Quantifying the Benefits of Geosynthetics Reinforcement in Flexible Pavement Design Using Cyclic Plate Load Testing

Abstract

Development of pavement design over the past decades has focused on moving from empirical design equations to more powerful and adaptive design schemes. The AASHTO mechanistic–empirical pavement design guide (MEPDG) has been developed to model pavement structure and predict its service life more accurately. Although MEPDG has been widely implemented to design conventional pavement, it is not yet capable of predicting the service life of pavement reinforced with geosynthetics. Given the above concerns, seven full-scale test sections that were constructed at Louisiana Transportation Research Center-Pavement Research Facility (LTRC-PRF) were devoted to a structural experiment to investigate the performance of geosynthetic reinforced pavements. The benefits of using geosynthetics to enhance the performance of pavements constructed over soft subgrades was evaluated using cyclic plate load testing. Cyclic load at a frequency of 0.77 Hz was applied through a 305 mm diameter steel plate. The test results clearly show the benefits of geosynthetics in significantly reducing pavement rutting. The test section with double geosynthetic layers performed better than the other six sections studied. After eliminating the effect of variations in construction, the benefits of geosynthetic reinforcement are quantified within the context of the AASHTOWare Pavement ME Design Guide. The developed design procedure is capable of quantifying the contribution of geosynthetics in pavements to base reinforcement as well as subgrade stabilization. A design methodology is proposed that falls within the context of MEPDG.

Keywords

infrastructure geology and geoenvironmental engineering geosynthetics geosynthetic reinforcement

Weak subgrade 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. The use of cementitious materials to treat/stabilize poor subgrade is a practice conventionally adopted by many state highway agencies in the U.S.A. However, geosynthetics offer an environmentally friendly and potentially economical alternative solution for reinforcing/stabilizing roads built over weak subgrade soil.

The concept of using geosynthetics as reinforcement in roadway construction started in the 1970s. Since then, numerous studies have revealed that using geosynthetic reinforcements in pavement structures can extend the pavement’s service life, reduce the thickness of the base layer, or both ( 1 – 11 ). The geosynthetic type, the location/layers of geosynthetics, the base thickness, and the subgrade strength have significant effects on the performance of geosynthetic reinforced flexible pavements (5, 10 –17). Among various techniques in experimental studies, the cyclic plate load test, with its low cost and time savings, has been widely used by researchers to evaluate the performance of geosynthetic reinforced pavements (1, 2, 5, 7, 10, 15, 17). This type of test has also proved to be a very good performance indicator test for the evaluation of pavement test sections ( 10 ). With the benefits of geosynthetics in pavement performance improvement being widely recognized, much research on geosynthetic reinforced flexible pavement has centered on understanding the mechanism of geosynthetic reinforcement ( 12 – 23 ).

The literature reveals considerable research aimed at developing design guidelines, empirical relationships, and/or design methods for certain conditions and design requirements for pavement reinforcement. However, no method or specification for design of flexible pavements with geosynthetic reinforcement has been universally agreed on, and the design/analysis of such structures is still being investigated. Several articles and reports are available that summarize the state of the art and current practice in this area (4, 5, 14, 15, 20, 24). However, many of these mechanistically based studies tended to employ complex finite element or discrete element methods, particularly for the reinforcement. The latest AASHTO mechanistic-empirical pavement design guide (MEPDG) ( 25 ) does not consider geosynthetic reinforced pavement sections either. This is mainly attributed to lack of understanding of the reinforcing mechanisms of geosynthetic reinforcement, especially failure to rigorously quantify the benefits of geosynthetic reinforcement. The other factors inhibiting the quantification of the contribution of geosynthetic reinforcement include: unavailability of geosynthetic characteristics that relate their properties to the performance of geosynthetics in pavements; inappropriate modeling of the geosynthetic interlayer systems; the interaction between geosynthetics and surrounding layers; the response of geosynthetics to a moving load under confined conditions; and the effect of the environment on the long-term performance of geosynthetics in pavements ( 13 ). The limited availability of design methods that mechanistically incorporate geosynthetics in pavements, and the unclear cost effectiveness and efficiency of geosynthetics when used in pavements, have resulted in reluctance to accept the technology.

These limitations provide the motivation for this research study. The main objective of this research was to evaluate the benefits of using two types of geosynthetics—geogrid and high strength geotextile—to reinforce base aggregate layer and/or to stabilize weak subgrade soil in flexible pavement applications and to quantify these benefits within the framework of AASHTO PavementME design. This was achieved through conducting cyclic plate load testing on full-scale geosynthetic reinforced pavement test sections. Different types and configurations of geosynthetics were considered for reinforcement. The test results were used to quantify the structural benefits of geosynthetic as a reinforcement effect and subgrade stabilization.

