SECTION 1.   BACKGROUND, OBJECTIVES, AND TASKS

I.   BACKGROUND

Phase I of the Consumer Braking Information Initiative developed brake test procedures for passenger cars and light-duty trucks in order to expand the scope of the New Car Assessment Program (NCAP). Phase II of the Consumer Braking Information Initiative provided the execution of the procedures with a round-robin series of service brake tests at three facilities to measure and evaluate variability in the results obtained at the different facilities and among different drivers. The Phase II effort additionally refined the brake test procedures and reporting methods which evolved from the Phase I program. This report documents Phase II of this program. The final report for Phase I of this program can be found on the National Highway Traffic Safety Administration's (NHTSA's) web site (www.nhtsa.dot.gov) under the Testing Results heading.

II.   OBJECTIVES

The test objectives were:

a.    Task 1.    Obtain and prepare test vehicles.

b.    Task 2.    Prepare facilities and test equipment.

c.    Task 3.    Perform brake burnish.

d.    Task 4.    Perform stopping distance tests.

e.    Task 5.    Gather environmental measurements.

f.    Task 6.    Perform the second round of testing.

g.    Task 7.    Perform the third round of testing.

h.    Task 8.    Perform the fourth round of testing.

i.    Task 9.    Conduct data analysis and reporting.

III.   TASK 1 - OBTAIN AND PREPARE TEST VEHICLES

        a.     Four model year 1999 commercial vehicles were leased from Enterprise Car Rental to support the test program. Table 1 identifies the four vehicles used throughout the Phase II effort. All four vehicles featured four-wheel antilock brake systems (ABSs). The Statement of Work (SOW) for the test program indicated that the vehicles should not have more than 10,000 miles of use prior to test initiation. However, due to the test program starting late in the automotive model year, it was difficult to locate low-mileage vehicles at that time. This deviation from the requirements in the SOW was coordinated with the NHTSA. Photographs of the test vehicles are provided in Appendix A, Figures A-1 through A-4.

TABLE 1.   TEST VEHICLE IDENTIFICATION

        b.     Prior to test initiation, the foundation brake components at each wheel position were replaced with new original equipment manufacture (OEM) hardware. The tires on each test vehicle were also replaced with new OEM tires of the specified make and size for the respective vehicle. A listing of all hardware replaced on the vehicles is presented in Appendix B.

        c.     During installation of the new brake hardware, type-K thermocouples were installed, as specified in Federal Motor Vehicle Safety Standard (FMVSS) 135, in the approximate center of the front brake linings of each vehicle. A tube-type decelerometer was affixed to the windshield of each vehicle as an aid for the driver during the brake burnish procedure. A force transducer, GSA Incorporated, model No. 114350, was affixed to the service brake pedal to measure pedal actuation force during the stopping tests. An indication of the brake pedal actuation force was provided to the driver in the form of three lamps. Yellow, clear, and red lamps would illuminate at pedal actuation forces of 125, 150, and 175 lbf, respectively. The service brake activation event was captured by monitoring the electrical signal at a rear service brake light on each vehicle. A fifth wheel, Nucleus model No. NC8, was affixed to the side of the test vehicle and was used to measure vehicle speed and stopping distance. Quadrature capability in the fifth wheel was utilized to ensure accuracy in the stopping distance measurement. This capability accounts for the fifth wheel directional changes resulting from the pitching motion of the vehicle as it comes to rest and nullifies any counts resulting from the rocking of the vehicle. The fifth wheel was calibrated by rolling it through a known 1000-foot distance and comparing the displayed distance to the known distance. A scale factor was created to correct the display value. The calibration process was repeated prior to the second round and after the fourth round of testing. Indication of vehicle speed was provided to the driver in the form of a digital display affixed to the windshield.

