Search

A compact high temperature superconducting (HTS) magnet, which consists of a stack of 500 HTS thin film annuli, was constructed and tested. Each thin film annulus, manufactured by the AMSC using the Rolling Assisted Bi-axially Textured Substrate (RABiTS) method, has a square cross-section of 40 mm × 40 mm with a thickness of 80 m. It has a 25-mm center hole created by machining. This paper reports a study on the anisotropic J c issue due to the rolling procedure of the Ni substrate direction and its impact on field homogeneity. Also, three different stacking methods with rotation angles of 22.5°, 90°, and 0° against the rolling direction have been tested to study their impacts on strength, spatial homogeneity, and temporal stability of trapped fields. Finally, the 500-annulus magnet was tested at 21 K under a cryogen-free environment using a GM cryocooler. The spatial field homogeneity and temporal stability were measured at 21 K and compared with those obtained in a bath of liquid nitrogen at 77 K.

We have studied the levitation force of a Gd–Ba–Cu–O bulk superconductor 60 mm in diameter and 20 mm in thickness with a newly developed superconducting magnet system that can provide high field gradient. The magnet system was composed of two superconducting coils that are co-axially placed, and currents are passed in the opposite direction. To study the effect of radial and vertical magnetic field components on the levitation force and its time relaxation, the superconductor was placed at several locations. It was confirmed that the levitation force is strongly dependent on the radial component of magnetic fields. It was also found that the time relaxation was reduced without sacrificing the levitation force by increasing the radial component of the external field.

The unique features of HTS offer opportunities and challenges to a number of applications. In this paper we focus on NMR and MRI magnets and present a brief description of some of the NMR/MRI magnet programs at FBML that are currently on-going and expected to begin in the near future, specifically a 1 GHz NMR magnet comprising an LTS magnet and an HTS insert wound with BSCCO-2223 tape, an NMR magnet assembled from YBCO annuli, and an MRI “demonstration” magnet wound with MgB2 wire. Two conclusions are presented for HTS as applied to NMR/MRI magnets: (1) impact of HTS on NMR/MRI magnet technology is great, because HTS offers many opportunities and challenges, some of which are described here; and (2) impact of NMR/MRI magnets on HTS manufacturers is mixed—for high-field NMR magnet, it is small because the required amount of HTS in these magnets is negligible compared with that in electric power applications; for “low-field” NMR magnets, it may be potentially large if MgB2 replaces Nb–Ti and Nb3Sn; and for MRI magnets, it is potentially huge if MgB2 replaces Nb–Ti and Nb3Sn.

We developed a shielding current distribution model in a bulk superconductor with the aim of estimating the levitation force. First, we directly measured the magnetic field in axial and radial directions above a bulk Y–Ba–Cu–O superconductor levitated by an electromagnet using a three-dimensional Hall sensor. We then calculated the field distribution by assuming two types of the shielding current distributions and compared the results with the measured values. The field distributions based on the model in that the shielding currents are flowing concentrically at the surface were not in good agreement with the measured data. We then assumed that the shielding currents are not uniform and the boundary has the form of n th hyperbola, which could reproduce the measured field distribution reasonably well.

We have proposed a new cryocooler system for the high temperature superconducting (HTS) bulk magnet. In this system, the cyclic temperature variations are generated. A change of magnetic trapped flux under the cyclic temperature variation is examined in detail. As a result, it has been found that the trapped flux mainly deteriorates at the first temperature rise, but it has been kept almost constant after the first temperature descent and following thermal cycles when the bulk is excited at 100 mT. Moreover it has been found that the change of the trapped flux at the excitation of 130 mT becomes same as the case of 100 mT. The deterioration at the first temperature rise can be explained by the Bean model, but we can find a difference between calculated results by the Bean model and experimental trapped flux distributions near the center of the bulk. Therefore, we take into consideration a two dimensional invasion of the magnetic field to explain the difference. Consequently calculated distributions of the trapped flux become in goal accordance with experimental ones.

The paper discusses protection issues for HTS magnet. For protection, the HTS magnet must rely on an active technique. Closed-form expressions of the matrix metal operating current density based on overheating and internal voltage criteria for protection, under very simplifying assumptions, are presented. Perhaps the most important conclusions of these criteria are that: (1) HTS (and LTS) magnets must be wound with composite conductor having a significant portion of its overall cross section occupied by normal metal generally of high electrical conductivity and (2) HTS windings must possess “high” NZP velocities to make the resistive zone occupy as large a fraction of the winding volume as possible. The paper also derives an analytical expression, under another set of simplifying assumptions, to determine the minimum resistive voltage level, dictated by the maximum hot-spot temperature set at 150 K, required to initiate an active protection process. Remarkably, this minimum detection voltage is nearly independent of the matrix metal current density, I op / A m . For a set of operating parameters used in a numerical example, a computed minimum detection voltage, at I op / A m = 5 × 10 4 A/cm 2 , is ∼30 mV, which, considering it must be discerned in the presence of extraneous voltage signals likely to be present in real world operating conditions, would be non-trivial. To satisfy the overheating criterion at a level of I op / A m , which keeps the winding overall current density “viable” and at the same time to raise the minimum detection resistive voltage, the winding volume occupied by the resistive state must be expanded. The paper concludes with discussion of challenging new areas of research for protection of HTS magnets.

