Consumer prototype first drive!

Working tirelessly since their X-Prize victory, Edison2 reaches a coveted milestone February 27, 2013. 

Photos on Flickr

A Look Back at 2011

2011 was a very good year for Edison2.

Ron Mathis lectured on the Very Light Car at NASA Langley and the NASA Goddard Space Center.  Brad delivered the keynote address at the worldwide launch of Siemens Solid Edge ST4. Oliver was as panelist in the Jefferson Innovation Summit, and spoke to the Society of Allied Weight Engineers and at the Automotive Weight Reduction Conference. The VLC visited the Detroit Auto Show, the DC Auto Show, the Louisville Auto Show and the Insurance Institute for Highway Safety, and was accepted into the permanent collection of the Henry Ford Museum. Edison2 was written about in the New York Times and featured on CNN International.

But it was in the shop, on the track and in the test lab that the Very Light Car really shone in 2011.

Edison2 won the XPrize through extreme platform efficiency and in 2011 we demonstrated how important a light-weight, low aerodynamic drag car can be. We fitted a Smart Car driveline into a VLC, and the 41 mpg (EPA highway) Smart engine tested at 89 mpg as a VLC. It is a fast, fun, very efficient machine. Ron Cerven, who led Li-ion to an X Prize Alternative class victory with an electric 2-seater, joined Edison2 in 2011, and helped create an electric Very Light Car. As expected, an eVLC combines acceptable range (114 miles), short recharge time (<7 hours) and outstanding efficiency (350 MPGe) with a small battery pack (10.5 kWh).

We also began safety testing of the Very Light Car. We ran numerous crash simulations using industry-standard software, and in November conducted our first actual crash test. These all confirm what we know from racing, where it is not uncommon for drivers to walk away from very high-speed crashes: that with the right architecture a very light car can be a safe car.

And we made a lot of progress in 2011 on the design of the next version Very Light Car. This prototype will be a car with bumpers, mirrors, production fit and finish, and more interior space. Wind-tunnel tests at Virginia Tech showed that the next-version shape is even more slippery than the X Prize car, so bumpers etc can be added without sacrificing efficiency.

2011 was a very good year… but 2012 promises to be even better. Stay tuned.



Squiggly Lines, Dots and the Art of Data Analysis

One of the cool things we get with official EPA test results for our electric car is instantaneous voltage and current draw readings. Since the eVLC runs on DC electricity, it’s simple to calculate power: it’s voltage multiplied by current.

Suppose at some point on the test the car draws 53 Amps at 107 Volts, the power is 53 x 107 = 5671 Watts or 5.671 kW. There’s a direct conversion from this to horsepower: 1 hp = 0.746 kW so in our example we’re using 5.671/0.746 = 7.6 horsepower.

The voltage and current readings taken by the lab allow us to plot the power consumption in different ways and study and learn from the results. Consider the graph below which plots power against speed for the FTP75 EPA City test

Speed in miles per hour is on the X-axis and Power in Watts is on the Y-axis. Inspection tells us lots of things. For example, the maximum speed of the test is about 56 mph, second, the maximum power required is just over 15 kW (about 20 horsepower), the majority of the test is run below 40 mph and, because the power is sometimes negative, we are using regenerative braking (regen).

On closer inspection it gets more interesting. The City cycle involves 23 stops and starts and each launch is reflected by its own loop in the power/speed trace. There is only one excursion above about 36 mph and the power required to accelerate the car is much greater than that required to maintain it in the cruise portion at about 50 to 55 mph.

Where it gets really interesting for an engineer is the convergence to a sloping line beneath zero power, meaning it’s to do with our regenerative braking. That this is a straight line indicates there is a torque limit to our regen: our relatively simple controller has a limit on how much it can cull from the motor and put back in the battery while the car is slowing.

Could we improve the regen and thereby make our numbers even better? That’s where the art comes in. Consider the graph below which is of exactly the same data as the first chart but with the measurements presented as points rather than a line.