        d.     An Advanced On-Board Computer System (ADOCS), with supplemental ADOCS signal conditioning system (ASCS), was used to provide a continuous recording of the various test events and associated data throughout the program. The data sampling rate of the ADOCS was established at 5 msec (200 Hz). The data recording rate for all data channels was established at 40 Hz, except for the thermocouples and cool-down distance, which were recorded at 10 Hz. All data recorded by the ADOCS were date and time stamped and included sufficient identifiers for association to a specific vehicle and test condition. In addition to being stored aboard the ADOCS, some data were telemetered to an on-site data acquisition van in order to provide information to assist in test control. Two identical data acquisition systems (serial numbers (SNs) 612 and 880), utilizing ADOCS, were used throughout the test program and were moved between test vehicles as needed. Photographs of the data acquisition systems are provided in Appendix C, Figures C-1 through C-4.

IV.   TASK 2 - PREPARE FACILITIES AND TEST EQUIPMENT

        a.     Testing for Phase II was conducted at the U.S. Army Aberdeen Test Center (ATC), Aberdeen Proving Ground (APG), Maryland; at the Transportation Research Center (TRC), East Liberty, Ohio; and at MGA Research, Incorporated, Burlington, Wisconsin. At each facility, brake testing was performed on relatively level, straight, and uniform test surfaces featuring asphalt pavement. The brake tests during the first and fourth rounds of testing were conducted at ATC runway No. 4 at the Phillips Army Airfield (PAAF). Brake testing conducted at TRC during the second round of testing was performed on the Vehicle Dynamics Area. The third round of testing conducted at MGA was performed on the Oval Test Track. It should be noted that the MGA test surface contained some irregularities/crevices in the stopping area that were filled with a type of asphalt tar/sealant. Photographs of the test surfaces utilized at each facility are presented in Appendix D, Figures D-1 through D-4. Table 2 presents the findings of the grade and slope surveys of the test surfaces at the various test facilities. The first and fourth rounds of testing at ATC were performed such that the vehicles were traveling upgrade. The runway was crowned down the centerline, with slopes ranging from 0.6 to 1.4 percent falling away on either side. The TRC surface was characterized by a slight upgrade to the midpoint of the stopping lane, followed by a slight downgrade beyond the midpoint. The test surface slope at TRC was left to right and ranged from 0.8 to 1.2 percent. The MGA stopping surface featured a downgrade for about three-quarters of the length followed by a slight upgrade. The slope of the MGA surface was from right to left and ranged from 1.5 to 5.3 percent. The surface profile measurements received from MGA after testing was completed indicated a 600-foot length of test surface had been measured; however, there was no indication of how the the actual stopping area used during testing was situated with respect to the surveyed area. It should be noted that the extreme slopes reported on the survey occurred at either end of the 600-foot long surveyed area, at entrances/exits of curves on the test track. Stopping tests were not performed in the vicinity of the curves; therefore, the slopes in the 4- to 5-percent range were not encountered during testing.

TABLE 2.   TEST SURFACE PROFILE

        b.     The dry and wet peak frictional coefficients (PFCs) of the test surfaces were measured prior to and following testing at each test facility. All PFC measurements taken during the test program were obtained in accordance with American Society for Testing Materials (ASTM) Method E1337-90, using a standard reference tire specified by ASTM E1336. ATC test surface PFC measurements were taken by the Department of Transportation (DOT), Eastern Federal Highway Lands Division (EFHLD) prior to the first and fourth rounds of testing and following the fourth round of testing. The MGA test surface PFC measurements obtained prior to and following the third round of testing were also taken by DOT, EFHLD. The skid measurement system used by DOT was identified as a K. J. Law, model No. 1290. TRC obtained the PFC measurements of their test surface using a modified K. J. Law, model No. 965 system, which TRC owns and operates. Because the surface friction measurement systems require calibration and the latest calibration dates of the systems used are not known, there may be some variation in PFC results obtained with the different systems used during the program.

