U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-05-054
Date: September 2005

Quantification of Smoothness Index Differences Related To Long-Term Pavement Performance Equipment Type

Appendix A: LTPP Profiler Comparison Studies

1990 COMPARISON: AUSTIN, TX

This comparison was conducted in February 1990. The information presented in this section was obtained from a paper written about this comparison.(10)

The four LTPP DNC 690 profilers were used in this comparison. Six test sections, each 322 m (1056 ft) long were established for testing. Reference profile measurements along both wheelpaths at the test sections were obtained using Dipstick. The profilers performed five repeat runs at each test section at speeds of 56 and 80 km/h (35 and 80 mi/h). The left wheelpath was marked at the test sections and the profiler drivers were asked to align the profiler along this path. Each test section was divided into two sections for analysis, resulting in the availability of 12 sections for comparing the profiler and Dipstick IRI. The following are the main findings from this study:

  • In most cases, better agreement between profiler IRI and Dipstick IRI was obtained for the left wheelpath than for the right wheelpath. It was noted that the IRI for the right wheelpath usually was higher than that for the left wheelpath. Poor agreement between profiler IRI and Dipstick IRI usually occurred at the rough sections.
  • Significant differences between profiler IRI and Dipstick IRI were noted for several cases, with the profiler IRI being much higher than the Dipstick IRI. Most of the cases where these significant differences occurred were along the right wheelpath at extremely rough sections that had Dipstick IRI values in excess of 4 m/km (254 inches/mi). An evaluation of the profiler data for such cases indicated saturation spikes in the data, which was the cause of the high IRI value. It was concluded that the side-to-side rocking motion induced on the profiler when traversing these rough sections caused sunlight to seep under the shroud covering the sensors and contaminate the profile data.
  • Evaluation of the profiles also indicated instances where lost lock occurred. This was another factor contributing to differences between Dipstick and profiler IRI values.
  • An evaluation of the Dipstick IRI values obtained for the 12 sections indicated the following IRI distribution:
  • Left Wheelpath: Seven sections had IRI values between 1.2 and 2.4 m/km (76 and 152 inches/mi), three sections had IRI values between 2.4 and 4.7 m/km (152 and 298 inches/mi), and two sections had IRI values exceeding 4.7 m/km (298 inches/mi).
  • Right Wheelpath: Five sections had IRI values between 1.2 and 2.4 m/km (76 and 152 inches/mi), five sections had IRI values between 2.4 and 4.7 m/km (152 and 297 inches/mi), and two sections had IRI values exceeding 4.7 m/km (298 inches/mi).

As seen from these IRI values, many sections had extremely high IRI values. The side-to-side rocking motion of the profiler on some of these extremely rough sections caused sunlight to seep under the shroud covering the sensors and contaminate the profile data. The majority of the LTPP test sections will not have such high roughness values. Thus, contamination of profile data by ambient light probably is not a major issue at the majority of the sections if the shroud covering the sensors is in good condition. However, at extremely rough sections, sunlight seeping under the shroud could contaminate the data by causing saturation spikes to appear in the data.

1991 COMPARISON: ANN ARBOR, MI TX

The four LTPP DNC 690 profilers participated in this comparison. Details about this comparison were obtained from reference 11.

Eight test sections were used in this study-four AC sections and four jointed PCC sections. When selecting the test sections, two levels of roughness were considered-IRI less than 2.0 m/km (127 inches/mi) and IRI between 2.0 and 4.7 m/km (127 and 298 inches/mi). The goal was to establish two sections for each pavement type that fell into each of these roughness levels. Dipstick measurements were obtained along both wheelpaths on all of the test sections. Profile testing was performed at speeds of 64 and 80 km/h (40 and 50 mi/h). The left wheelpath was marked on all of the test sections, and the profiler drivers were asked to align the profiler along this path when collecting data. Each profiler operator was instructed to obtain six error-free profile runs on each test section at each test speed.

The left- and right-wheelpath IRI values computed from the profiler data collected at the 80 km/h (50-mi/h) test speed and from the Dipstick data are presented in tables 17 and 18, respectively. The IRI values presented in these tables are the average IRI values computed from the six repeat runs. The Dipstick IRI values presented in these tables are the average IRI obtained for the two sets of measurements that were available for each wheelpath.

 

Table 17. IRI values along the left wheelpath (1991).
Device Left-Wheelpath IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4 5 6 7 8
North Atlantic Profiler 1.42 3.45 2.40 0.91 2.41 2.65 1.94 0.84
North Central Profiler 1.14 3.33 2.35 0.99 2.67 2.62 1.91 0.91
Southern Profiler 1.29 3.63 2.48 0.93 3.14 2.84 2.05 0.98
Western Profiler 1.18 3.31 2.27 0.87 2.74 2.62 1.86 0.84
Dipstick 1.20 3.44 2.10 .84 2.65 2.43 1.81 0.98

1 m/km = 5.28 ft/mi

 

Table 18. IRI values along the right wheelpath (1991).
Device Right-Wheelpath IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4 5 6 7 8
North Atlantic Profiler 1.29 4.64 2.63 1.63 3.09 2.62 1.77 0.96
North Central Profiler 1.34 4.29 2.54 1.01 2.89 2.65 1.74 0.96
Southern Profiler 1.26 4.37 2.59 0.99 2.97 2.59 1.72 0.91
Western Profiler 1.26 4.43 2.54 0.98 2.89 2.57 1.75 0.95
Dipstick 1.26 3.91 2.48 0.93 2.65 2.41 1.75 1.10

1 m/km = 5.28 ft/mi

The differences between the average profiler IRI and the Dipstick IRI at the test sections for the left and right wheelpaths are presented in tables 19 and 20, respectively.

