To use AFFTAC, you must be an authorized user. To receive authorization, provide the following information in an e-mail to Todd Treichel (firstname.lastname@example.org) at the AAR Tank Car Safety and Research Project:
Below are the links for downloading the AFFTAC package and for the documents associated with its release.
In the table below, "GUI" = "Graphical User Interface" and "CM" = "Computational Module".
|Release||Critical Bug Fixes||Upgrades, Minor Bugs, and Convenience Fixes|
|Corrected GUI-CM miscommuniation regarding the PRV elbow pressure value.||
(1) Fixed GUI bug that caused PRV settings to be forgotten sometimes.
(2) Fixed "Edit Table" in GUI button labels in the Ladings setup window.
(3) Added internal check in GUI to ensure user has a tank wall material selected.
(4) Increased numerical accuracy in two-phase flow model.
(5) Improved the logic and error message in the CM regarding solution concentrations.
|Corrected an error in the comptuation of the location of the liquid surface, which only affected the generalized TPS model.||
(1) Fixed a bug in the CM that would allow the fraction filled to become > 1, causing the code to crash under certain conditions.
(2) Corrected headings in the output, which in some cases would be missing in some columns.
(3) Corrected GUI-CM interface error regarding new PRV database.
Corrected critical bugs in the generalized TPS model's interface to the main part of the model,
(1) A bug that caused incorrect heat input calculations into the liquid,
(2) A bug that caused the wrong temperature to be reported for the wall temperature in the vapor space,
(3) A bug that used the wrong surface area in the generalized TPS model in the vapor space.
(1) Cleaned up DOS output of the CM.
(2) Added fatal error message for infinite slope in PRV open-close specifications.
(3) Corrected a CM bug in the vapor-region calculations for the generalized TPS model.
Corrected a critical bug in the generalized TPS model that, once fixed, enabled several
verification tests to render good results.
||(1) Cleaned up DOS output of the CM.
(2) Added fatal error message for infinite slope in PRV open-close specifications.
(3) Corrected a bug in the vapor-region calculations in the CM for the generalized TPS model.
Corrected a bug in the interaction between the generalized TPS model and the Larson-Miller creep and failure model.
(1) Made creep output log based.
(2) Modified the two-phase flow model calibration capability.
(3) Added internal data structure to support PRV calibration capability.
(4) Added a new "burst strength" indicator (an internal data structure) to support PRV calibration capability.
(5) Widened the display of lists in the generalized TPS editing window.
(6) Added usability enhancements to the valve calibration window.
(7) Updated values in PRV.db.
|Beta 06 Patch 01
(1) Corrected Lading properties editing window bugs that would lead to GUI crashes.
(2) Added usability and clarification enhancements to the Ladings editing interfaces.
(3) Corrected a non-critical GUI bug in the tank car material selection pull-down box.
(4) Made the computational module output in hundreths of minutes instead of tenths of minutes.
(5) Adjusted method of computing the temperature used for looking up thermal conductivity in the generalized TPS model.
|Transition Note||Starting at Beta 06, each released is accompanied by a significantly expanded Verification & Validation test document.|
Fixed a bug in the generalized TPS model that affects materials with significant changes in thermal
conductivity. The bug caused the lowest value of conductivity to be chosen regardless of temperature.
(1) Added "Analyze Confidence" capability.
(2) Made usability enhancements to window sizes and locations.
(3) Fixed buggy "Cancel" button in TPS setup window.
|Beta 07 Patch 01
(1) Corrected units to display "square feet" in the valve section of the Edit PRD window.
(2) Made the use of PRV area consistent in the CM.
|Transition Note||Starting at Beta 08, each release is optimized for fast computing performance.|
(1) Fixed a bug that was executed when the padding gas was used and a shell-full condition
existed that caused the tank wall temperature to not be synchronized with the lading
temperature as it should have been.
(2) Transitioned to releasing CM compiled for optimal speed.
|Beta 08 Patch 01
|Fixed a bug that could negatively affect simulations that only use the Strength model's UTS failure test and no others. An incorrect failure test in the new UTS model would result in non-conservative failure times. The combined used of other failure models would have greatly reduced or eliminated the impact of this bug.||None|
(1) Fixed a bug that prevented the GUI from recording changes to solute concentrations
in the Ladings database.
(2) Fixed a bug that prevented information about critical temperatures being recorded.
(3) Implemented the Analysis Import-Export capability.
