Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Computational electromagnetics books

  1. Mar 9, 2012 #1
    I would like to make an electromagnetic simulator, but I have no experience in this. I have written 3-dimensional ray tracing software for acoustic sources in MATLAB, but I did not include any boundary conditions, it was just propagation in free space through refractions.

    My only background in EM theory is griffiths and ulaby, and a student's handbook. They don't talk anything about modeling or simulating. But I want to learn more about how to model and simulate electrostatic and EM fields and simulate them graphically over time. I sat in front of my computer, and defined an object I want to model as a set of points in a grid . . and then I was at a loss of where to proceed after that without making something very complicated that my computer probably could not calculate efficiently or accurately (I can't model every charge carrier, and I don't know if a charge density over a surface will give me the kind of detail I'm looking for). Looking at wikipedia, I see that there are a lot of complex methods that I probably would not even think to do, so I feel like I'm wading into territory that is over my head.


    The books I've seen are expensive, so I want to make sure I invest in the right book. Does anyone have any experience or suggestions for a good reference book to get me started?
     
    Last edited: Mar 9, 2012
  2. jcsd
  3. Mar 12, 2012 #2
    The hard part is the boundary condition, where you have reflection, transmission, angle of reflection and transmission etc.

    I studied both your books and this is no where deep enough provide you info to do simulation. Griffiths is a vary good book to get your feet in the door but it has nothing on transmission lines and Smith Charts. Ulaby is quite good only in the section of co-ordinate systems and lossless transmission line. Rest of it are really too simple to be useful.

    For a more complete version following Ulaby, get the "Field and Wave Electromagnetics" by David K Cheng. This is the real version of Ulaby. Ulaby give you the necessary warm up for this book. I studied all three books mentioned......worked through most of the problems and re-studied again. I don't feel I am even close to writing this program you want.

    I have 8 to 9 books on this subject, I have not seen any talk about computer simulation. I think you really have to enroll in graduate classes. Books like Classical Electrodynamics by J D Jackson is the book by which all other books are being judged. This is a very very hard book to get into. Not only you need PDE, you need complex analysis, some real analysis and numerical analysis. I have not study this one yet.

    All the shape of the emitting bodies used in the three books are of simple shape like a sphere, a linear wire or cylinder or flat surface etc. None of the three books get into more challenging shape. I know Ulaby and Cheng do not even touch this pointy shape. I did not spend too much time studying the electrostatic part of Griffiths as I am more interested in the dynamics part. I don't recall it cover any pointy shape either. As we all know, more surface charges concentrate at the pointy surface on the a conductor, charge are not evenly distributed like in the ideal sphere shape. Just this one already rule out the basic books.

    The only other advanced EM book that is intermediate level and have few good review is "Classical Electromagnetism" by Jerrolf Franklin. I have this book but have not get into it yet.

    I suspect you are going to have to get very deep into EM before you can write the program. Post this question in the Classical Physics section in this forum, they have more people expert in EM.
     
    Last edited: Mar 12, 2012
  4. Mar 12, 2012 #3

    marcusl

    User Avatar
    Science Advisor
    Gold Member

    Do a search in Amazon on "computational electromagnetics" and take your pick--there are dozens. Read the reviews and pick one with lots of stars. I agree with Yungman that you need to understand E&M at a deeper level than Griffiths. Instead of physics texts like Jackson, you can look at Balanis, Advanced Engineering Electromagnetics for a more engineering and computational focus. For a nice presentation on selected topics, the two older books by Harrington (Time Harmonic Fields, and Moment Methods) are classics on fundamental E&M concepts as well as computation by moment methods, variational principles, etc. He doesn't cover finite elements, integral equations or FDTD, however.
     
    Last edited: Mar 12, 2012
  5. Apr 20, 2012 #4
    I know I'm kinda bumping this one but I'm pretty sure that computational EM is my dream niche. I'm a late sophomore EE so I still haven't taken an official EM course but I definitely plan on studying E&M this summer. Some questions:

    Would a book like Griffith's be over kill for an intro E&M book as an EE student?

    Should I study classical mechanics before E&M at Griffith's level?

    Is there a specific programming language that is heavily used in this field?

    This summer I plan to model as much stuff as I can within E&M, even if it's extremely basic. I already know several programming languages(C, Python, Matlab, etc.) so I would like to focus in on one language and one book for the whole summer. Thanks
     
  6. Apr 20, 2012 #5
    To me Griffiths is a must for EM regardless of major. I have many EM books for engineering, they don't explain the basic theory that well and more concentrate on EM wave and Tx lines which is of cause more important in EE. BUT Griffiths is so much better in the other part of EM theory. I studied Ulaby and Cheng first and worked out all the problems in their solution manual, still don't feel comfortable. I studied the third time with Griffiths and it's like a light bulb got turn on.

    On top, in chapter 9 and 10, the problem set is like an advance course of vector calculus.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook