GM-VOLT : Chevy Volt Electric Car Site

Lithium batteries and cold weather do not go well together. An incredibly important part of engineering the Volt, and electric cars in general, is to assure that they will function properly in very cold weather.

My MINI-E which is rudimentary and relies on cabin air to warm the pack, was just  in the shop having been towed for the third time in its short 7 month life. In temperatures below 32 degrees I regularly only get about 55 to 60 miles of range. I came out of my workplace two weeks ago to find the car unable to start with the battery registering zero.  A “battery module” was replaced.

The Volt battery has its own independently regulated software-controlled liquid thermal management system that ensures stability, reproducibility and longevity.

GM’s Voltec engineering team are on another road trip to Kapuskasing Ontario to cold-weather test the Volt.  Kapuskasing is 500 miles north of Toronto and this time of year has temps regularly below zero.

As Volt engineer Pam Fletcher writes:

We want to see how the vehicle responds in temperatures as low as -40 degrees C. Basically, we try to simulate customer behavior to be sure the vehicle responds exactly as a customer would expect.

Despite the frigid temperatures, the Volt is engineered to handle extreme conditions.  The battery is warmed up during plug-in charging, which is recommended particularly in cold climates, but we realize not everyone will do this.  So at night, we plug-in some vehicles and some we don’t.  We want to ensure the vehicles start in the morning, or if the battery is too cold, we want to be certain the engine-generator starts first to protect the battery.  The engine-generator system will provide energy to heat battery if it was not plugged in or to supplement battery temperature.  By the time you remote start the car, or remote cabin conditioning as we refer to it in the Volt, pack up your things and get in, the car is ready to go.

To learn more about cold weather testing the Volt, and to ask your own questions, tune in at 730 PM EDT for a live real time chat with Pam and Volt lead engineer Andrew Farah in the box below:

News about Toyota’s throttle and now braking problems have been filling the nation’s headlines.

The company had to recall 2.3 million vehicles and halt sales of 8 popular models of new cars.  The problem identified is that the throttle can get stuck in an unintended acceleration leading to loss of control, accidents, and death. Toyota has engineered a fix for the sticky throttle involving installation of a small steel beam that increases pedal resistance.  These are now just being outfitted to recalled vehicles by dealers, but may not tell the whole story.

Another problem we reported previously, is also surfacing more intensely and surrounds braking failure occurring in the 3rd generation 2010 Prius.  In this case, if a driver goes over a bump or pothole, the brakes can seem to disengage again leading to loss of control.  This too has resulted in accidents. Over 100 complaints about this have been lodged in the US and Japan.

Though there is no recall of the Prius yet, both the Japanese and US governments have launched investigations.

In both cases there may be an underlying theme.

Modern high tech vehicles rely on electronic controls to interface between the driver and the mechanical endpoints.  In the case of the accelerator, sensors measure the movement of the pedal and send an electronic signal to a control unit that operates the throttle.  In the Prius brakes, electronics and software monitor pressure on the brake pedal and wheel speed, and use the information to utilize regenerative braking whenever possible and friction braking when needed.

One unifying theory is that there may be failures in the electronic control units for these functions.  These failures potentially could be caused among other things by electromagnetic interference from other electronic components in the car.  US transportation secretary LaHood indicated that though Toyota denies it is the electronics at fault, those components will become the center of a new investigation.

So while all of this negative publicity is resulting in massive losses for Toyota in sales, quality perception and customer loyalty, and equivalent gains for competitors including GM, it may really be a wolf in sheep’s clothing.

After all, no car in the world will rely more on electronic controls than the Chevy Volt.  If Toyota couldnt prepare for the disaster they are now faced with in a high volume car like the Camry or the third generation Prius, can GM be sure the Volt will operate perfectly in just a few months based on the behavior of 80-odd preproduction prototypes?

“We should have no issues,” says GM Volt executive Tony Posawatz.

UPDATE: On Thursday Toyota issued a statement claiming the Prius brake problem was a software glitch involving the switch from regenerative to friction braking at the same time as anti-lock brakes kick in. They claim to have corrected the code in late January but haven’t decided how to manage the older cars. A recall hasn’t been ruled out.

The original Chevy Volt concept car was described as having twin 6 gallon gas tanks, 50 MPG in charge sustaining mode, and thus 600 miles of gasoline range.

With production, these parameters were changed.

GM has not announced the production Volt’s miles per gallon in charge sustaining mode, though the evidence suggests it will be somewhere between 30 and 50 MPG.

