U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
2023664000
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: FHWARD01166 Date: November 2003 
Structural Factors for Flexible Pavements—Initial Evaluation of the SPS1 Experiment
Final Report
Figure 1. Location of the SPS1 projects. Diagram. The figure is a U.S. map with boxes indicating SPS sites located in Alabama, Arizona, Arkansas, Delaware, Florida, Iowa, Kansas, Louisiana, Michigan, Montana, Nebraska, Nevada, New Mexico, Oklahoma, Ohio, Texas, Virginia, and Wisconsin.
Figure 2. LTPP data collection and data movement flowchart. Flowchart. The figure consists of a flowchart showing seven data types being collected. Deflection data first is collected with the falling weight deflectometer (FWD) and then goes through field control. The data then is sent to the regional offices where the quality control (QC) is done with FWD scan software, after which, data may be edited. Data is then loaded into the Regional Information Management System (RIMS) database. Longitudinal profile data is collected with the profilometer, and then it goes through field control by checking on International Roughness Index (IRI) variability. The data is then sent to the regional offices, where it is put through PROFCHK software to recompute profile parameters and to perform QC, after which, some profiles may be eliminated. The data then is loaded into the RIMS database. Transverse profile data is collected with both dipstick and film. Dipstick data is put through PROFCHK software by the regions, where some profiles may be eliminated. Film data is analyzed and electronically stored before being sent to the regional offices. All transverse profile data then is loaded into the RIMS database. Distress data is collected both manually and photographically. Manual data is reviewed and entered by the regional offices and then loaded into the RIMS. Photographic data is sent directly to the regional offices and then loaded into the RIMS. Climatic data, is collected at Automatic Weather Stations (AWS), either downloaded by modem or during site visits. The data is then reviewed with AWS check software. The data then is checked, extracted, and loaded into the RIMS. Traffic data is collected by weighinmotion (WIM) equipment and automated vehicle classification (AVC). Data is entered into the Regional Traffic Database. The QC then is performed and data is flagged. All data is entered into the Central Traffic Database. All traffic data deemed acceptable is used to compute summary statistics, and then the statistics are entered into the RIMS. Materials data is obtained from field and laboratory tests. Coring and sampling is done and tested by the LongTerm Pavement Performance (LTPP) program and State agencies. Test data and data forms are forwarded to the regional offices. For insitu field testing, test data and data forms are forwarded to the regional offices. All materials data is then reviewed by the regional offices and converted to machinereadable form before being loaded into the RIMS. All of the seven data types loaded into the RIMS are run through a series of dataspecific QC programs. Manual data upgrades are performed by the regions, if appropriate, and then loaded into the national Information Management System (IMS) database. Data that makes it to level E is released to the public.
Back to FigureFigure 3. Thickness histograms for the thin HMA layer (102 millimeters) from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction of 102millimeter hotmix asphalt concrete (HMAC), with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples have thicknesses of 95 to 115 millimeters, with frequencies greater than 20 percent. The remaining thicknesses have frequencies of less than 5 percent. In the second graph, L05B 102millimeter HMAC is shown, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. All samples have thicknesses between 85 and 135 millimeters. Between 95 and 115 millimeters, frequencies are between 25 and 35 percent. The remaining thicknesses have frequencies of less than 10 percent.
Figure 4. Thickness histograms for the thick HMA layer (178 millimeters) from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 178millimeter HMAC, and the second shows L05B 178millimeter HMAC. Both graphs show thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples for both graphs have thicknesses of 170 to 210 millimeters, with frequencies between 5 and 35 percent.
Figure 5. Thickness histograms for the thin ATB layer (102 millimeters) from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 102millimeter asphalt treated base (ATB), with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples have thicknesses of 95 and 105 millimeters, with frequencies of 35 and 40 percent, respectively. The remaining samples had thicknesses of 85, 115, and 125, with frequencies of less than 10 percent. In the second graph, L05B 102millimeter ATB is shown, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples had thicknesses between 95 and 115 millimeters, with frequencies between 25 and 45 percent.
Figure 6. Thickness histograms for the thick ATB layer (203 millimeters) from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 203millimeter ATB, and the second shows L05B 203millimeter ATB, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples on both graphs have thicknesses between 195 and 225 millimeters, with frequencies between 10 and 45 percent.
Figure 7. Thickness histograms for the PATB layer from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 102millimeter permeable asphalt treated base (PATB), and the second shows L05B 102millimeter PATB, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples on both graphs have thicknesses between 95 and 115 millimeters, with frequencies between 15 and 58 percent.
Figure 8. Thickness histograms for the 102millimeter DGAB layer from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 102millimeter dense graded aggregate base (DGAB), and the second shows L05B 102millimeter DGAB, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples on both graphs have thicknesses of 95 and 105 millimeters, with frequencies between 20 and 65 percent.
Back to FigureFigure 9. Thickness histograms for the 203millimeter DGAB layer from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 203millimeter DGAB, and the second shows L05B 203millimeter DGAB, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples on both graphs have thicknesses between 185 and 215 millimeters, with frequencies around 65 percent at 205 millimeters on each graph.
