Quantification of Smoothness Index Differences Related To
Long-Term Pavement Performance Equipment Type
Appendix B: Profiler Verification Studies
1996 VERIFICATION TEST
In 1996, FHWA purchased four K.J. Law Engineers T-6600 profilers
to replace the K.J. Law Engineers DNC 690 profilers that were used
in the LTPP program. Each region compared the DNC 690 and T-6600
profilers before using the T-6600 profiler to collect profile data
for the LTPP program.
Each region used four test sections for the verification test;
each test section was 152.4 m (500 ft) long. Two sections were
surfaced with AC, while the other two were jointed PCC pavements.
When selecting the test sections, the aim was to select two
sections for each pavement type so that one section will have an
IRI value of less than 1.6 m/km (101 inches/mi), and the other
section will have an IRI value of greater than 2.2 m/km (139
inches/mi). Dipstick measurements were performed on both wheelpaths
at all four test sections. Each profiler performed three
measurement sequences at each test section, with one measurement
sequence being performed on 1 day. A measurement sequence consisted
of obtaining a set of profile runs (a minimum of five good profile
runs) on a test section at test speeds of 64 and 80 km/h (40 and 50
mi/h). Therefore, at each section, each profiler collected six data
sets (two speeds x three sequences). During testing, each profiler
collected 24 data sets (4 test sections x 2 test speeds x 3
measurement sequences). The wheelpaths at the test sections were
not marked, and the profiler drivers judged the location of the
wheelpath when profiling the sections.
The guidelines for testing recommended that concurrent
measurements be made at the test sections with the two profilers
with a minimum time lag between measurements. The guidelines also
recommended that an effort should be made to obtain the three
profiler measurement sequences for a test section on 3 consecutive
days. Each RSC was asked to compute the IRI value of the profiler
data and Dipstick data using ProQual. The bias in the IRI value of
a profiler for a wheelpath on a test section was determined by
computing the difference between the average profiler IRI at a
specific speed for a data set (computed by averaging the IRI values
obtained from replicate runs) and the Dipstick IRI. The profiler
was deemed to have satisfied the IRI bias criterion if this
difference in IRI was within ±0.16 m/km (±10 inches/mi). For each
data set, the standard deviation of the IRI for each wheelpath was
computed by using the IRI values obtained for the replicate profile
runs. The precision of the profiler was evaluated by using this
standard deviation of the IRI. The profiler was deemed to have
satisfied the precision criterion if the standard deviation of the
IRI was less than 0.04 m/km (2.5 inches/mi).
Each RSC prepared a report documenting the results of this
comparison test. (See references 24, 25, 26, and 27.) A summary of
the findings contained in these reports is presented separately for
the four regions.
North Atlantic Region
For all profiler testing, the same driver operated each
profiler. A review of the report indicated that, in many cases,
more than five repeat runs were conducted for each data set.
Thereafter, five runs were selected for computing the average IRI
and the standard deviation of the IRI. It appears that the five
runs were selected after the IRI values from all available runs
were reviewed to select values that were similar to each other. The
average profiler IRI values obtained for the 80-km/h (50-mi/h)
profiler runs and IRI obtained from the Dipstick measurements at
the four sites that were extracted from the report are presented in
table 46. The differences between the average profiler IRI and the
Dipstick IRI for the profile runs obtained at 80 km/h (50 mi/h) are
presented in table 47. It should be noted that testing resulted in
six data sets at each section, and these tables only show the
results obtained for 80-km/h (50-mi/h) testing, which comprise
three data sets from a test section.
Table 46. Profiler IRI for 80-km/h (50-mi/h) runs and
Dipstick IRI: North Atlantic region.
