Index of content:
Volume 29, Issue 2, 01 February 1958
29(1958); http://dx.doi.org/10.1063/1.1723051View Description Hide Description
Measurements have been made for quench induced residual stresses in a ring of bearing steel. The stress distributions vary along the radial direction r and the axial direction z. The mean values of the axial stresses depend on r only and were obtained previously. The deviations of the residual stresses from the mean as a function of z are found at two radii. An interferometric method is used to detect the deflections of the ring, and the spatial variation is obtained by etching one surface in uniform steps. The stress distributions obtained had typical residual stress shapes, and were compression near the surfaces and tension in the middle. In a ring the mean values do not vanish, but the deviations from the mean are found to be small compared to the mean values. This indicates that the radial temperature gradient dominates in determining the stress distribution, regardless of the size of the axial temperature gradient.
29(1958); http://dx.doi.org/10.1063/1.1723052View Description Hide Description
``Dislocation Lichtenberg figures'' have been discovered on the surfaces of LiF crystals subjected to high inhomogeneous electric fields. The figures consist of thousands of dislocation loops arranged along  dendritic branches which are terminated by  branches. Etching is necessary to reveal the figures.
It is shown that electric fields can nucleate dislocations in LiF crystals.
29(1958); http://dx.doi.org/10.1063/1.1723053View Description Hide Description
Most previous treatments of long high‐density electron beams have assumed laminar flow. Recent experiments show that in many cases this assumption is unjustified. The present theory is based on a non‐laminar optical model which treats thermal velocities as an integral part of the motion. A Maxwellian distribution of initial transverse velocities is assumed at the cathode, and a first‐order focusing theory is applied in order to calculate trajectories at any point in the beam. It is shown that whenever a long beam is confined by a focusing field, images of the cathode are formed repeatedly along the axis.
When applied to uniform magnetic focusing fields, the theory predicts the periodic formation along the axis of cathodeimages and crossovers, and a relative rotation of successive images. Such effects have been reported. Values of required focusing fields are derived for given space‐charge and thermal‐velocity conditions, and relations are established between beam spread and transverse velocities. The resulting picture differs radically from the laminar model of beam flow.
29(1958); http://dx.doi.org/10.1063/1.1723054View Description Hide Description
A microwave absorption line of O2 near 60 000 Mc has been experimentally investigated as a possible atomic frequency standard of high precision. Theoretical design, actual design, and results for an oxygen microwave spectrometer for use either in observing the line frequency or as a discriminator in a frequency control loop synchronizing an oscillator are described. Essential characteristics are the rate of change of spectrometer output signal with frequency and the output noise level. General expressions for these quantities for a wide range of experimental arrangements are obtained, and may be used to predict the attainable frequency precision. Characteristics in complete agreement with the theoretical values, capable of precision to less than 1 part in 109, were experimentally observed with relatively crude apparatus. These characteristics involved line widths of around 560 kc and signal/noise ratios of 83 db. Improvement by a few orders of magnitude can be confidently expected by engineering refinements. An additional result was close confirmation of the theoretical absolute intensity of the N +=7 line and a precision measurement of its frequency as (60 434.70±0.05) Mc.
29(1958); http://dx.doi.org/10.1063/1.1723055View Description Hide Description
Preliminary measurements at 40 kc of the internal friction of 99.992% aluminum as a function of temperature yield a number of attenuation maxima at temperatures between 90°K and 300°K. These maxima are unlike the cold‐worked induced peaks, first reported by Bordoni. They appear to be of two types, a broad maximum at about 225°K and a series of very sharp peaks at lower temperatures. The latter are tentatively associated with the break away of dislocations from minor pinning points.
29(1958); http://dx.doi.org/10.1063/1.1723056View Description Hide Description
The rate of change of conductivity of n‐type single crystal plates of germanium at a temperature of ∼16°K during fast neutron irradiation is enhanced over that previously observed at higher temperatures. Annealing studies subsequent to irradiation reveal no evidence of thermally unstable defect states between ∼10 and ∼95°K. Annealing above ∼95°K indicates the presence of thermally unstable minority carrier traps of the type previously observed following irradiation at 120°K. The enhanced decrease in conductivity at 16°K compared to that observed at higher temperatures is believed to reflect a greater rate of introduction of defects, many of which anneal at temperatures above ∼95°K.
29(1958); http://dx.doi.org/10.1063/1.1723057View Description Hide Description
The Decker, Asp, and Harker technique and the Schultz technique for obtaining pole figure x‐ray data in the transmission and back‐reflection regions, respectively, are shown to give sufficient data for a complete pole figure, with consistent intensity contours in a 15° annulus of overlap. An improved integrating specimen holder, based on the design of Geisler, was found to perform satisfactorily for transmission and back‐reflection operation.
A powder method for making all‐metal sheet specimens with randomly oriented crystals is described. The use of this type of random sample for establishing background and 1X intensity levels for a corresponding textured specimen is shown to give results which are only semiquantitative when normal filtered radiation is used.