SuperPower is focused on scaling up second-generation (2-G) high-temperature superconductor (HTS) technology to pilot-scale manufacturing. The emphasis of this program is to develop R&D solutions for scale-up issues in pilot-scale operations to lay the foundation for a framework for large-scale manufacturing. Throughput continues to be increased in all process steps including substrate polishing, buffer and HTS deposition. 2-G HTS conductors have been produced in lengths up to 100 m. Process optimization with valuable information provided by several unique process control and quality-control tools has yielded performances of 6000–7000 A m (77 K, 0 T) in 50–100 m lengths using two HTS fabrication processes: metal organic chemical vapor deposition (MOCVD) and pulsed laser deposition (PLD). Major progress has been made towards the development of practical conductor configurations. Modifications to the HTS fabrication process have resulted in enhanced performance in magnetic fields. Industrial slitting and electroplating processes have been successfully adopted to fabricate tapes in width of 4 mm and with copper stabilizer for cable and coil applications. SuperPower’s conductor configuration has yielded excellent mechanical properties and overcurrent carrying capability. Over 60 m of such practical conductors with critical current over 100 A/cm-width have been delivered to Sumitomo Electric Industries, Ltd. for prototype cable construction.

We have studied the effect of preloading on the relaxation of the levitation force of bulk Y–Ba–Cu–O superconductors. The levitation force was measured by applying the external field with a superconducting solenoid. We set the standard magnetic induction ( B st ) to be 1.2 T supposing that the bulk Y–Ba–Cu–O disk experiences this field in operation. Preloading was applied such that the external field was first ramped to 1.1 B st and reduced to B st . We confirmed that the preloading was effective in reducing the relaxation of the levitation force, although the effective levitation force was largely depressed compared to that without preloading. The employment of an extra step to lower the field below B st before reaching B st was effective in improving the effective levitation force without sacrificing the ability to reduce the relaxation rate. This is qualitatively understood in terms of the field distribution inside the disk based on the critical state model.

It is known that the levitation of bulk superconductors could successfully be controlled using a Bi2223 superconducting electromagnet. In the previous study, we directly measured the levitation force and magnetic field on the surface of bulk superconductors levitated by this electromagnet. And the shielding current distribution was estimated from levitation force. But magnetic field was not in agreement with its simulation from this shielding current distribution. In the present study, we improved the formulas, and the shielding current distribution was estimated from trapped-field distribution. Experimental results of levitation force and magnetic field by the shielding current were compared with simulation from this shielding current distribution.

This paper presents the design and performance results of a pair of 6-kA high-temperature superconducting (HTS)-copper current leads, in which, over a short length at the warm end (e.g., 77 K) of each HTS section, comprised of paralleled Bi-2223/Ag–Au tapes, is operated in the current-sharing mode. Because of their reliance on vapor cooling, the leads are applicable only to liquid helium-cooled superconducting magnets such as those used in high-energy physics accelerators and fusion machines. The experimental measurements have demonstrated that key performance data of the new 6-kA HTS-copper leads agree reasonably well with those expected from design.

The high temperature superconductor (HTS) has very high J c value and can trap the very high magnetic field. The YBa 2 Cu 3 O x (YBCO) bulk superconductor is one of these attractive HTS and expected for many applications. In our previous study, to estimate the shielding current distribution in levitated YBCO bulks, we directly measured the levitation force and values of magnetic field on the surface of levitated YBCO bulk superconductor by using load cell and Hall probes. In this paper, we proposed shielding current distribution based on the axisymmetric 3-dimensional Bean-model and calculated the levitation force and magnetic fields on the surface of bulk superconductor. These calculated results were compared with experimental data.

The unique features that distinguish a high-temperature superconducting (HTS) magnet from a low-temperature superconducting (LTS) magnet are: (1) operating temperature and (2) temperature span over which the magnet remains superconducting. Thus, for two magnets, one HTS and the other LTS, both satisfying the same set of field specifications, the operating temperature and the operating range of temperature of the HTS magnet are generally an order of magnitude greater than that of the LTS magnet. In this CHATS 2002 Workshop, four issues, i.e., stability, protection, cryogenics, and economic issues will be examined as they are impacted by the two unique features, first as a general introduction to this presentation, between HTS and LTS magnets. This is followed by discussion of economics of large-scale HTS devices, specifically of fusion; electric power devices; high-energy physics; research-purpose NMR; and high- B DC magnets. The paper concludes with a presentation of preliminary results of stability/protection research presently being performed at the FBML.

