Final Report for the Methodology Study of the Consumer Braking Information Initiative

Work Performed by U.S. Army Aberdeen Test Center, Fall 1998

Note: The following document includes the exective summary and main text of the report. To view the complete report including appendices, go to the Department of Transportation docket website at http://dms/search/ and use Docket Number 6583.

NHTSA is investigating the feasibility of developing a braking performance measurement test procedure for light vehicles. The development of a suitable test procedure to evaluate the braking performance of light vehicles would enable NHTSA to provide braking performance information such as stopping distance, in addition to crash test performance information, as part of the agency’s New Car Assessment Program (NCAP), on those new vehicles that are purchased for use in crash tests under the NCAP.

The Aberdeen Test Center, a division of the U.S. Army Material Command, in Aberdeen Maryland, was contracted by NHTSA to conduct this research effort. Tests were conducted during the Fall of 1998 on ten light vehicles, using straight line stops on dry and wet asphalt, from an initial speed of 62 mph, with each vehicle in both lightly-loaded and fully-loaded conditions. The purpose of the tests was to determine if variability in stopping distance could be minimized, to collect sufficient data to permit statistical analysis of the results, and provide direction in developing a test procedure.

Braking tests were conducted on five passenger cars, two passenger mini vans, one full-size cargo van, one full-size sport utility vehicle, and one full-size pickup truck. All of the vehicles were equipped with a four-wheel antilock braking system (ABS), except for the pickup truck which had a rear-wheel only ABS. The vehicles were leased and were either 1998 or 1999 model year vehicles, with mileages between 2,300 and 18,000 miles. The tires on each vehicle were replaced with new tires of the same make, model, and size as the original tires. Each vehicle’s brakes were inspected for normal wear, but were not replaced or subjected to conditioning other than from normal, as-received use. The new tires were conditioned by driving at 50 mph for 50 miles.

Selecting vehicles that were equipped with four wheel ABS was a decision intended to minimize the variability in stopping tests. If a vehicle does not have ABS, then the test driver must skillfully apply the brakes to attain minimum stopping distance without locking the vehicle’s wheels. Conversely, it was reasoned that a vehicle with ABS acting on all wheels could be braked sufficiently hard to activate the ABS (i.e., at least some of the wheels would lock up if the ABS was not present), and as long as the brake pedal force remained high enough to keep the ABS activated for the duration of the stop, then the ABS would keep the vehicle at its optimal level of braking. The pickup truck that only had rear-wheel ABS was acquired inadvertently and could not be included in the final results, but did provide useful information on brake pedal force at the threshold of front wheel lockup.

A peak brake pedal force of 112 lbs. (500 N) was targeted to be consistent with pedal forces specified for certain tests in Federal Motor Vehicle Safety Standard No. 135, Light Vehicle Brake Systems. However, brake applications as high as 450 lbs. were experienced during early testing, generally with the peak brake pedal force occurring at the top of the initial pedal force ramp-up. Subsequent efforts were made to target a steady pedal force of 150 lbs., with emphasis on rapid achievement of this force. Exceeding the target pedal force was not found to affect the stopping distance, however, since the ABS took control of the braking forces to prevent wheel lockup. For the pickup truck that was equipped with rear-wheel ABS, pedal forces in the 25 to 35-lb. range were found to be the pedal force just prior to front wheel lockup, and the peak pedal forces could not be achieved as rapidly as for the vehicles that had four-wheel ABS.

For each condition of load (lightly-loaded and loaded to Gross Vehicle Weight Rating [GVWR])and road condition (wet and dry asphalt), ten stops were made for a total of forty stops per vehicle. The driver was permitted to first make several test stops to become familiarized with each vehicle, and to warm up the brakes. After each stop, the vehicle was driven around the test area to cool the brakes, and then the brake rotors and drums were checked with a hand-held pyrometer to check that front rotor temperatures (which were always hotter than the rear brake drums/rotors) were below 212 degrees F before the next stop was conducted. One of the passenger cars was used as a control vehicle to provide comparative stopping data throughout the test program, and this vehicle was instrumented with thermocouples in the front brake linings to provide additional lining temperature data throughout the testing.

