Table of contents

The Boltzmann transport equation for a pulsed source of neutrons in a non-multiplying medium is investigated by means of the ansatz ϕ(E, Ω) exp (iB.γ-λt). The resulting eigenvalue equation for λ is considered in the multi-group approximation.

The existence of a discrete decay parameter exceeding (νΣ(E))_min, identifiable as the `continuation' of the fundamental time eigenvalue λ₀, is demonstrated. A rapidly converging iterative scheme for computing the eigenvalues greater or less than (νΣ(E))_min is described.

The calculated and experimental values of λ₀ are compared for various polycrystalline moderators over a wide range of buckling. The analytic continuation of the dispersion law and the variation of λ₁ with buckling are also demonstrated.

The d.c. and a.c. behaviour of evaporated thin CeO₂ films sandwiched between metal electrodes is explained by the presence of high resistance barriers at each electrode-CeO₂ interface. In the case of Al electrodes the formation of an Al₂O₃ layer due to solid-state electrolysis accounts for electrical behaviour. For other metal electrodes Schottky-type depletion regions appear to govern electrical properties.

At room temperature high-purity synthetic single crystals of proustite were found to conduct electricity predominantly by motion of Ag⁺ ions; their transport number is 1·02±0·02. Evidence from photoconductivity observations shows that electronic conduction can also occur. The lattice dielectric constant is 21±1 independent of orientation at room temperature; at low frequencies higher dielectric constants are measured, due, it is thought, to space-charge effects involving ionic conduction and rate-limiting electrode processes. Effects at contacts and in the bulk during conduction of direct currents are described.

The elastic and piezoelectric constants of single crystals of rubidium dihydrogen arsenate grown in H₂O, and of rubidium dideuterium arsenate grown in D₂O, have been measured in the temperature range 20-80°C. The shear elastic constants S₄₄ and S₆₆ and the extensional elastic constants S₁₁ and S₃₃ increase slightly on deuteration while their temperature coefficients decrease slightly. The piezoelectric constants d₁₄ and d₃₆ increase by 48% and 61% respectively on deuteration at 25°C. For normal RbDA the elastic constants measured at 25°C are: S₁₁ = 22·8 × 10⁻¹² m² N⁻¹, S₁₂ = 8·5 × 10⁻¹² m² N⁻¹S₁₃ = −1·8 × 10⁻¹² m² N⁻¹, S₃₃ = 25·4 × 10⁻¹² m² N⁻¹S₄₄ = 96·0 × 10⁻¹² m² N⁻¹, S₆₆ = 232·0 × 10⁻¹² m² N⁻¹. The piezoelectric constants are d₁₄ = 9·9 × 10⁻¹² C N⁻¹ and d₃₆ = 28·5 × 10⁻¹² C N⁻¹.

The electro-optic and dielectric constants of single crystals of rubidium dihydrogen arsenate grown in H₂O, and of rubidium dideuterium arsenate grown in D₂O, have been measured. On deuteration the relative dielectric constant K₃₃ is increased from 28·5 to 42·5 and K₁₁ is increased from 54·5 to 71·5 at 25°C. The half-wave retardation voltage V_λ/2, measured at 500 nm is decreased from 3·8 kv to 3 kv and the electro-optic constant r₆₃ is increased from 16·7×10⁻¹⁰ cm v⁻¹ to 21·6×10⁻¹⁰ cm v⁻¹ on deuteration. The loss tangent tan δ₃₃, measured at 1 kHz and at 25°C, is about 2·6 for D-RbDA and 5·2 for RbDA. The specific resistivity ρ₃₃ for RbDA is 3·4×10⁶ Ω cm and that for D-RbDA is 3·6×10⁷ Ω cm, measured at 25°C. The input power requirement using these crystals in an electro-optic modulator is compared to that of KDP.