Potential performance advantages of a solid nitrogen cooled “permanent” high-temperature superconducting (SN2/HTS) magnet system over a liquid helium cooled low-temperature superconducting (LHe/LTS) system are explored. The SN2/HTS system design includes a second solid heat capacitor that cools a radiation shield. Recooling of the heat capacitors is performed with a demountable cryocooler. The SN2/HTS system offers both enhanced stability and improved portability over a LHe/LTS system. Design codes are constructed to compare the SN2/HTS system design with a LHe/LTS design for a general permanent superconducting magnet system employing a room temperature bore. The codes predict the system volume and mass that should be expected for a given set of design requirements, i.e. field strength and bore size, and a given set of conductor properties. The results indicate that present HTS conductor critical current and index are not yet sufficient for producing SN2/HTS systems of a size that is comparable with that expected for a LHe/LTS system; however, the conductor properties of Bi2223/Ag have been consistently improving, and new HTS conductors are expected to be developed in the near future. The codes are used to determine the minimum Bi2223/Ag conductor performance required for a SN2/HTS system to be competitive with a LHe/LTS system.

This paper presents results of a quench/recovery experiment for a BSCCO-2223/Ag composite tape in the presence or absence of a thin layer of solid nitrogen on each side of the tape. Voltage and temperature data were recorded for a 20-cm long BSCCO-2223/Ag tape operating in the range 20–55 K and subjected to a heat pulse of a 10–600 s duration applied over a short distance at its midpoint. The data clearly show that solid nitrogen is beneficial to the stability of high-temperature superconductors operating in this temperature range and subjected to transient heating disturbances.

A new design for a portable “permanent” superconducting magnet system is explored. The design involves a persistent-mode high-temperature superconducting (HTS) magnet that is cooled by a solid heat capacitor. The system is an alternative to permanent low-temperature superconducting (LTS) magnet systems where the magnet is cooled by a bath of liquid helium. An apparatus was constructed to demonstrate stable operation of a permanent magnet wound with Bi2223/Ag conductor while in thermal communication with a mass of solid nitrogen. The apparatus includes a room-temperature bore and can function while it stands alone, detached from its cooling source, power supply, and vacuum pump. The magnet is operated in the 20–40 K temperature range. This apparatus is the first to demonstrate the operation of a superconducting magnet with a permissible temperature variation exceeding a few degrees kelvin. Models are developed to predict the experimental system's warming trend and magnetic field decay. The models are validated with a good agreement between simulations based on these models and experimental results. The results indicate that present HTS conductor critical current and index are not yet sufficient to provide field strengths and field decay time constants that are required for typical persistent-mode applications.

This paper presents experimental and analytical results of a pair of helium vapor-cooled brass leads having a rated current, It, of 40 A, operated continuously below, at, and above 40 A. In Regime 1 (0

This paper describes the basic concept of and two microtesla flux measurement techniques for microampere flux pumps to be used in superconducting NMR magnets. Under construction presently at FBML are a prototype flux pump and an LTS/HTS NMR magnet, which will be coupled to a full-scale flux pump to be built after completion of the prototype. The two measurement techniques, independent and complementary, permit quantification of flux increments pumped into a magnet coupled to a microampere flux pump.

This paper describes a high-temperature superconducting (HTS) current lead, in which, over a short length at the warm end (e.g., 80 K), the lead, comprised of paralleled BSCCO-2223/Ag–Au tapes, is operated in the current-sharing mode. Because our leads rely on cooling by the effluent helium vapor, they are applicable only to devices such as superconducting magnets that operate in a bath of liquid helium. The current-sharing mode operation results in a significant saving of the superconducting materials (BSCCO, Ag, and Au) needed to construct our leads as compared with a “fully-superconducting” HTS lead of the same configuration that is rated for the same transport current and operates over the same temperature range. Detailed analysis of our four leads, each rated 6-kA and comprised of 240 paralleled BSCCO-2223/Ag–Au tapes, demonstrates that the cold-end heat input of each of our leads is equal to or can be optimized to be smaller than that of a completely superconducting counterpart comprised of 480 paralleled tapes. The paper also shows that a protection criterion under fault mode or heat exchange consideration ultimately determines the extent to which the material saving can be achieved in our leads.

