Improvements and Fixed Defects
Improvements
Issues | Short description of the source problem resolved in this version | |
---|---|---|
Solving | FX-8636 |
When the linear system solver selected in the solving process options box is “Solver automatically chosen”:
|
Solving | FX-12763 | When “Automatic” is selected in the advanced tab of the solving process options box for the initialization of state variables at the beginning of a time step, the prediction variable initialization, based on previous time steps, is now used by default for hysteresis projects in 2D and 3D. |
Solving | FX-10769 | In the solving process options box of Flux PEEC, there is a new FGMRes option when the compression solver is chosen, the residue can be relative or absolute. An absolute residue may lead to inaccurate results or long solving time and was the default option in the previous version of Flux PEEC. For any new projects using compression solver in Flux PEEC, the residue will be relative by default. |
Solving |
FX-12756 FX-13421 |
When the direct solver and the method without tree gauge for edge variables are selected in the solving process options box, the configuration of the MUMPS direct solver was changed to solve numerically deficient matrices (more unknowns than equations), which appears by using edge finite elements without tree gauge. |
FEMT | FX-14080 | A project is solved with FEMT,
but not postprocessed The user project contains some incoherences for an FEMT coupling. But FEMT say anything about the problem. Now with Flux 2020:
|
Display PyFlux Command | FX-12293 FX-13232 |
Before Flux 2020 the "Display PyFlux Command" didn't return the python command corresponding of the creation of a scenario. It was the same problem with the curve entities (the list of result values are returned, and mainly the user want the interval of the creation of the curve only). With Flux 2020, the "Display PyFlux Command" method return the constructor for these entities: solving scenario, 2D and 3D curves. The user can rebuilt easily a scenario or a curve like other entities.
Note: You can have the legacy behaviour if you use Flux in
advanced mode.
|
Iron losses | FX-14658 | A periodicity check is now performed in the evaluated field values to prevent potential errors in the computation of iron losses with Flux. This test is displayed in the result GUI box when an iron loss computation is performed. |
Import/Export context | FX-14049 FX-14182 FX-11394 |
The Import/Export context
dedicated to NVH simulations has been improved as follows:
|
Laminated regions | FX-12843 |
Laminated regions, which were already available in the 2D and 3D modules of Flux, are now also implemented in Flux Skew. As a consequence, new iron losses models such as Bertotti and Loss-Surface (LS) became available in the context of Flux Skew projects. The evaluation of iron losses in Flux Skew is also faster. |
Application change | FX-12272 | The application change algorithm
has been improved as follows :
|
Fixed Defects
Issues | Short description of the source problem resolved in this version | |
---|---|---|
Geometry Modeler extrusion |
FX-10440 |
In this extrusion from an imported sketch, the straight extrusion does not work, but if we try again it works without error. There is the same behavior in python if we replay the command. It is an ACIS issues. The problem is solved with the upgrade of the ACIS version from 2017.1.0.2 to 2019.1.0.1 |
Geometry CAD import |
FX-11262 |
Geometry import - Small
volumes / proximity detection - problem with the complete
infinite box. An update of ACIS (2017.1.0.2 to 2019.1.0.1) has been done in the Flux 2020 and CAD import seems to be better (heal, stitching, ...) - In this associated case "claw pole rotor", after the import, the complete infinite box is done without error message. |
Geometry CAD import |
FX-12620 |
Different behaviour of the
import between V11.2 and 2018.1 If we import the same geometry in V11.2 and in the modeler of the version 2018.0 and that we specified a coordinate system for the import, the location of the geometry is not the same, in V11.2 the geometry is shifted with the coordinate system specified, in 2018 it is not the case. Flux 2020 takes into account correctly the coordinate system in CAD import process with unit management. |
Geometry CAD import |
FX-12810 |
Position of the geometry
different after a same import in V11.2 and in V2018 The CAD import with Flux 2018 is correctly done but the position of the geometry is different to the result obtained with Flux 11.2. Flux 2020 takes into account correctly the coordinate system in CAD import process with unit management. |
Geometry CAD Import |
FX-13813 |
Geometry import (PARASOLID
from Simlab) + Infinity box : No connected domains With Flux 2020, during the Parasolid import from SimLab - a check and a convert of a closed sheet body in a solid body are done. |
Geometry Spline |
FX-12809 |
Impossible to open a project solved in V11.2
with Flux 2018.0 The project solved in V11.2 contains a complex geometry (complex because there are splines) imported in V11.2. The project solved is impossible to open with Flux 2018.0. If the project is not solved, it is possible to open it with Flux 2018.0. Flux 2020 allows spline lines modification when an associated geometric parameter defined a coordinate system field is modified. |
Geometry Arcs simplification |
FX-13874 |
Sketcher: There is some
problem of simplification of arcs don't built on the sketcher.
