Table of contents

A generalized form of the equations of linear viscoelasticity is presented which describes the behaviour of a material whose inherent mechanical properties are changing with time as a result of ageing or of changing external environmental constraints. The equations are used to successfully predict recovery from stress relaxation measurements for a Lincoln wool fibre subjected to large step changes in relative humidity during the experiment.

Etching studies of KCl crystals doped with Ca²⁺ are carried out. These solution-grown crystals contained 1·44±0·02 weight percent calcium as estimated by EDTA titration. Establishing the etchants that attack at the sites of dislocations, and at sites other than those of dislocations, and comparing this etching behaviour with that of pure KCl crystals, it is shown that non-dislocation etch pits are at the sites of Ca²⁺ ions.

Practical MOS capacitance characteristics at high frequency are given at temperatures of approximately 100 °C below ambient and the results are compared with the ambient characteristics for steam-oxidized Si devices with an Au electrode. Distinction is made between devices cooled while in accumulation to those cooled while in inversion and a quantitative explanation for the low-temperature characteristics is given. When devices are cooled by not more than 60 °C below ambient the shift in the C-V characteristic can be used to determine, very simply, the density of interface states and a typical result is given from 0·2 eV from the conduction band edge to within 0·1 eV of the valence band edge. This result is compared with data taken using two other techniques.

A filter glass with unique spectral transmittance properties has been developed. The glass is composed from oxides of Te, W and Pb doped with CuO. It shows reasonable durability and can be used successfully in the photometric field. The sharp cut-off in transmittance towards longer wavelengths, the complete suppression in the near infrared and the width and symmetry of the bandpass offer exciting possibilities. It would now appear possible that a near perfect standard photopic distribution might be obtained by a monofilter.

(011) cleavages of single crystals of calcium tungstate, grown by the flux-evaporation method in this laboratory, have been subjected to thermal etching between temperatures of 1110 and 1430 °C in a vacuum chamber at a pressure of 10⁻⁵ Torr. The etch figures thus obtained successfully reveal the emergence sites of dislocations in the crystals. An excellent correlation between thermal and chemical etch pittings has also been obtained on mirror cleavages. Quantitative measurements made on the etch rate and weight loss with temperature show that the process of the removal of ledges from the cleavage surface has an activation character. The activation energy for evaporation at dislocation sites is less than that responsible for overall evaporation from the exposed surface. The implications are discussed.

This paper outlines the difficulty of resolving two overlapping symmetrical peaks of the same unknown shape as might be seen in x-ray or electron diffraction studies of epitaxial layers.

Stress-strain tests have been carried out on continuous fibre-reinforced copper-tungsten composites at a strain sensitivity of 10⁻⁶. The material was prepared by the combination of filament winding, electrodeposition and hot-pressing using fibres between 10 and 50 μm diameter and volume fraction in the range 0·1-0·6.

The composite elastic modulus, microyield stress, 0·1% proof stress, stress at 1% total strain and fracture stress, all exceeded the values expected from the rule of mixtures, based on the properties of the tungsten wires and unreinforced copper prepared by a similar route to the composites.

The difference between the experimental and rule of mixtures modulus values is an order of magnitude greater than predicted by elastic constraint theory, but no satisfactory explanation could be found for this effect.

All the remaining parameters which involved matrix plasticity (and fibre plasticity at high strains), showed positive deviations from the rule of mixtures which have been interpreted in terms of matrix strengthening. This strengthening arises from effects due to the matrix grain size, and also to the abnormally high dislocation densities in the matrix as a result of differential thermal contraction on cooling from the pressing temperature. These effects are only likely to be observed in composites containing small diameter fibres, where the interfibre spacing falls below about 50 μm.

The mechanical behaviour of carbon-fibre-composite specimens of fibre volume fractions from 30 to 60% and various fibre orientation has been investigated experimentally for compression loading up to 200 MN m⁻² and duration approximately 30 μs. A split Hopkinson pressure bar technique has been employed with an optical recording method to monitor the displacements of the end faces of the cylindrical specimen. For specimens loaded along the fibre direction the deformation is characterized by two approximately linear regions of different slope, and an empirical relation has been derived between volume fraction and the modulus obtained from the initial slope.

The impulse, surface velocity and mass of dense, porous or powdered tantalum or lead spalled by an electron beam have been measured for energy fluxes in the range 10-500 J cm⁻². The highest energy flux gives an energy deposition sufficient to vaporize the tantalum and similar 4 times the energy required to vaporize the lead surface. In dense tantalum and lead the measured impulse agreed to within similar 10% with that calculated by a computer program using a single-phase equation of state.

The impulses from porous lead and tantalum of similar 0·92 solid density were measured to be within 10% of the impulse from dense material and are predicted reasonably well by a (pατ) equation of state if relaxation times greater than 50-100 ns are assumed.

Impulses from powdered lead or tantalum of similar 0·5 the full density are similar 0·6 the fully dense impulse and are also compared with a (pατ) equation of state.

Thin layers of alkali tantalates can be produced on tantalum heater ribbons by a very simple and rapid electrochemical process. Such ribbons then act as efficient, directly heated, dispenser-type sources of positive alkali metal ions.

The preparation of the active layers is described and their ion emission characteristics are investigated.

Linear magnetic birefringence in ferromagnetic thin films in the Voigt configuration leads to intensity modulation of the transmitted radiation during hard-axis switching.