Inspection shows us something interesting that’s not obvious in the smoothed line format – there are many more data points above zero power than below. Since the data points are taken at even time intervals, it becomes clear that even on the City cycle where you’re stopping and starting all the time, it’s how much energy you spend that dominates, not how much you recover through regenerative braking.

This has a number of profound implications for the design of the car, which have been a theme through everything Edison2 does - and you’ll see this if you read back through the various papers we’ve presented on this blog.

First, let’s consider whether we could do better if we had a better motor/regen controller, one that allowed us to recover the energy that might be available below and to the left of the present torque limit. The answer is, we could, but at what price? Right now we’re using inexpensive, off-the-shelf components and we’ve put up the best official EPA test numbers ever, by far. One of the core points we make at Edison2 is that we have superb performance with here-and-now technology. If we had a gold-plated, mil-spec controller we could do a bit better still but you’d have to be rich to afford it.

Second, the relative lack of points below zero power indicate that you can’t make great numbers, even with best in the world regen, if you’re spending energy like crazy to move a heavy square box along the road. The secret is to not spend it in the first place and that’s what the Very Light Car is all about

Third, regen has consequences. Most electric cars are front wheel drive because, since that’s where the weight goes when the car brakes, that’s where you can recover the most power. Front wheel drive has the disadvantage that the driving wheels must also steer so you have to move the power through articulating couplings in the driveshafts. This causes mechanical drag and it puts extra parts in the airflow, causing aerodynamic drag. At Edsion2 we realized the car spends much more time being driven than slowing so we deliberately chose rear wheel drive because we could package everything better for aerodynamics and it’s mechanically more efficient.

So, once again, testing by an impartial lab to a recognized standard has produced results that demonstrate the fundamental correctness of our methods and approach. We know we are on the right path and the results confirm it.



The Numbers Don't Lie

Our prototype electric Very Light Car (eVLC) was recently in an EPA-accredited lab for the full gamut of certified economy and range tests. Often this kind of thing is shrouded in secrecy but we thought it might be interesting to publish and discuss the actual results sheet (click here for a pdf).

Starting in the top left corner, we have Roush Industries’ logo. Jack Roush is well known as a successful NASCAR team owner and at the time of writing one of his drivers, Carl Edwards, is leading the Sprint Cup series standings. Less well known is that Roush has facilities all over the Detroit area providing engineering and testing services for the auto industry. We have enjoyed working with the professionals at their emissions lab in Livonia, MI.

The first block of data under the logo names the car tested, in this case VLC chassis #004, and, because it’s critical to the validity of the results, the tire pressure used. 44 psi is Continental’s recommended pressure for the DOT approved tires we use on the VLC.

The next block of information is the dyno settings. Because of its battery, the eVLC is heavier than our E85 powered cars, weighing 1140 pounds and, since the EPA standard is to add 200lb for the driver, Roush set the dyno for 1340 pounds. The car’s drag and rolling resistance characteristics are defined by 3 numbers determined by an SAE standard coastdown test. We’ve written extensively about this before and our numbers are A=6.31, B=0.1862 and C=0.00433. There’s a procedure to match the dyno to the coastdown ABC numbers and the compensating numbers are on the right. Once they’re set, the dyno provides exactly the same resistance to motion and acceleration as the real car on a straight and level road on a windless day.

Is “straight and level on a windless day” the same as “real world”? While you don’t have to deal with wind and corners in the lab, straight and level does prevent people measuring energy consumption coming down the mountain and ignoring what it would take to go back up. In any event it is the EPA standard and what everyone adheres to.

The CARB All Electric Urban Range test consists of running successive EPA Urban Dynamometer Driving Schedules (UDDS) with a 10-minute soak after each completed cycle. This continues until the car can no longer maintain the required speed trace or until the test is called off. The UDDS cycle, which can be found in 40 CFR, Part 86, Appendix I, is shown below. 

