Stiffness proportional damping does not seem to work

[divamva]

I have coded a simple SDOF system and subjected it to an earthquake
record and separately to a sine loading. All work just fine when using mass proportional damping. Unfortunately, if I try to use stiffness proportional damping (using any of the betaK, betaKinit, betaKcomm
or even all three in the command "rayleigh") the system behaves as totally undamped. I have discovered that the reason is my using a
zerolength element with "elastic material". It seems that this material
(and I am not sure about the other uniaxial materials as well) will not
generate damping (unless maybe the damping tangent "eta" is specifically provided). Now that I have realized that, my question is whether this
is a bug or a feature. I can easily see how someone with an MDOF system
with lots of such springs will not get from Rayleigh damping what he though he should get.

What do you guys think?

[fmk]

truss and zeroLength elements determine their own damping .. all other elements use the default provided by the Element classial will work .. that is why providing eta for the material will work .. it is something that obviously needs to be clearly documented and i apologise that it is not

[Guest]

Thanks for clarifying that, I didn't know this applied to trusses as well. Can you also tell me what the "damping eta" is? I need to know how to
set it so that I can get what I want.

Thanks for the help

[astrella]

I have a 3D steel frame. All beams are beamWithHinges, all columns are nonlinearBeamColumns. I am trying to implement Rayleigh damping. When I use just the mass proportional term, good results. When I add stiffness proportional terms, I get messages to the effect of:

incompatible member forces and deformations in the element: __

The elements are always column elements. Is there something between the distributed plasticity and the stiffness proportional damping that I am missing? Thanks

[gilgal77]

I tried several times to find out what eta means.

My conclusion is k=(1+i* eta)K. That is the coefficient of histeric damping.

But I'm not sure.