ChargeCar

Electric Garage Parking Regulations

CMU Police are now ticketing at the Electric Garage. To ensure everyone is aware of the parking rules, we are publishing the regulations. You can find them here: www.chargecar.org/parking

Cold Weather Driving

With an ambient outdoor temperature of 29 degrees, the Civic was taken for a test drive. Initial battery temperatures ranged from 3-5 C. After 4 minutes of driving, the temperature had risen to 4-8 C. After 8 minutes of driving, the temperature was 5-10 C. Finally, after 12 minutes of driving, the temperature was 8-11 C. Going up a steep slope (Shady Avenue in Squirrel Hill) with the accelerator all the way down caused the BMS to to have a low-voltage fault requiring a restart of the car. It seems that all of the voltages were below their normal ranges under load. Without load, the pack voltage was at its usual 110V, but with the voltage drop under load one could imagine a significant reduction in range. A preliminary benchtop test shows a 10-20% reduction in Ah capacity, but significant self-heating of the cells under load.

In summary, it seems prudent to design a heating system to keep the batteries at an acceptable range of temperatures. Additionally, it may be helpful to have a heater activated while the vehicle is charging to keep the cabin at comfortable temperatures as well! The effects of driving the car cold will likely become even more pronounced as the temperature outdoors drop further (20s are not far off!).

Heavy Car Experiment

We were curious to see if our conversion package would be compatible with heavier donor vehicles, so we conducted the following experiment:

We took the car out for some more load test driving on Friday; to Carnegie and back on I-376, 8.7 miles each way including severe motor loads on Greentree Hill at highway speeds. The Civic was loaded with 4 people, an extra 440 kg, bringing the gross vehicle weight to 1620 kg (3542 lbs) which is 46 kg over the original gross vehicle weight of 1564 kg (3440 lbs).

The car performed well, keeping up with 55-60 mph traffic. A high current of 550 A appeared once while accelerating up Greentree Hill but were normally approximately 300 A. There was no problem driving up the hill on Forbes Avenue towards Squirrel Hill. After driving, all the temperatures were acceptable, though the motor casing was as hot as 75 C (167 F); the maximum operating temperature is 120 C. The batteries were slightly warm as usual at about 95 F.

The increased weight did make the regenerative braking less effective at slowing the car and the steering was also much stiffer. The car was riding fairly low, with an extra 315 lbs on the front wheels which probably affected the steering geometry. The car was also rear-heavy (47/52) which is quite different from the original (internal combustion engine) weight distribution (60/40). This probably affects the feel of the car.

In conclusion, we believe the motor (and controller and battery pack) to be adequate for driving a substantially larger car, provided it is not too heavily loaded. Upgrading to a larger motor such as the AC-50 (very similar to our AC-31, it is slightly longer) would not require any major design changes. The resulting car would probably have a little more reserve for highway driving and a smaller temperature rise for a given drive.

Pittsburgh Public School Students

ChargeCar will be hosting a number of high school students for a few hours a week, starting October/November. They will be using the Electric Garage as a base for doing semester-long projects related to electric vehicles. We look forward to meeting our two students from Taylor Allderdice High School and our students from the Pittsburgh Science and Technology Academy (SciTech), from the Engineering Sciences and Environmental Sciences concentrations.

Hard Driving on the Highway

This morning the Civic was taken out for another round of hard driving, mostly highway driving up and down Greentree Hill. Four laps were made back and forth on I-376 between Route 51 and Carnegie, roughly 8 miles per lap. The car has no problem keeping up with traffic. Current went as high as 450 A when climbing the hill at 55 mph. Total mileage was 42.7 miles, using 10.62 kWh, 1.86 kWh regen (17.5%), giving us about 4.87 miles per kWh!

One reason for the hard driving test was to check the temperature increases on the motor, controller, and batteries so we can make good design decisions for future conversions. The starting temperature for the components was approximately 60-65 F (16-18 C). The outside temperature at 11 AM was about 55 F (13 C). Temperature data was recorded at the end of the drive from sensors we have installed ourselves (reported to the "display"), from the controller/BMS (reported to the Curtis programmer/"monitor"), and from a non-contact pyrometer:

	Display	Monitor	Pyrometer
Motor Temp.	93 F (34 C)	127 F (53 C)	129 F (54 C)
Controller Temp.	86 F (30 C)
Max. Battery Temp.	111 F (44 C)		105 F (41)

*The discrepancy in motor temperature measurements is explained by the different sensor locations; the "display" temperature is measured on the exposed end of the motor, "monitor" temperature is read from a sensor located, and the pyrometer reading is based off of the motor casing. The "display" temperature sensor should probably be relocated to a hotter location in the future.

