Evaluating iron losses à posteriori in magnetic cores

Introduction

The evaluation of iron losses in electrical steel sheets may be performed in Flux in an accurate and simple manner with the help of empirical à posteriori approaches while in post-processing. These methods greatly simplify the estimation of iron losses, since they do not require representing hysteretic phenomena during the physical description of a project and neither accounting for them during resolution. The results are a more straightforward project description, shorter computation times and in some cases the capability to evaluate iron losses after Steady State AC and Magneto Static solutions, in addition to Transient Magnetic application.

In this chapter, the basic workflow for evaluating iron losses in magnetic cores is presented, together with an outline of the two empirical à posteriori methods currently implemented in Flux, as well as the description of all the computation possibilities offered to the users.

Basic workflow

The essential steps for evaluating iron losses with à posteriori techniques in Flux are listed and discussed below:

In a project with a magnetic application, create one or more materials with a non-hysteretic magnetic B(H) property to represent the electrical steels used in the device.
Create one or more Laminated magnetic non conducting regions in the project and assign them to the surfaces (2D) or volumes (3D) representing the magnetic cores that are formed by stacks of electrical steel sheets.
Assign the previously created materials to the appropriate laminated magnetic non conducting regions.
Complete the physical description of the project, mesh it and solve a scenario.
After resolution and while in post-processing, launch the Computation of iron losses command from the Computation → Computation of iron losses menu.

The Computation of iron losses command gives access to the two à posteriori methods for evaluating iron losses in Laminated magnetic non conducting regions available in Flux, namely:

the modified Bertotti model and
the Loss Surface model.

In Flux, the Computation of iron losses command provides the user with the following three possibilities:

On regions: this choice enables to compute modified Bertotti or LS iron losses on a selection of laminated regions. The iron losses model can be defined on the material or in this dialog box via the Model for losses field. The computation provides instantaneous and average losses on the region(s) over the selected time interval for a set of fixed geometric and I/O parameters. This computation is more intended for in-depth analysis on a region.
Multi-parametric on regions: this choice enables to compute modified Bertotti or LS iron losses on a selection of laminated regions. The iron losses model can be defined on the material or in this dialog box via the Model for losses field. The computation provides instantaneous losses on the region(s) over the selected time interval for several sets of geometric and I/O parameters. This computation is more intended for optimizations or for efficiency maps.
On point with LS model defined in the material: this option, which is only accessible in Transient Magnetic application, enables to compute LS iron loss volume densities on a specific point of a laminated region. Iron losses model must be defined in the material, by means of an LS predefined sheets or by a user-defined model (see : LS model identification with MILS). The computation provides instantaneous and average loss volume densities over the selected time interval for a set of fixed geometric and I/O parameters. This computation allows displaying the hysteresis cycle of the material.

Note: To ensure very fast computations of iron losses, Flux needs to store - during the solving process - some data in .FLU folder and in memory. This means that in case of scenarios with many parameter values (e.g., some thousands), the size of the .FLU folder and the required memory could grow tremendously, thus limiting the performance of Flux for users who are not interested in very fast iron losses computations. For such situations, the option Data storage during solving exists in the Advanced tab of the Solving process options dialog box accessible via the Solving menu: it allows the user to disable the storage of data during solving and then to reduce the size of the .FLU folder and the required memory. On the other hand, the computations of iron losses slow down tremendously.

Note: In case of scenarios with geometric parameters, the data storage for very fast iron losses computations is not allowed by Flux. Consequently, the option described in the previous note has no effects.