Errors in the classical theory proposed by Lissberger and Parker (1971) are corrected, and the revised theory is used to calculate the magnitude of this modulation and the magneto-optic parameters Q(=Q₁-iQ₂) and f(=f₁-if₂) used in the phenomenological description of this second-order magneto-optic effect. The results are, in general, at variance with previously reported values but are both self-consistent and in good agreement with the available experimental data.

Profiles of magnetic orientation have been calculated for long reverse domains with cylindrical 180° Bloch domain walls in uniaxial material subject to an applied field H just sufficient to equilibrate the domains against radial collapse. The dependence of the wall energy, bending moment, support field, wall width and other profile parameters on mean wall radius R have been investigated numerically. As these properties depend on the radial variation of the support field, two different simple cases have been treated: H=H₀ independent of radius R; and H=H₁/R. Fortunately the wall profiles are not too strongly dependent on the radial variation of H: in particular the wall energies differ by about 1% at most. The results should be applicable to curved walls whenever local demagnetizing fields are not expected to modify the Bloch character of the wall severely.

It is the purpose of this paper to establish the frequency dependence of magnetic domain size in 50% nickel-iron cores. The core noise caused by the fluctuations in every cycle of the minor dynamic hysteresis loop has been measured for the change of magnetizing frequency in the range 10-5000 Hz. On the basis of both the measured values and the equation obtained from a statistical model of the core noise, the dynamic domain size can be estimated as a function of magnetizing frequency. The results show that below a threshold frequency of 50 Hz the domain size is constant and independent of the frequency, that in the range 50-200 Hz it varies approximately as f^−1.5, and that above a frequency of 200 Hz it varies as f^−2·2; under the static condition of 10 Hz the domain size can be estimated to be 4·9×10⁻⁴ cm³; above about 1000 Hz the domain size becomes independent of the frequency again. There thus appear four regions in each of which the domain size shows a simple power-law dependence on the frequency of the magnetizing field.

The response of magnetic domain walls in yttrium orthoferrite (YFeO₃) to a pulsed driving field was observed by a stroboscopic technique. The results demonstrate that the present technique is very useful for the observation of high-speed domain wall motion, and is capable of achieving a time resolution of less than a few nanoseconds. Some interesting responses, eg high speed shrinkage of narrow strip domains, domain annihilation, and critical velocities were found in this experiment. The observed maximum velocity of the domain wall was higher than the Walker's critical velocity of YFeO₃ (about 3·6×10⁵ cm s⁻¹) predicted by Hagedorn. Moreover, critical-velocity-like responses were found at two velocities, about 4·5×10⁵ cm s⁻¹ and about 7·7×10⁵ cm s⁻¹. A critical-velocity-like motion of about 6·5×10⁵ cm s⁻¹ was found for the motion of the tip of narrow strip domains during pulsed field driving.

A detailed study of the radiation-induced optical absorption spectra in LiF: Mg: Ti has revealed many more absorption features than previous work had indicated. Thermal annealing and optical bleaching were used to follow the growth and decay of at least seven optical absorption bands in the near ultraviolet spectrum between 200 and 400 nm. It is apparent that in addition to the intrinsic F, F₂ and F₃ centres there are a variety of impurity centres in the form of F_z⁻ and Mg-related absorption bands. The identification of the bands and their role in the thermoluminescence process are discussed.

The effects of Mg and Ti impurity concentrations on the thermoluminescence from LiF are discussed. The results are consistent with a model that Mg ion defects liberate electrons and Ti³⁺ ion defects emit light in emission bands at 420 and 650 mm in the process of hole capture. A consideration of the detailed processes shows that both Mg and Ti form a variety of defect complexes, and in particular one should consider the role of other impurities such as oxygen.

The glow peaks are resolved by measurements of the emission spectra with emission bands occurring at 300, 420, 520 and 650 nm. The 300 nm emission involves an intrinsic defect. Emission at 520 and 650 nm is enhanced by irradiation at high fluxes. However, room temperature storage of the crystals reduces the 520 nm emission.

Methods are discussed for distinguishing between volatile species in a direct evaporation flux from those produced by various background interactions, using mass-spectrometric techniques. It is shown that while flux modulation methods are essential, phase-sensitive detection at a single frequency is inadequate, and that measurements must be made over a wide frequency range. The improvement over frequency domain methods which can be achieved using time representation of the pulsed signal with signal-averaging and fast Fourier transforms is illustrated. In addition, a means of minimizing mass discrimination effects in the mass spectrometer is described.

Results obtained by the application of various forms of data acquisition and signal processing to the Knudsen and Langmuir evaporation of InP, GaP and GaAs are presented, and some thermodynamic implications of the new vapour pressure values examined.

The evaporation of InP under Knudsen and Langmuir conditions has been investigated mass-spectrometrically using a modulated molecular beam technique combined with phase-sensitive detection and signal averaging to separate evaporating species from background species in the vacuum system. P₂ was found to be the dominant species over InP under Knudsen conditions but significant evaporation fluxes of In and P₄ were also detected. The P₄ flux was attributed to P₂+P₂ association on the oven walls. A congruent evaporation point under these conditions was predicted at 629±5 K. Under Langmuir evaporation conditions the only evaporating species detected were In and P₂. Up to 638±5 K the (100) surface evaporated congruently but above this temperature there was a disproportionate loss of P₂ from the surface. The equilibrium vapour pressures of In, P₂ and P₄ over InP and of In over pure In are presented and enthalpies for the reactions InP(s)->In(s)+½P₂ (g); 2P₂ (g)->P₄(g); In (s)-> In (g) calculated.