Difference between revisions of "TUFLOW Version Backward Compatibility"

From Tuflow
Jump to navigation Jump to search
Line 1: Line 1:
  
=Backaward Compatibility Change Register=
+
=Backward Compatibility Change Register=
 
{|class="wikitable"
 
{|class="wikitable"
  
Line 238: Line 238:
 
|
 
|
 
|}
 
|}
 +
 
=Frequently Asked Questions (FAQ)=
 
=Frequently Asked Questions (FAQ)=
 
== Why are model results developed in an older release different to a newer release? ==
 
== Why are model results developed in an older release different to a newer release? ==

Revision as of 12:05, 30 May 2022

Backward Compatibility Change Register

2017-09-XX Builds
2017-09-AA New SX boundaries defaults. Set Defaults == PRE 2017 if similar results are required to the 2016-03-AE release.
XF files are now being processed for boundaries. Set Defaults == PRE 2017 or use XF Files Boundaries == OFF if similar results are required to the 2016-03-AE release.
Regions in 2d_bc layers now applied as regions (previously only cell over region centroid selected). No backward compatible workaround provided.
Material IL and CL now applied to gridded rainfall (previously not applied). No backward compatible workaround provided.
SA regions now always select a 2D cell even if there are no cell centres falling within the region (previously a SA region would not select any cell if no cell centres fell within the region). No backward compatible workaround provided.
If using “Reveal 1D Nodes == ON”, “Time Series Output Interval ==” must be specified. No backward compatible workaround provided.
2016-03-XX Builds
2016-03-AA New operational 1D structure defaults. Set Defaults == PRE 2016 if similar results are required to the 2013-12-AC release.
Primary upstream and downstream 1D channels now correctly take into account bed slope. Set Defaults == PRE 2016 if similar results are required to the 2013-12-AC release.
A new layered 2D FC calculation method has been implemented. Set Layered FLC Default Approach == CUMULATE or use Defaults == PRE 2016 if similar results are required to the 2013-12-AC release.
The SX Flow Distribution Cutoff Depth has been raised to 0.005m from 0.0m. Set Defaults == PRE 2016 or use SX Flow Distribution Cutoff Depth == 0.0 if similar results are required to the 2013-12-AC release.
The End After Maximum tolerance has been increased to 0.001m from 0.0m. Set Defaults == PRE 2016 if similar results are required to the 2013-12-AC release.
A new ERROR message has been added to cross-check the 1D timestep is a multiple of the timestep for all 2D domains. Set Defaults == PRE 2016 if similar results are required to the 2013-12-AC release.
WARNING 2460 has been escalated to an ERROR and stricter command line syntax rules have been introduced ("=" will now return an ERROR message if TUFLOW is expecting "==") Set Defaults == PRE 2016 if similar results are required to the 2013-12-AC release.
2016-03-AD The treatment of the eddy viscosity term in the GPU Solver has been enhanced with slightly improved results in areas of rapidly changing velocity patterns. Set GPU Viscosity Method == Method A or use Defaults == PRE 2016 to achieve the same results as Build 2013-12-AC and prior.
2013-12-XX Builds
2013-12-AA New default settings – see Defaults == PRE 2013-12 in the user manual for a list of the commands that have changed in their default setting. Set Defaults == PRE 2013-12 if similar results are required to the 2012-05 release.
2013-12-AC The default setting for Link 2D2D Approach has changed. Set Link 2D2D Approach == METHOD B to achieve the same results as Builds 2013-12-AA and 2013-12-AB. See Link 2D2D Approach for more information.
2011-09-XX and 2012-05-XX Builds
2012-05-AA New default settings – see Defaults == PRE 2012-05 for a list of the commands that have changed in their default setting. Set Defaults == PRE 2012-05 if similar results are required to the 2011-09 or 2010-10 releases.
The approach to the sizing of automatic manholes and the application of losses has been enhanced. Set Manhole Approach == Method A to achieve the same results as Build 2011-09-AA.
2011-09-AA The optimised compiler code is treated differently for Single Precision builds producing slightly different results (fractions of a mm) for some models. No workaround.
2010-10-XX Builds
2010-10-AA New Intel Fortran Compiler version produces slightly different results (usually fractions of a mm). No workaround.
w32 and w64 versions will give slightly different results for the same simulation. No workaround. Use the same platform (w32 or w64) for all simulations. Use Model Platform to force which platform should be used.
