Hi! I'm studying a bearing capacity problem with several layers of soils either isoelastic or MC. I asked my students to modify boundary conditions to study their influence in the results and we have noticed that when you run Solve it doesn't take into account the modifications, apparently runs over previous calculations without updating the new data. This problem is partially solved by imposing a number of Steps, but even with that sometimes it gets stuck.
I have attached the .adf file with an image of the results. You can see that it is still showing the results of a previous stage when the vertical left axis was free xy.
I can send you the .adf file, I can't attach it to this message.

please attach the related script file and in there mark the line that cause the problem to show up (use // for comments).

now I understand what you mean. this is not really a bug.

when you run the model with original BC (fix xy) and model reaches to equilibrium changing the boundary condition from xy to y is not going to change the unbalance force in the model and as a result when you solve the model again it converges in one step immediately and that's why you don't see any changes.

the step run is working because it's designed to run without considering the convergence criteria but if you look at legend window in the left side of main window you will see that "equilibrium ratio" is very small which indicates that model is converged and running more steps is not really necessary at this point.

you are going to get the similar result if you follow the same procedure using different commercial programs (which I suggest you to give it a try). the only way to force the program to start over is to somehow change the unbalance force by reducing the in-situ stress or adding additional force to system or completely removing the boundary condition which I don't think is the intention of this exercise. so I suggest you to start over your model with correct BC.

I hope it was clear. please let me know if you need additional information.

also, you can save the script, delete lines that you don't need, then modify manually the script and re-run calculations.