Volume 25, Issue 5, 01 May 1954
Index of content:
25(1954); http://dx.doi.org/10.1063/1.1721685View Description Hide Description
The scattering matrix of a 360‐degree circumferential slot in an infinitely long circular wave guide with an incident dominant circular‐electric wave is obtained by a variational principle. The theory which should by very good for small gaps is shown to be in good agreement with the obtained experimental results.
25(1954); http://dx.doi.org/10.1063/1.1721686View Description Hide Description
The problem of a vertical monopole situated over a circular perfectly conducting screen lying on a finitely conducting ground is considered. An approximate method employed originally by Abbott to calculate the self‐impedance is discussed. Using this formula explicit expressions are derived for the self‐impedance of a thin, top‐loaded vertical monopole. Sinusoidal current distribution is assumed.
25(1954); http://dx.doi.org/10.1063/1.1721687View Description Hide Description
Plastic deformations in beams caused by symmetric dynamic loads distributed over finite lengths are considered in this paper. The analysis is based on the assumption of ``plastic‐rigid'' behavior, so that the results should be valid when plastic deformations are sufficiently large. The final deformations for distributed loads are found not to differ much qualitatively from those to be expected for concentrated loads. However, the magnitude of the deformation decreases rapidly as the length of the loaded area increases from zero, for given load pulses. Thus the simplifying assumption of a mathematically concentrated load may cause a considerable overestimate of the final deformation.
25(1954); http://dx.doi.org/10.1063/1.1721688View Description Hide Description
That the noise convection current in a diode would be the same as the noise current in a triode under the conditions θ1≫1 and θ2≪1 (θ1 and θ2 are transit angles in radians of the potential minimum‐to‐grid and grid‐to‐plate regions, respectively) is shown by following the approach indicated by Rack and applying the Llewellyn‐Peterson small signal theory to study the noise in a triode. Although the general trend of the experimental results agrees very well with the prediction from the Rack approach, the measured value of noise power is found to be about 2 db below the predicted value. The experimental results are applicable to triodes as well under certain conditions.
25(1954); http://dx.doi.org/10.1063/1.1721689View Description Hide Description
Supersonic liquid jets were produced in air by means of a spring‐loaded injector. At high jet velocities another type of breakup seems to occur besides the classical Rayleigh surface tension breakup, and the sinuous aerodynamic breakup. Rotationally symmetric waves are formed, and appear to ``break'' analogous to wind‐produced waves on a body of water. The rate of cavitation of jets in water was studied for several jet velocities above the speed of sound in air.
25(1954); http://dx.doi.org/10.1063/1.1721690View Description Hide Description
The relativistic differential equations of motion of an electron in the field of a linear accelerator are considered. Treating only electrons near the axis, the field components are expanded in powers of distance from the axis r, and terms in r and ṙ of order two higher than the lowest order terms are neglected. The first integral of two of the equations of motion are then obtained, and together with the equation of motion for the r coordinate, give a second‐order nonlinear differential equation for r as a function of the phase of the electron with respect to the peak of the accelerating electric field. Utilizing earlier results giving the dependence of phase on distance along the axis of the accelerator, a choice of a particular dc magnetic focusing field allows this differential equation to be integrated twice, from which r as a function of distance along the accelerator is plotted.
The need for a focusing field is indicated, and it is shown that the particular field chosen has the advantage that with it, the electron will oscillate between known fixed limits of r. The necessity of shielding the electron emitter from the focusing field is shown, and an indication of the magnitude of the field and length of the accelerator for which it should be used, is given for the Purdue accelerator.
25(1954); http://dx.doi.org/10.1063/1.1721691View Description Hide Description
The friction coefficients of freshly cut surfaces in inert atmosphere are compared with values on the same surfaces later exposed to air. The kind of metal or oxide comprising the friction couple was also varied. From experimental data it was concluded that: (1) Solid phase welding accounts for almost the entire magnitude of the static friction coefficient for metal‐metal couples. (2) Fe‐Ag does not solid phase weld, other insoluble couples examined do weld. (3) A measure of solid phase weldability appears to be the quotient of the work of adhesion and the shear strength of the weaker couple. (4) Titanium does not solid phase weld to crystalline oxides to any greater extent than copper. (5) Freshly cut titaniumsolid phase welds to other metals in frictional contact. The degree of welding is independent of the atmosphere (air, nitrogen, or argon) or the time (less than 24 hours) at room temperature in such atmospheres. This behavior is contrary to that experienced with iron or copper With the latter couples less than five minutes exposure of a freshly cut surface to air is sufficient to markedly reduce the degree of solid phase welding. (6) The seizability of titanium is probably due to the lack of resistance, of the contaminant film formed on it in air, to penetration or wear in frictional contact.
