Search

This paper analyses the current distribution in no-insulation (NI) coils in a layer-wound configuration. The investigation aims at comparing the electrical characteristic of two coils wound from the same BSCCO tape, with or without electrical insulation between turns. Both coils are characterized by a very similar geometry, with the same number of turns and layers. Both coils, cooled in liquid nitrogen bath, are charged until the tape critical current is exceeded. To interpret the measurements and analyze the current distribution within the winding, a lumped-parameter equivalent electrical circuit is developed and solved numerically in a time-varying regime. The model results are compared with the signals acquired through voltage taps soldered at the same locations in both coils. Finally, the model is applied to investigate the impact of the transverse contact resistance and to estimate the most stressed locations of the coil during charging, in terms of power produced by Joule effect.

The measurement of transport current ac losses in high-temperature superconductor tapes with the electric method has been extensively discussed in the literature. It is well known that the configuration of the circuit used to acquire the voltage signal can affect the measured losses, potentially leading to inaccurate results. To reduce this undesired effect, the rectangular circuit arrangement, which is widely adopted, involves twisting the pair of voltage taps at a given distance from the tape middle axis. However, the explanations reported in the literature to identify the correct distance are not fully exhaustive, and an alternative interpretation is described here. Moreover, given that the conventional voltage taps arrangement creates a significant circuit area to which electromagnetic noise can be linked, an alternative configuration is proposed, aiming to reduce the linked flux. This article presents a theoretical analysis to quantify the impact of the measurement circuit configuration on the losses. To ensure clarity, the model equations involved are derived step by step. The numerical results are applied to a thorough investigation on the accuracy of the ac losses measured on a sample tape, at different frequencies and current amplitudes. Both the conventional and the alternative arrangements are studied, varying their main geometrical parameters, at the same operating conditions. The correction terms are then applied to the measurement results, highlighting the advantages and drawbacks of the different configurations.

This article presents a numerical model based on the definition of the scalar and vector potentials to calculate the ac losses generated in a high-temperature superconductor coated conductor tape by an ac transport current and by an ac external magnetic field applied with a generic orientation. The material characteristics of the different layers of a reference tape are included in the model equations, which allow analyzing the differences between a model including only the superconducting layer or all layers of the coated conductor. The model is applied to study the dependence of ac losses on the amplitude of the transport current and magnetic field (applied individually or combined) and their frequency. The numerical results are compared with analytical formulations, with a second numerical model based on the H -formulation, and with experimental results. The conditions at which the contribution of the nonsuperconducting layers to the total ac losses of the tape become relevant are analyzed. In particular, the dependence of losses on the frequency of the ac external magnetic field applied at different orientations, is investigated. The current density distribution computed on the conductor cross-section is used to explain the trend of the calculated losses.

The paper presents a numerical study on electrodynamic behavior of the inter-pancake twin-box joints of the Poloidal Field (PF) coils of the ITER Magnet System. This work was carried out with the THELMA code, developed at the University of Bologna. The THELMA model was adapted to describe both the superconducting wires and the resistive saddle of the joint in a set of reference configurations representative of a set of experimental tests performed on a joint sample at the SULTAN facility in Villigen, Switzerland. The model was applied to analyze both the DC joint resistance measurements and the AC loss vs frequency curve, in order to obtain a full representation of the joint electromagnetic behavior. The paper reports a comparison between the numerical and experimental results and parametric studies on the impact of the main model parameters on the loss vs frequency curve.

The ITER Central Solenoid (CS) will be manufactured by assembling a stack of six modules, which are under fabrication by the US ITER organization and its subcontractors. The tests of the first CS Module have been performed at the premises of the General Atomics (GA) facility in Poway (US), in order to check compliance to the ITER requirements. Among other tests, the magnet was submitted to exponential dumps of the transport current from different initial values (10, 15, 20, 22.5, 25, 35, 40 kA) down to 0 kA. These tests are aimed at conducting DC breaker commissioning of the test facility and were used to measure the AC losses in the coil during electrodynamic transients. This paper presents the results of these measurements, along with a comparison with analytical computations of the losses in the magnet.

The paper reports the results of the test and modelling activities carried out on large commercial supercapacitor cells at high current. Four commercials cells, with rated capacitance of 3000 F and rated voltage of 2.7 V, are considered. All cells are submitted to consecutive charge/discharge cycles at constant current. A test current in the range 80-130 A is used, which is comparable with the maximum operating current (up to 150 A) that the cells can reach in practical operations. A circuit model of the cells, able to reproduce the most relevant dynamic behavior, with a good compromise between accuracy, simplicity and robustness of the model's parameters, is also developed and validated against the experimental data. It is shown that all the investigated commercial cells are characterized by similar phenomena and comply with the same circuit model. Moreover, it is shown that the circuit parameters of the cells are in the same range and are weakly dependent on the test current. The obtained circuit parameters are accurately reported for all the cells and made fully accessible to users.

