MATERIALS AND METHODS
DGA 4.7 DGA 4.7
1. Only some of the measurement results (mainly total permanent deformation or rutting) are presented in this paper due to the length limit. More details can be found
An extended HVS test section is 15.0 m (49.2 ft) long and 1.0 m (3.3 ft) wide. A schematic in Fig. 1 shows an HVS test section along with the stationing and coordinate system. Station numbers (0 to 30) refer to fixed points on the test section and are used for measurements and as a reference for discussing performance.
Stations are placed at 0.5 m (1.6 ft) increments. A sensor installed at the center of the test section would have an x-coordinate of 7,500 mm and a y-coordinate of 500 mm.
Thermocouples (0 and 25mm) Thermocouple 80mm Paver 50mm Bedding Layer
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
y
x
Station Number (unit: 0.5m) 0.2 m
0.2 m 1 m
Subgrade
N
Legend
NOT TO SCALE
Permanent deformation Road surface deflectometer Thermocouple
Surface profile measured at every station between Station 3 and Station 27 15 m
Subbase
Pressure cell
Top View
Cross Section
100mm Base
450mm 650mm 950mm
Fig. 1: Proposed pavement structure for PICP test track and schematic of an extended HVS test section.
Test section instrumentation and measurements
Measurements were taken with the various equipment and instruments for the following variables: pavement and air temperature, water level in the pavement, surface permanent deformation (rutting), permanent deformation in the underlying layers, and surface deflection, vertical pressure (stress) at the top of the subbase and top of the subgrade, and jointing stone depth. Instrument positions are shown in Fig.
1. Only some of the measurement results (mainly total permanent deformation or
HVS test conditions
Environmental conditions
Pavement temperatures were not controlled in this experiment since temperature has a minimal effect on the behavior of the concrete pavers in terms of pavement response. All testing was therefore carried out at ambient temperatures.
Three tests were conducted: dry, wet and drained tests. During the dry test, a number of light rainfall events were recorded, but no change in subgrade moisture condition occurred. In the wet test, water was soaked through the pavers next to the test section and allowed to fill the pavement structure until it overflowed (Fig. 2).
The water level was then allowed to recede to the top of the coarse aggregate subbase, where it was maintained with controlled water flow until the end of the test.
In the drained test, no water was added and testing was started once the water level receded to the top of the subgrade (i.e., there was no standing water in the coarse aggregate subbase during the test).
Fig. 2: Flooded section during preparation for HVS wet testing.
Loading program
The HVS loading program for each test is summarized in Table 1.
Table 1: Summary of HVS Loading Program.
Section # Test Moisture Condition (Date)
Half-Axle Wheel Load1 (kN)
Repetitions ESALs2 678HC Dry
(02/12/2014 - 03/31/2014)
25 40 60
100,000 100,000 140,000
15,259 100,000 708,750
Section Total 340,000 824,009
679HC Wet (04/17/2014 - 05/21/2014)
25 40 60
100,000 100,000 140,000
15,259 100,000 708,750
80 40,000 640,000
Section Total 380,000 1,464,009
680HC Drained (06/11/2014 - 07/14/2014)
25 40
100,000 25,000
15,259 25,000
Section Total 125,000 40,259
Project Total 845,000 2,328,277
1 40 kN = 9,000 lb.; 60 kN = 13,500 lb; 80 kN = 18,000 lb.
2 ESAL: Equivalent standard axle load with 4th power.
All trafficking was carried out with a dual-wheel configuration, using radial truck tires (Goodyear G159 - 11R22.5- steel belt radial) inflated to a pressure of 700 kPa (101 psi), in a bidirectional loading mode with a one meter wide wander pattern (i.e., trafficking in both directions in line with standard procedures for testing base layer performance). Load was checked with a portable weigh-in-motion pad at the beginning of each test, after each load change, and at the end of each test.
TESTING RESULTS AND DISCUSSION Rainfall during APT testing
Fig. 3 shows the monthly rainfall data from January 2014 through August 2014 measured at the weather station close to the test track. This period spans construction of the test track and the three HVS tests. Rainfall was recorded during dry and wet testing, but not during the undrained test. Daily rainfall was very low with 6.4 mm (0.26 in.) being the highest recorded on any one day during testing. During the dry test, the test section was protected from direct rainfall by the HVS. The area surrounding the HVS was covered with plastic sheeting to prevent any infiltration of water.