Edison2’s Ron Cerven drove the test. Although Roush didn’t record the car’s odometer reading, that is not significant because the dyno has a very accurate odometer built in, and it shows that we went 114.781 miles on a charge.

A word about driving the test. The speed profile from the graphic above is presented as a rolling strip chart and the driver has to keep the car’s speed on the target line. There’s a permitted error band but it’s tight and you have to concentrate hard to stay in the acceptable range. If you go outside the error band for more than a few seconds, in a test lasting several hours, you fail. Our being issued a results sheet shows Ron drove the test within the specs.

Charging the battery must begin within 1 hour of the range test finish and the electricity used is measured and recorded. There’s a brutally simple way of making sure this is an honest full charge: you have to do the Highway range test the following day on the electricity you put in.

Roush measured 11.0 kWh of electricity, and the EPA says 1 gallon of gas is 33.705 kWh, so the Equivalent Fuel Economy is 114.781 * (33.705/11.0) = 351.7 mpg.

Note the DC energy used figure of 8.5 kWh. This is measured at the car’s electric motor and the 22.7% reduction is due to losses in the charger, the battery and the controller. If you were to cherry pick and use 8.5 kWh in the calculations, the miles per energy gallon would be higher but the EPA very properly insists the number of record is what goes into the charger.

The same test procedure is used on the highway range test, using the EPA’s Highway Fuel Economy Test (HWFET) cycle. This is run repeatedly with a 10 minute soak between each pass until the vehicle can no longer maintain the drive schedule or the test is called off. Once again, battery recharging must start within an hour of the test’s completion. The electricity used to recharge is measured and recorded.  The eVLC went 113.331 miles on 11.0 kWh for 347.3 mpg equivalent fuel economy.

EPA procedure is to determine range by averaging Urban and Highway mileage, weighted 55% Urban, 45% Highway and rounded to the nearest 10. Our 114.1 mile combined range therefore got rounded down to the 110 mile official number for Calculated Driving Range.

Further down on the sheet, the last two rows on the sheet involve the EPA’s recently introduced 5-cycle standard. Time was, a new car’s “sticker” mileage was determined solely by the 55/45 combination of the UDDS Urban and HWFET Highway cycles. Since 2008 has been determined by calculated values according to EPA formulae which factor in cold weather (meaning the heater’s on), hot weather (meaning lots of a/c use) and aggressive driving.

At the bottom is the number that really matters: MPG with 30% Cap (Combined) of 244.8 is the eVLC’s “sticker” energy mileage according to the current EPA methodology. It directly compares with the Leaf’s official 99 mpg and the Volt’s 93 when running on its battery. To restate this in different words, Edison2’s eVLC scores 245% and 261% of Nissan’s and Chevy’s energy mileage.

The EPA also applies their 5 cycle formulae to determine the official value for range. In our case, they calculate 79.9 miles. The corresponding Leaf number is 72.5 miles on a battery almost 2 ½ times the size.

Lastly, although the X PRIZE wound up more than a year ago, we thought it would be interesting to see what we would have scored with an electric VLC so we ran the X PRIZE test which consists of 4 successive UDDS + HWFET drive cycles with a 10 minute gap between each one. The whole test was a touch over 71 miles and we used 7.0 kWh of electricity to do it. That electricity is equivalent to 0.208 gallons of gasoline so, the way the X Prize would measure it, we did 71/0.208 = 341.8 MPGe.

341.8 MPGe is an interesting number because we’ve demonstrated 104 MPGe in the same tests with our E85 ethanol/gasoline blend powered cars. When the cars are very similar and the tests are the same, why the factor of 3+ difference? The answer is the motor of an electric car is greatly more efficient than even the best combustion engine. Against that, the electricity measurement is plug-to-wheels and ignores the upstream inefficiencies (generation and transmission losses) of the electricity supply chain.

Regardless, at Edison2 we’re energy source agnostic. Our efficiency is in the car, and whatever you run the car on, you just don’t need as much of it.