Battery temperature climbed higher than it ever has before, not surprising given the severity of the test. The battery pack is fully enclosed in a plywood box to contain as much of the heat as possible. The total increase in temperature was about 47 F (26 C), approximately 54 F (30 C) above ambient temperature. This heating corresponds to approximately 4% of the net energy consumption. Even if driven on a hot day (say, 100 F or 38 C), the battery temperature (perhaps 64 C) should be well below the 75 C operating limit recommended by Thunder Sky. This is more strong evidence that the our battery pack will not require cooling or ventilation.

We will be analyzing the data to see if there is a significant gradient in cell temperature between the middle of the pack and the outside. We will also be using the GPS data to extract topography information from the commute, specifically the grade.

The great takeaway message is that the Civic is suitable for highway driving, even with steep hills and high ambient temperatures!

[UPDATE]

Recharging consumed a total of 11.27 kWh. 8.76/11.27 = 78% efficiency. In units probably more useful to potential conversion drivers, this means that on hard drives, our car managed 42.7 miles using 11.27 kWh of energy (potentially from your wall outlet), giving us about 3.79 miles per kWh of charging. If we assume that cost per kWh is about $0.10 (Duquesne Light claims $0.0817/kWh for most residential customers; www.duquesnelight.com/ElectricChoice/calculate.html), then you can see that once your car is converted, $1 of electricity gets you about 38 miles of range. According to www.pittsburghgasprices.com, cheap gas in Pittsburgh costs $2.61/gallon; in a 25 mpg ICE car, you are spending at least $3.96 to go the same distance. Interesting.

Conversions Section Added

The ChargeCar website is in the process of being updated/redesigned to improve user-friendliness and aesthetics as well as to include more about our cars. Please check out our new "Conversions" tab on the main page to see pictures and a healthy dose of information about both of our vehicles! Please also check out the Scion and Civic photo galleries.

Endurance Test

Today, between 8:30 and 10:40 AM the car went through it's first prolonged endurance test.

From the Electric Garage, the car was driven (with two adult passengers) to 2nd Avenue and two laps were made between downtown and Bates. Very little voltage decrease was observed and it became clear that more rigorous driving would be required to run the batteries down in a time-effective manner (we have a weekly technical meeting at 11AM as well). So, the car was taken on the highway, at speeds of approximately 55 mph and driven between the Carnegie Science Center and Squirrel Hill. Twice. To finish off the battery, a few laps of the Panther Hollow/Schenley Golf Course were made, starting from Squirrel Hill. Finally, the minimum voltage sagged enough to cause the BMS to fault. This was after about 48.3 miles and 11 kWh of driving! The return to CMU was fortunately mostly downhill, coasting on regen and only causing two more faults from fairly conservative accelerations. The Civic is recharging in the Collaborative Innovation Center (CIC) Garage, having traveled a grand total of 49.5 miles on 11.26 kWh.

4.40 miles per kWh, with perhaps only 15% regen (due to highway driving). Also, the Civic drove nearly 50 miles, which is outstanding range! This should be sufficient to travel to Pittsburgh International Airport and back (to CMU campus)! It also appears that voltage vs. load can be used to estimate the state of charge, which will be a very useful tool for driving, replacing the fuel gauge. The voltage-load curve definitely shows gradual decrease with regular driving and a very much accelerated rate near full depletion. We suspect that we were very close to full depletion because of the additional 2 BMS faults on small hills driving back to campus.

This is very exciting news, we will have more endurance tests in the future, perhaps to the airport next!

Range and Efficiency

The Civic was taken out for a long drive Tuesday afternoon. From the garage, to downtown via Bigelow, back out to Squirrel Hill, Pocusset to Serpentine Drive, then approximately 10 miles through Schenley Park, and back to the garage via downtown and Bigelow. This was a reasonable combination of highway and urban driving. The total trip length was about 30.2 miles. The netbook in the passenger's lap reported a net charge usage of 6.62 kWh (with 19% regen from total), giving 4.56 miles per kWh! [Thought: Ah may be a better indicator of battery usage than kWh] The pack voltage did not sag a large amount and the test was terminated because of time.

