TIME-DOMAIN ELECTROMAGNETICS USING THE ELECTRIC-FIELD INTEGRAL EQUATION--A COLLECTION OF ARTICLES BY EDMUND K. MILLER AND COLLEAGUES (original) (raw)
Related papers
An integral equation is developed for determining the time-dependent current distribution on a wire structure excited by an arbitrary time-varying electric field. The sub-sectional collocation form of the method of moments is used to reduce this integral equation to a form that can be evaluated on a digital computer as an initial value problem. A Lagrangian interpolation scheme is introduced so that the dependent variables can be accurately evaluated at any point in the spacetime cone; thus, only mild restrictions on the space and time sample density are required. The integral equation relating present values of the current to previously computed values is presented in a form that can be directly converted into a computer code. Expressions are developed for the computer time and the relative advantages of time-domain and frequency-domain calculations are discussed, providing impetus for analyses in the time domain in certain cases. Part II of this paper will present well-validated numerical results obtained using the technique described.
Presented here are the numerical results from a computer solution of the time-dependent thin-wire electric-field integral equation described in Part I of this paper. Both radiation and scattering problems are considered. The present results are validated by their Fourier transform to the frequency domain, where they are compared with independently computed data. A space—time sampling criterion is derived for predicting the highest frequency to which the time-domain calculations are accurate and found to be in accord with the numerical results. The time domain results are also shown to grovide informative insights into the radiation characteristics of specific structures. Recommendations for further work are also presented.
Journal of Computational Physics, 1973
An integral equation is developed for determining the time-dependent current distribution on a wire structure excited by an arbitrary time-varying electric field. The subsectional collocation form of the method of moments is used to reduce this integral equation to a form that can be evaluated on a digital computer as an initial value problem. A Lagrangian interpolation scheme is introduced so that the dependent variables can be accurately evaluated at any point in the spacetime cone; thus, only mild restrictions on the space and time sample density are required. The integral equation relating present values of the current to previously computed values is presented in a form that can be directly converted into a computer code. Expressions are developed for the computer time and the relative advantages of time-domain and frequency-domain calculations are discussed, providing impetus for analyses in the time domain in certain cases. Part II of this paper will present well-validated numerical results obtained using the technique described.
TWTD-Exploring Electromagnetic Physics Using Thin Wire Time-Domain (TWTD) Modeling.pdf
emiller@esa.lanl.gov 0. ABSTRACT Among the various applications for which numerical electromagnetics models can be productively employed is that of conducting computer experiments for the purpose of exploring fundamental physics phenomenology such as radiation. A time-domain model is especially well-suited for this purpose since it provides a convenient way of isolating local effects in a manner not as readily achieved using its frequency-domain counterpart. This paper reports on some computer experiments using TWTD (Thin-Wire Time-Domain), an integral-equation, time-domain model based on the electric-field integral equation.
Journal of Computational Physics, 1973
An integral equation is developed for determining the time-dependent current distribution on a wire structure excited by an arbitrary time-varying electric field. The subsectional collocation form of the method of moments is used to reduce this integral equation to a form that can be evaluated on a digital computer as an initial value problem. A Lagrangian interpolation scheme is introduced so that the dependent variables can be accurately evaluated at any point in the spacetime cone; thus, only mild restrictions on the space and time sample density are required. The integral equation relating present values of the current to previously computed values is presented in a form that can be directly converted into a computer code. Expressions are developed for the computer time and the relative advantages of time-domain and frequency-domain calculations are discussed, providing impetus for analyses in the time domain in certain cases. Part II of this paper will present well-validated numerical results obtained using the technique described.
Modelling of straight thin wires using time-domain electric field integral equations
IEE Proceedings - Microwaves, Antennas and Propagation, 1994
The thin-wire time-domain electric field integral equation is formulated starting from the extended boundary condition theorem and including a first-order approximation for the charges at the end caps of the wire. The equation obtained is solved directly in the time domain by the moment method. The effects of this approximation and of changing the situation of the matching points on the segments of the wire are studied.
IEEE Transactions on Magnetics, 2008
For the modeling of electromagnetic scatter of thin metal wires located above or under real ground, a new numerical method based on time domain integral equations is presented. The influence of lossy ground is taken into account via Fresnel space-time reflection coefficient. The integral equations are solved efficiently by method of moments combined with marching on in time procedure. The calculated results agree with those of references. The method is also verified by the experiment in near field. Index Terms-Half-space, time-domain integral equation (TDIE), wire structure.
IEEE Journal on Multiscale and Multiphysics Computational Techniques
In this work, a straight forward method of moments (MOM) procedure to solve the time domain integral equation (TDIE) is presented and applied to a wire-grid model of an arbitrarily-shaped conducting body. The conducting body is illuminated by a Gaussian plane wave. Contrary to all the available time domain algorithms, the present procedure does not involve marching in time thus eliminating error accumulation, a major source for late-time instability problem. The procedure presented in this work is conceptually simple, numerically efficient, and handles multiple excitations in a trivial manner, all the while remaining stable. The numerical procedure utilizes pulse functions for space variable and time-shifted Gaussian functions for time variable, respectively. Further, the numerical procedure adopts Galerkin method of solution implying the usage of same time and space functions for both expansion and testing. The numerical results obtained in the time domain are validated by comparing with the data obtained from the frequency domain solution at several frequencies and performing inverse discrete Fourier transform.
Time-Domain Electromagnetics: The Short-Pulse Response of a Straight Wire.pdf
The result that the validity of the Rayleigh-FFT method, hence the Rayleigh hypothesis, is independent of the incident, wavelength agrees with the conclusion d r a m b y Pet,it and Cadilhac [20] and MiUar [Zl], [22]. However, Millar [22], by analyzing t.he singularities of solutions to the Helmholtz equation found t.hat, for a perfectly conducting sinusoidal surface the Rayleigh assumption is valid for precisely 2rA/Lz < 0.448. The current results set this limit of validity at. -0.6 and at larger values for penetrable surfaces ( U << 05 mhos/m). Most results used to define between 0.448 < 2rA/Lz < 0.6 do not differ significantly from the integral equation results or t,he physical opt.ics calculations where they are valid. An example in this range appears in . I n addition, the normalized rmse in matching t,he boundary conditions in t,he vicinity of 2rA/L,-0.448 are on t,he order of 10" percent when A' = 32, 2 M + 1 -31. Increasing the number of scattering modes (2M + 1) and matching points ( N ) to more than double this number cont.inually reduces this error.
IEEE Transactions on Electromagnetic Compatibility, 2005
The transient behavior of a single straight line embedded in a dielectric half-space and illuminated by the electromagnetic pulse (EMP) is analyzed directly in the time domain using the wire antenna approach. The formulation is based on the corresponding space-time Hallen integral equation. The effects of a two-media configuration are taken into account via the Fresnel reflection and transmission coefficient, respectively. The space-time Hallen integral equation is handled by the time-domain version of the indirect Galerkin-Bubnov boundary element method. The transient response obtained using the direct time-domain approach is compared with the results obtained via an indirect frequency-domain analysis method. Some illustrative numerical results are presented in the paper. Numerical results obtained via the different approaches agree satisfactorily, i.e., the maximum deviation between the results is around 6%. Index Terms-Antenna theory, dielectric half-space, electromagnetic transient analysis, Hallen integral equation, thin wire.