Except for its size, this commercial low-temperature superconductivity magnetic seperator is nearly identical to its high-temperature counterpart now in development. Hydraulically driven canisters shuttle through the cryostat.

After almost two years of development, DuPont's superconductivity business is finishing the detailed design of a pilot separator for processing kaolin clay and titanium dioxide ore. The separator will incorporate the largest high-temperature superconductivity magnet, in terms of combined size, magnetic field, and stored energy, that has ever been built. The 0.8m-diameter magnet will deliver a 2.5 tesla magnetic field and store 800 kilojoules of energy.

The program has been a joint effort with Sumitomo Electric Industry, the Osaka, Japan, fabricator of conventional copper wire and wire and tape for low- and high-temperature superconductivity.

As the world's largest supplier of titanium dioxide, DuPont has a keen interest in improving ore production.

Kaolin and titanium dioxide are the white pigments used in paint, paper, and plastics manufacturing. However, as ores they contain magnetic contaminants, such as hematite (Fe₂ O₃) and magnetite (Fe₃ O₄), which discolor the pigments, turning bright white into off-white. Kaolin mineral deposits also contain ilmenite, a discoloring compound of titanium dioxide and iron.

Most ore processing is done on-site, using either traditional copper or low-temperature superconductivity magnetic separators. Copper separators consume vast quantities of electricity, which is not a problem in Georgia or Florida where electricity is relatively inexpensive and readily available. However, newer mining sites in the Amazon jungle, China, and other locations usually lack quality power. And, if they do have it, it's very expensive.

Low-temperature semiconductivity magnetic separators shrink power consumption, but require liquid helium—something that remote suppliers can't readily deliver and with which local operators have had little experience.

High-temperature superconductivity, or HTS, can mirror the low power consumption of its low-temperature counterpart, but does it at a higher operating temperature—20 K versus 4 K, the boiling point of liquid helium. Since Carnot efficiency dictates that the higher the temperature, the easier it is to remove the heat, a higher operating temperature translates to a simpler, lower-cost refrigeration system. It's so simple that the equipment doesn't even need liquid cryogens.

"Our design uses only electricity," said Chris Rey, magnetics program manager for the DuPont superconductivity business in Wilmington, Del.

The heat load is handled by a two-stage, closed-cycle Gifford-McMahon cryo-cooler, a compact refrigeration system using a small amount of helium gas, a compressor and regenerator, plus a cold head to deliver refrigeration at temperatures low enough to reach superconductivity.

The system processes ore in batch mode by pumping unprocessed clay slurry through a nonmagnetic, 300 series stainless steel pipe. The pipe contains an in-line canister filled with a 400 series "soft magnetic" stainless steel-wool filter. Encircling the canister is an 0.8m high-temperature superconductivity solenoidal magnet wound from silver-sheathed bismuth-strontium-calcium-copper-oxygen (Bi-2223) powder-in-tube multifilament tape.

Drawing shows 1.8m x 1.8m x 1.5m rectangular iron yoke wrapped around the 0.8m- diameter high-temperature superconductivity magnet and batch mode canister assembly.

As the unprocessed slurry passes through the canister, the magnetic field is cycled on. When the magnet is energized, the magnetic contaminants are attracted to and trapped by the steel-wool filter, while the nonmagnetic particles pass through the filter and are collected as processed product. When the magnet is cycled off, the magnetic force drops, releasing the contaminants, which are then flushed out with water.

The cycle repeats with a batch every 10 to 20 minutes, depending on the processing volume and the magnetic fraction of the clay. Kaolin typically has a magnetic fraction of 2 to 5 percent, while titanium dioxide can measure as high as 50 percent. Even within a dig site, which may cover a half-mile or more, the magnetic fraction can vary widely and require adjustment of the duty cycle. A typical filter will last for approximately 8,000 hours of continuous operation.

Functionally, the HTS system operates the same way as a conventional copper or low-temperature superconductivity system, except that it uses a high-temperature magnet to generate the magnetic field. That small exception creates a host of tough technical and economic challenges.

"Any time we've applied HTS to magnets, there have been challenges because we are pushing the envelope," said Steve Van Sciver, director of magnet science and technology at the National High Magnetic Field Laboratory in Tallahassee, Fla. "The materials are difficult and every magnet is unique."

The separator's high-temperature superconductive coil consists of a stacked double pancake structure in which the pancakes are spliced together to form a continuous conductor. The assembly mounts inside the toroidal vacuum cryostat held inside an iron yoke.

"Scaling up from the 0.05m-diameter lab model to the 0.8m- diameter pilot magnet is a giant step for HTS," Rey said. "Some people doubt that you can scale like that. We think we can."

The size was not chosen arbitrarily; it matches the processing volume of the smallest commercial magnetic separator. Because processing volume is proportional to the radius squared, a smaller magnet would reduce capacity significantly.

There are internal forces in a magnet trying to pull it apart. As a magnet scales in size, the forces will increase proportionally, much like increasing gas pressure pushing on the walls of a container. Resisting these forces is made difficult by the relatively low tensile strength of the superconductive tape in the windings.

In high-temperature superconductivity magnets, two types of tensile strength are measured. One is the traditional tensile strength familiar to all mechanical engineers—the point at which the material will break under tension. The other is tensile strength as a function of critical current (J_c), the amount of superconductive current that an HTS wire will carry prior to going normal, which is to say, to lose superconductivity. It is a function of the ambient temperature, magnetic field, and stress in the wire. If the tensile stress is too high, it will reduce the critical current.

