Switched Reluctance Motors

Reluctance is an engineering term that (as far as I can tell) means that when you create an electromagnetic field (in the shape of a loop)…any steel part that is located near that field is “reluctant” to remain un-aligned with the field that is provided. The electromagnetic field will exert a strong force onto any steel part near it…to align with the invisible magnetic field.

The “switched” part just means that the location for the magnetic field that is performing the work is switched from one location to the next adjacent location [inside the motor] so that the rotor will continuously spin. I will sometimes use the term iron or steel interchangeably when mentioning a magnetically permeable material in a motor. Steel is simply iron that has less than one percent of its mass being some added carbon, which makes it harder, but doesn’t change its magnetic properties.

Note: when you google “reluctance” motors, there are two styles…the “switched” reluctance, and also “synchronous” reluctance. They use very similar principles. A Synchronous Reluctance motor has the same number of magnetic poles in the stationary stator and the rotating rotor. A switched reluctance typically has fewer poles on the rotor than the stator. Neither one is better or worse, they each have their benefits and drawbacks, depending on the application…

This website is about electric motorcycles, so…I must warn you ahead of time that I will be discussing electric car motors, and also stationary industrial motors. You may wish to take a break right now and drink a beer to calm your nerves before continuing.

The event that got this whole mess started was the fact that I read too much, and I stumbled across a wonderful article on the motor inside the Tesla Model-3 [written by cleantechnica, click here], which is a switched reluctance style (SR). The SR style isn’t just good, it’s better, and I hope this article will explain why.

Where to start?

The most popular legend describing the discovery of magnets is about an elderly Cretan shepherd about 4,000 years ago when iron was just beginning to be used more often (they were likely just iron meteorites the size of pebbles). Legend has it that this shepherd was herding his sheep in an area of Northern Greece called Magnesia. Suddenly, the iron tip at the bottom of his staff became firmly stuck, when he set it down on a large black rock that he was standing on. To find the source of the attraction, he dug it up to find what we now call lodestones.

It was believed that the invisible force was leading any iron bits towards itself. Lodestones contain magnetite, the naturally magnetic material Fe3O4. This type of rock was named magnetite after being found in Magnesia. Putting a sliver of lodestone on a small float that is suspended in a bowl of water formed a crude compass, so they instantly became very popular for ships. The word “load” in that area meant to “lead”, in the way that a compass leads the ship.

Of course, they didn’t understand that the Earth has a huge donut-shaped magnetic field, and the lodestone was aligning with it by always pointing north. However, it was so useful to have a compass, it didn’t really matter that they didn’t understand the physics behind it.

One of the interesting quirks of physics is the repeating patterns in nature. Electrons orbit the nucleus of an atom in the same way that planets orbit the sun. The fields in permanent magnets (and also electro-magnets which can be turned on and off) have a similar shape to the magnetic field of the Earth itself.

The most common permanent magnets (PM) are ferrite or neodymium. First, the materials that make up the magnet are mixed together and formed, and then they are heated up to their “Curie temperature”, which is a physical state where every atom is free to move around. Then, a strong electromagnet that is near it is turned on, and all the atoms in the magnet will align with the strong magnetic field. Then, the magnet is cooled down while the electromagnetic field is maintained. Doing this literally locks all of the atoms in alignment, so that all of their individual magnetic fields (similar to the earth), are aligned with each other, instead of pointing in random directions. Unless the magnet is heated up to its Curie temperature again, it will remain a permanent magnet.

Neodymium magnets are very strong and can survive a reasonably high temperature, which is why they are popular for the PM electric motors in EV’s. Ferrite magnets use a ceramic base that is high in Iron oxides (rust), so they are very affordable, but they are not as strong as Neodymiums, and they also retain their magnetism at a much higher temp than the neo’s. Ferrite magnets are the dark gray style that is found in cheap refrigerator magnets and are also used inside a microwave oven, in a large ring-shape (in case you see one being thrown away, and want a large free ferrite magnet).

There are several things that are important to notice about the bar magnet picture (posted above) in order to understand Switched Reluctance (SR) motors. Notice that the magnetic field is in the shape of two symmetrical loops. Also notice that in most of the field, there are few iron filings, indicating that the field is somewhat weak in those places. The most important part ot see is that there is a thick cluster of iron filings right next to the north and south poles of the bar magnet, and the size of the cluster is small and tightly packed.

