Performance Testing at the 1992 American Tour de Sol Dr. Robert Wills, P.E., Co-Director American Tour de Sol The second most frequent question asked to an electric vehicle owner is "How far can it go?" Electric vehicle range is determined by two things - the on-board energy storage available in the vehicle batteries, and the amount of energy needed to travel a unit distance (i.e. the energy consumption, or vehicle efficiency, in Watt-hours per mile). One of the major aims of the 1992 American Tour de Sol, at the request of the U.S. Department of Energy, was to measure both vehicle range and efficiency under controlled conditions. Efficiency Testing In 1991, we devised a means of measuring electric vehicle (EV) energy consumption that was simple, quick to install, portable, and easily adapted to any vehicle electrical configuration. This involved measuring average battery voltage and average motor current (at the motor controller/battery interface) using Fluke Model 87 Digital Multimeters. These innovative instruments are capable of sampling their input every 100 milliseconds and calculating a true average value for periods of up to 36 hours. We use two Fluke 87 meters in each test setup. One meter measures average battery voltage directly while the other measures battery current via a 0-1000 Amp clamp-on DC current probe (a Fluke 80i-1010). To facilitate installation, 15 foot extension test leads were made using flexible two-core rubber cable. The voltage leads were terminated in heavy duty, fully insulated alligator clips and fused in-line near the clips to protect against a cable short - for example from being compressed under the vehicle hood. Fluke fully insulated banana plugs were used for safety. The current probe leads were a simple extension adapter for the Fluke 80i-1010 leads. The power measurement is obtained by multiplying average battery voltage by average battery current. This is an approximation - the true energy consumed would be the average of the instantaneous power (P= V x I) - however as the battery voltage remains relatively constant during the tests, little error is induced. (This may not be the case for a testing cycle that has large amounts of acceleration and regenerative braking). A measure of this error can be obtained from the minimum and maximum data available from the Fluke meter. For example, the Solectria Force GT showed the following data for its 35 MPH laps at the Thompson Speedway: INSERT WillsForce.eps The voltage variation of 6% probably only occurred briefly during the constant speed laps, and so is unlikely to result in more than 3% error. When instrument and timing error is included, it is still likely that overall error is less than 5%. Another possible source of error is that some high frequency current components from the motor controller may not be sensed by the current probe, but again this should be minor. It is important to clamp the current probe on the battery side of the motor controller, as the actual motor current may be significantly higher than the battery current, especially at low speeds (the motor controller acts like a DC transformer). Similarly, the voltage connection must be made on the battery side of any contactors, otherwise zero, rather than the true battery voltage will be averaged during coasting or braking periods. The vehicle efficiency is calculated from: Av. Power (Watts) = Av. Volts x Av. Current (Amps) Energy Used (Watt-hours) = Av. Power x Time (Hours) Efficiency (Watt-hours/mile) = Energy Used/distance (miles) The average power consumption of the vehicle at a certain speed is related to the efficiency by: Average Power (Watts) = Efficiency (Watt-hours/mile) x Speed(mph) For example, using the data above for the Force, the average power is 5.175 kW, the Energy used (over 1.70 minutes) is 153.5 Watt-hours, and the efficiency (over the 1.04 mile course) is 141 Watt-hours/mile. Multiplying 141 Wh/mile by the actual speed of 36.7 mph gives the average power, again of 5175 Watts. The Efficiency Testing Courses Efficiency tests were performed at the initial scrutineering in Albany prior to the race, at the Thompson Speedway (Thompson, CT) on day 4, and in the Boston city traffic, at the end of the race. The Albany course was a 3.15 mile loop from the Rockefeller Plaza, across the Hudson River to Rensselaer, and back up to the Plaza. It ran on highway quality roads, had a height change of 160 feet down to 25 feet above sea level, and back again, and had several sharp turns and several possible stops. Entrants were required to travel this course at a minimum speed of 25 mph, with some allowance being made for necessary stops. Due to time constraints, not all cars could be tested on the Albany course. The Thompson course was the 5/8 mile banked oval track of the Thompson Speedway in Thompson, Connecticut. This track had an acceptable but not excellent road surface, and was surprisingly difficult to drive at speeds above 35 mph - the two end turns are steeply banked and quite tight. The measured distance around the inside of the track was 0.52 miles. Efficiency testing at Thompson was mandatory for all commuter class vehicles, and optional for the other categories. A second agenda for the Thompson testing day was to give all vehicles the opportunity to demonstrate their full range capability. After a 25 mile run to the speedway, each category was given a two-hour period to run to full range. The minimum allowable lap speed was 35 mph for the commuter categories. The Boston course was a 2.1 mile loop through Boston City streets to test vehicle performance under actual (worst case) stop-start driving conditions. Testing in Boston was optional for all categories. Results The Tour de Sol Commuter category includes practical commuting vehicles with a limit on battery capacity of 7,200 Wh for two-seat , 9,600 Wh for three seat and 12,000 Wh for four seat cars. PV array size (in peak Watts) is limited to 1/10th of the battery capacity. The results for the Tour de Sol Commuters