[atul1018]

Hello All

I want to simulate the flume experiment (open channel flow), where water flows along with fine sedimenst from inlet and infiltrates into the gravel bed.

To achive what I want, I am using CFDEM (OpenFOAM coupled with liggghts). First pure DEM simulation is performed using liggghts to prepare the gravel bed and then small particles are injected from inlet along with water. The simulation runs well and i can see particles infiltrating into the gravel bed. The only problem is that, CFDEM couples pisofoam with liggghts, that means I can't define a free surface (top boundary) in it.

I have maintained the same water elevation and discharge as in flume experiment, the inlet velocity is calculated accordingly. but this set up doesnt replicate the open channel flow because the top boundary is treated as wall and no slip velocity bc is applied there.

I wanted to ask if there is any way to perform open channel flow simulations using solver pisofoam (as it is coupled with liggghts). If not, is it justified to simulate flow as closed boundary instead of open surface when my focus is to model infiltration of small particles in gravel bed?

Thanks and looking farward to hear from your experiences

Atul

[Anup Singh]

Hello, Atul assuming you wish to reflect open boundary: you can elevate the level of your channel and define the top boundary as a slip boundary. In this way, the top boundary won't influence your flow. And if you want to you can try modifying your solver for tracking a free interface.
Also, it would be better for others to understand if you can attach a schematic of your problem as for me it's a bit hazy to visualize the problem properly.
Hope it helps

[atul1018]

assuming you wish to reflect open boundary: you can elevate the level of your channel and define the top boundary as a slip boundary. In this way, the top boundary won't influence your flow.

Yes, you got it right. I want my top boundary to be a free surface boundary. Your suggestion of slip bc sounds nice but I don'r teally get why shpuld i elevate the level of my channel. I need to replicate the xperiment in which constant water depth (14cm) and water discharge of 0.085m3/s are maintained (i provode inlet velocity v=Q/A) if i increase the height of my channel, it wont be no longer actual height, thus the inlet velocity will decrease. Additionally is it a general practice to use slip bc to replicate free surface in open channel flow using single phase solvers? what do you think?

And if you want to you can try modifying your solver for tracking a free interface.

I don't want to modify the solver to track the interphase, as it will become the multiplhase problem (solid, water, air) and frankly it will be too complicated for me.

Also, it would be better for others to understand if you can attach a schematic of your problem as for me it's a bit hazy to visualize the problem properly.

see the attchement, where the bed is first filled with large particles (pure DEM) the fine particles are injected along with water from inlet (CFD coupled with DEM). The inlet channel height is 14 cm (to get the water depth of 14cm), the bed height is 10cm. you can also see, small particles get infiltrated into the bed.

The problem/limitation of current setup is that it is a closed boundary problem opposite to open channel flume experiment which i am trying to model.

looking farward to hear from you.

Best Regards
Atul

[ali akbar75]

hi dear frieds, I wanna model a free surface vortex in a cylinder which has 4 tangential inlets to the cylnder in order to motivate and ease vortex formation that should discharge through a vertical power plant intake as seen in the attached picture. know my question is why i have velocity on atmosphere patch but there is no velocity in the water phase that has almost filled the tank...



i would be pleased if any dear friend could help me model this phenomenon