ABAQUS - the influence of changing kinematic and cycling hardening parameters

Hello,

I have recently done a simple hourglass model to check the influence of hardening parameters in Abaqus. The reason was the effects of the parameters' change in bigger models were a bit surprising, therefore I needed to investigate the reason using some simpler model. Please see file 1.

After performing tension/compression cycles I got results as shown in files 2, 3, 4 and 5. Please note that only the hardening parameter in the description changes, the other three stays the same for each series.

There is no problem with figures 2 and 3, the C_1 and gamma changes the behaviour as they should.

However, I have some problem with 4 and 5. Q_inf is 'maximum increase of elastic range' which clearly should only change the f_y of the material. And do nothing during the first cycle, as there is no back-stress. Point is, it changes the slope of the curve of the first cycle as well. Seems like Q_inf not only changes the f_y but somehow scales up (or down) the position of neighbour points of the yielding point (all the curves in plastic stage presented on page 4 seems to be the same but in a different scale).

Similar thing with the diagram 5. b should only change the 'rate at which f_y change with plastic strain', so also no influence on the curve itself and no influence during first loading. However, it changes the slope similarly to the effect of change of Q_inf. I.e. behaviour of the model showed in diagram 'b=0.5' for epsilon=<0.15;0.9> is nearly the same curve as in diagram 'b=10' for epsilon=<0.15;2.8>.   

Is it as it should be? Or is it the model or the programme problem? If you have any thoughts - please share them.

Regards,

Fujak