Chevron fined $422,500 for 3 instances of non-compliant fuels in California; ethanol, olefin and cetane issues
New Honda Fit Hybrid improves fuel economy 35% over outgoing model; best hybrid (non plug-in) fuel economy in Japan

Chevrolet Spark EV: speedy, smooth and quiet; the importance of motor control

2014-Chevrolet-SparkEV-049-medium
The Spark EV. Click to enlarge.

Chevrolet is hosting a series of media drives this week for the production version of its new Spark EV (earlier post). Featuring a GM-developed 105 kW electric motor that pushes out 400 lb-ft (542 N·m) of torque and a 21 kWh pack that supports an estimated 82-mile (132 km) range, the Spark EV is a speedy (0-60 mph in 7.6 seconds); efficient (combined city/highway 119 MPGe); affordable (as low as $17,495 in California net after Federal and state rebates); fun-to-drive electric vehicle targeted for the urban environment.

There are multiple interesting aspects to the vehicle, including GM’s rationale for offering fast charging as an option (it’s not to go across country), but one of the most manifest when driving the car (aside from the “Wheee!” of those 400 lb-ft in action) is how smooth and quiet the drive is. Much of this derives from the efforts GM has put over the years into motor control and active damping for hybrids and EVs.

Drive quality was a top priority during the development of the electric motor for the Spark EV, GM’s engineers on site all emphasized. While details on how the damping system works are proprietary and GM is not discussing specifics, a general outline of some of the control problems with electric drive can provide some background.

Electric motors have two primary parts: a stator and a rotor. The stator comprises a steel core that supports the motor winding, while intensifying and directing the magnetic fields between winding and rotor; GM uses bar wound configurations in its permanent magnet motors. The rotor—placed inside the stator—is built around a central hub that links to a shaft to transfer torque.

In a permanent magnet motor such as the Spark EV motor, the rotor comprises thin steel laminates that stack together into sections to form the rotor core. Each of these laminates is loaded with magnets that are placed in a carefully engineered pattern within the steel.

To generate torque from the motor, a controlled electric current is injected into the copper bar windings in the stator, creating a moving electromagnetic field. This pulls on an opposing magnetic field in the rotor, which creates the twisting force that is torque. In GM’s patented rotor design, each section is skewed to distribute efficiently the rotor’s magnetic field, while reducing noise. (Internal GM tools allowed the engineers to model how the motor could make noise and we were able to manage noise at the source.)

Optimizing the control of the current flow to the motor is the basis for efficiency, drive quality and noise. Basically, the controller requests a specific amount of voltage and current appropriate for driver demand and rotor position (commanded torque). However, voltage and current actually present may not be what is commanded, and the rotor position may not be what is expected. That results in actual torque.

The problem is non-trivial; traction motor control systems must deal with a wide range of requests based on speed and traffic conditions, route, weather, roadway conditions, mass load, etc. As a result, the problem has generated volumes of patents and research papers exploring different methods and algorithms.

The charts below show two scenarios for a generic traction motor. The top two charts depict a correctly functioning motor; the blue line is commanded torque, the green line actual torque. The bottom two charts depict a traction motor with a problem.

Known good
Known bad
Generic example of traction motor control (not from a GM vehicle). Blue is commanded torque; green is executed torque. The top two charts show a motor with no issue; the bottom two show a problem. The ideal for motor control is to match executed torque to commanded torque seamlessly (i.e., smooth). Click to enlarge.

In addition to basic motor speed control to deliver smoothly what the driver requests under all the varying conditions of a drive, there is damping control (active damping) intended to reduce transient driveline oscillations before they can reach the drive wheels of the vehicle, or to reduce the impact of an external perturbation.

As an example, take a situation in which you accelerate, and there’s a bit of unevenness due to a wheel slipping on a bit of gravel. If the sensors and processing of the input are sufficiently quick, active damping control of the motor can cancel out those effects in near real-time.

In one of GM’s more recent patent applications on motor control (2012), the inventors assert that speed control and damping control should be integrated, as decoupled damping and speed controls can produce discontinuities in the applied motor torque. In other words, damping and speed control torques should be designed together so that the damping torque does not “fight” a separate speed control—for example, when damping torque is in one direction and speed control torque is in the other direction.

In the approach described in this patent application, GM proposes calculating a speed control torque signal for the traction motor using a motor speed torque (MST) control block of the controller, and calculating the motor damping torque signal using a motor damping torque (MDT) control block of the controller.

The speed control torque signal and the motor damping torque signal can then be combined to generate a total motor control torque for the traction motor. The total motor control torque may be processed through a vehicle driveline model to generate an estimated damper torque, estimated axle torque for the axle, and an estimated wheel speed for the drive wheels. In the method described, the estimated damper torque, the estimated axle torque, and the estimated wheel speed back can then be fed back to the MDT control block as inputs to the MDT control block.

