Captain
18-06-2008, 09:02 PM
I came across this article, I though could be of interest to some RATS:
http://images.sportrider.com/tech/146_0603_08_z+traction_control+alex_barros.jpg
With the right electronics and a good software engineer, there is a lot of traction to be found in the wet. Alex Barros demonstrates at Shanghai.
If you attended any MotoGP event this year, you no doubt heard many of the bikes coughing and sputtering as they miss on acceleration. This is not because of some mechanical problem, but rather the engine is cutting out momentarily as the bike's electronics limit power to the rear wheel for traction control.
http://images.sportrider.com/tech/146_0603_01_z+traction_control+valentino_rossi.jpg
Traditional, simple traction control has been around for many years, and even in the motorcycle world: Honda's 1992 ST1100 had such a system, but it was discontinued after several years. In basic traction control, front- and rear-wheel speeds are monitored with sensors. A loss of traction is indicated by the rear wheel turning faster than the front, and a control unit decreases power by any one of several available means until both wheels are turning at the same speed. It all sounds simple enough, but in two-wheeled racing there are many additional factors to consider.
A quick perusal of traction control systems in the automotive world shows how they have evolved over the years. Some setups control power by retarding ignition timing momentarily, others by cutting fuel. Still others directly control the throttle plate with a servo-motor, much like Yamaha's 2006 R6. And some even apply the brakes to individual wheels rather than cut power. Much of the advances are in cars with four-wheel drive because it's hard to detect a difference in wheel speeds if they are all spinning, and some innovative ways of determining a slip condition must be employed.
Other details in cars that must be considered (and have been overcome) are the difference in the outside- and inside-wheel speeds when the car is turning, how to apply the control if the driver is applying the throttle and brake at the same time (evidently a common occurrence) and near-instantaneous changes in wheel speed when the car goes over bumps or a small patch of ice. Suddenly, it's not so simple anymore. The patents for the various systems are filled with flowcharts and logic diagrams rather than parts drawings and schematics.
The problems are similar in the motorcycle world, but what's changed in recent years to overcome them is not the basic knowledge or technology but rather the computing and electronic capabilities that can now be harnessed in a small and lightweight package. The problem now is not so much implementing traction control but rather taking advantage of all this computing power. Aftermarket companies are increasingly offering ECUs which provide ignition and injection control, as well as data acquisition. From there, adding traction control becomes a software exercise more than anything.
http://images.sportrider.com/tech/146_0603_02_z+traction_control+drawing.jpg
Honda's mid-'90s ST1100 was offered with a traction control option, which measured and compared front- and rear-wheel speeds to determine how much to retard ignition timing and cut back engine power.
Working with the basic concept of measuring wheel speeds and comparing them, we can consider the other aspects of a motorcycle's dynamics that must be taken into account and how to accommodate them. One important point is that riders use a spinning back tire to help steer, so a certain amount of slip must be enabled. Some wheelspin can actually increase traction when the bike is upright, but when the bike is leaned over then very little slip is desired. Conveniently, gyro sensors, a common part of any data acquisition system, can detect how far a bike is leaned over, and this value can be used as part of the control algorithm. Similarly, the throttle position sensor can be used--if the rider is using very little throttle, he is probably in the middle of a corner and wants as little wheelspin as possible. Use more throttle, and more spin can be allowed.
http://images.sportrider.com/tech/146_0603_03_z+traction_control+graph.jpg
A tire's traction level peaks at a given amount of slip, which is counterintuitive to no wheelspin at all for best traction. Traction and optimum slip value can change for different types of pavement. Some traction control systems can learn over time to adjust automatically for changing conditions.
More effective systems will take into account every sensor possible: gear position, gyro, rpm, throttle position, accelerometers (going uphill or downhill?) and even suspension travel (what if the front wheel is off the ground?). All the data is crunched using algorithms and look-up tables in the ECU, which then comes up with a number for the amount of slip required. The other side of the equation is controlling the power. Similar to the car applications, various methods can be used, including fuel or ignition tampering, or even control of the throttle in a fly-by-wire system. Heuristic systems can even adapt over time, learning how much traction is available at various lean angles and speeds and adjusting themselves accordingly. It really comes down to how savvy a software engineer is on your team more than anything.
http://images.sportrider.com/tech/146_0603_04_z+traction_control+throttle_plates.jpg
One way a traction control system can reduce power is by electronically closing the throttle. In this Honda setup, a servo motor and gears can advance or retard the throttle plates compared to the rider's inputs. A similar arrangement on the original Yamaha M1 gave throttle control for two cylinders to the ECU, and the most recent iteration has the ECU controlling all the throttle plates.