This means that there is a tiny space in a specific location where the field is very strong, and the rest of the field (just a short distance away) is relatively weak. An SR motor uses this phenomenon to create some movement between a stationary “stator”, and a rotating “rotor” in a motor.

Franklin, Volta, Maxwell, Oersted, and William Sturgeon

In the late 1700s, Benjamin Franklin performed many electrical experiments with a Leyden jar, which is a crude capacitor that was charged by spinning an electrostatic “friction-generator” (a disc made from wax, sulfur, and carbon soot, rubbed by a stationary brush made with the fur from a cat or a rabbit).

In  1799, an Italian named Alessandro Volta invented the Voltaic pile, which is a true chemical battery (copper and zinc discs, separated by paper soaked in salt-water as an electrolyte).

In 1821, a Danish scientist named Hans Oersted showed that a current that is passed through a wire created a mild magnetic field around it.

In 1825, the British inventor William Sturgeon created a true electromagnet. The soft iron bar was only seven ounces in weight, but it was able to lift nine pounds of weight when it was energized. This is a key milestone that has not been given the amount of recognition it deserves.

In 1827 the Frenchman Andre Ampere was conducting experiments with electromagnets, and he needed a way to measure current, since varying the Intensity (“I”) of the current changed the power of the magnetic field in the coil, the “Amp” was named for him. (This is also why in engineering formulas, the current is abbreviated as an “I”)

Also in 1827, a German physicist Georg Ohm published a book describing the relationship between circuit resistance, current, and voltage. As a result, a unit of measuring resistance was named for him, called the Ohm. it is abbreviated by the letter “R”, or sometimes by the Greek symbol Omega (Ω)

in 1831, a British genius with no formal university training named Michael Faraday made dozens of discoveries in a wide range of fields. His work on electromagnetism made him very famous, and he is credited with inventing the electric motor, although his “homopolar” motor was only an interesting desktop demonstrator, since it could not perform any useful work when scaled up in size. His other experimental work was so important, that he never pursued the creation of a working electric motor. Of course, at the time there were no generators, electrical grids, or practical batteries to use (voltaic piles were weak and expensive at the time).

In 1865, a Scottish scientist named James Maxwell explored and defined the ways in which electricity and magnetism were deeply related. His thorough set of formulas was so comprehensive and profound, they are still used today by electrical design engineers.

In 1871, a Scottish engineer named James Watt invented a workable steam engine that transformed industry forever. He didn’t do any work with electrical devices, but his comprehensive calculations about power creation and transmission ended up with the “Watt” being chosen as a widely used measurement of power and work being accomplished.

In 1888, a Croatian-born Serbian genius named Nikola Tesla patented several types of electric motors while living in the United States. One of the motors he described is a three-phase brushless induction motor that used alternating current. Such a design is still widely used today (130 years later). George Westinghouse hired Tesla, and also licensed the Induction motor patent.

That 1888 patent also includes a motor that appears to be the reluctance style, and that is the graphic that I have placed above the title at the top of this article.

Remember when I mentioned that the bar magnet had a small but very strong magnetic power concentrated at the two tips? Notice that in this graphic, when the electromagnets are energized, the corners of the salient poles on the rotor are almost touching the salient poles on the stator. The “air gap” between the stator tooth tips and the rotor tooth tips is very thin to take advantage of the strongest part of the magnetic field.

A few notes on motors, leading up to today

Nikola Tesla’s 3-phase brushless AC motors from 1888 were an incredible leap, rather than the slow step-by-step of inventions that came before it. However, they work well at a constant speed, and that speed is most efficient near 3,000-RPMs. They do require multiple loops of copper wire in the rotor, and there is currently a trend of converting these common motors to a “Synchronous Reluctance” motor when they wear out, need repair, and for new installations. By replacing the copper in the rotor with only steel, the price is much cheaper, without any reduction in performance.

They are similar in principle to the “Switched Reluctance” (SR) motors that this article is about, but with distinct differences. Both types of motors use a rotor that is only made from stacks of thin laminated steel, with no copper or expensive neodymium magnets. The “Synchronous Reluctance” style (SynR, in the pic below) uses a stack of laminations in the shape of a round disk, with curved air-slots to act as a flux “fence” that guides the flux loops that are created. They are shaped this way to interact with the existing induction motor stators, which use overlapping “distributed” windings around the stator teeth.