The car managed 70 mph on a level stretch of the Parkway, drawing ~280 A in 5th gear. This is only half of the available peak current, so the speed capability should be substantially higher (on level ground). The Civic also did 55 mph uphill on Bigelow coming from town without pushing too hard.

The total energy to recharge the Civic was 7.96 kWh; charging efficiency is around 83%.

Very encouraging results. We think that we can go even further than 40 miles - a 45 mile practical range would be very nice indeed. More range tests will be happening soon, we'll keep you posted!

Performance and Parameters

Recently, the Civic was modified by setting up the tachometer (motor rpm) and placing a force-sensing resistor on the brake pedal. We have mapped the regenerative braking to this sensor and taken it for drives. The result is a very sensitive first regime of braking (regenerative) that can then be easily continued into the normal mechanical brakes. Some more tinkering will be necessary to get the sensitivity right, increasing the "travel" to be about twice what it is now seems like the next step. It is much more intuitive than the temporary dial potentiometer mounted on the steering column.

The Civic was also taken for a total of 30 miles between charging cycles, split between approximately 5 miles of testing and 25 miles from 12 laps of the nearby Schenley golf course. Good news is, the pack voltage saw no drops in voltage until lap 10, and the drop was small. No BMS faults either! The 25 miles was estimated to deplete the pack approximately from a 90% state of charge to 30% in normal driving conditions with currents up around 300 A.

There have been some interesting problems with the motor torque at low speeds. We have heard that many other conversions are able to spin their wheels at the start. Previously, we had very gentle acceleration even when the accelerator was pressed to the floor.

However, our controller and motor came as a package deal, nominally an AC-31. It turns out that our controller was set to operate an AC-35 motor and the motor was labeled "AC-3501". However, it seems that the nameplate may be mistaken. By changing our controller parameters to those suitable for the AC-31 (motor setting 164), we have been able to achieve much better performance, even starting respectably in second gear. We are anxious to see the effects on our mileage and efficiency!

[UPDATE]

The new parameters also strongly affected the performance of regen. The FSR on the brake pedal seems like it could be potentially dangerous. The braking is far too strong in first gear, second and third are debateable. A major problem occurs when changing gears (technically, when you press the clutch), the regen cuts out and suddenly you have to press much deeper on the brakes (to activate the mechanical brakes). We are still waiting for a spring-loaded thumb-switch potentiometer to arrive, to be installed on the steering wheels near the right hand.

The car will be driven home this weekend by a ChargeCar member, we look forward to calculating efficiencies with the new controller parameters!

Weekend Driving and Weigh-In

Recently, one of our team members has been using the Civic as his primary commuter car and he has been recording data.

To home: 5.8 miles
2.407 kw-hr used
0.216 kw-hr regen
2.191 kw-hr total

To work: 5.8 miles
2.037 kw-hr used
0.316 kw-hr regen
1.721 kw-hr total

Round trip: 11.6 miles
4.444 kw-hr used
0.532 kw-hr regen
3.912 kw-hr total

==> 2.97 miles/kw-hr
==> 12% of energy used was given back to the car during regen

Perhaps certain hills of Pittsburgh are to blame?

Yesterday, thanks to people at CMU's Field Robotics Center, we were able to weigh our cars:

The Civic weighs 2588 lbs, 167 lbs more than its gas-fueled equivalent's curb weight (if one believes Wikipedia) and has a left-to-right weight distribution of 50.3%, and front-to-back weight distribution of 52.1%.

The Scion weighs 2615 lbs, 220 lbs more than the original curb weight (again, Wikipedia) and has a left-to-right weight distribution of 50.8% and a front-to-back distributiion of 54.7%.

Driving Evaluation Data

The bad LiFePO4 cell was replaced with a cell we charged last night and topped-off this morning. The BMS incorporated it perfectly and the cells are all within +/-0.02 V of one another. Another issue we encountered yesterday were potholes. The small sharp shocks seem to trip the contactor (essentially a large relay or electromagnetic switch) and the throttle could cut out for brief periods. We resolved this by raising the holding voltage on the contactor.

The Civic was just driven around the same 7.2-mile course (three laps of a circuit in Schenley Park) in the exact same manner as the Scion. Good news! There were no faults of either the contactor or battery pack. The batteries have still not shown any significant increase in temperature. We did experience some heating on the motor and controller - still far from the maximum specified temperatures. Perhaps this could be a solution to our windshield-defrosting and cabin-heating issue....