the result can be the opposite of the wished arc. The fixed in Flux 2020 is to reverse end points if necessary for non-sketcher arcs. |
Geometry Infinite box |
FX-12197 |
Impossible to complete the
infinite box of this 3D project The problem is solved with the upgrade of the ACIS version from 2017.1.0.2 to 2019.1.0.1 |
Geometry Infinite box |
FX-13269 |
Problems in infinity box
creation with imported, periodic geometry. This defect comes
from the offset on the X axis. Flux 2020 can complete infinite box with periodicity around X axis with an offset |
Mesh | FX-13555 | Wrong results using thin
conducting region. There is a problem of duplication node on
volume element at the intersection of region with one with the
type "thin conducting region". A correction has been implemented in Flux 2020 to fixed this specific configuration. |
Solving | FX-11824 | On a 3D project with shell elements in the mesh, the “delete results” command took more than one day. Now, it takes 22 seconds. |
Postprocessing | FX-12090 | The option "clear" in compute on
physical entity did not run correctly if there was only one
formula. With Flux 2020, it is now possible to compute on physical entity for only one formula or several formula and the result can be stored. |
Postprocessing Point computation |
FX-13572 | Flux crashed with point
computation following python command.
The computation point wished is not on the selected Volume region. The computation is not possible, but Flux shouldn't crash. Now Flux doesn't crash anymore and an user message is displayed on a dialog box. |
Postprocessing Grid 3D |
FX-13772 | Impossible to delete 3D grid in
Flux 3D or Flux Skew. Fixed in Flux 2020 |
User message | FX-13854 | misspelling in message after
importing material. Cleaned in Flux 2020 |
User message | FX-13476 |
In this project when we try
to delete the Infinite box when the geometry is meshed an
unclear message appear. The new message implemented is :A mesh exists : unable to delete the INFINITE_BOX. |
Labels | FX-12597 | Rename GeometryIncaView as GeometryFluxPEECView |
Labels | FX-13623 |
Problem of text in french on
the material manager whereas the selected language is
"English" |
Labels | FX-13706 | Problem of labels: replace selection by "surfacic" and "volumic" regions by selection by "face" and "volume" region. |
Flux-Activate coupling | FX-13968 and FX-13844 | Improved stability of the Flux-Activate coupling. |
Flux-Activate coupling | FX-14322 | Improved the documentation explaining how to couple Flux to Activate. |
Import/Export context | FX-13250 FX-13347 and FX-13344 | A previous bug correction provided in Flux HotFix 2019.1 obliged the user to recreate the computation supports used in the evaluation of forces with the Import/Export context. Recreating the computation supports is no longer necessary. |
Import/Export context | FX-11380 | The error message displayed by Flux when the user tries to create a data support from entities belonging to different mechanical sets was reformulated to become more intelligible. |
Import/Export context | FX-12165 | The creation of 3D coordinate systems is now allowed in the Import/Export context. |
Import/Export context | FX-12814 | Improvement of the data export methods used in the Import/Export context. The time required to export data from a selection of regions is now shorter. |
Import/Export context | FX-13340 | Corrected an issue that caused an incoherence between the numberings of evolutive curves and groups of calculations in the data Import/Export context. |
Import/Export context | FX-14023 | The creation of a surface forces collection is now only allowed on surface-type supports in the Import/Export context. Analogously, the creation of a volume forces collection is now allowed only on volume-type supports. |
Import/Export context | FX-14780 | Corrected a bug affecting the force computation in the Skew module. When forces are computed on rotor, the orientation of normal component was not well oriented. |
Iron losses | FX-13832 | The Python command used in FluxSkew to create a laminated magnetic region was modified to match the syntax of its Flux2D analog. |
Iron losses | FX-14336 | Iron losses computations using the “half-cycle” option are available now only in Advanced mode. |
Licensing | FX-14111 | Corrected an issue affecting the Linux version. Flux no longer crashes in the absence of a configured license. |
Materials | FX-13475 | Improved the convergence in the solving of 2D and 3D projects containing magnetic materials described by Preisach’s model. |
Materials | FX-14370 | Improved the consistency check performed by Flux during the creation of non-linear magnets described by their magnetic coercitivity (Hc) and remanent magnetization (Br) values. |
Sensors | FX-13166 | Improved the computation speed of the maximum magnetic flux density when using sensors in FluxSkew projects using the Steady State AC application. |
Sensors | FX-14447 | The error message displayed by Flux when a point-type sensor fails a “Check physics” test was reformulated to become more intelligible. |
Setup | FX-14556 | Fixed a bug that eliminated all Altair related shortcuts in Windows Start menu after an uninstall of Flux 2019.1.1. |
Thermal applications | FX-14577 | Improved the computation speed of the local radiation computation. |