Fixed Defects

Issues	Short description of the source problem resolved in this version
Circuit	FX-23842	In previous versions of Flux 3D, the evaluated information of unduplicated composed non-meshed coils (volume and resistance) was incorrect when the domain had symmetries or periodicities. Moreover, Flux displayed the wrong evaluated resistance for coil conductor components linked to non-meshed coils when those coils were duplicated according to the number of periodicities or symmetries. Now the correct values of this evaluated information are always properly displayed.
Data Import/Export context	FX-24994	Fixes an issue when visualizing and exporting values of a data collection created with versions older than Flux 2022.1.
Physics	FX-23109	When coil conductor regions with detailed description of losses were present in the project and when their geometry was not correct, a wrong message was displayed for the user. The geometry of these regions is not correct when it is not obtained as if the faces of the input electric terminal would have been extruded all along the way of current of the winding until the faces of the output electric terminal. Faces must exist between straight parts, where the current density is uniform, and cylindrical parts, where the current density is cylindrical. Now the right message is displayed: "COIL_A : On each volume of this region, the number of points on the input faces must be equal to the number of points of the output faces and to the number of lines between the input and output faces. Please, modify the geometry of each volume of this region to resolve correctly this problem."
Physics	FX-23589	In Transient Magnetic applications, in case of solving scenarios driven by the angular position of the mechanical set, plots of computed iron losses were mistakenly displayed with TIME quantity as abscissa, instead of ANG_POS_ROTOR values. The issue is now fixed with Flux 2022.2: ANG_POS_ROTOR quantity is correctly employed and displayed.
Physics	FX-23882	In 2D Transient Magnetic projects having coil conductor regions with detailed geometrical description, while evaluating Joule losses with sensors on coil conductor components or region, the losses could be wrongly unbalanced between phases at very low frequencies whereas they should be. In the same context but also for Steady State AC application, the zero frequency resistance (analitically computed) did not take in account the number of strands in parallel defined in the region. The number of strands in parallel is now well taken in account. Now that these two issues have been fixed, it is recommended for the user to delete the results and run again the scenario.
Physics	FX-25082	In AC Steady-State Magnetic applications coupled with Transient Thermal, the "type of equivalent B(H) curve" field which is available for certain isotropic non-linear material B(H) laws was not taken into account during solving, whatever the user choice was (i.e., the value "curve not modified" was always considered by Flux). The issue has been fixed, thus leading to better quality results.
Solving	FX-24865	With a project of a rotating machine obtained with an imported geometry, the solving process was stopped at the second step and the following message was displayed: RG_E_SECTION_BORNES_1_REG_BOB => Previous position not found The project was solved with the 3D magnetostatic application. The issue is due to - the presence of two I/O parameters piloted by the scenario by multivalues, and the angular position of the rotor piloted by monovalue with an angle of 14 degrees. - the use of the new "Method with electric conduction resolution + uniform J norm" to compute the current density J in the regions of type "coil conductor region" because the geometry of the project has been imported and these regions have non-perfect geometry. The issue has been corrected.
Solving	FX-24935	In Flux 2022.1 and previous versions from 2019.1 it was not possible anymore to create an FTS file for transient initialization after solving a project with the AC Steady State magnetic coupled with Transient Thermal application because of a regression. This bug has been fixed in the Flux 2022.2 release.
Solving	FX-23963	For some formulations that needed the generation of second order mesh, the continue to solve feature was not working correctly. This issue is now fixed.
Solving	FX-11652	With old Flux Transient Magnetic projects using solid conductor regions and transient initialization with .FTS files (generated by an AC Steady State project), an error message appeared in the transient project while attempting to solve the project. The issue is fixed; nevertheless, to avoid this error, the user is asked to regenerate the .FTS file with Flux 2022.2 or further version.
Solving	FX-24292	With a test case of a motor with 3 conducting magnets embedded in a conducting sleeve in the rotor, the solving process stopped, and the following error message was displayed: Message: NVENVN => Wrong number of positions The 3 magnets and the sleeve are described by solid conductors. They are in contact with each other and there are insulating face regions between them. There is an anticyclic periodicity. A wrong formulation was affected to the volume region of one of the 3 magnets. The issue was due to two of the 3 magnet volume regions which are in contact with a plane of periodicity, but not with the other one. The issue has been corrected.
Solving	FX-23885	In a specific Flux 2D project containing a solid conductor described by a superconducting model of power law for J(E) property, the solving process took a very long time with 2020 version. With 2022 Flux version, after having set the default solving options which have changed since version 2020, the solving process stopped, and the following error message was displayed: Message: MRC4DJ => Incorrect parameter n: n < 2 The issue has been corrected and the project can now be successfully solved. Moreover, the solving process is much faster by setting the solving options to "Maximal factor method" for the relaxation of Newton-Raphson algorithm, and the looseness and strengthening of the threshold deactivated.

