Jhno
Civil/Environmental
- Nov 10, 2015
- 13
Hi,
I am trying to create a soil model to conduct a total stress analysis. I use Opensees to do so.
I would like to know if my model constraints looks okay. I have a number of convergence issue and I wonder if those are due to my model being ill defined.
The model have loop that iterate to create soil node over column and row, so it can make variable size of deposit.
I define two vector, for each of the external vertical boundaries.
I fix all those node in the X direction, but I let the Y direction free.
I then fix all the bottom node in both X and Y.
I then apply incremental displacement at node at the surface of the model (number of node to match foundation size).
I let a distance of 5B between my loaded node and the external boundary to try to keep the effect of boundary out of the bearing capacity.
I use the penalty methods to enforce the constraints and a modified Newton algorithm (BFGS) to solve the problem.
I cannot get more than 0.0008 m displacement under the foundation.... I would hope to get to 20 mm, am far from there. Also, I have to use very small incr. displacement (0.000000005) which seems very very small. It takes forever to get to 0.0008 m in such a simple problems. The same kind of analysis takes like 15-20 min MAX in FLAC (Depend on mesh size).
My mesh size are 0.05m.
Any input would be greatly appreciated ! Thanks
I am trying to create a soil model to conduct a total stress analysis. I use Opensees to do so.
I would like to know if my model constraints looks okay. I have a number of convergence issue and I wonder if those are due to my model being ill defined.
The model have loop that iterate to create soil node over column and row, so it can make variable size of deposit.
I define two vector, for each of the external vertical boundaries.
I fix all those node in the X direction, but I let the Y direction free.
I then fix all the bottom node in both X and Y.
I then apply incremental displacement at node at the surface of the model (number of node to match foundation size).
I let a distance of 5B between my loaded node and the external boundary to try to keep the effect of boundary out of the bearing capacity.
I use the penalty methods to enforce the constraints and a modified Newton algorithm (BFGS) to solve the problem.
I cannot get more than 0.0008 m displacement under the foundation.... I would hope to get to 20 mm, am far from there. Also, I have to use very small incr. displacement (0.000000005) which seems very very small. It takes forever to get to 0.0008 m in such a simple problems. The same kind of analysis takes like 15-20 min MAX in FLAC (Depend on mesh size).
My mesh size are 0.05m.
Any input would be greatly appreciated ! Thanks