The volume integral equation method based on the current vector potential approximated by means edge element basis function is a well-established approach for 3-D eddy currents computation. The application of the method is straightforward when simply connected geometries and no connection with external circuits are involved. In this case, in fact, the solving system is easily obtained based on tree–cotree decomposition of the primal graph. However, when multiply connected geometries or external generators are considered, “additional” degrees of freedom, not related to interior cotree edges of the primal graph, need to be identified and involved to assure the consistency of the numerical solution. In this article, the link between the volume integral equation method and the circuit is investigated in detail, and the circuit view is used as a guide for systematically finding the additional degrees of freedom arising in case of multiply connected geometries and/or external circuits. In particular, the dual graph is introduced as the support of the circuit and it is shown that this is the natural frame for taking the topology into account. For multiply connected geometries, the additional degrees of freedom are related to loop currents crossing one only time (or an odd number of times) any cutting surface making the domain simply connected, and are found by applying a minimum path algorithm on the dual graph forming the circuit. In case of conducting domain connected to an external generator, an extended dual graph is introduced for finding the further additional degrees of freedom. This article also researches into the possibility to replace the usual current density of the elements, obtained via facet-element shape functions and exactly matching the current of the faces, with a uniform current density obtained by means of a minimum error procedure and approximately matching the current of the faces. The use of this uniform current density, besides improving the calculation time and the accuracy of the coupling coefficients, also allows the extension of the volume integral equation method to discretizations of the problem domain made of arbitrarily shaped polyhedral elements.

This work extends to second generation Rare-Earth Barium-Copper-Oxide ((Re)BCO) tapes an experimental procedure previously developed to analyze the impact of double bending at room temperature on the performance of Bismuth-Strontium-Calcium-Copper-Oxide (BSCCO) tapes. The modified procedure is applied to measure the critical current of a commercial (Re)BCO tape subjected to bending around a cylindrical mandrel first on one side, then on the other side, followed by the cooldown to cryogenic temperature. In the bending phase, mandrels of decreasing diameter are used to identify the minimum curvature leading to a significant reduction of the tape critical current. Furthermore, a novel finite element model is developed to complement the experimental results. The model simulates the double bending at room temperature, the following straightening of the sample, and its cooldown to cryogenic conditions. The coupled thermo-mechanical numerical model together with the temperature-dependent mechanical properties allow investigating the combination of thermal contraction effects and bending loads in the whole domain of the problem. The experimental and numerical results obtained help to give a better insight in the distribution of the strain and stress components inside the (Re)BCO tape and to evaluate their impact on the conductor electrical performance in relevant operating conditions.

The cable-in-conduit conductors (CICCs) that will be adopted for the ITER central solenoid (CS) coil of the ITER machine have been extensively characterized in the SULTAN facility in Villigen, Switzerland, by means of DC and AC tests. The AC measurements were performed superimposing a sinusoidal magnetic field, with the amplitude of 0.2-0.3 T and frequency in the range of 0.1-5 Hz, to a background constant magnetic flux density of 2 T and 9 T. This paper describes the analysis of the AC loss SULTAN tests of one CICC, identical to that used for the manufacturing of the CS Insert tested in 2015 in the CSMC facility in Naka, Japan. The numerical analysis of these experiments was performed by means of the THELMA code, by developing three models at different levels of discretization (sub-cables of different cabling stages and strands). The comparison between measured and computed losses allows one to estimate the interstrand contact conductances of the tested sample. The model provides a useful insight on the time and space distribution of the power dissipated in the conductor.

The levitation force between a single grain YBCO cylindrical bulk superconductor and a permanent magnet is investigated. The force measured both in zero field cooling and field cooling condition in the temperature range 20 K - 80 K is reported. The dependence of peak levitation force and power dissipation on the temperature is analyzed and the saturation of the peak force with increasing the critical current of the material is pointed out. A numerical analysis is also carried out for understanding the electrodynamics of the levitation mechanism. Good reproduction of the measured data is obtained and relevant observed phenomena, such as hysteresis of the levitation cycle at high temperature and saturation of the peak force at low temperature, are explained by means of the model. Characterization of the material in terms of critical current density in the field range 0 T - 0.5 T is also obtained by means of the numerical model by means of the fitting of the measured data.

The possibility of REBCO Roebel cables to carry a large amount of current in high external magnetic fields makes them suitable for application in future particle accelerator magnets. The physical phenomena related to the initiation and propagation of a quench event in Roebel cables have not yet been fully analyzed. In particular, given the peculiar architecture of these conductors, the features of the current redistribution and normal zone propagation after quench in individual tapes still require experimental and theoretical investigation. In this paper, an innovative quasi-3D electrothermal FEM model is developed to analyze the effect of quench in HTS Roebel cables. The model is based on a reduced dimensionality approach, under the assumption of a negligible thickness of the individual REBCO tapes. While the tapes are meshed in an identical 1-D pattern, the noncontinuous electrical and thermal contacts among them are accounted for as a sources in the corresponding thermal and electrical equations. Notwithstanding these significant assumptions, the main physical aspects involved in the quench initiation and propagation are retained and correctly described by the reduced dimensionality approach. The quench energy and the normal zone propagation velocity are analyzed for different cable configurations.