When it comes to variable-speed electric motors (especially portable tools), the vast majority of motors have been brushed DC (an AC supply can be run through a transformer to change its voltage if needed, and then rectified into DC at the motor). The simplest form of DC motor uses two large stationary PM’s on opposing sides, with a set of electromagnets mounted on a central rotor.

The way to have a powerful electromagnet on the spinning rotor was to connect their electric coils to a set of spinning switches that use rubbing contacts called “brushes”. The modern motor brushes are blocks of graphite (a form of carbon) that will slowly rub away and need to be replaced occasionally, however, the earliest brushes were bundles of bronze wires that actually looked like a paintbrush. If a rotor has electromagnets, it is called an “Armature”, and the spinning shaft-mounted contacts that the stationary brushes rub against is called a “commutator”

To “commute” something meant to change its direction, and a commutator switched the DC flow from one direction to the other in the armature, as it spun.

One of the basic principles of magnets is that the identical poles push away from each other (N/N…S/S), and opposite poles are attracted to each other (N/S…S/N). If you reverse the current (negative/positive) that is flowing through an electromagnet, you also reverse its magnetic field. Using these basic principles, it is not difficult to design a simple and powerful electric motor that can be scaled up to just about any size. It works.

In 1982, General Motors and Sumitomo each independently developed a powerful permanent magnet, which turned out to be the Nd₂Fe₁₄B “neodymium” magnet (which replaced expensive AlNiCo and SmCo magnets). Although the price and strength of Neo magnets were better than the previous options, they use “rare earth” elements that are not common, and can easily have their access restricted by international trade disputes. Despite this concern, their use expanded rapidly.

The use of neo’s in mass-produced computer hard drives in the 1980s dropped their prices to more affordable levels, and their use expanded even more. This included the motors in electric bicycles, which are very sensitive to the size of a motor, requiring them to be both small in size and still have high power. Which brings us back to the modern era…

In 2003, Martin Eberhard and Marc Tarpenning test-drove a new EV car prototype called the T-Zero, which had absolutely stunning performance, and was nothing like the failed General Motors EV-1. They commissioned a battery pack made from the most recent lithium-based cells, which made the car lighter and increased its range. It could now accelerate to 60-MPH in 3.6 seconds. When T-Zero decided to not convert their original production lead-acid battery pack to lithium, Eberhard and Tarpenning saw this as an opportunity and immediately formed the Tesla car company. Whatever anyone might have previously complained about when it came to electric cars, they could no longer claim that they were all ugly and slow.

An early wealthy American investor in Tesla was named Elon Musk (born in South Africa), and it wasn’t long before he became CEO of the company. Under his direction, the engineering staff has improved their motors and batteries with each version. However, if the company was to grow and become profitable, there were two bottlenecks that loomed on the horizon.

The first is being able to get enough cells to make the battery packs, and the 18650 format cells that they had designed the pack around were made in Asia. Musk decided to form a partnership with Panasonic, with the requirement that an entirely new factory for the Tesla cells would be built in the USA (the “Gigafactory” was built in Nevada).

The second bottleneck was the early models were expensive and somewhat exclusive. All of the parts of the Tesla cars are fairly conventional, except for the battery and the motor. The Gigafactory ensured that Tesla would now have a steady supply of batteries, but the second problem is that…other dramatic changes were needed to lower the costs of the 2017 Model-3 so that they could generate many more sales than the earlier Roadster and Model-S. The Model-3 motor design was targeted for a major reduction in costs, but it could not sacrifice the performance that the Teslas had become known for.

And that leads us back to this article in Cleantechnica, which just blew me away. The motor engineers started with a clean slate and explored every possible option. For the past 100 years, the existing types of motors worked well enough that nobody needed to reinvent the wheel, and the Switched Reluctance (SR) principle was not impressive enough in the early days to warrant spending money and effort in trying to improve it. To be fair, the controllers of the day were simply not sophisticated enough to make up for its shortcomings, and SR motors were known for having torque-ripple that was bad enough to create significant vibration and noise.

Remember earlier when I talked about how industrial inductance motors were being converted to Synchronous Reluctance (SynR) motors? That started around 2011 when a German firm began advertising a more sophisticated controller, which allowed existing Induction stators and cases to simply substitute a SynR rotor for significant savings in cost for new motor installations. Well…those controllers mean that if anyone wanted to dust off the 1888 SR motor design, it just might have a chance of working well now…and it does.

