Calvin Bradley & Arnaud Delaval Presented at the 2011 meeting of the Tire Society

1. On-Road Fuel Consumption Testing to Determine the Sensitivity Coefficient Relating Changes in Fuel Consumption to Changes in Tire Rolling Resistance Calvin Bradley & Arnaud Delaval Presented at the 2011 meeting of the Tire Society

2. Agenda • Environmental and regulatory context • Rolling resistance basics • Previous results • Derivation of sensitivity coefficient • On-road testing • General procedure • Methods to monitor fuel consumption • Data Analysis • Results • Conclusions

3. Rolling resistance role in production of greenhouse gases • Rolling resistance contribution to over the road transport • Passenger vehicles: 20% of forces opposing motion • Heavy duty trucks: 30% of forces opposing motion • United States transportation • Transportation accounts for 29% of CO2 production • The majority comes from over the road vehicles • Passenger Vehicle: 59% • Heavy Truck: 19% • In the U.S. alone 337 million metric tons (MMT) of CO2 are produced to overcome rolling resistance • 1 MMT of CO2 is 200,000 Hot air balloons by volume • 1 MMT of CO2 requires almost 800,000 acres of Pine forest to offset

4. Tire Efficiency Labeling Around the World Final Label TBD NHTSA plans an on-line calculator to help consumers valorize low rolling resistance tires on their vehicle

5. Tire Deformations • As a tire is deformed to carry the load three types of deformations are occurring • Bending • Compression • Shearing

6. Definition of Tire’s Rolling Resistance Ztire 4-8 kg/t 7-12 kg/t • Tire’s rolling resistance is defined as the energy dissipated by a tire per unit of distance traveled Driving with 10kg/t tires is as if the vehicle was climbing a permanent 1% slope • Tire’s rolling resistance is characterized by a rolling resistance coefficient :

7. Previous results • J. Barrand and J. Bokar established a first order estimate for predicting differences in fuel consumption • The sensitivity coefficient α was further demonstrated to be relatively independent to drive cycle • Empirical prediction • Simulations • Closed circuit stabilized speed testing

8. Rolling Resistance & Resistance to Movement aerodynamic drag inertia rolling resistance α • A vehicle requires energy to move forward and improve the driving comfort • Resistance to Movement characterizes the effort to be overcome • Rolling Resistance is one of the force acting on the vehicle accessories internal friction gravity Example of accessories : air conditioning, power steering, on-board entertainment…

9. Derivation of Sensitivity Coefficient • Fuel consumption for at a given moment can be described by; • From previous discussion we know • Engine efficiency is a function of required torque • So efficiency ηis also a function of CRR Force Resisting Motion Fuel Consumption (volume per distance) Engine Efficiency Energy Density of Fuel

10. Derivation of Sensitivity Coefficient • Taking the derivative of fuel consumption with respect to CRR and simplifying we obtain • Relating this to previous publications then we see • If the change in efficiency with respect to CRRis very small • However, this approximation proves to be insufficient and the second term for α cannot be neglected

11. Fuel Consumption Testing

12. Test Overview • All tires were machine tested to determine their coefficient of rolling resistance • The impact of these tires on fuel consumption is measured • Real world conditions • Drive cycle with both urban and city portions • Typical E10 gasoline • 3 different vehicle segments tested • Compact: Toyota Corolla • Midsize: Chevrolet Impala • Light Truck: Chevrolet Silverado • Wide range of tires evaluated • 30 tire sets • 10 different brands • Approximate range rolling resistance: 7 to 13 kg/t

13. Test Overview • Significant variation can occur during testing • Sources of variation are controlled for consistency as much as possible • Fueling procedures • Vehicle factors such as alignment, AC, windows, lights, weight • Other sources of variation are permutated through all tire sets • Convoy position • Driver • Vehicle • Fuel consumption is measured by multiple independent methods • Test length must be sufficient to provide enough data for significant differences between tire sets • Duration of each test ranged from 12,500 km to 23,500 km • Results come from approximately 587,000 vehicle kilometers of testing

14. Methods to monitor fuel consumption • Fuel pump records • Defines what the consumer actually pays for • Measures volume of fuel at ground temperature • Data cannot be taken as frequently as other methods • Requires a precise fueling procedure • Fuel injector information • Measures fuel by injector pulse frequency and duration • Vehicle on board fuel economy displays • Injector data is calibrated by vehicle manufacturer • Can include error from tire diameter variations • Scan Gauge II OBD tool • Injector data must be calibrated for each vehicle • Independent of tire diameter • Inline volumetric flow meters • Significant cost • Requires cutting of fuel lines for install • Measures volume of fuel at fuel line temperatures

15. Methods Summary • Injector Data (Vehicle display & Scan Gauge) • Cost: Low • Complexity: Low or Medium • Precision: High • Accuracy: Low • Fuel Meter (fuel line volumetric meter) • Cost: High • Complexity: High • Precision: High • Accuracy: High • Fuel Pump (Service Station Records) • Cost: Low • Complexity: Medium • Precision: Low • Accuracy: Highest

16. Results • ANOVA analysis was used to correct for driver and vehicle effects • All methods were normalized to fuel pump levels • Results demonstrate agreement between methods

17. Results • ANOVA revealed within a test, fuel consumption changes greater than 1.2% were critically different at 95% confidence • Repeats of testing on each vehicle show similar relationship to CRR

18. Results • Testing did not reveal significant differences between sensitivity coefficients for each vehicle • Thus a single value for α was determined sufficient for all tested vehicles • Value of α confirms that is not small enough to be completely neglected ∆FC with lowest RR tire set as reference Measured FC for all testing

19. Conclusion • Tire rolling resistance has a significant impact on vehicle fuel consumption • Changes in fuel consumption can bepredicted by the linear empirical model • The sensitivity coefficient α primarily depends on fuel energy density and effective engine efficiency • The sensitivity coefficient α is not strongly a function of vehicle (outside of vehicle weight) or drive cycle • For typical American usage with E10 gasoline fuel savings can be predicted with: • α = 0.082 with ∆FC in L/100km, CRR in kg/t, and Mg in metric tons • α = 1.58E-5 with ∆FC in gal/100miles, CRR in kg/t, and Mg in lbs