Figure 10. Thickness histograms for the 305millimeter DGAB layer from tables SPS1_LAYER (construction data) and TST_L05B. Graphs. The figure consists of two bar graphs. The first graph shows construction 305millimeter DGAB, and the second shows L05B 305millimeter DGAB, with thickness in millimeters on the horizontal axis and percent frequency on the vertical axis. The majority of samples on the construction graph have thicknesses between 285 and 315 millimeters, with a frequency of about 55 percent at 305 millimeter thickness. The majority of samples on the second graph have thicknesses between 295 and 315 millimeters, with a frequency of about 65 percent at 305 millimeters.
Figure 11. Histogram of air voids measured on the HMA surface layer. Graph. The figure consists of a bar graph, with percent air voids on the horizontal axis and frequency on the vertical axis. The frequency is rather evenly distributed across air voids between 2 and 12 percent, with frequencies between 5 and 15 percent.
Figure 12. Histogram of air voids measured on the HMA binder layer. Graph. The figure consists of a bar graph, with percent air voids on the horizontal axis and frequency on the vertical axis. The frequency is distributed between air voids of 1 to 13 percent, with frequencies between 2 and 15 percent.
Figure 13. Histogram of air voids measured on the ATB layer. Graph. The figure consists of a bar graph, with percent air voids on the horizontal axis and frequency on the vertical axis. The frequency is distributed between air voids of 2 to 9 percent, with frequencies between 5 and 20 percent.
Figure 14. Histogram of the material passing the number 4 sieve, PATB layer. Graph. The figure consists of a bar graph, with percent passing number 4 sieve on the horizontal axis and frequency on the vertical axis. For percent passing of 5, 10, 15, 20, 35, and greater than 50, there were frequencies of about 30, 52, 5, 5, 3, and 3 percent, respectively.
Figure 15. Histogram of material passing the number 200 sieve, PATB layer. Graph. The figure consists of a bar graph, with percent passing number 200 sieve on the horizontal axis and frequency on the vertical axis. For percent passing of 1, 2, 3, 4, and 7, there were frequencies of about 10, 20, 45, 24, and 3 percent, respectively.
Back to FigureFigure 16. Histogram of material passing the number 200 sieve, HMA surface layer. Graph. The figure consists of a bar graph, with percent passing number 200 sieve on the horizontal axis and frequency on the vertical axis. For percent passing of 3, 4, 5, 6, 7, and 8, there were frequencies of about 7, 10, 4, 40, 20, and 20 percent, respectively.
Figure 17. Histogram of material passing the number 200 sieve, ATB layer. Graph. The figure consists of a bar graph, with percent passing number 200 sieve on the horizontal axis and frequency on the vertical axis. Samples are rather evenly across 2 to 9 percent passing the number 200 sieve, with frequencies ranging between 5 and 25 percent.
Figure 18. Area of fatigue cracking measured over time comparing test sections with and without permeable base layers for all SPS1 projects combined. Graph. The figure consists of a graph of drained sections versus undrained sections, with the age in years on the horizontal axis and fatigue in meters squared on the vertical axis. The majority of fatigue measured occurred on sections between 2 and 6 years old. The undrained sections showed fatigue on about 16 sections, ranging between 1 and 130 meters squared. The drained sections showed fatigue on about 37 sections, ranging between 1 and 190 meters squared.
Figure 19. Total length of transverse cracks measured over time comparing test section with and without permeable base layers for all SPS1 projects combined. Graph. The figure consists of a graph of drained sections versus undrained sections, with the age in years on the horizontal axis and transverse cracking in meters on the vertical axis. The majority of cracking occurred on pavements between 1 and 3 and onehalf years old. Measurable crack lengths ranged between 0 and 85 meters, with no apparent difference between the drained and undrained sections.
Figure 20. IRI values measured over time comparing test sections with and without permeable base layers for all SPS1 projects combined. Graph. The figure consists of a graph of drained versus undrained sections, with age in years on the horizontal axis and IRI in meters per kilometer on the vertical axis. All sections graphed were between 0 and 7 years of age. The majority of undrained sections and all of the drained sections had IRI between 0.5 and 2. Approximately 6 of the undrained sections had IRIs greater than 2.
Figure 21. Rut depths measured over time comparing test sections with and without permeable base layers for all SPS1 projects combined. Graph. The figure consists of a graph of drained versus undrained sections, with age in years on the horizontal axis and rut depth in millimeters on the vertical axis. All sections graphed were between 0 and 6 years of age. Rut depths ranged between 1 and 30 millimeters, with no apparent difference between drained and undrained sections.