29(1958); http://dx.doi.org/10.1063/1.1723058View Description Hide Description
The Hall effect is shown to be well fitted for electronic multiplication of voltages—one of the most cumbersome operations in analog computation. The use of this effect yields a very simple instrument with good accuracy and speed. The rise time—corresponding to a 0.2% error in amplitude—for the response of a step voltage of maximum value (100 volts) applied to the slower of the two inputs, can be kept down to about 1 msec. This time is dictated only by the feeding circuits and is linearly dependent on the amplitude of the step. Thus a step of 5 volts gives a rise time of 50 μsec. Most semiconductors are investigated with respect to their capabilities of accurate multiplication. Silicon is found to be the best choice. The choice of dimension of the semiconductor crystal is discussed from the solution of a potential problem with skew boundary conditions which determines the Hall voltage. The output of a silicon crystal may be as high as about 0.5 volt for an applied magnetic induction of 5000 gauss, a current through the crystal of about 5 ma, and with a simple thermal contact between the crystal and the core of the magnet, without having errors greater than 0.1%. The output voltage is amplified with a chopper amplifier to 20 volts.
29(1958); http://dx.doi.org/10.1063/1.1723059View Description Hide Description
The explosive production of oblique shock waves in solids is analyzed in the approximation that third and higher order terms in the shock strength can be neglected, and a procedure is developed for calculating the attenuation of the shocks. Application is made to the problem of determining the equation of state of the burned explosives gas.
29(1958); http://dx.doi.org/10.1063/1.1723060View Description Hide Description
Dislocations introduced into single crystals of silicon by plastic bending at an elevated temperature have been studied quantitatively by the etch‐pit technique. The average etch‐pit density after deformation is approximately two to three times higher than the calculated dislocation density. Prolonged annealing of the deformed samples at a temperature close to the melting point results in polygonization or the alinement of etch pits perpendicular to the active slip planes. Satisfactory agreement was obtained between the densities of etch pits in the annealed and polygonized specimens and the predicted dislocation densities. Piled‐up horizontal arrays of etch pits have been observed in lightly bent samples. The value of yield stress obtained from the geometrical distribution of these arrays is an order of magnitude below the reported experimental values. Dislocation lines in single crystals deformed at 1000°C have been revealed by the technique of copper decoration. Examination of the decorated samples by infrared transmission microscopy revealed that the dislocation lines lie in the  direction.
29(1958); http://dx.doi.org/10.1063/1.1723061View Description Hide Description
Measurements are reported on noise in germaniumsingle crystals at temperatures between 300°K and 450°K. In this range the crystals were near‐intrinsic or intrinsic. The 1/f noise component was negligible in all cases above 1 kc providing the noise was measured between probes. The noise above thermal noise measured in this way could entirely be attributed to spontaneous fluctuations in generation and recombination of carriers. From the spectra lifetimes varying from 2 μsec for poorly etched crystals up to 100 μsec for the best etched crystals were found. The results are compared with the theory considering generation and recombination by means of direct transitions or via recombination centers located either in the bulk or at the surface. Very reasonable agreement is found in most cases. In some cases the spectrum falls off rather weakly at higher frequencies which indicates that the recombination centers do not lie at a single sharp energy in the forbidden gap but are distributed within a small range.
Method for Measuring Magnetostriction Corrected for Initial Domain Distribution and its Application to Nickel and Iron29(1958); http://dx.doi.org/10.1063/1.1723062View Description Hide Description
A method has been developed for the measurement of magnetostrictive strain that overcomes the effect of initial domain distribution in isotropic materials. A cubic specimen is used, placed between two iron rods of the same cross section. The combination is put along the common axis of two like solenoids connected in series. Strain is measured with a resistance strain gauge, once parallel to the magnetic field, and a second time normal to the field after the specimen has been rotated 90°. At saturation the algebraic difference gives 3/2 of the saturationmagnetostriction. Use of the difference eliminates or reduces a number of experimental errors in addition to that arising from nonrandomness of initial domain distribution.
The method has been applied to grade A nickel and armco iron.
29(1958); http://dx.doi.org/10.1063/1.1723063View Description Hide Description
X‐ray and electron diffraction and electron microscopy have been used to study the composition and distribution of the precipitated phases in an aluminum—4% silicon—1% cadmium alloy as a function of thermal history. Thin precipitated crystallites are detected in extraction replicas of specimens tempered above 400°F and have been identified by selected area electron diffraction as elemental silicon. The amount of silicon remaining undissolved after solution treatment at 1000°F can easily be distinguished from that precipitated upon tempering, and the amount of silicon observed to dissolve and reprecipitate after solution and tempering treatment is in agreement with reported equilibrium data. The aluminum oxide replica was found to be unsuitable for this investigation. A direct carbon replica technique, however, was found to be satisfactory. The extent of precipitation of the minor phases as a function of tempering temperature, as determined by quantitative x‐ray diffraction studies, was found to correlate with a mechanical property of the alloy.