We present Part 4 results of a comprehensive theoretical study of an ‘‘electromaglev’’ system, in which a high-temperature superconducting bulk YBCO sample is levitated stably in a DC magnetic field generated by magnet system underneath the floating object. An electromagnetic analysis, based on a three-dimensional finite element technique (FEM) applied to the current vector potential method, has been developed to determine the supercurrent distribution in a field-cooled (and hence trapped-flux) YBCO disk that levitates stably in a magnetic field generated by the magnet system. The supercurrent distribution thus determined was in turn used to compute trapped-flux-induced field profiles of the disk and predict a ‘‘levitation current’’ in the magnet system at which the disk, initially resting on a support plate, begins to levitate. Agreement between computed field profiles and levitation currents and those measured in the experiment was excellent, validating the analysis itself and the method used to derive solutions. The analysis demonstrates that the supercurrent distribution within a trapped-flux disk is far more complicated than that derived from the Bean model for a long cylinder under a uniform axial magnetic field. It is used for a parametric study of the eAects of disk dimensions (radius, thickness, radius/thickness ratio) and trapped-flux strength on supercurrent distribution and lift-to-weight ratio. The magnitude of the Br component generated by the magnet system is very important for lift and it is shown that thinner disks rather than thicker disks can improve lift-to-weight ratio. Because accuracy of the analysis is disk-size independent, small disks are time-eAcient for performing the analysis. ” 2000 Elsevier Science Ltd. All rights reserved.

We present Part 2 results of a comprehensive study, both theoretical and experimental, of an electromaglev (`active-maglev') system, in which a YBCO bulk sample is levitated stably in a DC magnetic field generated by a magnet system located underneath the floating sample. The 0th-order theory presented in Part 1 is used to interpret experimental results of levitation stability, lift-to-weight ratio, and lateral stiffness for four YBCO samples: 1) disk; 2) annulus; 3) the same annulus with a permanent magnet disk placed in the hole; and 4) ring. Also presented is a procedure to induce in a sample a trapped flux that is shown experimentally to be essential for the sample to levitate tilt-free. Trapped flux is another requirement for levitation stablity; it is in addition to at least two degrees of freedom required on spatial supercurrent flow and certain spatial profiles imposed on the field generated by the magnet system.

We present experimental and analytical results of quenching for three-coil systems wound with multifilamentary Nb–Ti composites, demonstrating that AC loss, due mostly to interfilament coupling in the composite during quenching, promotes normal-zone propagation (NZP). The experiment used two three-coil systems, each comprised of inner, middle, and outer coils connected in series and shunted with resistors. Both systems shared the same outer and middle coils. Two versions of the inner coil were used, both versions wound with composites that were otherwise identical except for twist pitch lengths. One composite's twist pitch length was chosen sufficiently short to make the interfilament coupling loss to be negligible during quenching, while the other composite's twist pitch length was chosen sufficiently long to make the loss significant. Our model for quenching analysis includes the coupling loss and has been used to simulate the experiment. Agreement between experiment and simulation is excellent. The computed temperature distributions during quenching show a clear effect of heating by coupling loss in the inner coil wound with a composite having a long twist pitch length and virtually no effect of heating in the inner coil wound with a composite having a short twist pitch length. The difference in quenching behaviors between the two systems, different only in twist pitch length in the composites for the inner coils, is so compelling that we conclude that interfilament coupling loss is the key to this difference. We may further conclude that coupling loss may be used to benefit protection of multicoil systems. The twist pitch length that controls this coupling loss is thus a key design parameter; it is best determined through simulation.

We present results of a comprehensive study, both theoretical and experimental, of an ‘electromaglev’ system, in which a high-temperature superconducting bulk sample, e.g. YBa 2 Cu 3 O 7 − δ (YBCO), is levitated stably in a DC magnetic field generated by electromagnets placed underneath the floating object. Results of the zeroth-order theory agree quite well with experimental results on lift-to-weight ratio and suspension stiffness for three bulk samples: (1) a solid YBCO disc; (2) a YBCO annulus; and (3) a YBCO annulus with a neodymium-iron-boron (Nd-Fe-B) permanent magnet disc (PMD) filling the centre. The experiment has also verified the need to satisfy two requirements to achieve stable levitation with a DC magnetic field only: (1) the spatial flow of the supercurrent in the sample must have at least two degrees of freedom; and (2) the electromagnets must generate a magnetic field profile that satisfies spatial requirements for lateral and pitch stability. A permanent magnet disc has only one degree of freedom for its Amperian current, thus it cannot be levitated stably in this system; the experiment has also demonstrated that an HTS solenoid (wound with silver-sheathed BSCCO-2223 tape) cannot be levitated stably, because the solenoid supercurrent flow is also restricted to the azimuthal direction only. The zeroth-order theory together with the Bean model shows that the supercurrent induced in a YBCO sample is independent of the critical current density, J c , of the material but is directly proportional to the axial component of the field and that the lift of the sample is directly proportional to the product of the axial and radial components of the magnetic field generated by the electromagnets.

