Computational Electromagnetic

Discussion in 'General Electronics Chat' started by haditya, Nov 1, 2007.

  1. haditya

    Thread Starter Senior Member

    Jan 19, 2004
    220
    0
    hi..
    any one here workin in this field? i know what the cousre involves but any info wud help..
    i intend to pursue my post-graduation in this field..
    can some1 please tell me what job prospects in both academic research and industry are like
    i am an undergrad EE student.
    my course is heavy on power engg.
    so i was wondering if computational EM with power engg application wud be the correct approach.. say things like EMI/EMC

    aditya

    ps- feels good to be back at aac
     
  2. Dave

    Retired Moderator

    Nov 17, 2003
    6,960
    144
    Hi haditya, nice to see you posting back here at AAC.

    A lot of the work I do is in the field of computational electromagnetics, more specifically electromagnetic tomography and dielectric heating technologies. These fields are different than power engineering which I have nothing more than an appreciation of the fundamentals (I suppose dielectric heating is loosely tied to power engineering) .

    As for career prospects, in UK (I cannot really comment about other parts of the world) there is a large scope for academics in the general field of computational EM. There are two main drivers for this, firstly technology has improved such that modelling and simulation of quite complex EM applications are available due to both advanced and networked computer systems and better simulation software; and secondly the focus of many EM applications has grown in recent times, particularly as we move towards a connected and wireless world.

    Sadly, it doesn't appear as though industry has thrown its full support behind computational EM as part of its R&D cycles. Why? Basically it comes down to cost and time. Systems for EM modelling and simulations are very costly and the licenses for simulation softwares such as Ansoft's HFSS (my personal preference) or CST MWS can run into the $10,000 per annum. The options are develop your own in-house package, something we did using a TLM solver, however this is time consuming and the process of simulation verification can be quite substantial. Furthermore, EM simulation can be quite time consuming in order to get accurate and credible results, and the interpretation of the results can be a lengthy process (experimental validations etc).

    Its a great industry to be in, and is only going to grow as the dependence on EM-technologies grows. From personal experience, the main players in this field have been through the academic route typically to PhD level, therefore this is a natural progression for those wishing to get into this industry.

    Dave
     
  3. haditya

    Thread Starter Senior Member

    Jan 19, 2004
    220
    0
    hey thanks dave
     
  4. Dave

    Retired Moderator

    Nov 17, 2003
    6,960
    144
    No problems.

    One piece of advice I will give you is get familiar with common EM modelling tools, this will be particularly helpful as and when you look to go into industry - they love people which experience of using the industry standard tools. You can do this rather cheaply by going to the downloads section of Ansoft's website (http://www.ansoft.com/downloads.cfm) and trying some of their free software. It is limited, and it is often related to electric/electronics circuit specific simulations, however the fundamentals are the same if you migrate up the Maxwell 3D or HFSS which are the two main EM solvers. The migration process will be quite easy from these free versions of the software. Other companies do the same, look at CST, Comsol and Ansys.

    Dave
     
Loading...