Superconducting electric machines (SEMs) have the potential to be commercially available in the coming years. This commercialization depends on the availability of high-temperature superconductors (HTS) produced on a large scale. HTSs have high current densities and low losses, making them the leading technology choice for future light and compact high-power-density superconducting rotating machines, with a particular niche for high torque at low frequency. The advantages of SEM in its fully superconducting design or hybrid configuration (conventional stator, superconducting rotor) inherit from the characteristics of the superconductor material. So, they can show greater efficiency at a higher power density and lighter frame than their conventional counterparts for an equivalent power rating. Applications like electric aircraft, naval propulsion, and wind turbines, among others, are likely to use SEMs if the rated power has to be increased beyond what is technically available with conventional technology. In this context, this paper reviews SEMs and their applications. However, it also aims to highlight the main the literature projects with a minimal Technology Readiness Level (TRL) larger than three. Due to the diversity of the superconductors’ characteristics and the variety of machines, the modes of operation of SEMs can be quite distinct from conventional machines. Taking into account such diversity, SEMs are presented and sorted out by their operational principles and the choice of superconducting material. Finally, the future perspectives of SEM are discussed.

Numerical models based on the finite-element method are popular tools for investigating the macroscopic electromagnetic behavior of high-temperature superconductor (HTS) applications. This article explains how to use the $T$-$A$ formulation in COMSOL Multiphysics for modeling HTS coils in 2-D with different coupling scenarios between the turns. First we consider a racetrack coil wound from one piece of superconducting tape. Then, we consider a coil obtained by winding a cable composed of different HTS tapes. In the latter case, the tape turns are either electrically connected along their entire length or just at the two ends of the coil: in the model, these two different types of electrical connection are implemented with the help of the electrical circuit module. The current density distributions and the ac losses of the coils in the different coupling scenarios are compared and discussed. The limits of applicability of the presented approach are pointed out. The model is developed for the straight section of racetrack coils, but can be easily adapted to axisymmetric geometries.

In this paper, we investigate a topology of superconducting electric machine, which uses stacks of second-generation (2G) superconducting tapes. These stacks have been used as an alternative to replace bulks in trapped field superconducting machines. In this context, we built prototypes of superconducting machines with 2G tapes spirally disposed on the rotor. These prototypes were tested in liquid nitrogen bath at 77 K for operation in two regimes: synchronous and asynchronous. The operation mode is according to the load torque. To perform the torque tests under these operating conditions, a measurement system was also built. Quasi-static torque, locked rotor torque with stator frequencies up to 60 Hz, and dynamic torque up to 1200 rpm were measured. The prototype with 2G high temperature superconductor (HTS) stacks presented a torque, very close to nominal value of this machine with permanent magnets in the rotor. Promising results have been found showing that 2G HTS may be an alternative to superconducting bulks.

High-temperature superconducting coils are used in various large-scale applications, like rotating machines and high-field magnets. However, modeling these coils is a complicated and time-consuming process, especially due to the non-linearity of the current–voltage characteristics of the superconductors and the complex multiphysics involved. In this work, we used a fast homogenized method to model the coupled electromagnetic and electrothermal properties of racetrack and pancake coils for different applications. For this purpose, various formulations wielding homogenization methods are used and benchmarked with each other, as well as with models considering the detailed structure of the HTS tapes. We observe a very good agreement between different models (homogenized and detailed), and we discuss the pros and cons of the inclusion of insulating layers between the turns in homogenization. This work was performed under the collaboration of the COST action modeling teams and can be used as a review of the state-of-the-art superconductor modeling techniques, and a source for the development and benchmark of future numerical methods.

This work consists on the first finite element analysis of quasi-static test of a superconducting machine with stacks of coated conductors arranged spirally around the rotor. The behavior of the machine as a generator is also investigated. Hybrid A-V-H 2D formulation and moving mesh were used as modelling framework. The stacks are represented with the homogenization technique. Quasi-static test is simulated considering a field cooling initial state and the generator simulation, a pulse field magnetization. Current density and magnetic flux density along the stack surface as well as speed and torque results are analyzed. When the machine is not moving, major current density loops are induced in the stack, following the magnetic poles. These loops are deformed by the current density induced by speed differences between stator and rotor. The quasi-static torque does not change with position. The simulation results prove the potential this topology has of operating as a generator. The quasi-static results are in agreement with measurements made with the prototype.

This article proposes a new approach using the T-A formulation coupled to lumped parameters in the electrical circuit, considering the skin-depth approach. The finite-element method models the superconducting layer, the electrical circuit models the stabilizer and substrate layers as lumped parameters, and a coupling engine links both of them. The ac losses have been calculated by both T-A 1-D and T-A 2-D formulations: the first one uses the approach proposed in this article, and the second, two current vector potential components. For the three formulations, we have studied the average power dissipation as a function of frequency and current, and the peak current as a function of the frequency. We have compared the T-A 1-D and T-A 2-D formulations to H formulation results previously studied. Comparing the results of the average power dissipation, we have found the maximum error to be around 10.5% in T-A 1-D and 35% found in T-A 2-D. The efficiency of the proposed technique was evaluated as 16 times less costly when compared to the T-A 2-D formulation.