FLAC3D Fastflow Troubleshooting

When attempting to solve semi-coupled or coupled fluid-mechanical problems in FLAC3D for tunnels, I frequently attempt to use the fastflow option to improve solution times. I find that the default fastflow parameters result in an error:

*** An accumulated value has become greater than the maximum allowed.

Upon inspection, the model frequently shows unrealistic pore pressures or displacements. This also frequently occurs when the fastflow timestep rapidly increases to unrealistic values. It is as if the fastflow algorithm is not sensing out of control behavior and reducing the timestep appropriately.

I try to prevent the error by reducing the fastflow-relaxation value, reducing the maximum automatic timestep or improving my mesh quality, but in some scenarios this does not fix the issue.

The fastflow examples in the manual work well, but when I try to apply them to tunnel or excavation style problems I fail to be able to apply fastflow.

I am interested in hearing from other users how they have applied fastflow to their models and what watch-its you have picked up through experience with fastflow.

Hi @pjkottke, Is your model fully saturated? I understand the fastflow applies to saturated media. Most of my cases hae been partially saturated and had to use the basic flow scheme.

Gian

Good question, saturation stays at 1 throughout my model, though pore pressure does drop to zero above the water table. I think this satisfies the fastflow requirements, but am not certain.

It 's hard for us to have an investigation without the datafile. You can send the datafile to flac3dsupport@itascacg.com.