I am carrying out some analyzes of a RVE with prolate pore, in order to obtain a yeld locus graph (Hydrostatic Stress x Equivalent Stress), following the work of Gurson (1993). For this, I implemented an MPC routine in which I inform the triaxiality and thus obtain the curve points. But for high triaxiality, the behavior of the routine is strange, not fixing the triaxiality value that I imposed. The time for the simulation I am using 1, in the figure I show the evolution of the points in relation to the frames. Note that when the equivalent stress rises in a certain part, changing the value of the imposed triaxiality. This way I am not getting the points to complete the graph (shown in the figure) Would anyone know what could be going on?