Tutorial Module02 Archive: Difference between revisions

Content deleted Content added
Created page with "<ol> =Introduction= A purely 2D model was developed during the Module 1. You may have noticed, when reviewing the results, that some road embankments were causing a damming e..."
 
No edit summary
 
(11 intermediate revisions by 4 users not shown)
Line 1:
<ol>
=Introduction=
A purely 2D model was developed during the Module 1. You may have noticed, when reviewing the results, that some road embankments were causing a damming effect across the main channel. In reality there are culverts beneath these embankments. We will add 1D culverts to the 2D model created in [[Tutorial Module01Module01_Archive|Module 1]] in this Module.<br>
The main steps in the process are outlined in the Table of Contents at the top of the page.<br>
The only new file type that is introduced in this module is the 1d_nwk GIS layer. This layer defines the 1D or quasi-2D (branched 1D) domain network of flowpaths (channels) and storage areas (nodes).<br>
Line 7:
=Define 1D Structures=
There are three culverts that will be modelled in this module. The locations of these structures has been provided in the '''Module_Date\Module_02\GIS\''' folder. To define the 1D network for TUFLOW input, please select your GIS package from the list below:
*[[Tute_M02_MI_1d_nwkTute_M02_MI_1d_nwk_Archive | MapInfo]]
*[[Tute_M02_Arc_1d_nwkTute_M02_Arc_1d_nwk_Archive | ArcGIS]]
*[[Tute_M02_QGIS_1d_nwkTute_M02_QGIS_1d_nwk_Archive | QGIS]]
*[[Tute_M02_SMS_1d_nwkTute_M02_SMS_1d_nwk_Archive | SMS]]
 
=Define 1D/2D Linkages=
We now need to specify how the 1D culverts is linked to the 2D domain. For this tutorial we will use two different methods to highlight the two main approaches to creating the linkage.<br>
The setup of the link is done in your GIS package, please select your chosen GIS package from the list below:<br>
*[[Tute_M02_MI_1d2d_LinkTute_M02_MI_1d2d_Link_Archive | MapInfo]]
*[[Tute_M02_Arc_1d2d_LinkTute_M02_Arc_1d2d_Link_Archive | ArcGIS]]
*[[Tute_M02_QGIS_1d2d_LinkTute_M02_QGIS_1d2d_Link_Archive | QGIS]]
*[[Tute_M02_SMS_1d2d_LinkTute_M02_SMS_1d2d_Link_Archive | SMS]]
 
=Simulation Control Files Updates=
Line 35:
</ol>
 
For <u>SMS</u> users, you do not need to modify these files manually, it is automatically done when you export TUFLOW files from SMS. Follow the instructions on page "[[Tute_M02_SMS_SimulationTute_M02_SMS_Simulation_Archive|Export TUFLOW Simulation]]"<br>
<br>
 
Line 72:
 
<br>
'''<u>SMS Users</u>''', you do not need to modify these files manually, it is automatically done when you export TUFLOW files from SMS. Follow the instructions on page "[[Tute_M02_SMS_SimulationTute_M02_SMS_Simulation_Archive|Export TUFLOW Simulation]]"<br>
 
=Run the Simulation=
Using your preferred method for starting TUFLOW, run '''M02_5m_001.tcf'''. Please refer to <u>[[Tutorial_Module01Tutorial_Module01_Archive#Running_TUFLOWRun_Simulation | moduleModule 1]]</u> for a detailed description of the various methods for running a TUFLOW simulation. Example batch files have been included in the complete model download for reference. <br>
 
=Error 2050=
Line 84:
The reason for this is that the "SX" type boundary is a source boundary for the 2D cells (the source comes from the 1D model), the 1D model has a water level boundary applied from the 2D cells. Even when the 2D cells are dry, the level is above the 1D channel invert and mass could be created. TUFLOW does not allow this to happen and Error 2050 is reported.<br>
To locate the area that where this is occurring, we will load the messages layer into our GIS package, please pick from the list below:<br>
* [[M02_MI_Error_2050M02_MI_Error_2050_Archive | MapInfo]]
* [[M02_Arc_Error_2050M02_Arc_Error_2050_Archive | ArcGIS]]
* [[M02_QGIS_Error_2050M02_QGIS_Error_2050_Archive | QGIS]]
* [[M02_SMS_Error_2050M02_SMS_Error_2050_Archive | SMS]]
 
To correct the incorrect downstream invert level for culvert "FC04.1_C" in your GIS package set the "DS_invert" attribute to 42.75 and save the layer. For MapInfo users, don't forget to re-export the layer to .mif format.</li><br>
Line 99:
This will take a few minutes for the simulation to run. As mentioned in Module 1, now would be a good time to fill in your TUFLOW modelling log ([[TUFLOW_Modelling_Log | TUFLOW Modelling Log]]), have a stretch or get a cup of coffee....<br>
 
If the model fails to start correctly please refer to the <u>[[Tutorial_Module02Tutorial_Module02_Archive#Troubleshooting |troubleshooting section]]</u> at the end of this page.
 
