swash-users Mailing List for SWASH (Page 3)
Brought to you by:
mzijlema
You can subscribe to this list here.
2011 |
Jan
|
Feb
|
Mar
(1) |
Apr
|
May
(2) |
Jun
|
Jul
|
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
(1) |
---|---|---|---|---|---|---|---|---|---|---|---|---|
2012 |
Jan
|
Feb
(1) |
Mar
|
Apr
|
May
(3) |
Jun
|
Jul
(1) |
Aug
|
Sep
|
Oct
(3) |
Nov
|
Dec
|
2013 |
Jan
|
Feb
|
Mar
(1) |
Apr
(2) |
May
(3) |
Jun
|
Jul
|
Aug
(1) |
Sep
(1) |
Oct
(1) |
Nov
(5) |
Dec
(4) |
2014 |
Jan
|
Feb
(1) |
Mar
|
Apr
(2) |
May
(2) |
Jun
(4) |
Jul
|
Aug
|
Sep
(2) |
Oct
(2) |
Nov
(1) |
Dec
|
2015 |
Jan
(2) |
Feb
(1) |
Mar
(1) |
Apr
(2) |
May
(2) |
Jun
(11) |
Jul
(6) |
Aug
(6) |
Sep
|
Oct
(1) |
Nov
(1) |
Dec
(1) |
2016 |
Jan
(6) |
Feb
(1) |
Mar
(3) |
Apr
(4) |
May
(2) |
Jun
(3) |
Jul
(2) |
Aug
|
Sep
|
Oct
(3) |
Nov
|
Dec
|
2017 |
Jan
(1) |
Feb
|
Mar
(1) |
Apr
(2) |
May
(7) |
Jun
(2) |
Jul
(3) |
Aug
(2) |
Sep
(2) |
Oct
|
Nov
(1) |
Dec
(2) |
2018 |
Jan
(1) |
Feb
(4) |
Mar
(1) |
Apr
|
May
(2) |
Jun
(1) |
Jul
(1) |
Aug
|
Sep
(1) |
Oct
|
Nov
|
Dec
|
2019 |
Jan
(5) |
Feb
|
Mar
(1) |
Apr
(8) |
May
(1) |
Jun
(2) |
Jul
(2) |
Aug
(2) |
Sep
(1) |
Oct
|
Nov
(4) |
Dec
|
2020 |
Jan
|
Feb
|
Mar
|
Apr
(2) |
May
(10) |
Jun
|
Jul
|
Aug
(1) |
Sep
|
Oct
(1) |
Nov
(4) |
Dec
(3) |
2021 |
Jan
|
Feb
|
Mar
(1) |
Apr
(5) |
May
(8) |
Jun
(8) |
Jul
|
Aug
(1) |
Sep
|
Oct
|
Nov
|
Dec
|
2022 |
Jan
(1) |
Feb
(3) |
Mar
(8) |
Apr
(6) |
May
(1) |
Jun
(5) |
Jul
(3) |
Aug
(6) |
Sep
(11) |
Oct
(2) |
Nov
(3) |
Dec
(1) |
2023 |
Jan
(1) |
Feb
|
Mar
(6) |
Apr
|
May
|
Jun
(1) |
Jul
(3) |
Aug
|
Sep
|
Oct
(3) |
Nov
(3) |
Dec
(2) |
2024 |
Jan
(1) |
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: D. L. <d.l...@gm...> - 2022-06-07 09:01:12
|
Dear All, *Features of SWASH* page of SWASH site has the following: *LimitationsAt present, SWASH does not account for:hotstarting (temporary limit)unstructured mesh computations in parallel using MPI (temporary limit)* On *Modifications *page it is stated that *The current version number of SWASH is 7.01.The use of unstructured mesh is included. The mesh may be comprised of triangular cells only. The numerical approach is described in Zijlema (2020).* Does the latter mean that SWASH 7.01 is able to execute triangular mesh computations in parallel using MPI? With best regards, Dmitri |
From: Laura S. <l.s...@qu...> - 2022-06-02 12:21:45
|
Hello ~ SWASH ~ Users! We have another monthly SWASH Users Meetup on Monday June 6th at 1 pm ET (10 am PT). This is an opportunity to learn about different modeling applications, pre-/post- processing strategies, and troubleshoot challenges with a group of fellow modelers. In past meetings we have learned about: harbor applications, directional spectra, methods to calculate runup, mitigating pesky reflections, variable friction inputs, and more! These meetings will occur on the first Monday of each month --- looking forward to chatting SWASH! ZOOM INFO: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Laura Szczyrba (she/her) is inviting you to a scheduled Zoom meeting. Topic: SWASH Users Meetup Time: Jun 6, 2022 01:00 PM Eastern Time (US and Canada) Join Zoom Meeting https://queensu.zoom.us/j/98985811483?pwd=YUpmdmhKWFpDdmNFRGJMcjBhSHFhUT09 Meeting ID: 989 8581 1483 Passcode: 324949 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Cheers! Laura Szczyrba (she/her) PhD Candidate Geological Sciences & Civil Engineering Queen’s University |
From: Clayton H. <cl...@ca...> - 2022-05-27 19:18:59
|
Hello SWASH Users, I have been experimenting with the new unstructured grid version of SWASH for tsunami propagation. In these experiments I have found the radiation and sponge layer boundaries to be problematic. I have not experienced these problems with regular grids, so I'm guessing it has something to do with the implementation with unstructured grids. For radiation boundary conditions, it appears that an instability can start at some nodes and progress out of control until it spoils the entire simulation (see figure below). It appears more likely to happen when the wave crest is not parallel to the radiating boundary. I have examined the mesh, and did not find anything out of the ordinary at the location of the problematic nodes. When I attempt to use sponge layers, the simulation does not make it past the initialization, and just hangs on SwashUSpongeLayer. Has anyone else experienced either of these issues? Any guidance on how to avoid these problems would be appreciated. Thanks, Clayton [image: image.png] |
From: Laura S. <l.s...@qu...> - 2022-04-28 14:02:44
|
