As far as I can see Robot starts to have difficulty to converge the last of many non-linear cases for the BFGS method and the multi-threaded solver (when you restart calculations or delete some of the previous cases then divergent cases will converge). Thank you for pointing this situation out - we will check what exactly happens.
Corrected in Robot 2015.
If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.