User Forum of Software BASEMENT

LydiaSeitz

User

Registered: 2014-09-08

Posts: 18

Instabilities in the morphodynamic model

Dear Basement Team,

I can only run my morphodynamical model stable when I set the bedload factor = 0,2. In this case I don´t get satisfying results (as expected).
I am sure there is some major mistake from my site in the model, but so far I can´t find it.

The model needs to have a
bed load control volume thickness of 2 m in order to run stable. If it is lower the following warning message appears and the model crushes:

"Bedload control volume mixtures differ more than 10 percent, is somofbeta = 1.0089, thickness = 0.15"

I think control volume thickness of 2 m is quite high. For the 2D model I usually had a control volume thickness of 10-20 cm.

As soon as I try to vary some other parameters such as theta crit or the critical shear stress calibration the model gives me the following error message:

A fatal error occurred:
internal error in fluxCal().edge->tmpVar_.fluxL_[ gg ] >= 500 because its 500.432 for edge 53 and gg = 3
in file BASEchain\BCInternalFlux.cpp on line 129
Version: 2.4R1692, compiled at Sep 10 2014, 11:01:08

We are sorry but BASEMENT must be terminated now.

Could you please give me any hints which other parameter I could vary in order to increase the stability of the model?

I would be really happy to get some hints from you.

Cheers,
Lydia

Last edited by LydiaSeitz (2015-03-04 12:26:13)

Matthias

Advanced_User

Registered: 2014-12-16

Posts: 29

Re: Instabilities in the morphodynamic model

Dear Lydia
In order to get rid of the "Bedload Control Volume Mixtures.." Error-Messages you can set the log-level to 0 (in your Basement GUI Main page where you see the console, go to Options - Log-Level and set it to 0). This blocks the output of such warning-messages.

You are correct, the control volume thickness should not be that big. It should be in a range of around 1.5 - 3x the d90 of your grain size distribution. (Correct me if I'm wrong!)

Regarding the fluxCal error: are you working with a SedRestart.dat file at the beginning of your simulation? How many layers have you defined?

Hope I could help you!

Best regards
Matthias

LydiaSeitz

User

Registered: 2014-09-08

Posts: 18

Re: Instabilities in the morphodynamic model

Dear Matthias,

thanks for the fast reply!
Setting to the Log-Level to 0 help to get rid of the Warning-Message.

Regarding the fluxCal error: I do start from hydraulic steady conditions with a restart file. For the morphology I start from initial beginning. Two layers are defined. I already tried to use only one layer, but the same problem occurred.

I tried to vary the upwind coefficient, the type for the inflow boundary condition (sediment discharge, transport capacity, IOUp), theta_crit without having any success.

The hydrograph was smoothed in order to increase stability.   

Do you have any idea what else I could check?

Best wishes
Lydia

Matthias

Advanced_User

Registered: 2014-12-16

Posts: 29

Re: Instabilities in the morphodynamic model

Hi Lydia

I need some more information about your model, maybe then I can help you

How does your longitudinal profile look like? Do you have an "oscillating" erosion/deposition/erosion/deposition etc. pattern?
How many grain classes have you defined?
How long does the model run until you get the fluxcal error?
Are you observing the morphology with baseviz? If yes, do the morphological changes appear very shortly (direct after the start of the model) or do they happen steadily?

You might try to increase the grain sizes of each class by say 10 - 20% to get a more stable run.

LydiaSeitz

User

Registered: 2014-09-08

Posts: 18

Re: Instabilities in the morphodynamic model

Dear Matthias,

I was just about to post that I found the problem.
It was (as expected) my own mistake. I found a transposed digit of the orientation angle in the geometry file. So the end of the cross section crossed another cross section. I already checked the cross sections a lot of times if there is any mistake but I never payed attention to the position of the cross section after modifying them (lesson learned!!!!) It is strange that the hydraulic part was running with this error without problems but now I am quite happy that it is working so far.

Now it is running for almost 10 years, usually it crashed after 5 years of modelling. So I will keep you updated if it finish the run. 

But thanks a lot for all your effort!

Best wishes
Lydia

Matthias

Advanced_User

Registered: 2014-12-16

Posts: 29

Re: Instabilities in the morphodynamic model

Hi Lydia
I am glad that you found your error and that your model is running now (completely?)!

However I was told that the orientation angle is only used for the interpolation of cross sections and for the presentation and later export to a GIS software. I thought the only parameter regarding the geometry between the cross sections was the given distance.

Could any of the developers comment on this? Is it i.e. morphologically impractical if there are sharp bends in the model?

PS OT: How can I upload images to this forum?

Lukas Vonwiller

Developer

Registered: 2014-09-04

Posts: 130

Re: Instabilities in the morphodynamic model

Dear Lydia, dear Matthias

The orientation angle determines, in combination with the global coordiantes, how the cross section is located in space.
However, this is used just for visualization and should not affect your simulations.

Thanks Matthias for your interactive participation on this forum, we really appreciate your effort!
It is in fact the goal of this forum that experienced users help other users to improve their models.

Cheers, Lukas

P.S. Right now I think there is no possibility to upload images on the forum, but we will check the possibility of uploading images to the forum in the future