One major obstacle with traction control in four-wheel drive cars is how to detect the car's speed when all the wheels are under power and could be spinning. MotoGP bikes have a similar problem. On many straights, the front wheel is rarely on the ground and the front-wheel speed sensor may not be accurate. One solution for autos is a small radar gun pointing at the ground. Another is GPS, which is accurate enough these days to measure small changes in speed very quickly. Most, however, poll the various gyros and accelerometers and calculate the car's ground speed based on the sensor readings, and this is surely used on the MotoGP bikes in conjunction with the wheel speed sensors.
Many four-wheel series allow traction control simply because policing any rules outlawing it is nearly impossible. The required sensors can be easily hidden on a car and, in some cases, wheel speed is not even measured. Some add-on units using only engine rpm as a guide are as small as a disposable lighter and can be easily removed after a race. Disallowing the use of front-wheel speed sensors, as the AMA has done in an effort to stop traction control being used, is not much of a deterrent if the system is part of (and tucked away inside of) the ECU.
As ABS (and the wheel speed sensors and computing power that comes along with it) becomes standard equipment on more production motorcycles, traction control and other electronic trickery won't be much further behind.
http://images.sportrider.com/tech/146_0603_05_z+traction_control+wheel_sensor.jpg
More pick-up points for a wheel speed sensor to detect gives a higher resolution to the data, enabling a traction control system to react faster. The solid disc used here is a magnetic ring element, into which a strip of small magnets is embedded for more data points and accuracy than a toothed ring. The Yamaha M1s have been seen with sensors on each side of the wheel for redundancy. Also note in this picture the small opening in the fender with gradations, used to quickly measure maximum fork travel using an O-ring on the tube.
http://images.sportrider.com/tech/146_0603_08_z+traction_control+alex_barros.jpg
With the right electronics and a good software engineer, there is a lot of traction to be found in the wet. Alex Barros demonstrates at Shanghai.
If you attended any MotoGP event this year, you no doubt heard many of the bikes coughing and sputtering as they miss on acceleration. This is not because of some mechanical problem, but rather the engine is cutting out momentarily as the bike's electronics limit power to the rear wheel for traction control.
http://images.sportrider.com/tech/146_0603_01_z+traction_control+valentino_rossi.jpg
Traditional, simple traction control has been around for many years, and even in the motorcycle world: Honda's 1992 ST1100 had such a system, but it was discontinued after several years. In basic traction control, front- and rear-wheel speeds are monitored with sensors. A loss of traction is indicated by the rear wheel turning faster than the front, and a control unit decreases power by any one of several available means until both wheels are turning at the same speed. It all sounds simple enough, but in two-wheeled racing there are many additional factors to consider.
A quick perusal of traction control systems in the automotive world shows how they have evolved over the years. Some setups control power by retarding ignition timing momentarily, others by cutting fuel. Still others directly control the throttle plate with a servo-motor, much like Yamaha's 2006 R6. And some even apply the brakes to individual wheels rather than cut power. Much of the advances are in cars with four-wheel drive because it's hard to detect a difference in wheel speeds if they are all spinning, and some innovative ways of determining a slip condition must be employed.
Other details in cars that must be considered (and have been overcome) are the difference in the outside- and inside-wheel speeds when the car is turning, how to apply the control if the driver is applying the throttle and brake at the same time (evidently a common occurrence) and near-instantaneous changes in wheel speed when the car goes over bumps or a small patch of ice. Suddenly, it's not so simple anymore. The patents for the various systems are filled with flowcharts and logic diagrams rather than parts drawings and schematics.