Back to FigureFigure 22. Longitudinal cracking in the wheel paths measured on different dates for the core test sections of the Alabama project. Graphs. The figure contains two line graphs showing the Alabama SPS1 sections constructed March 1, 1993, with survey dates on the horizontal axis and longitudinal cracking in the wheel path in meters on the vertical axis. The survey dates begin January 1, 1994, and end December 31, 1999. The first graph shows 6 test sections with no drainage. Section 1 had crack measurements that increased steadily to 70 meters by 1997 and dropped to 0 by 1998. Section 2 had crack measurements of 15 meters in 1995, then dropped to 0 by 1996, and then increased steadily to 45 meters in 1997 and dropped to 0 by 1998. Section 4 had 1 measurement of 40 meters in 1996. Sections 3, 5, and 6 had no measured longitudinal cracks. The second graph shows 6 sections with drainage. Sections 7–11 had measurements that start at 0 and jump up to lengths between 5 meters and nearly 40 meters. Section 12 had no measured cracks.
Figure 23. Longitudinal cracking outside the wheel paths measured on different dates for the core test sections of the Alabama project. Graphs. The figure contains two line graphs showing the Alabama SPS1 sections constructed March 1, 1993, with survey dates on the horizontal axis and longitudinal cracking outside the wheel path in meters on the vertical axis. The survey dates begin January 1, 1994, and end December 31, 1999. The first graph shows 6 test sections with no drainage. All sections had measurements of 0 before and after January 1, 1996, but had lengths as high as 100 meters for most of the sections on January 1, 1996. In the second graph, there were 6 test sections with drainage. In general, the test sections had longitudinal cracks that were measured around 100 meters in 1994, then measured at 0 in 1995, then measured between 0 and 80 in January 1996, and then at 0 later in 1996.
Figure 24. Transverse cracking measured on different dates for the core test sections of the Alabama project. Graphs. The figure contains two line graphs showing the Alabama SPS1 sections constructed March 1, 1993, with survey dates on the horizontal axis and transverse cracking in meters on the vertical axis. The survey dates begin January 1, 1994 and end December 31, 1999. The first graph shows 6 test sections with no drainage. Two of the sections had transverse cracking measured between 1 and 3 meters in January 1996 and at 0 at all other survey dates. The remaining sections had measurements around 0 on all survey dates. In the second graph, there are 6 sections with drainage. In 1994, 1 section had a measurement of 6 meters, and the other sections had measurements of 0. In 1996, crack lengths between 1 and 3 meters were measured for sections 8, 9, and 10, and 0 for the remaining sections.
Figure 25. Graphical illustration of the average amount of fatigue cracking observed on each of the projects, as of January 2000. Graphs. The figure consists of two graphs. In the first graph, the age in years is on the horizontal axis and fatigue cracking in meters squared is on the vertical axis. Measurements for fatigue cracking were around 0 for the first 2 years, and then increased between 0 and 20 meters squared between 3 and 5 years. Around 6 and 7 years, measurements ranged between 0 and 40 meters squared. In the second graph, the age in years is on the horizontal axis and the number of sections with fatigue cracks is on the vertical axis. The number of sections with cracks is around 0 for the first 2 and onehalf years. Beginning around 3 years, the number of sections increased to 1, and up to 10 by 7 years.
Back to FigureFigure 26. Percentage of the core test sections that exceed an IRI value of 1.2 meters per kilometer. Graph. The figure consists of a bar graph. The bar graph shows the experimental site factorial features on the horizontal axis and percentage of total test sections on the vertical axis. For coarsegrained, nofreeze; finegrained, nofreeze; coarsegrained, freeze; and finegrained, freeze sites, there were 0, 0, 2.1, and 31.9 percent of the total test sections, respectively.
Figure 27. Percentage of the core test sections that exceed 8 millimeters of rutting. Graph. The figure consists of a bar graph that shows drainage condition and base type on the horizontal axis and percent of test sections exceeding 8millimeter rut depths on the vertical axis. Permeable ATB sections were 6.7 percent of test sections exceeding 8millimeter rut depths, dense ATB were 11.7 percent, permeable aggregate was 13.3 percent, and dense aggregate was 20 percent.
Figure 28. Percentage of test sections that exceed an IRI value of 1.2 meters per kilometer. Graphs. The figure consists of two bar graphs. The first graph shows HMA surface thickness in millimeter and base type on the horizontal axis and percent of test sections that exceed an IRI of 1.2 meters per kilometer on the vertical axis. 178millimeter ATB sites were 3.8 percent of test sections. 102millimeter ATB sites were 13.5 percent. 178millimeter aggregate sites were 18.4 percent. 102 millimeter and aggregate sites were 21.1 percent of sites. The second graph shows drainage condition and base type on the horizontal axis and percent of test sections that exceed an IRI of 1.2 meter per kilometer. Permeable ATB sites were 6.7 percent of test sections. Dense ATB were 10 percent. Permeable aggregate was 20 percent. Dense aggregate was 20 percent.
Figure 29. Percentage of core test sections that have fatigue cracking. Graph. The figure consists of a bar graph. The bar graph shows drainage condition and HMA surface thickness in millimeters on the horizontal axis and percent of test sections with fatigue cracking on the vertical axis. For permeable 178millimeter, dense 178millimeter, permeable 102millimeter, and dense 102millimeter test sections, the percents with cracking were 8.3, 7.4, 10.2, and 14.8, respectively.