Index of content:
Volume 22, Issue 5, 01 May 1951
22(1951); http://dx.doi.org/10.1063/1.1700002View Description Hide Description
The radiation from fast electron beams passing through a succession of electric or magnetic fields of alternating polarity is examined. The radiation of maximum frequency is emitted in the forward direction. If the deflecting fields are not too large, a semiqualitative argument shows that the maximum frequency is the lowest possible harmonic. The frequencies emitted are determined by studying the Doppler effect, and the angular distribution of radiated energy as well as the total radiation are calculated in a simple straight‐forward manner with reference to well‐known formulas of special relativity. The question of the coherence of the radiation is discussed. The spectral distribution of radiated energy is then calculated more exactly. It is concluded that several applications of the radiation appear possible. A scheme for obtaining millimeter‐waves of considerable power is outlined. The upper limit of the power in a band extending down to a wave‐length of 1 millimeter is calculated to be of the order of several kilowatts for a beam of one ampere and an energy of 1.5 megavolt. The use of the radiation for speed monitoring of beams with energies up to 1000 megavolt is discussed.
22(1951); http://dx.doi.org/10.1063/1.1700003View Description Hide Description
In the investigated range of 5×10−3 to 10−7 mm Hg, the breakdown voltage over insulators in high vacuum is independent of pressure. Currents of 10−11 to 10−8 ampere were observed in the region below breakdown voltage by detecting x‐ray quanta with a Geiger‐Mueller counter. Pinhole camera x‐ray pictures revealed that practically all radiations originate from an area on the anode a distance from the insulator, with a weaker radiation coming from a ring immediately adjacent to the insulator. Current‐voltage relationships as usually observed in these experiments indicate a roughness factor and an emitting area on the cathode similar to previous findings in vacuum gaps. Current bursts were observed which did not develop into complete breakdown. Oscilloscopic observations revealed that sometimes at breakdown over insulators the voltage on the test sample drops to 2.5 kv; in other cases it falls to less than 100 volts. The low voltage arc‐like discharge extinguishes at a current of about one ampere for copper electrodes in contact with Pyrex glass. When a resistance in series with the test sample is increased to keep the maximum current below one ampere, no stable discharge is observed. As in a vacuum gap, the breakdown voltage over an insulator is increased by successive breakdowns. Part of this ``conditioning'' is permanent. The nonpermanent part is dependent on the state of the test sample prior to conditioning. The anode does not appear to influence conditioning. When the resistance in series with the test sample allows a discharge current above one ampere to flow, a fast conditioning usually occurs which results in a high permanent breakdown voltage.
An Electrical Method for Investigating the Nature and Behavior of Small, Airborne, Charged Particles22(1951); http://dx.doi.org/10.1063/1.1700004View Description Hide Description
A theoretical and experimental analysis of the current‐voltage (dc)characteristic of a parallel‐plate condenser through which the small, airborne, charged particles of an aerosol are passed continuously shows that a charge‐weighted distribution of radius (or radius squared)‐charge ratio may be obtained when certain experimental conditions are satisfied. Examination of such a distribution for an aerosol (particles less than a micron in diameter) before and after passage through a small (20‐liter) settling chamber shows that the classes of particles in the distribution that are easiest to collect in the condenser are also the classes that disappear most rapidly in the chamber. It is pointed out that the methods employed would become less tedious and more powerful when used with a dispersion of constant‐size particles, and that they are well adapted for certain fundamental investigations of aerosol properties.
22(1951); http://dx.doi.org/10.1063/1.1700005View Description Hide Description
Fracturing, or scabbing, of a material near a free surface as the result of a transient compressional stress wave of high intensity impinging on that surface has been observed for many years; however, little quantitative data that relate the fracture to the nature of the stress wave and the physical properties of the material seem to exist. The phenomenon has been investigated for five metals, 1020 steel, 4130 steel, 24S‐T4 aluminum alloy, brass, and copper, by using an explosive charge to induce a high intensity stress wave in the metal. The distribution of pressure within the wave was determined by a modified Hopkinson‐bar type of experiment.
Scabbing has been found to be governed principally by the spatial distribution of pressure within the wave and a critical normal fracture stress σ c that is characteristic of the material and perhaps the state of stress. Numerical values of σ c were obtained for each of the five metals.
