Index of content:
Volume 25, Issue 1, 01 January 1954
25(1954); http://dx.doi.org/10.1063/1.1721489View Description Hide Description
The thermoelectric powers and dc and 200‐mc resistivities of pure and metal‐doped selenium have been studied as functions of temperature. The properties of materials in the two important microstructures (colony and equiaxed) were distinguished. The influences of the metals Na, Cu, Ag, Mg, Zn, Cd, Hg, Ga, In, Tl, Pb, Sb, Bi, Te, Fe, Ni, Co, and Ce were noted.
A working model of the selenium semiconductor was developed. In this model there is an acceptor level <0.15 ev from the full band caused by nonmetals and possibly lattice defects. In strained crystals or material in the colony structure a second acceptor level 0.27 ev from the full band is postulated from the behavior of dc and ac resistivity curves as a function of temperature. The density of the first level varies in the range 1014/cm3 to 6×1016/cm3 with amount of nonmetal impurity. The density of the second level depends on the crystal structure, decreasing as the colony structure recrystallizes to the equiaxed. In the specimens crystallized at 110°C the density was about 2×1016/cm3. Both metals and non‐metals are largely segregated at grain boundaries. Metals produce effective donor levels at these places.
The dcproperties reflect the presence of internal barriers produced by the segregation. The donor levels enhance barriers existing in the purest selenium tested increasing resistivities and thermoelectric powers.
With increase in temperature >160°C the density of acceptor levels decreases as defects are removed by structural changes in the lattice.
25(1954); http://dx.doi.org/10.1063/1.1721494View Description Hide Description
25(1954); http://dx.doi.org/10.1063/1.1721514View Description Hide Description
25(1954); http://dx.doi.org/10.1063/1.1721515View Description Hide Description
25(1954); http://dx.doi.org/10.1063/1.1721517View Description Hide Description
A magnetically confined sheet electron beam filling the space between two plane parallel electrodes is discussed. It is assumed that the average electron space charge is not neutralized by positive ions so that the potential at the center is depressed, and the beam is inhomogeneous in velocity and charge density. The objective of this theoretical treatment is to determine whether or not exponentially growing space charge waves can be supported by such a beam.
For small inhomogeneity and continuous behavior of all quantities as functions of the inhomogeneity, it is shown that growing waves are not possible. This conclusion is supported by discussion of the analogous adjacent beam problem and the analogous velocity distributed filamentary beam. Although the adjacent beams can support growing waves, it is argued that the analogy is not valid because the adjacent beams cannot be obtained by a continuous perturbation of a single homogeneous beam. The analogy of the velocity distributed beam does not suffer from this deficiency. In this case it is shown that growing space charge waves cannot exist unless the distribution function has more than one relative maximum.
On a Property of a Family of Equiangular Spirals and its Application to Some Problems of Wave Propagation25(1954); http://dx.doi.org/10.1063/1.1721518View Description Hide Description
It is shown that all the normals drawn from any point in the plane to the members of a family of equiangular spirals may be simply located geometrically. The locus of all the end points of the normals is a circle. It is also shown that the standing wave pattern in a uniformly attenuating medium may in simple cases (as with an attenuating transmission line) be represented by a vector diagram using a symmetrical pair of spirals. Then the foregoing theorem makes it possible to determine the voltage and current maxima and minima by constructing the locus circle referred to. This is illustrated by several examples. Finally it is pointed out that the locus circle may also be projected into the ``Smith Chart.''
25(1954); http://dx.doi.org/10.1063/1.1721519View Description Hide Description
A numerical method is presented for describing either the one‐dimensional or radial transient flow of gases through a porous medium in which the initial and terminal pressures and/or rates are specified. The purpose of this investigation was to evaluate the effect on transient gas flow of a pressure‐sensitive gas permeability resulting from gas‐slip effects. The gas‐slip permeability or apparent permeability is assumed to vary as a linear function with the reciprocal pressure. The computations were carried out by means of punch card machines.
The results demonstrate that under certain conditions, the transient flow system is greatly influenced by gas‐slip permeability. However, the results also indicate that the effect of gas slip on transient flow behavior is negligible for pressure conditions usually associated with gas or petroleum reservoirs, and would be of most importance in the conduct and analysis of gas flowexperiments in the laboratory. A specific laboratory application of the calculations is a possible transient method for the determination of porosity and/or gas‐slip corrected permeability.
25(1954); http://dx.doi.org/10.1063/1.1721520View Description Hide Description
Taylor's analysis of the intense spherical explosion has been extended to the cylindrical case. It is found that the radius R of a strong cylindrical shock wave produced by a sudden release of energy E per unit length grows with time t according to the equation , where ρ0 is the atmospheric density and S(γ) is a calculated function of the specific heat ratio γ. For γ=1.4, S(γ) is found to be approximately unity. For this case, the pressure p 1 behind the shock wave decays with radius R according to the relation p 1=0.216E/R 2. Applying the results of this analysis to the case of hypersonic flight, it can be shown that the shock envelope behind a meteor or a high‐speed missile is approximately a paraboloid given by where D and V denote the total drag and the velocity of the missile, respectively, and x is the distance behind the missile.
