OK - to verify the model :


Make a beam model also and compare the first global mode (first image in your post - Load Multiplier =18.689) to the shell model - then you will know that the model is OK (does not mean reality is like this - only validation can tell)


Realised you have 3D parts - at most use shells or beams (for global modes only - no web buckling). 


Also have in mind that the load multiplier multiplies the total load (so gravity too - which is not completely correct - you can take it of)