 

Table 19. Differences between profiler IRI and Dipstick IRI: Left wheelpath (1991).
Device Profiler IRI – Dipstick IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4 5 6 7 8
North Atlantic Profiler 0.22 0.02 0.30 0.08 -0.24 0.22 0.13 -0.14
North Central Profiler -0.06 -0.11 0.25 0.16 0.02 0.19 0.09 -0.06
Southern Profiler 0.09 0.19 0.38 0.09 0.49 0.41 0.24 0.00
Western Profiler -0.02 -0.13 0.17 0.03 0.09 0.19 0.05 -0.14

1 m/km = 5.28 ft/mi

 

Table 20. Differences between profiler IRI and Dipstick IRI: Right wheelpath (1991).
Device Profiler IRI – Dipstick IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4 5 6 7 8
North Atlantic Profiler 0.03 0.73 0.16 0.09 0.44 0.21 0.02 -0.14
North Central Profiler 0.08 0.38 0.06 0.08 0.24 0.24 -0.02 -0.14
Southern Profiler 0.00 0.46 0.11 0.06 0.32 0.17 -0.03 -0.19
Western Profiler 0.00 0.52 0.06 0.05 0.24 0.16 0.00 -0.16

1 m/km = 5.28 ft/mi

The standard deviations of the IRI values for the 80-km/h (50-mi/h) testing for the left and right wheelpaths are presented in tables 21 and 22, respectively.

 

Table 21. Standard deviations of IRI for 80-km/h (50-mi/h) runs: Left wheelpath (1991).
Device Standard Deviations of IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4 5 6 7 8
North Atlantic Profiler 0.08 0.07 0.08 0.02 0.05 0.03 0.01 0.03
North Central Profiler 0.03 0.03 0.03 0.02 0.04 0.07 0.04 0.01
Southern Profiler 0.00 0.02 0.02 0.01 0.01 0.05 0.03 0.01
Western Profiler 0.01 0.02 0.00 0.02 0.02 0.05 0.04 0.03

1 m/km = 5.28 ft/mi

 

Table 22. Standard deviations of IRI for 80-km/h (50-mi/h) runs: Left wheelpath (1991).
Device Standard Deviations of IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4 5 6 7 8
North Atlantic Profiler 0.03 0.11 0.04 0.01 0.04 0.04 0.04 0.02
North Central Profiler 0.06 0.06 0.04 0.01 0.05 0.02 0.02 0.02
Southern Profiler 0.03 0.08 0.03 0.01 0.02 0.03 0.02 0.03
Western Profiler 0.04 0.15 0.06 0.01 0.05 0.03 0.02 0.01

1 m/km = 5.28 ft/mi

The main findings from this study are:

  • An ANOVA performed on the IRI values indicated that the left-wheelpath IRI from the Southern profiler was different from the IRI values from the other three profilers. An ANOVA also showed that the left-wheelpath IRI computed from the data collected by the North Central, Western, and North Atlantic profilers were similar. The overall mean IRI values for the profilers along the left wheelpath computed by considering all of the profile runs at all of the test sections at the 80-km/h (50-mi/h) test speed were 1.98, 1.97, 2.00, and 2.10 m/km (126, 125, 127, and 140 inches/mi) for the North Central, Western, North Atlantic, and Southern profilers, respectively. These values clearly show that the left sensor of the Southern profiler is collecting data that have higher IRI values than the data collected by the other three profilers.
  • In the right wheelpath, the ANOVA indicated that the IRI values obtained from profile data collected by the profiler combinations of North Central, Western, and North Atlantic; North Central, Western, and Southern; and North Central, North Atlantic, and Southern, were similar. Although the right sensor of the North Central profiler did not follow the same path as the other three profilers (because of the difference in sensor spacing), all three cases in which the profilers were not significant in the ANOVA involved this profiler. Although the North Atlantic, Western, and Southern profilers have similar sensor spacing, the ANOVA indicated that at least one of the profilers was different from the rest. The overall mean IRI values for the profilers along the right wheelpath computed by considering all of the profile runs at all of the test sections at the 80-km/h (50-mi/h) test speed were 2.18, 2.16, 2.26, and 2.15 m/km (138, 137, 143, and 138 inches/mi) for the North Central, Western, North Atlantic, and Southern profilers, respectively. Although the ANOVA indicated that the IRI values for the four profilers were not similar, an examination of the IRI values obtained by the profilers at the test sections did not indicate clear evidence that a particular profiler had a bias when compared to the other profilers.
  • The results of the ANOVA for the left-wheelpath IRI showed that the speed of testing was not significant. For the right wheelpath, the speed of testing was a significant factor for two of the profiler combinations out of a total of five combinations that were analyzed.
  • All profilers showed excellent repeatability in IRI values along both wheelpaths, except at a few sections that had spikes in the profile data. It was observed that the repeatability of the profilers was not affected by the surface type (asphalt vs. concrete), the level of roughness, or the two speeds selected for testing. Although the Southern profiler was producing higher IRI values when compared to the other three profilers along the left wheelpath, the repeatability of IRI values obtained by this profiler was comparable to those obtained for the other three profilers.