Road friction measurements of the test area were made eight times during the test period using a skid trailer. On each day that road friction was measured, ten measurements of the dry asphalt and ten measurements of the wet asphalt were made, and average dry and wet values were derived. The average peak coefficient of friction ranged from 0.89 to 0.95 for dry pavement and 0.85 to 0.88 for wet pavement. These measurements indicate that the asphalt surface was in good condition.

For each set of ten stops, the mean stopping distance was calculated along with the standard deviation and 95^th percentile stopping distance. Analysis of the pedal force attained during the first 0.3 seconds of brake application was used to develop the classification of a stop as Class A, B, C, and D, with Class D representing the slowest ramp-up of pedal force. Elimination of the slowest, Class D stops was found to have some effect on reducing the standard deviation (and hence 95^th percentile stopping distance) for some of the vehicles, while for other vehicles there was not an appreciable difference in eliminating the Class D stops. Appendix D provides an analysis of the effects on eliminating each successively slower class of stops from the ten stops for each condition of road and load. Appendix E provides final statistics for each vehicle with Class D stops removed. Note that in some cases, the remaining number of Class A, B, and C stops is small thus statistical significance of the mean and standard deviation is reduced. Also note that the Class A through D stop classifications do not apply to the pickup truck since much lower pedal forces were maintained in order to prevent front wheel lockup. Future research will be useful in determining what class of stop (e.g., Class C or better) can be consistently attained for most or all light vehicles equipped with four wheel ABS, now that these classifications have been identified.

NHTSA does not intend this report to provide comparative stopping distance information for the vehicles tested. Rather, the research effort is a preliminary effort to develop a test protocol that could be used in the future to measure the braking performance of NCAP vehicles. Further research is anticipated to further develop the test protocol, and determine, for example, if stopping tests can be replicated at other test facilities with consistent results. NHTSA is also coordinating this effort with European and Japanese governments with a goal of having a harmonized, international method that could be used to rate the braking performance of light vehicles.

THE MAIN TEXT OF REPORT BEGINS BELOW

AUTOMOTIVE INSTRUMENTATION TEAM

DATES OF TEST: 20 September 1998 through 20 November 1998

PREPARED BY:	Gregory A. Schultz, U.S Army Aberdeen Test Center
	Michael J. Babinchak, Dynamic Sciences Incorporated

U.S. ARMY ABERDEEN TEST CENTER

ABERDEEN PROVING GROUND, MARYLAND 21005-5059

TABLE OF CONTENTS 

I.	BACKGROUND

II.	OBJECTIVE

III.	TASK 1 – Perform Braking Performance Tests and Investigate the Causes of Stopping Distance Variability

	Procedure
	Test Results
	Analysis
	Conclusion
	
IV.	TASK 2 – Provide Details on Methodology to Address Variability

	Pedal Effort
	Vehicle Parameters
	Environmental Test Conditions
	Instrumentation and Measurement Techniques
	Test Sample Size
	
V.	TASK 3 – Develop a Test Protocol for the Braking Initiative

	General Test Conditions
	Procedural Conditions
	Required Test Data
	Measurement Techniques
	Road Test Procedures

VI.	TASK 4 - Identify a Method to Report Braking Performance to Consumers
	
VII.	TASK 5 – Develop a Test Report Format

	APPENDIX A	Vehicle Photographs
	APPENDIX B	Individual Brake Stop Results
	APPENDIX C	Sample Pedal Effort Plots
	APPENDIX D	Brake Stop Statistics with Pedal Effort Breakdown
	APPENDIX E	Final Performance Statistics for Each Vehicle
	APPENDIX F	Brake and Tire Temperature Data Sheets
	APPENDIX G	ATC Meteorology Data
	APPENDIX H	ASTM Frictional Skid Resistance Test Data
	APPENDIX J	Sample Wind Force Calculation
	APPENDIX K	Consumer Performance Measures
	APPENDIX L	Test Report Format

i

1. BACKGROUND

The U.S. Army Aberdeen Test Center (ATC) has performed a methodology study on passenger vehicle brake testing in support of an effort by the National Highway Traffic Safety Administration (NHTSA) to develop an effective consumer braking information program. With the implementation of this program, consumers would have access to brake performance information obtained from standardized test procedures, in addition to the collision safety information currently available.