(4) Expanded the CM's AFFTAC.plt capability, which allows for more detailed plotting using a third-party plotter.
|Beta 09 Patch 01
Corrected errors in the concentration values in the solutions of Ladings.db
(1) Upgraded the GUI so that it prints documentation at the top of the Ladings.db file to guide
users who want to modify the Ladings.db file by hand.
(2) Maintenance upgrade to the computational module, with no change in functionality.
|Transition Note||FORMAL RELEASE OF 4.00 FOLLOWS|
| Beta 10
(1) There was a potentially non-conservative bug in the legacy TPS model's temperature-dependant
thermal conductivity model. The units inside the Computational Module were not converted
correctly. This bug was fixed for the formal release.
(2) The polynomial coefficients for lading properties as a function of temperature were not being converted to the proper units. This will only affect users who have set up lading properties using polynomials; the tabular lookup capability was not affected.
(1) Minor inconstencies regarding the use of AFFTAC's Usum value were fixed. These resulted
in small changes to AFFTAC's predictions on its regression tests.
(2) A bug was fixed in the legacy TPS model where, if a tank liner was selected by the user and that liner did not deteriorate with time, it was left out of the calculations. Also, the GUI was not properly capturing deterioration time when entered by the user; that bug was also fixed.
(3) The part of the AFFTAC GUI used to display property data, either for strength models or ladings did not always display the variable name or units properly along the vertical axis. That was fixed for the formal release.
(4) There was an inconsistency that arises when the user chooses to use the Strength Database. In that situation, it was not definite which material properties would be chosen for the tank wall's thermal expansion coefficient and thermal conductivity. That was fixed for version 4.00; the values default to those of carbon steel and that fact was documented in the User's Manual.
(5) The angular dependence model within the Generalized TPS model does not automatically rotate if the user selects a tilted configuration. Rather, the user must tilt the angular dependence manually. This is an inconvenience that should, perhaps, be addressed in the next AFFTAC upgrade. For now, that fact was captured in the User's Manual.
(6) A memory leak that was exercsied when the user chooses to model a property using more than 20 points was addressed.
(7) The Computational Module did not always handle the printout of values for the final time step in a constent way. That was addressed for version 4.00.
(8) A bug was fixed in the Generalized TPS model that was due to the use of uninitialized values.
(9) The Computational Module's computation for the surface area of the tank wall touching the liquid phase of the lading was slightly off and was made right for version 4.00.
(10) A bug in the PRV open-close model was fixed. This long-standing bug was only exercised in situations where the PRV starts to close before it was fully open.
(11) Logic was added that makes AFFTAC consistent in the way it limits padding gas pressure between the use of valves and vents.
(12) The label for emissivity in the window where the user specifies the tank car geometry was made more clear.
(13) A minor bug was fixed wherein a transfer calculation for the Generalized TPS model included the legacy's spray-on coverage factor.
(14) A minor bug was fixed in the part of the Computational Module that checks the limits of data provided for the lading properties.
(15) A bug in the legacy TPS model was fixed; it was not actually causing the insulation to deteriorate but instead started the simulation with deteriorated insulation.
(16) The USum value was not being handled consistently by the GUI; this bug was fixed and might have minor impact on users' results.
(17) There were some long-standing bugs in the part of the Computationl Module that uses polynomials to model the tank wall strength decrease due to temperature rise. This was a long-standing bug in the legacy strength model, but its fix resulted in little to no changes in the regression test suite results.
| Beta 11
(1) There was a non-conservative error in the initial temperature of newly exposed tank wall that was fixed.
(2) The "ASTM A516 Non-Normalized" strength model entry appeared twice in the Strength Model database. The first entry was always used by the GUI whenever a model by that name was selected. But the first entry was incorrect and non-conservative. It was removed.
(1) Minor problems in the AFFTAC GUI were fixed having to do with finding files while in debug mode.
(2) The value of pi in the GUI was not accurate enough and was improved to be compatible with the value in the computational module.
(1) The Compressibility Factor and the ratio of specific heats, which have no units, were displayed with
blank parentheses. The parentheses now have "No units" inside them.
(2) The error handlers for the output file reader in two of the GUI plotter windows were not properly handling the end-of-file.
(3) The upper-right output plot always displayed the Larson-Miller result, even with the UTS model was active. Now, it displays the minimum of the two.
(4) There were two errors in the PRV database for valves labeled FRA lab test - 81 STD 32000 SCFM, FRA lab test - 280 STD 32000 SCFM, and FRA lab test - 294 STD 32000 SCFM.