Roebel cables assembled with high temperature superconducting tapes are a promising technology for several AC and DC applications. Their mechanical flexibility and compact design, combined with the capability of REBCO tapes to carry high transport currents in intense magnetic fields make them suitable for the application both in power devices and high field magnets. In this paper, an electro-thermal finite-element model developed at the University of Bologna (Italy) is described. The model allows computing the current and heat redistribution between the strands of the cable and from turn to turn inside a winding through non-uniform distributed thermal and electrical resistances between strands. The tape is 'homogenized' so as to create an anisotropic continuum model through a previously developed homogenization technique. The model is validated by comparison with quench tests performed on a well instrumented 7-turn pancake coil wound with a 2 m long Roebel cable composed of 15 REBCO tapes. The experiments were performed at the University of Southampton (UK) in the frame of the R&D activities of the EuCARD-2 project. The quench decision time, the temperature and electric potential evolution, the current and heat redistribution between strands in the event of a quench are analysed and discussed in the present study.

The National High Magnetic Field Laboratory (NHMFL), Tallahassee, FL, USA, has developed a 32-T all-superconducting user magnet system combining two series-connected high-field high-temperature superconductor (HTS) nested inner coils (insert) wound with SuperPower REBCO tapes and a low-temperature superconducting (LTS) outer magnet (outsert) composed of five coils (broken in 17 electrical sections). As a part of the magnet development work, a number of quench tests were done on the HTS insert dual-coil prototype in the actual LTS outsert, mimicking the actual magnet configuration and operation. The tests on the coupled insert/outsert system were performed at 4.2 K, using a set of tailor-made quench protection heaters to be employed in the actual insert. In these tests, in the transition to the normal state, a noticeable part of the outsert stored energy could be transferred to the insert through the inductive coupling. The experimental results are analyzed here by means of a quasi-3-D COMSOL FEM model. The model, developed at the University of Bologna in collaboration with the NHMFL, enables one to calculate the coil temperature distributions, the internal voltages, and current evolution through the increase of normal zone resistances over the insert coils. A lumped parameter equivalent model of the magnet electric circuit permits including appropriately the effect of mutual inductive coupling between the insert and the outsert coils on the currents in them. Also, the model accounts for a time-varying 2-D distribution of the magnet field generated by both the insert and outsert within their coils in the event of a quench. The REBCO insert coils are described in terms of an anisotropic homogenized medium, albeit with due regard for the essentially nonlinear characteristics of the superconducting tapes and for the interturn stainless steel reinforcement and sol-gel alumina insulation. A detailed comparison between the numerical and experimental results provides detailed insights into the modes of propagation of the normal zones.

MgB2 material operating at moderate field can conveniently be considered for short-term cost-effective implementation of superconducting magnetic energy storage ( SMES). Due to the intrinsic performance of this material, moderate field operation and cryogen free cooling in the range 10–20 K need to be considered. For investigating the feasibility of MgB2 based SMES technology a 10 kW—2 s system is under development. A small MgB2 coil was developed and tested as the first step of this study. The cooling was obtained by means of a cryogen-free test facility based on a two-stages Gifford–McMahon regenerative heat exchanger. A power conditioning system (PCS) was also developed in order to connect the coil with the power grid. The PCS is based on power mosfet components and consists of two parallel dc/dc boost converters and an inverter connected to a common dc bus at 100 V. Continuous operation of the coil at a charge/discharge rate up to 24 kW can be achieved by means of the PCS. Details of the layout of the coil and the thermal connection with the cryocooler are discussed in this paper. Results of cooling test are reported. Possible SMES operation is also discussed.

A new geometrical model has been implemented in the THELMA code, aiming at permitting the study of multistage, forced-flow cable-in-conduit conductors with rectangular cross-section, like that proposed for the toroidal field (TF) magnets of the future fusion reactor DEMO. This model gives verisimilar strand trajectories which should enable the adoption of a more detailed interstrand electrical contact model. This paper shows the results of the first application of these new code features to a set of contact resistance and ac loss experiments carried out by Twente University on a DEMO TF prototype conductor. The results of this analysis present some unforeseen aspects that need further work to be fully understood, to improve and validate the model.

