Air Bearings and Magnetic Bearings Compared

Every conventional bearing shares one compromise: solid contact, however precise, means friction, wear and heat. Two technologies delete the contact itself. Aerodynamic and aerostatic bearings levitate the rotor on a film of gas, and active magnetic bearings suspend it in a magnetic field under electronic control. Both deliver near-zero friction and no wear at the contact, at the price of complexity, cost and, for magnetic systems, a dependence on power and control.

This guide covers how each principle works, the supporting system each one needs, where each earns its keep, and when a conventional rolling bearing remains the wiser choice. The claims are kept physical: a magnetic bearing has no oil film but it is not maintenance-free, and an air bearing has almost no friction but very little load capacity.

Two ways to delete the contact

A gas bearing replaces the oil film of a hydrodynamic bearing with the viscosity of air. A magnetic bearing replaces any film at all with a controlled magnetic field. Both remove the solid contact, and with it the fatigue, wear and lubrication machinery that dominates a rolling bearing's life. The difference is what holds the rotor up: a gas bearing needs pressure or motion to build its film, while a magnetic bearing needs a live control loop, awake and powered at every instant.

Aerostatic air bearings: pressure supplied from outside

In an aerostatic bearing, compressed gas at 4 to 7 bar is fed through small restrictor orifices into a shallow gap between the bearing and the rotor. The pressure builds a film that floats the shaft even at standstill, so the bearing is stiff, precise and repeatable from zero speed. Because the film is supplied rather than generated, the stiffness is roughly constant with speed, and the design is used where a machine must move with nanometre precision and cannot tolerate a speed-dependent behaviour.

Restrictors and the micron gap

The gap is typically 2 µm to 20 µm. That thin film is what makes the bearing both precise and fragile: the orifice restricts the flow so the film pressure rises where the load pushes the shaft closer by 1 µm or so, and this self-compensation gives the bearing its stiffness. A blockage or a scratch changes the flow and the film collapses, which is why filtration to sub-micron levels is part of the bearing specification, not an accessory.

Why the film demands near-perfect surfaces

An air film of a few microns does not average the errors of the parts around it; it transmits them. The rotor journal, the bearing bore and the thrust faces must be finished and round to a level a rolling bearing never asks for, and the whole structural loop must be stiff, because the film cannot absorb deflection the way a compliant oil film can.

Aerodynamic and foil bearings: the rotor pumps its own film

In an aerodynamic bearing, no gas is supplied. The rotor's own motion drags gas into a wedge-shaped gap, and the pressure that builds carries the load. It therefore needs speed to work and cannot support a standing load at rest, but needs no compressor, plumbing or filtration, and dominates in high-volume turbo machinery such as small blowers that spin at 50,000 rpm to 200,000 rpm.

The foil bearing is the practical version of the aerodynamic idea for machines that must start and stop. The bearing surface is a thin compliant foil, 0.1 mm to 0.5 mm thick, that rides on a cushion of air at speed and rests on a sacrificial coating at start-up, so the rotor survives the low-speed regime where the film has not formed. Foil bearings are increasingly common in oil-free compressors and blowers, where their whole appeal is that they delete the oil system rather than the bearing.

Active magnetic bearings: a control loop in place of a lubricant

An active magnetic bearing suspends the shaft in a magnetic field. Position sensors read the rotor's location 2,000 times per second or more, a digital controller computes the correction, and electromagnets push the shaft back to the centre line. Because nothing touches, friction and wear at the bearing are zero and the rotor speed is limited mainly by the strength of the rotor material rather than by a lubricant or a cage, which is why magnetic bearings routinely run above 30,000 rpm.

Position sensors, controller and amplifier

The control loop is the bearing. Sensors, an amplifier that drives the coils, and a controller that is tuned to the rotor's dynamics all have to be working for the shaft to float. The same loop is also a measurement system: position, orbit, unbalance and force data stream out of the controller, which is why turbomachinery builders value magnetic bearings for condition monitoring that is built into the suspension. ISO 14839 gives the vocabulary for the vibration of machines with active magnetic bearings.

Touchdown bearings and the fail-safe design

The dependencies are structural. A magnetic bearing needs power and control to exist at all, and a failure of either drops the rotor, so the design includes touchdown bearings: rolling bearings with a small clearance, typically 0.1 mm to 0.3 mm, that catch the shaft during power loss or overload. Bias magnets or permanent-magnet bias reduce the steady-state power, but the system remains an electromechanical product with sensors, software and redundancy, specified as a system rather than a component.