1992 COMPARISON: AMES, IA TX

The four LTPP DNC 690 profilers participated in this comparison. The information presented in this section was obtained from a document written about this comparison.(12)

Eight test sections were used in this study-four AC sections and four PCC sections. When selecting the test sections, two levels of roughness were considered (IRI less than 2.0 m/km (127 inches/mi) and IRI between 2.0 and 4.7 m/km (127 and 298 inches/mi)). The goal was to establish two sections from each pavement type that fell into each of these roughness levels. Dipstick measurements were obtained along both wheelpaths at all of the sections, with replicate measurements obtained at seven sections. At six sections, two sets of Dipstick measurements were obtained, with three sets of Dipstick measurements being obtained at one section. Rod-and-level measurements were also obtained at these test sections; however, at some sections, only one wheelpath was surveyed. Two types of levels were used to perform the rod-and-level measurements. The profilers collected data at speeds of 64 and 80 km/h (40 and 50 mi/h). It was requested that each profiler operator obtain six good profile runs at ssection for each test speed.

During data analysis, it was discovered that one of the levels that was used to obtain measurements did not meet the resolution requirements for rod-and-level measurements for determining IRI that are outlined in ASTM Standard E1364-95 (2000).(31) Measurements obtained from this level produced high IRI values.

The left- and right-wheelpath IRI values computed from the profiler data collected at 80 km/h (50 mi/h), Dipstick data, and rod-and-level data are presented in tables 23 and 24, respectively. The IRI values presented in these tables for the profilers are the average IRI values computed from the six repeat runs. The Dipstick IRI values presented in these tables are the average IRI obtained from the replicate measurements for cases where more than one set of data were available. For the rod and level, only the IRI values obtained from the data collected with the level that met the resolution requirements outlined in ASTM Standard E1364 are shown.

 

Table 23. IRI values along the left wheelpath (1992).
Device IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
3 5 6 7 1 2 4 8
North Atlantic Profiler 1.42 1.18 0.66 1.53 1.59 2.21 4.23 1.42
North Central Profiler 1.39 1.25 0.65 1.75 1.62 2.13 4.59 1.28
Southern Profiler 1.39 1.23 0.60 1.55 1.74 2.37 4.37 1.47
Western Profiler 1.39 1.23 0.60 1.53 1.69 2.33 4.34 1.44
Dipstick 1.47 1.31 0.66 1.85 1.81 2.32 4.12 1.39
Rod and Level N/A 1.45 N/A N/A N/A 2.13 4.15 N/A
N/A = Measurements not obtained.

1 m/km = 5.28 ft/mi

 

Table 24. IRI values along the right wheelpath (1992).
Device IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
3 5 6 7 1 2 4 8
North Atlantic Profiler 1.39 0.93 0.79 3.50 2.02 1.67 5.68 1.09
North Central Profiler 1.36 0.85 0.74 3.64 1.97 1.59 5.74 1.09
Southern Profiler 1.32 0.88 0.74 3.08 1.97 1.59 5.55 1.09
Western Profiler 1.45 0.93 0.76 3.61 2.02 1.61 5.68 1.12
Dipstick 1.37 0.93 0.76 3.52 2.13 1.69 5.63 1.10
Rod and Level N/A N/A 0.95 N/A N/A 1.66 5.57 1.20
N/A = Measurements not obtained.

1 m/km = 5.28 ft/mi

The differences between the average profiler IRI and the Dipstick IRI at the test sections for the left and right wheelpaths are presented in tables 25 and 26, respectively.

 

Table 25. Differences between profiler IRI and Dipstick IRI: Left wheelpath (1992).
Device Profiler IRI - Dipstick IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
3 5 6 7 1 2 4 8
North Atlantic Profiler -0.05 -0.13 0.00 -0.32 -0.22 -0.11 0.11 0.03
North Central Profiler -0.08 -0.06 -0.02 -0.09 -0.19 -0.19 0.47 -0.11
Southern Profiler -0.08 -0.08 -0.06 -0.30 -0.08 0.05 0.25 0.08
Western Profiler -0.08 -0.08 -0.06 -0.32 -0.13 0.02 0.22 0.05

1 m/km = 5.28 ft/mi

 

Table 26. Differences between profiler IRI and Dipstick IRI: Right wheelpath (1992).
Device Profiler IRI - Dipstick IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
3 5 6 7 1 2 4 8
North Atlantic Profiler 0.02 0.00 0.03 -0.02 -0.11 -0.02 0.05 -0.02
North Central Profiler -0.02 -0.08 -0.02 0.13 -0.16 -0.09 0.11 -0.02
Southern Profiler -0.05 -0.05 -0.02 -0.44 -0.16 -0.09 -0.08 -0.02
Western Profiler 0.08 0.00 0.00 0.09 -0.11 -0.08 0.05 0.02

1 m/km = 5.28 ft/mi

The standard deviations of the IRI values obtained from testing performed at 80 km/h (50 mi/h) for the left and right wheelpaths are presented in tables 27 and 28, respectively.