The National High Magnetic Field Laboratory (NHMFL), Tallahassee, FL, USA, has developed, built, tested, and commissioned a 32 T all-superconducting user magnet system combining two series-connected high-field high-temperature superconductor (HTS) nested inner coils (insert) wound with SuperPower, Inc. REBCO tapes and a low temperature superconducting (LTS) outer magnet (outsert) composed of five coils (subdivided into 17 electrical sections). Protected-quench tests were performed at the NHMFL to analyze the reliability of the 32 T magnet during the superconducting-to-normal transition. The quench tests were performed at different values of transport current both in the HTS insert and in the LTS outsert. The University of Bologna, in collaboration with the NHMFL, has developed a quasi-3-D FE model suited for the analysis of quench in HTS magnets. The model was previously applied to the analysis of the experimental results of the quench tests carried out on the prototype coils, manufactured in the framework of the 32 T magnet R&D activities. In this paper, the numerical model is applied to analyze the quench initiation and propagation in the 32 T HTS insert. The dump of the transport current during quench is computed and compared with the experimental result. The most stressed regions within the insert windings are identified.

Purpose – A conveyor device is studied with the aim to reduce the friction between the inner surface of the beam and the chain. The lower is the friction between the chain and the beam, the lower is the surface wear. The magnetic repulsion force among permanent magnets (PMs) placed on the beam and on the chain is utilized to reduce friction. The paper aims to discuss these issues. Design/methodology/approach – The considered magnetic suspension is realized with PMs in repulsive configuration; it is designed by solving a constrained optimization problem, with reference to the geometry of the 90° horizontal bend FlexLink WL322 conveyor. Flux density field and its gradient are evaluated using volume integral equation method, allowing to calculate the forces acting on the chain and the stiffness of the magnetic suspension. Findings – The magnetic suspension prototype was manufactured and tested. The experimental and calculated values of the forces acting on the chain compares well. A stable horizontal equilibrium of the chain was obtained during both static and dynamical tests. Research limitations/implications – The quasi-static model used neglects the dynamical interactions among the elements of the chain, the PMs and loads weight during motions and the eddy current losses in the aluminium beam. However the dynamical tests on the prototype show that the chain motion is regular up to the nominal velocity all along the conveyor with the exception of the trailing edge of the 90° curve. Practical implications – The tests on the prototype show the possibility of a removal or at least a reduction of the friction force between the chain and the inner side of the beam by means of a passive magnetic suspension. As a consequence a reduction of noise and vibrations and an increase of the mean-time-to-failure is expected. Originality/value – Prototype testing shows that the unavoidable vertical instability of the magnetic forces has no practical consequence since, reducing the allowed vertical gap, the chain is stabilized by the gravitational force.

This paper describes the experimental results of a test method for determining the retained critical current after double bending at room temperature for short and straight Ag- and/or Ag alloy-sheathed Bi-2223 superconducting tapes. The critical current measurements were performed after double bending samples of a Bi-2223 tape manufactured by Sumitomo with different bending diameters set to 50, 60, and 80 mm. The results allow one to determine the critical bending diameter of the tape. An experimental technique was also developed for the measurements of the quench energy (QE) of the tapes in boiling liquid nitrogen. The impact of the bending on the QE was experimentally determined through measurements performed on the same samples previously used for the critical current measurements.

A three-year research project, called DRYSMES4GRID, was recently funded by the Italian Minister of Economic Development, Italy. The project, which involves five Italian partners (companies, universities, and research institutes), is aimed to demonstrate the feasibility of cost-competitive SMES based on magnesium diboride (MgB2) with a cryogen-free cooling by means of the manufacturing and the testing of a demonstrator with an objective rating of 500 kJ/200 kW. This rating is deemed suitable for disclosing the critical technological aspects of all components in view of practical applications. A further goal of the project is the assessment of the technical and economic benefits that the SMES can bring to real-world electric grids. The general outline and technical objectives of the project are presented in this paper. Preliminary design assumptions and results concerning the MgB2 coil and the power condition system are also discussed. © 2018 IEEE.

The transport performance of Nb3Sn cable-inconduit conductors (CICCs) depends on the strain distribution along the superconducting filaments determined by the combination of electromagnetic and mechanical forces applied to the strands. Experimental studies on the effect of bending strain were performed at the University of Twente by means of the Test Arrangement for Strain Influence on Strands (TARSIS) facility. The aim of this paper is to verify the agreement between a detailed electromechanical model of the wire and the experimental results obtained in TARSIS. A numerical model of two Nb3Sn internal tin strands was developed, which describes the wire through a distributed parameter nonlinear electrical circuit. The model requires a single strand to be discretized into a number of elements, connected by transverse conductances and subjected to a given strain distribution. The strain distribution maps were computed at the Ecole Centrale Paris by means of the MULTIFIL code at different experimental conditions in the TARSIS facility. The simulation results show good agreement with the experimental ones in terms of both critical current and n-value degradation.

