Bitte wählen Sie ihr Lieferland und ihre Kundengruppe
In this paper, a visualization method for large quantities of non-ideal parts is presented. Purpose of this method is the support of the product designer in tolerance definition, to give him better understanding of the impacts of tolerance definition. The three-dimensional display of simulation results improves the interpretation of tolerance effects and is intuitively understandable for the product developer, in contrast to the statistical results of commercial computer aided tolerancing software. A mapping of statistic values to a graphical representation (i.e. transparency) establishes the direct connection between abstract simulation results and part geometry. By the integration of an arbitrary number of simulation results in a single volume dataset, it is possible to visually process and interpret great amounts of data. The visualization in a scene is an advance compared to the ordinary way of analyzing distribution curves for chosen measurement points. It is no longer necessary to choose measurement points for the analysis, instead, a whole area of interest can be analyzed. For fast and efficient use of the proposed method, the deviating geometry has to be generated automatically by a Monte Carlo simulation that preserves the defined tolerances. Also, the positioning of non-ideal geometry, which is currently the bottleneck of our method, has to be accelerated to allow faster analysis. This can be done by parallelizing the needed calculations. The combined visualization of volume dataset and nominal mesh of the examined geometry is not yet solved satisfying. It has been tried to render the nominal mesh as a semi-transparent face, but this interfered with the visualization of the volume. In the current implementation, a surface can either intersect a voxel and increase the counter by one, or not. The hit value is increased by one even if the intersection area is very small, for example through a voxel corner. A better quality of the visualization could be reached by raising the voxel value dependent on the size of the intersection surface, comparable to anti-aliasing. Another possible analysis method that will be tested is the calculation of differences between two calculated volume datasets. This would enable the visualization of the geometric position and degree of variations when changing the tolerance definition. The volume visualization could also be applied on real measured surface data. Results of thousands of measurements can be combined in a single image for easier interpretation. The volume datasets of surface measurements and simulations can be compared to check if tolerance simulations produce realistic output.