EMI/EMC Computational Modeling Handbook, 2nd edition, 311 pp.
Authors: Bruce Archambeault, Colin Brench, and Omar M. Ramahi
Publisher: Kluwer Academic Publishers, 2001

There are many books in computational electromagnetics, but this book specializes in the usage of computational electromagnetics for EMC problems. Most EMC/EMI engineers come from the testing world (where EMC actually got started with the nascence of testing requirements and standards) and some come from the abstract world of electromagnetic theory. Both groups have learned to apply what they know to the often-misunderstood concepts of EMC and EMI, but the complexities of EMI problems have rendered the analytical approaches to the solution of interference and compliance problems very difficult to implement.

Now comes to the rescue the art of computational electromagnetics. For years, such computational techniques have been used for different types of EMI problems with different degrees of success (unless you understand the EMI problem, you canÕt model it properly). This book serves as a bridge for those in the EMC world who want to learn about modeling of EMI problems via computational electromagnetics (CEM). It may also be useful to the experienced CEM modeler who wants to know about the EMI world. The book is introductory in the level of difficulty and light in the mathematics. The purpose of the book is to teach the art of modeling for EMI rather than detail the CEM techniques.

The book is divided into 11 chapters. Chapters 4, 5, and 6 deal with the most commonly used CEM techniques such as FDTD, Method of Moments, and FEM, respectively. The rest of the book (Chapters 7 through 11), deals with the application aspects of EMC. The first three chapters are introductory in nature. For the experienced CEM modeler who wants to know about the applicability of CEM to EMI, I suggest starting with Chapter 1 and proceeding with Chapters 7 through 11. For the EMC engineer who is starting to learn CEM, a cursory reading of the book from Chapters 1 through 11 is recommended.

Chapter 1 introduces the reader to the basic principles of EMI and what is the state of the art in EMI modeling. The Òtool boxÓ approach, as described in the chapter, allows the EMC engineer to look at an EMI problem and choose the appropriate CEM tool (FDTD, MoM, or FEM) to analyze the situation. The choice of the appropriate CEM tool requires some intuition and experience. A successful modeler is one who has an intuition of what is the nature of the interference/compliance problem and one that has some experience on what each CEM is capable of delivering for that particular situation. The chapter ends with a brief description of FDTD, MoM, and FEM. Chapter 2 introduces the reader to the basic concepts of electromagnetic theory and eases the reader into the CEM techniques resulting from the manipulation of the mathematics embedded in Maxwell equations.

The introduction to CEM techniques starts with Chapter 3 where the FDTD method is discussed. The chapter is devoted to the details of FDTD. Two and three dimensional FDTD are discussed, with the emphasis of course on the 3D approach as outlined by the Yee cell. The chapter covers the modeling of radiation sources and the inherited dispersion issues that are part of every FDTD modeling approach. The chapter ends with mesh truncation techniques and the sources of FDTD modeling errors. Chapter 4 covers the method of moments which is used less in EMI analysis than FDTD, mostly because it is more complex mathematically, mostly used for surface currents, and more complex when using inhomogeneous media. Therefore, from the EMI point of view, only those problems in EMI where there is a need to calculate current distributions (e.g antenna emissions problems) does the method of moments provide an advantage. The material discussed in Chapter 4 is for the method of moments of perfectly conducting surfaces. Chapter 5 addresses the finite element method. The chapters covers the basic principles of FEM, such as creating the finite element matrix, matrix assembly, matrix solution and the solution of a two-dimensional Helmholz equation.

Chapter 6 prepares the reader for the world of EMI modeling. The first aspect of this is the consideration of how modeling tools can be used. Depending upon the frequency range and the physical size of the device to be modeled, quasi-static or full wave tools may be more appropriate. If the geometry of the problems permits it, two-dimensional models may be used to avoid the model complexity and higher computer resources necessary for full three-dimensional models. If detailed frequency responses are desired, the use of time domain tools is advantageous, as a wide bandwidth is modeled with a single run instead of multiple runs that are required with frequency domain methods. Therefore, it is up to the reader to weigh the pros and cons of the main CEM techniques for a given EMI problem based upon the data available for the problem and what is really wanted as a solution.

Chapter 7 presents the steps required to create practical EMC models for different computational techniques. In addition, examples of practical problems are presented to illustrate the use of modeling and to show some of the most critical areas. Good geometries are important to the construction of models, especially if such geometries are well defined. However, it is always important to realize that EMI models are often unusual in their needs and applications, and often require further attention to make sure the problems being solved are the ones of interest, and not those from a perfect model.

Every modeling task has its own priorities and criteria and the creation of EMI models can often be a difficult process. However, a guide to the steps needed to prepare for modeling, using the three main CEM techniques, is provided in the next chapter. Chapter 8 covers a wide range of modeling topics of interest to EMC engineers. Multiple stage models can be used when several sections are electromagnetically separable, and the modeling techniques can be varied to allow each individual stage to be optimized. Test sites can be evaluated before construction to show effects of changes to typical recommendations. Antennas and other measuring probes can be modeled as part of the evaluation process or as a method to allow their effects to be included in the overall results. Chapter 9 talks about the EMI modeling validation and the usage of different techniques, depending upon which is most appropriate. Validation is important in order to ensure the correctness of the model and help understand the basic physics behind the model. Measurements can be used to validate modeling results, but extreme care must be used to ensure the model correctly simulates the measurements made.

Chapter 10 covers standard EMI/EMC modeling problems. A set of well defined and designed modeling problems can be used as test beds for new software and it can serve as an important tool in the process of selecting an appropriate modeling technique. Taking time to consider the actual uses for such software and creating suitable problems is the key to getting the most value from such problems. The objective is to have a problem that is not only representative of the challenges of an EMC engineer, but also one that is not so complex that the answers can not be verified. Several examples are shown in the chapter that can be used as standard EMI modeling problems. These benchmark problems can be used to evaluate present and future modeling tools that may show up in the market. The last chapter in the book, Chapter 11, addresses advanced modeling techniques. The chapter describes the PEEC and TLM modeling techniques. These techniques are relatively new to EMC modeling activities, but can be extremely useful for certain EMC modeling applications. The PEEC technique is an equivalent circuit technique much more suitable than SPICE. PEEC is a full wave simulation tool because it does include the propagation delay. It is suited for EMC problems that include lumped circuit elements, such as power/ground plane decoupling, including capacitors, via inductance, and other printed circuit board related EMI problems. PEEC is also very suitable to interface directly with traditional quasi-static TEM based signal integrity tools to provide the full wave part of the problem for traces running over splits in their ground-reference planes, or traces with connectors between boards. The TLM converts the electromagnetic problem into a series of transmission lines. It is similar in some aspects to the FDTD, however, there is the ability of TLM to have the voltage and currents in each node located at the same point in space. This is an advantage when changing cell size or modeling very thin objects.

EMC