open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I am facing some problems to run simulations on parallel mode when I use multiple sediment class (3 classes) and a rigid bed, that in my case is a concrete channel, witch means that my ZF=ZR.

For only one processor it works well, but in parallel it simply does not work.

In parallel, if a run a simulation without computing a rigid bed (no fortran file for rigid beds, noerod.f for example) it works well and the 3 sediment classes presented on the results.

But to compute the rigid bed, first, I faced a problem to use defined ZR values in the noerod routine, that didn't splits the mesh and simply didn't start the parallel simulation. I fixed this issue by two different ways.
1 - Editing the default from ZR=ZF-100 to ZR=ZF
2 - Calling a Selafin file as a rigid bed (some programming that a guy from my lab wrote)

For both cases it fixes the way to set the rigid bed in parallel mode, and I have no erosion as it should be, but for both cases just the sediment class 1 was computed, and I have zero values for class 2 and 3.

Additionally, I tried to make ZR closer to ZF (for my case a small difference as a 0.001 m should be aceptable), but it also gave me just the class 1.
But, the weird thing is that When I set 0.1 meters or greater difference between ZR and ZF, it works as it should, the 3 classes of sediment and just the erodable layer being eroded.
It seems that by using a small layer of erodable material everything works fine, unfortunately such difference affects my results.


Does anyone have already faced similar problem and figured out how to deal with it?

Cheers

hey guys,


I've not solved the problem but I found a way to skip it.

Actually, I was using the "equilibrium inflow concentration = YES".
Apparently, all these factors combined cause some problem in this routine and give me just the sediment Class 1 ans zero for all other classes.

When I impose the concentration at boundaries it works.



Cheers,

Would you like to share your multiple study cases ?

I met the problemes when define the nonuniform sediment in the layer ?