3 Now with MOTHESIM. Le Plessis-Robinson. France.

A complete set of experiments on near-field antenna measurements is presented. This labwork serves two purposes: first, it introduces students to the difficulties associated with such a measurement technique, and second, they gain the confidence to work in high tech, elaborate environments. The technical goal in this labwork is to obtain the radiation pattern of an antenna through measurements of the near-field using a complete experimental set-up driven by a Macintosh computer, and performing a near-to-far field transformation using a combination of Helmholtz equation and the Fourier Transform. Actually, the knowledge students acquire in this labwork is built up in stages which are both independent and experiments in their own right: they are described in detail in the paper, along with the knowledge built up at each stage by students as they advance into the labwork, and the objectives achieved. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 66-76]

Electrical Engineering Department

University of Utah

Salt Lake City, UT 84112

Computers provide an exciting opportunity for boosting electromagnetic education and corporate training. Animated graphics of the wave propagation phenomenon, visualization of the abstract and highly mathematical subjects, one-on-one and self-paced tutoring, and the ability to mimic often unavailable and expensive laboratory experiments are among the often-cited benefits of a computer-based electromagnetic education. In this paper, we review the activities of the NSF/IEEE Center for Computer Applications in Electromagnetic Education (CAEME). This Center was established to stimulate and accelerate the use of computers and software tools in EM education. A reflection on the extensive software package developed and distributed by the CAEME Center is described and examples of the developed software are presented. To help integrate available EM software in classroom teaching and corporate training, CAEME developed four multimedia lessons for instruction. These interactive media lessons integrate and allow individuals to interactively manipulate information from multimedia sources such as video, software, and animated graphics and also include instructional information such as quizzes and tutorials to help evaluate the students' performance. Features of these lessons are presented, and future developments in the Center's activities are described. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 77-107]

I.E.M.N, U.M.R 9929

Departement Hyperfrequences et Semiconducteurs, Bat P4

Universite des Sciences et Technologies de Lille

59655 Villeneuve d'Ascq Cedex

The purpose of this article is to present an original pedagogical product for teaching microwaves. It contains 9 lessons and is composed of 2 media:

- a handbook

- interactive educational software

It is a multimedia self-training product recommended for technicians and engineers working in conventional electronics and wishing to acquire advanced knowledge in microwaves in connection with their working structure. It can also be used in self access at the university by undergraduate students. In this last case, the authors will state in the conclusion, their observations following an experiment in computer aided learning (C.A.L.) carried out at the university of Lille. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 108-115]

University of Missouri-Rolla

Rolla, Missouri 6540l

No Abstract provided. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 116-124]

Electrical Engineering Department

Polytechnic Institute of Bucharest

Bucharest, Romania

The paper presents the actual interests in computer-aided electromagnetic education (CAEE) at the Electrical Engineering Department from Polytechnic Institute of Bucharest, Romania.

The actual state of computer use on different levels of undergraduate and graduate education is presented. Specifically, an overview of the undergraduate theoretical training and practical applications is described, and research topics for graduate reports and doctoral dissertations are discussed. Issues related to hardware utilities are reviewed, and future projects aimed at improving the CAEE capabilities in Romania are described.

The material is based on significant selected references. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 125-137]

Ecole Nationale Superieure Polytechnique

B.P. 8390 Yaounde, CAMEROUN

Tel.: (237) 231226, Fax (237) 231841

CAL-ANTENNAS is a tool containing graphics ( 2D and 3D) and sounds coded in the Turbo Pascal 5.5 language, for the implementation of educational software on antennas. From the Units files, a data base (frequency bands, antenna forms, antenna dimensions, formulae, characterising radiation) and a repertory of numericals methods (integration, graphs plotting, etc...) have been developped, complying with speed contraints. The necessary fundamental principles are contained in text files. Thus, this is one of the first structured software packages developed on the computer in the domain of antennas that treats the fundamental principles and the methodology of design.

