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

Delft3D-WAVE stops unexpectedly

JH
Julia Hopkins, modified 5 Years ago.

Delft3D-WAVE stops unexpectedly

Youngling Posts: 14 Join Date: 1/1/13 Recent Posts
Hello!

I am running a coupled WAVE-FLOW model on nested grids using Domain Decomposition. I have been able to run this particular set-up (using four nested grids of varying resolution) successfully. However, recently I changed the WAVE boundary conditions (simple 3 parameter TPAR files, just changed the dates and some of the wave values) and the corresponding dates in the FLOW files, only to discover that the model now quits during WAVE computations at seemingly random times.

I get the error:

*** ERROR: file 'norm_end' expected to signal a correct SWAN calculation

The system cannot find the file specified.
The system cannot find the file specified.
The system cannot find the file specified.

Right before the model is about to compute wave conditions on the smallest grid. I am not sure what is wrong -- this happens somethings halfway through a run, sometimes only 1/3 of the way, and so on.

Any help on interpreting this error would be very appreciated!

Sincerely,

Julia
Ruairi MacIver, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Padawan Posts: 45 Join Date: 5/1/13 Recent Posts
This sounds very similar to a recurring error I have been experiencing (see thread http://oss.deltares.nl/web/delft3d/general/-/message_boards/view_message/533532).
Haven't resolved it yet. Would be helpful if you could you check a few things to see if it is a similar problem?

In my cases the last few screen outputs generally look something like
+SWAN is processing output request 1
+SWAN is processing output request 2
+SWAN is processing output request 3
+SWAN is processing output request 4
+SWAN is processing output request 5
*** WARNING: unable to run SWAN using "swan.sh". Trying with "swan.bat" ...
>>...End of SWAN run
*** ERROR: file 'norm_end' expected to signal a correct SWAN calculation


Also you should have a file named 'Errfile' in the model folder. In my case it generally contained one or more of the following message
Severe error : incorrect code in b.c. file: ??? ??? ???
where the ??? are an integer triplet.

Also check the content of the NEST??? (probably NEST004 in your case) to see if it contains any '******' (six *) entries.

Regards
Ruairi
JH
Julia Hopkins, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Youngling Posts: 14 Join Date: 1/1/13 Recent Posts
I do not get the "swan.sh" bit of the error message, but everything else is very similar (or identical) to what I see in my Errfile and NEST004.

I will check out the thread you linked to. Hopefully we can figure this out!
Ruairi MacIver, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Padawan Posts: 45 Join Date: 5/1/13 Recent Posts
I'm executing my simulations on a linux platform.
The 'swan.sh' part may not appear if simulations are executed on a windows platform.
Regards
Ruairi
Adri Mourits, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Yoda Posts: 1224 Join Date: 1/3/11 Recent Posts
Hi Ruairi and Julia,

When SWAN finishes normally, it generates a file named "norm_end". Delft3D-WAVE checks whether is file is generated. If not, it stops with an error message.

Possible problems can be classified as:
1. WAVE is not able to start SWAN
A message like "The system cannot find the file specified." points in that direction.

2. SWAN has run, but generates errors
Then you have to check the SWAN output. All SWAN output is written to ASCII files in the working directory with names like Errfile, OUTPUT. If you sort the files on date/time, you can check the few last files generated.

If this does not help you further, please add the (last part of the) full contents of the WAVE window output to a post on this forum.

Regards,

Adri
JH
Julia Hopkins, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Youngling Posts: 14 Join Date: 1/1/13 Recent Posts
Thank you for the prompt response.

There is an Errfile and associated generated, with the text

"Severe error : incorrect code in b.c. file: ____ ____"

repeated several times with different numbers filling the blank spaces.

I think there is an issue with how SWAN is processing the boundary conditions on the nested grids, but I am unsure of where to go from here.

Do you have any suggestions on how to fix the error?

Thank you, again!

The last part of the WAVE log is given below:

iteration 10; sweep 1
+iteration 10; sweep 2
+iteration 10; sweep 3
+iteration 10; sweep 4
accuracy OK in 98.70 % of wet grid points ( 98.00 % required)

+SWAN is processing output request 1
+SWAN is processing output request 2
1 file(s) copied.
1 file(s) copied.

C:\Delft3D\dd_Katama_4_2014_75>rem exit
>>...End of SWAN run
*** ERROR: file 'norm_end' expected to signal a correct SWAN calculation

The system cannot find the file specified.
The system cannot find the file specified.
The system cannot find the file specified.
Ruairi MacIver, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Padawan Posts: 45 Join Date: 5/1/13 Recent Posts
From my investigations I'm convinced the "Severe error" that both Julia and I report as appearing in the Errfile
"Severe error : incorrect code in b.c. file: ____ ____"
occurs while SWAN is reading the 2D spectra from the boundary condition file, the NEST??? file. This occurs somewhere around line 5572 of the swanmain routine of the source code.

In my case the code encounters the "******" entry in the data block following a FACTOR heading in the boundary condition (bc) file which it appears to interpret as the end of the current 2D spectrum. The code then reads the next line from the bc file and looks for one of the headings NODATA, ZERO or FACTOR in the first few characters. If the code doesn't find one of these headings it writes the Severe error message together with the first 20 characters of the line that has been read (hence the integers).

Haven't worked out why SWAN writes the "******" to the NEST??? file in the first place which I think is the root of the problem.

Regards
Ruairi
Ruairi MacIver, modified 5 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Padawan Posts: 45 Join Date: 5/1/13 Recent Posts
I've updated the linked post with some further observations (http://oss.deltares.nl/web/delft3d/general/-/message_boards/view_message/533532).
My present thinking is that it is some kind of spurious memory issue rather than a coding issue.

