Yes, I used the HAZ-values for approximating the Ramberg-Osgood relation.
Mmmh, the approximation is more or less good.
The question is, is it smart to reduce the yield for the complete body?
For steel there is no problem, the base material allways yields first,
I can neglect the weldings, so I can use the bilinear stress-strain relation for the complete part.
But for alu? Here it is the opposite. Can I use the reduced yield for the complete body?
After your questions I am no longer sure.
Or you define 2 nonlinear materials for ccx. If possible, I have never done this.
Here is the Smath-file for self-investigating, download Smath here (free):
https://en.smath.com/view/SMathStudio/summary
Rename file extension to "sm":