Validation discrepancy

 

 

I’ve been working on the FE validation for the flexural buckling of press-braked stainless steel Z-section columns, but I’m noticing that the error increases as the column length gets longer (attached a table for reference).
I used linear buckling analysis for initial imperfections and included those in the nonlinear Riks analysis. I also tried Ramberg-Osgood material model.
Any thoughts on what might be causing this issue?