Many designers know about magnetic bearings and have some understanding of how they work. Still, most have never specified them; except for special applications, magnetic bearings usually aren't considered. But now magnetic bearing use is on the rise because it offers advantages that improve traditional machine designs and address emerging motion challenges.

How exactly do they work? Magnetic bearings use electromagnetic coils to levitate rotating shafts. Sensors monitor the shaft's position and then feed that information to a digital controller. This controller changes the current in the coils to adjust electromagnetic forces on the shaft, and keep it balanced. These actions, which happen very quickly in terms of control-loop speed, precisely maintain shaft position at high rpm.

In fact, the need for high shaft speed is one of the factors that call for magnetic bearings over more traditional types. No metal-to-metal contact means that magnetic bearing systems spin faster, without lubrication. Instead, the shaft is suspended in whatever the surrounding medium might be — a vacuum, air, concentrated gasses, or liquid, possibly even a substance being processed.

When one compares the initial cost of a magnetic bearing to a simple ball bearing in, say, a small pump motor, the magnetic bearing system is a magnitude higher in initial cost. The difference is not because of hardware, which is fairly inexpensive. (Electromagnetic coils can be wound as economically as motor stators.) The control system is to blame: This includes the cost of a sophisticated digital controller, sensors to monitor shaft position, cables to carry shaft-position data to the controller, and cables to move power from the controller's amplifier to the electromagnetic coils.

If, instead, one compares a magnetic bearing's initial cost to that of a hydrodynamic bearing system on a large industrial machine, the cost difference narrows. The traditional system is likely to need bearings made of expensive materials, and require a sophisticated and expensive oil lubrication system.

Maintenance and reliability

Currently, there is a push to reduce overall cost of ownership for industrial equipment, even though it often requires larger capital investments. Magnetic bearing technology fits the bill here; it comes with a higher initial price tag, but (with no major friction components) saves on energy, maintenance, and rebuild costs. How do they fail then?

Air must flow through controllers to cool resident power electronics, and so the filters on these cooling systems can require occasional cleaning or changing — especially if they operate in dirty, dusty industrial environments. But this and similar maintenance tasks are basic, common sense items.

That said, controllers do have finite life. The mean time until controller failure is about eight to ten years. One factor is how hard the power electronics are pushed in a particular application. In any case, end users can expect to make changes in the power and control systems for a magnetic bearing after eight to 15 years of service.

In addition to filter changes, certain control diagnostics checks should be done regularly. In particular, recording calibration checks on the bearing's internal clearances — the so-called air gap between the shaft and coils — helps maintenance personnel ensure that bearings are running normally. (Since the control system offers constant feedback about system status anyway, these checks are fairly easy.) If there is change in performance, records help technicians determine its extent and cause. This function is particularly useful on very large machines, which are often critical to the operations they support. By monitoring potential problems, these intelligent machines alert operators to threatening failures before they occur.

Like an electric motor, magnetic bearings last 20 to 30 years, depending on the environment and how fast insulation breaks down in that setting. For example, in laser-cutting machines, gas blowers equipped with magnetic bearings let laser optics work at peak efficiency continuously and for extended periods without a need to service friction components — sometimes tripling the time between rebuilds.

The load limitation

Rolling bearings in machines can usually handle temporary overload conditions. While overloading may compromise the bearings, in most cases the machine mechanically forces through the problem. The overload doesn't result in a stoppage.

Magnetic bearings are not so forgiving of overloads; they have a definite load-carrying capacity. A machine equipped with an active magnetic bearing system that's undersized for the load will not function. If the electromagnets are unable to support the shaft and load, the shaft will no longer be levitated and the machine will shut down. For this reason, designers specifying magnetic bearings must clearly define the shaft loads — both static and dynamic — for the proposed application.

When designers go to higher speeds, they can also make machines smaller. To illustrate, a 1,000-hp compressor that runs at 1,800 rpm is a big machine — about twice the size of an office desk. By contrast, a 1,000-hp compressor running at 30,000 rpm is about the size of a microwave. The latter is more compact and has higher power density. Given the high speed at which magnetic bearings can run, designers can put more power into smaller packages.

In short, designing effective magnetic bearing systems is a balancing act. The cost of an oversized system may be too high, while undersized systems fail. A general recommendation, then, is to spend more time prototyping applications with magnetic bearings than those with standard bearings. Start with a conservative design, and then reduce the magnetic bearing's capacity in subsequent designs until optimized for the application. Fortunately, this design method is made easier by magnetic bearing self-measurement of shaft forces.

For more information, call (403) 232-9267 or e-mail the editor ateeitel@penton.com.

Keeping it clean

The semiconductor industry represents today's largest market for machines equipped with magnetic bearings. Cleanroom silicon-wafer manufacturing environments benefit from their low-vibration, lube and particle-free operation.

For example, turbo-molecular pumps, which create the vacuum required for manufacturing wafers, usually need replacement every few years — because of mechanical bearing failure. They last three to four times longer when equipped with magnetic bearings.

Cleanliness is also important in medical and cryogenic industries, where magnetic bearing use is on the rise. Lubrication-free equipment in cryogenic processes that handles extremely cold or liquefied gases — air-separation operations, for example — eliminates gumming and deteriorated overall efficiency in these cold environments.