 

Table 27. Standard deviations of IRI for 80-km/h (50-mi/h) runs: Left wheelpath (1992).
Device Standard Deviations of IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
3 5 6 7 1 2 4 8
North Atlantic Profiler 0.05 0.02 0.02 0.05 0.05 0.03 0.03 0.02
North Central Profiler 0.02 0.02 0.02 0.04 0.04 0.02 0.11 0.03
Southern Profiler 0.01 0.01 0.01 0.04 0.02 0.02 0.02 0.01
Western Profiler 0.02 0.01 0.01 0.04 0.04 0.02 0.03 0.03

1 m/km = 5.28 ft/mi

 

Table 28. Standard deviations of IRI for 80-km/h (50-mi/h) runs: Right wheelpath (1992).
Device Standard Deviations of IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
3 5 6 7 1 2 4 8
North Atlantic Profiler 0.04 0.02 0.02 0.07 0.03 0.03 0.04 0.02
North Central Profiler 0.02 0.01 0.01 0.22 0.02 0.02 0.05 0.02
Southern Profiler 0.02 0.01 0.01 0.06 0.02 0.02 0.04 0.02
Western Profiler 0.02 0.02 0.01 0.09 0.02 0.02 0.06 0.02

1 m/km = 5.28 ft/mi

The following are the main findings of this study:

  • All profilers showed excellent repeatability in IRI for both the left and right wheelpaths, except for the right wheelpath of section 7, which is an AC pavement with high-severity longitudinal cracking along the right wheelpath. Very high standard deviations in IRI were observed at this section. It was observed that the repeatability of the profilers was not affected by the two speeds of testing or the surface type (AC and PCC).
  • Generally, good agreement was found between the IRI computed from the Dipstick data and the profiler data for all four of the profilers along both wheelpaths at the majority of the sections.
  • An ANOVA performed separately on the left- and right-wheelpath IRI values obtained from Dipstick and the four profilers showed that the device type was not significant.
  • An ANOVA performed separately on the left- and right-wheelpath IRI values obtained from the profilers indicated that the speed of testing was not significant.
  • One of the levels that was used to obtain elevation measurements did not meet the resolution required for profile measurements that are specified in ASTM Standard E1364-95 (2000).(31) IRI computed from profiles measured by this level showed poor agreement with both profiler IRI and Dipstick IRI, with the rod-and-level IRI being higher.
  • IRI obtained from the measurements recorded by the level that met the ASTM resolution requirements showed good agreement with the IRI computed from the Dipstick and profiler data for cases where the IRI was greater than 1.6 m/km (101 inches/mi). However, there was poor agreement in the IRI for cases where the IRI was less than 1.6 m/km (101 inches/mi). Errors in leveling because of instrument errors (repeatability errors) and the deviation of the rod from the vertical will introduce random variations (noise) into the profiles. An analysis indicated that random variations in a profile have a much greater effect on the IRI for smooth pavement than for rough pavement. It was concluded that the effects of random variations in profiles caused by the previously described factors caused the rod-and-level IRI to have poor agreement with both the Dipstick IRI and profiler IRI for wheelpaths that had an IRI of less than 1.6 m/km (101 inches/mi).
  • Results from a study at one section indicated that the IRI for a concrete pavement could be affected considerably by slab curling. At that section, the morning and afternoon IRI values obtained by the same profiler for the left wheelpath were 2.40 and 2.11 m/km (152 and 134 inches/mi), respectively, while the corresponding values for the right wheelpath were 1.85 and 1.59 m/km (117 and 101 inches/mi), respectively.

1998 COMPARISON: URBANA, IL TX

The four LTPP T-6600 profilers participated in this comparison. This profiler comparison was the first-ever comparison of the four T-6600 profilers since profile data for the LTPP program was first collected in late 1996. The information presented in this section was obtained from a document written about this comparison.(13)

Four test sections were used in this study. Two of the sections were asphalt-surfaced, while the other two were PCC sections. When selecting the test sections, the goal was to select for each surface type one section with an IRI of less than 1.6 m/km (101 inches/mi) and one section with an IRI greater than 2.2 m/km (139 inches/mi). Dipstick measurements were obtained along both wheelpaths at all of the sections. Profile testing was conducted at a speed of 80 km/h (50 mi/h). Each profiler conducted five error-free profile runs on a test section.

The left- and right-wheelpath IRI values computed from the profiler data and the Dipstick data are presented in tables 29 and 30, respectively. The IRI values presented in these tables for the profilers are the average IRI values computed from the IRI for the five repeat runs.


The differences between the average profiler IRI and the Dipstick IRI at the test sections for the left and right wheelpaths are presented in tables 31 and 32, respectively.


The standard deviations of the IRI values for the profilers for the left and right wheelpaths are presented in tables 33 and 34, respectively.

The following are the main findings of this study:

  • The precision of a profiler along each wheelpath was evaluated by computing the standard deviation of the IRI for that wheelpath using the IRI values obtained from the five repeat runs. A profiler was considered to have failed the precision criterion if the standard deviation of the IRI for a wheelpath exceeded 0.04 m/km (2.5 inches/mi). All of the profilers met the precision criterion along both wheelpaths at all of the sites.
  • The IRI bias of a profiler along each wheelpath at a section was evaluated by computing the difference between the profiler IRI (average IRI from five runs) along that wheelpath and the IRI obtained by Dipstick. A profiler was considered to have satisfied the bias criterion if this difference in IRI was within ±0.16 m/km (±10 inches/mi). The North Atlantic, Southern, and Western profilers passed the bias criterion for both wheelpaths at all of the test sections. The North Central profiler passed the bias criterion for all of the cases, except for the right wheelpath of the rough AC section (section 2). The variability in the IRI values obtained by the North Central profiler for the repeat runs is considered to be the cause of the profiler failing the bias criterion at this section.