Measurements have been made of the small signal gain of CO₂-N₂-He laser amplifiers excited either by short pulses or a 50 Hz rectified a.c. waveform. A model, based on the known vibrational energy-level scheme appropriate to this laser, is developed and applied to the data. The relaxation times of the upper and terminal laser states, the relaxation rate of the first vibrational state of nitrogen and the nitrogen to carbon dioxide energy transfer rate are determined. Gas heating which limits laser gain is described in terms of a phenomenological thermal time constant, and the value of this quantity is deduced from the a.c. data under various conditions.

The data presented permits the excitation conditions for optimum amplification of pulses from a Q-switched oscillator to be determined under non-saturating conditions.

Holograms have been successfully recorded on sensitized lead iodide layers by using the 4880 Å radiation from an argon ion laser. These holograms require no processing and a reconstructed image can be observed immediately after exposure. The properties of sensitized iodide layers are compared with those of conventional photographic emulsions and it is suggested that they may have advantages for certain holographic applications.

A modified integrating sphere, with a ceramic fibre lining, was used to determine the diffuse reflectivities of various materials such as paints, metals, brick, slate and tree bark in the 0·8-2·5 μm spectral band. The absorptivity and emissivity normal to the specimen can also be deduced from these results.

Values of the Townsend coefficients α and ω/α have been obtained over the range 100 < E/p0 < 510 v cm⁻¹ torr⁻¹ for contaminated nitrogen samples; the contaminants used were the radioactive isotopes tritium and krypton-85. Two impurity concentrations were used for each contaminant: 6 μCi cm⁻³ and 60 μCi cm⁻³ respectively. Values obtained for the primary coefficients in these contaminated samples were significantly higher than those in pure nitrogen. This increase in α/p was attributed to the preferential ionization of long-lived metastable states of nitrogen, which were produced by the decay radiations of the included isotopes. This conclusion was supported by the breakdown of the similarity principle found to occur for Paschen curves during the present work. No significant difference in the secondary coefficients for the pure and contaminated nitrogen was observed.

From measurements of pre-breakdown currents across plane-parallel electrodes, and with the aid of a simple analysis developed here, ionization and attachment coefficients have been evaluated for 50 < E/p0 < 250 v cm⁻¹torr⁻¹. Values of the attachment coefficient are found to reduce to practically zero at about 70 v cm⁻¹ torr⁻¹, which could indicate a detachment process. Sparking potentials were also measured in the range 0·1 < p0d < 5 cm torr, and this enabled the evaluation of generalized secondary coefficients.

The development of plasma parameters in a spark discharge is studied using spectroscopic techniques. It is observed that a transition from a molecular glow discharge to a fully ionized arc channel plasma takes place in the discharge column after initiation of a radial shock wave. Finally the transition from a γion process to a cathode spot mechanism initiates an axial shock wave of Mach number M=1·38 moving through the fully ionized channel from cathode towards anode.

The theoretical d.c. current-voltage characteristics of a `collisionless' alkali plasma diode are derived. It is assumed that electrons and ions are created only at the two incandescent plane electrodes, the latter being at the same temperature. The electrons and ions are assumed to leave the electrodes with half-Maxwellian distributions of velocities. Some ions are trapped in the potential well between the electrodes.

It is found that the potential distribution is associated with a double sheath in front of the negative electrode, a plasma region of constant potential and an electron sheath at the positive electrode. The theoretical (I, V) characteristics are compared with experimental results obtained with a caesium plasma. At low applied voltages good agreement is obtained between theory and experiment; at higher voltages (where oscillations occur) the d.c. current falls to a lower value. Possible mechanisms for this drop are described.

The extraction of ions from the plasma of a gas discharge and their formation into a beam is investigated theoretically. The results are shown to be applicable to a gridded ion gun. The results are also applicable to the design of canal extraction systems, but it is shown that in this case a simpler method exists for obtaining the necessary information. The theory is compared with the results of experiments and some agreement is found for sufficiently high accelerating voltages. The discrepancy is believed to be the result of the partial neutralization of the beam by secondary electrons in the experiments.

