Fixed Defects
| Issues | Short description of the source problem resolved in this version | |
|---|---|---|
| Circuit | FX-22898 | In Flux 2D, the cumulative manipulation of FE coupling components assigned to coil conductor regions with losses and detailed geometrical description could corrupt the project under certain circumstances. More specifically, after repeated deletions and reassignments of a FE coupling component to a coil conductor region of type Foil-wound coil, the Check Physics command would mistakenly advert the user about a non-existent circuit description error. At this point, trying to delete the coupled circuit to redescribe it led to successive errors that prevented the project resolution. This issue has been solved in Flux 2022.1. |
| Co-Simulation | FX-23468 | During a co-simulation with Activate, Flux is now automatically launched from the folder containing the Flux to Activate coupling component. This new strategy avoids undesired behavior that may take place when the working directory indicated in Activate's built-in file browser is set to a protected folder. |
| Data Import/Export context | FX-22951 | In Transient Magnetic projects with scenarios controlled by the position of the mechanical set, while creating a Derived Data collection of type "FFT of a time collection" from a previously defined Force Data collection in the Import/Export context, Flux mistakenly triggered an error when the interval employed for the Force Data collection did not correspond to the whole interval solved in the scenario. It is worth mentioning that the FFT Data collection was nevertheless correctly computed on a restricted interval when the Force Data collection was defined on the whole scenario interval. In any case, this issue has been solved in Flux 2022.1. |
| Examples | FX-23163 | In the skew example "Induction motor with skewed rotor", the step by step document is not the good.
The corresponding step by step document has been added. |
| Examples | FX-22734 | The 2D example "Brushless IPM motor (embedded magnets)" contains wrong value for the peak torque ans the max speed in the step by step document.
The right values have been added in the document. |
| Flux e-Machine Toolbox | FX-22548 | In previous FeMt versions, maps extrapolations may leads to incorrect values near the maps borders.
In version 2022.1, maps computations have been modified to fix this kind of problem. |
| Macros | FX-22216 | Incoherent Circuit obtained by the macro circuit_collector_2D. Some terminals are not correctly connected when you open the circuit editor. The macro has been updated to correct this issue. |
| Material Manager | FX-23160 | In the Material Identification tool, the coefficients of the Modified Bertotti model were named Ke (classical losses coefficient), Kh (hysteretic losses coefficient) and Ka (excess losses coefficient) and appeared in that order from the top to the bottom of the graphic interface. On the other hand, in Flux the same coefficients are named differently and appear in a different order. More specifically, in Flux the dialog box used to describe the Modifed Bertotti model displays, from top to bottom, coefficients K1 (hysteresis losses coefficient), K2 (classical losses coefficient) and K3 (excess losses coefficient). Thus, the Material Identification tool was modified to match both the order and the terminology established in Flux and to avoid any possible confusion. Moreover, the pyFlux command generated by the Material Identification tool after the identification of the Modified Bertotti model of a material had the values of coefficients K1 and K2 inverted, leading to a wrong description of the model in Flux. This issue has also been solved in version 2022.1. |
| Physics | FX-22835 | When a coil conductor volume region consists
of several parts, and when these latter are in contact with the plane of
symmetry of type "Tangent magnetic fields, normal electric field", they are in
contact with the parts of the region in 3 places or more. Indeed, for the
moment, a coil conductor volume region must be in one part, and must be in
contact with none, one or two places with the symmetry planes of type "Tangent
magnetic fields, normal electric field". If not, a message is displayed at the
beginning of the solving process or when choosing the Check physics command.
There were two issues: 1. There was an unclear message. For instance, what follows could be displayed: Volume region PHASE_A : (Then, two lines with any characters) 2. After building the project with a PyFlux file, saving the project, closing it, then opening it again, the message was not displayed. The bugs are now corrected, and this message is displayed if necessary: The volume region PHASE_A is in contact with the planes of symmetry of type "Tangent magnetic fields, normal electric field" in at least 3 places. It must be in contact with these planes of symmetry in none, one or 2 places. Please review the geometry or physical aspects of the project. |
| Physics | FX-18407 | The problem came from, on a device from a mesh import, the check of solid conductors when closing the circuit editor.
In fact, when evaluating the terminals of solid conductors, if there was more than one type of faces (ListOfFacetPlane, ListOfFacetPoly,...), only one type was kept on terminals. Now with Flux 2022.1, the check is well handled and all terminal faces are persistent. |
| Physics | FX-23003 | Some inconsistencies were detected between the use of one composed coils with multiple segments and multiple coils with one segment to describe the same coil. Now, these two cases are giving the same results. |
| Postprocessing | FX-22731 | Some minor issues have been fixed in the ElectroStatic applications: they impacted the computation in parallelized mode of the Streamer Criterion when the Flux project contained a geometry coming from CAD import. The functionality is now more robust. |
| Postprocessing | FX-20044 | While using magnets with a material model accounting for demagnetization, the BrDemag spatial quantity isovalues - intended to show the magnetization state of the magnet - always displayed the results of the last scenario step instead of showing a different result at each step. The issue is now fixed and the isovalues are automatically updated: they show the demagnetization state of the magnet according to the selected scenario step. |
| Solving | FX-23619 | There was an error due to rotating mechanical set while solving in a particular case.
Two defects have been fixed: - Fix error on unsorted faces list at boundaries of mechanical sets. - No creation of Data Support when geometric parameters exist in scenario. |
| Solving | FX-11970 | There was a bug after the solving process, with the following message written in the file Flux3D.report:
RG_E_SECTION_BORNES_1_REG_BOB => No EVAL_COND_FIL_I found The project on which the bug occurred consists of 2 volume regions of type "coil conductor region" with the same geometry and is solved with the 3D transient magnetic application. Each region has a racetrack shape. The "method for the computation of J in coil conductor regions in 3D" (Solving process options > Edit > Advanced tab) is "Automatically specified method". For each of the two regions, the horizontal surfaces which cross the two vertical straight legs have a difference of 9.5 %. Thus, the method chosen by the software at the beginning of the solving process is "Method with electric conduction resolution (with finite element method) + uniform J norm". This bug is now fixed. |
| Solving | FX-22650 | An issue while exporting FTS file in 2D Magneto Static application has been fixed, in order that such FTS file can now be used to initialize Transient Magnetic simulations. |
| Solving | FX-23670 | A regression related to the transient initialization by FTS files has been detected in Flux 2021.2 as a side effect of correction FX-14127 (see Flux 2021.2 release note for additional details). In some cases using imported FTS files, quantities / unknowns on coils were not correctly initialized for Transient Magnetic solving, thus leading to abnormal oscillations / instabilities before outputs reached the expected steady state.
This issue is now fixed with Flux 2022.1; nevertheless, users are strongly advised to regenerate their FTS files with Flux's latest version before using them to initialize Transient Magnetic computations. |
| Solving | FX-23074 | Fix the issue where the order of the optimization constraint in the problem affected the optimization process. |
| Solving | FX-22772 | On some transient solving, the initialization strategy selected by the user were not correctly applied to all the variables. This issue is now fixed. |
| Solving | FX-22035 | Fix the use of second-order meshes in the free-shape optimization using the finite differences method which could have led to singular jacobian matrices. |
| Solving | FX-23474 | The free-shape optimization was limited to 9,999 nodes in the design variables (nodes moving). This limitation is now at 9,223,372,036,854,775,807 nodes (64-bits integer limit). |
| Solving | FX-23359 | In some use cases of the free-shape optimization, a multi-objective optimization was computed instead of a mono-objective optimization. In this version, all mono-objective problem is resolved as such now. |