- Running a simulation on Windows in parallel across more than one machine could fail when reading the materials database.
- The simulation would fail if the fan model was enabled.
- The k-epsilon model did not work when gravity and porous models were selected.
- Removed checking for a Java SDK in the
/opt directory during startup on HP machines. This was new in 2.04 and was only done if an SDK could not be found in the default location.
- Fuse boundary would fail if the boundaries were planar and aligned with one of the XYZ axes.
- File import was not reliably journaled as objects could be in a random order each import making playback unreliable if the macro referred to object names.
- Installation instructions incorrectly referred to "j2sdk" when that has changed to "jdk" in Java 1.5.
- Copy/paste of certain objects could fail when the target object already existed and had non-SI units.
- Copy/paste of mesh settings from one mesh continua to another could ignore base size settings and generate incorrect cell sizes.
- Could not import any file formats on Itanium based machines using the GUI. It is only IGES, STEP and Parasolid transmit files that cannot be imported on Itanium based platforms.
- Two boundaries would not be fused if they were planar and also aligned with one of the XYZ axes.
- An error would be encountered if iterating a simulation that included a moving mesh and live visualization scenes were also displayed.
- Two development-only properties in an interface condition node were hidden.
- Executing the action could crash the code if a particular (unlikely) sequence of steps was followed.
- Plots could not renamed.
- The startup script did not support the
-arch command line argument when looking for valid JDKs.
- The pressure could be incorrect on periodic interfaces with rotational transformation.
- There could be anomalous results on walls at partition boundaries when using the k-epsilon high y-plus wall treatment.