Hello and thank you for your help. nFor Peter : I did it. I performed a Shock response spectrum with RESP command of APDL (also with Matlab and th differences are minimum) , and I did the simulation with response spectrum approach from Modal simulation. I did also a MSUP simulation. The transient simulation is very important for me because I need to know the beahviour of some contacts during the shock. nFor cs: Do you mean to fix the basis with fixed support and applying the accelerations to the entire system? Is it correct? I thought to apply for each support a Remote displacement that is the double integration of my shock acceleration (in this way I have absolute results). nMy doubt is:nI have a complex model, where there are many components. The test process of shock impulse is conducted with the application of acceleration (with shaker) to the fixed points (screws points). To apply the acceleration to the entire model fixing the base nodes (the same acceleration to all the components - the rapid way inside workbench) is the same of to apply the acceleration to only the fixed points (screws points ). I mean in theoretical term. If the body directly connected to the shaker is not rigid I think that I would have different results with the two approaches. But maybe I lost this lesson during my University course. nThank you.