A PROCEDURAL INTERFACE FOR MULTIRESOLUTIONAL VISUALIZATION OF GENERAL NUMERICAL DATA

T. Geßner - B. HAASDONK - R. KENDE - M. LENZ - M. METSCHER - R. NEUBAUER - M. OHLBERGER - W. ROSENBAUM - M. RUMPF - R. SCHWöRER - M. SPIELBERG - U. WEIKARD

Together with a rapid development of computer hardware, sophisticated, efficient numerical algorithms allow simulation computations of complex physical phenomena. Methods, such as Finite Volume, Multigrid Finite Element schemes, Sparse Grid, Wavelet approaches, and Particle Methods or Gridless Discretizations all carry their own, tailored data structures, which reflect the decomposition of the function spaces as well as the decomposition in physical space.
Multiresolutional visualization on numerical data is described as an indispensable ingredient of real time interactive post processing. The typically enormous data bases are locally resolved on different levels of detail to achieve a significant saving of CPU and rendering time.
For efficient data analysis and graphical post processing the method of spatial, hierarchical subdivision combined with the recovery of the local function spaces is presented. To manage a variety of different numerical data a general procedural interface to arbitrary large numerical data sets is presented. This leads to a visualization beyond prescribed data formats. Discrete numerical solution data is directly addressed in the user's data structures. Furthermore the procedural interface supports a flexible method of local error measurement, again encapsulated in certain user supplied functions. The software conception, its data classes and methods are described and the setup of the corresponding procedural user interfaces is discussed in detail.
Examples from various numerical methods and different data bases underline the applicability of the proposed concept.