The above is by way of illustration; we have no idea whether or not the company is using any or all of this approach in the Spark EV.

The firsts of the Spark EV. As noted earlier from the preview in November (earlier post) the Spark EV is offering a few firsts. It will be the first vehicle on the market to use the recently approved SAE combo charger for DC Fast Charging (J1772). (Available as a ~$900 option later this year.) Also, its liquid cooled battery pack, developed with tailored A123 Systems Li-ion iron phosphate cells and GM controls, can handle multiple DC fast charges daily without impact on battery life. A fast charge can bring a pack depleted to the lower range of its state of charge window up to 80% capacity in approximately 20 minutes, with lower charge levels in correspondingly smaller amounts of time.

GM is eyeing the fast charge option as one way to address a potential roadblock to urban EV adoption: the lack of a home charging option for many urban dwellers—i.e., no garage or car port in apartments, etc. The company really doesn’t expect Spark EV drives to stutter across the country in 80-odd mile increments punctuated by 20-minute recharges, suggested Britta Gross, Director of Advanced Vehicle Commercialization Policy, during a briefing at the Spark EV drive in Portland, Oregon.

Currently, DOE data suggests that 90% of charging is done at home. However, the ability for a public fast recharge in cities could enable those urban dwellers without a viable home charging option to buy the Spark EV, she said.

Propulsion system. The Spark EV propulsion system comprises the motor, the gearbox, and the power electronics. GM placed a great deal of emphasis on reuse of components, said Stephen Poulos, Global Chief Engineer for eAssist and Battery Electric Propulsion at the briefing. The high-performance motor is shared with another—unnamed—electrified product, and the gearbox is right out of the production 6-speed automatic transmission.

When we are developing a new propulsion system, we have a lot of requirements to meet: safety requirements, performance requirements, all kinds of control interfaces, we have electrical requirements. For this particular product, the big priorities were, one, we needed to go fast and limit the new investment, exploit what we had. Reuse as much as possible. There is a lot of reuse in what we put together here.

Second, we needed to be very efficient. For the same reason that we are using a smaller vehicle that doesn’t require as much battery as a larger car, if we can have a very efficient drive system, we can use the energy of the battery more efficiently. We don’t need as much battery. There was a lot of pressure to keep this as efficient as possible.

The third priority was fun to drive. We wanted to be this car to be a blast to drive.

—Stephen Poulos

Driving. The Spark EV offers a P R N D L indicator; “L” (low) is for maximum regeneration. The Spark EV also offers a Sport mode that delivers improved pedal feel to the driver (without actually increasing torque). After experimenting with the different settings, we found it most comfortable to drive in Sport mode in L.

In the city, especially on the flat, having the vehicle in L basically means never having to touch the brake pedal. The regen is aggressive but smooth, and is more than capable of bring the vehicle to a stop in slow city traffic without brake intervention. Driving on hills, L does a splendid job of re-topping the battery. In California, one of the key initial markets, the Spark EV would be a slam dunk for the hills of San Francisco.

At the same time, the electric acceleration is a quiet delight. Even at the higher end of the speed range on the highway, there is ample power for overtaking. The Spark EV’s maximum speed is limited to 90 mph.

2014-Chevrolet-SparkEV-067-medium
The Spark EV instrument cluster. The range indicator is on the left, with the likely range displayed in the blue circle, along with a graphic indicator of level of battery charge as well as max and min range possibilities given the driver’s recent behavior. As you drive, the indicator will display little bubbles (not shown in the picture) moving up along on the min/max curves to show that you are trending toward one extreme or the other. The goal of this, said GM, is to provide a great deal of confidence with respect to range.

The indicator on the right is a direct readout of power output, or input (in the case of regen). The green “Ready” icon and light at the bottom of the display is simply an indicator that the car is powered up and good to go.

Assistance and information. The Spark EV is far from a spartan car. It is very comfortable to sit in and to drive (especially with the 2LT upper trim level and its steering wheel), but it also is a bit pared down in certain areas in an effort to keep both weight and cost down.

As one example, there is no array of advanced driver assistance functions such as collision warning and avoidance, parking assistance, and so on. Those all add weight and cost.

However, the Spark EV is fitted with the standard Chevrolet MyLink Infotainment system, including a seven-inch color touch screen that shows critical vehicle EV functions. OnStar with three years of Directions & Connections plan is included. Other functions (such as the Bringo navigation app) are made available via a Bluetooth link to your smartphone. (Bringo also features charging spot locations.)

Resources

Comments

Engineer-Poet

Tell me how the NRC's mandate for more (and better armed) security on the site has any relation to the problem solved by an additional pipe for SNF pool makeup water.

That security is very expensive.  I want to know what problem it solves, or if it's just another way for a nuke-phobic bureaucracy to make nuclear power less competitive.