Atlantic, Southern, and Western profilers passed the bias criterion for both wheelpaths at all of the test sections. The North Central profiler passed the bias criterion for all of the cases, except for the right wheelpath of the rough AC section (section 2). The variability in the IRI values obtained by the North Central profiler for the repeat runs is considered to be the cause of the profiler failing the bias criterion at this section.

 

Table 29. IRI values along the left wheelpath (1998).
Device IRI (m/km))
Surface Type and Section Number
Asphalt Concrete
1 2 3 4
North Atlantic Profiler 1.02 2.49 1.30 2.77
North Central Profiler 1.04 2.55 1.28 2.90
Southern Profiler 1.02 2.56 1.27 2.83
Western Profiler 1.02 2.59 1.28 2.90
Dipstick 0.95 2.52 1.17 2.87

1 m/km = 5.28 ft/mi

 

Table 30. IRI values along the right wheelpath (1998).
Device IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4
North Atlantic Profiler 1.08 2.38 1.26 3.09
North Central Profiler 1.05 2.62 1.24 3.17
Southern Profiler 1.04 2.39 1.24 3.05
Western Profiler 1.04 2.50 1.22 3.20
Dipstick 0.96 2.46 1.13 3.17

1 m/km = 5.28 ft/mi

 

Table 31. Differences between the profiler IRI and Dipstick IRI: Left wheelpath (1998).
Device Profiler IRI -Dipstick IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4
North Atlantic Profiler 0.07 -0.03 0.13 -0.10
North Central Profiler 0.09 0.03 0.11 0.03
Southern Profiler 0.07 0.04 0.10 -0.04
Western Profiler 0.07 0.07 0.11 0.03

1 m/km = 5.28 ft/mi

 

Table 32. Differences between the profiler IRI and Dipstick IRI: Right wheelpath (1998).
Device Profiler IRI-Dipstick IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4
North Atlantic Profiler 0.12 -0.08 0.13 -0.08
North Central Profiler 0.09 0.16 0.11 0.00
Southern Profiler 0.08 -0.07 0.11 -0.12
Western Profiler 0.08 0.04 0.09 0.03

1 m/km = 5.28 ft/mi

 

Table 33. Standard deviations of IRI: Left wheelpath (1998).
Device Standard Deviations of IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
North Atlantic Profiler 0.02 0.02 0.02 0.02
North Central Profiler 0.01 0.03 0.01 0.01
Southern Profiler 0.01 0.03 0.01 0.01
Western Profiler 0.01 0.02 0.03 0.01

1 m/km = 5.28 ft/mi

 

Table 34. Standard deviations of IRI: Right wheelpath (1998).
Device Standard Deviations of IRI (m/km)
Surface Type and Section Number
Asphalt Concrete
1 2 3 4
North Atlantic Profiler 0.01 0.02 0.02 0.02
North Central Profiler 0.01 0.04 0.02 0.00
Southern Profiler 0.01 0.02 0.01 0.01
Western Profiler 0.01 0.02 0.01 0.03

1 m/km = 5.28 ft/mi

  • The evaluation of the bias and precision values showed no distinct trend for a profiler that would indicate that it was different from the other profilers.
  • An ANOVA that was carried out separately for the left- and right-wheelpath IRI values for the four profilers indicated that there were no differences in IRI values obtained for the four profilers in both the left and right wheelpaths.
  • A visual review of the profile data plots indicated excellent repeatability of the profiles for the North Atlantic, North Central, and Western profilers for the following wavebands: (1) 1 to 30 m (3 to 100 ft), (2) less than 10 m (33 ft), (3) 10 to 20 m (33 to 66 ft), and (4) 20 to 30 m (66 to 100 ft). The Southern profiler also exhibited similar results, except for the rough PCC site. At this site, the Southern profiler showed poor repeatability for wavebands between 10 and 20 m (33 and 66 ft), and 20 and 30 m (66 and 100 ft), with the repeatability being poorer for the latter waveband. This may be caused by the condition of the connection between the profiler bar and the vehicle. After completion of the comparison test, the connections between the profiler bar and the profiler vehicle in the Southern profiler were inspected and tightened. Subsequent testing indicated that this profiler was able to collect repeatable profile data on rough sections.

2000 COMPARISON: COLLEGE STATION, TX

The four LTPP T-6600 profilers participated in this comparison. The information presented in this section was obtained from a document written about this comparison.(13)

Five pavement sections were used for profile testing: (1) a smooth AC site, (2) a medium-rough AC site, (3) a chip-seal section, (4) a smooth PCC site, and (5) a rough PCC site. The smooth AC and the medium-rough AC sections were located within the Riverside Campus of Texas A&M University and are used as calibration sections to certify the Texas Department of Transportation (TxDOT) profilers. The IRI values computed from measurements made with an ARRB walking profiler were available for these two sections. However, the walking profiler data for these two sections were not made available by TxDOT because these sites are used for certifying profilers. Measurements made with a reference device (e.g., Dipstick or a walking profiler) were not performed at the other three test sections. Profile testing was conducted at a speed of 80 km/h (50 mi/h). Each profiler was required to obtain five error-free profile runs on a test section.

