User Specified Damping?

[mpnewcombe]

Hello,

I am just coming to grips with Opensees and have some questions regarding the Rayleigh Damping formulation.

Firstly, the sort of structures I am analyzing have a non-linear elastic reponse. I have found that the best way to model the inherent (or elastic) damping in other programs (RUAUMOKO) is to use User Specified Constant Damping proportional to the initial stiffness... I reduce the first mode damping to account for the change in stiffness for non-linear first mode response (see Smyrou et al, 2011, Bull of Earthquake Engineering). This matches with shake-table test results for damping. As stated by many researchers, I do not think it is appropriated that higher modes are damped more than the fundamental mode, as is unavoidable with Rayleigh Damping. Furthermore, the error created by using Rayleigh Damping is significant, for systems where there is no hysteretic damping component.

Is there a way to 'trick' the Rayleigh Damping approach in Opensees to achieve a constant damping coefficient? I thought about using the region command and specifing damping only to elements that remain elastic. This would pick up the only the mass of the structure, which is connected to these elements. But I assume the stiffness of the complete structure is considered, which is non-linear, and the higher mode response would still be over-damped.

Alternatively, can local viscous damping model be used to simulate a constant damping coefficient?

I would very much like to continue using Opensees but the errors associated with the damping model may be too large. Are there any plans to introduce a user specified damping model in the future?

Best Regards,

Mike Newcombe