[jasonneuswanger]

This seems to be by far the most comprehensive open-source CFD system around, but I can't find any examples of a fairly commonplace CFD task: simulating the flow of a river.

I would like to be able to simulate 3D steady state flow in a 100m reach of river based on about 2000 field-surveyed points defining the bathymetry. Is OpenFOAM appropriate for that, and if so can somebody point me to resources involving a similar application? At first glance it doesn't look like the meshing format is very conducive to inputting a 2000-point survey.

I also need to track the density of invertebrates drifting in the river. They need to be able to have positive, negative, or neutral "buoyancy" (depending on behavior). I'm considering two ways to do this: tracking them as individual particles, or tracking their probability density like a plume of ink or something. Is either option possible in OpenFOAM, and if so, what components should I read up on?

Thanks in advance for any advice.

[wolle1982]

hi,

have a look on the "damn-break" tutorial in OpenFoam. this mostly fits your problem. you can mesh your river with Salome which also is OpenSource or use the snappyHexMesh Tool which would fit perferct on your problem.

[liu]

We did some 3D CFD modeling of a river.

Here are two pictures

[liu]

I tried many method to model the river bathymetry. The way used to generate the previous two plots is a little bit unusual.

First a flat bottom river mesh is generated. Then an "scour" mesh deformation process is done to move the mesh according to the bathymetry.

This mesh deformation method is used in the sediment transport model I developed for my Ph.D. thesis. The details can be found in

[1] X. Liu and M.H. García (2008). A 3D Numerical Model with Free Water Surface and Mesh Deformation for Local Sediment Scour. Journal of Waterway, Port, Coastal, and Ocean Engineering. 134(4): 203-217

[2] X. Liu. Numerical Models for Scour and Liquefaction around Object under Currents and Waves. Ph.D. thesis, University of Illinois at Urbana and Champaign, 2008

[ngj]

Hi Xiaofeng

The pics look really nice ... where in the world are we?

The meshing routine: did you complete that in one "time step", or was it more stable to achieve the final bathymetry using several updating steps? Here I am thinking about the conservation of the discretization of the bed boundary layer, and how the discretization changes under large deformations given a certain diffusivity in the mesh-moving algorithm.

Did you do morphology in this setup as well or was it restricted to a flow analysis?

Best regards,

Niels

[liu]

It's a river connecting two of the Great Lakes.

The mesh is deformed to its final status through many steps. The meshing routine is not ideal especially when you have steps or holes. And you don't have direct control on the grid space above the bed, i.e., the boundary layer. What I do is to try different diffusivity schemes and different parameters.

Maybe someone else has better idea to mesh a river with real bathymetry. I read somewhere people used the "projection" operation in Gridgen to project a flat bottom mesh into the bathymetry database. I tried but no luck. These things depend a lot on the experience of the user.

[jasonneuswanger]

Thanks Xiaofeng. I got your paper and I'll be studying it tonight.

Does anyone have any thoughts on my invertebrate "tracers"?

[fuji]

Dear all

I want to simulate the 3D steady state flow in river (about 2km) based on about 1000 field -surveyed point defining the bathymetry.
The boundary inlet is the water level.
Could you help me to do it.
send me all reference document about it please!


Thank you very much.
Fuji

[GrivalszkiP]

Originally Posted by liu

We did some 3D CFD modeling of a river.

This looks awesome!

I also trying to simulate river flow, but I have difficulties with the proper BC settings. In classic river modeling, we use discharge inlet and fixed depth outlet, but I did not find a proper way to fix the downstream depth without weird physical phenomea. How did you set your BCs?