=Check Files=
Line 106:
* In the 1D block: <font color="blue"><tt>Write Check Files </tt></font> <font color="red"><tt>==</tt></font><tt> ..\check\1d\</tt><br>
For a model such as the tutorial model which runs in a few minutes, the check files are less important as any issues should show in our results. However, larger models may take hours to run and it is good practice to review the check files so time isn't wasted. The check files are output in GIS format, please choose your package from the following list:<br>
* [[M02_MI_Check_FilesM02_MI_Check_Files_Archive | MapInfo]]
* [[M02_Arc_Check_FilesM02_Arc_Check_Files_Archive | ArcGIS]]
* [[M02_QGIS_Check_FilesM02_QGIS_Check_Files_Archive | QGIS]]
* [[M02_SMS_Check_FilesM02_SMS_Check_Files_Archive | SMS]]
 
=Review the Results=
==2D Results==
The 2D model results can be viewed using the same methods described in <u>[[Tutorial_Module01Tutorial_Module01_Archive#Viewing_the_ResultsViewing_Results | Module 1]]</u>. You should notice the flow now passes under the road embankments through the 1D culverts. However, at the peak of the flood event the flow in the creek exceeds the capacity of the culverts and two of the three road embankments are over-topped.
 
==1D Results==
Line 119:
===GIS Format===
To view the results in your GIS please select the package from the list below:<br>
* [[M02_MI_1D_ResultsM02_MI_1D_Results_Archive | MapInfo]]
* [[M02_Arc_1D_ResultsM02_Arc_1D_Results_Archive | ArcGIS]]
* [[M02_QGIS_1D_ResultsM02_QGIS_1D_Results_Archive | QGIS]]
* [[M02_SMS_1D_ResultsM02_SMS_1D_Results_Archive | SMS]]
 
===CSV Format===
Line 144:
We have added 1D culvert channels to represent the culverts underneath the road embankments. The 1D elements have been dynamically linked to the 2D model using three different methods.<br>
The model was simulated and the 1D results reviewed using a combination of GIS software and Excel.<br>
In the next module of the tutorial we will be modifying the 2D topography to incorporate surveyed breaklines. This can be found on the [[Tutorial_Module03Tutorial_Module03_Archive | Module 3]] page.<br>
In Module 4 we will model the creek as 1D channels, linked to the 2D model. This module has not yet been updated to the online wiki format, the previous version (MapInfo only) can be found [http://www.tuflow.com/Tutorial%20Model.aspx?Model2007 on the TUFLOW website].
 
=Troubleshooting=
This section contains links to some possible issue that may occur when progressing through the second tutorial module. If you experience an issue that is not detailed please detail the issue on the <u>[[Talk%3ATutorial_Module01 | discussion page.]]</u><br>
 
If you are unsure of why the model has failed to start, check the TUFLOW Log File (this is in the '''TUFLOW\runs\log\''' folder and has the name '''M02_5m_001.tlf'''. This can be opened in a text editor, the error is generally at the end of the file, however you can search for "Error" if you can not see the error.
*[[Save_Date_ErrorSave_Date_Error_Archive|Save date of .tab file is later than .mif or .mid]]<br>
*[[Unsnapped_SX_ErrorUnsnapped_SX_Error_Archive |Error 2024 - Could not find a 1D node snapped to CN line]]<br>
*[[Unsnapped_SX_ErrorUnsnapped_SX_Error_Archive |Error 2028 - Could not find a CN object or 1D node snapped to 2D SX object.]]<br>
*[[Only_TUFLOW_DOS_WindowOnly_TUFLOW_Console_Window_Archive | Why does the TUFLOW DOSConsole Window open, though the TCF isn’t read and the simulation doesn’t execute?]]<br>
 
= Advanced - HPC Solver (Optional) =
In this optional section we will run the model using the TUFLOW HPC (Heavily Parallelised Compute) solver. TUFLOWPlease HPCsee can[[HPC_Features_Archive run| betweenHPC 10Features andArchive]] 100for timesmore fasterinformation thanon TUFLOW ClassicHPC usingfeatures NVidiasupported Graphics Processing Units (GPU)(depending onin the model configuration and hardware performance). We will review relevant log messages and2017 resultsrelease.
 
TUFLOW HPC can run between 10 and 100 times faster than TUFLOW Classic using NVidia Graphics Processing Units (GPU)(depending on the model configuration and hardware performance). We will review relevant log messages and results.
 
==Model Setup==
Line 171 ⟶ 173:
 
==Results==
Using the methods described in the <u>[[Tutorial_Module02Tutorial_Module02_Archive#Viewing_the_ResultsReview_the_Results | ViewingReview the Results]]</u> section above:
* Check the simulation logs in the DOSconsole window, .tlf and .hpc.tlf log files. What proportion of the compute effort was spent on the 0D (initialisation, input / output communication between CPU and GPU, etc.), 1D modelling (CPU) and the 2D modelling (GPU)? Example compute percentage output is highlight below from the DOSconsole window. <br>
[[File:Tute M02 HPC 001.png|frame|none]]
* View the results in your preferred package. How do they compare to the TUFLOW Classic results?