25(1954); http://dx.doi.org/10.1063/1.1721521View Description Hide Description
A modified Ångström method for measuringthermal diffusivity and hence thermal conductivity of metals has been developed. Like previously reported dynamic methods, this method uses a heat source, whose temperature varies sinusoidally with time, located at one end of an effectively infinite rod. Unlike these methods, only one period of the heat pulse is required to eliminate the unknown coefficient determining the heat lost by radiation since both velocity and amplitude decrement of the heat pulse are measured. In addition to providing greater reliability at high temperatures by using only one period, the new method is faster in taking data and simpler in computation. The thermoelectric potentials from two thermojunctions are amplified and plotted on a Brown electronic recorder in order to obtain a permanent record of all necessary data for computing the thermal diffusivity. Results for copper,nickel, and thorium over the temperature range 0–500°C are given.
Brightness Waves and Transitory Phenomena in the Quenching of Luminescence by Alternating Electric Fields25(1954); http://dx.doi.org/10.1063/1.1721522View Description Hide Description
Zinc or cadmium sulfide phosphors, especially phosphors having short persistence, are less sensitive to ultra‐violet light or to x‐rays when they are submitted to the action of an alternating electric field. With a half‐wave x‐ray generator and an alternating field of frequency equal to the frequency of the potential applied to the x‐ray generator (50 cps), and in phase with it, the brightness waves (a curve giving the variation in the instantaneous brightness as a function of time in the interval of a period) are greatly modified. A sharp secondary maximum appears when the field passes through zero in each cycle; the amplitude of this maximum varies quickly during the cycles which immediately follow the application of the field.
If the light is filtered, the shape of the brightness wave varies according to the spectral region which is observed.
Direct Determination of the Flow Curves of Non‐Newtonian Fluids. III. Standardized Treatment of Viscometric Data25(1954); http://dx.doi.org/10.1063/1.1721523View Description Hide Description
A scheme is proposed for systematizing the handling of viscometric data through use of the variables apparent fluidity φ a and shearing stress F. The rate of shear g(F) is obtained from the relation: g(F)/F =φ a [1+Δ(F)]. Methods are presented for obtaining the correction term Δ(F) from capillary and concentric‐cylinder viscometric data. Use of the proposed scheme is illustrated.
25(1954); http://dx.doi.org/10.1063/1.1721524View Description Hide Description
The purpose of this paper is to attempt to present known results on detection in simplified form and to relate these to new results found by the author. For a signal of known form and nonrandom in nature (sure signal) upon which a background noise is linearly superimposed, several criteria for an optimum pre‐detection filter are compared. The background noise may be of very general nature and restricted merely to possessing a continuous covariance function and need not, in general, possess even a frequency spectrum. Specifically, it is shown that, when the input noise to a pre‐detection filter is Gaussian (with no restrictions concerning its possible spectrum), the filter which maximizes—for a fixed ``false alarm'' probability—the probability of detecting the signal when it is present is identical with the linear filter which maximizes the output signal‐to‐noise ratio. For a Gaussian noise input an explicit expression is obtained for the probability of detection. The stability of the optimum filter is discussed in some detail and some restrictions on the form of the input signal necessary for stability are given. The techniques used throughout limit one, in general, to the practical case of finite memory filters. Finally it is shown how an optimum signal shape of given energy content and duration can be chosen.
25(1954); http://dx.doi.org/10.1063/1.1721525View Description Hide Description
An expression for the slowing‐down length of neutrons in a hydrogenous mixture in which inelastic collisions occur is derived from the rigorous transport equation under the assumptions that (1) the masses of the nonhydrogen nuclei are infinite and (2) the inelastic scattering is of the type in which only discrete well‐defined energy levels of the target nucleus are involved. The result is applied to an iron‐water mixture.
25(1954); http://dx.doi.org/10.1063/1.1721526View Description Hide Description
A technique is described whereby fall‐out particles from dust clouds, commonly called aerosols, may be prepared for observation with the electron microscope. Specimens are collected by means of various types of precipitators on molecular filters which are made from cellulose esters. During the process of dissolving this material in acetone, the particles adhering to the filter are transferred to the Formvar‐coated specimen screens of the electron microscope. Losses are negligible, residual filter background is eliminated, and particles are clearly visible for measuring.
25(1954); http://dx.doi.org/10.1063/1.1721527View Description Hide Description
A method for measuring the velocities of explosive‐produced air shocks using an electrical ionization probing system is described. The results of such measurements indicate that under varying conditions of confinement of explosive and air shock, that the reciprocal of the air‐shock velocity varies linearly with the distance to the charge surface.
25(1954); http://dx.doi.org/10.1063/1.1721528View Description Hide Description
Torsion tests are made on wires of various ferromagnetic materials. It is found that the relationship between shear stress and shear strain is different for magnetized and unmagnetized wires. The difference is explained in terms of wall movement and rotation of ferromagnetic domains, and magnetostriction values are derived from it. This method of determining magnetostriction has several advantages. It is simpler to carry out than most other methods, its accuracy is 4×10−7, and it can be used at elevated temperatures. Its main limitation is that saturation magnetostriction values are not obtainable for alloys which have a large crystal anisotropy energy.
25(1954); http://dx.doi.org/10.1063/1.1721529View Description Hide Description
The conditions for the stability or instability of the interface between two immiscible incompressible fluids in radial motion are deduced. The stability conditions derived by Taylor for the interface of two fluids in plane motion do not apply to spherical flows without significant modifications.
25(1954); http://dx.doi.org/10.1063/1.1721530View Description Hide Description
The exact solutions to the problems of the diffusion of minority carriers involved in the measurement of surface‐recombination velocity in a semiconductor with a sample geometry bounded by two infinite planes are presented. The reduction of the exact solutions to simple forms used in the analysis of experimental data is shown.
25(1954); http://dx.doi.org/10.1063/1.1721490View Description Hide Description