Authored by: Alan Crapo,
Electric motors are by far the most common method used to convert electrical energy into mechanical motion. To provide motion, most motors use attraction and repulsion of magnetic fields to create the force needed for movement. But the generation of those fields and their method of interaction gives rise to literally dozens of motor designs.
Standard induction motors have been in use for over 100 years and are reaching the limit of improvements in both performance and cost. In an induction motor, a stator winding wound on the outer frame of the motor induces a voltage into the conductors of the rotor. The rotor voltage creates a current flow through the rotor forming a magnetic field that interacts with the magnetic field of the stator to make the rotor turn.
Induction motors have several inherent advantages. First, they are simple to build and are robust and reliable. Second, they naturally adapt to the load placed on the motor. Current rises as the load increases while the motor’s natural slip accommodates torque spikes. Through agreements between motor manufacturers, induction motors come in standard sizes and mounts that promote interchangeability.
The primary speed of an induction motor is based on the frequency of the ac power line, rather than the applied voltage. This means for fixed-speed applications, ac-induction motors can be driven directly from the ac-power line without the need for any electronic drive. However, most applications can show sizable energy savings by switching to variable speed operation to match load requirements. To accomplish this task, induction motors need an electronic-drive module to vary its speed.
In applications where variable speeds are advantageous, a variable-frequency drive is mandatory for induction motors. Modern electronic drives can operate both induction and permanent magnet motor, so once a motor drive has been added, the cost differential between an ac-induction motor and that of a permanent-magnet motor is only in the cost of the motor. The cost of the drive for either motor type is the same.
Induction motors operated with an electronic drive can reach speeds many times the motor’s rated base speed while still maintaining a constant horsepower. This is possible because ac-induction motors do not generate a fixed back or counter-EMF (electromotive force) in the rotor that builds with higher rpms. Upper speed limits are based purely on the mechanical design of the motor, with centrifugal force being the prime limiter.
Generally, making induction motors more efficient means adding more copper and steel. The drawback is that this approach makes a motor bigger and more expensive. In the last five years, these basic motor materials have significantly increased in cost. However, motors must meet higher efficiency standards to qualify for the NEMA Premium rating in the U.ŒS. And recent legislation mandates use of motors having NEMA Premium ratings. In Europe, the proposed standards are even higher, rating a “super” premium efficiency.
One recent move for improved efficiency replaces the standard die-cast aluminum conductors in induction motor rotors with rotors using copper as the conductor. However, both the cost of copper and the high-temperature processing of copper entails higher manufacturing costs.
Unfortunately, induction motors are not without disadvantages. One is that across-the-line starting lets induction motors draw high peak currents that can reach seven to 10 times normal operating currents. Obviously, these current surges can affect other devices on the power grid. Most modern installations use soft starters to mitigate this condition.
Induction motors also typically have a poor power factor in normal operation. Utility companies may require the installation of power-factor correction equipment for larger motors. Much of the power loss in an induction motor is in the rotor, which typically relies on air for cooling. The act of moving air through the motor housing for cooling leads to windage losses. Other losses include resistive or copper losses as current flows through the stator winding resistance, magnetic losses through stray magnetic paths, and the effect of hysteresis in the motor iron.
This low-cost, easy-to-use motor still dominates many applications. Traditional dc-brush motors are by far the most economical design for small devices. They’re used in numerous applications including toys, auxiliary motors in cars, hand-operated devices, and other consumer items. However, industrial usage for this type of motor has been on the decline.
In a dc-brush motor, power is delivered to the rotating armature coil via a commutator, a segmented copper disk that mechanically connects the proper winding with the proper polarity to keep the motor turning. Power is delivered to the commutator through brushes typically made of carbon in small motors or copper-impregnated carbon in larger motors.
Operational issues with this motor include: brushes and commutators wear out and must be replaced periodically; most of the power loss is in the rotor, which has a poor thermal path; and the efficiency curve peaks at one operating speed and load for a fixed-brush commutation angle. The efficiency falls off rapidly as the motor operates away from this point and is quite poor at speeds of one-half and twice the value of this operating point. Efficiency at the maximum power output of the motor is far below that of its peak efficiency.
