User Forum of Software BASEMENT

FabKS

User

Registered: 2023-11-28

Posts: 2

sediment transport balancing in crosssections and boundaries

Hi everyone,
I am currently working with Basement Version v2.8.2 on 2D-HN-simulations of river sections in which I am modeling bedload transport over a fixed bed. The following questions came up, and I hope that you can help me with them or, if you have had similar experiences, perhaps share your solution or workaround:

1. I feed my sediment into the model via an "external source". To evaluate the bedload transport, I have defined various cross-sections in the model area that cover the entire width of the channel. However, in many of my simulations, the amount of sediment transported via these cross-sections permanently exceeds the amount of sediment added and is sometimes even more than double the amount added. I inserted the cross-sections into the model both, without a breakline ("zig-zag") and with a breakline (straight course) and evaluated the parameter "QSed" as "stringdef_history".

2. the balancing at the downstream boundary of the model (QSed) also permanently provides a larger value than the sediment addition at the "External Source". This happens both for the evaluation as "boundary_history" and as "stringdef_history", whereby the values differ slightly from one another.

Both also occur if I replace the "External Source" with a "Sediment Discharge" boundary condition at the model inlet.

Below is some more information about my model, which I hope will help you:
I use the Engelund-Hansen approach for modeling bedload. The problem occurs for different grain diameters and porosities. I checked the amount of sediment added by stopping the sediment discharge from my model and comparing the deposited sediment volume with the integrated sediment input. The integrated input corresponds almost exactly to the deposited volume (taking porosity into account). So the input should work correctly.
To simulate the solid bed, I defined the Bed Material block (Grain Class, Mixture, Soil Assignment, Soil_Def), but did not add a moving layer to Soil Def. So the bed should be assumed to be solid based on the given mesh. There is also no negative height change, i.e. no erosion in the model. So the excess sediment in the cross-sections cannot result from erosion from the model area.
The functions "Direction - curvature effect dynamic" and "Direction - lateral bed sloap" as well as "Gravitational Transport" are activated.
Unfortunately, I cannot share the model itself externally. However, if necessary, I can try to reproduce the problem for example in a simple channel.

I'm looking forward to your ideas.

Many thanks and best regards
Fabian