We have updated to FLAC 9.0 and would like to confirm the following two questions:
Is UDM models (i.e. UBCSAND or PM4SAND) available in FLAC2d 9.0?
If grid configured for water flow? Can I use zone water set to set up initial pore pressure? It seems I could not get this to work. I used “model configure fluid” and “model fluid active off” for the mechanical initial condition and tried to use “geometry import”, “zone water set” and “zone initialize fluid-density” to set up the water pressure. “model fluid active on” and “model fluid mechanical off” are also used.
From manual:
"An initial gridpoint pore-pressure distribution is assigned the same way for the model configure fluid mode as for the non-model configure fluid mode (i.e., either with the zone gridpoint initialize pore-pressure command or the zone water set or zone water plane command). "
UBCSand and PM4Sand models are currently under development/update for FLAC2D 9. As these are UDM models and they are developed by the outside parties, we do not have control about exact timing of their availability. However, we expect them to become available within few months.
Pore pressure can be easily initialized in FLAC2D grid.
If you run mechanical analysis only and want to include pore pressure to effective stress calculations (model is not configured for fluid analysis), use command zone water density, which alongside with gravity command initializes pore-pressure (so do not forget to set gravity first). Here’s a small example: model new model large-strain off zone create quad size 10 10 zone cmodel assign elastic zone property bulk 1e9 shear 5e8 density 2000 model gravity 10 zone water density 1e3 zone water plane origin 5 8 normal 0 1 ;;alternatively, set water plane via geometry logic or import a geometry/dxf ;geometry set 'water_surface' ;geometry edge create by-position (0, 8) (10, 8) ;zone water set 'water_surface'
If you want to run full fluid analysis, use the same commands as above to initialize pore pressure or use command zone gridpoint initialize pore-pressure. I think you mixed it up with zone initialize fluid-density, which only sets the mass density of the fluid but does not initialize pore pressure.
If using zone gridpoint initialize pore-pressure command, make sure that the distribution is consistent with gravity, if it is specified (see second from the bottom paragraph here)
Thanks for your reply. I also found that if I use zone water density then I could have pore pressure, but didn’t realise it is because I removed [zone initialize fluid-density].
also, I am doing a coupled flow/mechanical calculation. The time step is very small. I believe I need to change the mesh size if I don’t change other parameters? any other recommendations? I could reduce the water modulus, but I need to consolidate it according to real-time, so I don’t want to reduce the water modulus.
just wondering if I set up the initial water pressure with zone water density and zone water set, then can I use “model configure fluid” and change the flow boundary condition, for example, with a raised water level, can I use “zone gridpoint initialize pore-pressure 0 range …” and " zone gridpoint fix pore-pressure range …“. It seems the previously used water table by” zone water set" is still in there. Can I remove it and use the now water pressure calculated by flow?
If you increase mesh size and especially make zones with higher geometric quality, the timestep should somewhat increase, but of course it also depends on fluid parameters. You can also use implicit solvers if you operate in fully saturated mode (implicit solver allow increasing timestep quite a bit, but you have to manually specify it).
@apyatigorets thanks a lot for your comments. I previously used a similar mesh size in FLAC 8.1, and it is quite fast, with stable explicit flow timestep in the order of 10^5 seconds (with the command "solve uncoupled phreatic age ??? nsub 20), however, now in FLAC2D9, the timestep is much smaller (in the order of 10^-3 second), it is now not possible to calculate it. is there a way to find out which mesh controls the timestep, I tried but couldn’t find any mesh size that is very small.
my problem is unsaturated/saturated (“solve uncoupled phreatic age ??? nsub 20” in fLAC8.1 before), so I am not possible to use the implicit? I tried it, but it is not stable.
@song, the explicit timestep difference between FLAC8.1 and FLAC2D 9 looks to be unreasonably large, I think something might be wrong with model set up. It is a bit hard to say what it is until we see a grid and datafiles. Could you please send a support request to flac2dsupport@itascacg.com? Please attach your original model and FLAC2D model (at least those datafiles needed to reproduce the issue).