intro story D-Flow FM

 

D-Flow Flexible Mesh

D-Flow Flexible Mesh (D-Flow FM) is the new software engine for hydrodynamical simulations on unstructured grids in 1D-2D-3D. Together with the familiar curvilinear meshes from Delft3D 4, the unstructured grid can consist of triangles, pentagons (etc.) and 1D channel networks, all in one single mesh. It combines proven technology from the hydrodynamic engines of Delft3D 4 and SOBEK 2 and adds flexible administration, resulting in:

  • Easier 1D-2D-3D model coupling, intuitive setup of boundary conditions and meteorological forcings (amongst others).
  • More flexible 2D gridding in delta regions, river junctions, harbours, intertidal flats and more.
  • High performance by smart use of multicore architectures, and grid computing clusters.
An overview of the current developments can be found here.
 
The D-Flow FM - team would be delighted if you would participate in discussions on the generation of meshes, the specification of boundary conditions, the running of computations, and all kinds of other relevant topics. Feel free to share your smart questions and/or brilliant solutions! 

 

=======================================================
We have launched a new website (still under construction so expect continuous improvements) and a new forum dedicated to Delft3D Flexible Mesh.

Please follow this link to the new forum: 
/web/delft3dfm/forum

Post your questions, issues, suggestions, difficulties related to our Delft3D Flexible Mesh Suite on the new forum.

=======================================================

** PLEASE TAG YOUR POST! **

 

 

Sub groups
D-Flow Flexible Mesh
DELWAQ
Cohesive sediments & muddy systems

 


Message Boards

d_hydro.exe abort after almost complet sediment simulation

RH
Roland Hesse, modified 5 Years ago.

d_hydro.exe abort after almost complet sediment simulation

Youngling Posts: 21 Join Date: 11/21/14 Recent Posts
Dear all,

I experienced an error after a "complete" multi domain (5 Domains) simulation, starting with <...>.bat-file. The d_hydro.exe aborts (windows error message: "d_hydro.exe does not work" [attached screen shot in German]) at the end of the simulation run. For some domains the writing of the tri-diag-<...>-files was not completed and the "TMP_<...>-files still exist. The simulation ran basically till the end; almost last time step. I already checked the available disk space. Though, I got a "FINISHED" message in the command line. This error occurred after I added/switched on the Processes "sediment" and "morpho" (with providing corresponding files). Before I had no problems. Any advice? A screenshot and 5/5 tri-diag-files are attached.

kind regards,

Roland
Adri Mourits, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts
Hi Roland,

The subdomains hinder eachother during cleaning/memory de-allocation. This happened before, was solved and now it is shown again.

If your output files contain all the data that should be written, you can neglect this error. Some debugging is needed to solve this. You can do that yourself or attach your input files (ideally stripped as much as possible) to a post on this forum. Then I will take care of it. You can also mail the files directly to me.

Thanks in advance.

Regards,

Adri
RH
Roland Hesse, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation

Youngling Posts: 21 Join Date: 11/21/14 Recent Posts
Hi Adri,


I think the output is mainly complete. In Quickplot/ for qpread.m I get an error when I try to access all time steps for the parameter "available sediment mass" . Though, I am not sure if there is an output for each time step for this parameter in general. The output of other parameters seems to be complete.

Adri Mourits:

Some debugging is needed to solve this. You can do that yourself or attach your input files (ideally stripped as much as possible)
Adri


To which input files do you refer? All Input files and the whole simulation folder respectively? I have in total 64 files/ 40 MB (due to domain decomposition) for this scenario. What do you mean exactly by stripped as much as possible? Just the files that are necessary to run the the basic simulation with sediment process?

kind regards,

Roland
RH
Roland Hesse, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation

Youngling Posts: 21 Join Date: 11/21/14 Recent Posts
I attached a simplified simulation (constant attribute parameters and boundary values, removal of "surplus" processes/ input files). I get the same error for this simulation . I did not upload the trim-<...>.dat/def files due to their size. Thank you in advance for the help!

regards

Roland
Adri Mourits, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation (Answer)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts
Hi Roland,

Thanks for uploading the input files.

I ran the model with revision 4970 without any problems. I'll try again with revision 4654.

Regards,

Adri
RH
Roland Hesse, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation

Youngling Posts: 21 Join Date: 11/21/14 Recent Posts
Hi Adri,

thanks for your answer. Though, I did not manage to get the executable for the newest version (tag 4970) due to following linker errors:
Error 37 error LNK2001: unresolved external symbol _NF_INQ_LIBVERS \Delft3D_tag_4970\engines_gpl\flow2d3d\packages\flow2d3d\f90_netcdf.lib(netcdf.obj) flow2d3d_intel13
The link error is related to the netcdf library ...\Delft3D_tag_4970\third_party_open\netcdf\lib\win32\Release\ifort13\f90_netcdf.lib. This file was manually specified in linker properties of vs-project.

it seems to be no compiler errors else.

kind regards,
Roland
Adri Mourits, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation (Answer)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts
Hi Roland,

NetCDF is added recently. For Windows, precompiled libraries are added, build with a few compiler versions (Intel 10, 12 and 13, if I'm correct). If you are using another compiler (version), you will get a compile error. In that case you have the following two options:
1. Use the latest Delft3D tag without NetCDF: revision 4440
2. Download the source code of NetCDF and compile it with the compiler you are using for Delft3D

Regards,

Adri
RH
Roland Hesse, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation

Youngling Posts: 21 Join Date: 11/21/14 Recent Posts
Hi Adri,

thanks for that advice. Thought, it is not a big problem as long as d_hydro.exe aborts at the last time step as it was the case before. Unfortunately, I experienced that the program also aborts after e.g. 30% of the whole simulation time for longer simulations. I also get no error message. Might this also be due to the cleaning/memory de-allocation?

Adri Mourits:
The subdomains hinder eachother during cleaning/memory de-allocation.


Or does this must have other reasons? I just want to make sure that this is not the case before I go on with searching for other possible errors. So far I did not find anything. I already started to check disk space, time series for boundaries etc.

kind regards,
Roland
Adri Mourits, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation (Answer)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts
Hi Roland,

No, that crash can not be caused by de-allocation. You'll have to check the diagnosis file (tri-diag.<runid>) and/or check your output.

Regards,

Adri
RH
Roland Hesse, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation

Youngling Posts: 21 Join Date: 11/21/14 Recent Posts
Hi Adri,

thanks. It seems that this problem is due to the size/ output time step of the history files (http://oss.deltares.nl/web/delft3d/general/-/message_boards/view_message/703268) Do you have an idea why this could happen?

regards,

Roland
Adri Mourits, modified 5 Years ago.

RE: d_hydro.exe abort after almost complet sediment simulation (Answer)

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts
See thread referred in post above.