Hello! I am having difficulty calculating the A1, B1, A2, B2 Fourier coefficients from SWASH results. Currently, I am using the outputs: WATLEV (water level), VKSI (U velocity), and VETA (V velocity). Then, I compute the auto-spectra and cross-spectra and apply those to Lygre and Krogstad (1986) Eqtns. 10. At this point, I am uncertain if I need to modify my approach since Lygre and Krogstad (1986) was designed for heave/pitch/roll or if I have a mistake in my calculations. Any suggestions would be much appreciated. Once I have coefficients correctly calculated, I have no problem calculating the directional spectra from there (which is the ultimate goal). Cheers, Laura Szczyrba (she/her) PhD Student Geological Sciences & Civil Engineering Queen’s University |
From: Laura S. <l.s...@qu...> - 2022-04-25 13:50:15
|
Hello ~~SWASH~~ users! We have our next SWASH user group meetup on Monday, May 2 at 1 pm ET (10 am PT). Our last meeting was amazing -- we discussed variable friction inputs, reflections at lateral boundaries, rotating bathymetry, runup, etc. I know I walked away with some new ideas! Zoom invite copied below. We are continuing to meet the first Monday of each month, feel free to stop by! ________________________________________________________________________________ Laura Szczyrba (she/her) is inviting you to a scheduled Zoom meeting. Topic: SWASH User Group Meeting Time: May 2, 2022 01:00 PM Eastern Time (US and Canada) Join Zoom Meeting https://queensu.zoom.us/j/91040662612?pwd=eGFoZlFHUEYxeXVRYVNhQUF6N1FOZz09 Meeting ID: 910 4066 2612 Passcode: 443517 _________________________________________________________________________________ Thanks, Laura Szczyrba (she/her) PhD Student Geological Sciences & Civil Engineering Queen’s University |
From: shaiashe <sha...@te...> - 2022-04-18 14:01:01
|
Hi Theo, It would help if you modeled the harbor structures as near-impervious porous zones, i.e., porosity and stonesize of 0.001, and keep the bed level low. Shai From: theofilos koronakis <tko...@gm...> Sent: Monday, April 18, 2022 3:25 PM To: swa...@li... Subject: [swash-users] Project 3D simulation using SWASH Hello , I'm new to swash program and i would like some assistance on how to handle it . What im trying to do is to simulate a two dimensional wave episode in a harbour area using monochromatic waves with height = 2 m and period = 10 sec. The area contains a steep slope (1/460) starting at max depth 20 m and ends at 10 m depth , on where the harbour is located. Since i want to simulate what is happening inside the harbour I use sponge layers at the upper and lower sides on the computational grid. Here is the code im using for the test PROJ 'Harbour_test' 'test' $ $***********MODEL INPUT********************************* $ MODE NONST TWOD $ CGRID REG 0. 0. 0. 1000. 860. 100 86 $ VERT 2 $ INPGRID BOTTOM 0. 0. 0. 100 86 10. 10. $ READINP BOTTOM 1. 'Bott' 3 0 FREE $ INIT zero $ BOU SIDE W CCW BTYPE WEAK CON REG 1.5 10 $ SPONGE UP LO 240.0 240.0 $ NONHYD BOX PREC ILU $ DISCRET UPW UMOM MOM H TIMEI METH EXPL 0.2 0.6 $ $************ OUTPUT REQUESTS ************************* $ QUANT DIST WATL dur 60 SEC $ GROUP 'HARBOUR' 1 101 1 87 TABLE 'HARBOUR' HEAD 'TEST.tbl' TSEC XP WATL BOTLEV $ TEST 1,0 COMPUTE 000000.000 0.02 SEC 000500.00 STOP The problem that i'm having is that i get steady water level -99.00 m and a steady tsec 296,5 sec and i don't know why. I really hope that you can assist me and guide me to right results Thank you , for your time External e-mail, be judicious when opening attachments or links |
From: theofilos k. <tko...@gm...> - 2022-04-18 12:25:39
|
Hello , I'm new to swash program and i would like some assistance on how to handle it . What im trying to do is to simulate a two dimensional wave episode in a harbour area using monochromatic waves with height = 2 m and period = 10 sec. The area contains a steep slope (1/460) starting at max depth 20 m and ends at 10 m depth , on where the harbour is located. Since i want to simulate what is happening inside the harbour I use sponge layers at the upper and lower sides on the computational grid. Here is the code im using for the test PROJ 'Harbour_test' 'test' $ $***********MODEL INPUT********************************* $ MODE NONST TWOD $ CGRID REG 0. 0. 0. 1000. 860. 100 86 $ VERT 2 $ INPGRID BOTTOM 0. 0. 0. 100 86 10. 10. $ READINP BOTTOM 1. 'Bott' 3 0 FREE $ INIT zero $ BOU SIDE W CCW BTYPE WEAK CON REG 1.5 10 $ SPONGE UP LO 240.0 240.0 $ NONHYD BOX PREC ILU $ DISCRET UPW UMOM MOM H TIMEI METH EXPL 0.2 0.6 $ $************ OUTPUT REQUESTS ************************* $ QUANT DIST WATL dur 60 SEC $ GROUP 'HARBOUR' 1 101 1 87 TABLE 'HARBOUR' HEAD 'TEST.tbl' TSEC XP WATL BOTLEV $ TEST 1,0 COMPUTE 000000.000 0.02 SEC 000500.00 STOP The problem that i'm having is that i get steady water level -99.00 m and a steady tsec 296,5 sec and i don't know why. I really hope that you can assist me and guide me to right results Thank you , for your time |
From: H. G. G. <go...@gm...> - 2022-04-17 17:40:57
|
