Learning Forum

[Maryam Tabbakhh]

Hi,

I implemented the DRM method and could verify the results but the problem is that there is something like step like line in the curve and it's not smooth. I explicit method for analysis and damping frequency range deform. Acually I also removed the damping but results is still the same. Do you have idea how can I fix it? 

 

Thanks

[Ushnish Basu]

This looks like there's some permanent displacement in the soil, possibly due to non-linearity. In DRM/ESI, all non-linearity must be inside the soil-structure interface, as we would need to do a static analysis first. The card INTERFACE_SSI_STATIC is used to capture the static reactions in the static analysis, for later application in the transient analysis to support the static load and isolate the PML from static stress.

[Maryam Tabbakhh]

Dear Ushnish,

 

The soil is linear elastic. This is the displacement in x direction. There is PML layer around but I did some analysis without the PML and the results was the same. if you zoom in you see that the opensees result is a nice smooth curve while in LSDYNA, I could get almost the same results but the curve is not smooth and is like step-wise. I applied acceleration here. Do you believe that I need to change the INTERFACE-SSI to static in order to get the better results? There is no gravity initialization here. 

 

Thanks 

[Ushnish Basu]

There are two different issues here:

1. The non-zero static displacement is of concern, as it won't be supported by the absorbing boundary. Please make sure that your acceleration is baseline-corrected.
2. I can't see the jaggedness of the output from the picture, but you can try reducing the output timestep in LS-DYNA to get a smoother curve

[Maryam Tabbakhh]

For number 1:

1- I used the acceleration output from SW4 software and got the same results with Opensees linear elastic material drm method. So the input is correct. I see, it means that with absorbing boundary (PML), I can't have this displacement. So how the free-field analysis is done? I applied the acceleration at the boundary of inner soil and outer soil and in the next step I am going to add structure on top of the soil and probably modeling nonlinear soil.

 

2- I attached the figure here for you again: I reduced the time step to 0.0001 (the time step of analysis si 0.0028) and again got the same thing. I am going to reduce it one more time. When I compare the velocity and acceleration, I see more difference because of this jagging. Is there any otherway than increasing the time step to get rid of this?

 

Thanks 

[Ushnish Basu]

Unfortunately, it's difficult to help further in this forum as I would need to look at the input file to figure it out, but we are not supposed to download anything from here.

I would suggest that you discuss with your Ansys account manager whether you can get more detailed help directly.