U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-04-032
Key Findings From LTPP Analysis 2000-2003
Pavement projects start out with a given set of site conditions,
including traffic, climate, and subgrade/foundation. LTPP data analyses have shown that each of these site conditions
affects pavement performance. While these conditions cannot be controlled, they always should be considered. In critical
situations, the pavement design features should be selected to mitigate the adverse effects of site conditions on performance.
The following key findings from several LTPP site conditions analyses are grouped into three
areas: traffic, climate, and subgrade/foundation.
- Report No. FHWA-RD-00-054
Information on cumulative truck axle loads plays an important role in pavement design and performance analysis. It is especially
crucial for mechanistic design methods and load-related distress prediction models. A comprehensive traffic load spectra projection
methodology was developed using a corridor-assignment model and evaluated at 12 LTPP test sections. Initial results
indicate that the proposed methodology can provide feasible traffic load projections.
- Report No. FHWA-RD-00-054
The new traffic load projection methodology provides a way to
predict annual axle load spectra. These are the frequency distributions
of axle weight of a given axle type into weight ranges for:
The cumulative axle load spectra can be obtained by summing
the annual axle load spectra to the year of interest.
- All in-service years of a roadway segment.
- Single, tandem, and tridem axles.
- All trucks combined (Federal Highway Administration
(FHWA) vehicle classes 4 through 13).
- Report No. FHWA-RD-03-094
Annual axle load spectra projected by the traffic load projection
methodology for all in-service years up to 1998 were concluded to
be reasonable (i.e., falling into expected ranges) for the majority
(558 or 63 percent) of the 890 LTPP traffic sites. The traffic load
spectra projected for the remaining 332 (37 percent) of traffic sites
were considered unreliable because of inadequate or missing data
collected at those sites.
- Report No. FHWA-RD-03-094
The LTPP Pavement Loading Guide (PLG) was developed to
overcome the difficulty of estimating traffic loads for the remaining
332 (37 percent) of the 890 LTPP sites. The document contains
guidelines for the development of the PLG along with two
examples using the PLG to obtain traffic load projections for LTPP
sites without site-specific truck class and/or axle load data.
- Report No. NCHRP 20-50(5)
In pavement design, the vehicle class distribution and the axle load
spectrum cannot be assumed using a default or single load distribution
for either the roadway functional class or a region.
- Report No. NCHRP 20-50(5)
To make predictions that can be used with confidence, research
quality traffic survey data of at least 5 years is recommended,
which should include accurately measured vehicle classes,
number of axle loads, and load configurations for a given roadway
- General Traffic Pattern Findings
Report No. FHWA-RD-03-094
Based on the 558 LTPP traffic sites with reasonable axle load
projection results, general traffic pattern findings obtained are
summarized as follows:
- The leading 4 traffic load contributors are 5-axle single
trailer trucks (FHWA vehicle class 9), 2-axle-6-tire single
unit trucks (class 5), 3-axle single unit trucks (class 6), and
4-axle or fewer single trailer trucks (class 8). These four
vehicle classes comprise 90 percent of the vehicles projected.
- The projected percentages of all vehicles contributed by
5-axle single trailer trucks and 2-axle-6-tire single unit
trucks are listed, respectively, for four roadway functional
- For urban principal interstates: 50 percent and 25 percent, respectively.
- For urban principal freeways and expressways: 45 percent and 20 percent, respectively.
- For rural principal interstates: 65 percent and 10 percent, respectively.
- For rural principal freeways and expressways: 50 percent and 20 percent, respectively.
- The minimum average daily traffic truck volume ranges from
30 trucks per day for a site located on a rural minor arterial
highway to 6,310 trucks per day on a site located on an
urban interstate highway. Between 1994 and 1998, the projected
mean annual growth rate in truck volumes was:
- For urban freeways and expressways: 6.5 percent.
- For rural interstates: 4.6 percent.
- For rural minor arterial highways: 3 percent.
- Vehicle Characteristics
Report No. FHWA-RD-00-054
- As of 1998, the nearly exclusive use of radial truck tires
was observed at all LTPP sites. By comparison,
74 percent of all truck tires were radials in 1988.
- The use of air suspension in trucks has increased. As of
1998, about 80 percent of all new truck tractors and about
60 to 70 percent of all new semi-trailers were equipped
with air suspension.