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
Dipstick |
T-6600 |
DNC 690 |
Dipstick |
Smooth AC |
1 |
10/07/96 |
0.78 |
0.86 |
0.83 |
0.89 |
0.87 |
0.92 |
2 |
10/09/96 |
0.74 |
0.86 |
|
0.88 |
0.86 |
|
3 |
10/10/96 |
0.75 |
0.87 |
0.87 |
0.86 |
Rough AC |
1 |
10/07/96 |
2.44 |
2.26 |
2.30 |
1.92 |
1.89 |
1.91 |
2 |
10/09/96 |
2.11 |
2.27 |
|
1.67 |
1.92 |
|
3 |
10/10/96 |
2.28 |
2.28 |
1.72 |
1.87 |
Smooth PCC |
1 |
10/03/96 |
1.19 |
1.25 |
1.27 |
1.37 |
1.30 |
1.39 |
2 |
10/04/96 |
1.18 |
1.22 |
|
1.34 |
1.31 |
|
3 |
10/04/96 |
1.20 |
1.21 |
1.40 |
1.33 |
Rough PCC |
1 |
10/07/96 |
1.92 |
1.81 |
1.93 |
2.11 |
2.12 |
2.22 |
2 |
10/09/96 |
1.88 |
1.89 |
|
2.23 |
2.15 |
|
3 |
10/10/96 |
1.88 |
1.88 |
2.18 |
2.18 |
Note: Only one data set was obtained for Dipstick
at each site. |
1 m/km = 5.28 ft/mi
Table 47. Differences between the profiler IRI and the
Dipstick IRI: North Atlantic region.
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
T-6600 |
DNC 690 |
Smooth AC |
1 |
10/07/96 |
-0.05 |
0.03 |
-0.03 |
-0.05 |
2 |
10/07/96 |
-0.09 |
0.03 |
-0.05 |
-0.06 |
3 |
10/10/96 |
-0.08 |
0.04 |
-0.05 |
-0.06 |
Rough AC |
1 |
10/07/96 |
0.14 |
-0.04 |
0.01 |
-0.02 |
2 |
10/07/96 |
-0.20 |
-0.04 |
-0.24 |
0.00 |
3 |
10/10/96 |
-0.02 |
0.02 |
-0.20 |
-0.05 |
Smooth PCC |
1 |
10/03/96 |
0.08 |
-0.02 |
-0.03 |
-0.09 |
2 |
10/04/96 |
-0.09 |
-0.05 |
-0.07 |
0.04 |
3 |
10/04/96 |
-0.07 |
0.06 |
-0.08 |
-0.06 |
Rough PCC |
1 |
10/07/96 |
0.08 |
0.20 |
-0.01 |
0.05 |
2 |
10/09/96 |
0.11 |
0.15 |
-0.06 |
-0.02 |
3 |
10/10/96 |
0.08 |
0.19 |
0.01 |
0.01 |
1 m/km = 5.28 ft/mi
The findings from the study are:
- Both profilers passed the IRI bias criterion for all data sets,
except for the followin cases: (1) at the rough AC site, the T-6600
profiler failed the criterion for two both the left and right
wheelpaths, and (2) the DNC 690 profiler failed the criterion at
the rough PCC site along the right wheelpath for one data set.
- Both profilers passed the precision criterion for all data
sets, except for the following cases: (1) at the rough AC site, the
DNC 690 profiler failed the criterion for two data sets in the
right wheelpath; the T-6600 profiler failed the criterion along the
left wheelpath for four data sets and along the right wheelpath for
three data sets, and (2) at the rough PCC site, the DNC 690
profiler failed the criterion for two data sets in the right
wheelpath; the T-6600 profiler failed the criterion along the right
wheelpath for three data sets.
- For both profilers, it was found that the IRI values for both
the left and right wheelpaths were insensitive to the two testing
speeds. This held true for both smooth and rough pavements, as well
as for AC and PCC surfaces.
- For most cases, the IRI values obtained by the profilers were
less than those obtained by Dipstick
- The IRI values varied much more for the rough sections when
compared to the smooth sections. The majority of these occurrences
involved the T-6600 profiler. Overall, the IRI values obtained from
the DNC 690 profiler appeared to be more repeatable than those
obtained from the T-6600 profiler. This might have been the result
of the T-6600 profiler being operated by an inexperienced driver
who was not as skilled at consistently tracking the wheelpath as
the driver of the DNC 690 profiler.