An experimental study was carried out on the tribological behaviour of materials of interest in cryogenic applications, focusing on diamond and graphite. Both natural diamond (referred in the text as diamond) and chemical-vapour-deposition (CVD) diamond (CVD-diamond) were used. The experiment was carried out using a pin-on-disc tribometer capable of operating at cryogenic temperatures, from 4.2 to 293 K. Two basic scenarios of testing were used: (1) frictional coefficient (μ) versus velocity ( v ) characteristics at constant temperatures; and (2) μ, versus temperature ( T ) behaviour at fixed sliding speeds. For diamond, graphite, and stainless steel pins, each set of pins against CVD-diamond disc, values of μ are virtually velocity independent. For each of diamond, alumina, and graphite pins, each set against graphite disc, the δμ/δv characteristic is favorable i.e. positive. For diamond and graphite pins against CVD-diamond discs, values of μ are nearly temperature independent in the range 77–293 K. Each μ( T ) trace for pin materials sliding on graphite discs has a peak at a temperature in the range 100–200 K. As is the case with diamond/stainless steel pairs, there is a definite correlation between friction and material hardness: μ is inversely proportional to material hardness. Thus μ is higher at higher temperatures, because hardness generally decreases with temperature.

We present the results of an experimental and analytical study of two-dimensional normal zone propagation in pancake test coils, wound with silver-sheathed BSCCO-2223 tapes. Two test coils were studied in detail, one having three and the other eight layers. Each test coil was housed in an adiabatic environment whose temperature (20–70 K) was controlled and maintained by a two-stage G-M cryocooler and placed in a background field (0–6 T) generated by a Bitter magnet. With a test coil carrying a transport current (0–200 A), a local heat disturbance was applied by a heater attached to the outermost layer of the coil. The resulting electrical and thermal responses of the coil were recorded with voltage taps and thermometers attached to the coil. A normal zone propagation code was developed to accurately simulate the voltage and temperature responses of each coil for both quenching and recovering events. The code solves the nonlinear transient heat diffusion equation in two-dimensional cylindrical coordinates with a finite difference method. As an application of this code, a two-coil system, with each coil comprised of one double pancake wound with silversheathed BSCCO tape, was studied for its quench behaviour as one of the coils was driven normal locally. The simulation results indicate that the value of a shunt resistor connected across the terminals of each coil has a profound effect on the level of hotspot temperature reached in the quench initiation spot.

A small coil, wound with a multifilamentry Nb3Al conductor, was tested over the temperature range 4.2–12 K. The experiment not only encompassed measurements of the conductor's critical currents but also the coil's performance, e.g. 4.8 T (180 A) at 4.2 K and 3.2 T (120 A) at 12 K, and the winding's turn-to-turn normal-zone propagation (NZP) velocities. The coil's stable performance of 120 A at 12 K nearly matched the conductor's short-sample critical current. The coil's terminal voltage and current have also been measured and found to agree well with results of simulation incorporating measured NZP velocities.

The friction coefficients for copper pairs lubricated with fatty acids and fluorinated fatty acids have been measured over a wide range of sliding speeds and temperatures. Sliding speeds in the range 10−7−10−2 m s−1 and temperatures in the range 4.2–300 K were used. The friction coefficients near 300 K are generally low and increase with sliding speed, while the friction coefficients at low temperatures are markedly higher and relatively independent of velocity. Each lubricant's friction vs. velocity behavior over the temperature range 150–300 K can be described by a friction-velocity master curve derived from a thermal activation model for the lubricant's shear strength. The activation energies deduced from this friction model are identical to those obtained in the same temperature range for a vibrational mode associated with low temperature mechanical relaxations in similarly structured polymers. These results suggest that thermally activated interfacial shear is responsible for the fatty acids' positive-sloped friction vs. velocity characteristics at low sliding speeds near room temperature.

Key aspects of the design and operation of hybrid magnets are reviewed. For a hybrid system to become a successful “facility” magnet, the system must have the following operational properties: (1) a “high” central field; (2) good experimental access; (3) can be scheduled for operation on a regular basis; and (4) readily replaceable water-cooled inserts. Because, as in most engineering problems, these requirements result in conflicting options, no solution is straightforward; solutions to one requirement must not dominate solutions to the other requirements. The paper also proposes and discusses design issues for the next generation of hybrids, in which the superconducting magnets would be wound entirely of high-temperature superconductor and operated at ∼ 20 K. Silver-sheathed bismuth-based (Bi-2223) tapes appear to be most promising for this application; new field records of 24.0 and 23.4 T have recently been established by a combination of Hybrid III at the MIT Francis Bitter National Magnet Laboratory and a test coil wound with Bi-2223 tapes operating, respectively, at 4.2 and 27 K. The paper concludes with a number of challenging problems that must be faced and overcome before Bi-2223 tapes can be used in large magnets such as those required for hybrid systems.