        c.     Atmospheric weather conditions were monitored and recorded during testing at all facilities. ATC and TRC used weather stations located in proximity to their test areas to record air temperature, barometric pressure, relative humidity, wind speed, and wind direction. These measurements were obtained at 15- and 30-minute intervals at ATC and TRC, respectively. The MGA facility did not have a weather station located at the test area. Air temperature was obtained periodically using a hand-held temperature indicator, and wind speed was measured periodically using a hand-held anemometer. General wind direction relative to the test surface heading was noted while testing at MGA. Throughout all testing, notations of sky conditions were recorded and road surface temperature was measured periodically using a hand-held temperature indicator.

         d.     Water delivery to the test surfaces for the wet road stopping tests was accomplished using a truck with spray nozzles at both ATC and TRC. This method produced a fairly uniform distribution of water onto the road surfaces during the tests. Prior to each series of wet surface testing at ATC and TRC, the water truck made two side-by-side passes over the test surface and followed those with a third water application centered over the initial two applications. Thereafter, the water truck made one water application down the middle of the stopping lane prior to each brake stop. At both ATC and TRC, the water used for test surface wetting was obtained from a potable water supply. Personnel wetted the test surface at MGA using a fire hose. Prior to a wet surface test sequence, the entire test surface was thoroughly wetted as personnel traversed the length of the test surface while spraying water. Personnel continued to wet the test surface in the same manner following each brake stop, ceasing spraying only when the test vehicle approached for another stop. The water used to wet the MGA test surface was pumped from a nearby stream. Photographs of the water application methods at ATC, TRC, and MGA are presented in Appendix D, Figures D-5 through D-7, respectively.

        e.     During initial preparation of the test vehicles at ATC, the weight distribution of each vehicle was measured on a calibrated platform scale (accuracy +0.5 percent of reading). The weight distribution of each vehicle was obtained at curb-loaded condition, which consisted of all fluid reservoirs and the fuel tank filled to capacity. Table 3 presents the findings of the weight distribution measurements of the test vehicles at curb weight.

TABLE 3.   CURB-LOADED WEIGHT
DISTRIBUTION

        f.     The desired test weight of the vehicles was curb weight plus 400 pounds. During performance of the stopping distance tests at all test sites, each vehicle's mass included the data acquisition system, a water-filled mass simulator in the front passenger seat, and the driver. The actual test weights exceeded the desired test weight on all vehicles. This was due to the passenger seat mass simulators being filled with water and installed in each vehicle prior to the final test weight being recorded. In an effort to expedite this Methodology Study, no effort was made to readjust the vehicles' weights prior to testing. Prior to each round of brake testing, all fluid reservoirs and the fuel tank of each vehicle were filled to capacity. Table 4 presents the as-tested weights of the vehicles, which were measured on the platform scale at ATC. The weights of the drivers used during testing ranged from approximately 160 to 230 pounds. Vehicle wheelbase measurements were also obtained during the initial vehicle preparation at ATC.

TABLE 4.   AS-TESTED WEIGHT DISTRIBUTION
AND WHEELBASE

V. TASK 3 - BRAKE BURNISH

    a.  Following installation and adjustment of the new foundation brake hardware, the vehicles were loaded to gross vehicle weight (GVW) in preparation for brake burnish. The vehicles were loaded with 50-pound sandbags, distributed as necessary to attain the specified gross axle weight ratios. Table 5 presents the manufacturers' recommended GVW ratios. Table 6 presents the vehicles' actual GVW distribution as loaded for the brake burnish.

TABLE 5.   RECOMMENDED GVW

TABLE 6.   BRAKE BURNISH WEIGHT
DISTRIBUTION

        b.     Each vehicle was subjected to brake burnish in general accordance with FMVSS 135, S7.1. This procedure required that 200 burnish stops be performed from an initial vehicle speed of 50 mph. From this initial speed, the vehicle was brought to a complete stop at a deceleration rate of 3.0 m/s² (9.8 ft/s²). As discussed earlier in Task 1, a tube-type decelerometer was affixed to the windshield to provide deceleration indication to the driver. Between each burnish, the vehicle was operated at 50 mph for a distance of approximately 2 km (1.24 mi).

        c.     At the completion of the burnish procedure, each vehicle's foundation brakes were inspected to ensure adjustment to the manufacturer's specifications. None of the vehicles' foundation brake hardware required manual adjustment following the brake burnish.

