2.2.5 Visualization

Technical data differ strongly from that used in other fields. While for many applications such as databases a huge amount of data of the same data-type is stored technical data is hierarchically structured. Proper visualization of three-dimensional data for device simulation is not an easy task. Thus, standard visualization tools are not applicable. Since the demands for proper visualization tools rise continuously, powerful visualization hardware is necessary.

The main problem of three-dimensional visualization is that two-dimensional data can be shown as three-dimensional graphs where the shown quantity is drawn in the third coordinate. For the visualization of three-dimensional data a four-dimensional space would be necessary which is beyond the imagination of most people.

Three-dimensional data, therefore, can only be visualized on the surface of the device, by two-dimensional cuts through the device, or on discrete values on a specific grid. For the analysis of simulation results values at different regions in the device must be queried. Several algorithms for presentation of data must be available, like iso-surfaces and iso-lines, cuts, multiple cuts or projections.

As a precondition, an unhampered motion inside the device must be possible which is usually the weak point of most visualization tools. For an efficient motion, additional input devices like slide bars or wheels would be useful alternatives to conventional input devices like mouse our keyboard.

Useful tools for visualizations seem to be TECPLOT [92] of AMTEC used by ISE and Open Visualization Data Explorer [29,93] of IBM.

As a toolkit the Visualization Toolkit (VTK) [94] is very useful which provides several algorithms like iso-surfaces, iso-lines or cuts. Nearly all three-dimensional pictures shown in this thesis are drawn using VTK.

Robert Klima 2003-02-06