methods it is possible to obtain accurate reactor para meters with a partially loaded exponential assembly in cases where only a limited amount of fuel is available.

The elastic compliances of PbF₂ have been determined using the composite oscillator technique on a set of five single-crystal specimens. The values obtained are s₁₁=1·534, -s₁₂=0·49 and s₄₄=4·756 (10⁻¹² cm² dyn⁻¹).

A new spinel CuMnCoO₄ has been synthesized. Its crystal structure has been determined by the powder method. It is found that this compound is tetragonal with a^*=8·31 Å and c^*=8·58 Å. The oxidation states of the copper, manganese and cobalt ions appear to be two, four and two respectively, from crystallographic considerations and x-ray absorption spectroscopy.

The problem of bulk doping inhomogeneities over the cross section of a Gunn device is approached on a large-signal basis. Gradients of a transverse type induce like components into the field pattern of a dipole domain: they, in turn, modify the laws of propagation and growth. The hypothetical but simple case of a periodic impurity distribution is treated semi-quantitatively to show that there can be a significant change in the response time of immature dipoles. It is concluded that the performance of an oscillator which relies on the transient mechanism, such as a diode operating in the hybrid régime, might be quite sensitive to this class of non-uniformity.

Conventional barium aluminate-impregnated tungsten matrix cathodes were coated with a metallic layer such as iridium, osmium, platinum or tungsten sputtered in an argon atmosphere. This simple practical procedure has been shown to lead to an enhancement in emission capability in the case of an iridium-coated or osmium-coated cathode, comparable with that reported in the literature using very much more complex procedures. The iridium- or osmium-coated cathodes exhibit effective work functions of 1·93 to 1·98 ev over a range 1200 to 1370 K. Results on coated (Os, Ir, Pt, W) and uncoated cathodes with 45% porosity matrices are presented and compared with an iridium-coated cathode with a 20% porosity matrix.

An examination is made of the intensity of light scattered nearly backwards by spherical dielectric particles illuminated by natural light. It is shown that, for particles larger than the wavelength, the intensity varies with direction in a characteristic way, agreeing well with an approximate theory at small angles.