North Central Region
For all profiler testing, the same drivers operated each
profiler. The drivers of both profilers used a camera system to
help them align the vehicle along the right wheelpath. The DNC 690
and the T-6600 profilers had different sensor spacings (1,676 and
1,422 mm (66 and 56 inches), respectively).
The report indicated that the first five acceptable runs
obtained at a section for each data set were used to compute the
average IRI and standard deviation of the IRI for each wheelpath.
The average profiler IRI values obtained for the 80-km/h (50-mi/h)
profiler runs and IRI obtained from the Dipstick measurements at
the four sites that were extracted from the report are presented in
table 48. The differences between the average profiler IRI and the
Dipstick IRI for the profiler runs obtained at 80 km/h (50 mi/h)
are presented in table 49. It should be noted that testing resulted
in six data sets at each section, and these tables only show the
results obtained for the 80-km/h (50-mi/h) testing, which comprise
three data sets from a test section.
Table 48. Profiler IRI for 80-km/h (50-mi/h) runs and
Dipstick IRI: North Central region
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
Dipstick |
T-6600 |
DNC 690 |
Dipstick |
Smooth AC |
1 |
10/07/96 |
1.07 |
1.00 |
0.98 |
1.12 |
1.05 |
1.08 |
2 |
10/09/96 |
1.05 |
1.01 |
|
1.05 |
1.07 |
|
3 |
10/10/96 |
1.02 |
1.00 |
1.08 |
1.05 |
Rough AC |
1 |
10/07/96 |
3.94 |
3.95 |
3.82 |
4.73 |
4.98 |
4.86 |
2 |
10/09/96 |
3.96 |
3.97 |
|
4.78 |
4.92 |
|
3 |
10/10/96 |
3.94 |
3.95 |
4.79 |
4.90 |
Smooth PCC |
1 |
10/03/96 |
1.11 |
1.11 |
1.05 |
1.07 |
1.09 |
1.15 |
2 |
10/04/96 |
1.11 |
1.10 |
|
1.08 |
1.07 |
|
3 |
10/04/96 |
1.11 |
1.11 |
1.08 |
1.09 |
Rough PCC |
1 |
10/07/96 |
2.64 |
2.75 |
2.55 |
3.00 |
3.06 |
3.01 |
2 |
10/09/96 |
2.67 |
2.70 |
|
2.95 |
2.99 |
|
3 |
10/10/96 |
2.63 |
2.74 |
3.02 |
3.02 |
Note: Only one data set was obtained for Dipstick
at each site. |
1 m/km = 5.28 ft/mi
Table 49. Differences between the profiler IRI and the
Dipstick IRI: North Central region
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
T-6600 |
DNC 690 |
Smooth AC |
1 |
10/07/96 |
0.09 |
0.03 |
0.04 |
-0.03 |
2 |
10/07/96 |
0.07 |
0.04 |
-0.04 |
-0.01 |
3 |
10/10/96 |
0.05 |
0.02 |
0.00 |
-0.03 |
Rough AC |
1 |
10/07/96 |
0.12 |
0.13 |
-0.12 |
-0.12 |
2 |
10/07/96 |
0.15 |
0.15 |
-0.07 |
0.07 |
3 |
10/10/96 |
0.12 |
0.14 |
-0.07 |
-0.04 |
Smooth PCC |
1 |
10/03/96 |
0.06 |
0.05 |
-0.08 |
-0.06 |
2 |
10/04/96 |
0.06 |
0.05 |
-0.08 |
0.04 |
3 |
10/04/96 |
0.06 |
0.04 |
-0.07 |
-0.06 |
Rough PCC |
1 |
10/07/96 |
0.08 |
0.20 |
-0.01 |
0.05 |
2 |
10/09/96 |
0.11 |
0.15 |
-0.06 |
-0.02 |
3 |
10/10/96 |
0.08 |
0.19 |
0.01 |
0.01 |
1 m/km = 5.28 ft/mi
The findings from the study are:
- Both profilers satisfied the IRI bias criterion along both
wheelpaths on the smooth AC and smooth PCC sections. On the rough
AC section, the DNC 690 profiler passed the bias criterion for both
wheelpaths for all of the cases, while the T-6600 profiler failed
the bias criterion along the left wheelpath for all data sets. On
the rough PCC section, the DNC 690 profiler failed the bias
criterion for four data sets along the left wheelpath, while the
T-6600 profiler passed the bias criterion for all of the cases
along both wheelpaths.