Brushless permanent-magnet motors
The dominant design for brushless-dc permanentmagnet motors (BLDC) is a radial design specified as to whether it’s an inner-rotor-radial or outer-rotor-radial. Basically, this means the inside of the motor rotates (innerrotor) similar to your typical motor or the outside of the motor rotates (outer-rotor) as used on some fans with the blades attached directly to the outer rotor. Permanent magnets embedded in the rotor interact with the magnetic field generated by the stator windings to produce torque. Basically, it’s the exact opposite configuration of the brushed-dc motor.
The design concept of BLDC motors is mature, well developed, and supports many different applications and construction materials. Brushless PM motors need an electronic drive controlled by magnetic sensors in the motor to replace the brushes and commutator, but this is not as big a disadvantage as it used to be. For the most part, drive costs have dropped significantly in recent years.
One issue any PM motor has is a pronounced back or counter-EMF that rises with motor speed. When motor speed reaches a point where the back-EMF matches the drive’s available bus voltage, motor speed cannot rise any further. (There is a way to get slightly more speed with a timing adjustment that will be covered later.) This speed limitation presents a problem for traditional PM motors.
The electronic commutation for these motors can be altered to attain higher speeds, similar to changing the timing on a spark plug in your automobile. However, current demand increases dramatically when operated in this mode. This timing advance is most often used for applications that need high speed at low torque combined with a lower-speed/high-torque requirement. This type of operation is typical of a washing machine with a low-speed wash and a high-speed spin cycle.
A relatively new design which gets around this speedlimitation problem is called an interior permanent-magnet (IPM) motor design. IPM motors typically employ less-expensive rectangular magnets rather than magnets curved to fit the circular rotor surface. The magnetic blocks are slid into slots in the laminated rotor core. Magnet retention is enhanced and it simplifies the manufacturing process. IPM motors use both the permanent-magnet torque and reluctance torque. This reduces or eliminates the limitation of motor speed imposed by back-EMF and drive-bus voltage.
The standard radial motor has a single stamping for the stator that includes slots for the windings. Recently, the development of segmented or cut-core designs with concentrated windings has produced motors that perform better than the standard design. High-pole-count motors with radial-facing magnets and poles are also gaining acceptance. High-pole-count brushless motors need control electronics that switch at higher rates to maintain the same rpm as lower pole-count motors.
Some radial BLDC PM motors have their rotor on the outside. This design was first introduced in small fans operating on a low dc voltage, such as the 12-Vdc cooling fans commonly used to remove heat from computers and other electronics. In these small fan motors, the rotor consists of magnets or a magnet ring inside a cup. The cup fits over the stator. This same concept has been used for disc-drive spindle motors for almost 30 years. Putting the rotor magnets on the outside intensifies the total magnet flux and also enlarges the air-gap diameter for greater torque density.
Though torque density improves due to the larger radius of the rotor, the stator with the heat-generating windings is now on the inside of the motor, where it is more difficult to cool. This configuration works best with short axial lengths having enough surface area on the stator and a conduction path to remove the heat. Long stack lengths prove a challenge to cool when the only heat path from the stator is through the stationary shaft.
An application that is well suited for this type of motor is the direct-drive washer motor which is about 12 in. in diameter and 2-in. long with ferrite-magnet arcs on the inside diameter of the outer-rotor assembly. The stator bolts to the stationary part of the washing machine and the rotor is attached to the rotating drum. The stator mounting provides a good heat path to the washer, and there is a large surface area, so it’s not difficult to cool the motor. This direct-drive approach keeps motor speed and noise low, even with spin-cycle speeds over 1,200 rpm. A beltdrive motor running at 15,000 rpm for the spin cycle is noisier than the 1,200-rpm direct-drive motor. The quiet operation and elimination of belts, pulleys, and transmissions has made direct-drive washer motors the preferred design of most Asian appliance manufacturers.
