open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone,
i got this error message

when I tried to run a simulation calling regional tpxo data base.
I am using the H/UV.AO_2008.out.
and my steering file is ...


and also my boundary condition file ...


1 liquid boundary and 1 solid boundary

I don't know if it is relevant but my area of interest is in 29N UTM WGS84 but fudaprepro shown an error and only has 30N... but I tried both 29N and 30N and still got the same message ...

thanks for your help

josias

sorry, my steering file and bc2 file didn't load.
here is

Hi,

Do you have a version of TELEMAC compilationg in debug mode ? This might help you directly. In fact, it's hard to see the problem if there's no .cli file with the mesh file. I think the problem comes from the mesh file.
Besides, you can try keyword MINIMUM DEPTH TO COMPUTE TIDAL VELOCITIES INITIAL CONDITIONS and MINIMUM DEPTH TO COMPUTE TIDAL VELOCITIES BOUNDARY CONDITIONS in order to have non-negative calculate water depth. Sometimes it causes errors when calculating U and V because of the use of TPXO.

Hope it helps.

hello,

thank you sorry maker and pham,

Pham: my problem is not already solved, I use V8p2r1.

I decreased my time step from 10s to 20s to 60s to 120s to 240s same result.

for Sorry maker keywords I tested it like :

MINIMUM DEPTH TO COMPUTE TIDAL VELOCITIES INITIAL CONDITIONS =0.1 also 0.5

MINIMUM DEPTH TO COMPUTE TIDAL VELOCITIES BOUNDARY CONDITIONS =0.1 also 0.5

same result

my cli file

, my case file

, and slf

file for input

maybe my mesh is too fine or else. What mesh size do I have to use?
thanks and best wishes

Hello,

I think your mesh has to be changed as, at least, San Pedro River is not well meshed: indeed, you have overconstrainted elements with 3 nodes of triangles belonging to boundaries. One of them should be at least inside your domain.
You also have a few other overconstrained elements (see e.g. at the bottom corners (South-east).

After your mesh is redone, you can change:
SCHEME FOR ADVECTION OF VELOCITIES = 1
and reduce the time step e.g. to 20 s.

Hope this helps,

Chi-Tuan

Hello,

Have you solved your issue?

If not, which release do you use? v8p2r1?

Depending on your mesh resolution, you may have to decrease your time step.

Otherwise, the 2 keywords mentionned by Sorrymaker may help if you have big velocities at the open boundaries or at the initial time step.

Hope this helps,

Chi-Tuan

hi pham, hi sorrymaker,
thanks for your comments and replies, I have reduced the default edge length of 200 m of my mesh, also I deleted the second liquid boundary of the river and it did work, I mean computed. I introduce it latter with constant flowrate, it did work. I have two other questions but I will open a different topic, thanks for all.

Hi everybody, Hi pham, hi sorrrymaker,

following a chat about K-Epsilon results in Bluekenue, in the Bluekenue Topic chat, I tried to compute this same simulation (previously discuss here TPXO WITH T3D) this time with k-epsilon model, to get appeared K-EPS results. and it did fail.

precisely it did run for duration under 6h or 7h but No later,

I changed multiple time the

SCHEME OPTION FOR ADVECTION OF K-EPSILON = 1,2,3,4

I also tried to work in K-W model, same result.

I decreased the mesh from 200m to 100m in the port around the guilty point, same result.

I change the type of liquid boundary of the ocean side I used 544 and 566, same result, even slightly different (8h of computation duration).





here the latest files I used

Hello,

In such cases, what is advised is to run the same simulation until the last iteration before crashing, but write the results of these last time steps. Visualising the results in such cases may help to understand what is wrong with your computation.

Sometimes, reducing the time step is enough.

Anyway, I have tried to change the schemes for advections for both velocities and k-epsilon (switch to caracteristics for which larger time steps may be used) and the computation could be run during the number of time steps you wrote in the steering file:
SCHEME FOR ADVECTION OF VELOCITIES = 1
SCHEME FOR ADVECTION OF K-EPSILON = 1
SCHEME OPTION FOR ADVECTION OF K-EPSILON = 1

I let you check if the results are good or not.

Hope this helps,

Chi-Tuan

hi everybody, hi chi-tuan,

thanks chi-tuan, my simulation did run.

josias