Search

High temperature superconductors, including Bi:2212, are being considered for high magnetic field magnets to be used in particle accelerators. A knowledge of HTS magnetization properties are needed for calculating field errors in such accelerator magnets. Here we present magnetic measurements of a 17 strand Bi:2212 Rutherford cable. The Bi:2212 strand was a Bruker OST PMM170123 non-twisted wire wound into a cable with dimensions 1.46 mm x 7.8 mm and cable pitch of 50.8 mm. M-H loop measurements were made with a Hall probe at 4.2 K. Two calibration methods were compared, (i) a flux exclusion approach, and (ii) a Ni replacement technique. The latter was found to be more reliable in this case. An effective filament diameter of 358 μm was calculated for the strand at 4.2 K and a perpendicular applied field of 3.33 T, significantly larger than the diameter of the filament bundle. Magnetization values at 1 T on the shielding branch were seen to be 100 kA/m when normalized to the total strand in the cable volume. The penetration field ≅ 0.4 T.

Interconductor contact resistivity ( ICR ) is a key property in determining the stability and current sharing of coated conductor cables. Most coated conductor cables have relatively high contact resistivity and low current sharing as fabricated because of surface roughness and an oxide layer that forms on the Cu-stabilizer. Here we work to quantify the differences in using three methods to modify ICR : sample diffusion bonding, deposition surface modification, and thin conformable inserts. At first a stack of two coated conductors was used to simulate a cable. This stack was put under transverse pressure and exposed to moderate temperatures to promote diffusion bonding via the removal of the unstable Cu-oxide layer ( ICR measurements were performed on stacks before and after diffusion bonding, for stacks with deposited surface layers, and for stacks with a smart material insert. In addition, Roebel cables were prepared with deposited layers on the individual strands and atmospherically controlled diffusion bonding, and ICR was compared between the as-received and modified Roebel cables as well as coated conductor stacks.

We present measurements on high-temperature superconducting (HTS) superconducting cables using a NbTi-wound, 3-T dipole ac magnetometer. Our goal is to measure the magnetization properties of the cables for use in accelerator magnet field error calculations. YBCO-wound CORCTM and Roebel cables were measured in pool boiling helium with no epoxy impregnation. The Roebel cable was a 9–5.6 style, while the CORCTM sample had 16 tapes, and was 3.21 mm OD. For measurement, the samples were inserted in a G-10 holder with pickup and compensation coils, and subjected to a ramping dipole field. The signals were read from a set of nanovoltmeters using data aquisition (DAQ) to monitor the pickup coils, and the data were imported into a LabVIEW software program. The Roebel and CORCTM cable measurements were compared to one another and to the tapes they were made with. Maximum cable magnetization was 1000 kA/m for the CORC and 2200 kA/m for the Roebel when normalizing to total cable volume. These values increased to 2860 and 2640 kA/m, respectively, when normalizing to tape volume.

HTS cables proposed for next-generation main, hybrid, and insert magnets will have very high I c (>10 kA) at 14 T and above. Determining the quench, current-sharing, and other properties of these cables will require measurement systems, which will likely incorporate a superconducting transformer. The utilization of high-field persistent mode solenoids will make size and cost less prohibitive, allowing a higher frequency of measurements within a larger number of research facilities. Additionally, a system designed for use with solenoids will allow for experiments at higher maximum fields than those achievable using dipoles and split coils. Proposed in this document is a bifilar sample probe, which fits within a 77-mm bore solenoid capable of measuring up to a 6-mm outer diameter conductor-on-round-core REBCO cables or wires up to 20 kA in transverse fields at 12 T and 4.2 K. Splices and mechanical considerations will be discussed.

We measured stability and quench in a YBCO pancake coil in liquid nitrogen bath (77.3 K) and in liquid helium bath (4.2 K) with externally applied magnetic fields up to 10 T. The pancake coil was instrumented with voltage taps and thermocouples in the winding for measurements of radial and azimuthal quench propagations. A heater was placed on the inner-most part of the winding. Heat pulses of various powers and durations were generated at different coil currents to measure quench and normal zone propagation. A protection circuit with a possibility of setting different quench voltage limits was used to turn off the coil current after the voltage limit was exceeded.