 

2. OBJECTIVE

The objectives of this methodology study were the following:

Task 1 - Perform braking performance tests and investigate the causes of stopping distance variability.

Task 2 - Provide details on a test methodology to minimize variability.

Task 3 - Develop a test protocol for the braking initiative.

Task 4 - Identify a method to report braking performance to consumers.

Task 5 - Develop a test report format.

 

3. TASK 1 -Perform Braking Performance Tests and Investigate the Causes of Stopping Distance Variability

1. Procedure

Service brake effectiveness tests were conducted on 10 vehicles with anti-lock brake systems (ABS). Testing consisted of straight-line brake stops from 100 km/hr (62 mph). The brakes were applied so that the ABS was activated as quickly as possible and fully invoked throughout the brake stop until the vehicle came to rest. Vehicle speed, stopping distance and pedal force were measured and recorded during each stop event. The vehicles were operated by professional test drivers with brake test experience ranging from low to high. Each vehicle was tested under two payload configurations on both wet and dry asphalt surfaces.

The initial criteria for vehicle selection was for each test item to be less than one year old with between 8,000 and 16,000 km (5,000 and 10,000 miles) of usage. However, some exceptions to this rule were allowed based on vehicle availability. A list of the vehicles used during testing is presented in Table 1 and a photograph of each vehicle is included in Appendix A. All of the vehicles were equipped with four-wheel ABS except for the Dodge Ram 1500 4x4, which was equipped with only rear ABS. Additionally, each vehicle selected had an automatic transmission.

Vehicle No. 1 (Pontiac Grand Am) was also used as a baseline vehicle throughout testing. This vehicle was subjected to three instrumented brake stops each day of testing. These data were used to investigate variations in stopping distance caused by changes in environmental test parameters such as road surface friction, wind speed and ambient temperature.

Prior to testing, the OEM tires on each vehicle were replaced with new tires of the same make, model and size as the originals. Tire inflation pressures were set and maintained at the suggested levels shown on the tires. In retrospect, the inflation pressures recommended by the vehicle manufacturers should have been used. This change is reflected in Task 3, the test protocol. Following the tire replacement, each of the vehicles was operated for 80 km (50 miles) at approximately 80 km/hr (50 mph) to provide a limited break-in for the tires. No additional brake burnish procedure was conducted.

The weight distribution of each test item was determined with the vehicle empty and after being payloaded. The payload was configured based on the recommended gross vehicle weight (GVW) and maximum axle ratings reported on the driver-side door. Sand bags and body weight simulators, as shown in Figure 1, were used as payload.

Figure 1. Body Weight Simulator.

Testing was conducted at ATC’s Phillips Army Airfield facility, near the intersection of Runways 17 and 22. The longitudinal test course grade was 0.1 percent, with each brake stop performed upslope. Road surface frictional coefficients were measured and recorded by the Eastern Federal Lands Highway Division of the Federal Highway Administration per ASTM E1337, both with and without water delivery. Frictional data were taken prior to testing and re-measured periodically to ensure consistent conditions throughout testing. The test rig is shown in Figure 2.

Figure 2. Frictional Coefficient Test Rig.

Four test configurations were implemented:

1. No payload, dry asphalt
2. Full payload, wet asphalt
3. No payload, wet asphalt
4. Full payload, dry asphalt

Test data collected included:

• Vehicle road speed
• Stopping distance
• Pedal application force
• Brake rotor temperatures
• Ambient temperature
• Tire temperature
• Road surface temperature
• Tire pressure.