In recent years, numerical models have become popular and powerful tools to investigate the electromagnetic behavior of superconductors. One domain where this advances are most necessary is the 3D modeling of the electromagnetic behavior of superconductors. For this purpose, a benchmark problem consisting of superconducting cube subjected to an AC magnetic field perpendicular to one of its faces has been recently defined and successfully solved. In this work, a situation more relevant for applications is investigated: a superconducting parallelepiped bulk with the magnetic field parallel to two of its faces and making an angle with the other one without and with a further constraint on the possible directions of the current. The latter constraint can be used to model the magnetization of a stack of high-temperature superconductor tapes, which are electrically insulated in one direction. For the present study three different numerical approaches are used: the Minimum Electro-Magnetic Entropy Production (MEMEP) method, the $H$-formulation of Maxwell's equations and the Volume Integral Method (VIM) for 3D eddy currents computation. The results in terms of current density profiles and energy dissipation are compared, and the differences in the two situations of unconstrained and constrained current flow are pointed out. In addition, various technical issues related to the 3D modeling of superconductors are discussed and information about the computational effort required by each model is provided. This works constitutes a concrete result of the collaborative effort taking place within the HTS numerical modeling community and will hopefully serve as a stepping stone for future joint investigations.

The circle-in-square cable-in-conduit conductor that will be adopted for the 3L module of the ITER Central Solenoid (CS) coil has been recently tested to assess its performance in conditions relevant for the real operation of the magnet. The tests on a ~43-m-long conductor were performed in the CS Insert Project carried out under the responsibility of the US ITER Domestic Agency, in charge of the CS modules manufacturing. In this paper, we present an evaluation of the ac losses of this conductor aiming at the assessment of the nτ value for the conductor coupling losses. The electromagnetic (EM) and thermal-hydraulic (TH) analysis of these experiments is then performed. A detailed distribution of the coupling losses in the conductor is computed, with the electromagnetic part of the THELMA model. The obtained power density distribution is then input to the TH model (the 4C code) to compute the TH parameters (e.g., the temperature evolution at selected locations along the conductor) during Central Solenoid Model Coil current dumps. The EM/TH model is then validated against experimental data and used to extract relevant information on the conductor ac losses.

The development of high temperature superconducting (HTS) DC power cables requires a detailed information on the impact of the main characteristics of the cable architecture on the electrical parameters of the superconducting tapes. This paper presents the experimental dependence of the critical current and n-value of BSCCO-2223 and (Re)BCO commercial tapes on the cabling twist pitch. DC characterizations were conducted on single tapes twisted around an insulated multistranded Cu former from an actual HTS power cable having a typical diameter of 3 cm. For these characterizations, the twist pitch was varied from 5 to 25 cm. To check that permanent degradation occurred, the critical current Ic and the n-value were measured on a flat sample holder prior to conduct the twist pitch measurements on the former. The results can be used to determine the available range of wire twist pitches to help designing HTS DC power cables.

The paper presents a new thermal model of a forced-flow supercritical helium cooling channel, used in cable in conduit conductors. This model is based on a lumped network approach, is integrated in a more general thermal model for the description of the cable strands, and is coupled with the existing cable and joint electromagnetic models. The model is applicable when quasi-steady-state thermal conditions are present for the cooling helium and no phase transitions take place. The results of the model application to a SULTAN sample current sharing temperature test are presented, showing a good agreement between the measured and computed thermal and electrical variables.

The transverse resistivity among filaments is a crucial parameter of composite superconducting wires, since it affects the coupling losses in ac regime, the current transfer length of the wire subjected to bending strain, and the current redistribution among filaments in the case of a quench. A direct measurement of the transverse interfilament resistance as a function of temperature in several Nb3 Sn multifilamentary wires was performed by means of a 4-probe method. This paper deals with the numerical analysis of the measurements performed on a OCSI wire manufactured by Luvata, containing 82 superconducting subelements each made of 150 filaments. The measurements are analyzed by applying a 2-D FEM of the wire cross-section and a 3-D electrical circuit model of the wire sample. The results obtained have been compared with those found on two other Nb3Sn wires with different architecture and a NbTi wire taken as a reference.

The ability of second-generation YBCO coated conductor (CC) tapes to transport high current densities at high temperature, i.e., up to 77 K, and at very high magnetic fields, i.e., above 20 T, are pushing the use of these tapes in various applications, from magnet to power system technologies. An accurate study of their quench behavior is mandatory for the design and safe use of cables and magnets manufactured with this ceramic superconducting material. A new 2-D finite-element method (FEM) numerical quench model, which is called anisotropic model of YBCO CCs, was built for this purpose in the COMSOL Multiphysics environment. One of the most difficult issues in the modeling of the YBCO tapes with the FEM is their high aspect ratio due to the very small thickness of the YBCO layer, about 1 μm. In the model developed, the problem of the high aspect ratio of the tape is tackled by multiplying the tape thickness by a constant factor and then compensating the heat and electrical balance equations through the introduction of material anisotropic properties. The FEM model is validated by comparison with literature experimental data on minimum quench energy and normal zone propagation velocity.

