Hello,
Again, this is a numerical issue due to small roundoff errors created by not fully forcing the positive/negative conduction inflection point to be at zero current/zero voltage. Please change the 0A column voltage values for the conduction losses to all be 0V.
Thanks,
Kris