- Both profilers satisfied the IRI precision criterion along both
wheelpaths on the smooth AC and smooth PCC sections. On the rough
AC section, the T-6600 profiler failed the precision criterion
along the left wheelpath for three data sets, and along the right
wheelpath for one data set, while the DNC 690 profiler failed the
criterion along the right wheelpath for two data sets. On the rough
PCC section, the DNC 690 and the T-6600 profilers failed the
precision criterion for one data set along the left wheelpath and
one data set along the right wheelpath
Southern Region
A review of the report showed that the average IRI and standard
deviation of the IRI along each wheelpath for each data set were
computed using all available runs for that data set. The average
profiler IRI values obtained for the 80-km/h (50-mi/h) profiler
runs and IRI obtained from the Dipstick measurements at the four
sites that were extracted from the report are presented in table
50. The differences between the average profiler IRI and the
Dipstick IRI for the profile runs obtained at 80 km/h (50 mi/h) are
presented in table 51. Testing resulted in six data sets at each
section; these tables only show the results obtained for 80 km/h
(50 mi/h) testing, which comprise three data sets from a test
section.
Table 50. Profiler IRI for 80-km/h (50-mi/h) runs and
Dipstick IRI: Southern region.
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
Dipstick |
T-6600 |
DNC 690 |
Dipstick |
Smooth AC |
1 |
11/06/96 |
0.72 |
0.73 |
0.70 |
0.74 |
0.73 |
0.75 |
2 |
11/08/96 |
0.71 |
0.74 |
|
0.74 |
0.73 |
|
3 |
11/08/96 |
0.70 |
0.73 |
0.73 |
0.73 |
Rough AC |
1 |
11/04/96 |
1.88 |
1.66 |
1.84 |
1.87 |
2.12 |
2.18 |
2 |
11/04/96 |
1.91 |
1.63 |
|
1.90 |
2.12 |
|
3 |
11/04/96 |
1.96 |
1.61 |
1.92 |
2.12 |
Smooth PCC |
1 |
10/30/96 |
1.76 |
1.79 |
1.67 |
1.78 |
1.69 |
1.77 |
2 |
10/31/96 |
1.74 |
1.77 |
|
1.76 |
1.69 |
|
3 |
11/01/96 |
1.83 |
1.81 |
1.75 |
1.69 |
Rough PCC |
1 |
10/30/96 |
2.05 |
2.15 |
2.28 |
2.48 |
2.09 |
2.35 |
2 |
10/31/96 |
2.02 |
2.16 |
|
2.50 |
2.09 |
|
3 |
11/01/96 |
2.12 |
2.18 |
2.52 |
2.09 |
Note: Only one data set was obtained for Dipstick
at each site. |
1 m/km = 5.28 ft/mi
Table 51. Differences between the profiler IRI and the
Dipstick IRI: Southern region
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
T-6600 |
DNC 690 |
Smooth AC |
1 |
11/06/96 |
-0.01 |
0.00 |
0.05 |
-0.01 |
2 |
11/08/96 |
-0.02 |
0.01 |
-0.08 |
-0.01 |
3 |
11/08/96 |
-0.03 |
0.00 |
-0.08 |
-0.02 |
Rough AC |
1 |
11/04/96 |
-0.24 |
-0.45 |
-0.33 |
-0.31 |
2 |
11/04/96 |
-0.21 |
-0.49 |
-0.38 |
-0.28 |
3 |
11/04/96 |
-0.15 |
-0.51 |
-0.37 |
-0.26 |
Smooth PCC |
1 |
10/30/96 |
0.07 |
0.10 |
-0.10 |
0.01 |
2 |
10/31/96 |
0.05 |
0.08 |
-0.07 |
0.01 |
3 |
11/01/96 |
0.14 |
0.11 |
-0.08 |
-0.02 |
Rough PCC |
1 |
10/30/96 |
-0.04 |
0.06 |
-0.08 |
0.13 |
2 |
10/31/96 |
-0.07 |
0.07 |
0.02 |
0.15 |
3 |
11/01/96 |
0.03 |
0.10 |
0.04 |
0.17 |
1 m/km = 5.28 ft/mi
The findings from the main study are:
- Both profilers passed the IRI bias criterion at the smooth AC
site. At the rough AC site, both profilers failed the IRI bias
criterion for all of the cases, except where the T-6600 profiler
met the criterion along the left wheelpath for one data set. At the
smooth PCC site, the DNC 690 profiler passed the bias criterion for
all data sets, while the T-6600 profiler passed the bias criterion
for all data sets along the right wheelpath, but failed the bias
criterion along the left wheelpath for three data sets. At the
rough PCC site, the DNC 690 profiler passed the bias criterion for
all data sets, except for three data sets along the right
wheelpath; the T-6600 profiler passed the bias criterion for all
data sets, except for two data sets along the right wheelpath.
