Patrick Carpentier had a tough time during morning practice   for the Miller Lite 250 in Milwaukee.

Patrick Carpentier had a tough time during morning practice for the Miller Lite 250 in Milwaukee.


The Blue Box impact sensor is installed in every Champ   Car. The information gained helps Ford engineers understand the affects   of high <I />g</i>-loads on the human body. Ultimately, the goal is to not   only help engineers design safer Champ Cars, but production vehicles as   well.

The Blue Box impact sensor is installed in every Champ Car. The information gained helps Ford engineers understand the affects of high g-loads on the human body. Ultimately, the goal is to not only help engineers design safer Champ Cars, but production vehicles as well.


Computer animation shows a Ford Champ car driver in   the cockpit.

Computer animation shows a Ford Champ car driver in the cockpit.


Based on crash data from the Blue Box, CART added energy-absorbing   headrests (see highlighted section) in the cockpits.

Based on crash data from the Blue Box, CART added energy-absorbing headrests (see highlighted section) in the cockpits.


When racing legend Dale Earnhardt slammed into a wall going upwards of 160 mph at last year's Daytona 500, most expected him to walk away. Bruised and battered, certainly, but walk away nonetheless. It happens all the time. Sadly, not in this case: Earnhardt suffered a fatal skull fracture upon impact.

Ironically, the accident "didn't look that bad," according to witnesses. But, as the old saying goes, looks can be deceiving. The fact is, drivers in a crash can experience anywhere from 15 to 100 gs on impact. And head injuries are almost always to blame for disability or death in motorsports, according to Dr. Steve Olvey, medical director for Championship Auto Racing Team (CART).

Fortunately, technology can now help explain what happens to a driver's head and neck in a crash. The "Blue Box," a crashsensor system in cars from the CART Fed Ex Champ Car Series, is letting CART and Ford Racing Advanced Technology engineers predict the responses and injury potential to drivers in high-g environments. The resulting data may go a long way toward improving vehicle safety both on the track and on the highway.

NEW BOX, MORE DATA
As CART's Official Safety Technologies Provider, Ford Racing has been supplying Blue Boxes to the CART field since 1997. This year CART race cars began using a new box built by Delphi Automotive Systems called the ADR 2 (Accident Data Recorder). The crash-hardened box, made from anodized aluminum, weighs about 2 lb and is small enough to hold in your hand. It sits on the floor of the cockpit, below the driver's knees, and draws power from the car's electrical system. If the vehicle loses power, the unit's internal uninterruptible power supply kicks in.

Using a series of accelerometers, the ADR 2 senses and records key vehicle parameters at 1k-samples/sec before, during, and after an accident. Analog information from the sensors gets converted to digital information by a microprocessor and stored in RAM. The memory logs about 2 Mbytes of data in a FIFO fashion, explains Delphi. Engineers retrieve the recorded data via a high-speed data link to a PC.

One of the biggest distinguishing factors between the new system and CART's earlier-generation models is its ability to record more than just crash-impact data from its internal sensors. Certain parameters from the car's telemetry system, such as speed, throttle, steering angle, and lap number, also get recorded.

"The old box basically had three accelerometers and only provided information upon impact," explains Jason Douglas, Ford safety research and development engineer. "In an accident we had only a 2-second window to capture information about the crash. Now, we're able to see a 3-minute window of data. The box continues recording for about 1 minute after the crash so in the event the car gets reimpacted, we can get that information as well."

Another big difference between previous models and the ADR 2 is the triggering factor: Early-generation boxes only began recording at impacts of 20 gs. The ADR 2 has two three-directional accelerometers and data logging begins in two scenarios. In the first, the low-g accelerometer reads 2 gs either side-to-side, or front-to-back, and the high-g accelerometer reads 15 gs front-toback, explains Douglas. (The high-g accel senses up to 500 gs while the low-g accel goes to 50 gs.) The second criterion is speed: When cars reach 35 km/h, the box turns on.

According to Douglas, this type of back-up system works well. "We are able to get speed and other parameters from each car's serial port data," he says. "But, if something isn't working, the box can read the acceleration and initiate data logging." The unit is also equipped with an angular rate sensor which measures yaw at 200°/sec.

Ford safety researchers get even more data from Delphi's earpiece sensor system which connects directly to the ADR 2. This system works in essentially the same way as the box: Small sensors are built into the left and right sides of radio ear-pieces worn by drivers. According to Delphi, six accelerometers, one for each of the three axes on each side, measure head acceleration during an accident. Proprietary software especially made for the ADR lets engineers watch the sensor data in real time on a laptop.

"The earpiece sensor system lets us track g forces on a driver's head as well as from where the box sits on the floor," adds Douglas.

The ADR 2 also can record data from external speed or rpm sensors, a lap indicator, earpiece accelerometers, seat-belt load sensors, or other sensors.

Engineers augment sensor data with video footage of the crash, photographs, medical reports, and debris. The total package paints a clear picture of what happens during impact, and gives enough information to reconstruct the accident.