The size of the gas tank has also not been released yet, though in November Edmunds claimed it was 8 gallons, and went on to speculate that the Volt would thus get 38 mpg in charge sustaining mode.

“That’s interesting speculation on their part,” said Volt chief engineer Andrew Farah.  ” But I haven’t told anybody who’s asked how big it is.”

“We’re not releasing the size of the tank yet,” he added.

Previously, then Volt executive Frank Weber told GM-Volt.com that the tank would be between 6 and 10 gallons, and Farah confirms at least that it is less than 10 gallons.

But why is GM being so vague about this figure?

“The reason we’re not (announcing it yet) is we want to make sure we get over 300 miles of fuel range,” says Farah.  ”We’re going to tweak it as such and I’ve got plenty of time to do that.”

Thus it seems GM is aiming for 300 miles of gasoline range, and therefore average real world charge sustaining miles per gallon will be the principle determinant of how many gallons is needed to reach that goal.

The graphic above shows the underside of the Volt after a crash test.  The light blue object behind the battery is the fuel tank.  In the graphic below the tank can be seen from above sitting behind the T-shaped battery pack.

How big is it? You decide.

The Chevy Volt’s gasoline engine/generator has a unique role in this car. It doesn’t operate as do gas engines in conventional cars in that it does not turn the driveshaft. The Volt is first and foremost an EV and the engine’s only job is to spin a generator to make electricity whenever the battery reaches its low point.

We have learned that the generator will operate through an ideal RPM range anywhere from around 1400 to 4000 RPMs. The RPM and the engine load are varied by the car’s power electronics based on the need for power, which is assessed and adjusted continuously once the pure EV range is depleted.

Combustion engines run at one of two four-stroke cyles. The most commonly used is the Otto cyle. The Atkinson cycle is more efficient as it allows all four strokes (intake, compression, power and exhaust) to occur in a single turn of the crankshaft. This allows the power stroke to be longer than the compression stroke thereby generating more work for the same  amount of energy.

Toyota, for example, uses an Atkinson cycle engine in its Prius to maximize efficiency and thus fuel economy.  Atkinson cycle engines tend to be lower power and therefore slower off the start, though that is offset in hybrids by electric motor supplementation.

Many people have therefore asked whether GM will use the Atkinson cycle for the Volt’s generator.

GM-Volt has learned from reliable sources it will not. The generation one Volt engine is an Otto engine.

We have no more information than that or the reasoning why, but feel free to speculate in the comment section.

Keep in mind, GM of course has access to Atkinson engines as they are used in their 2-mode hybrid trucks currently on the market.  The Volt generator will also be used as a common part as it will also be the powerplant for the high volume Chevy Cruze.

The two recent media test drives of the Chevy Volt’s in charge sustaining mode were seen as positive by 85% of GM-Volt.com readers.

Both reporters mentioned however that they could occasionally notice the generator revving after charge sustaining mode was well under way, though neither coud detect it when it first came on.

I was able to communicate with one of the reporters, Lindsay Brooke of the New York times, about his experience.

“Yes, when the ICE first kicked in (on an uphill climb) I could neither hear nor feel its engagement,” he said. “It was completely seamless. I only noticed it because I was keeping a close eye on the cluster—the icon noting battery charge changed over to the icon showing a gas pump which denotes the switchover and how much range remains using the generator.”

“The reason the ICE generator engages at random times is due to its current control regimen for charge-sustaining mode,” he said. “As the quote from Tony noted, the controls engaged the generator, then shortly thereafter called for another of the pre-set charging speeds (rather like, “Oops, I needed more juice than I previously anticipated.”).”

“It was not in situations where I was flooring the pedal, which as you know has no connection to the ICE’s throttle control. Nor was it necessarily in heavy-load situations. A couple times the ICE engaged when the car was going downhill, under what would be light load in a conventional vehicle,” he said.

Brooke added, “I think we’ll be pleasantly surprised when we drive the production-spec car.”

Andrew Farah, the Volt’s chief engineer explained to GM-Volt,” there are still points at which operation of the engine generator is more aggressive than we want it to be, and we want to make it operate less aggressively.”

He says the engine never generates more power than the car needs but may generate it more quickly than necessary. Engineers are able to vary power production by both varying the engine’s RPM and its load, and that along the RPM-load plane there is the third dimension of efficiency which has to be taken into account.

These variables are distinct from the NVH (noise-vibration-harshness) component which is what the customer actually perceives. The team must work within the constraint boundaries of the “NVH ceiling” at the high end and the permissible limit of dip into the battery reserve at the low end to achieve the lowest NVH possible at all times.

It is still at this point a work in progress.