Standard axial permanent-magnet motors
Axial permanent-magnet motors have been around for several decades and typically have a form factor like a pancake: large diameter and short length. The advantages of axial motors over radial motors include an output torque that rises with the cube of the motor diameter while the length of the motor has no appreciable effect on output torque. This gives axial motors a big volume advantage for high torque, higher power applications.
Axial motors have their windings exposed to the outside, boosting thermal performance. In addition, they can use a relatively inexpensive bobbin-style winding while substantially enhancing motor reliability. The flux path is straight through the poles of an axial motor without intervening case metal as seen on standard motors. This reduces flux leakage and lets motor designers use better performing grain-oriented magnetic steels. The winding incorporates a more optimal shape in that the winding perimeter is small for the enclosed area. More winding volume may be added by extending the length of the motor to boost efficiency.
Despite all these advantages, axial motors are still a limited niche, mostly because they need rare-earth magnets. Until the price of neodymium-iron-boron (NdFeB) magnets dropped to reasonable levels, these motors were far too expensive for most applications. The price of these magnets spiked a few years ago, then dropped. But the price is rising again. Supply concerns regarding these magnets are rising as well.
The pancake form factor of standard axial motors limits them to applications where that form factor is advantageous. Most applications need (or, are typically designed for) a motor with a radial form factor, where the motor is longer than the diameter. The axial forces developed on the rotor of this motor requires both axial and radial rotor positioning by the bearings. Also, many axial-motor designs use a high-pole count that produces many individual parts to manufacture, inventory, and assemble.
At this time there doesn’t appear to be any IPM standard axial-motor designs available. There is, however, quite a bit of development work going on in this area.
NovaTorque Inc., Sunnyvale, Calif., takes a different approach to its axial PM-motor design. Instead of a typical pancake cylindrical rotor shape, the NovaTorque motor uses two cone-shaped rotors, one mounted at each end of the motor shaft with the cones pointing inward. Instead of the conventional perpendicular flat cross section seen in most stator poles, the end of the NovaTorque stator poles match the conical shape of the rotors. The rotor hubs use an IPM arrangement that concentrates the magnetic flux. The saliency of the IPM designs offers better sensorless motor control as the motor reflects passage of the salient pole back to the motor drive.
The design of the rotor and field poles maximizes the surface area available for magneticflux transmission between the stator and rotor, while minimizing material volume. In fact, in some designs the area between the rotor and stator surfaces is twice the perpendicular cross-sectional area of the stator-field pole. This larger surface area concentrates the magnetic-flux density, letting the NovaTorque design use less-expensive ferrite magnets. Yet motor efficiency and performance equals or exceeds more-expensive motors that use neodymium rare-earth magnets.
Another special feature of the NovaTorque design is its axial flux path. The flux flows straight through the axially oriented field poles of the stator, meaning the flow is parallel to the motor shaft rather than at right angles as seen in more-conventional designs. This orientation lets the motor use grain-oriented transformer-grade steel, which lowers eddy-current losses and improves efficiency.
The axial orientation of the stator-field poles also allows the use of bobbin-wound field coils. This type of coil is easier to wind and uses a shorter overall conductor length than coils wound in the slots of a typical, radially oriented motor. This coil arrangement also creates a better thermal path. One face of each coil is next to the external motor case that lets it directly conduct heat away from the coil. Conventional induction motors typically find these coils buried inside the lamination stack, making heat dissipation difficult.
The conical-hub geometry lets the NovaTorque design overcome most of the primary disadvantages of other axial motors. Instead of a pancake, NovaTorque motors resemble radial PM motors and induction motors and can be made in the same standard sizes commonly available for those types of motors.
Switched-reluctance (SR) motors
While reluctance motors were one of the first types of motors to be developed in the early to mid 1800s, they fell out of favor as newer designs provided greater power and efficiencies. This class of motors was revived in the 1970s with the promise of performance equal to that of permanent-magnet motors without the cost of permanent magnets. However, several problems arose that hampered their development.
