The Electric Motorcycle, Part 3

Exploring the types of electric motors and their related power losses.

Mission R power unit

An electric motor is just a system by which magnetic force is converted into mechanical power. Linear motors may one day propel trains, but conventional rotating motors are more familiar to us.

DC series motor

As children, many of us made simple electromagnets by winding a coil of insulated wire around a nail. Sending battery current (DC) through the coil-induced magnetism in the nail enabled it to attract small ferrous objects. To use this principle to create a rotary motor, we can make two circular arrays of electromagnets, one inside the other.

The non-rotating part is called the stator, and the rotating part, the rotor or armature. The cylindrical stator is made of stacked thin, round iron laminations, each of which looks rather like a motorcycle clutch plate. Fitted inside this and rotating on bearings with a very close clearance (air gap) to the stator is a second cylindrical set of laminations. Facing each other across this gap are slots, cut into the stator’s inside diameter and the rotor’s outside diameter, parallel to the bearing axis. Between each pair of slots is a “pole,” that part of stator or rotor not cut away to make a slot. Insulated wire coils are wound into these slots, such that each coil magnetizes a particular pair of poles, which are 180 degrees opposed to each other.

If we now pass DC through stator and rotor coils, their respective poles will be magnetized just as was the nail. North and South magnetic poles will attract each other, and the poles of stator and rotor will align with each other. Now we need to find a way to make the rotor spin, not just align its poles to those of the stator and sit there, getting hot.

To do that, we need a system of switches that can send current through only the coil whose poles are about to align with the stator’s poles. We attach to the rotor a cylindrical arrangement of straight, parallel copper bars called a commutator. Each bar is separated from its neighbors by thin layers of insulation. One-hundred-eighty-degree pairs of these bars are wired to similarly 180-degree rotor windings. We can send current through any 180-degree pair of these bars by placing a pair of low-friction carbon “bushes” (contact blocks) in contact with the cylindrical array, one on each side.

We position these brushes so that, as a particular set of rotor poles and windings approaches the stator's poles, DC is sent through it by the brushes, and no other rotor coil receives power from them. Being magnetized by coil current, the two rotor poles are attracted to the next pair of stator poles, but just as they are about to align, the rotor has turned enough that the commutator bars connected to that coil have rotated past the brushes, and it is now the next pair of poles that receives power through its pair of bars, which have rotated into contact with the bushes.

What this system of rotary switches does is to create on the rotor a magnetic field rotating backward, which is always pulling the next pair of rotor and stator poles towards alignment. Just as they get close to each other, they are de-energized and power is now sent to the previous pair of poles and so on.

This is called a series motor because the coils of the stator and the brushes are connected in series.

Permanent magnet motor

Once more powerful permanent-magnet materials became available in the early 1950s, it became possible to do away with the complexity of a wire-wound stator and to provide the same stator magnetic field from properly oriented permanent magnets, creating the same circular array of magnetic poles, alternating north and south as before. As the brushes send power into the rotor windings, they operate as described above.

AC single-phase motor

In single-phase AC motors, which can run on household current, the 60-cycle variation in polarity sent through the stator coils causes the stator’s magnetic poles to continuously reverse their polarity: North, South, North, etc. This variation by itself cannot start a rotor turning, only make it dither back and forth. Help comes in the form of a starting coil whose phase is delayed by switching in a capacitor (a sort of “electrical spring”), producing enough torque to get the rotor turning in the right direction, giving it enough speed to couple with the varying stator field and accelerate to operating rpm. A centrifugal switch then turns off the starting coil (you can hear it click as the motor coasts to a stop).

Rotating within the stator is an armature or rotor. There is no commutator as in a DC motor, for the stator's alternating field can induce currents in conductors on the rotor which, at speed, couple with the stator's field to produce torque and rotation. This is, very reasonably, called an induction motor. By varying the number of poles, these motors can be made to spin at the common speeds of 1800 or 3600 rpm.

Three-phase induction motor

While single-phase AC produces a stationary but constantly reversing stator field that does not rotate, three-phase AC can produce a rotating field that will start the rotor without added gadgetry.

2015 Zero DS static 3/4 view

Why do we even have AC?