New default settings – see Defaults == PRE 2010-10 for a list of the commands that have changed in their default setting. Set Defaults == PRE 2010-10 if similar results are required to the 2008-08 or 2009 07 releases.
Generation of TINs for polygons in Read GIS Shape layers is more robust and uses an improved approach. In rare cases, the TIN would fail and TUFLOW would abort the start-up. No workaround.
2008-08-XX and 2009-07-XX Builds
2008-08-AC The default setting for Shallow Depth Stability Factor has changed. Set Shallow Depth Stability Factor == 3 for models without direct rainfall to achieve the same results as Builds 2008-08-AA and 2008-08-AB. See Shallow Depth Stability Factor for more information.
2008-08-AA
Uses a new set of defaults for a number of commands (see Defaults ). The new defaults produce slightly different results, and very slight differences also occur between the three versions offered. For established models run using the 2007-07-XX builds, use Defaults == PRE 2008-08 to use the default settings used by the 2007-07-XX builds. Testing of a range of models has shown zero change in results if Defaults == PRE 2008-08 switch is set, and the Compaq Fortran compiled version (cSP) is used. Each of the new default settings and their effects are discussed in the rows below.
The method for interpolating n values where the 2D Manning’s n varies with depth has been enhanced from a linear interpolation of the M (1/n) value to a spline interpolation of the n value. See Bed Resistance Depth Interpolation. Generally has little effect other than when the flow is predominantly in the depth range that the n value is varying. The new approach offers a smoother transition in n values from one depth to the other.
The default viscosity coefficient is now a combination of a 0.2 Smagorinsky and 0.1 constant coefficient, and there are some enhancements to the application of the viscosity term. See Viscosity Coefficient. This has slight effect for the majority of models. For fine grid models (<2m cell size) with low bed resistance and significant variations in velocity vectors the effect is more pronounced but is still slight.
Inertia and viscosity terms are now not transferred across dry cell sides when constructing the coefficients for the solution arrays. This was having the effect of generating a circulation on the other side of the wall (albeit a very weak one), which of course shouldn’t happen! Generally little effect, but can have some minor influence for urban models where buildings and fences are modelled as solid thin Z lines.
1D weir flow has been improved as the water level difference across the weir approaches zero. The new method is more stable. See Weir Flow. Very little difference other than improved stability.
Incorporates minor improvements for transitioning between Regimes A and B, and between inlet and outlet controlled regimes, for circular culverts. Very little difference other than improved stability.
The new automatic selection of cells for 2D SX connections using the 1d_nwk Conn_1D_2D attribute may choose more than one 2D cell. Very little difference other than improved stability at the pit 2D connections.
2007-07-XX Builds
2007-07-AA
Uses a new set of defaults for a number of commands (see Defaults). The new defaults may produce slightly different results. For established models run using the 2006-06-XX builds, use Defaults == PRE 2007-07-AA to use the default settings used by the 2006-06-XX builds. Each of the new default settings and their affects are discussed in the rows below.
Change Zero Material Values to One == OFF (previously ON) Will not cause different results if a Set Mat == 1 is specified before other material settings in the .tgc file, or if every cell has been assigned a material value.
Inside Region == Method B (previously Method A) Testing thus far has not shown any difference between the two methods (other than the substantial gains in processing time of polygons).
Line Cell Selection == Method D (previously Method C) May change results slightly, but improved stability and a smoother water levels along HX lines result.
VG Z Adjustment == MAX ZC (previously ZC) May change results slightly, but stability should be significantly enhanced in some situations.
Bed Resistance Cell Sides == INTERROGATE (previously AVERAGE M) Will influence results, usually slightly, but more pronounced where there are sudden changes in Manning’s n values such as in the urban environment.
Culvert Flow == Method D (previously Method C)

Culvert Critical H/D == OFF (previously Culvert Critical H/D == 1.5)