Hi Toby, It's been some time since your question; however, I have just seen your email. Please see my comments below: 1. In my experience, the wave direction should be as perpendicular as possible to the boundaries. (I also remember a recommendation similar to that from the manual). To do that, you should rotate your bathymetry file. 2. We conducted several test simulations a few years ago to understand which L/dx ratio is appropriate for SWASH simulations (to resolve the waves properly) with monochromatic waves. We have seen that the L/dx ratio should be around 30. In your case, your L/dx ratio is approximately 21 (considering Tp=5.4 sec, h=11.6 m), which might cause unrealistic decay of your wave height in your simulation. Although our study was conducted for monochromatic waves, I believe it is still plausible to consider the result of our study as a rule of thumb. Of course, you should consider all the individual waves in your wave series (at least a representative amount of them), i.e., the L/dx ratio is much smaller for the shorter individual waves in your wave series. I hope you find my answers helpful, and good luck with your simulations. Kind regards, Gokhan Toby Johnson <tob...@gm...>, 28 Mar 2022 Pzt, 14:11 tarihinde şunu yazdı: > Hi, > > Hoping that some experienced SWASH users might be able to give me some > feedback regarding a couple of issues I'm running into as a first time user: > > 1. When using weakly reflective boundary conditions around the corner of > the domain, I notice that there is a shadow zone radiating from the > boundary corner where significant wave heights are much lower - seems like > there is some kind of discontinuity in the boundary condition at the > corner. Boundary was set up as a single SEGMENT. This effect is more > noticeable for conditions with lower directional spreading, however it can > kind of be observed in the results below as well. > > 2. I'm attempting to model relatively low period waves (Tp = 4-5s.) I have > reduced the bathymetry to bring the model within the recommended validity > range using 2-layers and reducing the loss of evanescent modes. Courant > remains less than 0.4 and grid size is 2x2 m. The issue I'm having is that > I notice a strong decay in the significant wave height from the boundary > which strangely is even worse using 2-layers compared1-layer which is meant > to be removing far more evanescents. I'm not sure what is going on here and > it's really holding me back from using the model model, particularly using > SWAN spectral boundaries which are not so easy to artificially scale to > account for the damping. I have included a snapshot of the comparison in > simulations for 1 and 2-layer cases below. Boundary Hs = 1.94 m. Tp = > 5.4.s, m = 3, Dir = 315 deg. JONSWAP gamma = 3.3, maximum depth = 11.6 m. > Between the boundary and the breakwater the depth is constant so breaking > processes should not be having an effect. > > [image: image.png] > > Any advice or feedback here would be greatly appreciated. > > Regards, > Toby > _______________________________________________ > swash-users mailing list > swa...@li... > https://lists.sourceforge.net/lists/listinfo/swash-users > |
From: Toby J. <tob...@gm...> - 2022-04-02 10:39:49
|
Hi, Upon finishing my simulation in MPI SWASH recombines the MPI files into a single file as expected. However, I notice that this often takes a very long time to complete. For instance exporting a BLOCK .mat file of the computation grid with 3 varies and 60 frames seems to take upwards of 15-20 minutes in some cases (20 cores) Is this normal behaviour and does anybody know if it's possible to speed up this process? It seems unusually slow to combine a file of a few gigabytes. Regards, Toby |
From: Toby J. <tob...@gm...> - 2022-03-30 09:03:26
|
Hi Dr. Menno, Thanks for the feedback. I noted in some previous user posts that this was a potential solution. According to the Implementation Manual there are only 3 options (1,2 and 3) that can be used for 1-Layer mode. Unfortunately I have tried all of these without any success. For now, I've decided to reduce my grid extent in order to run on more cores but it'd be nice if this issue could be looked into as a future bug fix (if indeed it's SWASH related and not MPI). Regards, Toby On Wed, Mar 30, 2022 at 10:27 AM Menno Eelkema | Aqua Vision < m.e...@aq...> wrote: > Hi Toby, > > > > I’ve found that in parallel mode the grid partitioning options in the > Swashinit-file are usually a source of run-errors in case of large grids. > Finding the right partitioning option is a bit trial and error, but it > might be a solution. > > > > Kind regards, > > > > > > *Dr. Ir. Menno Eelkema* > > Coastal Engineer > > [image: Afbeelding met tekst Automatisch gegenereerde beschrijving] > > *Aqua Vision BV* > > Servaasbolwerk 11 > > 3512 NK Utrecht > > The Netherlands > > aquavision.nl <http://www.aquavision.nl/> > > +31 (0)6 81850067 > > +31 (0)30 245 9872 > > *m.e...@aq... <m.e...@aq...>* > > > > > > *Van:* Toby Johnson <tob...@gm...> > *Verzonden:* dinsdag 29 maart 2022 15:37 > *Aan:* swa...@li... > *Onderwerp:* [swash-users] MPICH2 Failure - Large Grid Sizes > > > > Hi all, > > > > Running into model crashes (see below) when trying to execute a simulation > with MPICH2 on more than 10 cores and a high grid size (3000 x 3000). Tee > model executes fine in serial and on a lower number of cores so its not > setup related. > > > > Does anyone know the cause of this? Given that one would want more cores > for larger grids this is proving to be quite a limitation. > > > > Thanks, > > Toby > |
