I modeled an explosion effect on two masonary walls seperated by a layer of sand in Autodyn. The walls were modeled by brick elements and the sand was modeled via Euler multi material solver. After 450000 cycles I got a message "negative density in state" and the solver stopped. I wish that any one can suggest the reason of this error and how can I deal with it. In the Ansys troubleshooting guide it is mentioned that this message propably means that the material properties were specified inaccurately. In this guide it is suggested that to deal with such error, we can switch the variable density OFF to allow any swings in the solution to stabilize and then restart the analysis. However, I don't know how to switch the variable density OFF.