were much as expected. Most vehicles used between 150 and 250 Wh/mile for speeds ranging from 25 to 45 mph. An interesting comparison can be made between Solartech II and S-CAR-GO which are similar vehicles with different drive trains. Solartech is a little lighter (2000 lbs vs. 2800) and has a Solectria 16 HP brushless motor, while the S-CAR-GO has a 17 HP Advanced DC series brush motor. They showed similar performance on the Thompson track, but S-CAR-GO's lack of regenerative braking showed up as significantly worse results in the Albany test which included a long decent followed by a climb back to the start. The concept cars in this category - Solectria's Flash, New Hampshire Technical Institute's Sungo and Unatego high school's Solar Bullet demonstrated what can be done with ground up design. All showed energy consumption of only 50-70 Wh/mile at 25 mph. The Sungo, in particular, speaks to a new, practical class of small commuter vehicle that could well become a commercial product. With its solid aluminum chassis, four-wheel design, and dual 11 HP Solectria brushless motors, (one for each rear wheel), the Sungo could satisfy most people's commuting needs while consuming less than 100 Wh/mile. Tour de Sol Commuter Results The American Commuter category is also aimed at practical commuting vehicles, but has no limit on battery capacity, apart from performance tradeoffs and the manufacturer's Gross Vehicle Weight. Vehicles in this category tended to be heavier (2500 to 3500 lbs), being conversions of conventional IC engined vehicles, and had accordingly higher energy consumption results. Notable cars were the very low energy use shown by the Solectria Force GT, the E96 Saab conversion and the two St. Johnsbury academy cars, the Jewel and the Hilltopper. Solar Car Corporation's Chevrolet S-10 pickup conversion, despite being the heaviest vehicle in the race, showed excellent performance with 177-270 Wh/mile (and a 90 mile range test run). The Sorrel Chevy S-10 conversion was not entered in the race but came for the Thompson display and was tested there. It is a similar vehicle to the Solar Car Corp. conversion, (they in fact used a Solar Car Corp. conversion kit) and the efficiency results showed good agreement between these cars. An interesting comparison here is between Solectria's Force GT, and the Force (which ran in the TDS Commuter category). The heavier (2350 vs. 2142 lbs) and more powerful (22 vs. 16 HP) Force GT had significantly lower efficiency results than the Force at higher speeds. This may simply be a result of driving style - slowing down for the tight corners of the Thompson Speedway - but may also be an indication of increased frictional losses, for instance from a wheel alignment problem. An interesting point is that these extremely efficient electric vehicles will require regular maintenance, such as checking tire pressure and alignment, for their performance to be sustained. American Commuter Results The Tour de Sol Racing Category is for solar powered vehicles that are limited to 480 Wp of solar charging and 4800Wh of battery storage. Most vehicles performed in the 50 to 60 WH/mile range, both in the Thompson and the Albany tests. Trenton State's Sunlion had obvious problems in testing at Albany that turned out to be a dragging parking brake. (Again, vehicles that rely on efficient performance for range must be well maintained). Two exceptional vehicles were the Conval Sol Survivor II and Dartmouth's Sunvox IV. The Sol Survivor II tested at such a low value in Albany that we initially thought it to be instrument error. The Sol Survivor team has spent three years refining their car, and in addition, has carefully measured and modeled vehicle efficiency themselves. This level of knowledge, together with an excellent car, enabled Conval to win the Tour de Sol Racing category, defeating many experienced college level teams. Conval also received the prize for most efficient Racing Category car. The prize was based on data collected in Albany. Tour de Sol Racing Results The Cross Continental Category is for solar powered vehicles that conform to the World Solar Challenge (Australia) and GM Sunrayce (USA) rules of unlimited battery capacity and PV array peak power, but a PV maximum area of 2 x 4 meters. Only one measurement of efficiency was made on these cars, as all use Uniq mobility controllers and motors which appear to have extremely high peak battery currents, or emit electromagnetic interference that upsets the operation of the Fluke current probe. For other vehicles, the Fluke 80i-1010 current probe converts measured current to 1 mV per Amp. The Fluke model 87 meter is set to average on its mV scale, with a maximum reading of 999 mV (or 1000 Amps peak). The Uniq Mobility drive systems evidently draw currents in excess of this 1000 Amps as normal operation causes a meter overrange and average data is lost. Switching the Fluke 87 to the Volt range, rather than the mV range, solved this problem, but this was only discovered at the end of the Albany testing and no Cross-Continental cars opted to be tested at Thompson. Cross-Continental Category Results The Open Category is for vehicles that do not meet the requirements of other categories, but that promote the aims of the event. This category contains a wide variety of vehicles, from bicycles, to motor-cycles, to vehicles that for one reason or another decided not to run in a standard category. An example of the latter is the MIT Aztec which had suspension problems and decided to run in Open with only a single occupant, rather than in the Tour de Sol Commuter category with two people aboard. Their testing results show them to be potentially the most efficient of the ground-up commuters (perhaps because of their innovative aerodynamic design and MIT race bred components) using only 60 Wh/mile vs Flash's 72 Wh/mile at 36 mph. Electric motorcycles are a mode of transportation that have not been paid much attention at this time, but will have to be addressed in a future