A solution so obvious, it must be wrong.
August 17, 2011 8:48 PM   Subscribe

Why don't electric cars use smaller motors with multiple-speed transmissions?

Everything I've read while researching an electric car purchase says that, since electric motors produce the most torque at rest, and produce such copious amounts of it, that (with one exception) electric cars don't use multiple-speed transmissions -- because they're too fragile, and because you can get away without it. For instance, Tesla tried like the dickens to work out a two-speed transmission, but eventually shipped the cars with a single-speed (the higher gear.) Similarly, to avoid halfshafts breaking along with transmissions, it is apparently typical to limit current to the motor initially, so that it puts out less power. Unfortunately, running single speed transmissions means that the motors have to run at high (and significantly less efficient) speeds as the car's speed increases.

Given that combination -- motors too powerful at low speeds, and not efficient enough at high speeds -- it seems like a no-brainer to run a much smaller electric motor and a multiple-speed transmission. The full low-speed torque output (without current limiting) would presumably be sufficient to get the car moving, but low enough not to break the gearbox, and the inability to produce as much power at higher speeds would be solved with the now not-imploding multiple-speed gearbox. Plus, smaller motor and lower speeds equals better battery life and less heat.

It seems so obvious, in fact, that I don't understand why manufacturers aren't taking this approach...but they're the geniuses, I'm not, so there must be a good reason. Please advise.
posted by davejay to Travel & Transportation (15 answers total) 1 user marked this as a favorite
More parts == higher cost and lower reliability.
posted by Bruce H. at 8:52 PM on August 17, 2011

The Chevrolet Volt uses your approach (actually, it does something more clever - one of the motors can be disengaged from the drive train and used as a generator - even while the car is driving)

The bit that is applicable to your question, is that the primary motor drives the car at low speeds, and at high speeds, both motors drive the car, and they geared in such a way that they add their speeds to each other, instead of their strength, so both motors are running in their sweet spot even though the car is going fast.

GM got a number of patents on this drivetrain, which may or may not (I wouldn't know) hinder other manufacturers from using similar arrangements.
posted by -harlequin- at 9:06 PM on August 17, 2011

Here's a useful post on the Volt's drivetrain. As Bruce H. alludes to, and as you sense reading the linked post, it's actually quite technically challenging to do what you're describing. Electric cars are so expensive by virtue of their batteries that manufacturers want to keep costs down.

Also, the Volt and the Leaf are the first cracks that major, technically sophisticated automakers have taken at electric propulsion since the EV1, so you haven't had the resources behind electric vehicle development that are needed to develop a drivetrain as complex as the Volt's. I would note that Nissan stuck with the less-complex single-speed drivetrain (and that they also underengineered their battery packs with air cooling instead of the Volt's liquid cooling - their customers will pay for that in about five years).
posted by Dasein at 9:17 PM on August 17, 2011

More parts == higher cost and lower reliability.

Speculation: The added weight may also go some distance towards eroding any efficiency gains. A transmission capable of transmitting fair amounts of torque is a heavy thing.

posted by -harlequin- at 9:22 PM on August 17, 2011

Best answer: AC electric motors can be designed to have very flat efficiency curves over speed. The general rule for big PMSM and similar AC motors is that the curve is flat above 20% of their maximum speed. For small motors it's more like 40% max speed. At around 10% max speed they're running at about half efficiency. Imagine a hypothetical electric car with a max speed of 80 mph. How often do you spend at full load (max torque/acceleration) below 8 mph? Sure, you might do some light-load driving down there in traffic, parking lots, etc., but very little foot-to-the-floorboards work. How often do you spend driving above 16 mph?

So why add a gearbox that's always expensive, complicated, and a couple percent inefficient when you can get away without it?