22(1951); http://dx.doi.org/10.1063/1.1700006View Description Hide Description
Permanent magnets have been successfully utilized to energize magnetic lens pole pieces for high resolution electron microscopy at direct magnifications of 7000 times with 30‐kilovolt electrons, and 6000 times at 50 kilovolts. Stray magnetic fields, which have previously limited high magnification and high resolution, have been minimized by using two circuit gaps in parallel and by surrounding the assembly with a magnetic shield connected to the external pole pieces of the two gaps. Magnetomotive forces of 1400 gilberts have been obtained. A single magnet,magnets in bucking arrangement, and configurations involving magnets disposed either radially or parallel to the optical axis of the lens, have been successfully utilized. The effects of various magnetic circuit parameters on the design and operation of the lens system have been studied.
22(1951); http://dx.doi.org/10.1063/1.1700007View Description Hide Description
It is common knowledge that in cyclically stressed ferromagnetic materials an energy loss due to internal friction is observed; this part of loss disappears when the material is brought to saturation or above the Curie temperature. It is also known that a stress, applied to a ferromagnetic material, produces a motion of Weiss domains from their rest position.
The scope of this paper is to verify experimentally the following hypothesis: in cyclically stressed ferromagnetic materials,energy loss because of magneto‐elasticinternal friction is induced by a domain motion due to the applied stress itself.
The ratio Ir/Is (where Ir =residual induction, and Is =saturation intrinsic induction) has been chosen as an index of the domain position; its measure has been carried out by normal ballistic method on a permeameter rigged with a torsion applying device.
Energy loss was measured by means of a torsion pendulum; a coil allowed the saturation of specimen in order to separate magneto‐elastic from purely mechanical losses. Plotting curves of Ir/Is and of magneto‐elastic losses as a function of applied stress and comparing them for each tested material, it has been found that they, besides having the same shape, show abrupt slope variations at practically the identical value of stress. It has further been observed that a specimen presenting only a low percent variation of Ir/Is ratio (i.e., in which domains had slightly moved from their rest position) showed also moderate magneto‐elastic losses. These results appear to confirm assumptions made about the magneto‐elasticinternal friction in ferromagnetic materials. The reason why motion of domains takes place in an essentially irreversible way still remains unknown.
22(1951); http://dx.doi.org/10.1063/1.1700008View Description Hide Description
Existing calculations on the tunnel effect through thin films in contacts apply to very weak and very strong electric fields. It is the main purpose of the present paper to complete the picture by the treatment of intermediate cases, which are important for many contact applications.
22(1951); http://dx.doi.org/10.1063/1.1700009View Description Hide Description
The problem of the diffraction of a plane electromagnetic wave by the convex surface of a paraboloid of revolution is examined. Expressions for the components of the incident wave, the scattered wave, and the refracted wave are obtained and are written in such a form that boundary conditions can be satisfied over a paraboloidal surface ξ=ξ0. The analysis is then restricted to the case of a perfectly conducting paraboloid and a wave whose wave front is perpendicular to the axis of rotation. The amplitude of the scattered wave is plotted as a function of the distance along the axis of rotation for three paraboloids. One of these curves is compared with the scattered field produced by a sphere whose radius is equal to the radius of curvature of the paraboloid at the nose.
22(1951); http://dx.doi.org/10.1063/1.1700010View Description Hide Description
A steady‐state temperature distribution is found for the case of a periodically varying point source of heat, immersed in a homogeneous isotropic infinitely extended fluid, moving with a constant velocity. The phase velocity of the generated temperature wave is discussed. If the phase velocity in a stationary medium is V 0, and the fluid velocity v>V 0, the phase velocity in a moving medium is given by V=v[1+½(V 0/v)4 −(5/8)(V 0/v)8+…]. For practical cases one can usually assume V=v, with an error of less than 0.1 percent, even when V 0 is as large as 20 percent of v.
22(1951); http://dx.doi.org/10.1063/1.1700011View Description Hide Description
The accuracy of a V‐2 type ballistic rocket which is guided to the termination of propulsion and is in free flight thereafter is investigated. The effect of the rotation of the earth and the re‐entry into the atmosphere on the accuracy is considered and the errors due to lack of thrust control are analyzed.