FROM TRACK TO SCREEN
After a crash, CART safety crews pull the car off the track and photograph the damaged chassis components. At the same time, Ford engineers download data from the ADR 2 to a laptop, getting a plot of operating parameters during the accident. According to Douglas, the most important data comes from the accelerometers. "Immediately following a crash, we download the g forces in X, Y, and Z directions to get the typical pulse of what happened," he explains. In CART cars, the data port usually sits inside the cockpit tub by the driver's knee or near the rollover hoop behind the head.

Ford then simulates the crash using CAE modeling software. Basically, engineers see a computer model of the driver in the cockpit and the forces he or she endured on impact. According to Douglas, g force alone is not a telling assessment of an impact. Duration of the load is just as critical: High gs in a short spurt may lead to injuries like those from lower gs over a longer time period.

Engineers use all this data to make predictions about driver injuries and validate them from medical reports. "Doctors are also becoming more and more interested in the data," says Douglas. "It gives them an idea of the injuries they should look for."

Ford keeps every downloaded crash on file for reference. Results so far have led CART to add an energy-absorbing headrest in the cockpit. Drivers sit in a cocoon of 1-in.-thick carbon fiber. Explains Douglas, "The headrest mates to the tub and essentially controls the movement of the driver's head during an accident, absorbing the energy on impact. If it wasn't there, the head would make direct contact with the rigid carbon-fiber tub."

The data also gets shared with Ford's production community.

The hope is to one day use it to improve production-vehicle designs. "The ability to recreate a crash on the computer is a useful tool," says Douglas. "Imagine crash-testing a regular car. Engineers set up a prototype car, crash it into a wall, and take information from the crash-test dummies and different accelerometers throughout the car. If they don't like the results and think they can make it safer by changing the dashboard or steering-wheel material, for example, they have to build a new prototype and crash it all over again. On the computer, engineers can change material properties instantaneously and run the crash repeatedly. They get faster results and can make quicker decisions on how the car might be made safer."

Besides CART, other racing circuits employ the ADR 2 to study crash data. Formula 1 initiated the program and also uses it in its F3000 feeder series. Ditto for Indy Racing League (IRL) and its feeder series, the Infiniti Pro Series. And Delphi is currently working on the next-generation ADR 3. The biggest challenge, say engineers, is in boosting capabilities while deploying a light, small package that can withstand intense vibrations.

Traction control: from roadway to racetrack
Road and race cars share many technologies. Among the most recent is traction control — basically a function of electronics used to sense and control wheel spin. CART is now allowing traction control for the 2002 season and beyond.

On the road, traction control is generally a safety feature. But on the track, it's all about performance.

The engines in most race cars and many high-performance road cars produce more torque than the tires can transmit to the road at low speeds. When this happens, the wheels spin excessively and can't get proper traction. Ideal traction, leading to optimal forward acceleration, implies a certain amount of slip between the tire rubber and the road surface. If the wheel and road speed are the same, there's no slip. Cars can't quickly accelerate and have less control. Similarly, there's a loss of control and lack of acceleration if the tires spin at a significantly higher speed than the road speed, say 50%. Research shows there is optimum grip when tires spin 10% faster than road speeds, depending on those road speeds, tire temperatures, and other variables. Traction-control systems work by determining the optimum tire slip based on such variables and keeping the driver from applying more torque than the tires need.

Current traction-control systems work by comparing the drivenwheel speed with the theoretical wheel speed for zero slip, as explained by Bruce Wood, Cosworth Racing's CART program director.

"In a conventional rear-wheel-drive car and in Champ cars, a sensor measures the front wheel speed with zero slip and relays this information to the engine computer," he says. "The computer extrapolates the rear wheel speed from the engine speed and compares front and rear speeds. This gives the slip value that the computer then compares with the optimum value for the conditions as stored in a table."

Also, notes Wood, nearly all systems consider lateral or cornering force when calculating optimum engine torque. For example, if the optimum tire slip is 10% when the vehicle moves in a straight line, the computer will control the driving wheels to deliver that slip along the vehicle axis. But if the vehicle is cornering with a given g force or lateral acceleration, there will be some lateral tire slip. As a result, a lower slip percentage will be required to compensate for lateral slip, allowing for maximum safe acceleration in turns.

If the driver adjusts the throttle in a way that will make wheel slip exceed the calculated optimum value, the computer will override the driver's throttle request for more torque, limiting torque to an amount that will give tires maximum grip.

How the computer overrides the throttle differentiates the systems, says Wood.

"In road cars, drive-by-wire throttles or cable extenders may be used, but in race cars where speed is paramount, the computer usually cuts the ignition to instantaneously limit torque," he explains.

While a driver can achieve optimum traction control without such systems, doing so lap after lap can prove very difficult.

"Hence, traction control does not really make the perfect lap any faster but does let the driver lap more consistently," explains Wood. "This was demonstrated in the 2002 season-opening CART race at Monterey, Calif. Here several drivers strung together several very fast and consistent laps one after another. Without traction control, we would have seen only one such lap in every four or five."