29(1958); http://dx.doi.org/10.1063/1.1723064View Description Hide Description
It is shown that the transit time in a photoconductor cannot be less than the charge relaxation or storage time if the photoconductor has contacts of the ``space charge,'' or ohmic, type. The principal consequence of the observation is that the photoconductor cannot simultaneously act as a detector, an amplifier, and a storage element and still have a response time as short as the storage time. This puts a restriction on the performance of high‐resistivity photoconductive devices. For example, with present television standards and acceptable response times, the maximum quantum efficiency possible for a photoconductive camera tube of the vidicon type with space charge limited dark current is less than unity. [Note added in proof.—These same general conclusions have also been reached by A. Rose, Helv. Phys. Acta 30, 242–244 (1957)].
29(1958); http://dx.doi.org/10.1063/1.1723065View Description Hide Description
The formula for the pressure‐drag coefficient of a sphere which moves at extremely high speeds is derived. In the derivation it is assumed that the nose of the shock contour follows exactly the frontal half of the spherical surface and the local pressure on the frontal spherical surface corresponds to the pressure behind the shock wave after the statistical equilibrium between the various degrees of freedom of the molecule has been reached. The chemical dissociation of the molecules behind the shock wave is taken into account in the analysis. The theoretical results of the drag coefficient of the spheres agree with the corresponding available measured value within the experimental error for the range of Mach number between 5 and 10.
29(1958); http://dx.doi.org/10.1063/1.1723066View Description Hide Description
Filings from a molybdenum rod, pressed into a briquet, were used as the cold worked sample. Satisfactory peak shapes for 110, 200, 211, 220, and 310 were obtained with filtered CuKα. Peak shapes for 400 and 422 obtained with filtered MoKα were not satisfactory, partly because of fluorescence radiation. Peaks were recorded with a Norelco diffractometer, and instrumental broadening corrected by the Stokes method. The Fourier coefficients indicate approximately isotropic incoherent domains L=260 A, and strains inversely as Young's modulus.Molybdenum is similar to tungsten in that there is no evidence for faulting on (211) as a result of cold work.
29(1958); http://dx.doi.org/10.1063/1.1723067View Description Hide Description
Cation distribution in many ferrites has been difficult to determine by x‐ray diffraction since the scattering factors of such cations as Ni, Co, Mn, and Cu are close to that of Fe. By using radiation of a wavelength close to the absorption edge of one of the cations, as suggested by Bertaut, sufficient difference in x‐ray scattering factors can be developed because of anomalous dispersion, thus allowing the cation distribution to be ascertained. This method was employed in the case of nickelferrite.
As a modification of Bertaut's technique, ratios of intensities of the diffracted peaks for FeKα and FeKβ radiation from nickelferrite were used to eliminate certain factors in the intensity equation which are difficult to calculate, for example, the absorption and temperature factors.
A least‐squares technique was used to determine the best values of the cation distribution and the oxygen position for the intensity ratios obtained for seven reflections. It was found that the cation distribution coefficient was 0.48±0.02, in agreement with the results obtained by magnetic‐moment measurements and neutron diffraction. The oxygen positional parameter was found to be 0.240±0.002 of the lattice parameter with an octahedral position at the origin, which is in good agreement with the value obtained using neutron diffraction.
The method appears to be of general value in distinguishing between two elements distributed over nonequivalent positions in a crystal lattice when their x‐ray scattering factors are almost equal.
29(1958); http://dx.doi.org/10.1063/1.1723068View Description Hide Description
The Born‐Green theory, employing the Kirkwood superposition approximation, is investigated as a means of calculating the intrinsic bulk viscosity—a bulk viscosity due solely to Van der Waals interactions—in a monatomic gas. In the framework of the Born‐Green Theory, it is found that the Kirkwood approximation does not introduce irreversibility and leads to homogeneous integral equations with no normalizing conditions. By the neglect of terms of the order of contributions from triple collisions, expressions for the bulk viscosity are derived in terms of other thermodynamic quantities which yield rough values of the order of one micropoise for argon and nitrogen at 0°C, 1 atmos.
29(1958); http://dx.doi.org/10.1063/1.1723069View Description Hide Description
The surface tensions of liquid Ti, Zr, and Hf at their melting points were measured in pure argon atmosphere by the drop‐weight method. A uniform longitudinal field induction coil was used to melt the sample rod tip. The average values obtained were 1390±40, 1400±40, and 1460±40 dynes/cm, respectively.
29(1958); http://dx.doi.org/10.1063/1.1723070View Description Hide Description
A study is made of the etch effects produced by sputteringgermanium crystals and bicrystals under normal incident low‐energy (100 ev) Hg+‐ion bombardment in a low pressureplasma (1 micron gas pressure). The surfaces show preferentially etched pits which readily reveal the crystal orientation. These pits become more abundant at lower target temperature and truncated at higher target temperature or at higher energy of the bombarding ions. A sequence of micrographs from the same surface area reveals that these pits cannot be correlated with the pits which are persistently developed by CP‐4. A 6° tilt boundary however always develops into a sharply indented furrow just as was the case with a CP‐4 etchant.