radarfdtd.txt

a full 3D simulation of electromagnetic waves with efficient absorbing boundary a full 3D simulation

NAME
    radarFDTD - a full 3D simulation of electromagnetic waves with
    efficient absorbing boundary conditions

SYNOPSIS
    radarFDTD directory [configuration file]

DESCRIPTION
    When started the program needs a directory as a command line
    parameter. If there is no second parameter defined it assumes
    the file is named sim.cfg. After starting it will output quite a
    few informations about the simulation it is going to start. The
    most important part is the section about memory, do not try to
    run the program in the swap area, since this will surely slow it
    down and may risk your hard disk which may start to smoke after
    a few weeks of hard work :-(.

CONFIGURATION FILE
    In this section there will be a short description of all
    sections within the file sim.cfg. Feel free to experiment with
    those settings, but do not blame me, if something goes entirely
    wrong... ;-)

    Since I am not a native English speaker, I hope you are not
    offended if I left some mistakes in here.

    Last thing, before we start: all lines which are empty or have a
    leading pound sign '#' are ignored (BTW, can someone explain to
    me why this double cross is called pound sign?).

  VERSION NUMBER

      # version number is...
      3

    This just states the version number this configuration file was
    written for. This is checked by the program and if it does not
    match the internal layout, it will not use it.

  MEMORY

      # maximum number of bytes the program may allocate
      100000000

    This is the number of bytes the program is allowed to allocate.
    Please change this according to your own need (or the memory
    limits of your computer; remember: this limit should be less
    than the physical amount of memory installed)

  MAXIMUM PML CONDUCTIVITY

      # insert maximum pml conductivity here [S/m]
      0.025

    There is a fixed equation which rules the rise of conductivity.
    If you are interested in a more detailed explanation, have a
    look at the paper of Fang and Wu from 1996 (citation can be
    found in the README file). As a good first guess try something
    between 10 and 50 mS/m, which should be reasonable good.

  MAXIMUM STRETCHING

      # insert maximum pml-stretching here [1]
      4.0

    Stretching should prevent evanescent waves to keep on moving
    within the perfectly matched layers, do not use too high values
    here, 2 to 4 seems to be quite reasonable.

  STEEPNESS

      # insert steepness here
      2.0

    The higher the value for steepness is, the less reflection will
    occur at the boundary between the simulation space and the
    absorbing layer which is generally a good thing. But
    unfortunately a high value here means that there will be a high
    reflection before a wave manages to reach the layer, which is a
    PEC (perfectly conducting layer)

  SIMULATION LENGTH

      # Length of Simulation [s]
      30.0e-9

    This just states the overall time duration of the simulation in
    seconds. Do not try to run a simulation too long, i.e. for a few
    seconds if your simulation space is only a few meters wide... it
    will take ages

  SIZE OF SIMULATION SPACE

      # size of simulation space (number of boxes in x-, y- and z-direction, respectively
      41 41 41

    These three numbers specify how man cells are used in each
    direction (let us call them x,y and z for the time being). I
    usually use odd numbers, so I can place the transmitter in the
    very middle of the simulation box, but handle this as you like
    to

  CELL SIZES

      # real size of a box [m] (x,y,z)
      0.025 0.025 0.025

    Here you specify how big (or small) a single Yee-cell should be,
    in this example this is about 2.5 cm (which is roughly an inch
    ;-)). Be careful with this selection, since you should have at
    least 10 boxes per main wavelength you want to simulate,
    otherwise you might get strange wavy things out of the
    simulation.

  TIME STEP SIZE

      # this number is a bit special: if positive it is the size of a time steps in seconds,
      # if negative, a matching time step is computed via the Courant-criterion
      # and finally divided by this number, i.e.