axial load ratio in RC members

[yalham]

I have modeled few tests from PEER Column database.
As axial load ratio is less than about 1% (or about it), there is good agreement between test and numerical data using fiber models (steel02 and concrete02 models), however increasing the load ratio to 2% or more, there is considerable difference. Is there anyone has note or any experience. It is true even for tests conducted by the same team for the same specimens, in which only axial load change.

thanks in advance

[vesna]

It all depends on the way you define parameters of your materials. Try to calibrate parameters of your material models by looking at different columns with different axial load ratio.