From: Menno E. | A. V. <m.e...@aq...> - 2022-03-30 07:00:42
|
Hi Toby, I’ve found that in parallel mode the grid partitioning options in the Swashinit-file are usually a source of run-errors in case of large grids. Finding the right partitioning option is a bit trial and error, but it might be a solution. Kind regards, Dr. Ir. Menno Eelkema Coastal Engineer [Afbeelding met tekst Automatisch gegenereerde beschrijving] Aqua Vision BV Servaasbolwerk 11 3512 NK Utrecht The Netherlands aquavision.nl<http://www.aquavision.nl/> +31 (0)6 81850067 +31 (0)30 245 9872 m.e...@aq...<mailto:m.e...@aq...> Van: Toby Johnson <tob...@gm...> Verzonden: dinsdag 29 maart 2022 15:37 Aan: swa...@li... Onderwerp: [swash-users] MPICH2 Failure - Large Grid Sizes Hi all, Running into model crashes (see below) when trying to execute a simulation with MPICH2 on more than 10 cores and a high grid size (3000 x 3000). Tee model executes fine in serial and on a lower number of cores so its not setup related. [cid:image002.png@01D84410.0B8EC0E0] Does anyone know the cause of this? Given that one would want more cores for larger grids this is proving to be quite a limitation. Thanks, Toby |
From: Toby J. <tob...@gm...> - 2022-03-29 13:37:57
|
Hi all, Running into model crashes (see below) when trying to execute a simulation with MPICH2 on more than 10 cores and a high grid size (3000 x 3000). Tee model executes fine in serial and on a lower number of cores so its not setup related. [image: 60bb2e01-05ef-4d63-93a4-02b43b6ce631.png] Does anyone know the cause of this? Given that one would want more cores for larger grids this is proving to be quite a limitation. Thanks, Toby |
From: Cassandra H. <che...@uc...> - 2022-03-29 00:18:52
|
Hi fellow SWASH users Some of us SWASH users who met at Ocean Sciences 2022 have decided to start up a discussion group on zoom for SWASH users. Our next meeting is April 4, 10 AM PT (1 PM ET), and our ongoing schedule will be the first monday of the month. (April 4, May 2, June 6, etc.) Zoom details: Topic: SWASH Users Group Time: Apr 4, 2022 10:00 AM Pacific Time (US and Canada) https://ucsd.zoom.us/j/97707491033 Meeting ID: 977 0749 1033 One tap mobile +16692192599,,97707491033# US (San Jose) +16699006833,,97707491033# US (San Jose) Dial by your location +1 669 219 2599 US (San Jose) +1 669 900 6833 US (San Jose) +1 213 338 8477 US (Los Angeles) Meeting ID: 977 0749 1033 Find your local number: https://ucsd.zoom.us/u/abOS2Bx6qQ |
From: Toby J. <tob...@gm...> - 2022-03-28 10:57:56
|
Hi, Hoping that some experienced SWASH users might be able to give me some feedback regarding a couple of issues I'm running into as a first time user: 1. When using weakly reflective boundary conditions around the corner of the domain, I notice that there is a shadow zone radiating from the boundary corner where significant wave heights are much lower - seems like there is some kind of discontinuity in the boundary condition at the corner. Boundary was set up as a single SEGMENT. This effect is more noticeable for conditions with lower directional spreading, however it can kind of be observed in the results below as well. 2. I'm attempting to model relatively low period waves (Tp = 4-5s.) I have reduced the bathymetry to bring the model within the recommended validity range using 2-layers and reducing the loss of evanescent modes. Courant remains less than 0.4 and grid size is 2x2 m. The issue I'm having is that I notice a strong decay in the significant wave height from the boundary which strangely is even worse using 2-layers compared1-layer which is meant to be removing far more evanescents. I'm not sure what is going on here and it's really holding me back from using the model model, particularly using SWAN spectral boundaries which are not so easy to artificially scale to account for the damping. I have included a snapshot of the comparison in simulations for 1 and 2-layer cases below. Boundary Hs = 1.94 m. Tp = 5.4.s, m = 3, Dir = 315 deg. JONSWAP gamma = 3.3, maximum depth = 11.6 m. Between the boundary and the breakwater the depth is constant so breaking processes should not be having an effect. [image: image.png] Any advice or feedback here would be greatly appreciated. Regards, Toby |
