Hi Niki,
Looking at your parameters I would not expect significant differences in any of the models as they should all be dominated by the core, a very thin 20nm diameter disc (of thickness 0.08nm to 0.34nm depending on the existence or not of faces). Adding a 0.1nm shell (rim) shouldn't make much difference unless the core is close to contrast matched I would think? For the face thickness one might expect some oscillations at unrealistically high q that I would normally expect to be in the background.
Reading the model documentation I now see the problem(s). The main problem being our interpretation of the model.
1) The core shell bicelle model states that it is a general case of the coreShellCylinder that allows you to have different SLD on the face and the rim! I then looked at the cylinder documentation (and if you blow up the picture in the excel sheet you will see it) where the graphic clearly shows that the shell surronds the core of the cylinder: both a face and a rim but the thickness and sld is the same all around (the bicelle allows both for a different t and a differnet SLD between the rim and face). hence the graphic shows that Ltotal= 2H +2t where L(of core)=2H as also shown on the graphic.
2) I notice that you in fact are going out to 10 1/A. At these qs you are looking on the fraction of an Angstom lengths and the average SLD approximation made in all these models starts to break down as implied in my first paragraph.
3) Ignoring now whether the continuum approximation is correct at these very high qs the sandwich structure in the normal direction fo a 0.14nm coating layer followed by the 0.08nm core thicknes followed by the 0.14nm coating on the other side shows up as the oscillation you see in 1 and 3 (your model graphic is of course wrong for 3 as discussed in 1 above - it should have a 0.14nm layer on the face as well as the rim with same SLD on both). Adding just a tiny rim to the giant disk as I suggested in the beginning does not matter much unless you make the core match the solvent perfectly .. then you will see some ringing in the pattern. In fact I was surprised at how close you have to be to matched before seeing anything.
Hope this helps?
Paul