My model does what he wants and I want to change it; Axial Distribution Problem; Master Thesis

So, let’s see if anyone can help me out here.

I am currently working on a structural-informed parametric model of a parking garage for a thesis. For this, I started with only one module of the second floor and created the first model, which can be seen in the first picture (idealized it is a frame structure with supporting ramps in longitudinal direction. The “ramp beams” should thus transfer the loads in compression).

The next step creates the problems for me. Now I want to stack the modules to get a multi-storey object. What I expected is that the pattern of compression transfer in the “ramp beams” would be continued downwards. But instead I get tension forces in the outer beams and generally different results compared to the single module. See the result in the next picture.

To show what I want to have in the end I exploded the floors and translated the reaction forces as well as vertical and horizontal forces always downwards. See picture (In principle, the individual floors should only pass on the individual forces of the supports).

Well I have been trying for some time how to manipulate the stacked model to get the desired distribution of axial forces. I think the problem is that the model is overconstrained. I have also tried varying the cross sections of the beams to give certain elements more stiffness. I have also tried with springs and joints, but nothing has led to the expected effect.

Im attaching also my grasshopper files if you want to have a deeper look at my work.
Garage_Grasshopper - SingleLoadFolded - MultiFloor2.gh (437.1 KB)
Garage_Grasshopper - SingleLoadFolded.gh (195.1 KB)
Garage_Grasshopper - SingleLoadFolded3.gh (253.4 KB)

I know it is too much at once but maybe I will be lucky and get some input.

Thanks in advance