From: Denny R. <den...@gm...> - 2022-03-08 13:07:12
|
I have faced this issue when I run SWASH in Windows using the binary installation and MPICH2 executable. It is solved if you compile the software by yourself; compiling it works flawless either for Windows or Linux under Intel or AMD architectures. Best, Denny Rivas Hydrologic and hydrodynamic modelling consultant El mar., 8 mar. 2022 6:45 a. m., Menno Eelkema | Aqua Vision < m.e...@aq...> escribió: > Hi Erik, > > > > I’ve found that in parallel mode the grid partitioning options in the > Swashinit-file are usually a source of run-errors in case of large grids. > Finding the right partitioning option is a bit trial and error, but it > might be a solution. > > > > Kind regards, > > > > > > *Dr. Ir. Menno Eelkema* > > Coastal Engineer > > [image: Afbeelding met tekst Automatisch gegenereerde beschrijving] > > *Aqua Vision BV* > > Servaasbolwerk 11 > > 3512 NK Utrecht > > The Netherlands > > aquavision.nl <http://www.aquavision.nl/> > > +31 (0)6 81850067 > > +31 (0)30 245 9872 > > *m.e...@aq... <m.e...@aq...>* > > > > > > > > > > > > *Van:* erik young <eri...@gm...> > *Verzonden:* maandag 7 maart 2022 23:16 > *Aan:* swa...@li... > *Onderwerp:* [swash-users] Problem initiating PRINT-00X files in MPI > SWASH run > > > > I am trying to run a 2d wave simulation, with everything working in > serial. When I run the simulation in parallel: > > > > > swashrun [filename] 8 > > > > I get the following error on all processors other than the first: > > > > forrtl: severe (47): write to READONLY file, unit 4, file C:\...\PRINT > Image PC Routine Line Source > swash.exe 00007FF7438CC5E8 Unknown Unknown Unknown > swash.exe 00007FF7438859BA Unknown Unknown Unknown > swash.exe 00007FF74380A269 OCPINI 290 > ocpids.for > swash.exe 00007FF743482CB9 SWASHINIT 85 > SwashInit.f90 > swash.exe 00007FF743481B69 SWASHMAIN 119 > SwashMain.f90 > swash.exe 00007FF743481AB4 MAIN__ 63 > Swash.f90 > swash.exe 00007FF74394C8C2 Unknown Unknown Unknown > swash.exe 00007FF74394CCB4 Unknown Unknown Unknown > KERNEL32.DLL 00007FFF46F37034 Unknown Unknown Unknown > ntdll.dll 00007FFF47A82651 Unknown Unknown Unknown > > > > Based on documentation the run should initiate 8x print files for 8x > processors, but instead it attempts to write them all in the same file, and > fails. > > > > Any help much appreciated! > _______________________________________________ > swash-users mailing list > swa...@li... > https://lists.sourceforge.net/lists/listinfo/swash-users > |
From: Menno E. | A. V. <m.e...@aq...> - 2022-03-08 12:43:19
|
Hi Erik, I’ve found that in parallel mode the grid partitioning options in the Swashinit-file are usually a source of run-errors in case of large grids. Finding the right partitioning option is a bit trial and error, but it might be a solution. Kind regards, Dr. Ir. Menno Eelkema Coastal Engineer [Afbeelding met tekst Automatisch gegenereerde beschrijving] Aqua Vision BV Servaasbolwerk 11 3512 NK Utrecht The Netherlands aquavision.nl<http://www.aquavision.nl/> +31 (0)6 81850067 +31 (0)30 245 9872 m.e...@aq... Van: erik young <eri...@gm...> Verzonden: maandag 7 maart 2022 23:16 Aan: swa...@li... Onderwerp: [swash-users] Problem initiating PRINT-00X files in MPI SWASH run I am trying to run a 2d wave simulation, with everything working in serial. When I run the simulation in parallel: > swashrun [filename] 8 I get the following error on all processors other than the first: forrtl: severe (47): write to READONLY file, unit 4, file C:\...\PRINT Image PC Routine Line Source swash.exe 00007FF7438CC5E8 Unknown Unknown Unknown swash.exe 00007FF7438859BA Unknown Unknown Unknown swash.exe 00007FF74380A269 OCPINI 290 ocpids.for swash.exe 00007FF743482CB9 SWASHINIT 85 SwashInit.f90 swash.exe 00007FF743481B69 SWASHMAIN 119 SwashMain.f90 swash.exe 00007FF743481AB4 MAIN__ 63 Swash.f90 swash.exe 00007FF74394C8C2 Unknown Unknown Unknown swash.exe 00007FF74394CCB4 Unknown Unknown Unknown KERNEL32.DLL 00007FFF46F37034 Unknown Unknown Unknown ntdll.dll 00007FFF47A82651 Unknown Unknown Unknown Based on documentation the run should initiate 8x print files for 8x processors, but instead it attempts to write them all in the same file, and fails. Any help much appreciated! |
From: erik y. <eri...@gm...> - 2022-03-07 22:16:03
|