Circuit

FX-23842

In previous versions of Flux 3D, the evaluated information of unduplicated composed non-meshed coils (volume and resistance) was incorrect when the domain had symmetries or periodicities. Moreover, Flux displayed the wrong evaluated resistance for coil conductor components linked to non-meshed coils when those coils were duplicated according to the number of periodicities or symmetries. Now the correct values of this evaluated information are always properly displayed.

Data Import/Export context

FX-24994

Fixes an issue when visualizing and exporting values of a data collection created with versions older than Flux 2022.1.

Physics

FX-23109

When coil conductor regions with detailed description of losses were present in the project and when their geometry was not correct, a wrong message was displayed for the user.

The geometry of these regions is not correct when it is not obtained as if the faces of the input electric terminal would have been extruded all along the way of current of the winding until the faces of the output electric terminal. Faces must exist between straight parts, where the current density is uniform, and cylindrical parts, where the current density is cylindrical.

Now the right message is displayed:

"COIL_A : On each volume of this region, the number of points on the

input faces must be equal to the number of points of

the output faces and to the number of lines between

the input and output faces.

Please, modify the geometry of each volume of this region

to resolve correctly this problem."

Physics

FX-23589

In Transient Magnetic applications, in case of solving scenarios driven by the angular position of the mechanical set, plots of computed iron losses were mistakenly displayed with TIME quantity as abscissa, instead of ANG_POS_ROTOR values. The issue is now fixed with Flux 2022.2: ANG_POS_ROTOR quantity is correctly employed and displayed.

Physics

FX-23882

In 2D Transient Magnetic projects having coil conductor regions with detailed geometrical description, while evaluating Joule losses with sensors on coil conductor components or region, the losses could be wrongly unbalanced between phases at very low frequencies whereas they should be.

In the same context but also for Steady State AC application, the zero frequency resistance (analitically computed) did not take in account the number of strands in parallel defined in the region. The number of strands in parallel is now well taken in account.

Now that these two issues have been fixed, it is recommended for the user to delete the results and run again the scenario.

Physics

FX-25082

In AC Steady-State Magnetic applications coupled with Transient Thermal, the "type of equivalent B(H) curve" field which is available for certain isotropic non-linear material B(H) laws was not taken into account during solving, whatever the user choice was (i.e., the value "curve not modified" was always considered by Flux). The issue has been fixed, thus leading to better quality results.

Solving

FX-24865

With a project of a rotating machine obtained with an imported geometry, the solving process was stopped at the second step and the following message was displayed:

RG_E_SECTION_BORNES_1_REG_BOB => Previous position not found

The project was solved with the 3D magnetostatic application. The issue is due to

- the presence of two I/O parameters piloted by the scenario by multivalues, and the angular position of the rotor piloted by monovalue with an angle of 14 degrees.

- the use of the new "Method with electric conduction resolution + uniform J norm" to compute the current density J in the regions of type "coil conductor region" because the geometry of the project has been imported and these regions have non-perfect geometry.

The issue has been corrected.

Solving

FX-24935

In Flux 2022.1 and previous versions from 2019.1 it was not possible anymore to create an FTS file for transient initialization after solving a project with the AC Steady State magnetic coupled with Transient Thermal application because of a regression.

This bug has been fixed in the Flux 2022.2 release.

Solving

FX-23963

For some formulations that needed the generation of second order mesh, the continue to solve feature was not working correctly. This issue is now fixed.

Solving

FX-11652

With old Flux Transient Magnetic projects using solid conductor regions and transient initialization with .FTS files (generated by an AC Steady State project), an error message appeared in the transient project while attempting to solve the project. The issue is fixed; nevertheless, to avoid this error, the user is asked to regenerate the .FTS file with Flux 2022.2 or further version.

Solving

FX-24292

With a test case of a motor with 3 conducting magnets embedded in a conducting sleeve in the rotor, the solving process stopped, and the following error message was displayed:

Message: NVENVN => Wrong number of positions

The 3 magnets and the sleeve are described by solid conductors. They are in contact with each other and there are insulating face regions between them. There is an anticyclic periodicity. A wrong formulation was affected to the volume region of one of the 3 magnets. The issue was due to two of the 3 magnet volume regions which are in contact with a plane of periodicity, but not with the other one. The issue has been corrected.

Solving

FX-23885

In a specific Flux 2D project containing a solid conductor described by a superconducting model of power law for J(E) property, the solving process took a very long time with 2020 version. With 2022 Flux version, after having set the default solving options which have changed since version 2020, the solving process stopped, and the following error message was displayed:

Message: MRC4DJ => Incorrect parameter n: n < 2

The issue has been corrected and the project can now be successfully solved. Moreover, the solving process is much faster by setting the solving options to "Maximal factor method" for the relaxation of Newton-Raphson algorithm, and the looseness and strengthening of the threshold deactivated.