|
 |
 |
| Thin-Wire Moment Method Codes of Antennas |
| Source: Author: Published:1268175223 |
Thin-wire codes that assume only filamentary currents are readily available. We have experience with NEC, the Richmond code (ASAP), and AWAS, a commercial code. All have advantages, but they take time to  learn. A commercial code with a graphical interface makes the input and output easier: for example, for NEC. These pay for themselves quickly by saving time. NEC can include plates, but since it uses a MFIE (magnetic field integral equation) for them, it is limited to closed bodies. When accuracy becomes important, it is necessary to decrease the segment length and increase their number. These codes use matrix inversion with calculation time proportional to and a matrix fill time proportional to . Run time increases enormously as the number of segments increases.
The commercial code AWAS determines the segmentation, while the user of NEC must specify it. The rule is to use at least 10 segments per wavelength, but initial analysis can tolerate the errors due to using fewer segments. The segments should be longer than the diameter, and care must be taken that the segments do not overlap because the radius of the wires is too large. Solid objects, such as plates, can be modeled as wire frames, with the rule that the perimeter of the wire equal the spacing between the wires. This rule can be violated, but a test of the convergence should be made. When we model slots in a solid object, we cannot apply the perimeter equalspacing rule because the slot will disappear. These codes compute the radiation pattern more accurately than the input impedance due to simplistic source models, and we may have to build the antenna to determine the true input impedance. Of course, an antenna with a good input impedance response that does not have the required pattern is useless.
We can reduce NEC run time if the antenna has symmetry with multiple inputs. The code reduces input by  allowing the user to specify symmetry. For example, a multiarm spiral analysis requires only the input of one arm. The various mode voltages are entered after the basic structure impedance matrix has been solved. If an object has M-way symmetry, the matrix fill time is reduced by  and the solution time by . The various voltage modes can be applied afterward. If we add another wire segment after specifying symmetry, the symmetry is destroyed and the program uses the full matrix. The only advantage we gain is in specifying the model because the program solves the full matrix instead of the reduced matrix. |
| [Print]
[Close] |
|