To study the role of liquid helium trapped within the winding of ‘dry’ superconducting magnets, energy margins were measured in a dry-winding superconducting test coil. The test coil comprised five non-inductive layers, each wound with circular cross-section multifilamentary NbTi superconducting wire. A pulsed coil, applied only through the two middle layers of the test coil carrying a transport current in a background magnetic field, was used to simulate a transient disturbance induced within the conductor over a confined region of the magnet. Measured energy margins, with the test coil immersed in liquid helium or in gaseous helium, agree well with the conductor's enthalpy densities required to drive the conductor normal. Experimental results show that the liquid helium occupying the void space within the winding has no beneficial effect against transient disturbances induced internally in the conductor; it does, however, slow down quench propagation. Because of the presence of an insulation layer at the conductor surface, the winding is effectively adiabatic against internal disturbance pulses. The trapped liquid helium is expected to be beneficial against transient heating applied external to the conductor surface, as would be the case with mechanical disturbances.

Normal zone propagation (NZP) within a small cryocooler-cooled, Nb3Sn tape-wound magnet operating at 10–12 K is analysed experimentally and theoretically. In the experiment, the NZP is induced by electric heaters attached around the magnet perimeter, and the voltages across each double-pancake coil comprising the magnet are monitored and recorded. Based on the experimental results, a theoretical propagation model is presented in this paper, using lumped-parameter linear circuit elements (i.e. inductors and resistors). A computer simulation of the model predicts transient response of the magnet terminal voltages during induced NZP, which agrees with the results observed in the experiment. The present study has been conducted in preparation for a NZP study of tape-wound high Tc superconducting pancake coils.

In a multicoil adiabatic superconducting magnet composed of nested solenoids, each shunted with a resistor, an initial quench in one solenoid can induce quenches in the other solenoids even when they are thermally isolated. The present work investigates induced quenching in such magnets. Specifically, we have identified that coupling a.c. loss is responsible for induced quenching in thermally isolated coils. An existing computer code has been modified accordingly to incorporate a.c. losses. The modified code has been tested against experimental results of premature quenches recorded for a 750 MHz NMR magnet presently under development at MIT; agreement between the results of the code and experiment is excellent. In multifilamentary composite conductors, the coupling loss produced by a time-varying magnetic field depends on the square of the filament twist pitch. The new finding implies that filament twist pitch is an important parameter for protection of high current density multicoil superconducting magnets; by properly adjusting the twist pitch of the conductor in each coil, it may be possible to make each coil self-protecting without reliance on, for example, heaters for protection.

The interpretation of cryogenic temperature friction is enhanced by the concurrent evaluation of the low-temperature mechanical properties of sliding materials. Absolute sliding stability is greatly desired but difficult to attain in cryogenic systems. Experiments performed to investigate the material factors necessary for smooth motion indicate that interfacial shear creep plays a key role in sliding stabilization. Because viscous creep is typically not observed at cryogenic temperatures, we conclude that it is not possible to guarantee sliding stability based solely on material selection criteria.

A tandem magnetic refrigerator has been developed for continuous operation between 4.2 and 1.8 K. This paper presents the design, computer simulation and construction of the prototype machine. The regenerative concept is employed in the magnetic refrigerator to obtain a cascaded magnetic Carnot cycle effect along the temperature axis from the cold end to the warm end. Entropy pumping action from the cold heat reservoir to the warm heat reservoir occurs in the active magnetic regenerator. The inherent sources of irreversibility in regenerative magnetic refrigerators, i.e. heat capacity imbalance between magnetic refrigerant and heat transport medium, helium entrainment and dead volume effect, have been minimized by design optimization of the system components. The magnetic system of the tandem refrigerator has two virtually identical units, each consisting of a gadolinium gallium garnet (GGG; Gd 3 Ga 5 O 12 ) magnetic core, a superconducting magnet, a warm end heat exchanger and a cold end heat exchanger. These components are united by a cryogenic displacer which shuttles the heat transport medium, subatmospheric 3 He gas, between the two units. The prototype of the designed magnetic refrigerator has operated continuously, producing a net refrigeration rate of 12 mW per magnetic core at 1.8 K.

The quenching process in composite high Tc and low Tc superconductors has been studied in detail. Specifically, thermal stability and quench protection issues were addressed to gain a better understanding of how high Tc superconductors may be best used in superconducting magnet applications. A series of experiments has been performed to measure quench propagation velocities - a parameter that quantifies the quenching process - along short samples (≈12 cm) of both low Tc (niobium tin) and high Tc [BiPbSrCaCuO(2223)] composite superconductors as a function of operating temperature, transport current and background magnetic induction. A computer simulation has also been developed to predict the behaviour of these conductors during a quench. The simulation includes the temperature and magnetic field dependence of material properties and generates temperature profiles that are marched out in time. Simulated voltage traces are also generated for comparison with the experimentally recorded voltage traces. The assumption that quench propagation velocities are constant, generally used in quench analysis for low Tc conductors, is also reasonable for high Tc conductors and thus, velocities may still be experimentally extracted from high Tc conductors. In general, quench propagation velocities are two or three orders of magnitude slower in high Tc conductors than in low Tc conductors. This indicates that magnets wound with these conductors are unlikely to be self-protecting. For the same reasons that they are difficult to protect, they are likely to be much more stable in response to thermal disturbances than their low Tc counterparts.