- Both profilers passed the precision criterion at the smooth AC
site. At the rough AC site, the DNC 690 profiler met the precision
criterion for all data sets, while the T-6600 profiler failed the
precision criterion along the left wheelpath for all data sets and
along the right wheelpath for two data sets. At the smooth PCC
site, the DNC 690 profiler met the precision criterion for all data
sets; the T-6600 profiler passed the criterion for all data sets,
except for three data sets along the left wheelpath. At the rough
PCC site, both profilers met the precision criterion for all data
sets, except along the left wheelpath where the T-6600 profiler
failed the criterion for four data sets.
- In most cases where the bias or the precision criteria were not
met, testing was performed at 56 km/h (35 mi/h). The drivers
appeared to have difficulty running the same wheelpath consistently
at the slower test speed of 56 km/h (35 mi/h).
Western Region
For all profiler testing, each profiler was operated by the same
driver. A review of the report indicated that nine profile runs
usually were conducted for each data set. When computing the IRI
bias for a particular wheelpath for a test speed, the average
profiler IRI was computed by averaging the IRI values that were
obtained for all profile runs for the three data sets and then
subtracting the Dipstick IRI from this value. When computing the
standard deviation of the IRI at a site for a particular wheelpath
and a specific test speed, all profile runs obtained for the three
test sequences were considered.
The average profiler IRI values obtained for the 80-km/h
(50-mi/h) profiler runs and IRI obtained from the Dipstick
measurements for the four sites that were extracted from the report
are presented in table 52. The differences between the average
profiler IRI obtained for the 80 km/h (50-mi/h) runs and the
Dipstick IRI for each test sequence are presented in table 53. The
values presented in table 53 were computed from the values shown in
table 52. The report on this comparison did not give the values
shown in table 53, since the differences between profiler IRI and
Dipstick IRI were computed differently, as described
previously.
Table 52. Profiler IRI for 80-km/h (50-mi/h) runs and
Dipstick IRI: Western region.
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
Dipstick |
T-6600 |
DNC 690 |
Dipstick |
Smooth AC |
1 |
10/10/1996 |
0.78 |
0.79 |
0.86 |
0.94 |
0.92 |
0.97 |
2 |
10/10/1996 |
0.78 |
1.15 |
|
0.93 |
0.91 |
|
3 |
10/10/1996 |
0.75 |
0.80 |
0.91 |
0.93 |
Rough AC |
1 |
10/15/1996 |
2.63 |
3.42 |
1.96 |
2.58 |
2.55 |
2.10 |
2 |
10/16/1996 |
2.53 |
3.50 |
|
2.52 |
2.63 |
|
3 |
10/17/1996 |
2.60 |
3.13 |
2.67 |
2.56 |
Smooth PCC |
1 |
10/11/1996 |
1.10 |
1.18 |
1.16 |
1.04 |
1.08 |
1.04 |
2 |
10/11/1996 |
0.97 |
0.95 |
|
0.87 |
0.90 |
|
3 |
10/16/1996 |
1.83 |
1.81 |
1.75 |
1.69 |
Rough PCC |
1 |
10/10/1996 |
2.24 |
2.33 |
1.99 |
2.46 |
2.41 |
2.12 |
2 |
10/11/1996 |
2.30 |
2.32 |
|
2.37 |
2.40 |
|
3 |
10/16/1996 |
2.58 |
2.55 |
2.60 |
2.70 |
Note: Only one data set was obtained for Dipstick
at each site. |
1 On smooth PCC data collected with T-6600 on
10/10/96.
