Meshing Air Bearings
This case report is written after a case of Schut Geometrische Meettechniek bv Groningen. Herein it was the assignment to mesh an air bearing, made by Schut. This assignment rose after the conclusion of an earlier case, in which the mesh was made by SALOME, but did not seem to be as suitable as needed. Therefore the mesh should have been created more manually.
In order to make the mesh as suitable as possible, we have chosen to create the mesh between the points with a Delaunay triangulation, first in 2 dimensions and then layer every triangle to 3 dimensions. This Delaunay triangulation was already implemented in a program called Triangle, which we used for better results. However before creating the mesh in Triangle, we needed to give in the coordinates for which we wanted to create the triangles. This could be done manually, but that would not be efficient enough. Therefore we wrote a script in C, in which we gave in the geometry we wanted and we would then receive some coordinates on the boundary of this geometry. In this the geometry remains variable and we can choose how many coordinates we want to receive, which was one of the preferences of Schut.
After running Triangle, a lot more needed to be done. Triangle only makes lists of points and edges for the 2 dimensional mesh, but for meshing an air bearing we need 3 dimensions, so we needed to layer the 2 dimensional results. Also in 3D points and edges are no longer sufficient enough to create the mesh, we also needed faces and cells. This is what took the most time, since with this we had to rearrange and order the results and the output had to be rewritten. This however gave the nicest visual results (see picture). With this output the case came to an end and Schut was able to use these results for further investigation and computations regarding fluids and pressure distributions in the air bearing.
Maike Jaspers