(I'm not saying that we won't see transmissions in electric cars eventually, but we'll be chasing those little diminishing returns at that point.)
posted by introp at 9:35 PM on August 17, 2011

Best answer: I grabbed the first promising link from a google books search; note the efficiency curve graph upper-left. This book is referring to optimization methods so that's just a generic motor curve, but it gives some idea of just how fantastically efficient synchronous motors can be over speed.
posted by introp at 9:47 PM on August 17, 2011

I always wonder what happened to the CVT transmission.
posted by Wild_Eep at 9:50 PM on August 17, 2011

Everything I've read...says that...electric cars don't use multiple-speed transmissions

Multiple speed transmissions are old hat anyway; CVTs are the new hotness. Continuously variable transmissions are perfect for electric cars as they maintain the electric motor at a relatively constant speed (selected at an optimum efficiency/torque point for the motor) and adjust the vehicle's forward speed by varying the transmission ratio over a smooth continuous range.

Power sharing transmissions have been developed specifically for hybrid engine models, such as the Prius. PSTs can not only continuously adjust the transmission ratio for forward movement, but also dynamically vary the ratio of the contributions of the two engines in a continuous range.

On preview: beaten by minutes by Wild Erp!
posted by ceribus peribus at 9:55 PM on August 17, 2011

CVTs are useful for parallel hybrids (where the IC engine is coupled to the wheels); a generic IC engine can easily give up 15-30% efficiency when you double its operating speed. A really good toroidal CVT is somewhere around double the inefficiency of a manual transmission (the actual numbers are closely-held secrets but the things have been generically modeled and practically examined in many papers) but you make that back before too long in IC fuel savings.

Note that the gains aren't huge, though; available CVT cars generally only have slightly higher practical efficiencies than manual transmission cars (see the Nissan Cube at 25/30 mpg for a manual and 27/31 for the CVT with the exact same engine; admittedly the Cube uses a belt CVT which is less efficient than a toroidal-type, but cost is a practical matter).

If you're worried about keeping an eletric motor in a series hybrid or pure electric vehicle at an optimal constant speed you're either at a considerably higher energy price-point than a car (like a power plant where chasing 1% efficiency means megabucks) or you need to go study up on electric motor theory and practice.
posted by introp at 10:33 PM on August 17, 2011

This Economist article about the characteristics of induction motors, and their relative attraction to manufacturers and users alike in electric cars, might interest you. Note also the mention of neodymium - required for designs with permanent magnets. Since this material is hard to get - with a supply controlled by China - it has become a factor which is influencing the choice of motors put into designs - and hence everything else such as possibilities for transmission.
posted by rongorongo at 10:42 PM on August 17, 2011

I always wonder what happened to the CVT transmission.

My understanding is that Formula One banned them, on the grounds that they removed too much of the human element from the racing.
F1 is the automotive technology trendsetter. Once CVT was a non-starter in F1, it was a non-starter everywhere else. If F1 cars have buttons that you push to change gears, then (ten years later) that's what all the hottest street cars will have too, and it's what the car enthusiasts will want.

But CVT seems to be superior and so will have its day. Electric cars might be how that happens, because they aren't as influenced by the race track as gas cars are.
posted by -harlequin- at 10:54 PM on August 17, 2011

My understanding is that the Tesla roadster EV was initially designed with a 2 speed transmission, but the torque of the electric motor kept breaking it. They worked with several transmission companies to engineer a transmission that could handle the torque of the electric motor, but ultimately found the best course of action was to remove the transmission entirely.
posted by j03 at 1:02 AM on August 18, 2011

Best answer: I think you are over estimating the high speed inefficiency. Electric motors are most efficient at their designed speed, and less efficient at slower speeds due to whatever the speed control mechanism is.

Transmissions, however, are always inefficient. They add drag, period. They also add weight. If you gain 5% efficiency by adding a transmission, but lose 7%, it's a loss.

Better to design a single motor that does what you need than add in messy gears.
posted by gjc at 6:10 AM on August 18, 2011

CVTs aren't gone. I have one in my Suzuki Kizashi, and my friend has one in his Subaru Outback. The one in my Suzuki is built by JATCO, which is a subsidiary of Nissan but makes transmissions for a large number of car makers.
posted by Quonab at 8:08 AM on August 18, 2011

Response by poster: Thanks, folks! One question answered in a very useful way. Much obliged.
posted by davejay at 12:08 PM on August 19, 2011

« Older Experience with Stewart, Cooper, & Coon?   |   Do I have a potential problem with aggressive... Newer »
This thread is closed to new comments.