I’ve found references to the Tesla Model-3 SR motor being designed by an engineer who was hired in 2012, so I imagine that was about the time that the motor engineers settled on SR as a motor type, and then began experimenting to see how much better they could improve it. The entire reason I decided that this new Tesla SR motor deserved its own article was that…the Tesla motor engineers made a motor that not only had the same performance as before (*which was excellent), but…they did it in a motor that was actually smaller and lighter, and MUCH cheaper. This means that we are all going to be seeing more products with SR motors in them, so…let’s take a look at what I’ve been able to dig up so far.

Switched Reluctance motors (SR)

As soon as I started researching SR motors, I came across several terms I hadn’t heard before. One of these terms is “Salient” Poles. It means that the focus of the magnetic pole is formed as a protrusion, with the air around it. The stator poles in the PM motors (that we are all familiar with already) have always had salient poles, but they didn’t need a special name to distinguish them during discussions over the years. By also making the poles of the rotor salient, these types of motors are now sometimes referred to as “doubly salient” motors by some design engineers.

The coils in the SR motor graphic above are energized as opposing-location pairs, with both of them flowing their magnetic fields in the same direction. The magnetic flux is shown with blue lines, but only the right side is shown here, there is a left side to the field that is symmetrical in shape (the double loop seen in the bar magnet pic).

Another feature is the number of poles between the stator and the rotor. For the motor to be easily self-starting, they cannot have the same number of poles (if they were, when one set of aligned poles are de-energized, the rest of the poles would already be aligned). I have seen pictures of working SR motors with more poles on the rotor (compared to the stator), but most design proposals have fewer poles on the rotor to emphasize the benefits of having a rotor that is as light as possible.

If a rotor has less mass to accelerate and decelerate, having that feature can be desirable. There have been 2500-RPM SR motors that have fully reversed to full speed in less than one second. The 8/6 pole configuration shown above is symmetrical and easily reversible.

Also, I mentioned earlier that industrial induction motors have a distinct efficiency profile that has them running at a near-constant 3,000-RPMs. SR motors are able to run at unusually high RPM’s if desired. The Dyson company has developed a home-appliance vacuum where it’s SR motor spins up to over 100,000-RPMs.

Another characteristic is that these SR motors run cooler than a PM motor for several reasons. first, there are no flux reversals, like the kind that are typically found on PM motors used in EV’s. As the SR energized coil-pairs are activated in sequence around the motor, they all create magnetic fields that flow the same direction. There are still some eddy-currents created, but nowhere as much eddy current heat compared to constantly reversing the fields in the stator back and forth.

The best way that I can explain eddy current heat is to say that all elements have loose electrons that can be pried out of one atom and shoved into the next atom by a moving magnetic field. Each Iron atom might theoretically have 26 electrons, but in the presence of a moving magnetic field, one atom might have 25, and another might have 27. The farther these loose electrons can be dragged, the more heat is generated (this is a horrible explanation of eddy currents, but its the best I can do right now). This is why the iron inside an electromagnet is not a solid rod, but it is made of a stack of thin plates called “laminations”.

In the pic above, the green arrows indicate the direction of the magnetic flux. The red arrows on the circular shapes roughly show the paths of loose electrons. The individual laminations are electrically separated by a thin and clear varnish that insulates them. Laminations that are 0.50mm thick are common, but the thinner the lamination, the less eddy current heat you will experience. Although, the thinner laminations are more expensive since it takes more of them to fill the stack size. I have seen 0.35mm thick lams as an upgrade, and you can even find laminations as thin as 0.20mm

In the pic below, rotor-(A) is a laminated rotor from an SR motor (6/4). The four salient poles on the rotor use a significant amount of air between the poles to make their steel protrusions a more attractive path for the magnetic flux loops. Examples (B) and (C) are two variations of a SynR rotor. Both (B) and (C) use four magnetic field groups, and both use flux fences to guide the flux loops. (B) uses curved axial fiberglass sheets between the steel lams (which perform slightly better), and (C) uses curved air-slots [which are lighter and cost less].

The Tesla car company, Zero Motorcycles, and Alta Motors all use the outrunner configuration of the motor, and this is where the rotor is located in the center, and the stator is connected to the outer shell. This makes it easy to use some type of external active-cooling (such as an air-fan or a liquid cooling pump) to remove the heat from the stator. If you have the stator at the core and make the outer shell the spinning rotor, that style of motor is called an “outrunner”.