- Compared to bias ply tires, radial tires operate at higher tire
pressures and, thus, generate more sharply defined imprints
on pavements, which represent more concentrated loads.
- Compared to the traditional spring-leaf suspension, air
suspensions are considered to generate lower dynamic
pavement loads. However, air suspensions result in a
common high dynamic load frequency regardless of load
magnitude. The spatial concentration of traffic loads leads
to accelerated localized pavement damage.
- Climatic Effects on Pavement Performance
Report No. FHWA-RD-01-167
- In the SPS-2 experiment, the highest transverse cracking
was observed in the slabs built in the dry no-freeze
climates, followed by the wet-freeze climates, and then by
the dry-freeze climates. The slabs built in the wet nofreeze
climates have the lowest transverse cracking. The
data support similar findings from earlier studies that, in
drier climates (the western United States) where high thermal
gradients exist, it is important to design for resistance
to transverse cracking (shorter joint spacing minimizes the
adverse effect of climate).
- The largest longitudinal cracking lengths of SPS-2 sections
occurred in the dry no-freeze climates, followed by the
dry-freeze climates, and then by the wet-freeze climates.
The lowest longitudinal cracking lengths were observed
in those sections built in the wet no-freeze climates.
- Estimating Climatic Parameters Using Virtual Weather Stations Data
Report No. FHWA-RD-03-092
- The daily, monthly, and yearly LTPP Virtual Weather Station
(VWS) climatic estimates obtained by a newly developed
model were found to be reasonably accurate for locations
across North America. The climatic conditions (including
air temperature, precipitation, humidity, freezing index, and
wind speed) for 880 SPS and GPS pavement sections are
estimated using data from as many as 5
nearby national weather stations. These VWS estimates
are compared to onsite data for the same time period
measured by the Seasonal Monitoring Program (SMP) at
63 GPS and SPS sections and by Automatic Weather
Stations at 35 SPS test sections. Results of this comparison
have verified that the model for developing VWS
estimates can be a useful tool to predict climatic conditions.
- The LTPP VWS climatic estimates also were found to compare
well to the National Climatic Data Center's (NCDC)
measurements. These data were collected from 1994
through 1996 for the NCDC Cooperative Program; they
covered 8,000 weather stations scattered over 5,347
NCDC sites throughout the United States.
- A difference in elevation between a project site and the
nearby weather station(s) of more than 250 meters (m) (825
feet (ft)) significantly affects the climatic estimates. In this
case, temperatures must be corrected to reduce the bias of
the estimate. A model was developed for correcting the
maximum temperature for elevation difference.
- Within a range of 60 kilometers (km) (37.5 miles), the
distance of the contributing weather stations from a project
site does not affect the VWS estimates at any project site.
- Variability of Climatic Parameters
Significant year-to-year variability was observed in climatic data,
an important factor for pavement design procedures. The yearto-
year variability of annual precipitation is 21 percent; and of the
freezing index, 34 percent. On average, the year-to-year variability
of monthly temperature data is 6 percent.
- Environmental Effects in the Absence of Heavy Load
Report No. FHWA-RD-02-087
It is very important to study the effects of environmental factors such as climate and subgrade on the performance of flexible and
rigid pavement with a reduced number of heavy axle loads. The SPS-8 experiment is designed to emphasize the effects of siterelated
factors (temperature, precipitation, and subgrade) and structural factors (pavement type and layer thickness) on flexible and
rigid pavements with no more than 10,000 18-kip equivalent singleaxle loads (ESALs) per year in the study lane. The SPS-8 experiment
can be considered as an extension of SPS-1 (new flexible pavements) and SPS-2 (new rigid pavements) with limited traffic effects.
- Temperature and Precipitation
For SPS-8 AC sections, the most prevalent early distress is
longitudinal cracking outside the wheel path. The distress is most commonly observed for sections built in the wetfreeze
climates and for sections on an active subgrade (frost-susceptible or swelling soils due to freeze-thaw
cycles). Fatigue, longitudinal cracking in the wheel path, and transverse cracking are present on just a few sections. The
mean rut depths for all AC sections are below 6 millimeters (mm) (0.24 inches).