5 87
35 41
0 0 0 0
0 10 20 30 40 50 60 70 80 90 100 110
Jan Feb Mar Apr May Jun Jul Aug
Rainfall (mm)
Month (2014) 678HC - Dry
680HC - Drained 679HC - Wet
Fig. 3: Measured rainfall during the study period (25 mm = 1 inch).
Dry test (Section 678HC)
678HC: Test summary
The HVS loading history for testing on the dry section is shown in Fig. 4.
Measurements showed no water in the subbase for the duration of the dry test.
0 50 100 150 200 250 300 350 400
2/10/14 2/20/14 3/2/14 3/12/14 3/22/14 4/1/14 4/11/14
Number of Load Repetitions (x1,000)
Date
Loading Schedule Number of Load Repetitions
25kN 40kN 60kN
Breakdown
Fig. 4: HVS loading history (678HC dry test).
678HC: Permanent deformation on the surface (rutting)
Fig. 5 shows the development of surface permanent deformation (average maximum total rut) with load repetitions for the three subsections. Observations of surface rutting for each wheel load during the dry test include the following:
• 25 kN (4,500 lb) Wheel Load
+ During HVS testing, rutting usually occurs at a high rate initially, and then it typically diminishes as trafficking progresses until reaching a steady state. This initial phase is referred to as the “embedment” phase. The initial embedment phase in this test, although relatively short in terms of the number of load repetitions (i.e.,
± 5,000), ended with a fairly significant early rut of about 5.0 mm (0.2 in.) that was attributed to bedding in of the pavers under the wheel load. The rate of rut depth increase after the initial embedment phase was uniform until the load change.
• 40 kN (9,000 lb) Wheel Load
+ A second small embedment phase was recorded after the load change to 40 kN.
The section with the 450 mm subbase was most sensitive to the load change, as expected. After embedment, the rate of rut depth increase was again uniform, but faster than the rate recorded with the 25 kN load, indicating that the pavement was sensitive to very heavy loads (i.e., at or above legal design loads.
• 60 kN (13,500 lb) Wheel Load
+ A third embedment phase was recorded after the load change to 60 kN. The change in rut rate was more severe during this embedment, and the rate of rut depth increase accelerated. The change in rut rate was larger on the 450 mm and 650 mm subbase sections compared to that on the 950 mm subbase section. After completion of trafficking, the average maximum rut depth (average of the total rut
recorded at each station) for the 450 mm, 650 mm, and 950 mm subbase subsections was 24.5 mm (0.96 in.), 21.4 mm (0.84 in), and 17.7 mm (0.70 in.), respectively.
The test was stopped after 340,000 load repetitions (equivalent to 824,009 ESALs) when the average maximum rut on the 450 mm subbase subsection reached 25 mm (1 in.), which was the terminal rut depth set for the test.
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
0 50 100 150 200 250 300 350 400
Average Total Rut (mm)
Load Repetitions (x 1,000) 450 mm 650 mm 950 mm
25kN 40kN 60kN
Fig. 5: Average maximum total rut depth (678HC dry test).
Wet Test (Section 679HC)
679HC: Test summary
The HVS loading history for testing on the dry section is shown in Fig. 6.
0 50 100 150 200 250 300 350 400
4/16/14 4/26/14 5/6/14 5/16/14 5/26/14 6/5/14 6/15/14
Number of Load Repetitions (x1,000)
Date
Loading Schedule Number of Load Repetitions
40kN 60kN
25kN 80kN
Fig. 6: HVS loading history (679HC wet test).
679HC: Water level in the pavement
The water level was approximately maintained at the approximate top of the subbase for the duration of the wet test. A plot of the average water level measured in the 950 mm subbase subsection is shown in Fig. 7. The average water depth below the track surface was 274 mm (standard deviation of 62 mm), or 44 mm below the top of the subbase. This confirms that the subbase served as a reservoir layer during traffic loading. This condition can be considered a “very worst-case” traffic loading scenario since most permeable pavements are designed to drain standing water within
the base/subbase within 2 to 3 days after a single rain event; whereas loading was applied for 35 days while the subbase and subgrade were immersed in water.
0 200 400 600 800 1,000 1,200 1,400
0 50 100 150 200 250 300 350 400
Depth (mm)
Load Repetitions (x 1,000) Water Level Top of Subbase Pavement Surface
Top of Subgrade
25kN 40kN 60kN 80kN
Fig. 7: Water level in the pavement (950 mm subbase subsection) (200 mm = 8 inches) (679HC wet test).
