9.1 Background
The Mississippi Department of Transportation (MDOT) has been evaluating thinlays on the NCAT Test Track for 15 years. In 2003, a ắ-inch thick, low volume road, 4.75 mm nominal maximum aggregate size (NMAS) mix was placed on the track (Section W6) with the expectation that it would only last for a half-million equivalent single axle loads (ESALs);
however, this mix has proven to be one of the most versatile surface layers in the history of the Test Track. To date, this section has supported over 50 million ESALs with no cracking, rutting, roughness, raveling, or friction deficiencies noted. The 2003 Mississippi thinlay consisted of 69%
imported limestone screenings, 30% hard sand, 1% hydrated lime antistrip agent, 6.1% polymer modified liquid asphalt, and no reclaimed or recycled materials.
In order to address the need to maintain more lane miles with limited budgets, MDOT decided to use the 2012 research cycle to redesign thinlays by adding reclaimed asphalt pavement (RAP), changing from polymer modified to neat asphalt, eliminating imported stone screenings, and relying completely on locally available surplus sand stockpiles in Mississippi. The result of this effort was a new thinlay mix placed on the surface layer of Section S3 in 2012. The
redesigned preservation thinlay surface consisted of hard sand locally available in Mississippi, RAP, Portland cement filler, hydrated lime antistrip, and neat virgin liquid asphalt. The target lift thickness on the track was 1 inch, but this mix could be used for preservation in lifts as thin as ắ inch. Both W6 and S3 thinlay mixtures were placed on the originally constructed Test Track sections with approximately two feet of asphalt structure. A low cost per mile can be achieved as a result of the use of all local materials, RAP, and neat liquid asphalt in a relatively thin surface layer (i.e., low spread rate).
9.2 Objective
The objective of Section S3 for the 2012 research cycle was to evaluate the rutting performance of the redesigned thinlay mix on the NCAT Test Track. After the application of 10 million ESALs at the end of the 2012 cycle, no significant rutting or surface cracking was observed. MDOT chose to continue traffic through the 2015 research cycle in order to expand the scope of the mix evaluation to include cracking and durability.
9.3 Mix Design
Table 1 provides information about the redesigned preservation thinlay surface, which
consisted of 72% hard local sand, 25% processed RAP, 2% Portland cement filler, 1% hydrated lime antistrip, and 5.1% neat virgin liquid asphalt. The asphalt binder contributed by the RAP source was 1.6%, resulting in 6.7% total asphalt content. Quality control data for the binder and surface mixes sampled during production are shown in Table 2. No problems were noted during production, placement, or compaction. A loose mix sample was taken during construction for laboratory performance testing. A final density of 94.2% was measured in the compacted mat
with no tenderness observed at any temperature. Tables 1 and 2 also include information for the original thinlay mixture placed in Section W6 for comparison.
Table 1. Mix Design Information for Mississippi Thinlay Mixtures
Design/Materials S3 Thinlay Mixture W6 Thinlay Mixture
Gyrations Ndes 50 Ndes 50
Asphalt Virgin PG 67-22
(unmodified) 5.1% Virgin PG 76-22
(SBS-Modified) 6.1%
Asphalt from RAP 1.6%
Total Asphalt 6.7% Total Asphalt 6.1%
Aggregate -3/8” Bailey Coarse Gravel 40% Cherokee Limestone 69%
Bailey Crushed Fines 8% Guntown Crushed Gravel 19%
Bailey Coarse Sand 24% Mississippi Natural Sand 11%
RAP Blaine RAP 25%
Others Hydrated Lime 1% Hydrated Lime 1%
Portland Cement 2% Sttucture
Table 2. Gradation and Volumetric Properties of Mississippi Thinlay Mixtures
Mix Properties S3 Thinlay Mixture W6 Thinlay Mixture
Sieve Mix Design
(Percent Passing)
Quality Control
Mix Design (Percent Passing)
Quality Control
1/2” (12.5 mm) 100 100 100 100
3/8” (9.5 mm) 99 99 100 100
No. 4 (4.75 mm) 83 86 99 98
No. 8 (2.36 mm) 61 62 72 75
No. 16 (1.18 mm) 46 46 43 50
No. 30 (0.60 mm) 35 34 30 35
No. 50 (0.30 mm) 17 17 18 22
No. 100 (0.15 mm) 9 9 11 15
No. 200 (0.075 mm) 7.5 6.4 8.0 11.3
Volumetrics
Total Asphalt (%) 6.7 6.4 7.5 6.1
Virgin Asphalt (%) 5.1 4.8 7.5 6.1
RAP Asphalt (%) 1.6 1.6 0.0
Lab Air Voids (%) 5.0 4.1 4.0
Voids in Mineral Aggregate (VMA) (%) 18.4 17.3 16.0
Voids filled with Asphalt (VFA) (%) 73 76
In-Place Density (% of Gmm) 94.2 92.2
9.4 Laboratory Performance Testing
As part of the 2012 cycle, the loose mix sample taken during construction was reheated in the NCAT laboratory to prepare test specimens for evaluating rutting performance of the
redesigned thinlay mix. The Asphalt Pavement Analyzer (APA) was used in accordance with AASHTO T 340-10 (2015). The APA is a modification of the Georgia Loaded Wheel Tester (GLWT) and follows a similar rut testing procedure where a wheel is loaded onto a pressurized linear hose and tracked back and forth over a testing sample to induce rutting, as shown in Figure 1.