The most significant influences are the selection of upstream or downstream controlled regimes depending on the H/D ratio, and the bug fix relating to Regime E if Structure Losses == ADJUST. Offers improved stability, better convergence for Regime C and smoother transitioning between some regimes.
Changed the setting of the default width (if eN1 < 0.001) of automatic weirs over R and C channels (i.e. RW and CW) to be the diameter/width multiplied by the number of culverts (previously, the width was not multiplied by the number of culverts). For backward compatibility, original weir width can be set by manually setting the eN1 attribute to the Diameter_or_Width attribute value of the culvert.
Bug fix that when using a restart file TUFLOW occasionally set the 2D FC bridge deck additional loss value incorrectly. No backward compatible workaround provided.
Bug fix that incorrectly set the water levels on dried VG cells (only applies to simulations with source inflows, e.g. SA or RF, somewhere within in the model). May cause slight changes in results. Backward compatibility provided if Defaults == PRE 2007-07-AA is set (noting that setting this command reinstates the bug). This bug also causes the mass error calculations to falsely give a mass error that is not occurring.
Fixed bug that did not correctly apply the reduction in conveyance for a FC BD (bridge deck) of FD (floating deck) cell using the 2d_fc Mannings_n attribute. Backward compatibility applied if Defaults == PRE 2007-07-AA is set, however, note that this reinstates the bug and the resistance to flow at FC BD and FD cells may need to be reviewed. Indications are that only minor changes in results occur. The flow area under 2D FC BD and FD cells is correctly calculated.
2006-06-XX Builds
2006-06-AA
Uses a new set of defaults for a number of commands. The new defaults will produce different results. For established models run using the 2005-05-XX builds, use Defaults == PRE 2006-06-AA to use the previous default settings. Each of the new default settings and their affects are discussed in the rows below.
Cell Wet/Dry Depth == 0.002 previously 0.05) and Cell Side Wet/Dry Depth == 0.001 (previously 0.03) The most pronounced effect of the shallower wet/dry depths is likely to occur in areas that are still filling at the flood peak, such as behind a levee that is only just overtopped. The shallower wet/dry depths provides a greater flow depth for a longer period over the levee.
Adjust Head at ESTRY Interface == OFF (previously ON) Usually does not have a major influence on results except where very high velocities occur.
Boundary Cell Selection == Method C (previously Method A) and Line Cell Selection == Method C (previously Method A) May select slightly different cells along boundary/link lines. This may cause a difference where the line is along the top of levee, possibly creating a “hole” in embankment.
Viscosity Formulation == Smagorinsky (previously Constant) and Viscosity Coefficient == 0.2 (previously 1.0) Can have a significant effect where the viscosity term is influential. This occurs where the friction term is less dominant (i.e. low Manning’s n and/or deeper water such as the lower, tidal, reaches of rivers).
Structure Losses == ADJUST (previously FIX) Can have a significant affect in the vicinity of structures within a 1D network and for culvert networks. Does not affect 1D structures linked to a 2D domain or at the structure ends not connected to another 1D channel.
Storage Above Structure Obvert (%) == 5 (previously CHANNEL WIDTH) Usually negligible effect unless the model storage is predominantly within 1D closed sections (i.e. B, C and R channels). The 1D domain is likely to be more sensitive to instabilities due to the much smaller storage above the top of the closed sections, therefore, a smaller 1D timestep may be required and/or the Storage Above Structure Obvert (%) increased.
Depth Limit Factor == 10 (previously 1) No effect as previously the model would have become “unstable” as the trigger for an instability was the top of the channel/node.
Culvert Flow == Method C (previously Method B) Usually only minor effects plus improved stability.
Culvert Add Dynamic Head == ON (previously OFF) Minor influence.
Bridge Flow == Method B (previously Method A) Negligible influence plus improved stability. However, note the different treatment of energy losses once the bridge deck obvert/soffit is submerged if a BG or LC table is specified.
WLL Approach == Method B (previously Method A) Only affects the presentation of results. Note, that Method A is no longer recommended or supported.
Apply All Inverts == ON (previously OFF)

Does not affect hydraulic calculations, however, if a Blank, B or W channel is now lowered/raised because the inverts are now used, this will affect results/stability - see note at end of Apply All Inverts).