679HC: Permanent deformation on the surface (rutting)
Fig. 8 shows the development of permanent deformation (average maximum total rut and average deformation) with load repetitions for the three subsections. This was significantly quicker compared to the rut depths recorded during testing under dry conditions.
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
0 50 100 150 200 250 300 350 400
Average Total Rut (mm)
Load Repetitions (x 1,000) 450 mm 650 mm 950 mm
25kN 40kN 60kN 80kN
Fig. 8: Average maximum total rut depth (679HC wet test).
679HC: Visual assessment
Apart from rutting, no other distress was recorded on the section. No cracked pavers were observed. Some darkening of the paver surfaces was noted, attributed to rubber deposits and polishing from the HVS tires. Photographs of the wet test section after HVS testing are shown in Fig. 9.
Close up of 450 mm subbase subsection.
View of test section looking from north to south Close up of 650 mm subbase subsection.
Fig. 9: Test section photographs (679HC wet test).
Drained Test (Section 680HC)
680HC: Test summary
This test was included to compare rate of rut increase on the section with no water in the subbase with rate of rut increase on the dry and wet tests. A limited number of load repetitions were applied, sufficient to compare surface rutting trends with those on the other two sections. The HVS loading history for testing on the dry section is shown in Fig. 10. No water was measured in the subbase for the duration of the drained test.
0 50 100 150 200 250 300 350 400
6/10/14 6/15/14 6/20/14 6/25/14 6/30/14 7/5/14 7/10/14 7/15/14 7/20/14
Number of Load Repetitions (x1,000)
Date
Loading Schedule Number of Load Repetitions
40kN 25kN
Fig. 10: HVS loading history (679HC drained test).
680HC: Permanent deformation on the surface (rutting)
Fig. 11 shows the development of permanent deformation (average maximum total rut and average deformation) with load repetitions for the three subsections. The plots show that rutting trends and rut depths were similar to those recorded on the dry section. Permanent deformation in the underlying layers was not measured in this drained test.
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70
0 50 100 150 200 250 300 350 400
Average Total Rut (mm)
Load Repetitions (x 1,000) 450 mm 650 mm 950 mm
25kN 40kN
Fig. 11: Average maximum total rut depth (679HC drained test).
CONCLUSIONS AND RECOMMENDATIONS
This paper briefly summarizes the main results from HVS APT testing and reveals the structural performance of PICP under heavy traffic loading on ports, highway shoulders, rest areas, and maintenance yards, etc. Key observations from the APT testing include:
• There was a significant difference in rutting performance and rutting behavior between the wet (i.e., water level maintained at the top of the subbase) and dry tests, as expected.
• A large proportion of the rutting on all three sections occurred as initial embedment in the first 2,000 to 5,000 load repetitions of the test and again after each of the load changes, indicating that much of the rutting in the base and subbase layers was attributed to bedding in, densification, and/or reorientation of the aggregate particles.
This behavior is consistent with rutting behavior on other types of structures.
• The number of load repetitions and equivalent standard axles required to reach the terminal rut depth (25 mm [1 in.]) set for the project is summarized in Table 2 for both dry and wet conditions. The sensitivity of the pavement structure to water (i.e., standing water in the subbase) and to load is clearly evident.
Table 2: Repetitions and ESALs Required to Reach Terminal Rut.
Test Load repetitions at terminal rut (25 mm) ESALs at terminal rut
450 mm 650 mm 950 mm 450 mm 650 mm 950 mm
Dry Wet
340,000 95,259
Rut < 25mm 180,000
Rut < 25mm 210,000
824,009 165,884
Rut < 25mm 220,000
Rut < 25mm 216,519
• Although only limited testing was undertaken under drained conditions (i.e., wet subgrade but no standing water in the subbase), rutting behavior appeared to show similar trends and behavior to the test under dry conditions.
• The increase in rate of rut depth increased with increasing load, indicating that the pavement structure was load sensitive, especially at load levels close to and above the legal load limit. Care should therefore be taken when designing projects that will carry large numbers of heavy or overloaded trucks.
• No distress was noted on any individual pavers and no pavers were dislodged from the pavement during testing.
• The thickness of the subbase influenced rutting behavior and rut depth in the different layers. The specific effect was not presented in this paper due to length limit. The general finding is that, a large proportion of the permanent deformation measured on the test track occurred in the subbase, and that increasing the thickness of the subbase did not reduce this rutting. Details can be found in the final report (Li, Jones, Wu and Harvey, 2014).