The development of coils that can survive a quench is crucial for demonstrating the viability of MgB(2)-based main magnet coils used in MRI systems. Here we have studied the performance and quench properties of a large (outer diameter: 901 mm; winding pack: 44 mm thick × 50.6 mm high) conduction-cooled, react-and-wind (R&W), MgB(2) superconducting coil. Minimum quench energy (MQE) values were measured at several coil operating currents (I(op)), and distinguished from the minimum energy needed to generate a normal zone (MGE). During these measurements, normal zone propagation velocities (NZPV) were also determined using multiple voltage taps placed around the heater zone. The conduction cooled coil obtained a critical current (I(c)) of 186 A at 15 K. As the operating currents (I(op)) varied from 80 A to 175 A, MQE ranged from 152 J to 10 J, and NZPV increased from 1.3 to 5.5 cm/s. Two kinds of heater were involved in this study: (1) a localized heater (“test heater”) used to initiate the quench, and (2) a larger “protection heater” used to protect the coil by distributing the normal zone after a quench was detected. The protection heater was placed on the outside surface of the coil winding. The test heater was also placed on the outside surface of the coil at a small opening made in the protection heater. As part of this work, we also developed and tested an active protection scheme for the coil. Such active protection schemes are of great interest for MgB(2)-based MRIs because they permit exploitation of the relatively large MQE values of MgB(2) to enable the use of higher J(e) values which in turn lead to competitive MgB(2) MRI designs. Finally, the ability to use a quench detection voltage to fire a protection heater as part of an active protection scheme was also demonstrated.

Purpose/Objective(s) To explore the potential application of medium temperature superconductors (MTS) and high temperature superconductors (HTS) to the development of background magnets for MRI systems for use in combination with particle beam therapy (PBT) and the concept design for MRI-guided PBT. Materials/Methods In proton/heavy ion therapy, because of the sharp dose drop off owing to the Bragg peak effect, imaging guidance becomes crucial. Various imaging techniques have been considered, but MRI, with its soft tissue contrast, 3D imaging, high spatial resolution, and potential on-live monitoring feature would be particularly useful in combination with PBT. One of the key challenges is that the charged particle beam will be deflected in a transverse magnetic field. The deflection of the particle beam in the presence of the background B0 has received significant interest, and both in-line and transverse configurations have been considered. For in-line configurations, smaller deflections are seen, with some rotation, and for transverse, greater deflection, and momentum change. With the use of a pencil beam PCT approach in combination with computation, such deflections can in principle be compensated. However, if “real time” imaging is to be enabled, minimum computation and reconstruction time is desirable. For these reasons, minimizing and simplifying the deflection is desirable. This can be enabled with the use of the in-line configuration, but in this case an in-line system compacts enough to mount on a gantry may be of interest. This can be made more achievable with the use of MTS or HTS where less bulky magnets are possible, cooling is simplified, and systems are much less quench sensitive. At the same time, the minimization of fringe fields can also substantially simplify the particle beam trajectories and minimize their deflection. This may also be enabled for open configuration transverse configurations with the use of less bulky HTS and MTS magnets. For this work we performed magnetic modelling of several different configurations of MTS and HTS magnets, both in-line and transverse, using Vector fields and other modelling software. The sizes and configurations of the MTS and HTS systems as compared to potential competing LTS systems were then compared. Particle beam trajectories were then computed with GEANT software and compared. Results (1) MTS and HTS systems can allow the development of in-line MR which are sufficiently compact, open, and robust (to quench) to allow mounting on a gantry. This reduces particle beam deflection, while at the same time allowing rotation of the beam around the patient; (2) Several designs for minimized fringe fields are given, both for in-line and transverse configurations. Conclusion MTS and HTS have a significant potential for application to MRI-guided Particle Therapy based on their potential for mounting on a gantry, and enabling of more open background magnet configurations. Magnet designs with minimized fringe fields are also possible and useful.

