Search

This paper presents a comparative study of the treatment of lumped elements in two different full-wave electromagnetic formulations: the finite-difference time-domain (FDTD) technique and the technique of integral equations. Although these two approaches are basically different, they are seen to exhibit the same kind of numerical troubles when lumped elements are involved. More precisely, lumped elements are systematically associated with numerical parasitic elements that can seriously alter the calculated results. As a solution, this paper proposes a general alternative formulation, which permits one to suppress these numerical parasites. This formulation is successfully applied to the FDTD technique. A simpler numerical deembedding procedure, which does not fundamentally transform the original formulation, is preferred for the integral-equation technique. Index Terms - Finite-difference time-domain, integral equations, lumped-circuit element, numerical deembedding.

An efficient procedure is developed for simultaneously characterizing the time-domain and frequency-domain behavior of electromagnetic devices. The procedure works in the complex-frequency plane - called the s-domain - and provides an analytical expression for the behavior of the device at any frequency and for any transient excitation. This analytical expression is obtained by first evaluating a reduced-order model of the poles and zeros of the device. These poles and zeros are then used to characterize the device in terms of rational polynomials in the s-domain. Two different methods for evaluating reduced-order models are presented. One is called asymptotic waveform evaluation (AWE) and is combined with the finite-element method; the other is called adaptive Lanzcos-Pade sweep (ALPS) and is combined with the boundary-element method. The resulting reduced-order models provide the frequency-domain behavior of the device over a broad bandwidth. Using the Laplace transform, these reduced-order models also provide the time-domain behavior. Several numerical examples have been run using commercial electronic design automation (EDA) software to demonstrate that this solution procedure is a highly efficient and accurate way to characterize the electromagnetic performance of real-life devices. Index Terms - Electromagnetic analysis, electromagnetic transient analysis, finite-element methods, Maxwell's equations, reduced-order systems.

A two layer model is proposed for the examination of photothermal radiometric signal's dependence on a two layer semiconductor's thermophysical and optoelectronic properties. The semiconductor's infrared emission caused by pump modulated laser beam was interpreted through thermal and electronic effects. Results verify the model's potential to extract electronic and thermal parameters and provide a quantitative analysis of photothermal radiometric experiental data.

The authors of a previously published paper on the dependence of electromagnetic (EM) energy absorption concluded that homogeneous modeling of the human head is suited for assessing the spatial-peak absorption for transmitters operating at 900 MHz or below. Additional studies became necessary for the frequency bands utilized by new mobile communications systems (i.e., 1.5 and 2.5 GHz) since some peripheral tissue layers have a thickness of the range of [mathematical expression not reproducible in ASCII]/4-[mathematical expression not reproducible in ASCII]/2. The results of the simulations combined with worst-case considerations confirmed the anticipated and more complex relationship between absorption and anatomical details at these higher frequencies. Nevertheless, a homogeneous representation of the head is suited for assessing the maximum specific absorption rate (SAR) in the head of the user of mobile telecommunication equipment (MTE) if the appropriate dielectric parameters are chosen.

After being used extensively for remote sensing, microwave radiometry is also applied for non-invasive thermometry, in medecine and industry. Several works have already shown the interest of such technique for thermological applications.

In this paper the authors compute the magnetic fields generated by a lot of typical power line configurations and compare these values with the results given both by analytical models and by measurements. The computations have been made by means of a flexible calculation code developed by the authors. This code overcomes some simplifying assumptions introduced in previous analytical studies. In fact, it is able to account the effective line route of the power line under analysis in a three-dimensional space. There are no limits on number and position both of energized conductors and of test points.

A robust new algorithm for electromagnetic (EM) optimization of microwave circuits is presented. The algorithm (TRASM) integrates a trust region methodology with the aggressive space mapping (ASM). The trust region ensures that each iteration results in improved alignment between the coarse and fine models needed to execute ASM. The parameter extraction step is a crucial part of the ASM technique. The nonuniqueness of this step may result in the divergence of the technique. To improve the uniqueness of the extraction phase, we developed a recursive multipoint parameter extraction. This suggested step exploits all the available EM simulations for improving the uniqueness of parameter extraction. The new algorithm was successfully used to design a number of microwave circuits. Examples include the EM optimization of a double-folded stub filter and of a high-temperature superconducting (HTS) filter using Sonnet's em. The proposed algorithm was also used to design two-section, three-section, and seven-section waveguide transformers exploiting Maxwell Eminence. The design of a three-section waveguide transformer with rounded corners was carried out using lip HFSS. We show how the mapping can be used to carry out Monte Carlo analysis using only coarse model simulations. Index Terms - CAD, electromagnetic simulation, Monte Carlo analysis, microwave filters, optimization methods, space mapping.