22(1951); http://dx.doi.org/10.1063/1.1700012View Description Hide Description
Discontinuities in the slopes of Langmuir probecharacteristics for discharges in mercury vapor in the presence of argon or krypton gas were eliminated by heating the wire probe by electron bombardment before each reading. The results indicate that in many cases nonlinear probe plots are due to change in the contact potential of the probe and do not mean a non‐maxwellian distribution of electron velocities. Typical probe plots are given for discharges with and without striations, and also for a discharge in gas contaminated by a trace of CO or CO2. Electron temperature data are listed for representative pressures of gas and mercury vapor and for various discharge currents.
22(1951); http://dx.doi.org/10.1063/1.1700013View Description Hide Description
The short‐time tensile breaking strength of various grades of graphite was measured as a function of temperature from room temperature to the sublimation point. A characteristic common to all the strength versus temperature curves is that the strength approximately doubles in going from room temperature to about 2500°C and then decreases rapidly to zero near the sublimation point. Graphite can be used as a structural material at temperatures much higher than those permissible with other materials. The density distribution within an original block is given for grade EBP graphite. The creep characteristics of grade ECA graphite under tensile stress at elevated temperatures were investigated. The testing temperature ranged from 2200 to 2900°C while the applied stress varied from 500 to 5500 psi. The range of ``steady'' creep rates measured extended from 3×10−8 to 2×10−4 in. per in. per sec. The activation energy and change in entropy defined by rate theory were calculated for the above conditions. Low frequency dynamic measurements of the Young's modulus of grade ECA graphite made at temperatures from 1000 to 2000°C showed an increase of the modulus with temperature.
22(1951); http://dx.doi.org/10.1063/1.1700014View Description Hide Description
By using suitably defined logarithmic derivative functions and the recurrence formulas due to Infeld, the problem of obtaining numerical answers for the scattering of a plane electromagnetic wave by a sphere can be simplified. This is particularly true if the sphere has a complex index of refraction. The nature of these logarithmic derivative functions is shown for one special case involving water spheres.
Experimental results are given for the back‐scattering cross sections of water spheres in the electrical size region 0.74≤2πa/λ≤5.90. These results were obtained at a wavelength of 16.230 cm using the standing‐wave method of measurement. It is believed that these were the first laboratory measurements of the back‐scattering cross section made on individual water spheres in this critical size region.
Comparison is made between the experimental and theoretical results.
22(1951); http://dx.doi.org/10.1063/1.1700015View Description Hide Description
The reactions which occur at silverelectrodes in a normalglow discharge in air have been determined. These are: (1) formation of AgNO2 and some Ag2O at the anode at the rate of 3.4 μg/coulomb; (2) loss of metal from the cathode by chemical action at the rate of 3.5 μg/coulomb (probably the same reaction as (1) with subsequent loss of the reaction products by the greater heating of the cathode, but this hypothesis has not been established); and (3) normal sputtering loss at the cathode at the rate of 0.4 μg/coulomb. These processes result in building a conducting layer on the anode. If the electrode separation is so small that the anode extends into the region of the cathode fall, then the high electric field pulls the newly formed and not very coherent growth upon the anode across into a bridge between the electrodes.
22(1951); http://dx.doi.org/10.1063/1.1700016View Description Hide Description
In measuring small microwave reflections back to the transmitting antenna, a finite reflecting surface which simulates the reflection of an infinite plane is desirable for calibration. Treating a flat reflecting surface by the methods of physical optics, the reflection can be represented as the sum of the contributions of Fresnel zones. Owing to the slow rate of decrease of reflections from successive zones, the shape of the surface has considerable effect on the accuracy with which its reflection simulates that of an infinite plane.
By means of the calculus of variations, an optimum shape is designed which, for a given antenna pattern, reflector size, and range of antenna to reflector distances, will give the closest simulation of the reflection of an infinite plane. Such an optimum surface of 23 wavelengths extreme dimensions has been constructed for use at 25 wavelengths from a dipole antenna. The maximum simulation error was calculated to be ±3 percent. The reflection from this surface has been measured and is within ±4 percent of that of an infinite plane over the range of 15 to 35 wavelengths from the antenna.