• Given that a large proportion of the permanent deformation measured on the test track occurred in the subbase and that increasing the thickness of the subbase did not reduce this rutting, it is recommended that the specifications of the aggregate properties used in this layer and the methods used to construct it are reviewed to determine whether any reductions in rutting can be achieved by changing them. Further research into stabilization of the subbase aggregate using geogrids, geocells, or cement should also be considered.
REFERENCES
Chow, L. C., Mishra, D., and Tutumluer, E. "Framework for Improved Unbound Aggregate Base Rutting Model Development for Mechanistic-Empirical Pavement Design." Proc., Transportation Research Board 93rd Annual Meeting.
Huang, H., Tutumluer, E., and Dombrow, W. (2009). "Laboratory characterization of fouled railroad ballast behavior." Transportation Research Record: Journal of the Transportation Research Board, 2117(1), 93-101.
Jones, D. (2005). "Quality management system for site establishment, daily operations, instrumentation, data collection and data storage for APT experiments." Pretoria, South Africa: CSIR Transportek.(Contract Report CR-2004/67-v2).
Jones, D., Harvey, J., Li, H., Wang, T., Wu, R., and Campbell, B. (2010). "Laboratory Testing and Modeling for Structural Performance of Fully Permeable Pavements under Heavy Traffic: Final Report." University of California Pavement Research Center, Davis, California.
Jones, D., Li, H., and Harvey, J. (2013). "Development and HVS Validation of Design Tables for Permeable Interlocking Concrete Pavement: Literature Review."
Kim, I. T., and Tutumluer, E. (2006). "Field validation of airport pavement granular layer rutting predictions." Transportation Research Record: Journal of the Transportation Research Board, 1952(1), 48-57.
Li, H., Jones, D., and Harvey, J. T. (2010). "Summary of a Computer Modeling Study to Understand the Performance Properties of Fully Permeable Pavements."
Li, H., Jones, D., Wu, R., and Harvey, J. (2013). "Development and HVS Validation of Design Tables for Permeable Interlocking Concrete Pavement: Field Testing and Test Section Structural Design."
Li, H., Jones, D., Wu, R., and Harvey, J. (2014). "Development and HVS Validation of Design Tables for Permeable Interlocking Concrete Pavement: Final Report."
Metcalf, J. (1996). "Synthesis of Highway Practice 235: Application of Full-Scale Accelerated Pavement Testing." NCHRP Synthesis, 235.
Smith, D. R. (2011). "Permeable Interlocking Concrete Pavements: Design." Specifications, Construction, Maintenance, 4th Edition, Interlocking Concrete Pavement Institute, Washington, DC.
Thompson, M., Gomez-Ramirez, F., and Bejarano, M. "ILLI-PAVE based flexible pavement design concepts for multiple wheel heavy gear load aircraft." Proc., Ninth International Conference on Asphalt Pavements.
Tutumluer, E., Kim, I. T., and Santoni, R. L. (2004). "Modulus anisotropy and shear stability of geofiber-stabilized sands." Transportation Research Record: Journal of the Transportation Research Board, 1874(1), 125-135.
Tutumluer, E., and Seyhan, U. (1999). "Laboratory determination of anisotropic aggregate resilient moduli using an innovative test device." Transportation Research Record:
Journal of the Transportation Research Board, 1687(1), 13-21.
Ullidtz, P., Harvey, J., Basheer, I., Jones, D., Wu, R., Lea, J., and Lu, Q. (2010). "CalME, a Mechanistic-Empirical Program to Analyze and Design Flexible Pavement Rehabilitation." Transportation Research Record: Journal of the Transportation Research Board, 2153(1), 143-152.
Wnek, M. A., Tutumluer, E., Moaveni, M., and Gehringer, E. (2013). "Investigation of aggregate properties influencing railroad ballast performance." Transportation Research Record: Journal of the Transportation Research Board, 2374(1), 180-189.
Expressway in Guangdong
Xiaoliang Mei1,2; Shuo Lin3; Xinwei Li4; and Xiaoge Tian5
1Chang’an University, Xi’an, Shaanxi 710064, China. E-mail: [email protected]
2Guangdong Communication Department, Guangzhou, Guangdong 510101, China.
3Guangzhou Expressway Co. Ltd., Guangzhou, Guangdong 510288, China.