Our nation’s electric power system is AC rather than DC because voltage transformers, being dependent upon rapidly changing magnetic flux, work with AC but not DC. This allows AC to be easily transformed up to high voltage for long-distance transmission (line loss is proportional to the square of the current, and at constant power, the higher the voltage, the lower the current). Another point is that the most maintenance-intensive part of a DC machine is its brush/commutator assembly.

Single-phase AC vs. three-phase AC

If we display the varying voltage of single-phase household AC on an oscilloscope screen, we see a classic sine curve, centered on the zero voltage line, rising and falling to peaks at 155V above and below it (this, averaged out, delivers the same power as 120V DC). It’s clear from this curve that little power is delivered near the zero line because voltage and current are so low there. Most of the power is therefore delivered intermittently, in the vicinity of the peaks.

To deliver power continuously, three-phase power arrives on three wires, and each phase is offset from its neighbor by one-third of a cycle, that is, 120 degrees. Thus, while one phase is passing through zero and delivering no power (or motor torque), the other two are at very much non-zero values and are delivering motor torque in their respective windings. This allows a three-phase AC motor to deliver 180 percent of the power of a single-phase motor.

Single-phase household current saves money and complexity for the moderate loads it drives, but industry requires the greater power density of three phase. Single-phase is provided to households by tapping just one of the three phases on the 1100V local line and transforming it down to 120V. Look up at the power pole to which your house’s drops are connected and you will see the transformer.

Variable-speed motors save power

For many years in the long era of cheap energy, an estimated 50 percent of industrial electric motor power in the US was wasted in driving pumps, blowers, or other varying loads with constant-speed electric motors. The coming of cheap semi-conductor switching devices made it possible to replace such constant-speed motors with variable-speed motors that could run at the speed actually needed by the load at the moment.

One such control technology allowed variable-frequency three-phase AC to drive powerful, long-lasting variable-speed induction motors at whatever speed the load actually needed, not just at the standard 1800/3600 speeds.

Although electric vehicles can be built using DC motors, the “fuel economy” of AC motors makes them the choice where vehicle range is an issue (i.e. not drag racing, Bonneville, etc.).

Efficiency

When I was a lad, I enjoyed our local library’s old, dark-green textbooks on things like power stations or marine diesels. One useless info tidbit that impressed me was the very high efficiency of large electrical machines, such as the AC generators in electric power stations or huge electric motors in pumping stations. Their efficiency has been very high since 1910, and today the numbers range above 99 percent.

Small electric motors typically offer 90 percent efficiency, with special high-efficiency motors (usually heavier and/or longer) at 94 percent. Motors in sizes for electric vehicles can be built to higher efficiency, but because this involves providing more copper and more high-quality magnetic iron for pole structures, such heavier motors cost more.

Bear in mind that electric motor efficiency compares the electric power delivered to the mechanical power output. The efficiency of electricity generation does not appear in this number. As I’ve noted elsewhere, this varies from 35 percent to 60, depending on energy source (coal, nuclear) and heat cycle (straight thermal, combined cycle, etc.). Overall efficiency is computed by multiplying together all efficiencies in the system.

2015 Zero DS motor

ELECTRIC MOTOR LOSSES

Restrictive heating, or OHMIC loss

In any device that uses electric power, the largest loss is typically that from resistance, proportional to current, squared, multiplied times winding resistance (Ohms). When I think of resistance, I think of people (electrons) trying to rush through a crowd (atoms in a solid). The many collisions that result cause the crowd to become angry (hot).

This Ohmic resistance in the stator windings is called primary Ohmic loss, and that in the rotor is called secondary Ohmic loss. The only practical way to reduce this is by reducing current (amperes) by raising voltage, or by using conductors with lower resistance. The higher the operating voltage, the more of the motor’s volume must be devoted to wire insulation, limiting what can be done in this way. Silver conducts somewhat better than copper but is not a money-saving path to reduced loss.

To avoid making the motor bigger, square-section wire may be used in place of circular section, giving an increase in wire area of 27 percent. In flights of fancy, futurists propose that room-temperature superconductors are always "just around the corner" (Maybe GizMag is announcing it right now!), and such a development would indeed allow packing much more power into a motor of a given size (with zero Ohmic loss, there would be no Ohmic heating to limit how much power you could send to the motor).