VI.   TASK 4 - STOPPING DISTANCE TESTS

a. Procedure.

        (1)    For each round of stopping tests, the vehicles were configured with all fluid reservoirs filled to capacity, a water-filled mass simulator positioned in the front passenger seat, the data acquisition system secured in the approximate center of the rear passenger seat, and a driver.

        (2)    The drivers at each test facility were instructed in the brake application method to be used throughout testing. This brake application method requested that the driver attain a brake pedal force of at least 100 lbf within 0.1 second of brake application and target a pedal force of 150 lbf thereafter. Visual feedback of pedal force was provided to the driver in the way of indicator lamps affixed to the vehicle dashboard. This pedal force indicator system was described earlier in Section III. Subsequent to each stop, the brake pedal application was categorized in accordance with the pedal application classes established during Phase I testing, which are presented in Table 7. The Phase I report provided some discussion on the adoption and use of these classes to categorize brake pedal applications.

TABLE 7.   BRAKE PEDAL EFFORT CLASSES

        (3)    Prior to performing stopping tests, the vehicle brakes were warmed to an initial temperature between 149 and 212 ^oF by performing light brake snubs. Personnel in the on-site data acquisition van communicated brake temperature information to the driver. Once the brake linings were at appropriate temperatures, the driver was instructed to perform a stop from a nominal vehicle speed of 62 mph. For each stop, vehicle speed at initial brake pedal application, stopping distance, and brake temperature were recorded. Stopping distance results were normalized, per Society of Automotive Engineers (SAE) J299, for a vehicle speed of 62 mph. Between stops, the vehicle was operated at 50 mph for a sufficient time to cool the front brake linings to a temperature at or below 212 ^oF. The distance and time traveled for these brake cool-down segments were recorded.

        (4)    A total of ten stops meeting the aforementioned brake pedal application criteria were requested for each vehicle and test condition at each test facility. During testing of a particular vehicle, test results were reviewed to ensure they were within specified criteria prior to proceeding to another test condition. Additional stops were performed, if needed, to ensure a sample of ten valid stops. At each facility, ten valid stops were required for each vehicle on both dry and wet pavement. When it became obvious that the drivers were having difficulty meeting the brake pedal application criteria for a Class A stop, stops falling within Classes B and C were accepted as valid.

b. Test Results.

The first round of testing was performed at ATC during the period 5 through 12 October 1999. Stopping distance results for all stops are presented in Appendix E, Tables E-1 through E-4. Plots of brake pedal activation force for the dry and wet surface stops are presented in Appendix E, Figures E-1 through E-4. The brake lining temperatures at the time of brake pedal application for each stop are presented in Appendix E, Tables E-5 through E-8. Distance traveled and time elapsed during the brake cool-down segments are presented in Appendix E, Tables E-9 and E-10. The environmental data recorded during the stopping distance tests are presented in Appendix E, Table E-11.

c. Analysis.

        (1)    To assist in the analysis of the stopping distances, brake pedal activation forces for all stops were categorized according to the criteria developed during Phase I of the Consumer Braking Information Initiative; the categories are presented in Table 7. Figures 1 through 4 present comparisons of the average stopping distance and standard deviation of data sets containing stops falling within the various pedal effort classes.