A 32 T all-superconductive magnet with high field REBCO inner coils is under development at the National High Magnetic Field Laboratory, Tallahassee, Florida, USA. As part of the development activity, two prototype coils with full scale radial dimensions and final design features, but with reduced axial length were constructed. The prototype coils consist of six dry-wound double pancakes modules with uninsulated conductor and insulated stainless steel cowind. Quench studies on one of the prototype coils at 4.2 K in self-field and in a background magnetic field of 15 T were performed by activating a set of quench protection heaters. In this paper, we present a numerical analysis of the experimental results of the quench tests of one of the prototype coils. The numerical analysis was carried out through a coupled electro-thermal FEM model developed at the University of Bologna. The model is based on the coupling with distributed contact resistances of the coil pancakes described as 2D elements. A homogenization procedure of the REBCO tape and other coil materials is presented, which allows reducing the number of degrees of freedom and the computational effort. The model is applied to the analysis of the current and voltage evolutions during the experimental quench tests on the prototype coil.

An international round robin test was carried out in order to establish a test method for retained critical current after double bending at room temperature of Ag-sheathed Bi-2223 superconducting wires. Tests for commercial Bi-2223 tape were conducted by six laboratories using the same guidelines. The standard uncertainties (SUs) of measurands were evaluated for these four quantities: I C0, I C/I C080, I C/I C060, I C/I C050, where, I C0 is initial critical current and I C /I C0XX is critical current after XX mm bending. Using an F test to determine where the most scatter was generated in the test results it was found that the greatest scatter in the normalized critical current measurements came from inter-laboratory scatter. In a type-B uncertainty evaluation, the major contribution was from the bending diameter and measuring temperature. The relative SU tended to increase as the bending diameter decreased. A specific mandrel diameter corresponding to a retained critical current of 95% could be determined with a relative SU of 1.3%. In order to reduce the overall scatter, the temperature difference between the critical current measurements before and after bending should be small.

Since year 2009, the joint of the toroidal field (TF) conductor samples tested in the SULTAN facility at PSI Villigen, CH, is solder-filled, so that the helium coolant can only flow axially inside the central channel. The latter is however plugged starting about 40–45 mm downstream of the joint. The helium has then to pass from the central channel into the annular cable region over such a short length that the desired homogeneity of the flow distribution among the petals in the high field region is not guaranteed a priori, since central helix and petal wrappings act as azimuthally non-uniform obstacles to the radial flow. In the paper we first present a geometrical model for estimating this interference to the radial flow, and combine it with a CFD (ANSYS-FLUENT) model of the hydraulic effect of the plug, in order to estimate the mass flow rate imbalance among the petals at the beginning of the plug. This is then used as boundary condition by the THELMA code, to parametrically assess the results of a TCS measurement.

A magnetization-based formulation of the volume integral equations method for 3-D magnetostatics is discussed. The magnetization of each element is related, via the constitutive law of the material, to the average magnetic flux density within the element rather than to the value at the center as is usually done. This assumption leads to more accurate distribution of magnetization and allows a faster convergence of the solution. Moreover, it leads to more symmetric matrix of the coefficients and reduces the numerical instability due to looping patterns of magnetization, which is inherent to integral methods. The formulation is made effective by the use of a hybrid numerical and analytical approach, which allows for the fast and accurate calculation of the coefficients. The proposed model is validated and compared with the usual model both for saturable and linear material with high susceptibility.

This paper presents the models of three new-generation high-performance samples of ITER full-size Nb3Sn Cable-in-Conduit conductors (CICCs). These samples were recently tested in SULTAN, at CRPP (Switzerland), and demonstrated the technical feasibility of CICCs that satisfy the ITER specifications. The improvements of these samples regard not only the cable strand and the cabling parameters, but also the manufacturing details of their SULTAN samples. A set of changes have been adopted on the SULTAN sample termination and joint, to try to achieve a more uniform current distribution among the strands and a more accurate measurement of the strand longitudinal voltages. This paper presents a comparative numerical analysis of the effects that these different joint technologies have on the DC current distribution among the strands of the SULTAN samples cable. A correlation between the current distribution in the sample and the resistance distribution at the joint is also shown and discussed.

The THELMA code is used to study the coupled thermal-hydraulic electro-magnetic problem of the current and temperature distribution inside the TFPRO2 Nb3Sn SULTAN sample, which was tested in 2007. The code computes self-consistent voltage and temperature values both on the jacket, where they are measured, and inside the cable, where they are more directly representative of the conductor performance. The measured temperature gradients, related to non-uniform Joule heating at the joint and in the high-field region, as well as to non uniform current distribution, are reasonably well reproduced by the model, together with the voltage-current characteristics.

