I'm not sure I completely understand your problem. You're saying the results are inconsistent, because of your convergence criteria? You can manually tweak your goal settings in the calculation control panel, setting your own convergence criteria.
I figured out why the program was not producing repeatable results. In the Calculation Control Options, I changed the Criterion to stop to "all satisfied" This let all the goals converge and be repeatable instead of finishing the solver early.
Thanks for your help.