I am simulating a two stage turbine,I am using arbitrary sliding interface, If I have defined the change grid file correctly and also the events, what can be the probabale cause for the unstabel solution, the time step is within the acceptable range, is the information in FILE.div useful? If so, how to open that?

May be it is better to explain the problem more,I am using a similar procedure to arbitrary sliding interface in tutorial 7, the change grid commands are ,If time get 0, then calculate the angular velocity in deg/s, activate cylindrical coordinate, multiply the time by angular velocity,get a vertex in which y coordinate is nearly 0. find the angular change. get the cells of first and second rotor, get their vertexes, change the position of them, set all cells and end. For attach boundaries I have defined all the boundaries in rotors as masters and the ones on stators and exit as slave, two events defined. there is no problem when writnig geomtry and problem file but the solution always give negative densities in first run and when I turn density as function of T, then solution unstable in fisrt run. Thanks