open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I did some runs with Telemac3D and Sediment, using the Gibson consolidation model.
And while playing a bit with the variables, I ran into quite a surprising result.

The variable MAXIMUM NUMBER OF LEVELS DISCRETIZING THE BOTTOM (refering to the dico it's default value is 2, although the Sedi3D manual 2.2 of the EDF states ist 30) was 2 in my first runs, then I changed it to 5 and in a consolidation step (exactly in the second consolidation step that occured) the bottom of the model is set to -18500and something.

Accordingly also the free surface goes down to that value, and due to the boundary conditions it soon crashes.

What I don't get is why the higher number in the levels for the bottom results in this consolidation problem. Has anyone seen this before?

It might be due to some wild combination of variables, since I used a really small time step for consolidation, a value of 100 for the concentration of fresh deposits and a high critical shear stress for deposition.
On the other side I don't see why this should produce problems.
Any hints or ideas where to search for an explanation?

kind regards,
csee

ok (just to let you know), I found in the tassem.f code a supposed to be good hind:

if(NPF(IPOIN).GE.5) then it says

so my guess it is something related to
WEIGHT SUPPORTED BY SOLID SKELETON
and I will take a closer look.

does anyone know about some documentation or anything related to this?

Kind regards
csee

2weeks later:
I know why I got that huge amount of consolidation, in the condis subroutine the erodeable bottom is set to 100m (the variable is named hdeb) and so the 100m+ a tiny little bit are consolidating in one timestep causing the huge bottom change.
That value needs to be adjusted if consolidation is used.
(So not directly related to the solid skeleton but a little further down in the code.)

So I found myself a new situation to wonder about, everytime the consolidation takes place now the change in the bottomlayer becomes bigger, and so if a lot of steps are taken the change is quite big again.

Just to let you know (if you're interested)
csee

Dear Csee,
thanks a lot for your hints and interest. In next future we are planning to revisit the consolidation model implemented in Telemac-3d. We will keep you informed about improvements/changes.

Many thanks again,

Pablo

Hello again,

I'm interested if there is anyone revisiting or working in the consolidation area by now or if there were any decisions regarding consolidation.

Since at the moment I still plan to understand/use some consolidation model but am unsure about the laws used in Gibson model [(I had a look at Been and Sills but that did not help much) for Permeability and effective stress] and even more about the multilayer functionalities and have a couple of other questions.

Maybe that would be easily solved if there are some documentation parts, code comments or such beeing produced.
Or it might turn out to better change some of my plans.

kind regards
csee

Dear Csee,
Soon we will be launching the new version of the documentation of Sisyphe in which we have included a full chapter dealing with consolidation modelling.

The documentation is being verified. As soon it will be ready I'll post a message in the Forum.

All the best,

Pablo