A 9-tape, 14 mm wide ReBCO Roebel cable was soldered onto a U-shaped holder. The critical current, Ic, was measured at 77 K and self-field. The cryostability of the cable was studied in response to the application of local pulses of 1–14 W at several values of i = I/Ic. A detailed analysis of the cable’s cryostability was presented. With a Stekly parameter α = G/Q « 1 and a heat generation margin of ∼190 kW/m2 the present ReBCO cable was shown to be ultra cryostable with respect to internally generated transport-current overload. However, the cable was much less stable against externally and locally applied disturbances because of the tendency to initiate local film boiling. A locally applied 10 W led to a prediction of a film-boiling-cooled zone with a temperature of 181 K. However, when cold-end cooling was considered, the predicted hot spot temperature decreased to 87–115 K depending on the surface-cooling efficiency. Predictions were compared to experiment extracting a cooling efficiency parameter representing the penetration of the cryogen into the cable. Experiment showed the generation of time stable normal zones which were a function of disturbance power. This led to the description of the cable stability in terms of minimum quench power; the results are presented in stability diagrams.

We have measured ReBCO coated conductor-based conductor on round core (CORC®) and Roebel cables at 77 K in a spinning magnet calorimeter, which subjected the tapes in the samples to a radial magnetic field of 566 mT (peak) at frequencies up to 120 Hz (272 T s−1, cyclic average) with an approximately sinusoidal waveform. The samples were oriented such that the field applied to the tapes within the cables was entirely radial, simplifying subsequent analysis. An expression for loss which included hysteretic, flux creep and eddy current losses was fit to both the CORC® and the Roebel cables. This expression allowed easy comparison of the relative influence of eddy currents and flux creep (or power-law behavior) effects. The loss of both the CORC ® and Roebel cables measured here were seen to be essentially the sum of the hysteretic loss, flux creep effects, and the normal metal eddy current losses of the individual tapes. The losses of these cables were measured at high B × dB/dt with no coupling current loss observed under the present preparation conditions. The influence of flux creep effects on loss were not negligible. The losses of the CORC® cable per meter of tape were seen to be reduced from the case of a flat tape because of the helical geometry of the tapes.

A new facility for the measurement of AC loss in superconductors at high dB/dt has been developed. The test device has a spinning rotor consisting of permanent magnets arranged in a Halbach array; the sample, positioned outside of this, is exposed to a time varying AC field with a peak radial field of 0.566 T. At a rotor speed of 3600 RPM the frequency of the AC field is 240 Hz, the radial dB/dt is 543 T/s and the tangential dB/dt is 249 T/s. Loss is measured using nitrogen boiloff from a double wall calorimeter feeding a gas flow meter. The system is calibrated using power from a known resistor. YBCO tape losses were measured in the new device and compared to the results from a solenoidal magnet AC loss system measurement of the same samples (in this latter case measurements were limited to a field of amplitude 0.1 T and a dB/dt of 100 T/s). Solenoidal magnet system AC loss measurements taken on a YBCO sample agreed with the Brandt loss expression associated with a 0–0.1 T Ic of 128 A. Subsequently, losses for two more YBCO tapes nominally identical to the first were individually measured in this spinning magnet calorimeter (SMC) machine with a Bmax of 0.566 T and dB/dt of up to 272 T/s. The losses, compared to a simplified version of the Brandt expression, were consistent with the average Ic expected for the tape in the 0–0.5 T range at 77 K. The eddy current contribution was consistent with a 77 K residual resistance ratio, RR, of 4.0. The SMC results for these samples agreed to within 5%. Good agreement was also obtained between the results of the SMC AC loss measurement and the solenoidal magnet AC loss measurement on the same samples.

This paper describes a detailed design of a 6-T Nb 3 Sn superconducting racetrack coil designed for conduction cooling. We then describe a bench test pursued as a proof of concept for one winding of an actively shielded air-core electric machine with superconducting field windings. Electromagnetic design selection is drawn from previous optimization work. The coil former design is then discussed. Numerical simulations of thermal and structural features are pursued to determine temperature distribution and strain within the winding. A coil instrumentation and experimental setup of a quasi-conduction cooled system is described. Finally, test results are presented; a maximum critical current of 480 A was reached at a peak temperature of 7.9 K, surpassing the operational current goal of 435 A. Future work and planned improvements to the test setup are discussed.