The IRI values computed from the profiler data and the walking profiler data for the left and right wheelpaths on the tests sections are presented in tables 35 and 36, respectively. The IRI values presented in these tables for the profilers are the average IRI values computed from the IRI for the five repeat runs.

The differences between the average profiler IRI and the Walking Profiler IRI at the test sections are presented in table 37.

The standard deviations for the IRI values for the profilers for the left and right wheelpaths are presented in tables 38 and 39, respectively.

 

Table 35. IRI values along the left wheelpath (2000).
Device IRI (m/km)
Test Section
Smooth AC Medium AC Chip Seal Smooth PCC Rough PCC
North Atlantic Profiler 1.13 1.93 3.00 1.54 2.94
North Central Profiler 0.98 1.81 2.72 1.52 2.68
Southern Profiler 1.01 1.85 3.31 1.48 2.78
Western Profiler 1.04 1.94 3.33 1.56 2.90
Walking Profiler 1.01 1.85 N/A N/A N/A
N/A = Measurements not performed.

1 m/km = 5.28 ft/mi

 

Table 36. IRI values along the right wheelpath (2000).
Device IRI (m/km)
Test Section
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic Profiler 0.72 1.77 2.02 1. 2.95
North Central Profiler 0.79 1.75 2.77 1.68 2.92
Southern Profiler 0.73 1.81 1.98 1.63 2.96
Western Profiler 0.66 1.72 2.11 1.63 2.88
Walking Profiler 0.67 1.75 N/A N/A N/A
N/A = Measurements not performed.

1 m/km = 5.28 ft/mi

 

Table 37. Differences between the profiler IRI and the walking profiler IRI (2000).
Section WheelPath Average Profiler IRI - Reference IRI (m/km)
Profiler
North Atlantic North Central Southern Western
Smooth Asphalt Left 0.12 -0.03 0 0.03
Smooth Asphalt Left 0.08 -0.04 0 0.09
Medium Asphalt Right 0.05 0.12 0.06 -0.01
Medium Asphalt Right 0.02 0 0.06 -0.03

1 m/km = 5.28 ft/mi

 

Table 38. Standard deviations of IRI: Left wheelpath (2000).
Profiler Standard Deviations of IRI (m/km)
Test Section
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic 0.03 0.02 0.10 0.01 0.01
North Central 0.05 0.03 0.06 0.01 0.03
Southern 0.01 0.02 0.07 0.02 0.03
Western 0.01 0.02 0.08 0.01 0.01

1 m/km = 5.28 ft/mi

 

Table 39. Standard deviations of IRI: Right wheelpath (2000).
Profiler Standard Deviations of IRI (m/km)
Test Section
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic 0.01 0.02 0.02 0.01 0.03
North Central 0.06 0.02 0.06 0.01 0.03
Southern 0.02 0.02 0.09 0.02 0.03
0.00 0.01 0.01 0.01 0.01 0.01

1 m/km = 5.28 ft/mi

The following are the main findings of this study:

  • At both the smooth AC and medium AC sections, the difference between the profiler IRI and the walking profiler IRI was within the LTPP-specified bias criterion of ±0.16 m/km (±10 inches/mi) for all of the profilers for both wheelpaths. With the exception of the left wheelpath on the smooth AC section for the North Atlantic profiler and the right wheelpath on the smooth AC section for the North Central profiler, the difference between the profiler IRI and the walking profiler IRI was less than 0.1 m/km (6 inches/mi).
  • At the chip-seal section, the North Central profiler had an IRI value that was 0.74 m/km (47 inches/mi) higher than the average IRI for the other three profilers in the right wheelpath, while for the left wheelpath, it had an IRI value that was 0.49 m/km (31 inches/mi) lower than the average IRI for the other three profilers. After testing was completed, the North Central profiler operator indicated that the sensor covers might not have been taken off when this section was profiled. Under such conditions, only the accelerometer data are used for computing the profile. This can explain why the IRI from the North Central profiler was different from the IRI values obtained by the other three profilers at this site.
  • All profilers showed excellent repeatability of IRI values at the smooth AC, medium AC, smooth PCC, and medium PCC sites. At these four sites, the precision criterion for IRI that is used in the LTPP comparison studies (a precision of less than 0.04 m/km (2.5 inches/mi)) was satisfied by all four profilers along both wheelpaths, except by the North Central profiler at the smooth AC site. At this site, the North Central profiler had IRI precision values of 0.05 and 0.06 m/km (3.2 and 3.8 inches/mi) along the left and right wheelpaths, respectively. The IRI precision at the chip-seal section that had significant cracking was variable, with precision values for a wheelpath for the four profilers ranging from 0.01 to 0.10 m/km (0.6 to 6 inches/mi).
  • An evaluation of the point-to-point profile repeatability values for the left, right, and center paths for the profilers did not show any trends that would suggest that a specific profiler was different when compared to the other profilers.
  • It was noted that at the smooth AC and medium AC sites, the left sensor of the Western profiler was noisy when compared to the other three profilers. This noise was not influencing the IRI values computed from these profiles; however, it may influence the ride indices that weigh shorter wavelengths more heavily.
  • The wheelpaths on the smooth AC and medium AC sections were marked with white paint dots. These marked wheelpaths had a spacing that was slightly greater than the LTPP profiler sensor spacing. Therefore, when profiling the sites, LTPP’s profilers profiled a path that was slightly inside the marked path. Small spikes were noted on some of the profiles collected at this site, and these appear to have been caused when the sensor traversed over the wheelpath markings. However, these spikes did not influence the computed IRI values.
  • A visual examination of the profile data plots showed good agreement in profiles collected by all of the profilers along the left, center, and right wheelpaths, except for two cases: The two cases involved data collected by the North Central profiler along the left and right wheelpaths at the chip-seal section. At this section, the data collected by the North Central profiler along both wheelpaths were different from the data collected by the other profilers. (Note: It is believed that the profiler had the sensor covers on when the profile data were collected at this site.) At the two PCC sites, the North Central profiler had some inconsistencies in the data collected with the center sensor. The North Central region indicated that they were aware of this problem with this sensor and were in the process of correcting it with the help of K.J. Law Engineers.