The problems are similar in the motorcycle world, but what's changed in recent years to overcome them is not the basic knowledge or technology but rather the computing and electronic capabilities that can now be harnessed in a small and lightweight package. The problem now is not so much implementing traction control but rather taking advantage of all this computing power. Aftermarket companies are increasingly offering ECUs which provide ignition and injection control, as well as data acquisition. From there, adding traction control becomes a software exercise more than anything.
http://images.sportrider.com/tech/146_0603_02_z+traction_control+drawing.jpg
Honda's mid-'90s ST1100 was offered with a traction control option, which measured and compared front- and rear-wheel speeds to determine how much to retard ignition timing and cut back engine power.
Working with the basic concept of measuring wheel speeds and comparing them, we can consider the other aspects of a motorcycle's dynamics that must be taken into account and how to accommodate them. One important point is that riders use a spinning back tire to help steer, so a certain amount of slip must be enabled. Some wheelspin can actually increase traction when the bike is upright, but when the bike is leaned over then very little slip is desired. Conveniently, gyro sensors, a common part of any data acquisition system, can detect how far a bike is leaned over, and this value can be used as part of the control algorithm. Similarly, the throttle position sensor can be used--if the rider is using very little throttle, he is probably in the middle of a corner and wants as little wheelspin as possible. Use more throttle, and more spin can be allowed.
http://images.sportrider.com/tech/146_0603_03_z+traction_control+graph.jpg
A tire's traction level peaks at a given amount of slip, which is counterintuitive to no wheelspin at all for best traction. Traction and optimum slip value can change for different types of pavement. Some traction control systems can learn over time to adjust automatically for changing conditions.
More effective systems will take into account every sensor possible: gear position, gyro, rpm, throttle position, accelerometers (going uphill or downhill?) and even suspension travel (what if the front wheel is off the ground?). All the data is crunched using algorithms and look-up tables in the ECU, which then comes up with a number for the amount of slip required. The other side of the equation is controlling the power. Similar to the car applications, various methods can be used, including fuel or ignition tampering, or even control of the throttle in a fly-by-wire system. Heuristic systems can even adapt over time, learning how much traction is available at various lean angles and speeds and adjusting themselves accordingly. It really comes down to how savvy a software engineer is on your team more than anything.
http://images.sportrider.com/tech/146_0603_04_z+traction_control+throttle_plates.jpg
One way a traction control system can reduce power is by electronically closing the throttle. In this Honda setup, a servo motor and gears can advance or retard the throttle plates compared to the rider's inputs. A similar arrangement on the original Yamaha M1 gave throttle control for two cylinders to the ECU, and the most recent iteration has the ECU controlling all the throttle plates.
One major obstacle with traction control in four-wheel drive cars is how to detect the car's speed when all the wheels are under power and could be spinning. MotoGP bikes have a similar problem. On many straights, the front wheel is rarely on the ground and the front-wheel speed sensor may not be accurate. One solution for autos is a small radar gun pointing at the ground. Another is GPS, which is accurate enough these days to measure small changes in speed very quickly. Most, however, poll the various gyros and accelerometers and calculate the car's ground speed based on the sensor readings, and this is surely used on the MotoGP bikes in conjunction with the wheel speed sensors.
Many four-wheel series allow traction control simply because policing any rules outlawing it is nearly impossible. The required sensors can be easily hidden on a car and, in some cases, wheel speed is not even measured. Some add-on units using only engine rpm as a guide are as small as a disposable lighter and can be easily removed after a race. Disallowing the use of front-wheel speed sensors, as the AMA has done in an effort to stop traction control being used, is not much of a deterrent if the system is part of (and tucked away inside of) the ECU.
As ABS (and the wheel speed sensors and computing power that comes along with it) becomes standard equipment on more production motorcycles, traction control and other electronic trickery won't be much further behind.
http://images.sportrider.com/tech/146_0603_05_z+traction_control+wheel_sensor.jpg
More pick-up points for a wheel speed sensor to detect gives a higher resolution to the data, enabling a traction control system to react faster. The solid disc used here is a magnetic ring element, into which a strip of small magnets is embedded for more data points and accuracy than a toothed ring. The Yamaha M1s have been seen with sensors on each side of the wheel for redundancy. Also note in this picture the small opening in the fender with gradations, used to quickly measure maximum fork travel using an O-ring on the tube.