This version of CAL-ANTENNAS for the microcomputer based on the Intel 386 and 486 Microprocessors contains more than a hundred illustrations. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 138-156]

Center for Electromagnetics Research

235 Forsyth Building

Northeastern University

Boston, MA 02115

As supercomputers become more accessible and as inexpensive personal computers become more powerful, the numerical modeling of electromagnetic fields in non-idealized geometries becomes increasingly practical. To enable graduate students to solve useful real-world problems, Northeastern University's course ECE 3347: Computational Methods of Electromagnetics teaches the important techniques of field and wave simulation, making use of a variety of programming languages, graphics packages, and computer systems which range from home computers to the most powerful supercomputers. Strong emphasis on algorithm design and computer testing helped motivate students to develop an understanding of the major issues involved in using computers to simulate electromagnetics problems. [Vol. 8, No. 1, (1993), Special Issue on Computer Applications in Electromagnetics Education, pp 157-165]

*Departrnent of Electrical and Computer Engineering

Mississippi State University

P.0. Drawer EE

Mississippi State, MS 39762

**Engineering Research Center for Computational Field Simulation

Mississippi State University/National Science Foundation

P.O. Box 6176

Mississippi State, MS 39762

A numerical technique is presented for computing the potential distributions surrounding power transmission and distribuhon lines of complex geometry. The technique employs a finite difference solution using boundary-fitted coordinates. A newly developed finite difference solver code is coupled with the existing EAGLE grid generation code to yield a system capable of solving for the electric potential andfield distributions surrounding complex configurations. A code validation example is presented which consists of a sphere-to-ground electrostatic solution. Sample results are also presented for a distribution line model. [Vol. 8, No. 2 (1993), pp 4-16]

Electronics Division

Defence Research Establishment Ottawa

Ottawa, Ontario, Canada

K1A 0Z4

This paper is concerned with the use of time- and frequency-domain methods for computing the interaction of electromagnetic waves with simple and complex structures. An example chosen for this study is a cubic box with the top open. The Finite Difference Time Domain (FDTD) method is used for computing time-domain responses to an electromagnetic pulse (EMP), a Gaussian pulse, and a sine wave. Frequency-domain results are obtained by using a moment method solution of the electric field integral equation (EFIE). Comparison is then made, both in the frequency and time domains, on corresponding quantities using Fourier transforms. Effects of various factors - the shape of the incident waveform, discretization of the structure, and Fast Fourier Transformation - on the CPU time and the accuracy of the solution are demonstrated. Guidelines are established for obtaining an accurate response. [Vol. 8, No. 2 (1993), pp 17-43]

The MITRE Corporation

Bedford, MA 01730-1420

This paper summarizes recent MITRE efforts to validate the NEC-3 and NEC-GS versions of the Numerical Electromagnetics Code (NEC) developed by Lawrence Livermore National Laboratory for predicting the performance of antenna wire elements in close proximity to flat earth. In an early version (NEC-l), the effect of the air-ground interface was included by applying a plane-wave Fresnel rei:lection coefficient approximation to the field of a point source. The NEC-2 version, while still retaining the Fresnel reflection coefficient model as an option, provides a more accurate ground model by numerically evaluating Sommerfeld integrals. The version NEC-3 extends the NEC-2 version to cases for bare wire segments below the air-earth interface. Version NEC-GS utilizes rotational symmetry to provide a more efficient version of NEC-3 for the case of a monopole element with a uniform radial wire ground-screen (GS).

Results of the various versions are compared with each other and with other mndels. The input-output format of the NEC-GS version is discussed. It is concluded that the NEC-3 Sommerfeld integral option in the NEC-GS version is the best available model for monopole elements with electrically small radial-wire ground planes. [Vol. 8, No. 2 (1993), pp 44-71]