A novel energy storage system based on a SMES cooled by cryogenic fuel is being developed at the Dept. of Electrical Engineering of the University of Bologna. The system represents an alternative solution with respect to nowadays battery packs for the storage of electrical energy with large specific power. The device may find suitable application on board of road vehicles fed by liquid hydrogen with hybrid power-train (internal combustion engine and electric motor/generation), where a cryogenic tank is already present and no additional cryogenic costs are needed. The housing of the SMES inside the tank used for fuel storage, allows the exploitation of the thermal power required for the continuous evaporation of fuel for feeding the engine to remove the AC losses from the magnet, thus avoiding the need of externally supplied cooling power. In this paper both the design of the magnet and its operation in combination with electronic converter and electric motor/generator during charge/discharge cycles are discussed. Moreover the ongoing activity for the development of a laboratory-scale prototype using MgB2 superconductor is presented.

We have analyzed the characteristics of dc-induction heating of aluminium billets both numerically and experimentally and developed a numerical model for the calculation of the power and temperature distributions inside the billet during the heating. We validated the model by comparison with the experimental results from a laboratory-scale prototype using permanent magnets. We studied the heating process of a real scale system by means of the model, taking into account the weakening of the mechanical properties during the heating. We provide an example of design of an industrial scale dc induction heater.

A superconducting magnet with 200 kJ stored energy and 75 kW output/input power has been designed and manufactured. The device, which is aimed to power quality enhancement in electric grid, consists of two coaxial coils, counter-series connected in order to keep the stray field below the limit of 0.5 mT. The SC wire is a copper stabilized Nb-Ti strand. Several copper slabs are placed within the coils in order to permit heat removal deep inside the coil, thus allowing cryogen-free cooling and operation at high AC losses regime. Several tests with different ramp rate have been carried out at the liquid helium temperature. Design and manufacturing characteristics as well as test results are discussed in this paper.

PurposeTo analyse the heating process of an aluminum billet rotating in a static magnetic field produced by superconducting coils.Design/methodology/approachThe idea is to force the billet to rotate in a static magnetic field produced by a DC superconducting magnet. Since, a static superconducting magnet has no losses, the efficiency of the system is the efficiency of the motor used. In order to evaluate the temperature distribution arising from the field profile produced by a given coil configuration, a numerical model, based on an equivalent electric network with temperature‐dependent parameters, is used.FindingsThe main heating parameters, i.e. heating time, total power injected and temperature difference, are evaluated for different values of angular velocity and magnetic field. The field profile suitable to meet the specifics of an industrial heating process in terms of temperature homogeneity and heating time is determined. Starting form this profile the layout of the magnet is arrived at and some considerations on the operating condition of the superconducting windings are reported.Research limitations/implicationsThe mechanical stress in the billet due to weight, centrifugal effects, applied torque and resonance is examined by taking into account the weakening of the material properties with the increase of temperature and the practical limits of the heating process are evaluated.Practical implicationsDC induction heating of aluminum billet using superconducting magnets can be done fulfilling the specifics of the industrial processes.Originality/valueA high‐efficiency induction heater for aluminum billets using superconducting coils in a novel scheme is investigated.

In this paper, a conveyor device aimed at moving objects placed on a translating chain is studied. Losses, friction, and noise reduction are the main objectives, which are obtained by magnetically levitating the chain with the aid of a system of permanent magnets placed on the chain and on the beam. A stable vertical levitation was obtained, while the horizontal instability was managed reducing the allowed gap between the chain and the beam. A customized three-phase drive unit was also realized in order to reduce the size and the inertia of the conveyor. A prototype was built and tested in order to verify the outcomes of the design.

The present work is focused on the demonstration of an innovative approach to a superconducting magnetic energy storage system by means of next generation superconducting wires. The device is thought to be integrated in a more complex biomass plant for green energy production which includes an anaerobic digester and a cogenerator for biogas and electrical energy production. Presented technology allows the storage of the green energy produced with a very high efficiency and with a better power quality respect to traditional counterparts.

The THELMA (Thermal-Hydraulic-ELectro-MAgnetic) code has been developed with the aim of simulating the main aspects of superconductors to be used in the coils of the International Thermonuclear Experimental Reactor (ITER). An application of the code is presented here, where THELMA is used to simulate a single strand by considering, as cable elements, groups of superconducting (SC) filaments and the corresponding portion of the resistive matrix. This approach is used to reproduce the voltage-current characteristic of a Nb3Sn bronze route strand when subject to a bending mechanical load. Attention is focused particularly on the effect of the applied mechanical load on the critical current, which is considered a relevant item in the explanation of the degradation of the coil performance observed in several ITER Model and Insert Coil experiments. The longitudinal strain of the SC filaments is calculated by means of a composite beam model of the strand, taking into account the nonlinear, temperature-dependent material characteristics of the components. The whole load history is simulated, computing first the thermal strain due to cool-down, and then the mechanical strain due to the bending at 4.2 K