Conveyance Calculation == ALL PARALLEL (previously CHANGE IN RESISTANCE) Will affect results as ALL PARALLEL can be around 10% more “slippery” than CHANGE IN RESISTANCE. For calibrated or established models developed using build prior to Build 2006-06-AA , recommend setting to CHANGE IN RESISTANCE
Flow Calculation == Method B (previously Method A) Negligible effect.
Builds prior to 2006-06-XX Contact support@tuflow.com

Frequently Asked Questions (FAQ)

Why are model results developed in an older release different to a newer release?

If comparing a Classic model with HPC, also check the Will TUFLOW HPC and TUFLOW Classic results match? page in addition to this answer.
In addition to the above, there are reasons why model results would be different between different TUFLOW releases, whether it is the Classic or HPC solver as follows:

  • General improvements and fine-tuning of the solution scheme, especially for the more complex hydraulic physical terms and situations such as: sub-grid turbulence representation; treatment of shocks (e.g. hydraulic jumps); and transitioning between sub-critical and super-critical flow on steep slopes.
  • Some new functionality can cause a significant change in results. For example:
    • Sub-Grid Sampling (SGS) applied to an existing model that used a too coarse cell resolution in high flow areas of highly variable topography (relative to the 2D cell size). SGS will greatly improve the model's ability to convey water accurately in these situations with vastly improved results.
    • New default sub-grid turbulence scheme in the 2020 release of TUFLOW HPC that is cell size independent and allows modellers to use cell sizes much smaller than the flow depth across all scales from flume to large rivers. For more information on differences between Smagorinsky scheme (HPC releases up to 2020) and the new Wu turbulence scheme (2020 onwards) see here.
  • Changes to the default settings and values, e.g.:
    • different default eddy viscosity formulation and/or coefficients,
    • improved data pre-processing approaches such as sampling materials on cell mid-sides instead of cell centres,
    • and many others.
    • For backward compatibility the Defaults == command is available to run old models on new releases to replicate past results (note, sometimes full backward compatibility cannot be catered for due to different code compiler and updates that can't be reverted, especially for several releases earlier).
  • New features that use GIS attributes previously reserved (i.e. unused). If these attributes were not populated with the recommended “reserved” value (usually 0 or blank), then they can cause unpredictable results in later releases.
  • Bug fixes noting that most bug fixes are input/output related and rarely affect the model's hydraulic calculations.
  • Change in timestepping can also produce a small change in results. HPC uses the Runge-Kutta 4th order integrator, which is usually fairly insensitive to time step provided the model is running stably. However when a region is filled by flow that only just overtops an embankment, a 10 mm difference in water levels upstream of the embankment can create a much larger difference in levels downstream. Hence small differences in time-stepping (along with many other aspects of model setup) can trigger local differences in model results.
  • Model orientation (if changed) could also mean slight change in results. This is mostly given by interpolating values from different calculation points. Every cell has nine calculation points. Based on the model origin, all or most of the calculation points would have different topography elevation sampled, which translates to slightly different results.
  • If using 1D channel, possibly different cells have been selected as HX boundary and might have different elevations. This can be reviewed in 1d_to_2d check file.

Generally, there should not be substantial differences as the fundamental equations being solved are unchanged and TUFLOW Classic and HPC solvers have always solved all the physical terms using a 2nd order spatial approach. The one exception is the turbulence (eddy viscosity) representation, which is the most complex and challenging to solve of all the physical terms (many 2D schemes simply omit this term). If significant differences (>10% of depth change across the whole model) are observed then it’s most likely due to the first four dot points above. To identify in which release(s) the significant changes occurred, the model can be run with the latest build and for past releases. The changes for each release are documented in their release notes. Past releases and release notes are all available here. Once the exact release where the changes occurred is tracked down, individual features can be turned off to narrow down the cause.

The recommendation is usually for new or reworked models to use the newest build to take advantage of the latest features and enhancements, some level of calibration might be required for reworked models. The new TUFLOW executable is not different from the previous ones in the meaning that any existing model should be re-calibrated if there are available calibration data. However, particularly if a model is already calibrated, using prior builds of TUFLOW or winding back default settings using Defaults == command is considered reasonable for established models that are to be used for minor tasks and an update of the model would not be cost effective.

Up
Go-up.png Back to TUFLOW Modelling Guidance