Six specimens were compacted with a Superpave gyratory compactor to target air voids of 7.0 ± 0.5 percent and a height of 75 mm. The specimens were then tested at 64°C (the 98% reliability
temperature for the high performance grade of the binder for the Test Track) using a vertical load of 100 lbs. and a hose pressure of 100 psi for 8,000 cycles. Rut depth readings were taken manually at two locations on each specimen after being seated for 25 loading cycles and at the end of testing (8,000 cycles in addition to 25 seating cycles) and automatically during the test.
Higher rutting depth indicates the less resistance to rutting.
Figure 1 Asphalt Pavement Analyzer
Table 3 shows the manual and automated rut depth measurements for the thinlay mixture.
Past research has shown that if a mixture has an average APA rut depth of less than 5.5 mm, it should be able to withstand 10 million ESALs of truck traffic at the Test Track without
accumulating more than 12.5 mm of field rutting. Based on the data shown in Table 3, the thinlay mixture met this criterion. These APA results agree with the field rutting results since the thinlay mix did not fail due to rutting during the 2012 research cycle.
Table 3. Summary of APA Test Results
Sample ID Air Voids (%) Manual Rut Depth (mm) Automated Rut Depth (mm) Mold Placement
5 6.9 5.2 3.8 Left Rear
7 7.0 4.6 4.6 Left Front
9 6.9 5.3 4.5 Center Rear
6 7.1 5.4 4.9 Center Front
10 6.9 6.2 4.8 Right Rear
8 7.0 5.5 5.1 Right Front
Average 7.0 5.4 4.6
Std. Dev. 0.1 0.5 0.5
9.5 Test Track Performance
Weekly monitoring of each test section was conducted on Mondays. Each section was inspected for signs of cracking and multiple measurements of rutting and surface roughness were made. Similar to the 2003 MDOT thinlay, this mix has now supported over 20 million ESALs with no cracking, rutting, roughness, raveling, or friction deficiencies noted. At the end of the 2015 cycle, the final rut depth measured was 0.05 inches, and the amount of cracking observed was 0.2% based on the total lane area. Figures 2 and 3 show two low-severity cracks observed in Section S3.
Figure 2 Crack Observed in Section S3 Near the Inside Edge of the Outside (Research) Lane
Figure 3 Crack Observed in Section S3 Near Center of the Outside (Research) Lane Figure 4 shows the texture change of the thinlay mixture in Section S3 through the 2012 and 2015 cycles. Macrotexture increased slightly during the 2015 cycle, but the mean texture depth was still good at around 0.6 mm after 20 million ESALs. Pavement roughness quantified using
international roughness index (IRI) during the 2015 cycle is shown in Figure 5. The IRI data are very consistent, indicating that the pavement remained smooth.
Figure 4 Macrotexture Results for Thinlay Mixture After 20 Million ESALs
Figure 5 Pavement Roughness Results for Thinlay Mixture After 20 Million ESALs 9.6 Conclusions
The objective of Section S3 in the 2012 and 2015 Test Track cycles was to evaluate the
production, placement, rutting, cracking, and durability performance of the redesigned thinlay mix consisting of all locally available surplus sand materials, RAP, and neat liquid asphalt (PG 67- 22) in a relatively thin surface layer (i.e., a target lift thickness of 1 inch or less). Through the 2012 and 2015 cycles, this mix has supported over 20 million ESALs of heavy truck traffic with no cracking, rutting, roughness, raveling, or friction deficiencies noted.