We have directly measured the levitation force of levitated bulk superconductors. The levitation force had a hysteresis loop during increasing and decreasing field processes. Such a hysteretic behavior can be explained in terms of hysteretic penetration of external field to the superconductor. Based on such magnetization, we performed numerical simulations of the levitation forces, which were fairly in good agreement with experimental data.

A normal zone propagation model has been developed for superconducting composites under adiabatic conditions. It is based on the Whetstone-Roos model, originally developed for normal zone propagation in adiabatic wires of unclad superconductor. The model takes into account the temperature and magnetic field dependent material properties, for both superconductor and matrix metal. Analytical results agree well with experimental data.

Experiments were performed to determine the friction and wear coefficients of various polymer pins slid against either AISI 304 stainless steel or oxygen-free high conductivity (OFHC) copper at temperatures of 293, 77 and 4.2 K. Three kinds of polymeric materials were tested: 1, unfilled; 2, solid-lubricant filled; and 3, fabric- and particle-reinforced plastics. The polymers' penetration hardnesses were measured at 293 and 77 K to facilitate the wear coefficient calculations. Although the polymers were much harder at the cryogenic temperatures, it was found that their non-dimensional wear coefficients remained relatively constant, generally in the range 5 × 10−7–5 × 10−6. Cross-plots of the polymers' wear coefficients versus the corresponding friction coefficients show good correlation and indicate a third- to fourth-power functional relationship between friction and wear for both the unfilled and the reinforced plastics. The magnitude of the wear coefficients and the relationship between friction and wear suggest that adhesive wear is the dominant wear mechanism for these polymers.

Applying the understanding gained through recent work in cryotribology (low temperature friction and wear) the author discusses stability issues for high performance magnets. As an illustrated example, the design basis for the dry-wound, adiabatic magnet of the 35 T Hybrid III at the Francis Bitter National Magnet Laboratory is examined.

Experiments were performed to determine the effect of operating temperature on the stability and protection parameters of a small adiabatic Nb - Ti magnet. Quench propagation velocity and minimum quench energy were measured for various combinations of temperature, magnetic field and transport current. The operating temperature ranged from 4.2 to 8 K; the transport current from 25 to 125 A; and the magnetic flux density from 0 to 6 T. An existing computer code was modified to predict the magnet's quench behaviour. The principal research objective was to improve existing analytical models of stability and protection and to use them as analytical tools in designing high T c superconducting magnets operating over a wide temperature range, up to ≈ 100 K.

This paper describes two key elements required in liquid hydrogen critical current measurements of superconductors: first, a technique to produce small quantities of liquid hydrogen and second, the design criteria for current leads.

Experiments were conducted to determine static friction coefficients for various epoxy mixes slid against copper at room temperature and at 77 K. Several of the mixes' mechanical properties were also measured as an aid to interpreting the friction results. At room temperature, mixes which exhibit a strong tendency towards viscoelastic behaviour also demonstrate a considerable time dependence in their static friction coefficients. At 77 K all of the epoxy mixes are in the glassy state, and their static friction coefficients are noticeably reduced from their room temperature values. Several Teflon ® -powder-filled mixes were also tested to examine the consequences of incorporating a friction modifier into the epoxy. The mechanical properties of the Teflon-filled mixes are similar to those of the unfilled epoxies, with the exception that the particle-filled mixes are generally much more brittle. The use of the Teflon filler has the additional effect of reducing the friction coefficients for the epoxy mixes to values comparable to that of bulk Teflon.

A superconducting magnet, consisting of concentric Nb 3 Sn and NbTi coils, will operate in a bath of subcooled helium at 1.8 K in a quasi-adiabatically stable mode. Water-cooled 10 MW insert magnets which fit within the 36 cm cryostat bore are based upon a design which has proven very successful. The system will provide researches with ultrahigh continuous magnetic fields.

A theoretical analysis of operating temperature dependent normal zone propagation (NZP) has been performed in an adiabatic superconducting magnet over the temperature range 4.2–80 K. Because of the inverse dependence of NZP velocity on winding heat capacity, which increases sharply with temperature, above ≈ 20 K the normal zone hardly grows beyond its initial size and the hot spot temperature reaches well above 300 K. A computer simulation of NZP was applied to a small, two-coil solenoid system that was assumed to operate at any temperature between 4.2 and 80 K. The results corroborate the conclusion of the analysis: that the solenoid system is self-protecting for operating temperatures up to ≈ 20 K. This ‘operational critical temperature’ up to which an adiabatic magnet is self-protecting depends also on matrix current density and magnet size.

