Hi, Disla,
but your model has a first eigenfrequency of 0.08 Hz and the one in the video is 1/T1=1/0.282 s = 3.54 Hz = f1 (more realistic for a building).
The video uses what in Nastran is called point normalization, setting a value of 1 to a specific dof, however if I do that with your model (the original one posted, not the latest) I get the following values for mode 1:
phi11’ = 1.00, phi21’ = 1.85, phi31’ = 2.25, very different to the values used in the video.
Regarding the participation factor calculation, modes are mass normalized so the issue has to be anywhere else, maybe this question is related to it seems B31 beams have issues (TBC): Total effective mass differs if using density or lumped masses