22(1951); http://dx.doi.org/10.1063/1.1700017View Description Hide Description
Direct‐current corona studies were made on coaxial cylinders with inner cylinder at negative potentials in pure N2, pure O2, and mixtures of 1 percent 10 percent O2 in N2, and on clean dry air from 27‐mm to 760‐mm pressure. Observations of current potential relations from field intensified ionization currents to as near breakdown as possible were made noting thresholds for current transitions, pulsed discharges and other oscillographic as well as visual appearances. Negative wire corona does not exist as such in truly pure N2. Unless a high series resistance was in the line once the critical region of potential was reached, the gap broke down to an arc. With current limiting resistor bombardment of the filament with positive ions from the low order pre‐discharge released enough gaseous impurity so that after six minutes the resistor could be removed and the internal resistance of the negative ion space charge near the anode held the current at 2000 μa. The behavior indicated that breakdown was preceded by a low current Townsend discharge that cleaned up or conditioned the filament. In all the negative wire studies the onset thresholds vary with the past history of the wire, and the threshold potential at higher pressures overshoots the minimum operating potential of the discharge once it has set in. Studies with 1 percent O2 indicate that the threshold discharge with clean gas is a continuous current with a diffuse glow which in time contracts to a cathode spot with Trichel pulses. Similar behavior applies to air at lower pressure. The negative wire in O2 shows very heavy predischarge currents with relatively small current jumps when random Trichel pulses appear, and further small current jumps when the fixed spot with regular Trichel pulses appears. There is evidence for a continuous heavy background of Townsend discharge along the whole wire until the visible spot and heavy pulsed discharge appears. The duration of Trichel pulses in pure O2 is two to ten times those in air with point to plane geometry and two to three times those in air with coaxial cylindrical geometry. Lower pressures in pure O2 show many current transitions and transitions between discharge forms.
22(1951); http://dx.doi.org/10.1063/1.1700018View Description Hide Description
Single crystals of 99.999 percent pure lead and copper were grown by the Bridgman method. A tensile creep apparatus is described. Measurements on copper at 23°C and at −190°C show that for resolved shearing stresses below 0.6 kg per mm2 the steady‐state creep rate is negligible (i.e., less than 10−6 per minute). The resolved yield stress for copper is 219±21 g per mm2 at both the temperatures used. In the case of lead, for resolved shearing stresses below 0.13 kg per mm2 the steady‐state creep is negligible at −190°C. The yield stress at that temperature is 96±8 g per mm2. Measurements were made on the steady‐state creep of lead at 25°C and 110°C. The results produced by annealing were investigated.
22(1951); http://dx.doi.org/10.1063/1.1700019View Description Hide Description
Four heavy lubricating oils, three light lubricating oils, and three aircraft engine fuels were tested for viscosity, surface tension, and density, and a relation was established between the time rate of evolution and solution of air and the viscosity of the liquids. A definite relation was also found between solubility constants and half‐lives as well as between half‐lives of evolution and solution.
The Electric and Magnetic Constants of Metallic Delay Media Containing Obstacles of Arbitrary Shape and Thickness22(1951); http://dx.doi.org/10.1063/1.1700020View Description Hide Description
The dielectric constant and permeability of a metallic‐obstacle medium are obtained in this paper by various static methods. The equivalent shunt capacitive susceptance and series inductive reactance of the individual obstacles are also determined. In the first part of the paper, a correspondence is established between an infinitely thin conducting obstacle and an aperture in an infinitely thin conducting wall. With the aid of this correspondence, delay medium formulas are given for several obstacle shapes. Formulas are then derived for the effect on the magnetic field of moderate obstacle thickness. Finally, it is shown how the electric and magnetic constants may be evaluated by electrolytic tank measurements in the case of obstacles having completely arbitrary shape and thickness. This has been done for a typical delay medium, and the effect of obstacle thickness on the index of refraction is shown.
22(1951); http://dx.doi.org/10.1063/1.1700021View Description Hide Description
The lattice and grain boundary self‐diffusion coefficients of high purity silver have been measured over extended temperature ranges. It has been found that the lattice diffusion coefficients are described by the equation, and are independent of the thermal history of the crystals within wide limits. The absolute grain boundarydiffusion coefficients fit the equation, and are independent of the grain size of the specimens.