Once you have settled on the largest size of motor that can fit on your design, the way to get more power is to give the stator a temporary burst of amps. This causes lots of heat, and if the magnets on the rotor get too hot (in a PM motor design) then the magnets will get demagnetized and ruined. Of course, these motors have temp sensors, but it just means that if you are running hard very frequently, then you may find that the controller suddenly starts limiting the power you can use until the motor cools down. SR motors can take much more amp-heat and they can do that much more often because they don’t have PM on the rotor.

Shared poles, and putting magnets back in

I mentioned earlier that in an SR design, it’s possible to have the direction of the electromagnetic fields all cycle in the same direction, with the benefit of producing less heat due to having no flux reversals. There is some permanent magnet (PM) motors that alternate the direction of current through the electromagnets, which alternates the direction of the flow in their magnetic fields. Rapid flux reversals cause additional eddy current heat, but it can be one way to make a compact PM motor very powerful.

However, if you settle on using an SR design and you also specify that you want to have all the fields cycle in the same direction (unipolar), you have an additional option you can incorporate. In order to achieve the “double-loop” magnetic field similar to the bar magnet pic (back at the top of this article), the aligned poles on either side of an energized coil in the stator do not “need” to also have coils, and those two poles can be made of steel laminations with no copper, just like the rotor.

In the pic above, the magnetic flux lines are shown from a Finite Element Method Magnetics analysis graphic (FEMM). The “half-motor” example shown is the same configuration as the motor just above it. In this graphic, you can see that the red arrow is pointing to an energized pole (copper wire coil not shown), and the blue arrow is pointing towards the unpowered pole. Since this is a symmetrical configuration using a 2-phase operation, the other mirrored half would be identical. The thin green lines represent permanent magnets that form a skin on the faces of the shared poles (which have no copper wire coils). The thin green PM’s in this location reduce torque ripple, and also smooth-out cogging.

Here is a pic of a prototype above, just so you will know that this isn’t just an engineer’s theoretical idea. If one of the major benefits of an SR motor is the elimination of PM’s, it seems to be a step backward to add them back again, but…take notice that two loose PM’s are shown in this pic. One is set on its edge at 11:00 O’Clock (on the stator), and another is laid flat at around the 8:00 O’Clock position.

Not only are the number of magnets significantly fewer in this type of motor, but they are also very thin. It was found that a magnet of a one-millimeter thickness provided the best results (but expensive), and ordering magnets to be made at a 3mm thickness was the thinnest that provided the most cost-effective option. That is only 1/10th of an inch thick. The PM in this style of motor doesn’t need to be strong, so cheap ferrite magnets are a definite option.

Elon Musk has tweeted that the Tesla Model-3 motor is a switched reluctance, and it uses six poles (in the stator?). That struck me as odd, but if three of the poles are shared, it would perform like a 9-pole stator. If that motor has three large electromagnets in the stator (and three shared poles), then what would the rotor look like? And by what sequence pattern would the stator coils be energized? I don’t know. I know that this may be a let-down after reading all the way through this long-ass article, but I am on the lookout for pics and verification of the Tesla Model-3 motor configuration. I will post them here when I find them.

I also found out that the Range Rover Defender also has an EV variant that uses an SR motor.

Speaking of clever configurations, remember I stated earlier that the shorter the path of the eddy currents caused by magnetic fields has to travel, the cooler they run? And remember how the double loops of the flux normally travel around the entire perimeter of the steel laminations in the stator? I found a brilliant design proposal that uses a 3-phase operation, but instead of six poles, the designer split the poles into six “siamese” pairs, for a total of 12 poles on the stator, and 10 poles on the rotor. You can see below how short the powerful flux loops are. This style is shown in the pic below.

Why are SR motors so hot right NOW?

Once controller electronics became more sophisticated in 2011, and then this advanced style recently became more affordable, it was only a matter of time before someone realized you could make a less expensive motor using the SR type from 1888.

This year, China announced that its production of permanent magnet motors of all sizes and types was growing so fast that they would not be able to export as many of the strong neodymium magnets that these motors needed in order to work.

There is always a pressure to make a better performing product that also costs less, but an announcement that your EV assembly line will no longer be able to get any PM motors is a seismic event. The SR design changes all of that. So now, we are all going to be hearing a lot about switched reluctance motors in the near future.