Trouble in initiating storm surge simulation using Delft3D FM (1.5.1.41875)

RE: Trouble in initiating storm surge simulation using Delft3D FM (1.5.1.41

2020-07-27T09:34:56Z

The problem is that sometimes the communication between flow and wave does not work immediately after the restart TheFreshMarketSurvey (the problem does not occur in the middle of the tidal cycle), but this happens at a random moment in the simulation (not always during the first restart for example). I cannot figure out what the problem is, and hope that you can help me. When the simulation crashes a com-file is present www.thefreshmarketsurvey.com

RE: Trouble in initiating storm surge simulation using Delft3D FM (1.5.1.41

2020-02-03T13:05:38Z

There are some small flow links on the D-Flow FM model mesh, which have to be removed. I've tried to run your test model and it gave me this error (see diagnosis file attached):

** ERROR : 4 small flow links discarded. Run 'merge circumcenters' to remove small flow links or increase threshold

This was previously discussed in this thread. Solution: merging small circumcenters of the mesh, using RGFGRID: Operations --> Grid --> Merge Circumcenters.
Kind regards,
Dmitry

Trouble in initiating storm surge simulation using Delft3D FM (1.5.1.41875)

2020-02-01T12:55:19Z

Hello everyone,

I've been trying to simulate storm surge case using Delft3D FM (1.5.1.41875). However, the program couldn't start the simulation showing a message "Initialization of tst05 has failed." ("tst05" is the name of the case).

I made a single irregular grid by utilizing Delft Dashboard and RGFGRID. A few quadrangular regular grids with different resolution were generated with DDB first, and they were transformed to irregular grids, and then merged together with triangular irregular grids in between. I tried to maintain the orthogonality of the grid below 0.5, even though it increased as I interpolate the grid with *.xyz bed level file (don't know why).As for bed level data, I combined GEBCO 2019, SRTM 4.1 (both obtained via DDB and a high resolution local data (obtained from an external source).Also, I generated a spider web file via DDB using JTWC best track file.

When I ran the model, FM gave me (!) messages, namely "Roughness&Viscosity&Diffusivity contains unspecified points, the calculation kernel will replace these with default values". I thought it was due to a kind of grid and bed level problem since at first the model gave the same message for Bathymetry "Bathymetry contains unspecified points, ~". However, Roughness&Viscosity&Diffusivity message were kept shown even after I modified the grid and solved bathymetry problem.

Because there is no relevant information in the manuals and websites, not to mention that the model does not allow me to look into it closely, I don't know how to solve this.Please take a look at the files I attached. Any help or discussion will be appreciated.

Thank you and Best,
Yongsu Na

---
tst05_net.nc - original grid file (w/o bed level).
tst05_rev2_net.nc - some grid cells were eliminated (parts of Yeosu, South Korea).
tst05_rev2_depCmbn_net.nc - bed level was interpolated.
---
P.S. I could not add any post at Delft3D FM Form (https://oss.deltares.nl/web/delft3dfm/general1/-/message_boards/category/217304). There weren't any buttons. Does anybody know?