General Electric

About: General Electric is a company organization based out in Boston, Massachusetts, United States. It is known for research contribution in the topics: Turbine & Signal. The organization has 76365 authors who have published 110557 publications receiving 1885108 citations. The organization is also known as: General Electric Company & GE.

Dual-rotor, radial-flux, toroidally-wound, permanent-magnet machines

Abstract: A novel machine family-dual-rotor, radial-flux, toroidally wound, permanent-magnet (RFTPM) machines-is proposed in order to substantially improve machine torque density and efficiency. After the principles of operation, configurations, and features are discussed, the machine design and optimization guidelines are given. A prototype has been designed, built, and tested. The measured torque density of the prototype, which well matches the design value, is almost three times of that of the induction machine with the same power of 3 hp and speed. Meanwhile the efficiency is still kept high and the material cost is kept low by using ferrite magnets. Three novel approaches are proposed to reduce the cogging torque in the RFTPM machines, whose validity is verified by finite-element analysis results and experimental measurements.

Fermi Surface and Energy Bands of Copper

Abstract: The Green's function method has been used to study the energy bands of Cu for two quite different potentials. It is found that the resulting $E(\mathrm{k})$ for the two cases are very similar throughout the Brillouin zone having the same ordering of the levels and comparable level separations in the conduction, low-lying excited, and $d$-band regions. This implies that the calculated band structure is not as sensitive as had been previously contended. The properties of the Fermi surfaces associated with the two theoretical band structures are compared with the results of experimental studies. It is found that the theoretical surfaces intersect the hexagonal zone face in accord with experiment. The computed radii of contact for the two cases are close to the measured values. Furthermore, the bellies for the two cases are shown to deviate appreciably from sphericity in agreement with the results of recent magnetoacoustic effect experiments. The origin of the distortions is explained in terms of the interaction between the conduction and $d$ bands. The "masses" defined in terms of the cyclotron resonances for various orbits on the Fermi surface, the lowtemperature electronic specific heat, and the dielectric constant in the infrared region are determined for the calculated $E(\mathrm{k})$. The calculated masses are all somewhat lower (by about 10-30%) than the corresponding measured masses. It is believed that these discrepancies reflect the contributions of the effects neglected in the individual-electron model. Finally, the sharp rise in the optical absorption, which on the basis of the theoretical $E(\mathrm{k})$ corresponds to the onset of interband transitions between the $d$ bands and the Fermi level, is found to occur at an energy in good accord with experiment.

Effect of Dissolved Gases on Some Superconducting Properties of Niobium

Abstract: The addition of interstitial solute atoms to niobium, in concentrations below the solubility limit, lowers the transition temperature. Interstitial oxygen has the largest effect, decreasing ${T}_{c}$ by 0.93\ifmmode^\circ\else\textdegree\fi{}K per at.%; while increasing the resistivity in the normal state by 5.2 \ensuremath{\mu}\ensuremath{\Omega} cm per at.%. Magnetization curves obtained on niobium ($\frac{{R}_{298\ifmmode^\circ\else\textdegree\fi{}\mathrm{K}}}{{R}_{10\ifmmode^\circ\else\textdegree\fi{}\mathrm{K}}}\ensuremath{\approx}500$) and on similar specimens containing interstitial oxygen or nitrogen are substantially reversible and are similar to the shape predicted by Abrikosov for superconductors of the second kind. The field first penetrating the sample, ${H}_{\mathrm{FP}}$, less than the thermodynamic critical field, ${H}_{c}$, decreases with increasing concentration of the interstitial atom, while ${H}_{\mathrm{N}}$, the field at which the normal state is restored (as determined from magnetization measurements), increases. The ratio, $\frac{{H}_{\mathrm{N}}}{{H}_{c}}$, is a linear function of ${\ensuremath{\rho}}_{n}$. When the solubility limit is exceeded, hysteresis effects become more pronounced similar to those predicted by Bean's model. Resistive measurements at low current density in longitudinal magnetic fields indicate that both niobium and its interstitial solid solutions exhibit some superconducting properties above ${H}_{\mathrm{N}}$.

A high-speed sorting procedure

Abstract: There are a number of methods that have been used for sorting purposes in various machine programs from time to time. Most of these methods are reviewed by Harold Seward [1] in his thesis. One tacit assumption runs through his entire discussion of internal sorting procedures, namely, that the internal memory is relatively small. In other words, the number of items to be sorted is so large that they cannot possibly all fit into the memory at one time. The methods of internal sorting which he discusses are sorting by: 1) Finding the smallest. 2) Interchanging pairs. 3) Sifting. 4) Partial sort. 5) Merging pairs. 6) Floating decimal sort. The first four methods all require a time proportional to n ~, where n is the number of items being sorted. The time for the fifth method is proportional to n(ln n). The time for the sixth method is proportional to n(ln r), where r is the largest number to be used in a key. As pointed out in Seward's paper, one would normally choose either method five or six for a rapid internal sort, especially if n is to be very large. The chief drawback of these two methods, however, is the fact that they require twice as much storage as the other four methods. The advent of very large high-speed random access memories changes the picture relative to sorting somewhat. It is 1now possible to have a very large number of items to be sorted in memory all at one time. It is highly desirable, therefore, to have a method with the speed characteristics of the merging by pairs and the space characteristics of sifting. If such a method were available it would be possible to sort twice as many items at one time in the machine and still do it at a reasonably high speed. Such a method is outlined in this paper. The idea is, in fact, to combine some of the properties of merging with some of the properties of sifting. The method is most easily described by reference to the block diagram, figure 1. Suppose we are given a sequence of elements f~ to be Yes ? [Y~ I FIG. 1. Block diagram to sort a set of elements f~ with keys rl, i = 1,2,...,n. sorted with keys ri, i = 1, 2, • •. , n. One begins by dividing the set of …

Composition Requirements for Glass Formation in Metallic and Ionic Systems

Abstract: WE have developed in recent papers the point of view that all liquids, ionic and metallic included, would form glasses when sufficiently undercooled, save for the intervention of crystallization1–3. We have shown2 that the glass-forming tendency (for a given rate of cooling and for substances of a given molecular type) is greater the lower is the reduced melting temperature : where T m is the thermodynamic crystallization temperature, and h v is the molecular heat of vaporization. This criterion is in semi-quantitative accord with experience on glass formation. In particular it encompasses Zachariasen's4 criterion, and accounts for the enhancement of glass-forming tendency by molecular asymmetrization, both being simply equivalent to the depression of τm. It is also clear, as we pointed out, that the addition of impurities which depress τm similarly enhances the glass-forming tendency. In this communication we point out striking new experimental confirmation of this last deduction.