Bearings that support shafts with electromagnetic force have been around for a long time. But the aerospace industry was the only user of these friction-free devices until the early 1980s. Then magnetic bearings began to appear on large turbomachines that process chemicals and natural gas. Though these devices run faster and longer than conventional bearings, control costs have kept them out of most manufacturing applications.
In the 1990s, designers broke the cost barrier by switching from analog to digital controls. As a result, magnetic bearings became practical for smaller, specialized equipment such as vacuum pumps, used in manufacturing semiconductors, and machine tool spindles. As controls continue to get better, and costs fall with increasing production, magnetic bearings are likely to find a home in even more applications.
An uplifting approach
Magnetic bearings perform the same functions as conventional oillubricated bearings on rotating equipment. Unlike their counterparts, though, magnetic versions support rotating shafts without physical contact. This lack of contact virtually eliminates friction losses, letting machinery run at previously unattainable speeds. Without friction, the bearings don't need lubricants, which can contaminate the operating environment.
The operating principle is simple: electromagnets in a bearing create a magnetic flux that levitates or suspends a shaft and lets it rotate freely in air. The system has three basic elements: a stationary bearing (called an actuator) that surrounds a rotating shaft, position sensors that measure radial movement of the shaft, and a controller.
Each bearing contains electromagnets with north and south poles oriented around the inner diameter. A small air gap separates the bearing and shaft radially. Applying electric current to the magnets creates a magnetic flux strong enough to suspend the rotating shaft within this air gap, so it doesn't contact the bearing.
Proximity sensors adjacent to the magnetic poles monitor radial shaft displacement in the X and Y axes (horizontal and vertical directions), usually by measuring inductance of the air gap between shaft and bearing.
Using shaft position data from the sensors, a controller calculates the forces needed to support the shaft and keep it centered within the bearing. Then it sends signals to power switching amplifiers (usually two for each axis), which feed enough current to the magnets to generate these forces in the form of electromagnetic flux. The controller repeats this process at least 10,000 times each second, continually monitoring and adjusting shaft position.
Most magnetic bearing systems incorporate two radial bearings to support the shaft, plus one thrust bearing to accommodate forces in the axial direction. A thrust bearing consists of two fixed components (stators), oriented on either side of a disc-shaped rotor attached to the shaft. Electromagnets in the stators generate magnetic flux to keep the shaft positioned lengthwise. A proximity sensor measures axial displacement of the shaft and sends data to the controller in the same way as the radial bearing sensors.
Digital control is key
The control system for a magnetic bearing consists of an analog or digital controller and a power supply (amplifiers). Analog systems have been around for 30 years, but digital systems are rapidly replacing them because they cost less and take less space. These lower-cost digital controls are making it easier to move into smaller industrial equipment such as machine tool spindles.
An analog controller typically consists of five circuit boards, whereas a digital controller may use only one digital signal processor (DSP) and cost about 1/10th as much.
A controller contains information on bearing characteristics (stiffness and damping) used to stabilize the suspended shaft over its operating range. An analog system provides this data in hard-wired control circuits that are custom designed for each application. Digital systems, on the other hand, hold the information in software, which is easier to reprogram.
Continue to next page
By constantly sensing and adjusting shaft position, a digital control system compensates for dynamic instabilities and vibration not only in the shaft but throughout the entire system. Reducing vibration caused by cutting forces in a machine tool, for example, increases the smoothness of machined parts and increases tool life.
The controller can also be programmed to counter specific events. For example, as the shaft passes through its critical speed, the controller may relax bearing stiffness (forces holding the shaft centered) so that less vibration is transmitted through the bearings into the machine.
In another case, an automatic tool changer that handles different size tools may cause dynamic imbalance of a rotating shaft. The control system can be set up to sense these changes and adjust for them on-the-fly. This capability eliminates the need to balance most tools and tool holders for speeds up to about 40,000 rpm.
When part of a tool breaks off, the controller can respond to vibration above a certain level and quickly stop the machine to minimize adverse effects.
The first commercial magnetic bearings were installed on large industrial turbomachines (50 to 29,000 kW capacity) in Europe, Asia, Canada, and the U.S. starting about 1985. More than 200 are operating in such applications as natural gas compressors for pipelines and cryogenic expanders for oxygen and ethylene production. Cryogenic expanders operate at temperatures as low as 2150oC, which would instantly freeze any oil leak, clogging the machine. Clearly, the elimination of lubricants with magnetic bearings was a key factor.
