Dynamic analysis of deep rock with implicit-explicit conversion method

I am doing the numerical simulation about dynamic analysis of deep rock. Firstly, the implicit solution for initial geo-stress. Secondly, change the boundary conditions goes to explicit model, and the initial geo-stress is transferred by "dynain" file.  When I removed the transitional constraints from the side of model, the initial stress field shows "collapse" before dynamic stress arrived. Is it necessary to apply the transitional constraint on the boundary of model?  For some cases, there must be free surface in the model, so, how to modeling these cases?