A new formulation of the Preisach-Mayergoyz model was developed to radically improve its numerical performance. This new computationally efficient vector hysteresis model was incorporated in a 3-D finite element software. Many improvements to the transient nonlinear solver were made in order to obtain a good convergence of the TEAM problem 32 while using only 125 time steps per cycle. The results of this simulation show good agreement with the experimental data.
Today MgB2 superconducting wires can be manufactured in long lengths at low cost, which makes this material a good candidate for large scale applications. However, because of its relatively low critical temperature (less than 40 K), it is necessary to operate MgB2 devices in a liquid or gaseous helium environment. In this context, losses in the cryogenic environment must be rigorously minimized, otherwise the use of a superconductor is not worthy. An accurate estimation of the losses at the design stage is therefore mandatory in order to allow determining the device architecture that minimizes the losses. In this paper, we present a complete a 3D finite element model of a 36-filament MgB2 wire based on the architecture of the Italian manufacturer Colombus. In order for the model to be as accurate as possible, we made a substantial effort to characterize all constitutive materials of the wire, namely the E–J characteristics of the MgB2 filaments and the electric and magnetic properties (B−H curves) of nickel and monel, which are the two major non-superconducting components of the wire. All properties were characterized as a function of temperature and magnetic field. Limitations of the characterization and of the model are discussed, in particular the difficulty to extract the maximum relative permeability of nickel and monel from the experimental data, as well as the lack of a thin conductive layer model in the 3D finite element method, which prevents us from taking into account the resistive barriers around the MgB2 filaments in the matrix. Two examples of numerical simulations are provided to illustrate the capabilities of the model in its current state.
Purpose
Surface heat treatment by induction heating (10-100 kHz) requires precise prediction and control of the depth of the induced phase transformation. This paper aims at identifying common issues with the measurement and modeling of magnetic properties used in induction heating simulations, and it proposes ways to improve the situation.
Design/methodology/approach
In particular, it is demonstrated how intrinsic magnetic properties (i.e. the B-H curve) of a sample can change during the magnetic characterization process itself, due to involuntary annealing of the sample. Then, for a B-H curve that is supposed perfectly known, a comparison is performed between multiple models, each one representing the magnetic properties of steel in time-harmonic (TH) finite element method simulations. Finally, a new model called “power-equivalent model” is proposed. This model provides the best possible accuracy for a known nonlinear and hysteretic B-H curve used in TH simulations.
Findings
By carefully following the guidelines identified in this paper, reduction of errors in the range of 5-10 per cent can be achieved, both at the experimental and modeling levels. The new “power-equivalent model” proposed is also expected to be more generic than existing models.
Originality/value
This paper highlights common pitfalls in the measurement and modeling of magnetic properties, and suggests ways to improve the situation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.