The test instrumentation installed on each vehicle consisted of ATC’s Advanced Onboard Computer System (ADOCS), a pedal force transducer, a rolling fifth-wheel, driver displays and brake-lining thermocouples (on vehicle No. 1 during baseline testing). A GSE Inc. model 114350 pedal effort transducer, Serial No. 90, was installed on the brake pedal to measure pedal force. A Nucleus model NC8 fifth-wheel, Serial No. 8479, shown in Figure 3, was used to measure vehicle speed and rolling distance. The resolution of the fifth-wheel and the force transducer were 0.01 m (0.03 ft) and 1.0 N (0.23 lb), respectively.

Figure 3. Grand Am Vehicle with Fifth-Wheel.

Other test instrumentation consisted of a hand-held, thermocouple-type pyrometer for measuring brake component temperature, tire temperature and ambient roadway temperature. Average wind speed, peak wind speed, average wind direction and wind direction standard deviation were obtained in 15-minute intervals using an anemometer provided by ATC’s Meteorology Team (MET).

Stopping distance, vehicle speed and brake pedal force data were sampled dynamically during each brake stop event. The stopping distance measurement was triggered by the vehicle brake light circuit and ended when the vehicle came to rest. The sampling rates for the fifth-wheel and pedal force transducer were 200 Hz and 10 Hz, respectively. To account for variability in the target speed at brake application, the measured stopping distances were normalized to 100 km/hr (62 mph) in accordance with SAE J299 (August 1987). All other vehicle-related measurements were obtained statically.

The following three sections describe the procedures used for brake application, brake temperature measurements and cool-down, and water application on the test surface. Each of the three procedures evolved to some degree during early testing and therefore, are being given separate consideration. While it is not typically desirable to modify procedures during testing, early results showed that some modifications were required.

Test drivers were initially instructed to perform brake stop events in a manner simulating a panic stop, with the transmissions left in drive. The goal was to fully invoke the ABS as quickly as possible, exceed the 500-N (112-lb) force limit used for compliance testing in FMVSS 135 and maintain a steady application until the vehicle came to rest. This brake application method emphasized vehicle performance, as opposed to driver performance, and ensured that all vehicle brake systems were controlled with sufficient force for peak ABS performance.

While testing the first two vehicles, average steady-state application forces typically varied from 1100 to 1500 N (250 to 350 lb), with peak forces as high as 2000 N (450 lb). Immediate generation of these high forces produced high initial application rates, generally exceeding 500 N (112 lb) in 0.1 seconds. While these high rates were desirable, the high steady-state forces were considered excessive. Therefore, a 660-N (150-lb) target was established for the steady-state force.

A different brake application method was required for the Dodge Ram 1500 4x4, since the vehicle was equipped with only rear ABS. In order to avoid lock-up of the front wheels, drivers had to perform brake stops with less pedal force than with the other vehicles, while still achieving optimum brake performance. This limitation resulted in brake stops with significantly lower initial ramp-up rates and subsequent steady-state force levels.

Prior to performing each brake stop, brake lining temperatures were required to be kept below 100 ^oC (212 ^oF). Since the use of thermocouples within the brake linings was not included in the scope of this test, thermocouple-type pyrometers were used to measure temperature. For disk brakes, the lining temperature on the exposed side of the outer brake pads was recorded, and for drum brakes the reading was taken on the outer surface of the drums, adjacent to the swept area of the brakes. The initial assumption was that the temperature gradient across the lining material and drum material was relatively small.

As brake temperature data was collected during testing of vehicle No. 1, brake rotor temperatures were also measured and recorded. A substantial difference was noted between the temperature of the front rotors and the temperature of the back of the pad linings. While the temperatures on the exposed side of the brake pads were found to be below the 100 ^oC (212 ^oF) limit, the rotor temperatures rose above 100 ^oC (212 ^oF) and reached as high as 196 ^oC (385 ^oF).

