Now I use 3 of them in the attached blender file
foil1: basic foil
foilInner: like basic foil but 1mm ‘shrinked’
foilInner2: like basicfoil but 1/2mm ‘shrinked’

I boolean subtract the foilInner and foilInner2 from a copy of foil1 and get
foilShell1 (1mm thick)
foilShell2 (0.5mm thick)

so far so good.

Next I use a cube to boolean intersect with the foilShell2 to get a grooveRing:

which part exactly do you find overly complicated?
boolean operations on simple geometries work for me - so I don’t know how to reduce the example further

Maybe this sounds silly, but perhaps it’s the order of the boolean operation? Try slicing to have a solid piece first, then subtract out to make it hollow last. I’m going to go with that because I know Blender’s booleans seem to invert some things depending on which way the normals are facing, and a hollow object has surfaces with inward facing normals.

Thats an idea; I think with boolean operations nothing sounds sillly.
I managed to get one wing sliced in several pieces by trying this and that way until it worked out. But changing the dimension of the wing a little bit, I need to start fiddling aroudn anew; this is not practical for all day use… so I need to find a ‘stable’ approach.
Keep everything solid as long as possible seems sensibel, thanks for the hint!
Alex

Here is a simplified file including the wing and a cube.
Try to boolean subtract the wing from the cube.
Neither carve nor bmesh works. debugCaseMould.blend (4.1 MB)

with bmesh, the rendered result shows the cube (screenshot1); the mesh (screenshot2) look weird.
With carve, the result lokks like the wing (screenshot3)

I found the source of the problem: the wing’s outer edge was not closed -> it was only a hull, open at the wingtips. This leads to trying to subtract an infinitesimal hull from the cube which is obviously not sensible.

With a completely closed hull, the boolean difference works as expected: