Geosynthetics 2015

February 15-18, Portland, Oregon

Evaluation of Geosynthetic Reinforced/Stabilized Pavement Built over Soft Subgrade Soil Using Cyclic Plate Loading Testing

M. Abu-Farsakh, P.E., Louisiana Transportation Research Center, Louisiana State University, USA, cefars@lsu.edu S. Hanandeh, Department of Civil and Environmental Engineering, Louisiana State University, USA Q. Chen, P.E., Louisiana Transportation Research Center, Louisiana State University, USA L. Mohammad, Louisiana Transportation Research Center, Louisiana State University, USA ABSTRACT The benefit of using geosynthetics to enhance the performance of pavement constructed over the soft subgrade was evaluated using cyclic plate load testing. A total of six test sections, with varying types and layers of geosynthetics and base thickness, were constructed inside a 2 m x 2 m x 1.7 m steel box. A cyclic load at a frequency of 0.77 Hz was applied through a 305 mm-diameter steel plate. The test sections were instrumented by a variety of sensors to measure load associated pavement response and performance. The test results showed the benefits of geosynthetics in significantly reducing the pavement rutting. The test section with double layer geosynthetics performs much better than all other sections studied in this paper. Geosynthetics placed at base-subgrade interface function more as weak subgrade stabilization than as base layer reinforcement. 1. INTRODUCTION Weak soil is a common problem in road construction. Whether it is a temporary access road or a permanent road built over a weak subgrade, a large deformation of the subgrade can lead to deterioration of the paved or unpaved surface. Due to the soft nature of the soil and the presence of a high ground water table, cement or lime is usually used to treat soil subgrade in the state of Louisiana. However, geosynthetics offer a potentially economical alternative solution for stabilizing roads built over weak soil. The concept of using geosynthetics as reinforcement in roadway construction started in the 1970s.Since then several experimental studies were conducted to evaluate the benefits of using geosynthetics in road construction (e.g., Hass et al.1988; Al-Qadi et al. 1994, 2008; Cancelli et al. 1996; Perkins, 1999, 2002; Berg et al. 2000; Leng and Gabr 2002; Tingle and Jersey 2005; Chen et al. 2009; Perkins et al. 2009; Abu-Farsakh and Chen 2011). Two types of geosynthetics products, geotextile and geogrid, are normally used in these studies. Literature results revealed that geosynthetics can extend the pavement service life and/or reduce the base layer thickness. The geosynthetic type, the location of geosynthetics, the base thickness, and the subgrade strength have significant effect on the performance of geosynthetic reinforced flexible pavement (e.g., Collin et al. 1996; Kinney et al. 1998; Perkings 1999; Al-Qadi et al. 2008). With the pavement design moving toward Mechanistic-Empirical based methods, quantifying the benefits of geosynthetics and incorporating these benefits into Mechanistic-Empirical Pavement Design Guide (MEPDG) has received a lot of attention recently (e.g., Perkins et al. 2009; Chen and Abu-Farsakh 2012) and should be further studied. 2. FORMATTING DETAILS The main objective of this research is to evaluate the benefits of using geosynthetics to reinforce base layer and/or stabilize weak subgrade soil in flexible pavement application. For this purpose, six large-scale laboratory cyclic plate load tests were conducted to examine the effect of geosynthetics types, base course thickness, and number of geosynthatic layers on the performance of geosynthetic reinforced flexible pavement. A variety of sensors were installed for each section to measure load-associated pavement response and performance, which could be used to quantify the benefits of geosynthetics in the framework of MEPDG. 3. TESTING PROGRAM 3.1 Test Sections Six Test sections were constructed in a steel box with dimension of 2 m (6.5 ft.) (Length) × 2 m (6.5 ft.) (Width) × 1.7 m (5.5 ft.) (Height). Figure1 shows a typical pavement test section with geometric dimensions and layout of instrumentations used in this study.

823

Each test section was constructed with 1.06 m of heavy clay to represent the weak natural subgrade soil. The subgrade layer was constructed by mixing the soil with a certain amount of water to achieve the target moisture content and then compacting it to the target dry density with compaction lift thickness of 152 mm. A 305 mm thick geotextile–wrapped sand embankment was then constructed for Section 1. Sections 2 and 3, which have same base layer thickness of 457 mm, were reinforced by the geogrid. While both sections 2 and 3 have a layer of geogrid placed at the base-subgrade interface, there is an additional layer of geogrid installed at the upper one-third of the base layer for Section 2. Section 4 is a typical unreinforced control section with 457 mm thick base layer. The high-strength woven geotextile, placed at the base–subgrade interface, were used to reinforce Sections 5 and 6, which have base layer thickness of 457 mm and 254 mm respectively. The summary of configurations of each test section is presented in Table 1.

Note: All units in mm

Figure 1 The indoor test box and load actuator for cyclic load testing

Table 1 Summary of Test Sections

Test Section

Geosynthetic Configuration

Nominal Base Thickness (mm)

Nominal HMA Thickness (mm)

Section 1 305 mm non-woven geotextile-wrapped sand embankment between the base and subrade

254 76

Section 2 One layer GG @ base-subgrade interface and one layer GG at the upper 1/3 of base

457 76

Section 3 One layer GG @ base-subgrade interface 457 76

Section 4 No reinforcement 457 76

Section 5 One layer GT @ base-subgrade interface 457 76

Section 6 One layer GT @ base-subgrade interface 254 76

GG: Geogrid, GT: Geotextile

824

3.2 Material Properties 3.2.1 Subgrade

The subgrade soil has a liquid limit of 88 and a plastic index of 53 with 96.6 % passing # 200. It is classified as CH per Unified Soil Classification System (USCS) or A-7-6 according to the American Association of State Highway and Transportation Officials (AASHTO) classification system. The subgrade soil has an optimum moisture content of 35% and maximum dry density of 1,250 kg/m

3 (78 Ib/ ft

3) according to the standard Procter test. To simulate weak subgrade

condition, the target moisture content and dry density of subgrade were set as 48% and 1,114 kg/m3, respectively, during

construction. 3.2.2 Base Course

The base course material used in this study is Mexican crushed limestone, which has a coefficient of uniformly (Cc) and uniformity coefficient (Cu) equal to 3 and 37, respectively. This crushed limestone is classified as GW and A-1-a according to the USCS and the AASHTO classification system. The maximum dry density, as determined by the modified Proctor test, was 2,066 Kg/m

3 (140 Ib/ft

3) at optimum moisture content of 9.4%. The target moisture content and

dry density during construction were set as 7.5% and1,983 Kg/m3, respectively.

3.2.3 Hot Mix Asphalt

The HMA used in the construction is a wearing course. It is design level 1 Level 1 (< 3 million ESALs) superpave mixture, which has a nominal aggregate size of 12.5 mm (0.5 in). The asphalt binder was classified as PG 76-22M according to the Performance Grade (PG) specification. The optimum asphalt binder content is 4.1 percent. The theoretical maximum density of HMA is 155 Pcf. 3.2.4 Geosynthetics

Two types of geosynthetics were used in this research, a Triaxial geogrid, GG, and a high-strength woven geotextile, GT. The physical and mechanical properties of these geosynthetics as provided by the manufacture are presented in Table 2.

Table 2 Physical and mechanical properties of geosynthetics used in this study

Index Property GG GT

MD* XMD*

Polymer Type Polypropylene Polypropylene Aperture Size (mm) 40×40×40 N/A N/A Ultimate Tensile Strength (kN/m) N/A 7 26.26 Radial Stiffness @ 0.5% strain (kN/m) 270 N/A N/A Junction Efficiency 93% N/A N/A