1 m/km = 5.28 ft/mi
The findings from the main study are:
- Very high IRI bias values that were outside of the acceptable
range were observed for both profilers at the rough AC and rough
PCC sites. The report indicated that the high bias values were
probably caused by the way Dipstick measurements were performed.
The report also indicated that the documentation for the Dipstick
measurement procedures being used at that time included the
following paragraph: " The footpads should be placed to avoid minor
localized cracks, holes, open joints, the edge of open joints or
wide cracks, and loose stones or debris." The report indicated that
this procedure was followed during Dipstick measurements. However,
when profilers collect data, they do not avoid these features, so
there were high bias values for the profilers at the rough AC and
rough PCC sites.
- High IRI precision values that did not meet the specified
criterion were obtained in many cases.
- After reviewing the report from this research project, the
precision criterion failure at many of the sections is attributed
to the procedure that was used for computing the precision. The
other regions computed the precision for each data set. In the
analysis performed by the Western region, all of the profiler runs
obtained for all three sequences were used to compute the
precision.
Table 53. Differences between the profiler IRI and the
Dipstick IRI: Western region.
Site |
Sequence |
Test Date |
Average IRI (m/km) |
Left Wheelpath |
Right Wheelpath |
T-6600 |
DNC 690 |
T-6600 |
DNC 690 |
Smooth AC |
1 |
10/10/96 |
-0.08 |
-0.07 |
0.04 |
-0.05 |
2 |
10/10/96 |
-0.08 |
0.29 |
-0.04 |
-0.06 |
3 |
10/10/96 |
-0.01 |
-0.06 |
-0.06 |
-0.04 |
Rough AC |
1 |
10/15/96 |
0.66 |
1.46 |
0.48 |
0.45 |
2 |
10/16/96 |
0.56 |
1.54 |
0.42 |
0.53 |
3 |
10/17/96 |
0.63 |
1.16 |
0.57 |
0.46 |
Smooth PCC |
1 |
10/11/96 |
-0.06 |
0.01 |
0.00 |
0.03 |
2 |
10/11/96 |
-0.18 |
-0.13 |
-0.16 |
0.11 |
3 |
11/16/96 |
-0.20 |
-0.21 |
-0.18 |
-0.14 |
Rough PCC |
1 |
10/10/96 |
0.25 |
0.35 |
0.34 |
0.29 |
2 |
10/11/96 |
0.31 |
0.33 |
0.34 |
0.29 |
3 |
11/16/96 |
0.59 |
0.57 |
0.48 |
0.58 |
1 m/km = 5.28 ft/mi
Overall Comment on the Results
The procedures used by the four regions for computing the
average IRI and the standard deviation of the IRI were different.
The following procedures are used by each region in computing the
average IRI and the standard deviation of the IRI:
- North Atlantic Region: Five runs from each
data set were used in the computations; however, these runs
appeared to have been selected based on the review of the IRI
values obtained from all of the replicate runs.
- North Central Region: The first five
error-free runs for each data set were used in the
computations.
- Southern Region: All available replicate runs
for each data set were used in the computations.
- Western Region: At each site, for a particular
test speed, all runs that were collected at the site for all three
test sequences were used in the computations.
Because of the different procedures used by the four regions in
computing IRI bias and IRI precision, the results from the regions
cannot be compared.
Overall, it appears that the T-6600 profilers are performing
satisfactorily and, generally, data collected by the T-6600
profilers appear to be similar to the data collected by the DNC 690
profilers from an IRI viewpoint.
2002 VERIFICATION STUDY
In 2002, FHWA purchased four ICC profilers to replace the K.J.