2003 COMPARISON: MN/ROAD, ALBERTVILLE, MN

The four LTPP ICC profilers were used in this comparison, which was conducted at the Mn/ROAD facility in Albertville, MN. This was the first comparison of four of LTPP's ICC profilers after they began collecting data for the LTPP program in August 2002. One of LTPP’s T-6600 profilers that was used to collect profile data for the LTPP program was still operational, and this profiler was also used in the comparison. The information presented in this section was obtained from a document written about this comparison

Five test sections were used for profile testing: (1) a smooth AC section, (2) a rough AC section, (3) a smooth PCC section, (4) a medium-rough PCC section, and (5) a chip-seal section. Dipstick measurements were obtained along both wheelpaths at all of the test sections. Profile testing was conducted at a speed of 80 km/h (50 mi/h). Each profiler was required to obtain five error-free profile runs on a test section.

The left- and right-wheelpath IRI values computed from the profiler data and the Dipstick data are presented in tables 40 and 41, respectively. The IRI values presented in these tables for the profilers are the average IRI values computed from the IRI for the five repeat runs.

 

Table 40. IRI values along the left wheelpath (2003).
Device IRI (m/km)
Section 1 Section 2 Section 3 Section 4 Section 5
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic:ICC 1.27 2.76 0.92 1.45 2.25
North Central: ICC 1.26 2.75 0.93 1.57 2.15
Southern: ICC 1.29 2.78 0.93 1.45 2.15
Western: ICC 1.28 2.75 0.91 1.43 2.20
K.J. Law Engineers 1.31 2.75 0.94 1.47 2.25
Dipstick 1.17 2.80 0.88 1.32 2.24

1 m/km = 5.28 ft/mi

 

Table 41. IRI values along the right wheelpath (2003).
Device IRI (m/km)
Section 1 Section 2 Section 3 Section 4 Section 5
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic: ICC 1.68 2.81 0.98 1.70 2.54
North Central: ICC 1.73 3.01 1.02 1.72 2.54
Southern: ICC 1.69 2.62 0.96 1.67 2.54
Western: ICC 1.66 2.54 0.97 1.71 2.50
K.J. Law Engineers 1.64 2.46 0.96 1.70 2.44
Dipstick 1.81 2.79 0.99 1.64 2.63

1 m/km = 5.28 ft/mi

 

Table 42. Differences between the profiler IRI and Dipstick IRI: Left wheelpath (2003).
Profiler Average Profiler IRI - Dipstick IRI (m/km)
Section 1 Section 2 Section 3 Section 4 Section 5
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic: ICC 0.10 -0.03 0.04 0.13 0.01
North Central: ICC 0.09 -0.04 0.05 0.25 -0.09
Southern: ICC 0.12 -0.02 0.05 0.13 -0.09
Western: ICC 0.11 -0.05 0.03 0.11 -0.03
K.J. Law Engineers 0.14 -0.05 0.06 0.15 0.01

1 m/km = 5.28 ft/mi

 

Table 43. Differences between the profiler IRI and Dipstick IRI: Right wheelpath (2003).
Profiler Average Profiler IRI - Dipstick IRI (m/km)
Section 1 Section 2 Section 3 Section 4 Section 5
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic: ICC -0.13 0.02 -0.01 0.05 -0.09
North Central: ICC -0.08 0.22 0.03 0.08 -0.09
Southern: ICC -0.12 -0.18 -0.03 0.03 -0.08
Western: ICC -0.15 -0.25 -0.02 0.07 -0.13
K.J. Law Engineers -0.18 -0.33 -0.03 0.05 -0.19

1 m/km = 5.28 ft/mi

The standard deviations of the IRI from the profilers for the left and right wheelpaths are presented in tables 44 and 45, respectively.

