Learning Forum

[Maryam Tabbakhh]

Hi, 

I am trying to use the LSDYNA to model the nonlinear soil using DRM method.

As shown in the attached figure, I have a DRM layer of solid elements (the red block) in which I have acceleration time history of each node. I wanted to use the interface ssi and load seismic ssi in order to calculate the effective seismic forces and run the analysis. 

 

My understanding is that: 

1-I need to define two interface ssi at the interior and exterior of the DRM layer (the lines between interior soil and DRM layer and DRM and exterior soil). 

2-Apply acceleration time history of each node while assigning the interface ssi (in the card) it belongs to with load seismic ssi node. Because in load-seismic-ssi I need an interface SSI and the coordination of each node and accleration time history. So I need to define the interface ssi for each node.

 

Do you think is that a righ approach? 

 

2-my second question is about the Rayleigh damping. In my original soil model, The elastic soil inside the domain has some frequency independent damping (Rayleigh) but the DRM layer and soil outside that have zero damping. How can I incorporate that in the model? Is is possible to some parts have damping and some part doesn't have damping? How about for frequency-independent-damping (Rayleigh)? 

 

Thanks 

[Ushnish Basu]

There should be only one INTERFACE_SSI for each physical soil-structure interface. Multiple LOAD_SEISMIC_SSI cards can be applied to the same interface. Please see the Remarks under the latter card in the Keyword manual.

LS-DYNA implements a different form of Bielak's method, where the free-field accelerations and the free-field forces at the interface are used, instead of specifying the motions at two layers of nodes. Please see the theory and implementation details here:

https://lsdyna.ansys.com/esi-theory/

https://lsdyna.ansys.com/esi-implementation/

For reference, see the 1980s papers by Bielak:

https://lsdyna.ansys.com/esi-references/

I believe you have already seen the examples.

[Ushnish Basu]

Further, Rayleigh damping is not frequency-independent, for that use DAMPING_FREQUENCY_RANGE_DEFORM, which can be applied part-wise. 

 

[Maryam Tabbakhh]

Thank you very much for explanations.