        (2)    In most cases during the first round of testing, the average stopping distance of the vehicles decreased with the removal of the Class D stops. Because of the limited number of Class A stops performed for some vehicle and road surface combinations, there were instances in which data sets with only class A stops had higher average stopping distances than other data sets. This also caused the standard deviation of some Class A-only data sets to be higher than data sets including the other pedal effort classes. However, in data sets containing a reasonable number of Class A stops, the differences in average stopping distance between the A, B, and C pedal effort classes were minimal.

        (3)    An anomaly was noted when reviewing the stopping performance of the Chevrolet Malibu. The wet stopping distance was shorter, although only slightly, than the dry stopping distance for all pedal effort categories. The dry and wet surface testing of the Malibu was performed on different days; however, atmospheric conditions, including surface temperature, were nearly identical between the days. It was unclear as to what might have caused these results.

        (4)    Environmental parameters were reviewed to determine their effect on the stopping distance tests. Air and road surface temperatures during the first round of tests at ATC ranged from 48 to 58 and 60 to 85 ^oF, respectively. An analysis of temperature effects was not possible given the limited number of stops performed for the various temperature conditions. Such an analysis would have required additional testing. During the wet stopping tests, no puddling of water deeper than 1/8 inch was observed, and there were no reports/observations of hydroplaning. An analysis of wind effects on two of the vehicles was performed, and those findings are presented in Figures 5 and 6. The analysis is presented in further detail in Appendix J. When considering wind effects, it should be noted that vehicle heading during the stopping tests at ATC was 040^o. A maximum tailwind of 13 mph was encountered while performing the dry stopping tests on the Chevrolet Astro. Those effects would equate to an approximate 1.5-foot increase in stopping distance. A maximum headwind of 16 mph was recorded during performance of the dry stopping tests on the Malibu. Those effects might have decreased stopping distance by approximately 1.3 feet.

        (5)    The effects of stopping surface grade on stopping distance were analyzed. Figure 7 presents the findings of that analysis, which is applicable to all of the vehicles tested during this program, and Appendix K describes the analysis in further detail. The 0.1-percent upgrade of the ATC test surface would have decreased stopping distance by approximately 0.1 foot.

VII.   TASK 5 - ENVIRONMENTAL MEASUREMENTS

        a.     As discussed in Section IV, PFCs of the test surfaces at each facility were obtained at both dry and wet conditions. A summary of the measurements is presented in Table 8. The values reported for ATC and MGA are the average of 15 measurements obtained for each site and surface condition. The data obtained during the PFC measurements at ATC and MGA are presented in Appendix L. There was some concern that the wet surface PFC values were little different from the dry surface values for the pretest measurements at ATC. Similar results were obtained at MGA for the pretest measurements. The post-test measurements revealed the wet surface PFC value was higher than that for the dry surface. This issue is discussed further in Section XI.

TABLE 8.   TEST SURFACE PFCs

        b.     Measurement of atmospheric conditions during testing was discussed in Section IV. Atmospheric data measured during the stopping distance tests at the various test facilities are presented with the stopping distance data in Sections VI, VIII, IX, and X. Table 9 summarizes the atmospheric conditions encountered at the various test sites. Because a weather station was not available during the third round of testing at MGA, the data presented are those periodically recorded using hand-held instruments. Generally, weather conditions were good throughout the test program, although the air and road surface temperatures were observed to be somewhat low in the mornings as a result of testing in late fall.

TABLE 9. SUMMARY OF ATMOSPHERIC CONDITIONS

VIII.   TASK 6 - SECOND ROUND OF TESTING

        a.     Procedure. The stopping test procedures discussed in Section VI were utilized at each test facility.

        b.     Test Results. The second round of testing was performed at TRC during the period 28 and 29 October 1999. The first stops for the Chevrolet Blazer and Ford F150, on dry asphalt, were not included in the analysis because the brake lining temperatures were above the allowable limit prior to these stops, resulting in a noticeable increase in stopping distance for the Blazer as compared to other stops in the series. Stopping distance results for all analyzed stops are presented in Appendix F, Tables F-1 through F-4. Plots of brake pedal activation force for the dry and wet surface stops are presented in Appendix F, Figures F-1 through F-4. Brake lining temperatures at the time of brake pedal application for each stop are presented in Appendix F, Tables F-5 through F-8. Distance traveled and time elapsed during the brake cool-down segments are presented in Appendix F, Tables F-9 and F-10. The environmental and weather data recorded during the stopping distance tests are presented in Appendix F, Table F-11.

c. Analysis.