To gain a better understanding of the heat transfer across the brake pads, thermocouples were installed in the front brake pads and rear brake shoes of vehicle No. 1 prior to its use as a baseline vehicle. The thermocouples were placed approximately 1/16 inch below the lining surface adjacent to the rotor. During testing of vehicle No. 2, brake temperatures of the baseline vehicle were monitored and recorded using the thermocouples as well as manually with the pyrometer.

An examination of the temperature results showed that the thermocouple data closely matched the temperatures obtained from the rotors with the pyrometer. The findings revealed that the temperature at the outer surface of brake pads was not an accurate representation of the brake lining temperature. It was noted that the rotor and brake pad combination could be modeled as a classic heat equation problem, with the rotor temperature assigned as one boundary condition of the pad. Using this approach, a continuous temperature gradient would be expected across the pad with the temperature on the rotor side of the pad being equal to the rotor surface temperature.

Based on this model and an analysis of the test data, it was concluded that measurements of the rotor surface temperatures yielded relatively accurate measurements of the lining temperatures at the lining/rotor interface. It was also noted that the rear brake shoe temperatures of vehicle No. 1, obtained using thermocouples, were significantly lower than the temperatures of the front brakes.

As a result of these findings, the brake temperature measurement procedure was modified. Throughout the remainder of testing (starting with the third test vehicle), the temperatures of the front brake rotors were measured with the pyrometer and these readings were used as the temperature indicator to keep below 100 ^oC (212 ^oF). Rear brake temperatures were also recorded, but were always significantly cooler than the front.

Typically, after each brake stop, the front rotor temperatures were above the 100 ^oC (212 ^oF) limit and the next stop could not be initiated. To cool the brakes, the vehicle was operated at approximately 80 km/hr (50 mph) for a short period of time after each brake stop. Experimentation showed that the temperatures could be controlled and stabilized with the cool-down procedure lasting between 6 to 10 minutes, depending on the ambient temperature.

For wet asphalt testing, water was applied to the road surface using the water tanker rig shown in Figure 4. The truck was operated at approximately 32 km/hr (20 mph), while water from the tanker was placed over the test surface using the distribution pipe shown on the back of the tanker. The water was released through holes placed along the longitudinal axis of the pipe with pressure generated from the pressure head in the tanker.

Prior to wet surface testing, three passes were made with the water tanker traveling longitudinally along the test area, as shown in Figure 5. The first two passes were made side-by-side, and the third pass was made overlapping the center of the lane created by the first two passes. The total length of the wet area was approximately 150 m (500 ft). Prior to each brake stop event, an additional pass was made with the water tanker along the center lane where the brake stops were conducted. Water was distributed with the intent to fully wet the asphalt surface without creating excessive standing water.

Figure 4. Water Tanker Rig.

Figure 5. Water Application Procedure.

When vehicles No. 2 and 4 (Ford Expedition and Chevrolet Malibu) were tested on wet asphalt, some hydroplaning was experienced. An inspection of the test area revealed that standing water as deep as 1/4 inch had collected in minor depressions on the test course. To avoid this condition in later testing, the test area was displaced approximately 45 m (150 ft) farther up the runway, while still remaining within the area where the frictional measurements were taken. During subsequent testing, the water depth generally remained below 3 mm (1/8 inch).

2. Test Results

The weight distribution of each vehicle without payload is presented in Table 2. The weight distribution of each vehicle when fully payloaded and its corresponding gross vehicle weight (GVW) rating is presented in Tables 3 and 4, respectively. All weights were taken with driver and ADOCS included and with the vehicle fully fueled.

The longitudinal CG locations of each vehicle both empty and fully payloaded are presented in Tables 5 and 6, respectively. All tests were conducted with driver weight (using sand bags) and ADOCS included and with the vehicle fully fueled.

Brake stop results from each vehicle in all four test configurations are presented in Table 7. Results from day-to-day baseline testing with the Pontiac Grand Am are presented in Table 8. Stopping distances and deceleration rates shown for each vehicle configuration are averages of all stops conducted that were considered to follow the guidelines presented in the test procedure. Brake stops not conducted properly were removed from the data set.