The experimental evidence for and against the existence of rheological abnormalities in liquids within about 1 μm from a solid boundary is critically reviewed. It is significant that, in those experiments where great care was taken to exclude dust, no anomalous effects were found. Work by the present authors has demonstrated conclusively that the apparent rigidity of thin films of kerosene, which has been reported by previous workers, is due solely to dust particles. It is concluded that rheological abnormalities are now known to extend no further than a few molecular diameters from a solid boundary in pure liquids and, whilst the situation in liquids containing surfactants is still a matter of controversy, there is no sufficient reason yet to believe that rheological abnormalities extend any further in these liquids.

In the course of fast reactor safety studies various problems associated with the transient boiling of liquid metals have arisen. Some of these of general interest, associated with nucleation, bubble growth and channel voiding, are discussed here. The differences between the behaviour of liquid metals and water are emphasized, and in particular the effects of dissolved gas and ionizing radiation are dealt with. An approximate method of calculating the rate of growth of a vapour bubble is described, which includes the effects of inertia, heat transfer and surface tension. This is applied to the voiding by boiling of a blocked reactor coolant channel.

Measurements are presented of the variation of stress and strain in metals as they are subjected to shock impulses of up to 200 kb and released. Results have been obtained for a magnesium alloy and for two grades of aluminium of very different yield strengths. The release characteristics of these metals are shown to be broadly consistent with a simple elasto-plastic model for metal behaviour. Elasto-plastic Hugoniots are fitted to published shock data for the metals used; a revised calibration is given for the manganin stress transducer employed, and a new method of construction is illustrated. Stress relaxation is observed in magnesium and is associated with the easy slip on the basal plane in hexagonal close-packed materials.

A viewing grid with equally spaced parallel lines (parallel to the y axis) combines with its mirror image or its shadow on the surface to form moiré patterns. Under the condition that the x-coordinate distance of the eye is much larger than the width of the viewing grid, the moiré patterns approximate to the equal-elevation lines of the surface. The general equations of the moiré patterns from a surface of known topography are given in both cases for the first time and are confirmed by experiment. On the other hand, from the experimental shape and spacing of the moiré patterns of an unknown surface, the topography of the unknown surface can be deduced.

The requirement that both the radial and axial components of the thermal conductivity of a rod be measured at temperatures in excess of 850°C with a reasonable degree of accuracy is the primary reason for the work which is described in this paper. The basic technique of Powell and Schofield was used for the radial heat-flow measurements. However, refinements in the measurement of small temperature differences and their implication so far as the experimental equipment is concerned, plus the extension of the technique to good thermal conductors, are discussed in some detail. The best available technique for the measurement of the thermal conductivity along the axis of a rod is that described by Jain and Krishnan. Additions to the technique as described originally wherein a pair of furnaces are used, and some of the necessary data required for the application of the Jain and Krishnan method, which are generated in the course of the radial heat-flow measurements, are described. The changes and modifications which are described permit improved accuracy and increased sensitivity for both methods of measurement.

A modified form of the Nelder and Mead simplex hill-climbing technique has been applied to the calculation of n, k and d for thin metallic films from measurements of R_p, R_s, T_p and T_s. In order to obtain consistent results it was found necessary, in addition, to include an error term in the objective function so that the known approximate value of d could be used in the calculations.

The silver chloride-silver iodide equilibrium diagram has been constructed. The eutectic temperature is 259°C and the eutectic composition is 47 mole% silver chloride. The mean thermoelectric potential of this mixture between 260 and 960°C has been measured as -0·498 mv degK⁻¹. Values of electrical conductivity of silver chloride-silver iodide mixtures are available in the literature, and the mean specific conductivity of the eutectic mixture between 260 and 960°C has been estimated as 2·91 Ω⁻¹cm⁻¹. The thermal conductivity has been measured as 1·83 × 10⁻³ W cm⁻¹ degK⁻¹ at 300°C. These values yield an approximate mean figure of merit for thermoelectric generation between 260 and 960°C of 0·4 × 10⁻³ degK⁻¹. This is a threefold increase over the estimated figures of merit for the pure components.