        (1)    All stops performed at TRC were Class A stops; thus, there was no comparison of stopping distance results made between the different pedal effort classes. Figures 8 and 9 present the average stopping distance and standard deviation, respectively, of the results obtained for the four test vehicles.

IX. TASK 7 - THIRD ROUND OF TESTING

        a.     Procedures. The stopping test procedures discussed in Section VI were utilized at each test facility.

        b.     Test Results. The third round of testing was performed at MGA, Incorporated, during the period 3 and 4 November 1999. During this round of testing, the vehicle drivers requested and were given approval to utilize the cruise control function on the test vehicles to establish the nominal vehicle speed for the stopping tests. The first wet surface stops for the Chevrolet Astro and Blazer were excluded from analysis because the stopping distance exceeded the available wet surface and the stop terminated on dry asphalt. The seventh wet surface stop for the Blazer was also excluded because the driver reported some hydroplaning during the stop. The first wet stop for the Ford F150 was not included in the analysis because the brake lining temperatures at the point of brake application were above the acceptable limit. Stopping distance results for all analyzed stops are presented in Appendix G, Tables G-1 through G-4. Plots of brake pedal activation force for the dry and wet surface stops are presented in Appendix G, Figures G-1 through G-4. Brake lining temperatures at the time of brake pedal application for each stop are presented in Appendix G, Tables G-5 through G-8. Distance traveled and time elapsed during the brake cool-down segments are presented in Appendix G, Tables G-9 and G-10. The environmental and weather data recorded during the stopping distance tests are presented in Appendix G, Table G-11.

c. Analysis.

        (1)    As was done with the stopping distance data in Section VI, the brake pedal activation forces for all stops were categorized according to the criteria developed during Phase I of the Consumer Braking Information Initiative. Figures 10 through 13 present comparisons of the average stopping distance and standard deviation of data sets containing stops falling within the various pedal effort classes.

X.   TASK 8 - FOURTH ROUND OF TESTING

        a.     Procedures. The stopping test procedures discussed in Section VI were utilized at each test facility.

        b.     Test Results. The fourth round of testing was performed at ATC during the period 17 through 29 November 1999. The fifth dry stop for the Blazer was excluded from analysis because the brake lining temperatures were above the specified limit prior to the stop. Likewise, the first and fifth dry stops and the first wet stop for the F150 were not included due to high brake lining temperatures. Stopping distance results for all analyzed stops are presented in Appendix H, Tables H-1 through H-4. Sample plots of brake pedal activation force for the dry and wet surface stops are presented in Appendix H, Figures H-1 through H-4. The brake lining temperatures at the time of brake pedal application for each stop are presented in Appendix H, Tables H-5 through H-8. Distance traveled and time elapsed during the brake cool-down segments are presented in Appendix H, Tables H-9 and H-10. The environmental and weather data recorded during the stopping distance tests are presented in Appendix H, Table H-11.

c. Analysis.

        (1)    As was done with the stopping distance data in Sections VI and IX, the brake pedal activation forces for all stops were categorized according to the criteria developed during Phase I of the Consumer Braking Information Initiative. Figures 14 through 17 present comparisons of the average stopping distance and standard deviation of data sets containing stops falling within the various pedal effort classes.