The Applied Research Laboratory

The Pennsylvania State University

P. O. Box 30

State College, PA 16804

Electrical and Computer Engineering

The Pennsylvania State University

University Park, PA 16802

The Applied Research Laboratory

The Pennsylvania State University

P. O. Box 30

State College, PA 16804

Field strength variations produced by an orbiting aircraft dual trailing wire VLF transmitting antenna are investigated. The towplane is assumed to be executing a circular orbit at a constant altitude and speed. A steady-state mechanical model is adopted for determination of the shape of the dual trailing wire antenna. The exact current distribution on this antenna is calculated using the Numerical Electromagnetics Code (NEC) which is based on a method of moments solution of the Electric Field Integral Equation (EFIE). A propagation code developed at the Naval Ocean Systems Center (NOSC) called TWIRE has been modified to be used in conjunction with NEC. This modified version of TWIRE has been called TWIRENEC. The TWIRENEC code uses the current distribution information provided by NEC to determine the dipole moments for a segmented antenna. The wire segmentation geometry and corresponding dipole moments are then used to calculate the electric field strength as a function of distance and azimuth in the earth-ionosphere waveguide. The waveguide can be considered as either horizontally homogeneous or inhomogeneous. It is demonstrated that the periodic variations in field intensity resulting from an orbiting transmitter are a function of receiver position. These periodic variations can range from a small fraction of a dB to several dB depending upon the location of the receiver with respect to the transmitter. A point dipole approximation of the dual trailing wire antenna is suggested for use in the study of VLF radiation excited by an orbiting antenna in the presence of wind shear. The point dipole approximation is applied to estimate the field strength variations caused by a yo-yo oscillation of the transmitting antenna as it orbits. These yo-yo oscillations are characterized in terms of the change in verticality of the point dipole which occurs over one complete orbit. [Vol. 8, No. 2 (1993), pp 72-92]

Electrical and Computer Engineering Department

The Pennsylvania State University

University Park, PA 16802

Electrical and Computer Engineering Department

Naval Postgraduate School, Monterey, CA 93943

The Numerical Electromagnetics Code (NEC) was used to evaluate the admittance and the electric near and far fields of a monopole antenna mounted on a cubical box over a perfectly conducting ground plane. Two models of the box, employing surface patches and wire grids, were evaluated. The monopole was positioned at the center, the edge, and at a corner of the box's top surface. NEC admittance results were obtained and good agreement was found with experimental data and with results from PATCH, another independent electromagnetic modeling code. Results are presented in contour and 3-D formats for the near fields and polar format for the far field radiation patterns using surface patch and wire grid models in NEC. Excellent agreement was obtained for both approaches in NEC after finding the optimum number of patches and wire grid segmentation to obtain convergence. This paper provides guidelines for convergence for both modeling approaches and indicates a six-fold savings in run-time for the surface patch method. Furthermore, results are presented in modern graphical format for near field comparisons of the two NEC techniques. [Vol. 8, No. 2 (1993), pp 93-113]

Department of Mathematics and Computer Science

University of San Diego

San Diego, CA

Research and Development Division

Naval Command, Control and Ocean Surveillance Center

San Diego, CA

As modeling systems mature, they become larger, more complex, and more difficult to maintain. Modeling tools increase in number and complexity. Frequently they are written in different languages and require data in different formats. Databases also increase in size as modeling systems are applied to new and more complex problems. Engineers spend large amounts of money trying to integrate tools and data that are basically incompatible. Unfortunately, budgets do not grow at the same rate as the complexity of our modeling systems and databases. To maintain productivity, it is necessary to design modeling environments that can handle large amounts of data in flexible ways and are simple to maintain and upgrade.

This paper describes a new environment developed by the authors for the modeling of communication antennas based on a relational database management system. This approach simplifies the task of integrating a set of heterogeneous programs with incompatible data formats. The relational database provides a common store for all modeling objects including the antenna, platform, ground, electromagnetic sources, currents, charges, and fields, and model history. The database management system provides the organization, storage, and retrieval functions and some of the data input, validation and display functions for the antenna models. The main advantages of this approach are its ability to grow as new tools and capabilities are added, its portability to other machines and operating systems, and the ability it provides engineers to easily share data among themselves and with other modeling applications.

This work was conducted for the Naval Ocean Systems Center as part of the Navy Summer Faculty Research Program, a cooperative program with the American Association for Engineering Education (ASEE). [Vol. 8, No. 2 (1993), pp 114-127]

GE Aircraft Engines M/D J185

1 Neumann Way

Cincinnati, OH 45215-6301

An automatic method of synthesizing resistive tapers is developed. This method embeds a hybrid moment method/Green's function inside a nonlinear optimization package. Using this technique, resistive tapers are rapidly synthesized for complex scatterers which can consist of multiple resistive strips, as well as large, arbitrary conducting regions. The method is applied to the optimization of resistive tapers that reduce the diffraction from conducting scatterers. [Vol. 8, No. 2 (1993), pp 128-143]

Dept. of Electrical and Electronic Engineering

University of Stellenbosch

Stellenbosch 7600

South Africa.