25(1954); http://dx.doi.org/10.1063/1.1721692View Description Hide Description
The objective of this work is to determine the orders of magnitude of the band width available from a cavity antenna. Fcr simplicity of calculation a piece of shorted square wave guide of side a, opening into a half‐space, is considered, and the resonance frequency and band width (1/Q) are calculated from transmission‐line theory. The analysis depends on an equivalent aperture admittance which is defined for the lowest‐order wave guide mode.
This type of antenna has a minimum Q which is equal to 0.424(a/λ)−3 when (a/λ)<0.35. The cavity depth required for the minimum Q is of the order of 0.4a. The required dielectric constant increases rapidly as a/λ decreases. The effect of dielectric loss is considered. Some experimental results which agree well with theory are presented.
25(1954); http://dx.doi.org/10.1063/1.1721693View Description Hide Description
When a shock is incident upon a corner from either the concave or convex side it is reflected by the walls forming the corner and diffracted at the edge. The mathematical problem corresponding to this phenomenon has not been solved analytically except in the following limiting two‐dimensional cases: a weak (acoustic) shock incident on any corner (Keller and Blank) and a finite shock incident either normal or parallel to a nearly flat wall (Bargmann, Lighthill, Ting, and Ludloff). In this article it is pointed out that there are certain special cases of a finite shock entering a corner from the concave side in which no diffraction occurs. In these cases it is therefore possible to solve the entire problem by using the theory of regular reflection, which yields an explicit solution by algebraic means alone.
25(1954); http://dx.doi.org/10.1063/1.1721694View Description Hide Description
This paper describes work function measurements of titanium and zirconium carried out by the conventional photoelectric method. Two samples of each metal were investigated before and after treatment with chlorine. The changes in work function due to chlorine absorption were of the same order of magnitude as those reported for other halogens on other metals.
25(1954); http://dx.doi.org/10.1063/1.1721695View Description Hide Description
A unified explanation of delayed yield and of strain rate and temperature dependence of yield point in iron was attempted on the basis of the Cottrell locking theory, taking internal stress into account. The energy of a dislocation required to fit existing data is small by an order of magnitude compared with the value theoretically estimated by Mott and Nabarro. The distance, calculated from the value of internal stress, between dislocations assembled at a grain boundary is of a reasonable order of magnitude. It is suggested that the mechanism of yielding may change from Cottrell type to Mott‐Nabarro or Frank‐Read type when applied stress falls below some limiting value.
25(1954); http://dx.doi.org/10.1063/1.1721696View Description Hide Description
The negentropy principle of information was discussed in a previous paper and applied to a variety of problems of scientific observation. The present paper considers the technical problems of telecommunication. Using a thermodynamical discussion one obtains the Shannon formula for the capacity of a channel with noise. This example proves again the connection between information and negentropy.
The problem of writing and reading is examined. Here we have a type of ``dead information'' which must be re‐energized for reading: The source of light needed for reading provides the negentropy which is changed into information by the reader.
The negentropy connected with ``organization'' is also discussed. All these examples show the connection between information and negentropy. They are of great importance for a theory of information, although the entropy amounts involved may be completely negligible.
25(1954); http://dx.doi.org/10.1063/1.1721697View Description Hide Description
The electrolytic analog transistor is an operating model of the junctiontransistor which substitutes the reduced and oxidized forms of ions in solution for electrons and holes in a semiconductor. A base electrode makes a low‐resistance contact to the solution and also serves to maintain the ratio of oxidized and reduced ions at an equilibrium value, thus establishing the potential of the solution. In P‐N‐P operation, a polarizable metalelectrode (the emitter) is held at such a potential above the base electrode that the majority ions, the reduced form, are readily oxidized at its surface. An identical electrode, the collector, is placed within a few tenths of a millimeter of the emitter. It is biased in such a way that the minority ions (the oxidized form) which diffuse from the emitter to its surface are reduced. Utilizing several oxidation‐reduction couples and different base electrodes, it has been found that values of α = − (∂Ic /∂Ie )vc in the neighborhood of unity can be obtained. Because of the relatively low mobilities of ions in solution, the device operates only at low frequencies. The device may also have some application as a constant current element in transistor bias circuits.
25(1954); http://dx.doi.org/10.1063/1.1721698View Description Hide Description
The intrinsic diffusivities of individual components in several binary systems have been determined with vapor‐solid type diffusion couples. In this type of couple a high vapor‐pressure component is diffused from the vapor phase into a slab of lower vapor pressure initially containing inert markers on its surface. Diffusivities may then be calculated from marker movement and diffusion‐penetration curve data. Equations necessary for analysis of such couples are developed.
The alpha brass, copper‐nickel, and silver‐gold systems were investigated. Observations of polygonization during diffusion and dimensional changes normal to the diffusion direction were also made. Porosity formation, which occurred only in the vapor‐solid silver‐gold couples, was investigated in detail and was compared to porosity formation in corresponding sandwich couples. The general use and limitations of such couples in diffusion studies are discussed.