Ruairi
KL
Khai Jing Liew, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Youngling Posts: 9 Join Date: 9/21/15 Recent Posts
Hi Ruairi

Greeting to you. I am just wondering whether you find the solution for this problem or not because I am running into the same issue right now.
I am running two 1-month simulations for winter season and summer season respectively. Both use the exact same domain configuration with different wind forcing and boundary inputs in which winter season has stronger wind forcing and rougher sea condition (higher significant wave height). My simulation on summer time completed its run without crashing. However, my simulation on winter time crash several times due to the severe error occur in SWAN. I checked my ERR written as

Severe error : incorrect code in b.c. file: 8 6 2
Severe error : incorrect code in b.c. file: 124 6
Severe error : incorrect code in b.c. file: 0 0
Severe error : incorrect code in b.c. file: 0 0
Severe error : incorrect code in b.c. file: 0 0
Severe error : incorrect code in b.c. file: 0 0

and there are ****** in my NEST002.

The swn-diag file and PRINT does not print any errors.

Any help and suggestion would be very much appreciated.

Thank you
KJ
Adri Mourits, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly (Answer)

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

Can you try the following:
1. Open the following file in an editor: delf3d_source_code\src\engines_gpl\wave\packages\data\src\swan_input.f90
And search for this piece of code, around line 4700:
    line(1:72) = '$***************************** OUTPUT REQUEST **************************'
    write (luninp, '(1X,A)') line
    line       = ' '
    line(1:2)  = '$ '
    write (luninp, '(1X,A)') line

Add two lines at the end of this block:
    line(1:72) = '$***************************** OUTPUT REQUEST **************************'
    write (luninp, '(1X,A)') line
    line       = ' '
    line(1:2)  = '$ '
    write (luninp, '(1X,A)') line
    line       = 'OUTPUT OPTIONS SPEC ndec=8'
    write (luninp, '(1X,A)') line

2. Compile the source code
3. Run your model

Does this solve your problem? Please add your findings here on the forum; if it works, I'll try to add it to the trunk.

Regards,

Adri
Adri Mourits, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly (Answer)

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

A colleague ran into the same problem. The fix in my previous post solves it. I committed this fix in the trunk in revision 7565 (no tagged version yet).

Regards,

Adri
HL
Huidi Liang, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Youngling Posts: 1 Join Date: 4/8/17 Recent Posts
Hi Adri,
I try to run delft3d on Linux cluster recently. I have some problems on wave-module and flow-wave coupling.
1) when I run the wave model, it crashed with the errors as follow:

- Delft3D
- Deltares, Delft3D-WAVE Version 3.05.01.000000, Mar 28 2017, 14:55:16
- Open source
-
--------------------------------------------------------------------------------
*** MESSAGE: Delft3D-WAVE runs stand alone
SWAN_INPUT: [Processes] WaveForces is set to "dissipation 3d"
To switch on radiation stresses: "WaveForces = radiation stresses <2013"
Done reading input
Initialising grids and grid mappings
Start loop over SWAN grids, time = 60.000
Allocate input fields
Allocate and read SWAN depth
Write SWAN depth file
Deallocate input fields
Write SWAN input
<<Run SWAN...
*** WARNING: unable to run SWAN using "swan.sh". Trying with "swan.bat" ...
>>...End of SWAN run
*** ERROR: file 'norm_end' expected to signal a correct SWAN calculation




2) what's kind of .sh document should I have when I run the flow-wave coupling?
Qinghua Ye, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Jedi Council Member Posts: 612 Join Date: 3/2/11 Recent Posts
Hi Huidi,

The problem you posted is not the same as those listed above. It may be from the wave run. Pls check the wave diagnose file.

Greetings,

Qinghua
SM
Seyed Moin091 Mojabi, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly

Youngling Posts: 4 Join Date: 3/18/13 Recent Posts
Dear Adri

I tired to make a nested grid model in DELFT3D-Wave. Unfortunately, the DELFT3D-Wave suddenly stops with the following message:


SWAN is preparing computation

forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
swan_4072ABCDE_de 000000000051FC4C newton_ 1723 swanser.f
swan_4072ABCDE_de 000000000051D459 cvmesh_ 1241 swanser.f
swan_4072ABCDE_de 000000000044464C sinupt_ 3883 swanmain.f
swan_4072ABCDE_de 0000000000443B26 sprcon_ 2868 swanmain.f
swan_4072ABCDE_de 000000000044208C swprep_ 2424 swanmain.f
swan_4072ABCDE_de 000000000043D33D swmain_ 406 swanmain.f
swan_4072ABCDE_de 000000000043CA34 MAIN__ 110 swanmain.f
swan_4072ABCDE_de 000000000040425C Unknown Unknown Unknown
libc.so.6 00007F7673257F45 Unknown Unknown Unknown
swan_4072ABCDE_de 0000000000404169 Unknown Unknown Unknown
*** WARNING: unable to run SWAN using "swan.sh". Trying with "swan.bat" ...
sh: 1: swan.bat: not found
>>...End of SWAN run
*** ERROR: file 'norm_end' expected to signal a correct SWAN calculation

I am sure the problem lies in the Nesting, because the model works very well with individual girds which are used to make the nested Grid


How can I approach and solve the problem ?

I attached the model if you need it for further information


Kind Regards
Moien

Attachments:

Adri Mourits, modified 2 Years ago.

RE: Delft3D-WAVE stops unexpectedly

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

When running the domains individually, errors are printed in the SWAN output file PRINT.

Regards,

Adri