* MD: machine direction; XMD: cross-machine direction. 3.3 Instrumentation Various types of instrumentations, as showing in Figure1, were installed at different locations within pavement layers to measure the load associated pavement response and performance. These include the pressure cell to measure the total vertical stress at the top of subgrade, the piezometer to measure the possible excess pore water pressure in the subgrade, the customized potentiometer to measure the compressive strain at the mid–height of the base course layer, the customized LVDT to measure the total deformation of subgrade layer, strain gauges to measure the strain distribution along the geosynthetics, and LVDTs to measure the surface deformation of pavement test sections. 3.4 Construction Control During construction of test sections, the nuclear density gauge, Geogauge, and vane shear test device were deployed to measure the in place properties of subgrade to ensure the quality of subgrade construction. The Dynamic Cone Penetrometer (DCP), Light Falling Weight Deflectometer (LFWD), nuclear density gauge, and Geogauge were used to measure the in-place properties for construction quality control of base course. The LFWD, Pavement Quality Indicator (PQI) and Portable Seismic Property Analyzer (PSPA) were used to measure the in-place properties of HMA. At least five measurement were made for each property.

825

For the subgrade layer, the dry densities varied from 1,092 to 1,112 kg/m3, with moisture contents ranging from 48.11%

to 48.9%. The Geogauge moduli and undrained shear strength were in the range of 21 to 23 MPa and 42 to 52 kPa, respectively. For the base course layer, the dry densities varied from 1,936 to 1,989 kg/m

3, with moisture contents

ranging from 6.9% to 7.9 %. The LFWD moduli, geogauge moduli, and DCP index were in the range of 80 to 125 MPa, 114 to 132 MPa, and 5.3 to 6.5 mm/blow, respectively. For the HMA layer, The LFWD moduli and PSPA shear moduli were in the range of 220 to 288 MPa and 16,556 to 23,476 MPa, respectively. The air voids and density from PQI measurement were in the range of 6 to 12% and 2,066 to 2,274 kg/m

3, respectively. The air voids, density and indirect

tensile strength of HMA were also obtained from the core sample taken after test. They were in the range of 6 to 10%, 2,242 to 2,402 kg/m

3, and 2,269 to 3,781 MPa, respectively.

3.5 Cyclic Load Testing A hydraulic actuator was used to apply cyclic loads to the pavement test section through a loading plate sitting on the surface of the HMA. The loading plate was a 25.4 mm thick steel plate with 305 mm in diameter. The cyclic loading in this study consists of repeated cycles of a loading pulse which has a linear load increase from 22 kN (500 Ib) to the maximum load in 0.3 second, followed by a 0.2-second period where the load is held constant at the maximum, followed by a linear load decrease to 2.2 KN (500 lb) over 0.3-second period, and then followed by a 0.5-second period of 2.2 KN (500 lb) before the next loading cycle. This results in a loading frequency of 0.77 Hz. The tests were originally designed to apply a maximum load of 40 kN (9,000 Ib) for 100,000 cycles, followed by a maximum load of 53 kN (12,000 Ib) for additional 100,000 cycles, followed by a maximum load of 67kN (15,000 Ib) for another additional 100,000 cycles, and then followed by a maximum load of 80 KN (18,000 Ib) for last 100,000 cycles or until reaching 25.4 mm rut depth. However, due to the breakdown of pump system during the test, the actual load applied never reached 67kN (15,000 Ib) in some tests and 80 kN (18,000 Ib) in other tests. Meanwhile, due to the power outage, several tests stopped and restarted a couple of times. As such, to better compare the performance of each section, the test data were converted to ESALs using the fourth power rule. 4. TEST RESULTS AND ANALYSIS 4.1 Pavement Surface Permanent Deformation The surface permanent deformation, which was determined by averaging the reading of two LVDTs rest on top of the loading plate, are shown in Figure 2 for the six test sections. The results show that the surface permanent deformation accumulated with the number of load cycles; Sections constructed with geosynthetics show less rut depth as compared to the unreinforced section; and more reduction in the pavement surface deformation was observed for double layer reinforcement section (Section 2). It can be seen from the figure that Sections 3 can sustain 3,356,587 at rut depth 25 mm and section 5 can sustain 3,439,878 ESALs at a rut depth of 25 mm which result a traffic benefit ratio (TBR) of 2.5 and 2.9, respectively. The TBR is defined as the number of load cycles carried by a reinforced section at a specific rut depth divided by that of an equivalent unreinforced section. Section 2, with double layer reinforcement, never reached 25 mm rut depth during the test and the maximum rut depth of 12 mm was obtained at 4,193,883 ESALs. Section 1 performed worse than section 4, which indicates that 203 mm thick Mexican limestone is more effective than 305 mm thick nonwoven geotextile-wrapped sand embankment in reducing the pavement rutting. Section 6 has the worst performance among all six sections because of its much thinner base layer (254 mm vs 457 mm).This also suggests that 203 mm thick base layer is more effective in reducing pavement rutting than a layer of high-strength woven geotextile at base-subgrade interface. 4.2 Permanent Deformation in Base and Subgrade As previously mentioned, in the base layer, the customized potentiometer was installed at the mid-height to measure the compressive strain. It is assumed here that the compressive strain at the mid-height represents the mean compressive strain of the whole layer. As such, the overall deformation of the entire base layer was determined by multiplying the measured compressive strain by the thickness of base layer. In the subgrade layer, a customized LVDT was mounted on a steel rod, which has an end fixed to the bottom of the steel box. Therefore, the deformation measured by the LVDT is the overall defamation of the entire subgrade layer. Figures 3 and 4 illustrate the development of base and subgrade permanent deformation with number of load cycles. As can be seen from the figures, for control section (Section 4), the subgrade layer makes much more significant contribution to the total permanent deformation, when compared to the base layer. The geosynthetics can reduce permanent deformation both in base and subgrade layer. However, when the geosynthetic is placed at the base–subgrade interface, significant reduction of permanent deformation occurred in subgrade layer while only small reduction of permanent deformation was observed in base layer. This means that the geosynthetic at the base-subgrade interface mostly enhance the performance of weak subgrade, i.e., it functions more as weak subgrade stabilization than as base aggregate reinforcement. For double layer reinforcemnt (Section 2), the permanent deformation were signicantly reduced both in base and subgrade layer. This suggests that while the