 

Table 44. Standard deviations of IRI: Left wheelpath (2003).
Profiler Standard Deviations of IRI (m/km)
Section 1 Section 2 Section 3 Section 4 Section 5
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic: ICC 0.02 0.03 0.01 0.03 0.02
North Central: ICC 0.01 0.04 0.01 0.04 0.02
Southern: ICC 0.02 0.02 0.01 0.03 0.03
Western: ICC 0.01 0.04 0.01 0.01 0.04
K.J. Law Engineers 0.02 0.03 0.01 0.01 0.06

1 m/km = 5.28 ft/mi

 

Table 45. Standard deviations of IRI: Right wheelpath (2003).
Profiler Standard Deviations of IRI (m/km)
Section 1 Section 2 Section 3 Section 4 Section 5
Smooth AC Rough AC Smooth PCC Medium PCC Chip Seal
North Atlantic: ICC 0.07 0.11 0.02 0.02 0.02
North Central: ICC 0.01 0.13 0.01 0.01 0.03
Southern: ICC 0.07 0.08 0.02 0.03 0.04
Western: ICC 0.08 0.09 0.01 0.01 0.03
K.J. Law Engineers 0.04 0.08 0.02 0.02 0.03

1 m/km = 5.28 ft/mi

The main findings of this study are:

  • Overall, all of the profilers appear to be obtaining repeatable IRI values. These data did not indicate that a particular profiler was behaving differently than the other profilers as far as IRI repeatability is concerned. The precision criterion for IRI that is used in the LTPP comparison studies is that the IRI standard deviations from multiple runs on a section should be less than 0.04 m/km (2.5 inches/mi). However, sometimes this criterion cannot be met if distresses are present along the wheelpath, because even a slight shift in the path profiled can have a significant impact on IRI. The IRI precision criterion was met for all of the cases, except for the following: (1) all profilers, except for the North Central profiler along the right wheelpath of the smooth AC section (section 1); (2) all of the profilers along the right wheelpath of the rough AC section (section 2); (3) left wheelpath of the rough AC section (section 2) and medium-rough PCC section (section 4) for the North Central profiler; and (4) left wheelpath of the chip-seal section (section 5) for the K.J. Law Engineers profiler. Distresses were present along the right wheelpath on the smooth AC section (section 1) and along both wheelpaths on the rough AC section (section 2), and variability on the paths profiled by the profilers was the most likely cause of the failure of the profilers to meet the specified criterion at these two sites.
  • Good agreement between profiler IRI and Dipstick IRI was obtained for the majority of the cases. The LTPP criterion used for the comparison studies is that the difference between the Dipstick IRI and the profiler IRI should be within ±0.16 m/km (±10 inches/mi). This criterion was met for all of the cases, except for the following: (1) right wheelpath of the smooth AC section (section 1) and the chip-seal section (section 5) by the K.J. Law Engineers profiler, (2) left wheelpath of the medium-rough PCC section (section 4) by the North Central profiler, and (3) right wheelpath of the rough AC section (section 2) by all profilers, except for the North Atlantic profiler. Extensive distresses were present on the rough AC section (section 2), and the failure of the profilers to meet the specified criterion at this site may be caused by the differences in the way downward features in the pavement are measured by Dipstick and the profilers.
  • An evaluation of the profile data collected by the North Central, North Atlantic, and Western ICC profilers indicated that the profile data collected by these profilers generally have similar repeatability. The K.J. Law Engineers and Southern ICC profilers showed much higher variability in the profile data for repeat runs along both wheelpaths when compared to the other three ICC profilers. A comparison of the profiles obtained by the four ICC profilers on the five test sections indicated that all four of the profilers are capturing similar profile features. A profile feature that appeared in any ICC profiler was also present on the profiles collected by the other ICC profilers. After the profiler comparison, another evaluation was performed to investigate the repeatability of the profile data collected by the Southern profiler. This investigation indicated that the Southern profiler was obtaining repeatable profile data that was comparable to the data obtained by the other three ICC profilers on the Mn/ROAD test sections. The poor profile repeatability obtained by the Southern profiler on the Mn/ROAD test sections may have been caused by problems with the operational procedures that were followed by the profiler operator (e.g., insufficient lead-in, not maintaining a constant speed, etc.).
  • Collection of profile data at speeds of 35, 50, 65, 80, 95, and 110 km/h (22, 38, 41, 50, 59, and 69 mi/h) was performed by the K.J. Law Engineers and Southern ICC profilers on one section. The analysis of the data indicated that the IRI value did not appear to be influenced by the speed of testing.
  • There were differences in IRI values computed for the Dipstick data using ProQual and RoadRuf. RoadRuf is a software program developed by UMTRI. However, a comparison of IRI values obtained by ProQual and RoadRuf for profiler data showed that the IRI values were similar. For the Dipstick data collected along 10 wheelpaths (from 5 sections), the differences in the IRI values from ProQual and RoadRuf ranged from 0.004 to 0.079 m/km (0.25 to 5 inches/mi), with the IRI values from ProQual being higher than those obtained by RoadRuf for all of the cases. When ProQual computes the IRI from the Dipstick data, it first applies a filter that has an upper-wavelength cutoff of 100 m (328 ft), then it uses the filtered data to compute the IRI value. When RoadRuf computes the IRI from the Dipstick data, the software uses the Dipstick elevation profile to compute the IRI without any prefiltering of the data. The filtering of the Dipstick data performed by ProQual may be the cause of the differences in the Dipstick IRI values between RoadRuf and ProQual.
Previous | Table of Contents | Next

 

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101