Don't get me wrong, my project is not cassette flooring system. a project modeled with a wall up to 1% of the building floor area. I don't know how you added an algorithm, but there are serious differences in the relative storey drifts. I think you can analyze and experience this in a model with sheared beams without waiting for us to send you a project, but if not, we can share the model via the portal[ /quote] It doesn't have to be cassettes or ribs. If the building has simple beams or gallery floors with very large spaces, it will affect those types of structures as well. If you share the project like our other colleague in order not to speak afaki, I will comment. I would say this without looking at the project. If it is a project that you are solving with a semi-rigid diaphragm model, the results of v10.59 are valid in any case. FEA (Semi-rigid diaphragm system) of floors, curtains and docket of the whole structure is created automatically by the program. For this creation process in each version, we make use of the sample projects that come to us in any structure made by the users. Users can create systems for this creation that we could never have foreseen, and we are constantly improving and updating this mesh creation process for each project. If there are no simple beams in the system, the results do not change between v10.20 and v10.59. If there are simple beams in the system, that is, even one simple beam that does not form a frame, very small results are noticeable. As the number of non-frame beams increases, the differences begin to increase. Again, there are differences in cassette and rib systems. If there are no simple beams in your structure and this difference occurs, it is necessary to examine it separately. In this case, we expect no difference. From v10.20 to v10.59 . That's the change we're making. Also, nonlinear solution options have been added. These things have nothing to do with the nonlinner solution. Please share via portal.