SMS:TUFLOW Model Parameters

From XMS Wiki
Revision as of 15:08, 19 June 2008 by Jhworth (talk | contribs) (→‎Water Level)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

Run control parameters are divided into two sections for TUFLOW simulations. The main model control dialog and the model control parameters that apply specifically to the 1D portions of the TUFLOW run. Right-click on the simulation and choose the appropriate menu command to edit either set of model parameters.

This document will try to highlight frequently used model control parameters. Please see the TUFLOW documentation for a description of those items left out or for more information on specific items.

Model Control

The main model control dialog sets parameters for the run in general and some 2D specific controls.

Time

This page is used to set the run times for the simulation. The optimum timestep size is determined by the courant number. The rule of thumb is that the timestep should be about half as many seconds as the grid cell size in meters. So a grid with 10 meter cells, should run with a 5 second timestep. The timestep may be reduced to increase stability, but generally problems indicate another type of problem.

Output Control

  • Interval - Identifies how many seconds between output times.
  • Minimums/Maximums - TUFLOW can compute minimum and maximum values or only the maximum values for the simulation. This can be very handy in visualizing flood extents.
  • Zero negative depths - By default, TUFLOW zeros depths that are computed as negative values at cell locations. If you turn this off, and setup the display options in SMS to only show positive depths you can see partially wet cells.
  • Display interval - By default TUFLOW outputs to the screen information for each timestep. It is often useful to see statistics less often so the text doesn't scroll so fast.
  • Data - This is used to specify the datasets for TUFLOW to include in the solution files.

Wetting/Drying

See the FESWMS manual for information on these options. Generally the default options are fine.

Restart Files

Restart files can be used to start a simulation part way through a run. They are generally used to generate an initial condition for several simulations that start with the same boundary conditions but are later changed for different events. For example to evaluate 10, 20, and 100 year events in a tidal area it might be useful to run a couple tidal cycles of a normal event and save these to a restart file. The restart file could be used with the 10, 20, and 100 year events without rerunning the first couple tidal cycles.

Water Level

  • Initial Water Level - All models need to have an initial water level. Any cells with a lower elevation than the initial water level will start out wet and the rest will start dry. A TUFLOW model does not require a fully wet domain when starting like FESWMS or RMA2. It is generally a good idea for the initial water level to match the downstream boundary condition at the start of a run. The global initial water level can be overridden on a local level using a 2D Spatial Attributes Coverage.
  • Override default instability level - By default TUFLOW will assume that if the water rises 10 m above the highest elevation in the grid then the run is unstable. This value may be overridden to allow the water level to get higher than this without causing TUFLOW to abort.

Eddy Viscosity

Eddy viscosity is used to compute energy losses due to turbulence not modeled at the scale of the simulation. TUFLOW supports specification of a constant eddy viscosity or using Smagorinsky specification. When using the Smagorinsky specification, the coefficient is generally between 0.06 and 1.0. It is recommended to use the default eddy viscosity of 1.0 when using constant eddy viscosities.

BC

In this tab, the boundary condition event to use is specified.

Materials

The material set to use is specified in this tab.

Misc

  • Write Check Files - TUFLOW creates a number of check files that can be used to verify model inputs. TUFLOW will run faster if not writing check files.
  • Check inside grid error setting - By default TUFLOW will generate an error of entities exist outside the grid domain. If you want to use data that extends beyond your grid, you can change this to warning or off.
  • Read External File - The SMS interface does not support every option available in TUFLOW. Unsupported options can be used by creating a TUFLOW command file (see the TUFLOW documentation) and specifying the filename here.

1D Control

General

  • Timestep - The optimum 1D timestep is based upon the courant number. The 1D timestep can generally be made quite small without affecting run times.
  • Output Interval - How often to write 1D solution data to the csv solution files.
  • Initial Water Level - The initial water level should generally be the same as the 2D water level.
  • Write Check Files - If this is off TUFLOW will not write check files for the 1D data.
  • Read external estry control file - The SMS interface does not support all of the options available for 1D controls. You can build an estry control file (see the TUFLOW documentation for instructions) and have it read at the end of the SMS generated files.

Network

  • Depth Limit Factor - This determines the depth that TUFLOW will consider unstable and abort the run. The default value is 10, meaning that if the channel depth is 10 times larger than the depth of the channel the simulation will be considered unstable.
  • Conveyance calculation - For 1D cross-section based channels, the conveyance calculations are divided into subsections. The subsections can be divided anywhere there is a change in resistance or a channel can be formed around each channel point(All parallel option). If the all parallel option is not used it is possible for the conveyance in a section to decrease when the depth increases which gives TUFLOW and error.
  • Minimum nodal area - Nodes with a small storage area can have stability problems. Adding a minimum nodal area for all nodes has the potential to make the run more stable but may attenuate the model solution.