Learning Forum

[Maryam Tabbakhh]

Dear all,

 

I have an important question about this topic. I want to apply the earthquake in z direction to the tank fill of water with the box floating on it like top figure. I add one more picture to show it better here. I need to first gravity initialize the box and after it reached equilibrium, I am going to apply earthquake just in Z direction. It is basically two step. Step one, from 0 to 5 second, I just apply graviry through a inclined curve and then apply earquake at 5 second ( the acceleration that I provided starts after 5 second). My problem is that I need the base of the tank to be fixed in z direction in order to gravity initiation, but for the next step that I want to apply acceleration in z direction, I need that DOF in z direction to be free. If I fix DOF in z direction, my tank doesn't more in z direction when earthquake applies but the first step (gravity initialization) reaches equilibrium and the problem with having the DOF in z direction free, it doesn't converge in the gravity initialization and box starts to rotate. Do you have any solution for me? I did the same thing for Y direction and the results were successful because for applying earthquake in y direction, I don't need to free the DOF in z direction free and I could get reasonable response. Appreciate any suggestions.

[Ushnish Basu]

You can use BOUNDARY_PRESCRIBED_MOTION with zero motion during the gravity initialization (set OFFA in DEFINE_CURVE to generate the initial zero portion automatically)

Alternatively, use BOUNDARY_SPC_BIRTH_DEATH to set a death time on the SPC

[Maryam Tabbakhh]

Thank you so much for your response. I'll try it now. 

[Maryam Tabbakhh]

Dear Ushnish,

I tried SPC with death/birth. I put an spc for the nodes at the bottom (except corners) of the tank during gravity initialization (0-5) then I removed that and applied earthquake to the bottom corner nodes. It got unstable and the run didn't go thorugh (by decreasing time step to -20). I tried to apply the earthquake to the same node set but it didn't work. It gives me an error that this node has spc birth death  *** Error 10902 (KEY+902)
     Node ID 81 has spc's applied with different
     death time with *BOUNDARY_SPC cards 0 and 0.

Apparently I can't apply prescribed motion set and spc birth/death to the same node set. So I devide the bottom nodes to two groups (one I applied x,y,z DOE via birth/death spc (which will finish after gravity initialization) and the other one x,y which is availale throughout the whole analysis. what is your recommendation? Thanks 

*BOUNDARY_PRESCRIBED_MOTION_SET_ID
         1earthquake z
         3         3         1       222       0.5         01.00000E28       0.0

 

*BOUNDARY_SPC_SET_BIRTH_DEATH
         3         0         1         1         1         1         1         1
       0.0       0.5

[Ushnish Basu]

Please see the first suggestion I had above

[Maryam Tabbakhh]

I did that as well but the problem is that the speciment moves and doesn't reach the equalibrium like when there is a real zero boundary condition. I applied the same loading in y direction (the boundary was fixed in z direction during the analysis) so it reaches equilibrium and then I applied the acceleration in horizontal direction. Than'ts why the first solution doesn't work. But by adding the spc birth/death the problem of gravity initialization is solved but it gets unstable after that.