As the technology got better, applications expanded into smaller equipment, especially turbomolecular pumps (typically less than 1 kW) that create vacuum environments for manufacturing semiconductor chips. Because magnetic bearings operate without oil, they keep the environment cleaner. About 50,000 of these pumps have been applied worldwide.
Other installations include about 250 machine tool spindles ranging from 1 to 50 kW and operating at speeds from 15,000 to 60,000 rpm. These spindles incorporate either ac induction or dc brushless motors.
The higher speeds made possible by magnetic bearings help machine tool spindles produce parts faster. Some of them run at 40,000 to 45,000 rpm to form grooves in copper tubes used for heat exchangers. Others are found in grinding machines that finish automotive gears at 30,000 to 180,000 rpm, and milling machines that generate thin sections in aluminum parts at 40,000 to 70,000 rpm. Light-duty spindles (less than 1 kW) for textile plants represent a small but growing application area for magnetic bearings.
You won't find magnetic bearings in conventional electric motors yet. But the similarity of a magnetic bearing and shaft to the stator and rotor in an electric motor may change the way tomorrow's motors work. Researchers believe that magnetic components may eventually perform two functions in a motor – supporting a shaft and supplying motion.
Meanwhile, many machine tool spindles and turbomolecular pumps incorporate magnetic bearings and dc brushless motors. Some large turbomachines take a similar integrated approach. A compressor for a Columbia Gas pipeline in Maryland comprises a 10,000 rpm adjustablespeed motor that drives overhung impellers at both ends. Natural gas cools the motor and bearings, so no other lubricant is needed.
Magnetic bearings also provide support for a 3,500-hp (2.6 MW) Reliance Electric motor for the Orange and Rockland utility in New York State. The motor bearings support an 11,000-lb shaft rotating at 900 rpm.
Information for this article was provided by Chet Farabaugh of S2M America, Roanoke, Va., and Geoff Clark of Revolve Magnetic Bearings Inc., Calgary, Alberta, Canada.
Continue to next page
The big attraction of magnetic bearings is that they eliminate contact between the fixed bearings and rotating parts (shaft) of a machine. Noncontact operation offers several benefits:
• There is no frictional wear or heat buildup, resulting in longer life than with conventional bearings. Service life is generally limited only by the control components.
• Machines run faster and with less vibration, producing more parts per hour. For example, one type of machine tool spindle operates at 45,000 rpm, whereas similar units with conventional bearings are limited to 20,000 rpm.
• Without friction, magnetic bearings don't need lubrication. This suits them for equipment operating in "clean" environments, such as refrigeration compressors, natural gas turboexpanders, and vacuum pumps used in semiconductor manufacturing. Lubrication systems are eliminated as well, cutting cost and saving space.
• Magnetic bearings function in severe environments such as in a vacuum or in steam, and in temperatures ranging from -250 to 450oC.
• Power losses due to friction, lubricant shear, and lubricant pumping equipment, are eliminated. The only real power consumption consists of resistance losses in the bearing coils and eddy current losses in the bearing laminations.
If a magnetic bearing system fails, magnetic flux may no longer support the shaft. When that happens, a set of backup bearings (also called auxiliary or touchdown bearings) catches the spinning shaft, preventing it from dropping onto the main bearings and damaging their surfaces. Most backup units are either plain or ball bearings that can handle several touchdowns, usually 5 to 20, from full speed. That is usually enough because such failures are rare.
Power interruptions that occur when the power supply fails or when a machine is intentionally shut off are handled differently. A backup battery provides enough current to levitate the shaft until the machine coasts to a stop. After a sensor detects no further shaft rotation, backup power shuts off and the shaft descends onto the backup bearings.
Battery power can also be used to levitate a shaft before turning a machine on to prevent wear caused by shaft rotation. Or it can support the shaft for 10 to 20 min after a power failure to permit an orderly shutdown.
he radial air gap between a shaft and main bearing usually ranges from 0.014 to 0.03 in. depending on size, whereas the gap for the backup bearing ranges from 0.007 to 0.015 in. This ensures at least 0.007 in. clearance for the main bearing, preventing contact at any time.