The Shape of Things to Come

Audi Urban ConceptThe Frankfurt Auto Show debuted some concept cars of particular interest to us: the Audi Urban Concept, the VW NILS and the Opel RAKe.  These have more than a passing resemblance to our 2010 XPrize winning Very Light Car and our new electric VLC. Light weight and low aerodynamic drag are absolute virtues for auto efficiency, and we are pleased to have Audi, VW and Opel on this path with us. But while the similarities are readily apparent, the differences are important.

These cars are all very light. However, the Very Light Car carries 4 passengers (with 12 ft3 of luggage) while the Audi and Opel seat 2 and the VW only one. Also, the Very Light Car achieves its low weight through a new automotive architecture, not expensive exotic materials. It is predominately a steel and aluminum car. 


The Audi, VW, Opel and the VLC all have a streamlined fuselage with wheels external to the chassis. But the VLC is supremely aerodynamic, recording the lowest coefficient of drag ever seen for a 4-seat vehicle at the GM Aero Lab. It has less than half the drag (CdA) of a Toyota Prius. Our wheels are enclosed in pods because of aerodynamics, but by doing so, and with our compact suspension, we add crushable space for safety. VW asserts that its NILS concept will meet worldwide safety standards; so will the Edison2 Very Light Car. Computerized simulations at Roush show us already meeting some US standards, and we will begin actual crash testing in the upcoming months.

Edison2's approach results in unprecedented platform efficiency. The VLC requires only 5.3 hp to cruise at 60 mph. While any drivetrain benefits from this, it's particularly important for electric cars, allowing small battery packs, realistic range and charging times, and high MPGe.

Opel RAK

Although our 10 kWh eVLC has not yet undergone EPA-certified testing, our internal results (detailed previously on our blog) indicate greater range, shorter charging time and much better MPGe than the 24 kWh Nissan Leaf.

We believe the Very Light Car, the Audi Urban Concept, the Opel RAKe and the VW NILS Concept are the inevitable future of the automobile, in which efficiency is essential and weight and aerodynamics are taken seriously. This approach will not be limited to "urban vehicles", however, but will extend to all cars.

Here is a photo of our new eVLC prototype. The shape of things to come, perhaps?




Making Our Case at the White House

Aneesh Chopra, Assistant to the President and Chief Technology Officer at the White House, is a no-BS guy who wants to get big things done.  

Needless to say, we all got along just fine.

We were very fortunate to meet with Mr. Chopra and his team and planned accordingly.  After introductions and establishing the groundwork we already understood - that even in the Executive Office of the President there are no magic pots of money and the triad of vision, execution and accountability reign supreme - the Edison2 team dove right into what's at stake.

Edison2 is building the most radically efficient car on the planet, with dramatic potential for adoption in the domestic and global consumer market alongside broad private and public fleet applications. Yet for the growing body of evidence proving the current and future value of our work, a redesigned vehicle architecture poses an immense challenge to the legacy automotive industrial complex.

Thus, our "ask" was simple.  We need the technology office in the White House to leverage its mandate and reach to catalyze a broader and deeper understanding of the fundamental process of our work.  The targets are both potential R&D partners within various agencies of the Administration, and long-view industry players who can and will grasp that our work represents a virtually inevitable future (and therein, Edison2's significant time advantage).

Once this validation occurs - or "lands" in the parlance of policy - we predict that the deep level of interest we've received from potential partners (OEMs, suppliers, manufacturers), public agencies, and potential scale investors will translate to action.

Since winning the X-Prize, Edison2 has steadfastly remained "car first."  We have favored refining our technology and the market readiness of the Very Light Car over positioning the company for a nine-figure loan. But all along the way, we have quietly formed a worldwide nucleus of supporters.  Edison2 is reaching a critical crossroads - and the time and attention graciously provided by Mr. Chopra and his team might have provided the beginning of the tipping point we've planned for all along.  

There are 124 days left in 2011.  Plenty of time for big things to happen this year to set the foundation for Edison2 for years to come.