Symmetrical hysteretic curves

[zhwang]

To whom it may concern,

For my current bare steel frame model, I set everything symmetrically, including symmetrical section, node position, history load, and restraints. However, the hysteretic curves of the model are not 100% symmetrical.
The program is attached: https://www.dropbox.com/s/h08d8gzmbht3lek/no.spf?dl=0
The hysteretic curves:
https://www.dropbox.com/s/n3wg2nidrwclw ... 3.jpg?dl=0

Could anyone suggest me what is the problem? Is the asymmetry caused by my mistake in the model? Or caused by the convergence of the seismostruct software itself? Or any other reasons?

Thanks very much.

[ruipinho]

Hi zhwang,

I have not checked your model, hence mine are blind suggestions (which also take for good your assurance on everything being fully symmetric):

- tighten your convergence criteria a bit

- introduce out of plane restraints (I am assuming that your analysis is 2D)

Regards,

Rui

[seismosoft]

You need to consider that this is non-linear analysis. For example during the first cycle in the positive direction the structure is undamaged, whereas in the negative direction the model is already damaged from the previous loading (in the positive side). For this reason, one should expect exectly the same behaviour.
Can this be the reason for your observations?

Seismosoft Support

[zhwang]

Dear Seismosoft Support,
Do you mean it is the damage of the structure caused the asymmetry of hysteretic curves in positive and negative direction? Can my attached hysteretic curves picture show it has same behaviour in positive and negative direction?

Thank you very much.