Parallel algorithms are presented that are suitable for the solution of the system of linear equations generated by moment method problems on local memory Multiple Instruction, Multiple Data (MIMD) parallel computers. The two most widely used matrix solution algorithms in moment method codes are described, namely the conjugate gradient (CG) method and LU decomposition. The underlying philosophy of parallelism is briefly reviewed. Suitable parallel algorithms are then described, presented in pseudo-code, their timing behaviour analyzed theoretically, and timing results measured on a particular MIMD computer--a transputer array --are presented and compared to the theoretical timing models. It is concluded that efficient parallel algorithms for both the CG and LU exist and that MIMD computers offer an attractive computational platform for the solution of moment method problems with large numbers of unknowns. [Vol. 8, No. 2 (1993), pp 144-175]

Swiss Federal Institute of Technology

Zurich

Many terms and ideas used in numerical methods have their origin in analytical mathematics. Despite the well-known discrepancies between number spaces of computers and those of good old mathematics, the consequences of applying mathematical theorems to numerical methods and the importance of physical reasoning are often underestimated. The objective of this paper is to demonstrate that introducing 'a priori' knowledge of a problem into a numerical code can lead to superior numerical techniques but it may violate analytic dogmas at the same time. [Vol. 8, No. 2 (1993), pp 176-187]

Microwave Device and Physical Electronics Laboratory

Department of Electrical Engineering

University of Utah

Salt Lake City, Utah 84112

The method of moments reduces to the boundary-residual method or the point-matching method with a suitable weighting function. This paper shows another means by which these three methods can produce equivalent results. Arguments are given as to why point matching can fail to converge, while the other two methods rigorously converge. An example is given to support these arguments. [Vol. 8, No. 2 (1993), pp 188-202]

Graz University of Technology, Kopernikusgasse 24, A-8010 Graz, Austria

Problem No. 10 of the TEAM Workshops is solved by three different finite-element formulations using a magnetic vector potential with the Coulomb gauge and an electric scalar potential. Allowing the normal component of the vector potential to jump at iron/air interfaces yields results in good agreement with measurement data. [Vol. 8, No. 2 (1993), Special Section in Team Benchmark Problem Solutions, pp 203-215]

Graz University of Technology, Kopernikusgasse 24, A-8010 Graz, Austria

Problem No.13 of the TEAM Workshops is solved by two scalar potential and one vector potential finite-element formulations. The results obtained by the different scalar potential methods are identical and their agreement with those yielded by the vector potential approach and also with measurement data is satisfactory. [Vol. 8, No. 2 (1993), Special Section in Team Benchmark Problem Solutions, pp 216-225]

Tampere University of Technology

P.0.Box 692

33101 Tampere 10

Finland

Argonne National Laboratory

Advanced Photon Source

Argonne, Illinois 60439

U.S.A.

Four solutions for the TEAM magnetostatic benchmark #13 are presented. The problem was solved with the three dimensional volume integral code CORAL, formerly called GFUNET. A series of models were solved with increasing discretization in order to study the convergence and the charged CPU-time. [Vol. 8, No. 2 (1993), Special Section in Team Benchmark Problem Solutions, pp 226-231]

SOLUTION OF TEAM BENCHMARK PROBLEM #9 (Handling Velocity Effects with Variable Conductivity)

School of Electrical Engineering

Georgia Institute of Technology

Atlanta, GA 30332-0250

Users often raise the question of whether it is possible to analyze eddy current problems with velocity effects within codes that are not programmed to account for movement. This paper looks at a technique for applying a conventional boundary element technique to the analysis of a velocity induced eddy current by altering the conductivity of the conducting medium as a function of position. Results of the predicted B fields for v=0 m/s and v=10 m/s are compared to the analytical solution of a coil traveling axially down the center of a conducting tube. Good agreement is achieved; further refinement could be realized by iterating on conductivity if necessary. [Vol. 8, No. 2 (1993), Special Section in Team Benchmark Problem Solutions, pp 232-243]

The assistance of Mrs. C. Freislich at the Department of Electrical and Electronic Engineering, University of Pretoria, Pretoria, South Africa in the preparation of these abstracts for publication, is gratefully acknowledged.