Filamentization has been shown to be an effective method in reducing magnetization ac loss of RE-Ba-Cu-O (REBCO) coated conductors. We have used a laser striation method followed by selective electroplating of copper to fabricate fully stabilized multifilamentary REBCO tapes with 12, 24, and 46 filaments. While expected levels of ac loss reduction has been confirmed in short segments of such multifilamentary tapes, electric coupling between the filaments of long tapes needs to be suppressed by transposition of the filaments. In this work, a technique was developed and implemented to reduce ac losses in REBCO multifilamentary coated conductors through a new design that allows us to transpose the filaments without any mechanical twisting. The process consists of patterning REBCO tapes by laser ablation followed by partial insulation and bonding. A twofold reduction of ac magnetization losses was achieved in the transposed multifilamentary REBCO tape when subject to an external perpendicular magnetic field at frequencies between 50 and 200 Hz and peak field values up to 88 mT. The filament-to-filament resistance and its effect on the coupling have also been investigated.

The high field superconducting magnets required for ongoing and planned upgrades to the Large Hadron Collider (LHC) will be wound with Nb3Sn Rutherford cables for which reason studies of Nb3Sn strand, cable, and magnet properties will continue to be needed. Of particular importance is field quality. The amplitudes of multipoles in the bore fields of dipole and quadrupole magnets, induced by ramp-rate-dependent coupling currents, are under the control of the interstrand contact resistances—crossing-strand, $R_{c}$ , adjacent strand, $R_{a}$ , or a combination of them, $R_{{\text{eff}}}$ . Although two decades ago it was agreed that for the LHC $R_{c}$ should be in the range 10–30 μ Ω, more recent measurements of LHC quadrupoles have revealed $R_{c}$ values ranging from 95 to 230 μ Ω. This paper discusses ways in which these values can be achieved. In a heavily compacted cable $R_{{\text{eff}}}$ can be tuned to some predictable value by varying the width of an included stainless steel (effectively “insulating”) core. But cables are no longer heavily compacted with the result that the crossing strands of the impregnated cable are separated by a thick epoxy layer that behaves like an insulating core. If a stainless steel core is actually present, $R_{{\text{eff}}}$ must be independent of core width. Since there is no guarantee that a fixed predetermined amount of interlayer separation could be reproduced from winding to winding it would be advisable to include a full width core.

The coupling magnetization of a Rutherford cable is inversely proportional to an effective interstrand contact resistance R eff , a function of the crossing-strand resistance R c , and the adjacent strand resistance Ra. In cored cables, R eff continuously varies with W, the core width expressed as percent interstrand cover. For a series of un-heat-treated stabrite-coated NbTi LHC-inner cables with stainless-steel (SS, insulating) cores, R eff (W) decreased smoothly as W decreased from 100%, whereas for a set of research-wound SS-cored Nb 3 Sn cables, R eff plummeted abruptly and remained low over most of the range. The difference is due to the controlling influence of R c - 2.5 μΩ for the stabrite/NbTi and 0.26 μΩ for Nb 3 Sn. The experimental behavior was replicated in the R eff (W )'s calculated by the program CUDI, which (using the basic parameters of the QXF cable) went on to show in terms of decreasing W that: 1) in QXF-type Nb 3 Sn cables (R c = 0.26 μΩ), Reff dropped even more suddenly when the SS core, instead of being centered, was offset to one edge of the cable; 2) R eff decreased more gradually in cables with higher R c 's; and 3) a suitable R eff for a Nb 3 Sn cable can be achieved by inserting a suitably resistive core rather than an insulating (SS) one.

For many accelerator magnets field quality at the bore is a critical requirement for which reason it is necessary to fully characterize the persistent-current magnetization of strands of the kind under consideration for these magnets. The magnetization of a strand is generally measured in a magnetometer. However, certain effects can differentiate such measurements from the true magnetizations of strands in magnets. This report focuses on persistent-current magnetization: 1) measured by vibrating-sample magnetometer on segments of strand extracted from a section of heat treated Nb 3 Sn cable as functions of angle of the applied field, and 2) calculated as function of applied transport current. It is found that the magnetization of a strand in a cable increases by ~10% as the field applied to the cable is shifted from edge-on to face-on, and that the difference between the current-on and current-off magnetizations is not significant until close to the operational field of a magnet.