Law Engineers T-6600 profilers. Each RSC compared these two
profilers before using the ICC profilers for data collection. A
minimum of five test sites were used in each region for this
comparison, and each region was asked to select at least one site
meeting the following requirements: (1) smooth AC section with an
IRI less than 1.6 m/km (101 inches/mi), (2) smooth PCC section with
an IRI less than 1.6 m/km (101 inches/mi), (3) rough AC section
with an IRI greater than 2.2 m/km (139 inches/mi), (4) rough PCC
section with an IRI greater than 2.2 m/km (139 inches/mi), and (5)
a chip-seal section. Dipstick measurements were not obtained at the
test sites. The wheelpaths at the test sites were not marked, and
the profiler operators judged the location of the wheelpath when
profiling the test sections.
The purpose of this profiler comparison was to compare the IRI
values and the profiles obtained by the two profilers. Each region
collected data at the test sections following normal LTPP data
collection procedures, except that the data collection at the PCC
sections was performed in the afternoon. Each region was asked to
submit IRI values for five error-free profile runs on each section
for each profiler. At many of the sites, the two profilers
collected measurements on the same day. However, at several sites,
there was a time difference of up to 1.5 months between the
measurements collected by the two profilers. Each region submitted
the results of the comparison to the LTPP technical support
services contractor, who prepared a report documenting the results
obtained from all four regions.(28) Table 54 shows the IRI values
that were obtained from the testing. This table shows the region,
site number, surface type, dates when the sites were profiled,
average left- and right-wheelpath IRI from each profiler, and the
standard deviation of the IRI obtained from each profiler along the
left and right wheelpaths.
The main findings of this study are:
- Overall, the mean IRI values (average IRI for left- and
right-wheelpath IRI) obtained by the two profilers showed good
agreement. The difference in the mean IRI between the two profilers
at the test sites was within ±0.05 m/km (±3.1 inches/mi) for the
majority of the sites.
- When individual wheelpath IRI values were compared between the
two profilers, the differences were higher in magnitude than the
differences observed when the mean IRI values were compared. When
individual wheelpath IRI values at the test sites were compared,
for 70 percent of the cases, the differences in the IRI were within
±0.10 m/km (±6 inches/mi). An evaluation of the results from the
LTPP profiler comparison study that was conducted in 2000 in Texas
indicated that the differences in the IRI that were observed
between the K.J. Law Engineers and ICC profilers were comparable to
the differences in the IRI that were observed among the four LTPP
K.J. Law Engineers profilers.(14) This indicates that the IRI
values obtained from the data collected with the ICC profilers show
reasonable agreement with the IRI obtained from the data collected
by the K.J. Law Engineers profilers.
- The comparison of the profile plots indicated that there are
differences in long wavelengths in the profile data recorded by the
K.J. Law Engineers and ICC profilers at many of the sites. Visual
reviews of the profiles showed that similar pavement features were
being recorded by both profilers.
Table 54. IRI values obtained from the 2002 verification
study.