I am trying to run a 2d wave simulation, with everything working in serial. When I run the simulation in parallel: > swashrun [filename] 8 I get the following error on all processors other than the first: forrtl: severe (47): write to READONLY file, unit 4, file C:\...\PRINT Image PC Routine Line Source swash.exe 00007FF7438CC5E8 Unknown Unknown Unknown swash.exe 00007FF7438859BA Unknown Unknown Unknown swash.exe 00007FF74380A269 OCPINI 290 ocpids.for swash.exe 00007FF743482CB9 SWASHINIT 85 SwashInit.f90 swash.exe 00007FF743481B69 SWASHMAIN 119 SwashMain.f90 swash.exe 00007FF743481AB4 MAIN__ 63 Swash.f90 swash.exe 00007FF74394C8C2 Unknown Unknown Unknown swash.exe 00007FF74394CCB4 Unknown Unknown Unknown KERNEL32.DLL 00007FFF46F37034 Unknown Unknown Unknown ntdll.dll 00007FFF47A82651 Unknown Unknown Unknown Based on documentation the run should initiate 8x print files for 8x processors, but instead it attempts to write them all in the same file, and fails. Any help much appreciated! |
From: Laura S. <l.s...@qu...> - 2022-02-17 18:42:09
|
Hello SWASH users! I am encountering difficulty when inputting a stillwater elevation above 0. If I maintain SET LEVEL = 0, the model runs beautifully. However, if I add any amount of stillwater elevation (e.g., SET LEVEL = 0.1), the model crashes and indicates that the water level is too far below bottom level. Reducing the time step and adding more vertical layers does not seem to help, but perhaps there is a balance that I am not finding. Does anybody have any advice? I have attached the .sws file for reference. Cheers, Laura Szczyrba (she/her) PhD Student Geological Sciences & Civil Engineering Queen’s University |
From: Paulo Avilez-V. <pva...@fe...> - 2022-02-02 01:08:32
|
Dear Patricia The answer by Dr Gokhan Guler is quite thorough and accurate. Nevertheless let me point out that sometimes even the usage of the EXCEPTION option in the INPGRID BOTTOM command can cause that kind of instabilities. In the solver settings NONHYDrostatic BOX 0.5 SOLVer 0.01 0.0 1000 0.55 PREConditioner ILU As pointed out by Dr Gokhan Guler, 0.5 corresponds to the Crank-Nicolson method which is 2nd order. However, this method is neutrally stable and thus it can not actively contribute to the stability of the solution. Using 1.0 corresponds to the Euler method, which is unconditionally stable and thus solves the problem. But the Euler method is 1st order and very dissipative, thus you might end up "eating" all the energy of your wave along its propagation. I would suggest using 0.6 or 0.67 for a stable and not too dissipative time integration. . Best regards Paulo Dr-Eng Paulo Avilez-Valente Assistant Professor Faculdade de Engenharia Universidade do Porto Rua Dr Roberto Frias P-4200-465 Porto Portugal On Tue, 1 Feb 2022 at 21:05, H. Gokhan Guler <go...@gm...> wrote: > Dear Patricia, > > I faced this error so many times. Previously, I have taken several notes > regarding the potential reasons for this error. I copy these notes below > for your information. I could not look into the details of your *.sws file. > However, I noticed that some parts of your input file are related to the > below-given items. Therefore, I hope that these notes might be helpful for > your case. > > Best regards, > > Gokhan > > > > "INSTABLE: water level ..." error, potential reasons: > > 1. The wave generating boundaries in 2D simulations: > > If the target wave is large compared to the water depths along the > boundary. > > Presence of land along the wave generating boundary. > > 2. Sponge layers might cause instability problems in 2D simulations. > > 3. Activating BREaking command might cause instabilities. > > 4. Solver settings: > NONHYDrostatic BOX 0.5 SOLVer 0.01 0.0 1000 0.55 PREConditioner ILU > For example, in this line, 0.5 corresponds to the Crank-Nicolson method. > It's more accurate; however, it might cause stability problems, especially > in 2D simulations. Using 1 corresponding to the Euler method may solve the > problem. > > 5. Discretization schemes: > Some schemes may cause wiggles in the free-surface solution (as indicated > in the manual). These might accumulate and cause instability problems. > My experience is using the default schemes in general or the following > schemes for monochromatic waves. > FROMM for horizontal advective terms > BDF for vertical advective terms > MUSCL for CORRDEP > > > > > > > > > > > Patricia Mares Nasarre <pat...@gm...>, 27 Oca 2022 Per, > 19:24 tarihinde şunu yazdı: > >> Hi, >> Does anyone know why I get this error? >> >> "** Terminating error: INSTABLE: water level is too far below the bottom >> level! >> ** Message : Please reduce the time step!" >> >> It appears during the first seconds of the simulation of a 2D case. The >> input file is: >> >> "PROJ 'name' 'O1' >> >> SET LEVEL -0.29 >> >> MODE NONST TWOD >> >> COORD CART >> >> CGRID REG 502400 3906100 0 17600 11900 879 594 >> >> VERT 3 30 40 30 PERC >> >> INPGRID BOT REG 502400 3906100 0 879 594 20 20 EXC -9999 >> >> READINP BOT 1 'file.bot' 1 0 FREE >> >> WIND 5.2 210 CON >> >> INIT zero >> >> BOUN SHAP JONSWAP 3.3 SIG PEAK DSPR DEGR >> >> BOUNDCOND SIDE N CCW BTYPE WEAK SMOO 30 SEC CON SPECT 0.5 4 210 30 >> >> BOUNDCOND SIDE E CCW BTYPE WEAK SMOO 30 SEC CON SPECT 0.5 4 210 30 >> >> BOUNDCOND SIDE S CCW BTYPE SOMM >> >> BOUNDCOND SIDE W CCW BTYPE SOMM >> >> SPON S 500 >> >> SPON W 500 >> >> FRIC MANN 0.025 >> >> BRE 0.6 0.3 >> >> NONHYD BOX PREC ILU >> >> DISCRET UPW NONE >> >> DISCRET CORRDEP NONE >> >> TIMEI METH EXPL 0.4 0.8 >> >> QUANT HSIG 'Hs' 'Hsignificante' 0 3 -99 dur 3 HR >> >> BLOCK 'COMPGRID' NOHEADER 'Hs.mat' XP YP DEP HS 1 >> >> POINTS 'CHECK' 505168 3908454 >> >> TABLE 'CHECK' HEAD 'TABLE_check1' TIME XP YP BOTLEV DEP WATL OUTPUT >> 160000 0.5 SEC >> >> TABLE 'CHECK' HEAD 'TABLE_check2' TIME XP YP VEL VMAG VDIR OUTPUT 160000 >> 0.5 SEC >> >> COMPUTE 160000 0.05 SEC 200000" >> >> I have tried to reduce the time step manually as well as modeling land >> area as both topography and exception. >> >> Any help would be much appreciated. >> >> Kind regards, >> Patricia >> _______________________________________________ >> swash-users mailing list >> swa...@li... >> https://lists.sourceforge.net/lists/listinfo/swash-users >> > _______________________________________________ > swash-users mailing list > swa...@li... > https://lists.sourceforge.net/lists/listinfo/swash-users > |
From: H. G. G. <go...@gm...> - 2022-02-01 21:05:52
|
Dear Patricia, I faced this error so many times. Previously, I have taken several notes regarding the potential reasons for this error. I copy these notes below for your information. I could not look into the details of your *.sws file. However, I noticed that some parts of your input file are related to the below-given items. Therefore, I hope that these notes might be helpful for your case. Best regards, Gokhan "INSTABLE: water level ..." error, potential reasons: 1. The wave generating boundaries in 2D simulations: > If the target wave is large compared to the water depths along the boundary. > Presence of land along the wave generating boundary. 2. Sponge layers might cause instability problems in 2D simulations. 3. Activating BREaking command might cause instabilities. 4. Solver settings: NONHYDrostatic BOX 0.5 SOLVer 0.01 0.0 1000 0.55 PREConditioner ILU For example, in this line, 0.5 corresponds to the Crank-Nicolson method. It's more accurate; however, it might cause stability problems, especially in 2D simulations. Using 1 corresponding to the Euler method may solve the problem. 5. Discretization schemes: Some schemes may cause wiggles in the free-surface solution (as indicated in the manual). These might accumulate and cause instability problems. My experience is using the default schemes in general or the following schemes for monochromatic waves. FROMM for horizontal advective terms BDF for vertical advective terms MUSCL for CORRDEP Patricia Mares Nasarre <pat...@gm...>, 27 Oca 2022 Per, 19:24 tarihinde şunu yazdı: > Hi, > Does anyone know why I get this error? > > "** Terminating error: INSTABLE: water level is too far below the bottom > level! > ** Message : Please reduce the time step!" > > It appears during the first seconds of the simulation of a 2D case. The > input file is: > > "PROJ 'name' 'O1' > > SET LEVEL -0.29 > > MODE NONST TWOD > > COORD CART > > CGRID REG 502400 3906100 0 17600 11900 879 594 > > VERT 3 30 40 30 PERC > > INPGRID BOT REG 502400 3906100 0 879 594 20 20 EXC -9999 > > READINP BOT 1 'file.bot' 1 0 FREE > > WIND 5.2 210 CON > > INIT zero > > BOUN SHAP JONSWAP 3.3 SIG PEAK DSPR DEGR > > BOUNDCOND SIDE N CCW BTYPE WEAK SMOO 30 SEC CON SPECT 0.5 4 210 30 > > BOUNDCOND SIDE E CCW BTYPE WEAK SMOO 30 SEC CON SPECT 0.5 4 210 30 > > BOUNDCOND SIDE S CCW BTYPE SOMM > > BOUNDCOND SIDE W CCW BTYPE SOMM > > SPON S 500 > > SPON W 500 > > FRIC MANN 0.025 > > BRE 0.6 0.3 > > NONHYD BOX PREC ILU > > DISCRET UPW NONE > > DISCRET CORRDEP NONE > > TIMEI METH EXPL 0.4 0.8 > > QUANT HSIG 'Hs' 'Hsignificante' 0 3 -99 dur 3 HR > > BLOCK 'COMPGRID' NOHEADER 'Hs.mat' XP YP DEP HS 1 > > POINTS 'CHECK' 505168 3908454 > > TABLE 'CHECK' HEAD 'TABLE_check1' TIME XP YP BOTLEV DEP WATL OUTPUT > 160000 0.5 SEC > > TABLE 'CHECK' HEAD 'TABLE_check2' TIME XP YP VEL VMAG VDIR OUTPUT 160000 > 0.5 SEC > > COMPUTE 160000 0.05 SEC 200000" > > I have tried to reduce the time step manually as well as modeling land > area as both topography and exception. > > Any help would be much appreciated. > > Kind regards, > Patricia > _______________________________________________ > swash-users mailing list > swa...@li... > https://lists.sourceforge.net/lists/listinfo/swash-users > |
From: Patricia M. N. <pat...@gm...> - 2022-01-27 16:23:41
|
