How can we stretch the rare-earth metals supply for vehicle motor magnets? Phillip Totaro examined this question in these pages earlier this year from the point of view of motor design internals and wind turbines.
But top-level systems and circuits design can also have a major impact on rare-earth consumption, particularly in the automotive arena.
Consider the Toyota Prius and the Honda Insight. The original Insight, that is: the aluminum two-seater. The Prius has about 80 kilowatts of installed motor generator (MG) capacity; the Insight, 10 kilowatts. What's more, it uses only 12 percent as much rare earth materials and boasts 40 percent more passenger capacity -- while getting 10 more miles per gallon (MPG).
Informed observers may object that the Insight is a 'mild' hybrid, unable to roll away or reverse in electric-only mode, and that the Honda mild hybrid system has done poorly in heavier vehicles such as the hybrid Accord. All true -- and all beside the point. As we shall see, there are general lessons to be learned from the comparisons of the two vehicles, and there are ways around the mild-hybrid limitations. The importance lies in thinking about systems engineering as means of conserving rare earth materials.
Here are six ways to do it:
1) Avoid series architectures like the plague.
The series architecture is regarded as the more “pure” architecture. It requires the ability to send the peak power through not one but two permanent magnet (PM) devices. The parallel architecture is less versatile. But it sends only a fraction of the peak power, and moves through a single device.
Various compromises have emerged between the two models, exemplified by the split-drive of the Prius, which uses a planetary gearset to apportion power between mechanical and electrical drive paths.
In general, the closer to a series architecture, the higher the rare-earth consumption. The Prius is probably about a 60-40 series/parallel mix (modeling the mix is not simple).
2) Gearshift the motors.
The elegance of stepless operation -- that is, no low gear or high gear -- is tempting. But while electric motors have a broader speed band than a combustion engine, they are by no means indefinite or truly flat. On the torque-speed map of a Prius, the motors are sometimes straining far from their sweet spots. By contrast, the Insight has the motor ahead of the transmission, so it benefits from the gearshift. This is also what keeps it from doing electric-only rollaway: it must drag the engine.
Tonier gearing systems fix that. GM’s hybrid Yukon/Tahoe has four gearing modes for its two motors, allowing them to haul this heavy full-sized SUV. It can start or reverse electrically as a full hybrid. (The limited 21 MPG is probably from the heavy NiMh battery triggering the engine upsize to 6.0 liters. Lithium to the rescue?) This system is also used in GM's hybrid buses; it shares a lineage with BMW and Mercedes-Benz -- all heavy, torque-thirsty vehicles.
A lighter example: Brammo Motorsports, of Ashland, Oregon, have been conquering the electric motorcycle racing circuit, recently winning the TTXGP North American Championship. They have now put a six-speed gearbox in their Empulse model.
3) Invest in body weight reduction.
Thirty percent weight saved is 30 percent torque saved is 30 percent samarium-cobalt (or neodymium-iron-boron) saved -- at least when it comes to acceleration. Again, comparing the Insight to the Prius, the Insight, with its aluminum body, is over a half-ton lighter. Lightweight body obstacles are well known, and they include cost, fabrication, and crash safety. There are also reasons for hope. Honda’s thixotropic forming attacks the inability to cold-roll aluminum. In Iceland, the Reydarfjordur aluminum mill harnesses the melting Vatnajökull glacier for electricity. This approach has not been immune to criticism, but at least it’s using global warming to fight global warming. On the composites front, the Tesla roadster and the Boeing 787 Dreamliner both bring composites further into the commercial body market.
4) Use different, or novel, types of motors.
Induction motors are the first fallback. They’ve been around since Nikola Tesla's time, using basic alloys of iron, aluminum, and copper. They’re capable of greater than 95 percent efficiency. But it’s trickier to get the best performance out of them. Also, conventional wisdom holds that an induction motor is twice as heavy. A good designer can get around that by spinning the motor much, much faster. This does require advanced geartrain design, but it’s definitely doable -- the Tesla roadster, for example, uses an induction motor.
Another rare-earth-free motor is awaiting its day in the sun: the switched-reluctance (SR) motor. It’s not easy to casually describe the difference between the SR and the PM, but the end result is that the SR is able to achieve very high speed, power density and efficiency -- perhaps greater than the PM -- using cheap iron materials. Obstacles to date, such as vibration and torque ripple, can be attacked, usually by putting more power transistors in the drive inverter (the more, the merrier, to an extent). Falling electronics costs help. I also wonder about a “compound” motor, combining the SR motor with an induction or PM motor on the same shaft.
Finally, I wouldn’t quite rule out some new miracle motor. I have an email in my inbox forwarding such a proposal. I haven’t had time to analyze it; maybe it will be a crock. It was, however, sent by a highly experienced switching power converter engineer.
5) Invest in keeping the motor-generators cool.
Magnets perform better when cool. Bolt them to a hot engine, and you have to salt them with rare, expensive dysprosium or terbium to keep them working. The alternative is better cooling, but the available choices are fraught with unresolved contradictions. A remote motor (axle or hub mount) is a cooler location, but implies a series architecture and has poor access to central cooling systems, whether oil or glycol (hence some limits on the very powerful rear-axle motor of the 4WD Hybrid Highlander). Casting air fins onto an oil-filled housing is the cheapest though not the coolest solution -- but if everyone does it, the magnets may become too expensive worldwide. Designers, roll up your sleeves.
6) Don’t bite off too much with the design specification; Don’t bite off too much range in a plug-in hybrid.
In engineering, incrementalism often pays. Consider the Chevy Volt. The 40-mile electric range came with a price: 435 lbs. of expensive batteries, hauled around by 166 kilowatts of expensive PM motor-generators. If GM had taken the Volt, cut the electric range by say 40 percent, and put the money saved into an aluminum body and a shrunken, turbo-charged engine (sometimes referred to as a Type I plug hybrid, instead of a Type II) -- we might have a more satisfactory package. All the various categories of savings -- battery, motors, and weight -- would snowball with one another. Ultimately, the Volt wouldn’t be nearly the rare-earth hog that it is.
Curiously, as hybrid sales went ballistic in about 2004, rare-earth metals underwent a major price fall driven from China. Demand has now reversed the price situation. Will non-mild hybrids prove to have been a passing phase, enabled by an ephemeral moment in the metals market? I hope not.
I do believe that we have many options and tools within reach to overcome the rare-earth metals crunch. These types of commonsense systems engineering solutions may contribute a great deal. Of course, there will be limits to the range of achievable improvements. But these innovations will complement the digging of new mines -- and are probably easier to achieve than dredging the ocean floor or recovering nodules from stray asteroids.
Doug Widney is a San Francisco-based energy consultant with a diverse electric greentech practice. In addition to strategic advising, he remains an active circuit designer. An occasional Calcars volunteer, he cut a modest amount of metal on the first Prius PHEV conversion. Reach him at firstname.lastname@example.org.