The second basic source of loss has the grand description “magnetic hysteresis,” but all it means is that when forward current magnetizes a set of stator poles to be North, then reverses itself to remagnetize them as South 1/60th of a second later, not all the energy put into this process can be recovered. Reversing all those tiny magnetic domains in the poles causes some molecular motion, and molecular vibration is heat.

Currents in the iron poles

Some power is also lost because, with all these magnetic fields coming and going, currents are induced not only in in the wire windings but also in the iron stator poles. Any such current flow also suffers Ohmic loss, but most of this is stopped by making the pole assemblies not as solid but as stacks of thin laminations that are shellac-coated (insulated) before stacking them up and riveting them into complete cylindrical stator pole assemblies, ready to have wire wound into their slots. This assembly of insulated laminations stops any large-scale induced currents from flowing because the farthest it can flow is the thickness of one lamination.

You will find the same laminated-pole construction used in transformers. It used to be that in a big room lit by fluorescent tubes, at least one of the lighting units would make an annoying hum. An experienced man showed me that this could often be quieted by pulling the ballast transformer and giving the rivets holding its pole laminations together a few good taps to tighten their grip, thereby preventing the laminations from vibrating audibly (and annoyingly) against each other at 60 cycles per second.

Bearing loss and windage

A third set of losses arises from the motor’s shaft bearings and the motor’s internal windage. Ball-bearing friction torque is typically 0.001 to 0.002 times the load, and unless the motor drives through gears or belt, the only load is the weight of the rotor. Bearing loss is small. That leaves windage, which in air-cooled motors is mainly the power consumed by a small fan attached to one end of the rotor. As you probably know, in flywheel energy-storage systems, fast-spinning rotors would rapidly slow down from churning air, so they are enclosed, pumped down, and kept under vacuum. Thus, in motors turning at usual speeds, some torque is lost in the vigorous air turbulence in the small air gap between stator and rotor.

Stray-load losses

Finally there is the catch all called stray-load loss, which allows me to use a favorite phrase, “parasitic oscillations.” Depending upon the details of the slotting of rotor and stator, high frequency currents and iron loss may be provoked, much of which can be avoided by care in design and manufacture. The cost of such care is not justified in the $50 motor you buy to power an attic fan that operates only on hot days, but if you are trying to make an electric vehicle go X miles at Y speed on a single charge, those single-digit losses become important.

Cooling

If an electric car’s motor is 94 percent efficient and is cruising at 18-kW load (24 hp), its six percent loss is 1000 Watts. Large industrial motors are normally air cooled, but as design makes motors more compact and dense, liquid cooling may be required.

In speaking with Zero engineer Abe Askenazi, I learned that their motor is presently air-cooled but that motor cooling sets a ceiling on continuous top speed. Because electrical machines are fairly massive, they can for a time act as their own “heat sinks,” allowing motor or generator power to exceed their normal ratings for limited periods.

Something similar happens with motorcycle internal-combustion engines. If the engine is air-cooled, you will notice that the cylinder head is quite heavy; this extra mass acts as its temporary heat sink. The head heats up down the straightaway, then cools down during braking and in corners. Same with liquid cooling, except that water is a far more efficient heat storage medium than are equal masses of metal.

As part of a complete powertrain system, the electric motor must, through sensors, report its temperature to the system ECU, which will act to prevent high temperature damage to wire insulation.

Motor torque

The torque of an internal-combustion engine depends strongly on its rpm, depending on how its design compromise is made. A sportbike’s long valve timing and large ports sacrifice torque at lower rpm in order to deliver more at higher revs. A cruiser is compromised in the reverse sense; its short valve timing and moderate port sizes deliver high torque at the bottom of its usable rpm range, but that torque slopes downward as the engine revs up because the limited intake system falls behind the engine’s air demand.

Electric motors deliver both. The torque of an electric motor comes from its internal magnetic fields, which have no such constraints. Many electric vehicles have only a single gear ratio for this reason.

Askenazi said that four years of working with electric powertrain problems have completely shifted his thinking. “A prime mover with one moving part,” he said, citing some of the complex details of the internal-combustion engine, with its many valves, valve collets, spring retainers, and valve springs, to name just a few. Great complexity that could, with the possibility of improved batteries, become as irrelevant as the Walschaerts linkage, invented in 1844 to control steam locomotives.

Complex, sophisticated, fascinating enough for a lifetime of study. But irrelevant. Maybe.