826

performance of subgrade was signicantly improved by the geogrid at the base-subgare interface, the performance of base layer was enhanced by the geoegrid placed at the upper one–third of base layer.

Figure 2. Development of Surface Permanent Deformation

Figure3: Development of Permanent Deformation in Base

Figure 4: Development of Permanent Deformation in Subgrade

827

4.3 Stress Distribution within Subgrade The vertical stress distribution at the subgrade/base interface and along the centerline of the loading plate for both reinforced and unreinforced sections are shown in Figure 5. The stresses measured were the total vertical stresses induced by the peak load during each cycle, and the stresses induced by the weight of the soil are not included. Because the customized LVDT used to measure the deformation of subgrade layer was installed at the location underneath the center of the loading plate, the stress measurement at this location was not available. However, from the figure, one can see that the magnitude of vertical stress was increased away from the plate in the reinforced test sections compared to unreinforced sections. This may suggest that the load was redistributed to a wider area in the reinforced test sections, resulting in an improved stress distribution on top of the subgrade layer when geosynthetics were installed.

Figure 5 Vertical Stress Distribution 5. CONCLUSIONS A large-scale plate load testing program was carried out to evaluate the benefits of two geosynthetic products, a triaxial geogrid and a high strength woven geotextile, on the performance of pavement built over the weak/soft subgrade. A total of six test sections, with varying base thickness and number of geosynthetic layers, were constructed and instrumented to measure the load associated pavement response and performance. The test results demonstrate that both geosynthetic products significantly improved the performance of the pavement section in term of reducing the surface permanent deformation. The geosynthetics also result in redistributing the applied load to a wider area, thus reducing the stress concentration and achieving an improved vertical stress distribution on top of subgrade layer. The pavement section with double layer reinforcement had the best performance among all six sections tested in this study. Instrumentation measurements indicate that geosynthetics placed at base-subgrade interface function more as weak subgrade stabilization than as base aggregate reinforcement. By placing an additional layer of geogrid at the upper one-third of the base layer, the permanent deformation in base layer can also be significantly reduced. ACKNOWLEDGEMENTS The authors acknowledge and appreciate the financial support provided by the Louisiana Department of Transportation and Development (LA DOTD), Tensar International., and TenCate. The authors also wish to thank the personnel at Pavement Research Facility of Louisiana Transportation Research Center (LTRC) and the graduate students at LTRC who helped with installing instruments and in-situ testing. REFERENCES Abu-Farsakh, M. and Chen, Q. (2011). Evaluation of Geogrid Base Reinforcement in Flexible Pavement Using Cyclic Plate Load Testing, International Journal of Pavement Engineering, Vol. 12 Issue 275-288. Al-Qadi, I.L., Brandon, T.L., Valentine, R.J., Lacina, B.A., and Smith, T.E. (1994). Laboratory Evaluation of Geosynthetics Reinforced Pavement Sections, Transportation Research Record: Journal of the Transportation Research Board, No. 1439, TRB, National Research Council, Washington DC. 25-31.