Results from each individual brake stop for each vehicle can be found in Tables B-1 through B-11 in Appendix B. Sample plots of applied pedal effort versus time can be found in Appendix C in Figures C-1 through C-10. Each figure contains pedal force plots from all brake stops conducted within a specific test configuration. One group of plots from each test vehicle is included.

An analysis of the data was conducted to investigate the variability in braking performance of each vehicle and to determine the sensitivity of the brake stops to variables such as pedal effort, brake temperatures, surface conditions, environmental variations and payload. The analyses consisted of the following:

• The effect of brake pedal effort.
• The effect of brake temperature.
• The effect of tire temperature.

• Day-to-day variation effects.
• The effect of free standing water on wet stops.
• Test configuration effects.

Initially, stopping distance results from each vehicle test configuration were compiled and the mean and standard deviation (s_n-1) were calculated for each data set. One-sided, 95% confidence interval estimates were also determined for each data set assuming a normal distribution of the measured stopping distances. The reported one-sided confidence intervals of each data set indicate to a 95% confidence that the actual average stopping distance is below this value. The results are shown in Appendix D.

The data obtained from Dodge Ram testing were excluded from the following analyses, since the vehicle was equipped without front ABS. The rear ABS was effective at eliminating rear wheel lock-up, and thus yaw, during the brake stops. However, brake applications had to be performed with significantly less pedal effort than the other vehicles in order to eliminate front wheel lock-up. Therefore, there was no basis for comparison.

The effects of pedal effort and brake temperature on individual brake stops were examined first using all of the brake stop data found in Appendix B. Criteria were then established for each variable based on trends found within the data that adversely affected the validity of the brake stop results. These criteria will be discussed in the following sections. Individual brake stops not meeting the established criteria were then removed and the average stopping distance and standard deviation of the data set were recalculated. The final statistics for each vehicle after removing brake stops not meeting established criteria are presented in Appendix E. A summary of the final results are shown in Tables 9 and 10 and Figures 6 through 9.

Figure 6. Comparison of Vehicle Stop Results on Dry Surface

Empty Versus Payloaded.

Figure 7. Comparison of Vehicle Stop Results on Wet Surface

Empty Versus Payloaded.

Figure 8. Comparison of Vehicle Stop Results without Payload

Dry Versus Wet Surface.

Figure 9. Comparison of Vehicle Stop Results with Payload

Dry Versus Wet Surface.

Throughout testing, brake stops were performed by applying a specified force instantaneously upon the brake pedal and maintaining a target pedal force until the vehicle came to rest. A typical plot of pedal force application versus time can be seen in Figure 10. The effect of pedal force on vehicle stopping distance when applied in this manner was analyzed throughout testing. Specifically, two factors were examined closely to determine if variations in applied pedal effort led to deviations in stopping distances. First, the initial spike application was analyzed to determine if slower rates in achieving the target pedal force led to greater deviation between individual brake stops for each test configuration. Second, the pedal force after the initial spike was examined to determine if the magnitude of the steady-state pedal effort led to variations in stopping distances.

Figure 10. Typical Pedal Effort Application, Pedal Force versus Time.

In assessing the initial pedal force application rate recorded during testing, each stop was placed into one of four classes - A, B, C or D - based on the applied pedal force recorded at 0.1, 0.2 and 0.3 seconds for each brake stop. The applied pedal force range at each time interval that define the four classes are shown in Table 11. All brake stops had to fall within one of the four classes to be considered a valid brake stop. Stops with pedal forces falling below class D were concluded to have too slow a rise time and not considered valid. A sample plot of pedal force versus time for each class is shown in Figure 11.

Figure 11. Sample Pedal Effort Application, Classes A through D.

