open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi All

I would like to know if TELEMAC uses depth dependent calculations for roughness, or is this possible while simulating sediment transport in GAIA does any specific bedload formulae account for depth dependent calculations for roughness.


Thanks
Mo

Hi Mo,

if you use Nikuradse friction law, the Nikuradse value is not depth dependent. Converting Nikuradse value to Strickler, Manning etc. you would get depth-dependent values.

In Gaia you have the possibility to calculate the roughness from the skin and ripple/dune friction and give it back to telemac.
COMPUTE BED ROUGHNESS AT SEDIMENT SCALE = YES
BED ROUGHNESS PREDICTOR OPTION = 1, 2 or 3 (see Gaia Manual)
The bed roughness predictor is using Nikuradse friction law.

Best regards,
Rebekka

Hi Rebekka

I didn't get this part; (Converting Nikuradse value to Strickler, Manning etc. you would get depth-dependent values).
So you mean every formulas (except Nikuradse's) accounts for the depth, meaning in larger discharge/head it will reduce the friction coefficients and in lower discharge/head it will increase the friction coefficient, or will it consider the given-for example- manning's coefficient as for the bottom.



thanks
Mo

Hi Mo,

the Nikuradse roughness can be chosen independently of the water depth. So you can choose the same value for low water or high water. The value describes directly the roughness, has a physical unit (m) and can be derived from grain sizes (typically 3*d50 for skin friction).
For Strickler or Manning you have to adapt the value for different water depths.
Therefore, I prefer using Nikuradse. My experience is if you know the geometry precisely (e.g. lab model) and you have no form roughness (no dunes) 3*d50 will give a very good guess and you can predict the water levels (low water, mean water, high water) very well with the same Nikuradse roughness coefficient.

Best regards,
Rebekka

Thanks Rebekka, that is well explained.

Hi Rebekka

I need please your opinion on my post number #43790; I will also copy it here:
(I want to know if the number of the inflow and outflow boundary nodes are supposed to be equal or relatively close, since I have mesh with 10 points in the inflow zone and 22 nodes in the outflow . Is equalizing the number of nodes a must (for stability and mass balance reasons)?)

Danke

Hi Mo,

the number of nodes for liquid boundaries are not connected. If you have trouble concerning mass balance or stability you should use uniform triangles with 60° angles (like a sawtooth). Usually you would choose a similar resolution for the inlet and outlet but it is not necessary as long as your coarser resolution is still fine enough for physics and numerics.
If you still have problems you could use a small part rigid bed at the inlet or/and a given distribution of the velocity (can be set at the cli-file) or you can give your sediment input with nestor instead of a liquid boundary for sediments...
But this is all related to your case and the problems you have.

Maybe one of the ideas can help you.
Best regards,
Rebekka