[Boone]

Thank you very much for your previous answer. I found a less aesthetic but sufficient workaround by computing the wallDistance using checkMesh -writeAllFields and the wallDist keyword in fvSchemes.

I am trying to simulate a case of periodic hills with a given sediment concentration throughout the domain, in order to analyze sediment distribution in different regions as well as erosion and deposition fluxes at the bed.

So far, I have been computing div(phi.a) to quantify erosion/deposition at the surface and I am not sure that this is the proper way to do this. I noticed that the code includes dedicated quantities called Erosion and Deposition, which are computed in createFavreAveraging.H.

When I run the simulation with FavreAveraging activated, I obtain fields like betaSqrUbMean, UbMeanF, etc., but I do not get Erosion or Deposition. Upon inspecting calcFavreAverage.H, I saw that the lines Erosion.write(); and Deposition.write(); are commented out.

Since these quantities are of primary interest in sedimentation studies, I don’t understand why they are not computed by default. I may have missed something. Could someone explain how to save the output related to deposition and erosion fluxes so that I can visualize them in ParaView?

Thanks a lot!

[eduard.puig.montella]

In the current code, “Erosion” and “Deposition” are only declared but they are never assigned anywhere in the loop so they remain zero. We might add this feature in the future, but at this point you can't quantify erosion/deposition using these variables.

For a resolved mobile granular bed, a better way to quantify erosion/deposition is either:
a) track bed/interface position using an isosurface of alpha (for instance alpha=0.2) and compute Δbed elevation over time (you can do that with paraview, fluidfoam or adapting some of the postprocessing scripts), or
b) compute solids flux across a reference surface (alpha*Ua·n (also doable with fluidfoam and paraview)) near the bed. This should be more appropriate than using div(phi.a), which measures local accumulation/depletion of sediment due to transport in any direction.

[zjy]

Originally Posted by eduard.puig.montella

I completely agree with zjy. In sedFoam, the near-bed momentum transfer is governed by a multiphase formulation and a resolved mobile granular layer (yplus assumes a single-phase wall-bounded turbulence model so it needs to be modified). Plus, yplus (and actually most of the functionObjects of openfoam) are not implemented in sedfoam. That said, this is not a fundamental limitation of sedFoam. It is simply that the yPlus functionObject is written for single-phase wall-bounded turbulence models and is therefore not directly applicable/available in the sedFoam multiphase context. If needed, it can be implemented as a custom functionObject, in the same way as we did with the forcesSed module already provided in sedFoam. If you really need to implement the yplus as it is in openfoam and you have problems with the implementation in sedfoam we can help you out.

However, the resulting y+ would only be meaningful for a pure fluid wall layer. If you are interested in what happens at the granular-fluid interface it is usually more relevant to compute the stress contributions in the cells/regions of interest.

Thanks a lot for the clarification，Quick question: do you know if there are any plans to add a built-in way (in future sedFoam versions) to compute bed shear stress at the water–sediment interface? For sediment transport applications this is a pretty key function

[julienC]

Originally Posted by zjy

Thanks a lot for the clarification，Quick question: do you know if there are any plans to add a built-in way (in future sedFoam versions) to compute bed shear stress at the water–sediment interface? For sediment transport applications this is a pretty key function

Dear zjy,


Thanks for your question.We have recently developped a methodology and postprocessing tools to do that and much more like computing bedload and suspended load fluxes as well as reconstructing Exner equation terms. This work has been published recently in IJMF and it can be found here: https://doi.org/10.1016/j.ijmultiphaseflow.2025.105457
or alternatively here for an open-access version : https://cnrs.hal.science/IGE/hal-05299771v1


As usual, the potsprocessing tools are also available freely online : https://github.com/SedFoam/exnerDiagnosis.git


By the way, I am curious to learn more about your research. Don't hesitate to give us some feedback about your work.

Cheers,
Julien