EM field modelling is an old and well established science; or should that be art? Algebraic analysis using Schwarz-Christoffel transformations were used by J.J. Thomson in 1883. Then in 1900, F.W. Carter published a paper on airgap induction, using conformal mapping. Many other papers followed. For the most part, these were all concerned with field solutions in parts of devices, not the whole device. Somewhere around the mid-twenties, finite differences (FD) appeared applied to both electrical and magnetic field problems. The labour involved was considerable, and it was not until the introduction of digital computers in the early sixties that the technique became widely used. Within ten years FD gave way to finite element modelling. Possibly the first paper in which the technique was applied to electrical problems was by Silvester in 1969. Since that time, the method has become extremely popular for the solution for both magnetic and electric field problems. More recently, boundary element methods have appeared. These have their own loyal following. It has to be added, that much of the early development of all the above numerical methods owes much to structural engineering. That situation has now changed, and there is a healthy flow of new ideas in both directions. (4 pages)

The stability characteristics of the Xi-core at constant airgap is examined by experimental and theoretical methods. The 2-dimensional field is solved numerically using the finite difference technique. It is shown that the stability characteristics can be deduced from stored field energy variations.

A circle diagram is described for the d.c. braking of an idealised linear induction motor with an iron-backed or air-backed conducting secondary. Braking thrust and normal force can be easily predicted from the diagram. Experimental confirmation of the method is described.

Fourier-transform theory is used to solve the longitudinal-edge-effect problem in linear induction motors, for constant-current conditions. Inversion of the Fourier transform is obviated by employing Parseval's theorem; calculations are thus considerably simplified. Experimental confirmation is given, together with an approximate method for sketching the thrust/slip curve. The ratio (goodness factor/pole pairs) is shown to be important.

A frequently recurring problem in induction device theory involves the determination of the input wave impedance to a device. This is commonly done by solving for the field components. In this paper, a completely different method based on the scalar Riccati equation is described. It is shown that the wave impedance can be obtained by the numerical integration of a Riccati-type differential equation, having wave impedance as the unknown variable. This is illustrated with three simple models each having an unknown field distribution in one direction only. Extensive computer testing was carried out and the results obtained were found to agree very closely with published analytical methods. The method is very simple, easily programmed and avoids the use of higher transcendental functions.

As an alternative to analysis, the principle of electromagnetic similitude is applied to linear induction motors. It is shown that exact similitude is not essential. Shadow windings are introduced as a means of determining power, impedance and voltage levels under constant-current conditions. Hence the constant-voltage performance of the device can be derived. Results obtained for both static and dynamic thrust, for real and scale-model machines, are compared. It is concluded that scale-modelling techniques can give an accurate indication of all the electromagnetic forces in l.i.m.s. Finally, an index of l.i.m.s is proposed, expressed in terms of dimensionless parameters.

It is shown that the normal force in a single-sided linear induction motor can be appreciable, and that it is highly dependent on the speed and terminal conditions. Two simple theories are presented, and these are compared with experimental results. Tests were performed on both static and dynamic machines showing the variation of normal force with frequency and speed, and the effects of pole changing and plugging. It is concluded that the normal force and the possible changes that can occur in its magnitude and direction are highly significant design factors.

Tooth-ripple harmonics in electrical machines are important in the calculation of losses and noise. S.Neville has shown that, by employing ‘wavelength spectra’, their calculation and presentation may be simplified. The idea is extended and the ranges of the variables are increased. The resulting form of the calculated data is more convenient for use in a computer program.

The problem of travelling electromagnetic waves in multiregion induction machines at power frequencies is examined. Using the concept of surface impedance, a method is derived whereby equivalent circuits can be established in a systematic manner. A method of determining levitation force from a knowledge of the equivalent circuit is developed. Calculation of input and output power is simplified using the approach outlined. The ferromagnetic portions of the multiregion system are assumed to have constant permeability.