        (2)    During the fourth round of testing, ATC drivers again had some difficulty executing Class A stops. However, the average stopping distances between the different pedal effort classes did not vary significantly. The largest differences in average stopping distance between the various pedal effort classes were 2.48 and 3.19 feet, respectively, for the dry and wet surfaces. If the Class D stops are excluded from the analysis, as was done in Phase I of the Consumer Braking Information Initiative, then the largest differences in average stopping distance between the remaining pedal effort classes became 1.44 and 3.19 feet, respectively. However, it should be noted that the number of Class A stops was quite low for some vehicles and surface conditions, which diminishes the significance of this analysis.

        (3)    Environmental parameters were reviewed to determine their effect on the stopping distance during the fourth round of tests. Air and road surface temperatures during this round of testing at ATC ranged from 35 to 52 and 43 to 60 ^oF, respectively. An analysis of temperature effects was not possible given the limited number of stops performed for the various temperature conditions. Such an analysis would have required additional testing. During the wet stopping tests, puddling of water deeper than 1/8 inch was not observed and there were no reports/ observations of hydroplaning during the tests. An analysis of wind effects was performed on the results from the fourth round of testing. During testing, the winds were observed to be predominantly crossing from left to right, although some headwinds were also experienced. A maximum headwind of 12 mph was recorded during performance of the stopping tests on the Malibu. Those effects may have decreased stopping distance by approximately 0.9 foot. Again, the insignificant grade of the ATC test surface had little effect on the stopping distance results.

XI.   TASK 9 - DATA ANALYSIS AND REPORTING

a. Procedures.

        (1)    Pedal application efforts. Brake pedal application efforts from all stops performed during the test program were analyzed to determine whether the pedal application criteria were met as well as whether the criteria should be modified or altered.

        (2)    Stopping distance results. The mean, standard deviation, and 95th percentile stopping distance results were determined for each series of stops performed during the program.

        (3)    Deviations in results between test sites. Stopping distance results obtained at the various test sites were compared and any substantial differences were analyzed to determine the cause(s).

        (4)    PFC measurements. PFC measurements of the test surfaces at the various sites were compiled and analyzed. A protocol was developed for the evaluation of road surfaces to ensure their adequacy for brake testing.

b. Test Results.

        (1)    Pedal application efforts. Time histories of pedal application force for all stops performed during the first through fourth rounds of testing are presented in Appendix E, Figures E-1 through E-4; Appendix F, Figures F-1 through F-4; Appendix G, Figures G-1 through G-4; and Appendix H, Figures H-1 through H-4, respectively. In addition, the force class categorization and maximum pedal force for each stop are presented in Appendix E, Tables E-1 through E-4; Appendix F, Figures F-1 through F-4; Appendix G, Figures G-1 through G-4: and Appendix H, Figures H-1 through H-4, respectively. Comparisons of the brake pedal efforts for each test site are presented in Appendix I, Tables I-1 through I-4.

        (2)    Stopping distance results. Tables 10 through 13 present the mean, standard deviation, and 95th percentile (1.645 standard deviations above the mean) stopping distances of all test series. Vehicles with high variability will have 95th percentile stopping distances significantly higher than the reported average, while those with small deviations between individual stopping distances will have values closer to the reported average. Because there was little deviation in the average stopping distance of the data sets representing the different pedal effort categories and because there were a limited number of Class A stops in some test series, data sets that include stops in the pedal effort categories A, B, and C are presented.

TABLE 10.   STOPPING DISTANCE STATISTICS,
CHEVROLET ASTRO

TABLE 11. STOPPING DISTANCE STATISTICS,
CHEVROLET BLAZER

TABLE 12.   STOPPING DISTANCE STATISTICS,
FORD F150

TABLE 13.   STOPPING DISTANCE STATISTICS,
CHEVROLET MALIBU

        (3)    Deviation in results between sites. Figures 18 through 21 present comparisons of average and standard deviation stopping distance between the test sites. Because there was little deviation in the average stopping distance of the data sets from the various test series and because there were a limited number of Class A stops in some test series, data sets that included stops in the pedal effort categories A, B, and C are presented.