Results of recent sink tests

A shift in the xy plane by 1/10th dx DID NOT form a sink (not in AMR, not in fixed). Evidently, in order to form a sink the potential minima in the grid must be well defined — slightly shifting the grid by 1/10 dx is NOT large enough to force the minima to be in a single center cell.

In contrast, shifting the grid by ½ dx in x and y successfully restricted the minima to reside in a single cell, at the center of the grid (in x and y). Though, it was along a face in the z plane. Since the potential was symmetric along x and y, a sink was able to form. However, since the potential was asymmetric along z, a kick happened.

The above tests were done using an even number of cells — 86 in each direction. They were prompted by the fact that a cell formed a sink with 85 cubed fixed, but not when using AMR with a similar eff. resolution. This indicated that sink formation rested on whether the potential was centered in the middle of the grid, located in a single cell, as it was in 853.

Comments

No comments.