Ernst,

Two issues play in these results – 1) spatial combination and 2) pseudostatic component

1) All four load cases define the same shock. But each separate spectrum entry (Shocks #1 & #4 each have one spectrum, Shock #2 had three and Shock #3 has two) is calculated independently. So while Shock #1 calculates response to everything happening at the same time, Shock #2 will calculate three independent events (10 wagging and 50 wagging and 70 wagging) that are then summed in some manner – either SRSS or absolute (here it’s SRSS) – to develop the total response. In my mind, the latter approach would consider node 10 deflecting to the right while node 50 deflects to the left; the former approach has everything moving in the same direction. So I would not expect your first three shock cases to produce the same results.
2) Now Shock #2 and Shock #4. When a single shock does not include the entire system, the relative deflection of different support groups (grouped by the shock spectrum) is a major contributor to the total response – usually much greater than the inertial component. Both Shock #2 and Shock #4 have the same three spectra. Shock #2 has 0 deflection defined for all supports in the group (entered in the last column) while Shock #4 has those three items undefined. Where no deflection is identified, CAESAR II will calculate the deflection. The lowest mode of vibration is controlled by displacement and this displacement is derived from the lowest frequency in the appropriate shock spectrum. You specify frequency versus acceleration. Displacement = acceleration/(2*pi*f)^2. Node 10 in Shock #4 shows a displacement of 506.6 mm; that equals 200(mm/sec^2)/(2*pi*0.1Hz)^2. The displacement report clearly indicates this total response is controlled by this pseudostatic component.

The Control Parameters tab has settings for the spatial summation and pseudostatic summation method and a pseudostatic on/off switch.