HTS magnet wire consists of ceramic materials packed in metallic tubing. The ceramics act as both structural and current-carrying members. If the ceramic cracks under tension, it no longer superconducts and the magnet converts to its normal resistive state, which can have serious consequences.

DuPont is addressing the strength issues with a mechanical support structure that reinforces and supports the windings and does double duty as part of the conduction cooling process. The construction details are proprietary, but combining functions helps to keep costs down, according to Rey.

The high cost of the HTS tape used to wind the magnets is a continuing concern. The problem is exacerbated by the tape's very low critical current density. Low current density means extra turns are needed to generate the same magnetic field as an equivalent low-temperature superconductivity magnet. More turns translate to more money.

Tests have shown that an iron yoke surrounding the coil assembly will concentrate the magnetic flux lines toward the inside diameter of the magnet where the canister is located, thereby reducing the amount of HTS conductor required to reach the desired field strength.

DuPont believes that expanded HTS usage will help drive down the basic tape costs. "When LTS wire and tape were first introduced, they, too, were costly," Rey explained. "As part of the project, we are forecasting where HTS costs are headed." The company is not depending on volume alone to slash costs; it has a materials development program to address this concern.

Another challenge for HTS magnets is ac losses, both in the form of magnetic hysteresis and heat-generating eddy currents that add to the refrigeration load. At present, reducing these heat losses represents the biggest hurdle in the commercialization of a batch HTS separator with a cyclical magnetic field.

"You have to make sure the heat can be dissipated before it significantly increases the temperature of the magnet," Van Sciver pointed out. "Otherwise, the magnet becomes partially resistive and eventually quenches." To quench is to lose superconductivity and, with it, the contained magnetic field.

Finally, the design has to address potential cyclical fatigue induced by the repeated magnetic cycling.

"We think we have a good handle on the fatigue issues, and we are designing the system to handle over 200,000 magnetic cycles," Rey said.

While the cyclical magnetic field, or batch, design moves forward, DuPont is hedging its bets with a second design effort based on what it calls "reciprocating canisters."

"Until 1990, most magnetic separators used a batch-type canister," Rey said. "Now the split is roughly 50-50 between batch and reciprocating designs. It is not obvious which technology will be the most marketable in the next century, so it makes sense to work on both."

The reciprocating canister system is a three-year project proceeding under the auspices of the U.S. Department of Energy's Superconductive Partnership Initiative. The $6.4 million development cost is split 50-50 between DOE and DuPont. DOE may seem like a strange partner for mining equipment development, but advances in HTS technology are bound to spill over into the power applications that are near and dear to DOE's heart.

Providing technical support to DuPont are its subcontractors, the National High Magnetic Field Laboratory, which is working on the magnet and cryostat design, and the Carpco Division of Outokumpu Technology Inc. of Jacksonville, Fla., a leading supplier of low-temperature superconductivity reciprocating separators for mineral processing. Carpco will build the physical separation equipment.

The reciprocating canister system uses a constant magnetic field and at least two movable canisters, one of which is positioned in the slurry flow stream. When the flow stream canister becomes full of magnetic contaminants, the flow is temporarily stopped and the canister is physically extracted from the magnetic field by either hydraulic force or rack and pinion drive. Simultaneously, the replacement canister is pulled into the field. While the second canister processes the material, the extracted canister is flushed with high-pressure water.

Reciprocating canisters are typically more complex to fabricate than a fixed canister batch system, but the design offers other major advantages. There is no cycling of the magnetic field, so there are no ac losses, a major source of heat. In addition, the constant magnetic field design eliminates any possibility of cyclical fatigue. Because the system operates semicontinuously, more ore can be processed for a given processing volume. Conversely, the same amount of ore can be processed by a smaller magnet diameter.

Even though the constant magnetic field design will use a smaller magnet (0.2m diameter) than the cyclical batch version, James Daley, manager of DOE's superconductivity program, believes the development will advance the knowledge base for larger-bore HTS magnets, in general.

The smaller magnet will handle a proportionally larger magnetic field with no increase in magnetic force on the structure. Also, the smaller magnet is less susceptible to magnetic instabilities that can be generated by something as simple as movement of the windings. These instabilities can lead to quenching, when all the electromagnetic energy is converted into heat.

The challenge is to handle quenching so that no permanent damage occurs, and the magnet can be recooled and started up again. That requires careful design, Van Sciver said.

DuPont's goal is to produce a low-cost, reliable, easy-to-maintain HTS magnetic separator that will meet the commercial needs of remote ore processing operations. Though no decision has been made on who would market any commercial system that results from this effort, Carpco admits it might be a likely candidate.

A successful HTS separator may have a potential future beyond kaolin and titanium oxide processing. Carpco's president, Frank Knoll, feels that there are other mineral applications worth investigating. In addition, some research work has been done in environmental remediation areas.

"It has been experimentally verified in the lab that it is possible to remove uranium oxide compounds from contaminated soils and biological contaminants from waste water," Rey explained.

Perhaps the most interesting spin-offs from these efforts are currently unknown. Pushing technological boundaries can open unexpected doors, and what is learned in mining may pay off in microchips.

home | features | news update | marketplace | departments | about ME | back issues | ASME | site search

© 1999 by The American Society of Mechanical Engineers