For dielectric breakdown analysis, a common procedure is applied which is based on obtaining the field data from FE electrostatic field analyses. An example is given of the simulation of the shielding ring and terminal lead of a typical power engineering transformer using CST EM STUDIO® (CST EMS). The model was provided by Dr. Beriz Bakija, Siemens Transformers, Nuremberg, Germany.
Important in this process of estimating the dielectric breakdown risk  is:
1) Accurate modeling of the geometry. Models from CAD data such as Pro/E can be seamlessly imported and easily modified for e.g. associating features of the geometry with parametric variables.
2) Application of appropriate mesh topology such as curved tetrahedral elements to ensure accurate modeling of components with rounded features. This is especially critical for breakdown estimation....
3) Mesh convergence of field results obtained by mesh adaption with supporting features such as "snap to geometry" which ensures that the simulated geometry remains true to the imported geometry during the mesh refinement process. This avoids the creation of artificial singularities which may otherwise arise.
4) Higher order elements are applied to obtain highly precise fields.
5) Integrated post-processing and extraction of field data along lines. Export of data for the necessary estimation of breakdown fields in external user tools.
The position of the shielding ring above the windings is shown in the transformer cut-away view in figure 1. The terminal lead is not shown (opposite side of the transformer). The model was fully constructed in CST EMS.
Figure 2 shows a 3D view of the transformer lead system. A brief summary of the components and material attributes is shown in figure 3.
CST EMS allows the user to efficiently designate materials to groups of components. In a similar manner, the potentials can be applied to groups of objects to reduce the effort in defining the potentials. CST EMS automatically detects electrical connections between components. Therefore, the potential only needs to be applied to one component of several touching objects. CST EMS ensures that these components inherit the same potential.
A further feature of CST EMS is the handling of conducting PEC components which have not been attributed a potential by the user. The user may determine what potential is applied to these components - grounded or floating. Figure 4 shows just the 250 kV definition since all other components are set to grounded in this model.
Automatic meshing may be supplemented by mesh refinement or manual meshing. A critical issue in the application of mesh adaption is the mesh snapping feature which ensures that the adapted mesh stays true to the original geometry. However, an experienced user may wish to avoid additional mesh refinement steps and manually apply local mesh refinement. The maximum mesh step size may be attribued to single objects or groups of components.
The accuracy of the results depends on the topology and order of the elements used. For this simulation, 2nd order basis function elements were applied. The Multi-Grid algorithms used in CST EMS' solver technology ensure good scalablity, efficient memory usage and fast solution of systems that have a large number of unkowns. Even larger and more complicated problems can be handled with CST EMS.
The calculated electric potential is shown in figure 5 with a superimposed view of the surface mesh.
Figure 6 shows the step required to generate the field lines. Multiple faces or even single points may be defined by the user on the objects of interest. The field lines are then computed automatically in the integrated visualization engine. Typical field lines are shown in figure 7.
The data may be exported and processed in a similar manner to that shown in references  and  to estimate the breakdown risk.
This article has served to demonstrate how CST EMS is critical in the simulation of transformer and other power engineering components and devices.