The data recorded from the pedal transducer showed an initial pedal force present at the brake event start time (t₀). The presence of this force can be attributed to the initial acceleration of the effective mass of the brake pedal and pedal force transducer. This observation is an application of Newton’s 2nd Law. Simply stated, the pedal can not move unless a force is applied to it. Under class A stop application rates, initial pedal accelerations of several g’s were present. Brake applications at these accelerations to the effective mass of the brake pedal and transducer resulted in the observed forces.

A statistical analysis of the data obtained from each vehicle under each configuration was done to assess the effect of the initial force spike on stopping distance. The statistical data can be found in Appendix D. The average, standard deviation and 95% one-sided confidence interval were determined for each group with all stops included in the population. The same analysis was conducted with stops from less desirable classes removed (one class at a time) from the population until only stops in class A remained.

The brake stop data and the corresponding statistical data showed that, in the majority of cases, improvement in standard deviation and average stopping distance was evident with the removal of stops included under class D. Figures 12 through 15 compare the average stopping distance and standard deviation calculated both with and without class D stops included for each vehicle configuration set that contained at least one "D" in the population. Of the 18 data sets, 14 sets showed a decrease in average stopping distance and 15 sets showed a decrease in standard deviation with the removal of class D stops. Only the Pontiac Bonneville on dry surface with payload had an increase in both categories. An overall analysis of the class D stops supports the trend in improved average stopping distance and standard deviation. Of the 43 total class D stops conducted during testing, 67 percent (29 of 43) placed in the longest three stops of a data set. Furthermore, of the 54 longest three stops from the 18 data sets containing at least one class D stop, 54 percent were class D. Based on these findings, class D stops were excluded from the final statistics data presented in Table 9 and Appendix E.

The statistics were recalculated after removing stops under class C and then class B . Generally, the removal of these stops produced no consistent trends in braking performance or left a population too small in size to examine statistically.

Figure 12. Comparison of Individual Data Sets with and without Class D Stops

Dry Surface without Payload.

Figure 13. Comparison of Individual Data Sets with and without Class D Stops

Wet Surface without Payload.

Figure 14. Comparison of Individual Data Sets with and without Class D Stops

Dry Surface with Payload.

Figure 15. Comparison of Individual Data Sets with and without Class D Stops Included

Wet Surface with Payload.

The steady-state pedal force after the initial spike was also examined to determine if its magnitude influenced stopping distance. In the analysis, 20 brake stops from Grand Am baseline testing were randomly selected and examined. Only stops included in the final results that met the class A pedal effort criterion were selected. The relevant data from each stop and the average stopping distance and standard deviation of the entire group is presented in Table 12.

An analysis of the data presented in Table 12 revealed that no significant difference in stopping distance was evident with varying levels of steady-state applied pedal effort. The average stopping distance of the six brake stops with average pedal efforts under 670 N (150 lb) was 43.9 m (144.1 ft), compared to 44.0 m (144.3 ft) for stops with average pedal efforts over 1110 N (250 lb). The remaining eight stops had an average stopping distance of 43.7 m (143.4 ft).

A further examination of Table 12 showed that the longest stop in the set [45.8 m (150.1 ft) on 7 October] had a higher average pedal force than the shortest stop [42.5 m (139.5 ft) on 15 October]. This observation supports the conclusion that the magnitude of the steady-state pedal force was independent of stopping distance.

No trends between recorded brake temperatures and brake performance data were noted when considering stops in which cool-down runs were conducted prior to the brake stop. However, initial stops performed with vehicles that sat stationary for extended periods of time, allowing brakes to cool to ambient temperatures, produced unfavorable results in some cases. Table 13 shows results from the eight stops conducted in which the brake temperatures were measured within 6 ^oC (10 ^oF) of ambient temperature before testing.

The data shows that eight out of the nine stops resulted in stopping distances at least one standard deviation from the average stopping distance of the data set. Based on these findings, stops with cold brakes were excluded from the final statistics presented in Tables 9 and 10 and Appendix E, in addition to the class D pedal effort stop exclusion. The remainder of the analysis was conducted with these revised statistics. All recorded brake temperature data can be found in Appendix F.

Tire temperature was measured to determine its effect on braking performance both w