Region |
Site Number |
Surface Type |
Profile Date |
Average IRI Left WheelPath(m/km) |
Average IRI Right WheelPath(m/km) |
Std.Dev IRI Left WheelPath(m/km) |
Std.Dev IRI Right WheelPath(m/km) |
K.J. Law |
ICC |
Law |
ICC |
Law |
ICC |
Law |
ICC |
Law |
ICC |
N.Atlantic |
251002 |
AC |
7/25/2002 |
7/25/2002 |
4.22 |
4.45 |
1.48 |
1.30 |
0.22 |
0.28 |
0.06 |
0.14 |
N.Atlantic |
361011 |
AC |
7/24/2002 |
7/24/2002 |
0.91 |
0.89 |
0.87 |
0.89 |
0.03 |
0.04 |
0.01 |
0.03 |
N.Atlantic |
364018 |
PCC |
7/25/2002 |
7/25/2002 |
2.64 |
2.87 |
2.22 |
2.18 |
0.09 |
0.16 |
0.06 |
0.05 |
N.Atlantic |
245807 |
PCC |
6/05/2002 |
8/01/2002 |
1.56 |
1.41 |
1.58 |
1.53 |
0.07 |
0.02 |
0.02 |
0.02 |
N.Atlantic |
872811 |
AC |
4/20/2001 |
7/31/2002 |
1.50 |
1.43 |
1.57 |
1.65 |
0.21 |
0.02 |
0.07 |
0.04 |
N.Atlantic |
360801 |
AC |
7/23/2002 |
7/23/2002 |
1.11 |
1.08 |
1.09 |
1.06 |
0.01 |
0.02 |
0.02 |
0.02 |
N.Atlantic |
360802 |
AC |
7/23/2002 |
7/23/2002 |
1.38 |
1.38 |
1.77 |
1.70 |
0.01 |
0.03 |
0.01 |
0.04 |
N.Atlantic |
360859 |
AC |
7/23/2002 |
7/23/2002 |
1.06 |
1.06 |
1.15 |
1.13 |
0.01 |
0.01 |
0.02 |
0.02 |
N.Central |
17A001 |
AC |
7/16/2002 |
7/16/2002 |
1.01 |
0.95 |
1.22 |
1.18 |
0.05 |
0.02 |
0.06 |
0.02 |
N.Central |
17A002 |
AC |
7/16/2002 |
7/16/2002 |
2.68 |
2.78 |
2.83 |
2.69 |
0.04 |
0.02 |
0.03 |
0.03 |
N.Central |
17A003 |
PCC |
7/17/2002 |
7/17/2002 |
1.06 |
1.08 |
1.15 |
1.16 |
0.02 |
0.02 |
0.01 |
0.02 |
N.Central |
17A004 |
PCC |
7/17/2002 |
7/17/2002 |
4.01 |
4.07 |
4.14 |
4.16 |
0.02 |
0.02 |
0.02 |
0.02 |
N.Central |
17A005 |
CS |
7/16/2002 |
7/16/2002 |
3.20 |
2.95 |
3.65 |
3.76 |
0.08 |
0.08 |
0.03 |
0.01 |
Southern |
481064 |
AC |
7/24/2002 |
9/10/2002 |
1.89 |
1.75 |
1.75 |
1.75 |
0.04 |
0.05 |
0.12 |
0.13 |
Southern |
481070 |
AC |
7/24/2002 |
9/10/2002 |
1.58 |
1.73 |
1.75 |
1.82 |
0.08 |
0.05 |
0.11 |
0.10 |
Southern |
48B350 |
CS |
7/24/2002 |
9/10/2002 |
1.84 |
2.32 |
1.75 |
2.45 |
0.16 |
0.13 |
0.10 |
0.25 |
Southern |
483003 |
PCC |
7/24/2002 |
9/11/2002 |
2.11 |
2.31 |
1.75 |
2.26 |
0.02 |
0.06 |
0.07 |
0.04 |
Southern |
485253 |
PCC |
7/24/2002 |
9/11/2002 |
1.29 |
1.67 |
1.75 |
1.56 |
0.01 |
0.02 |
0.02 |
0.02 |
Western |
320110 |
AC |
6/10/2002 |
7/31/2002 |
0.57 |
0.57 |
0.74 |
0.74 |
0.01 |
0.01 |
0.01 |
0.01 |
Western |
320209 |
PCC |
6/10/2002 |
7/31/2002 |
1.03 |
1.21 |
0.96 |
1.11 |
0.03 |
0.03 |
0.02 |
0.03 |
Western |
67454 |
AC |
7/26/2002 |
7/26/2002 |
2.24 |
2.27 |
2.32 |
2.19 |
0.02 |
0.03 |
0.04 |
0.06 |
Western |
69107 |
PCC |
7/25/2002 |
7/25/2002 |
2.58 |
2.50 |
2.29 |
2.38 |
0.03 |
0.07 |
0.03 |
0.08 |
Western |
169034 |
CS |
6/04/2002 |
7/29/2002 |
1.81 |
1.74 |
2.02 |
1.98 |
0.03 |
0.05 |
0.08 |
0.09 |
1 m/km = 5.28 ft/mi