Farah also notes that they don’t let the engine to run at all at low speeds because there is less ability to mask its noise.

Overall, 99% of driving in the integration prototypes is without audible engine noise, there are rare ocassional spikes of sound that the reporters noticed, which will be ironed out in time for production.

I had a discussion with Alex Cattelan, the Chevy Volt’s chief powertrain engineer about the engineering design and operation of the Chevy Volt’s charge sustaining mode.  This is the mode that occurs after the car has depleted the first 40 miles of  range and the gas generator has begun providing electric power.

When you first unveiled the Volt and it was a math model, the car was promoted as getting 50 MPG in generator mode. Now that there are real world parts and parallel hybrid like the Prius verse series. Can you speak about the efficiency difference between series and parallel hybrid operation?
We’re tuning our fuel economy right now. From an architectural perspective there are differences between series and parallel hybrids , there’s absolutely no doubt about that. The issue you mentioned the Volt is a series hybrid when we go into charge sustaining mode or when the engine comes on. We like to think of it not as a hybrid. You’ve got to understand that all of the decisions that we’ve made around this product are made because its an EV. That is the first and foremost thing that it needs to be. So because it is an EV some of the decisions that we’ve made around engine operation will be different than what Toyota makes in its parallel hybrid. For them they are always operating in hybrid mode so they need to optimize everything for engine operation.

In our case we’re optimizing everything for EV operation and the secondary is certainly going to be better than conventional vehicles, but were not necessarily totally optimizing the system for charge sustaining mode because we don’t want to compromise electric vehicle mode.

So to be optimally efficient in charge-sustaining mode you might compromise EV performance?
In the electric vehicle mode, and its not just performance, its efficiency in electric vehicle mode that we’re optimizing.

You mean those first 40 miles?
Right, so you’ve got to remember our principle promise is this is an EV and our engine is there as a range extender and so even when the engine is on, we operate as through we are in EV. All the primary propulsion is satisfied by the electric motor. The engine is really there to supplement power to keep the battery sustained. Now there are a couple of tricks of the trade that we do since we have the engine on more, but for example we don’t want to do a whole lot of gearing that you would do in a parallel hybrid, because none of that is beneficial to you in the EV state.

But doesn’t the fact that you could keep the engine at fixed RPMs also allow you better efficiency?
Actually we don’t keep it at a fixed RPM, we have a window of operation that is optimized. We have been able to optimize the engine for a window of efficiency but it is still best to charge your power and torque levels within that window as the customer torque request varies. We don’t want to always be operating at one state because really you may be putting too much energy into the battery or drawing too much energy out of the battery. It is still good to vary that engine power and torque. Not to follow exactly what the accelerator pedal does, but to optimize efficiency.

We actually have a very sophisticated efficiency calculator in our model within our software. It calculates on a very very short time scale what the driving conditions of the car are. Which mode you are in, whether you turning you engine on or off, and what power and torque you want to run than engine always with the optimization of efficiency in mind as well as managing trade offs for driveability and noise.

We took all of the model that’s in the two mode hybrid and we’ve basically been carving out pieces we don’t need, adding in pieces we do need for this architecture and optimizing that model for this particular vehicle. We didn’t build this from scratch, this is also software that we are using that is also on the 2-mode but we are modifying it for optimizing this architecture.

I’ve driven the 2-mode and notice you can see the switched in mode of operation without feeling it in the car.
Which is the goal, you don’t want you to feel it in the car, we don’t want the customer to know these transitions are taking place, but we need to be able to enable them for efficiency.

With the Volt, once you’re in CS mode you will have a few different windows of operation or just one window for the generator?
We’re optimizing the generator to have different power generating levels. But the beautiful piece of being able to decouple the engine or generator from the axle torque requirement is we can travel along and hit those power levels that we need to optimize the system for battery charging and discharging, we can maneuver across them at any rate of change we so choose.

So think of it as the beauty of being able to decouple the engine is we have a degree of freedom that we don’t have to follow the pedal at all. We can pick and choose the points that are most efficient, we can go between those point on the best path and the most pleasing path to the customer. Actually this is a lot of the work we are doing even on a Prius hybrid every hybrid does it to some extent but every engine is required to follow the pedal. It is much more coupled to the axle torque request than in our vehicle.

It seems to me then you should make CS mode even more efficient then in a car where the engine always has to turn the axle?
Right and it is more efficient than a conventional vehicle because they do have to have that engine coupled. Again were optimizing some of those efficiency point puts we are really doing is focusing on the optimization of the EV. There are trade offs because we absolutely consider this product an EV by nature.