25(1954); http://dx.doi.org/10.1063/1.1721699View Description Hide Description
The problem of finding the density of ac energy and the ac power flow in an electron beam, originally uniform in velocity and charge density, which interacts with a weak electromagnetic traveling wave, is considered. For the problem to be significant, the way in which the traveling wave is built up must be specified and the energy and power flow turn out to depend upon the method of establishment.
25(1954); http://dx.doi.org/10.1063/1.1721700View Description Hide Description
Distinction between data transmission from one person (or point) to another and the inception of new knowledge is analyzed in relation to information theory. The customary measure H = −∑ ipi logpi applies essentially to the former case, and after discussing the requirements appropriate to a yardstick for new knowledge an elementary measure for this concept is tentatively proposed. The intuitive requirements, however, are not all mutually compatible and the present measure can only be regarded as provisional; in particular some ``memory'' characteristics appear mandatory for future development.
25(1954); http://dx.doi.org/10.1063/1.1721701View Description Hide Description
It is shown that the magnitude of the local velocity and the surface pressure coefficient can be limited over the entire upper surface of a two‐dimensional hydrofoil as it approaches the free water surface. In this limiting condition the flow over the upper surface is governed by the shallow water wave propagation phenomenon, and the limiting pressure or velocity is determined by the maximum depth attainable from the hydraulic jump relations. The experimental data verifies this new theory and proves that the previous two‐dimensional hydrofoil theories are valid only for relatively deeply submerged bodies.
Fast Time Analysis of Intermittent Point‐to‐Plane Corona in Air. III. The Negative Point Trichel Pulse Corona25(1954); http://dx.doi.org/10.1063/1.1721702View Description Hide Description
Fast oscilloscopic time analysis of the negative point Trichel pulse corona in room air at various pressures and gap geometries reveals the following data. The very short rise and quenching time of the pulse at atmospheric pressure observed by English is confirmed. Under these conditions the secondary action is a photoelectric liberation from the cathode and discharge extinguishes by dissociative attachment to give O− ions. Decreasing pressures reduces space charge density, prolongs the discharge and brings in a secondary liberation by positive ion bombardment. Increasing potential at constant pressure at first decreases clearing time, leaving pulses unchanged and current increases proportional to repeat rate. At higher potentials, photon action is reduced relative to positive ion impact at the cathode, the discharge extends further into the gap, and quenching is very effective with less total ions reducing pulse size. With increasing repeat rate, current no longer increases proportional to repeat rate in consequence of smaller pulse height. The effects of increased point diameter are increased pulse size. Comparable clearing times for large and small points require higher potentials for the former. Thus, at constant pressure and potential, increase in point radius produces a decrease in frequency and increase in total charge per pulse. Conditioning of the point at high‐current densities and repeat rates decreases the number of ions per pulse, which is of the order of 109 under normal circumstances, and shortens the pulse by reducing cathodework function and increasing photoelectric liberation, so that choking occurs with little ion action. Since repeat rate is slightly decreased by the charge liberation, the reduced pulse charge yields a reduced current through a conditioning that decreases work function. Comparison of ion movement across the gap during clearing time, with that for O2 + ions of known mobility in air, indicates that the rapid ion transit in Trichel pulses is caused by a majority of O− ions with a mobility of about 4 cm2/volts second from near the cathode with some formation of O2 − ions in transit.
25(1954); http://dx.doi.org/10.1063/1.1721703View Description Hide Description
A study of bulk minority carrier lifetimes and surface recombination rates has been made, using experimental equipment similar to that of Navon, Bray, and Fan, wherein the time‐decay of carriers injected electrically by a pulse is observed. The diffusiontheory for carriers of bulk lifetime τ has been worked out, considering a definite probability of nonrecombination, or reflection coefficient, to be associated with the surfaces of the sample. A simple relationship between this quantity and the surface recombination velocity has been derived. If the surface recombination velocity is known, the effective trapping area at the surface may be calculated from this relation. Values of bulk lifetime and surface recombination velocity have been found for both n‐ and p‐type germanium samples, and for a wide variety of surface treatments. By suitable modifications in experimental technique, this experiment may be utilized to measure drift mobility of minority carriers.
25(1954); http://dx.doi.org/10.1063/1.1721704View Description Hide Description
Microwaves propagated between two doubly‐curved parallel conducting surfaces, in the geometrical optics approximation, follow rays obeying Fermat's principle. The optical distance in this case is determined both by the index of refraction of the material between the surfaces and by the curvature of the surfaces. The focusing properties of such a parallel plate lens is fully specified by giving the metric of the two‐dimensional optical space associated with the mean surface of the guide. A family of equivalent lenses is shown to exist, and formulas are obtained that permit one to go from any one of these lenses to any other. Furthermore, conditions are given for there to exist in a family of equivalent lenses a lens employing a unit index of refraction everywhere. Of particular interest is the demonstration that the Rinehart analog of the Luneberg lens and the RCA ``R‐2R'' lens are members of a two‐parameter family of lenses with analogous focusing properties.