The thesis addresses the problem of determining the electrostatic field in the case of axially-symmetric conductive bodies. For such structures analytical solutions are not always applicable, and therefore a numerical approach is required that allows handling
general geometries and multi-body configurations.
A matrix-based method was developed and implemented, rooted in the fundamental Green’s function $G_{0}$ and a local approximation $G_{\mathrm{apr}}$, where the body’s contour is discretised into arc segments of equal length. The assembled system of linear equations enables calculation of surface charges, on the basis of which the potential, electric field, and surface charge density are determined. In particular, the singularity in the diagonal entries $\mathbf{A}_{ii}$ of the matrix was addressed by isolating the singular part and using the Cauchy principal value.
The method was tested on simple bodies (sphere, ellipsoid, tube) as well as on a more complex surface of revolution, where it demonstrated numerical stability and reproducibility. Results are presented through quantitative graphs and qualitative visualisations, such as equipotential surfaces, field lines, colour maps of field magnitude, and charge distributions. It was found that charges concentrate in regions of sharp geometrical transitions, and that the surface charge density $\sigma$ attains its highest values where the field strength is greatest.
The method is limited to axially-symmetric bodies of revolution without vertical walls, and applies only to ideal conductors in a homogeneous electrostatic environment. Possible extensions include adaptive segmentation, treatment of different dielectric media, and the use of the Galerkin method.
For validation the method was also compared with the method of images in the case of two spheres. The analysis showed that the method of images converges more rapidly and for practical purposes a small number of auxiliary charges is sufficient; however, it is restricted to simple geometries. The developed method has proven to be a more universal tool for analysing a broader class of axially-symmetric structures.
|