At any rate, the data we have collected from this drive should be an excellent first pass at comparing the efficiencies between the Civic and the Scion. After the drive we recharged the battery pack with 1.80 kWh. This is fantastic news! That means that our 7.2 miles was driven at a rate of 4 miles per kilowatt-hour, pretty great as far as converted electric vehicles go! Granted, this was driving conservatively (keeping the current out of the 100 V pack under 100 A) and with regenerative braking. More updates soon.

Honda Civic Driving!

The Honda Civic is now a true conversion. It was completed last week (July 23rd) and it has been taken out for drives in our parking lot and on local roads. Some drives incorporated serious hills (i.e. Forbes Avenue towards Squirrel Hill, if you are aware of Pittsburgh's roads) with varying numbers of passengers (up to four adult males).

Here is a quick YouTube video of the first time our Civic was ever driven (without gas!):

www.youtube.com/watch

Just this morning, the Civic was taken on the parkway and driven up to 60 mph. There was still acceleration to be had at those speeds too!

We will be testing the range on practical commutes soon. So far, everything looks good. In spite of outdoor temperatures being reasonably warm (70's), we do not appear to have problems with the motor or LiFePO4 cells overheating. We do, however, have one cell reporting very low voltages, when the rest are all leveled perfectly by our BMS. We look forward to figuring out what the actual practical range of this vehicle is - stay tuned!

Honda Civic EV Update

A week ago, we received 36 TS-100 (100Ahr) Thundersky Lithium Iron Phosphate batteries, 33 to go into the Civic (with 3 spares). While we've been waiting for them, we've also been prepping the controller and motor to go in, including modifying the adapter plate to connect the AC motor to the transmission and clutch. With all the key pieces coming together, we've managed to mount the first components into the car and even made the wheels spin up! Now we're fine-tuning the hundreds of parameters on the controller, prepping the battery pack to actually be installed inside the vehicle, and getting the battery management system set up so that we properly instrument the pack. Once all that's done, we'll be ready to take the car out for its first trip as an EV!

After we've put the car through its paces, we'll pull it back into the garage to do some more work on it. We're still working to figure out the final battery configuration, what the user controls will look like, and what vehicle sensors we need in order to use the Civic as a scientific platform.

This is an exciting time for the garage! Come stop by to check out everything we've got going on!

ChargeCarPrize Contest Launched!

The ChargeCarPrize programming contest (beta phase) has launched! The software & data are available to download, and the leaderboard is live for submissions. We have real-world data collected by our secret set of drivers available for training and testing your algorithms.

For those who haven’t heard already, the ChargeCarPrize programming contest is a collaborative competition to design the best way to use the compound energy storage system of ChargeCar. Given GPS coordinates, previous drives, speed, acceleration, and other characteristics, contestants must develop an efficient strategy for using the supercapacitor. Should regenerative breaking go into the cap or battery? Should I trickle-charge the cap from the battery at red lights to handle the acceleration afterwards? Should I do full route-prediction to determine the future power demands of the vehicle? Let’s find out!

The contest will award its first monthly prize to the leading submission, a $100 gift certificate to Amazon.com, on September 1^st, 2010. Prizes will continue monthly for the duration of the contest, when a yet-unannounced Grand Prize will be awarded to the best overall submission! Check out chargecar.org/invent and get started today!

Battery Tech Talk

On June 23rd, we had Carnegie Mellon professor Jay Whitacre come to our weekly tech meeting to give us a quick overview of battery chemistry, performance characteristics, and tuning for high performance. We were able to record the video, and it's available at the Carnegie Mellon Robotics Institute website. From that site:

June 2010 - Length: 1:15:00

Jay Whitacre, a Carnegie Mellon professor with a joint appointment in Materials Science and Engineering and Public Policy, gives a quick overview of Lithium-ion batteries. Starting at the chemical level, he explains the properties and mechanics of the battery which give rise to macroscopic behavior, especially focusing on issues relevant to electric vehicles and other high-power systems. He then fields questions about the batteries themselves, ongoing research at Carnegie Mellon, and tricks for improving long-term performance, such as pulse-charging and pairing batteries with similar defects.

http://www.ri.cmu.edu/video_view.html?video_id=60&menu_id=387

Hope you enjoy watching it!