- Pavements (flexible or rigid) constructed on active subgrade have the highest mean initial International
Roughness Index (IRI) values and slopes (the smoothness rate of change over time), followed by
pavements constructed on fine subgrade, and coarse subgrade. The data support a similar finding
from previous studies that a good working platform (specifically, stabilized base and granular subgrade
or embankment) contributed to a smoother pavement construction.
- Initial IRI values for the SPS-8 test sections show that flexible pavements were constructed to be smoother
than the rigid pavements. The analysis of IRI slopes indicates that the subgrade is the most important
factor for flexible sections, while precipitation appears to be the most important factor for rigid
- Pavement Type and Layer Thickness
- The SPS-8 flexible pavements with thin (102-mm (4-inch)) AC surface layers were found to be
smoother than the sections with thick (178-mm (7-inch)) AC layers. Similarly, the SPS-8 rigid pavements
with thin (203-mm (8-inch)) concrete slabs were constructed to be smoother than the sections
with thick (279-mm (11-inch)) concrete slabs. This seems to contradict the general idea that thicker
surface layers can generate smoother pavements.Further investigations should be conducted.
- A few of the SPS-8 PCC sections have very limited transverse cracking and joint faulting. The mean joint
faulting on all PCC sections is insignificant, i.e., below 0.4 mm (0.02 inches). However, these
observations are based on 8 years of data (the oldest SPS-8 test section was 8 years old as of June 2001),
which is early in terms of pavement life.
- Comparisons of SPS-1, -2, and -8 Test Sections
- As expected, traffic loading is much heavier on SPS-1 and SPS-2 than on SPS-8 sites. As of June 2001,
the estimated accumulated ESALs on SPS-1 sites was about 1.46 million, compared to 0.043 million
ESALs on SPS-8 AC sites. Similarly, SPS-2 sites had accumulated 4.77 million ESALs, compared to 0.23
million ESALs on SPS-8 PCC sections.
- The average IRI slopes (the smoothness rate of change over time) for both SPS-1 and SPS-2 sections
are much higher than for the corresponding SPS-8 sections. The variability of mean IRI slopes is higher
for PCC than for AC sections.
- Overall, the much more heavily loaded SPS-1 and SPS-2 sections exhibit higher amounts of load-related
distresses. Such distresses include AC rutting, AC fatigue cracking, JPCP joint faulting, and JPCP
transverse cracking. However, the non-load-related distresses including AC transverse cracking and
non-wheel path longitudinal cracking are similar for SPS-1, SPS-2, and SPS-8.
- Moisture Contents
Report No. FHWA-RD-99-115
The Time Domain Reflectometry (TDR) technique measures the
dielectric constant of soils in the LTPP SMP. This constant can be
used to compute the in-situ moisture content of unbound base
and subgrade materials. This study was intended to develop procedures
to produce good estimates of in-situ gravimetric moisture
content using the TDR traces in the LTPP database
- In-situ gravimetric moisture content of unbound base and
subgrade materials can be determined using a two-step
- Volumetric moisture content of unbound base and
subgrade materials is determined using four proposed
models based on LTPP TDR traces and necessary
- In-situ gravimetric moisture content is then determined
using two newly developed methods based on volumetric moisture content.
- The two-step procedure was further developed into a userinteractive
computer program, MOISTER. The program is
used to determine moisture content of unbound base and
- Frost Penetration
Report No. FHWA-RD-99-088
The bulk resistivity of a soil increases dramatically when the soil
freezes. The electrical resistivity technique is used to measure the
electrical resistance, which is the voltage drop divided by the current
passing through a pavement depth, which is based on Ohm's law.
Together with soil temperature measurements, the electrical resistivity
(i.e., geometry-adjusted resistance) is used to estimate the
depth of frost penetration beneath a pavement section.
- A user-interactive computer program, FROST, was developed
to facilitate the determination of frost penetration
depth within a pavement structure by interpreting the electrical
resistivity and soil temperature data collected at the
- The moisture content of a soil determined by MOISTER
(FHWA-RD-99-115) based on the TDR data can be used
to confirm the freezing events as determined by FROST.
The rationale is that the moisture content computed by
TDR data does not include the frozen water (ice content).
Hence, when a soil freezes, its TDR-computed moisture
content drops because its unfrozen moisture content
Previous | Table of Contents | Next