gasoline-free society. This year, we had two motorcycle entries, the Envirocycle and the EM/120 which was not finished in time to run the race road course, but was tested in Boston. The Envirocycle, a converted Honda, could well be expected to perform a little better than 100 Wh/mile with its advanced DC 4 HP motor and aerodynamic fairing. More attention should perhaps be given to rolling resistance and drive losses. The EM/120 is a very exciting machine, beautifully faired and probably capable of more than 150 miles with its 8100 Wh of batteries. Boston Testing Results The final efficiency tests in Boston were performed in stop and go traffic. The tests were totally voluntary, and only a few teams took advantage of them after the five long days on the road. Envirocycle again showed a consistent 100 Wh/mile performance, while Kineticar showed just how much energy can be used in Boston traffic. In this stop and go driving environment, the lack of regenerative braking and the disadvantage of a heavy car (3281 lbs) are very apparent. The Audi conversion was brought for display on the last day of the race by a member of the Electric Auto Association. It used an Advanced DC motor and Curtis controller similar to the Solar Car Corp. cars, and showed very reasonable performance. Electric Vehicle Efficiency The following graph shows the full set of testing results obtained at Thompson. This graphical presentation brings out several interesting points: 1/ The cars are grouped in three area - Large vehicles with poor aerodynamics such as the VW Van Conversion, Kineticar and Poetry in Motion - Medium size conversions such as the Force and Solar Car Corp. cars - Ground up electric vehicles such as the TDS racers, the Sungo and the Flash. 2/ In most cases, and especially in the middle group of cars, aerodynamic drag effects (a Wh/mile increase proportional to velocity squared) are not seen to be significant. This may not be so for the Force GT and the Flash, which are highly optimized vehicles with minimal friction, but for the remainder, performance improvements are more likely from lower rolling resistance and drive train improvements than from better aerodynamics. 3/ The two tests on the Solar Tech car (a small converted BMW Izetta) at roughly the same speed , but with different driving style are interesting. The first test showed 102 Wh/mile while the second, with the speed pedal held at maximum (i.e. constant speed) resulted in only 70 Wh/mile. Driving style can have a large effect on performance. 4/ Similar results are evident for the Solectria Force which showed very low energy consumption at low speeds but jumped up dramatically at higher velocity. This is most likely a result of slowing for, and accelerating out of the corners. 5/ Some cars, notably Poetry in Motion and Kineticar showed lower energy consumption at higher speeds. This is an indication of incorrect gearing, or perhaps again an improved driving style as the driver gets used to the course. Range Testing The day four range information is presented in the tables above. Most notable were Solectria's Force GT , which was using nickel cadmium batteries, which drove 100.2 miles to receive the prize for greatest range. Two cars that used lead acid batteries also performed very will. The St. Johnsbury Academy's Electric 'Hilltopper drove 96 miles, and Solar Car Corp's Electro Chevy drove 90 miles on a single battery charge at a minimum speed of 35 mph. Other Tests The scrutineering process at the American Tour de Sol includes many other safety related tests including rules compliance, construction and safety, a cone driving test, starting on a incline, acceleration, and a brake test. The brake test was of interest as most cars end up carrying a rather full load of batteries, though all but one of the entrants managed to meet the race rule requirements. A standard automotive decelerometer was used for the testing. It proved to be an excellent indicator of brake performance. One interesting point that came out from the design of the Sungo was that dual motor drive systems require careful balancing both for forward acceleration, and especially for regenerative braking. There is a need for the motor controllers to communicate and match the wheel torque under these conditions. Conclusions Efficiency testing at the 1992 American Tour de Sol produced consistent and reasonable results that may be used as benchmarks for the design and testing of electric vehicles and EV programs. Because of the limited number of tests and variations in driving style, this data should not be used in deciding whether one car is "better" than another, but can be used as a guide as to what is possible, both today and in the future. There were two significant changes from the previous year: - A new class of highly efficient electric commuter vehicles, with performance in the 50 - 100 Wh/mile range is evolving - for example Solectria's Flash and NHTI's Sungo. - Performance in the racing categories is clearly improving with cars testing in the 28 to 42 Wh/mile range (Dartmouth's Sunvox IV and Conval High School's Sol Survivor II) compared to typical values of 47-60 Wh/mile last year. During the 1993 American Tour de Sol, we plan to refine our testing procedures further with a full day of testing at a race track, some form of start-stop urban driving cycle, an obstacle course and other tests of practicality and utility, and more refined efficiency measurements including electronics to calculate instantaneous power from our voltage and current probe inputs to increase the measurement accuracy. Technical Testing was made possible by a grant from the U. S. Department of Energy through the Argonne National Laboratory. Many thanks to U. S. DOE and Argonne National Laboratory staff, American Tour de Sol participants and the many volunteers that assisted NESEA in performing the tests and collecting the data. Access Author: Dr. Robert Wills, P.E., Skyline Engineering, POB 134, Temple, NH 03084 ù 603-878-1600 ù FAX 603-878-4643