First, to boost power density, these motors must operate at high flux levels and need to use small rotor air gaps. These two conditions created a noisy motor that was difficult and costly to manufacture. A second factor is that standard motor drives won’t power switched reluctance motors. The drive and motor must be an integrated design as drive characteristics greatly influence motor performance. This also leaves the user totally dependent on a single manufacturer, because the motors and drives are not interchangeable. Induction motors and permanentmagnet motors can use the same standard commercially available drives with a simple software switch, many alternative motor and drive sources are available.
The switched-reluctance motor has found some specialized applications, but even now is rarely used. SR motors are designed for specific needs or applications, so they are not available in standard sizes. But advances in IPM designs, which take advantage of both permanent magnet and reluctance torque, eliminates many of the reasons for using SR motors.
Stepper motors are yet another version of variablereluctance motors. They are high-pole-count motors, on the order of 50 or more poles. The most commonly used stepper motor carries 200 poles. The most widely used type, the hybrid stepper motor, contains a permanent- magnet assembly in the rotor. This type of motor takes discrete steps of motion and is generally limited to speeds below 2,000 rpm.
The advantages of stepper motors include high torque for the size of the motor, a simple low-cost drive and motor, and simple position control. Disadvantages include low speed, significant loss of torque as speed rises, strong resonances at specific frequencies, high noise, and low efficiency with respect to load variations. To get high torque, stepper motors are constructed with small air gaps, often in the range of 0.1 to 0.2 mm (0.004 to 0.008 in.).
In-hub wheel motors
Many axial-motor designs are focused on in-hub wheel motors for vehicletraction applications. The natural pancake form factor of standard axial motors has led to this focus. The in-hub simplicity appeals at first look, but there are significant challenges to its implementation.
First, remember that electric motors develop power as a product of torque times speed. As wheel rotations rarely exceed 1,000 rpm for normal vehicles, motors would need significantly higher torque to produce a reasonable power output. Therefore, low-speed wheel motors must be substantially bigger, use more materials, weigh more, and cost more than a high-speed motor. While higher-speed motors do need a gear drive to reduce speed, gear drives are reliable and relatively inexpensive.
Second, electric motors are precision devices with closely controlled mechanical air gaps which are sensitive to dirt and particles collecting in the gap. PM motors are especially prone to this issue because any magnetic particles (steel, for example) are directly attracted into the gap by the magnetic fields. Brake rotors on the wheels generate these magnetic particles as they wear, which means the motor must be perfectly sealed while still allowing free rotation.
Motors are also sensitive to water damage through wear and corrosion. Bearings especially do not take well to water and, therefore, need special sealing.
The amount of heat generated by these motors mean all need at least forcedair cooling, and some require liquid cooling. In fact, many automotive motors require liquid cooling and all need forced-air cooling at a minimum. How this cooling is to be accomplished when mounted on the wheel is also unclear. In addition, the power and control leads must be cabled out to the wheels.
The handling of a car depends on the sprung weight of the wheels. Motors are relatively heavy components and placing them on the wheels adds to the sprung-weight problem.
Radial motors designed with very short stack lengths and relatively large numbers of poles can create high-torque pancake motors that can be used for low-speed, high-torque applications such as in-hub wheel motors. But the disadvantages of this design are similar to the axial configuration. High torque requires the use of rare-earth magnets. The high-pole count also produces a large number of discrete parts that need assembly.
Other motor types
Linear motors take a traditional motor design, either radial or axial, and essentially unroll it so motion follows a straight line rather than circular. Linear motors are the fastest growing area within motor applications, and they are being used for machine automation, conveyors, and moving parts through factory warehouses.
Magnetic levitation (mag-lev) motors are a special version of linear motors. The magnetic field of the motor not only provides motion, but lifts or suspends the armature (the part that moves) such that it no longer touches the fixed windings of the guide track. The primary use of this type of motor is in transportation vehicles such as trains and monorails. However, today it’s also finding new applications in factory conveyor systems.
Trends in efficiency improvements
There is currently a strong trend in product design to change from induction and universal motors to using permanent-magnet motors with electronic drives, even in some fixed-speed applications. This is because permanent-magnet motors can have higher efficiency and smaller size for the same output power. This saves on energy costs and can reduce overall product size and weight.
(The text in this article has undergone minor edits to reflect changes made after the original article went to press.)