The modeling of current distribution and ac losses in multistrand superconducting cables is strictly related to the evaluation of the cable electric parameters. A perfect knowledge of these parameters is however impossible due to the complicated manufacturing process. Strand deviations from the nominal geometry and interstrand conductance changes occur in the cable compaction process. Statistical approaches can however be applied to deal with these issues. This paper reports a sensitivity analysis aimed to determine the effects of strand deviations from the nominal geometry. The deviations, however, are not directly imposed on the mutual inductances matrix, as this approach can lead to misleading results if the final matrices do not correspond to physically possible configurations. The statistical deviations are imposed on the strands geometry, keeping the strands in physically possible locations by means of geometric constraints. A code for the electromagnetic analysis of superconducting cables is then applied to the calculation of current distribution and ac losses in the different configurations, evaluating the relevance of the main parameters variation.

In order to investigate how a superconducting fault current limiter (SFCL) can enhance the performance of a power system, an accurate circuit model of the device needs to be introduced in power system simulators. In this paper, we present a finite-element numerical model to calculate the time evolution of the voltage across a magnetic-shield-type SFCL, when it is connected to an external circuit. The calculation of the voltage is carried out by using the energy conservation law, and requires the calculation, at any instant, of the current density distribution inside the superconducting tube and magnetization distribution inside the ferromagnetic core of the device. These distributions are determined by means of two coupled equivalent electric and magnetic circuits, whose topology and components are obtained through the spatial integration of quasi-static form of Maxwell equations. Comparisons between numerical and experimental results are shown.

The electromagnetic models developed in the recent years to study current distribution and ac losses in superconducting magnets have focussed on detailed descriptions of either cables or joints/terminations, usually considered separately. A more accurate physical description of the magnet system requires to model simultaneously the different parts of cable, the joints between them and the cable terminations. In this work we present the coupling of a distributed parameters circuit model of the cable with a simple lumped parameters resistive model of the joints and terminations. The principles and results of the coupling strategy are illustrated and validated by means of a comparison with the analytical solution of the problem found for triplex cables.

A numerical model to calculate current and losses distribution inside a multifilamentary superconducting strand, when AC transport current and AC magnetic field are applied, is presented. The model is based on the equivalent circuit network derived from the magneto-quasi-static form of Maxwell equations and takes into account the superconducting material properties by defining a continuous non-linear anisotropic relation which applies for any point of the composite strand. In order to reduce the number of equations of the solving system an algebraic algorithm is applied. Some cases for which an analytical solution is available are examined. © 2002 Elsevier Science B.V. All rights reserved.

Proton therapy for the treatment of cancers adopts a rotating gantry to irradiate the tumor from any direction. The gantry system consists of different beamline magnets that bend the proton beam towards the patient. The use of superconducting magnets allows reducing the size and weight of the last bending section. During the gantry operation, it is necessary to change the magnetic field of the last bending section in order to vary the proton penetration depth. This change determines electrodynamic transients in the superconducting strands and cables that generate losses. A thorough computation of these losses is essential for a safe design of the cryogenic system. Two main types of losses must be accounted for when dealing with multistrand superconducting cables, related to the magnetization and coupling of the superconducting filaments (intrastrand losses) and to the current loops induced between different strands during electrodynamic transients (interstrand losses). This work describes the methodologies and numerical codes adopted to compute the hysteresis and coupling losses in an innovative magnet system designed by the Paul Scherrer Institute for future superconducting gantries. In this design the superconducting coils are wound using Nb3Sn Rutherford cables. The validation of the numerical tools versus analytical results is presented for simplified cases with uniform magnetic flux density applied to the conductor. The results of the losses calculation and the impact of the different contributions are then presented for the actual proposed magnet system configuration.

Superconducting strands are composite wires made of fine superconducting filaments embedded in a metallic matrix. The transverse resistivity among superconducting filaments affects the coupling losses during electromagnetic transients and the electro-thermal behavior of the wire in case of a quench. A direct measurement of the transverse interfilament resistance as a function of temperature in NbTi multi-filamentary wires was performed at the ENEA Frascati Superconductivity Division, Italy by means of a four-probe method. The complexity of these measurements is remarkable, due to the current distribution phenomena that occur among superconducting filaments during these tests. A two-dimensional finite element method model of the wire cross section and a three-dimensional electrical circuit model of the wire sample developed at the University of Bologna are applied here to derive qualitative and quantitative information about the transverse electrical resistance matrix. The experiment is aimed at verifying the qualitative behaviors and trends predicted by the numerical calculations, especially concerning the current redistribution length and consequent length effects of the sample under test. A fine tuning of the model parameters at the filament level allowed us to reproduce the experimental results and get quantitative information about the current distribution phenomena between filaments.