A three-lead test rig with a built-in neon condenser has been constructed to measure fielddependent critical currents of high T c superconductors (HTSs) at 27 K. The test rig can accommodate two test samples and the current leads are good up to ≈ 250 A. This paper describes the test rig assembly and the experimental procedure for producing small quantities (≈ 100 cm 3 ) of liquid neon. The test rig has been used successfully to measure 27 K critical currents of HTS coils and short samples in background magnetic fields up to 20 T.

Large single-grain bulk RE–Ba–Cu–O (RE: rare earth elements) superconductors can trap large fields exceeding several teslas and thus can function as very strong quasi-permanent magnets. We have found that the resin can penetrate into a bulk superconductor, when the sample was immersed in molten resin. Hence, resin impregnation was effective in improving mechanical properties. Three bulk superconductors‘ annuli with resin impregnation, each 50-mm i.d. and 80-mm o.d. was built and energized, by a field-cool method, to generate, in a bath of liquid nitrogen, a persistent trapped field of 1.62 T.

This paper presents magnetoresistivity data for silver used in silver-sheathed BiSCCO tapes over the temperature range 4.2 – 159 K for magnetic inductions up to 14 T.

The irreversibility lines of BiPbSrCaCuO silver sheathed tapes were measured at temperatures from 33 K to 110 K. It was determined that the irreversibility magnetic field, H irr (81 K) did not change with the sweep rate of magnetic field from 0.16 to 0.8 T/min. It was found that the irreversibility line was improved to higher magnetic field and temperature with the increase of J co . This suggests that the irreversibility line has a strong relationship to the morphology of BiPbSrCaCuO.

A ‘critical-current-margin. (CCM) design criterion for superconducting magnet stability is presented. It is applicable to high-performance magnets that operate at high current densities; these magnets have cooled conductor surfaces, but their calculated heat fluxes are well beyond recovery values. For stability of a magnet based on this CCM design criterion, disturbances must be limited in size but need not be eliminated nor localized. The CCM criterion is compared with the MPZ/Cold-End Theory. It is shown that this criterion offers a much higher confidence level of stability than the MPZ/Cold-End Theory. Furthermore, and unique to this theory, stability is independent of current density in the stabilizer. The CCM criterion is based on two magnet winding elements: heat transfer and composite conductor. In this criterion heat transfer must always remain in the nucleate boiling region throughout the operation of a magnet. The design emphasis is on more superconductor and less stabilizer. It is also demonstrated that the CCM theory works particularly well with conductors such as cabled conductors that have high ratios of cooled perimeter to cross-section. Experimental results are presented supporting this CCM design criterion.

A 250 dm 3 subcooled cryostat system for superfluid helium was designed and built for a hybrid magnet. The cryostat, which houses a 430 mm bore, 12 T NbTi superconducting magnet, has a refrigeration capacity of ≈20 W at 1.8 K.

The theoretical correlation between shear stress and epoxy resin fracture developed in Part 2 was verified experimentally using a series of epoxy-impregnated, thin-walled superconducting test coils. A Lorentz shear stress field was created within the winding of each coil by placing it in a constant background magnetic field and energizing it with transport current, which was increased slowly at a constant rate. Because the boundary condition at each end of the test coil critically influenced the Lorentz shear stress field, which caused epoxy resin fracture, premature quench data were measured for test coils with different combinations of end boundary conditions. In test coils with both ends rigidly clamped, cracks occurred as transport current was increased; during a training sequence the test was terminated by a premature quench. Using acoustic emission and voltage signals, each premature quench was linked directly to a crack occurring near one of the ends. Test coils which had both ends unsupported, giving the winding freedom to expand radially, did not experience epoxy fracture and showed no premature quenches: these reached critical current at the first attempt.

The transient stability of a NbTi, cable-in-conduit conductor cooled with supercritical helium was investigated experimentally by inductively heating the conductor for short-time durations. Data were gathered for zero and nonzero imposed flow rates. In addition, some of the zero-flow measurements were taken while the conduit was surrounded by an external vacuum as opposed to the situation where the exterior wall of the conduit was in contact with atmospheric pressure liquid helium. Analysis was carried out by deriving a heat transfer coefficient having both transient and quasi-steady-state components. Results indicate that the theoretical heat transfer coefficients used in most computational studies are inadequate and emphasis should be placed on the rate of cooling between the conductor and the helium rather than on the helium heat capacity; an imposed flow rate of limited magnitude has the greatest effect in the region of multiple stability; an increase in the conductor surface area brings about an increase in both the stability margin and the limiting current; the presence of liquid helium external to the conduit has little or no effect on transient stability.