Half-wavelength dipole antenna (FDTD)

Half-wavelength dipole antenna (FDTD simulation)

The movies below show results of a 3D simulation of a center-fed half-wave dipole antenna. The antenna, from left to right, is added as a black line in the second and third movie. On the left, the E-field component in the direction of the antenna wire is shown and the other two movies show the poynting vector (components towards viewer and modulus).


Field E_x (horiz. comp.) blue=negative, yellow=positive	S_z (towards viewer) logarithmic scale black...white = 60 dB	Poynting vector \|S\| logarithmic scale black...white = 60 dB

All movies are in a plane about 1/20 wavelength above the antenna, which is seen from above. The excitation is a sine with frequency near the antenna's half-wave resonance. The E-field shown here is the field component that has to be zero at the location of the wire, since there can be no field parallel to a good conducting wire. Therefore we see only three regions of strong fields: near the middle, where the wire has its excitation gap, and outside the endpoints.

The plots of the Poynting vector might serve to show which part of the wire is mainly responsible for creating the radiated field. The movie of the modulus, |S|, does not give a clear answer, since at close distances the pointing vector reaches high values along the entire wire. The fact that |S| is high along the entire line just means that energy is shifted between the endpoints and the middle region (electric and magnetic field respectively).

The movie of S_z (the component towards the viewer) seems to indicate that most of the energy that leaves the antenna is coming from the middle part. At larger distances, the modulus |S| shows that the energy goes mainly to directions perpendicular to the wire, as would be expected for a dipole radiation pattern.

The simulation was done on an FDTD grid with size: nx*ny*nz*nt = 80*80*50 * 210. Absorbing boundaries are present (Berenger's technique) consisting of the outermost 10 layers. The code that computed it is in the following files:
dip_fdtd.f90
bin_io.f90
gif_util.f90

After that, the freeware tool gifsicle was used to make the two movies, combining the separate pictures of timesteps 200 to 310, with commands like:
gifsicle -l -d 25 S2??.gif S30[04].gif > Sz.gif
gifsicle -l -d 25 E2??.gif E30[04].gif > Ex.gif

For more FDTD results see Gouy phase-shift

Jos Bergervoet,

bergervo@[this_site].nl