Hi, Does anyone know why I get this error? "** Terminating error: INSTABLE: water level is too far below the bottom level! ** Message : Please reduce the time step!" It appears during the first seconds of the simulation of a 2D case. The input file is: "PROJ 'name' 'O1' SET LEVEL -0.29 MODE NONST TWOD COORD CART CGRID REG 502400 3906100 0 17600 11900 879 594 VERT 3 30 40 30 PERC INPGRID BOT REG 502400 3906100 0 879 594 20 20 EXC -9999 READINP BOT 1 'file.bot' 1 0 FREE WIND 5.2 210 CON INIT zero BOUN SHAP JONSWAP 3.3 SIG PEAK DSPR DEGR BOUNDCOND SIDE N CCW BTYPE WEAK SMOO 30 SEC CON SPECT 0.5 4 210 30 BOUNDCOND SIDE E CCW BTYPE WEAK SMOO 30 SEC CON SPECT 0.5 4 210 30 BOUNDCOND SIDE S CCW BTYPE SOMM BOUNDCOND SIDE W CCW BTYPE SOMM SPON S 500 SPON W 500 FRIC MANN 0.025 BRE 0.6 0.3 NONHYD BOX PREC ILU DISCRET UPW NONE DISCRET CORRDEP NONE TIMEI METH EXPL 0.4 0.8 QUANT HSIG 'Hs' 'Hsignificante' 0 3 -99 dur 3 HR BLOCK 'COMPGRID' NOHEADER 'Hs.mat' XP YP DEP HS 1 POINTS 'CHECK' 505168 3908454 TABLE 'CHECK' HEAD 'TABLE_check1' TIME XP YP BOTLEV DEP WATL OUTPUT 160000 0.5 SEC TABLE 'CHECK' HEAD 'TABLE_check2' TIME XP YP VEL VMAG VDIR OUTPUT 160000 0.5 SEC COMPUTE 160000 0.05 SEC 200000" I have tried to reduce the time step manually as well as modeling land area as both topography and exception. Any help would be much appreciated. Kind regards, Patricia |
From: Abushet S. <an...@gm...> - 2021-08-18 09:24:10
|
Dear SWASH users and support group As far as I know SWASH does not provide wave directions or directional parameters of waves. Does anyone help me estimate wave directions (and other directional parameters) over the whole computational domain from available SWASH output results? Any post-processing script would be greatly appreciated. Thank you in advance Abushet |
From: ANIKA M. <a6...@ua...> - 2021-06-30 20:12:00
|
Hello, I am a master's student and currently using SWASH to simulate wave overtopping during storms over a breakwater in a harbour. I have been testing different values for Manning friction of the breakwater, and am receiving warnings in the print file saying that the percentage of instable points for computing horizontal eddy viscosity is increasing, resulting in NaN output values for the overtopping discharge (DISCH command). This has happened changing manning from 0.05 to 0.049, one worked, the other did not. I am a new SWASH user and not sure how to solve this problem, has anyone had similar issues? Below my command file. Many thanks and best regards, Anika $ SET LEVEL 3.036 & MODE NONST ONED & $*********Model Input**************** $ GRID 0. 0. 0. 581 0. 1162 0 $ VERT 2 $ INPGRID BOTTOM 0. 0. 0. 1162 0 0.5 0. READINP BOTTOM 1. 'P9_bat_0.5_point12.bot' 1 0 FREE $ INPGRID FRIC 0. 0. 0. 1162 0 0.5 0. READINP FRIC 1. 'P9_friction_0.5_point12.dat' 1 0 FREE $ INIT ZERO $ BOU SHAP JON 3.3 SIG PEAK $ BOU SIDE W BTYPE WEAK CON SPECTRUM 4.9408 16.6892 SPONGE E 162 $ FRICTION MANNING $ BREAK $ NONHYD BOX PREC ILU DISCRET UPW MOM BOTCEL SHIFT TIMEI 0.1 0.5 $ $ $********Model Output********** $ $ POINTS 'GAUGE1' FILE 'wg1.wvg' TABLE 'GAUGE1' HEAD 'Point12_1_wg1.tbl' TSEC BOTL DISCH WATL OUTPUT 001858.000 0.05 SEC $ $ TEST 1 0 COMPUTE 000000.000 0.01 SEC 020626.000 STOP |
From: DEKNEUDT Q. <que...@ar...> - 2021-06-28 20:36:16
|
Dear all, I'm experiencing an issue related to a Swan/Swash coupling. I would like to use the outputs from SWAN as forcing for two Swash layout boundaries in input. I have checked, the outputs from SWAN are correctly defined to match the Swash inputs forcing. I can run Swash with one of the two boundaries forcing but when I'm using both, Swash doesn't want to run calculation despite no warnings or errors in the Print file. Did some of you experienced the same problem ? I am using Swash latest version 7.01 and Swan version 41.31 You'll find my print file attached and a screenshot of the message from Swash after the start of compilation. Best regards, [cid:image001.png@01D76C47.6CF125E0] [Description : Description : Logo_Artelia_signature] Quentin Dekneudt Stage ingénieur : Hydraulique Maritime ARTELIA EAMO 6 rue de Lorraine · 38130 Echirolles · France www.arteliagroup.com<http://www.arteliagroup.com/> |
From: DEKNEUDT Q. <que...@ar...> - 2021-06-28 17:39:17
|
My apologize, I forgot the attached file De : DEKNEUDT Quentin Envoyé : lundi 28 juin 2021 18:00 À : swa...@li... Objet : Swan/Swash coupling issue Dear all, I'm experiencing an issue related to a Swan/Swash coupling. I would like to use the outputs from SWAN as forcing for two Swash layout boundaries in input. I have checked, the outputs from SWAN are correctly defined to match the Swash inputs forcing. I can run Swash with one of the two boundaries forcing but when I'm using both, Swash doesn't want to run calculation despite no warnings or errors in the Print file. Did some of you experienced the same problem ? I am using Swash latest version 7.01 and Swan version 41.31 You'll find my print file attached and a screenshot of the message from Swash after the start of compilation. Best regards, [cid:image001.png@01D76C48.607F0060] [Description : Description : Logo_Artelia_signature] Quentin Dekneudt Stage ingénieur : Hydraulique Maritime ARTELIA EAMO 6 rue de Lorraine · 38130 Echirolles · France www.arteliagroup.com<http://www.arteliagroup.com/> |