A react-and-wind MgB2 coil segment for a conduction-cooled magnetic resonance imaging (MRI) machine has been fabricated and tested. The coil was developed as part of a collaborative effort on a conduction-cooled, MgB2-based, whole-body MRI image guided radiation therapy device. This study focuses on the fabrication, winding, instrumentation, cooling, and initial critical current (I c ) testing of this near-full-size MgB2 segment coil. The coil was 0.9 m in diameter; the winding pack, 44.0 mm wide × 50.6 mm high, used 1.7 km of an 18 filament MRI-style conductor with Nb chemical barriers, Cu interfilamentary matrices, and an outer monel sheath. The conductor was insulated and reacted before winding onto a stainless steel former. The coil was instrumented with Cernox and E-type thermocouple temperature sensors, strain sensors, and voltage taps. The conduction-cooled coil was mounted in a cryostat capable of accepting coils of up to 0.9 m in diameter and 0.5 m in height. Critical current measurements were performed as a function of temperature during a controlled heating of the coil. The operational target was I = 200 A at 13 K. The full magnet was designed to produce 0.75 T in the imaging area (at I = 200 A), with a maximum field of 1.93 T in the winding. The single segment coil measured here exceeded this operation specification, with an I c of 280 A at 15 K and a maximum field 1.93 T in the winding. The coil was modeled using a finite element method, and a load line plot showed that 100% of short sample was reached at 21.5 K and above. These measurements demonstrate the viability of conduction-cooled MgB2 background coils as replacements for liquid helium cooled NbTi background coils in future MRI devices.

There are a few key conductor-specific factors which influence the power loss of superconductors; these include critical current, geometry, and normal metal resistivity. This paper focuses on the influence of sample geometry on the power loss of superconducting strips and the effect of filamentation and sample length as a function of the field penetration state of the superconductor. We start with the analytical equations for infinite slabs and strips and then consider the influence of end effects for both unstriated and striated conductor. The loss is then calculated and compared as a function of applied field for striated and unstriated conductors. These results are much more general than they might seem at first glance, since they will be important building blocks for analytic loss calculations for twisted geometries for coated conductors, including helical (Conductor on Round Core, CORC), and twisted (e.g., twist stack cables) geometries. We show that for relatively low field penetration, end effects and reduced field penetration both reduce loss. In addition, for filamentary samples the relevant ratio of length scales becomes the filament width to sample length, thus modifying the loss ratios.

In this study, a two layer conductor on round core cable was tested for stability and normal zone propagation at 77 K in a liquid nitrogen bath. The cable was instrumented with voltage taps and wires on each strand over the cable’s central portion (i.e. excluding the end connections of the cable with the outside world). A heater was placed in the central zone on the surface of the cable, which allowed pulses of various powers and durations to be generated. Shrinking (recovering) and expanding (not recovering) normal zones have been detected, as well as stationary zones which were in thermal equilibrium. Such stationary thermal equilibrium zones did not expand or contract, and hit a constant upper temperature while the heater current persisted; they are essentially a form of Stekly stability. Overall, the cable showed a high degree of stability. Notably, it was able to carry a current of 0.45Ic cable with maximum temperature of 123 K for one minute without damage.

A Roebel cable, one twist pitch long, was modified from its as-received state by soldering copper strips between the strands to provide inter-strand connections enabling current sharing. Various DC transport currents (representing different percentages of its critical current) were applied to a single strand of such a modified cable at 77 K in a liquid nitrogen bath. Simultaneous monitoring of I–V curves in different parts of the strand as well as in its interconnections with other strands was made using a number of sensitive Keithley nanovoltmeters in combination with a multi-channel high-speed data acquisition card, all controlled via LabView software. Current sharing onset was observed at about 1.02 of strand Ic. At a strand current of 1.3Ic about 5% of the current was shared through the copper strip interconnections. A finite element method modeling was performed to estimate the inter-strand resistivities required to enable different levels of current sharing. The relative contributions of coupling and hysteretic magnetization (and loss) were compared, and for our cable and tape geometry, and at dB/dt = 1 T s−1, and our inter-strand resistance of 0.77 mΩ, (enabling a current sharing of 5% at 1.3Ic ) the coupling component was 0.32% of the hysteretic component. However, inter-strand contact resistance values of 100–1000 times smaller (close to those of NbTi and Nb3Sn based accelerator cables) would make the coupling components comparable in size to the hysteretic components.