0

5

10

15

20

25

30

35

40

0 0.5 1 1.5 2 2.5

Press

ure (

kP

a)

Relative Distance from the Center of Plate (D)

Section 2

Section 3

Section 4

Section 5

828

Al-Qadi, I.L., Dessouky, S.H., Kwon, J., and Tutumluer, E. (2008). Geogrid in flexible pavements: validated mechanism. Transportation Research Record: Journal of Transportation Research Board 2045, pp. 102-109. Berg, R. R., Christopher, B.R., and Perkins, S.W. (2000). Geosynthetic reinforcement of the aggregate base course of flexible pavement structures, GMA White paper II, Geosynthetic material Association, Roseville, MN, USA, 100. Cancelli, A., Montanelli, F., Rimoldi, P., and Zhao, A. (1996). Full Scale Laboratory Testing on Geosynthetics Reinforced Paved Roads, Proceedings of the International Symposium on Earth Reinforcement. Fukuoka/Kyushu, Japan, November, Balkema, 573-578. Chen, Q. and Abu-Farsakh, M. (2012) Structural Contribution of Geogrid Reinforcement in Pavement. GeoCongress 2012.1468-1475. Chen, Q., Abu-Farsakh, M., and Tao, M. (2009). Laboratory evaluation of geogrid base reinforcement and corresponding instrumentation program. Geotechnical Testing Journal, ASTM, Vol. 32, No.6, pp. 516-525.

Collin, J.G.., Kinney, T.C., and Fu, X. (1996). Full scale highway load test of flexible pavement system with geogrid reinforced base courses. Geosynthetics International, Vol. 3, No.4, 537-549. Haas R., Wall, J., and Carroll, R.G. (1988). Geogrid Reinforcement of Granular Bases in Flexible Pavements, Transportation Research Record: Journal of the Transportation Research Board, No. 1188, TRB, National Research Council, Washington, DC, USA, 19-27. Kinney, T.C., Abbott, J., and Schuler, J. (1998). Benefits of using geogrids for base reinforcement with regard to rutting Transportation Research Record: Journal of the Transportation Research Board, No. 1611, National Research Council, 86-96. Leng, J. and Gabr, M.A. (2002). Characteristics of Geogrid-Reinforced Aggregate Under Cyclic Load Transportation Research Record 1786, Paper No. 02-4091. Perkins, S.W. (1999). Geosynthetic Reinforcement of Flexible Pavements Laboratory Based Pavement Test Sections. Federal Highway Administration Report FHWA/MT-99-001/8138, Montana Department of Transportation, Helena, Montana, USA, 109 Perkins, S.W. (2002). Evaluation of geosynthetic reinforced flexible pavement systems using two pavements test facilities, Federal Highway Administration Report FHWA/MT-02-008/20040, Montana Department of Transportation, Helena, Montana, USA, 120p. Perkins, S.W., Christopher B.R., Cuelho, E.G., Eiksund, G. R., Schwartz, C.S., and Svanø, G. (2009). A Mechanistic-Empirical Model for Base-Reinforced Flexible Pavements,” International Journal of Pavement Engineering, Vol. 10, No. 2, 101–114. Tingle, J. and Jersey, S. (2005). Cyclic Plate Load Testing of Geosynthetic-Reinforced Unbound Aggregate Roads Transportation Research Record: Journal of the Transportation Research Board, No. 1936, Transportation Research

Board of the National Academies, Washington, D.C., 2005, 60–69.

829