TUFLOW May 2005 Release

This page lists any reported problems, status of fixes and/or any work arounds on the TUFLOW May 2005 release.

Known Problems and Workarounds in Release 2005-05-AN:

Z Line Misses a ZU or ZV Point

A Z Line may not modify the ZU or ZV points in a cell if the Z Line finishes exactly on the cell's side. This has been fixed for the next release. The problem is solved by slightly moving the end of the Z Line.

Selection of Z Line and HX Line Cells Differ

Cells selected by identical Z Lines and HX lines may still sometimes differ (even if using the various CC and HX Z Line options). This has been fixed for the next release with a new cell selection process that is used by both Z Lines and HX lines.

1D Channel Flow Appears to be Incorrect

In rare situations where a 1D channel is oscillating between sub and super-critical flow every half-timestep, the flow reported in the output maybe incorrect (the flow calculated within the computational scheme is correct). This has been fixed for the next release and does not effect the hydraulic computations.

1D Domain Messages Sometimes do Not Appear in MapInfo

1D domain messages to the _messages.mif file may sometimes not appear within a MapInfo Map Window. They can be viewed within a Browser Window. This has been fixed for the next release.

Storage Above Structure Obvert Value Not Applied Correctly

The Storage Above Structure Obvert == value is not applied correctly. This has been fixed for the next release.

Pit Channel Inlet Connected to 1D Network

If the inlet to a pit, which is for visual purposes automatically located by the Pit Channel Offset == value (default is 10m) to the north of the pit point lies exactly on another 1D node, it will snap/connect to that 1D node hydraulically. The workaround is to slightly move the pit or other 1D node, or to change the Pit Channel Offset == value.

Pit SX Cells Not Correctly Selected (Fixed in Build 2005-05-BC)

If the 2D grid is not on a north-south orientation, SX cells selected by pits maybe offset to the left of the pit. This has been fixed for the next release and in Builds 2005-05-BA and 2005-05-BC.

Poor Mass Error with Direct Rainfall and High Elevations

If the Zpt values are several hundred metres or higher, and the direct rainfall option is being used, large mass errors can result because of floating point inaccuracies. A double precision option has been included and is being tested. Alternatively, lowering the whole model to around sea level should solve the problem.

Poor Mass Error in Steep Shallow Flow Areas

In very steep areas with very shallow flow, high mass errors can occur if the cells are repetitively wetting and drying. Reducing the Cell Wet/Dry Depth and Cell Side Wet/Dry Depth values should solve this problem (a number of fine grid models are successfully using less than a mm, ie. down to 0.0002 and 0.0001m).

No PO in _TS.mif File for 2nd, 3rd, ... 2D Domains

2D PO output to the _TS.mif file only occurs for the first domain in multiple 2D domain models. This has been fixed for the next release.

Polygon not Selecting Correct 2D Cells

If a MapInfo polygon (region) has several vertices that lie exactly on a horizontal line that passes through the 2D cell centre, the selection of cells within a polygon for Read MI commands may not work correctly. Also, if the last two vertices in the polygon are duplicates, a similar problem may occur. Both of these (very rare!) situations have been fixed for the next release.

Simulation Crash with Write PO Online == ON

If using Write PO Online == ON and the Start Time Series Output is after the Start Time, the simulation crashes. This has been fixed for the next release. Resolve by setting the Start Time Series Output equal to the Start Time (which is the default).

1D Timestep Reset to Zero

If second line of the .ecf file is blank, the 1D start time is reset to zero and not correctly set to be the same as the tuflow .tcf start time. (Problem relates to providing backward compatibility for old .ecf fixed format files). Has been fixed for the next release. If the .tcf file start time is zero, this problem does not occur.

XP-RAFTS .tot and .loc 5g20 Formats (included in Interim Release 2005-08-AF)Builds 2005-08-AF onwards recognise this format.

SX Boundaries with a U Flag (fixed in Interim Release 2005-08-AF)

2D SX boundaries that utilise the U flag to simulate pumps do not work due to recoding introduced in 2005-01 builds relating to the different timesteps in different 1D and 2D domains.

SMS gives a Zero Node Error (fixed in Interim Release 2005-08-AF)

Any 1D WLLs that do not snap, or for 2 point WLLs, cross a 1D channel, are now reported as a WARNING (prior to Build 2004-11-AA these were reported as an ERROR and the simulation halted). These WLLs may cause zero node IDs in the SMS.2dm file causing an error in SMS. It is recommended to remove any of these WLLs - they are identified as a WARNING in the _messages.mif file.

Can't Write Zpts

When creating a new model following the steps in Chapter 6, the Timestep command will need to be included anywhere in the .tcf file before using the Write MI Zpts command in the .tgc file. This is a hangover from the new different timesteps in different domains feature and will be corrected in a future release