I must clarify here, that the staggered solver isn't at all doing the same thing here. It solves the same time step, for different fields. Here, the same field (and the same internal state) should carry over the the next timestep where a different analysis takes over instead.
Even if you only needed a single (static) time step, it would not be able to handle
static structural -> buckling analysis
since the internal fields are not mapped over at all.
And even for elastic materials, doing
static structural -> dynamic structural
wouldn't work, since there is currently no code in dynamic structural that checks a FieldManager for the initial state.
We need to add add the code that allows us to bring over initial guesses through fields (shouldn't be that difficult).
More difficult would be to bring over the *entire* internal state, which is more necessary than the primary field itself.