Kit P

"nuke-phobic bureaucracy "

E-P has in no way supported his allegation. My experience is that the NRC is not afraid of nuclear power. The NRC is very knowledgeable about the risk of nuclear power.

There have been a few things that change how we think about safety such as TMI, the Browns Ferry Fire, 911, and more recently, earth quakes that were larger than expected.

We evaluate those issues and the NRC reviews our responses. Utilities that come up with the best solutions often set the standard for the rest of the industry.

Nuke-phobic groups jump up and down but are pretty be the NRC because they do not have much positive to offer.

Engineer-Poet

I'd like to see Twït P explain the nuclear-phobic and ignorant positions of the ex-NRC chair, and the strongly anti-nuclear history of the current chair, plus everything about the agency's "regulatory ratchet", if it is not in fact anti-nuclear.

Larzen

@ BrotherKenny4:
Re: "dying in the desert for oil companies' profits". Oil not the only consideration here; Geopolitik is also in play. If you are concerned about where we get our oil, perhaps your dear leader should have made it easier to get the Canadian oil that is now going to China instead of to us. As long as we are still, mostly, an oil-based economy, perhaps he should be a little more flexible about exploration and recovery on Federal lands. So far, his subsidies of domestic solar and wind have been disastrous. Soviet-style top-down planning has never worked very well, but this man doubles-down on every mistake he makes. That being said, I think Kit's blanket statements about the technology of EVs being a total waste of time are misguided and counterproductive. As nascent as this technology is, like the Space Program, we have derived many benefits, if only incrementally and seemingly at a snail's pace. We all want instant gratification. It ain't gonna happen - it has never happened. technological advancements are won over painstakingly slow steps over rough terrain. This technology will pay off in the long term - the tree from these seeds can already be envisioned by the sprouts we see coming from the ground. Don't be such a party-pooper. No one likes a party pooper. Your negativism overrides your whole outlook. I'd much rather have a Tesla than my Corolla, and one day cars like it will be affordable to the middle class and they will not give us the range-anxiety that we now have. Be patient, grasshopper.

ai_vin

@Larzen
The pipeline you're talking is taking the oil south to be refined on the Gulf Coast, so it is destined for export. And because it could then be sold on the world market it will be sold at world prices so those Americans who are getting the oil now would have to pay higher gas prices.

http://indiancountrytodaymedianetwork.com/article/the-truth-about-keystone-xl%3A-few-jobs,-no-energy-security-60041

http://www.canadianprogressiveworld.com/2013/07/22/keystone-xl-tar-sands-pipeline-raises-american-gas-prices-report/

Kit P

“EVs being a total waste of time”

Of course I did not say that. I would love for the power industry to take market share. It is the silly things people say about BEV. For example,

“I'd much rather have a Tesla than my Corolla, …”

Why? Folks here are so shallow. Marketing is based getting people to want things that they do not need. If a BEV was a better choice than a Corolla, I would won one.

How does making informed decision turn into a negative?

Kit P

@E-P

“ex-NRC chair”

He is gone, I am glad. It would appear that he had little influence. I do think that a contrary opinion has value. We must be able to defend the technical merit of our design.

“plus everything about the agency's "regulatory ratchet", if it is not in fact anti-nuclear”

I would explain that safety is pro-nuclear. Over the years the NRC and industry has worked on issues based on importance to safety. Generic Safety Issue like GSI-191 gets a lot more attention than say small amounts of tritium.

While it may not be obvious from reading blogs, when the NRC audit has ‘findings’ that appear trivial it is an example that the NRC did not find serious issues. I once had an innovative idea but the NRC had another opinion that was not so complementary. Two inspectors showed up from Washington to provide oversight for my ‘radical’ test. The most serious written comment was that correct word was ‘ensure’ not ‘insure’. I was able to demonstrate that my plan including a test would provide a safer way of doing things.

I would rather work at nuke plant with two resident inspectors and army of visiting inspectors than a coal mine with overworked MSA covering many mines.

Engineer-Poet
He is gone, I am glad.

And many others too.  But McFarlane may be no better, just a lot quieter.

It would appear that he had little influence.

He had way too much influence, including deep-sixing the release of the report on Yucca Mountain and ordering US forces to stay at a ridiculous distance from the Fukushima meltdowns where they could not assist tsunami survivors.

More to the point, the influence of his political bosses has increased (Markey is now a Senator).  These people need to be tossed out of office (as a first step).

I would explain that safety is pro-nuclear.

Obsolete, redundant and needlessly expensive "safety" regulations reduce safety by driving nuclear power out of the market, replacing it with more dangerous power using more dangerous supply chains... like coal.

John Russi

New Spark EV is amazing car its speed of 0-60 mph in 7.6 seconds is wow! and average is also good.

The comments to this entry are closed.