Analysis of planar 4x4 microstrip antenna arrays installed on airframes

Kurzfassung

The assessment of the installed performance of antennas on aircraft is a very difficult task. Patterns can be measured by performing in-flight tests, which are very expensive and allow in general the determination of only few cuts. This costly method can be replaced by down scaling the dimensions of aircraft and antenna, whilst scaling up the frequency, so that all the electrical dimensions (in wavelengths) are kept constant. In such tests, a mock-up of the real aircraft is necessary. A scaled antenna model, which should represent accurately the radiation characteristics of the original antenna without the need to have exactly the same geometry, is also required. Alternatively to these experimental procedures, computer modelling can be used. The main advantage of using simulations for this purpose is the reduction of time and costs. Moreover, some electromagnetic characteristics, such as the current densities on the aircraft surface, can not directly be measured, but they can be estimated using appropriate computer models. In order to validate numerical predictions of installed performance, simulated results are compared with measurements performed with different mock-ups of aircraft used in the civil aviation. The antennas used in the experiments are 4x4 microstrip planar arrays, which served as radiation models of a smart GALILEO digital-beamforming antenna under development at the Institute of Communications and Navigation. Two arrays have been used: one broadside and the other with the main lobe pointing to 40° from the boresight. During the measurements, these antennas simulated the real smart array operating under different beamforming conditions. The single elements are right-handed circularly polarized patches that are coupled to the feed lines by means of slots. The optimal axial ratio is obtained at 22 GHz. The top part (patches and slots) is the same for both arrays. For producing different beam directions, different feed networks have been designed. In both cases, all the 16 patches are fed with the same amplitude. For the broadside array, the elements are fed with the same phase, whereas a progressive phase shift of 140° has been applied along the horizontal direction. The installed performance simulations have been done using the FEKO electromagnetic simulator. Physical optics (PO) and the multi-level fast multipole method (MLFMM) were the numerical techniques used in the analyses. Further information about the choice of the position to install the arrays on the mock-ups and details describing the simulations will be given at the conference. The measurements with the mock-up have been performed in DLR’s antenna test range. The details concerning the measurement setup will be discussed at the conference. A comparison of computed and measured radiation patterns will be presented. The numerical predictions are in good agreement with the experimental data, especially in terms of co-polarization.