4South China University of Technology, Guangzhou, Guangdong 510640, China (corresponding author). E-mail: [email protected]
5Changsha University of Science and Technology, Changsha, Hunan, China.
Abstract: The typical structures of frequently used semi rigid base, asphalt treated macadam laying on semi-rigid base and graded broken stone laying on semi-rigid base in Guangzhou-Heyuan Expressway were analyzed on structural stress.
Comparative analyses on the load response characters with asphalt pavement structure layers with different base were carried out. Combined effects of gradation, modulus, and thickness of Asphalt treated macadam layer with field investigation of composite base, design schemes of asphalt pavement with rigid-flexible composite base were proposed for expressway in Guangdong.
INTRODUCTION
Most area of Guangdong Province were in inter-tropical and subtropical climate, that annual mean temperature was 9~16℃ from north to south, and the summer mean temperature was 28~29℃. The average annual rainfall was 1500~2000mm in Guangdong belonging to rainy region with obvious seasonal changes. Base on practice and application experiences, asphalt pavement structure on semi-rigid base with reliable technique and reasonable economy, was widely adopted to expressways in Guangdong, that was suitable for various traffic environment, geological condition and all highways. So asphalt pavement structure on semi-rigid base was recommended forms in ‘Application Technique Guidelines for Typical Structure of Pavement in Guangdong Province’(Transportation department of Guangdong
185
Province 2008). Flexible base asphalt pavement was also used in some expressways, and firstly used in Guangzhou-Shenzhen expressway with a long trial under large traffic volumes (Ex: 2013, AADT= 480,000). In recent years, experiment sections of flexible base has been successively constructed and tested in Yue-Gan Expressway, Yu-Zhan Expressway, Guangzhou-Heyuan Expressway and Foshan 1st Ring Rd.
Three different types of structures，such as semi-rigid base，rigid-flexible composite base in Guangzhou and Huizhou sections of Guangzhou-Heyuan Expressway ， were chosen for structural stress analyses. To proposed typical design scheme of asphalt pavement with Rigid-flexible composite base suitable for Guangdong, comparative analyses on the load response characters of asphalt pavement structure layers with different bases were carried out.
TYPICAL STRUCTURES FOR ANALYSIS
Based on the typical structures of frequently-used semi rigid base, asphalt stabilized macadam laying on semi-rigid base in Guangzhou section and graded broken stone laying on semi-rigid base in Huizhou section of Guangzhou-Heyuan Expressway were chosen for analyses. Three typical forms of asphalt pavement structures were showed in table 1:
Table 1 Three typical structures of asphalt pavement with different base Pavement with semi
rigid base（I）
Asphalt stabilized macadam layer laying on
semi-rigid base（II）
graded broken stone laying on semi-rigid
base（III）
AC-13C (4cm) GAC-13C (4cm) AC-13 (5cm)
AC-20C (6cm) GAC-20C (6cm) AC-20 (7cm)
AC-25 (8cm) GAC-25 (8cm) AC-25 (8cm)
Cement Stabilized
Macadam (20cm) ATB-25 (8cm) ATB-25 (8cm)
Cement Stabilized Macadam (20cm)
Cement Stabilized Macadam (20cm)
graded broken stone (32cm) non-screening macadam
(15cm)
Cement Stabilized Macadam (20cm)
Cement Stabilized Macadam (15cm) subgrade
non-screening macadam
(15cm) earth subgrade
subgrade
The prescribed value of highway asphalt pavement design criterion was showed in table 2:
Table 2 Mechanics parameter of structural materials Name Modulus（MPa）
Modified Asphalt SMA-13 1400 Modified Asphalt AC-13 1200 Modified Asphalt AC-20 1400 heavy traffic asphalt AC-25 1200 heavy traffic asphalt ATB-25 1000 Cement Stabilized Macadam( 5%
cement)
1500
Cement Stabilized Macadam ( 4%
cement)
1300
graded broken stone 200
subgrade 40 STRESS ANALYSIS ON PAVEMENT STRUCTURE
By Finite Element Analysis program ANSYS( LI Sheng 2013; Xie Pengyu 2013;YUE Peng 2010; DENG Feng-xiang 2015; Long Liqin 2014), the maximum and minimum values of horizontal tensile stress σx and σy, vertical compression stress σz, shear stressτxz at different depth under edge and center of wheel clearance , center of single wheel, were obtained and showed in fig.1.
horizontaltensile stress σx horizontal tensile stress σy
Depth (m)
Depth (m)