All four support a shaft; the difference is what each one needs around it to keep working.

Where each technology is the right answer

Gas bearings dominate where precision outranks load: coordinate measuring machines, lithography stages, metrology spindles and precision optical machining, where runout of 0.05 µm or less is routine and any oil would spoil the product. Foil and aerodynamic bearings dominate where an oil-free high-speed shaft is the point, as in small turbo compressors. Magnetic bearings dominate where oil elimination is strategic: subsea and process-gas compressors, flywheels and vacuum pumps where a lubricant would spoil the process. Machines in this class span 100 kW to several megawatts.

In every case the surrounding system, clean gas or quality power and control software, is part of the bearing specification. A machine built on an air bearing is a machine with a compressor, and a machine built on a magnetic bearing is a machine with a control cabinet.

Maintenance moves rather than disappears

Contact-free bearings do not eliminate maintenance; they relocate it. Air systems need filtration checks, compressor service and orifice cleanliness, and a single blocked restrictor can end in a crash. Magnetic systems need controller and sensor health, cable integrity and periodic checks that the touchdown bearings are still free. What disappears is the greasing round; what appears is a control system that must be understood and a stream of monitored data that must be watched. A magnetic bearing whose controller alarms are ignored behaves exactly like a bearing whose lubrication was ignored. The vibration severity bands of ISO 10816 and the signal habits in the vibration analysis guide carry across both worlds.

Why the rolling bearing remains the default

For the overwhelming majority of machines the contact-free technologies are the wrong answer. A rolling bearing carries far more load per unit of size, works at standstill and through a power cut, tolerates real-world contamination, costs far less, and follows the ISO 281 L10 life method that lets an engineer predict its death months in advance. The honest framing is a continuum: rolling bearings for nearly everything, air bearings where nanometre precision or oil-free cleanliness is the product, and magnetic bearings where oil elimination and extreme speed justify an instrument-grade system. Hybrid designs blur the boundary, but the selection question still ends where it began: what does the load, speed and precision budget actually need?

Buyers meeting an air or magnetic bearing specification for the first time should treat the whole support system as the product. The gas quality, the control cabinet, the touchdown hardware and the commissioning data all ship as part of the capability, and the failure analysis differs too: an air bearing usually fails as a crash from lost air or a particle, and a magnetic bearing's diagnostics live in the controller logs rather than on a damaged part. The rolling-bearing fundamentals that underpin all of this remain in the technical library, the conventional families stay filterable in the bearing selector, and the speed and DN guide explains why a rolling bearing still reaches further than most engineers expect before a contact-free solution is justified.

Common misreadings of contact-free bearings

• Assuming zero friction means zero loss. The gas supply, the control electronics and the touchdown drag all consume energy even when the film carries the load.
• Forgetting the filtration. A single particle can block an orifice or scratch a micron film; clean gas is a hard requirement, not a preference.
• Treating a magnetic bearing as a passive part. Without power and control it has no stiffness at all, so the fail-safe design and the touchdown gap matter as much as the coils.
• Ignoring load capacity. An air bearing of a given envelope carries roughly 1 to 10 percent of what a rolling bearing of the same size carries, which decides many otherwise promising applications.
• Assuming the film is stiff in every direction. A gas film's stiffness varies with load and geometry, so a design that carries a radial load well may still be soft in tilt.
• Forgetting the duty cycle. A foil bearing that starts and stops 50 times a day wears its coating far faster than one that runs continuously.
• Overlooking the competences. A control system and a clean-gas system each need an owner; a machine is only as reliable as the people who watch its alarms. The polymer bearing guide and the angular contact range show the conventional alternatives at the other end of the spectrum.

Where a conventional solution is decided instead, the mechanics of noise, heat and life are the familiar ones, and the harsh-environment guide covers the sealing and material choices that let a rolling bearing survive a demanding duty. The catalogue organises the contact-free and conventional families together in the other bearings range, and the wider context of where these technologies appear across machine types is collected on the industry pages. Established suppliers such as SKF and NTN illustrate how the conventional and instrumented ranges sit side by side in one product programme.