Experimental Testing Program

Cyclic Plate Load Tests

A hydraulic actuator, which has a force rating of 22 kips (100 kN) and a dynamic stroke of 6 in. (152.4 mm), was placed between two I-beams of the crosshead. A field frame consisting of two I-beams was constructed to support the crosshead. Figure 1a presents a photograph of the cyclic plate loading test setup. The cyclic load was applied through a steel rod that fits into a concave-shaped hole on the loading plate that sits on the surface of the hot mix asphalt (HMA) layer. The loading plate was a 1 in. (25 mm) thick steel plate, 12 in. (305 mm) in diameter. The maximum applied load in tests was 9,000 lb (40 kN), which results in a loading pressure of 80 pounds per square inch (psi) (550 kPa) and simulates dual wheels under an equivalent 18,000 lb (80 kN) single axle load. The load pulse, as shown in Figure 1b, has a linear load increase from 500 lb (2.2 kN) to 9,000 lb (53 kN) in 0.3 s, followed by a 0.2 s period where the load is held constant at 9,000 lb (53 kN), followed by a linear load decrease to 500 lb (2.2 kN) over a 0.3 s period, then followed by a 0.5 s period of 500 lb (2.2 kN) (rest period), before the next loading cycle is applied. This load pulse results in a frequency of 0.77 Hz (Figure 1b).

Figure 1.

Cyclic plate load test: (a) testing facility, (b) load pulse, and (c) cross-section of test sections with target thickness.

Pavement Test Sections

A total of seven test sections were used in this study. Six full-scale test sections were 80 ft (24 m) long and 13 ft (4 m) wide. In addition to the six full-scale test sections, one more section (Section 7), measuring 12 ft (3.7 m) long and 12 ft (3.7 m) wide, was constructed as the control section of Section 6. The cross-sections of the test lanes are presented in Table 1. Sections 1, 4, and 7 are the unreinforced sections that were constructed without geosynthetic reinforcements, of which Section 1 was constructed over 1 ft (0.3 m) thick sand embankment, which is a common practice in Southern Louisiana. Sections 4 and 7 are two typical control sections without geosynthetics. Sections 2 and 3 were reinforced/stabilized by triaxial geogrid placed at the aggregate base–subgrade interface. An additional layer of geogrid reinforcement was also installed at the upper one-third of the aggregate base layer thickness in Section 2. High strength geotextiles were used to reinforce/stabilize Sections 5 and 6, which had different thicknesses of aggregate layer. A 3 in. (7.6 cm) HMA surface course was later constructed over the test lane sections.

Table 1.

Test Sections With Target Layer Thicknesses

Section	Description	Geosynthetic location	Base	Hot mix asphalt
Section 1	NA	12′′ (305 mm) compacted geotextile-wrapped sand layer	10′′ (254 mm)	3′′ (76 mm)
Section 2	Double geogrid-reinforced section	One layer at base–subgrade interface and one layer at the upper one-third	18′′ (457 mm)	3′′ (76 mm)
Section 3	Geogrid-reinforced section	One layer at base–subgrade interface	18′′ (457 mm)	3′′ (76 mm)
Section 4	Control section	No reinforcement	18′′ (457 mm)	3′′ (76 mm)
Section 5	Geotextile reinforced section	One layer at base–subgrade interface	18′′ (457 mm)	3′′ (76 mm)
Section 6	Geotextile reinforced section	One layer at base–subgrade interface	10′′ (457 mm)	3′′ (76 mm)
Section 7	Control section	No reinforcement	10′′ (457 mm)	3′′ (76 mm)

Note: NA = not available.

Pavement Layer Materials

Subgrade

The native subgrade soil consisted of a high plasticity clay, having 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 clay has an optimum moisture content of 35% and a maximum dry density of 78.1 pounds per cubic foot (pcf) (1,250 kg/m³) according to the standard Proctor test.

Base Course Material

Mexican crushed limestone material was used in the base course layer for all test sections. The crushed limestone had 1.56% passing No. 200 opening sieve, an effective particle size (D10) of 0.382 mm, a mean particle size (D50) of 3.126 mm, a D85 of 19 mm, a uniformity coefficient of 37, and a coefficient of curvature of 3. This crushed limestone is classified as GW and A-1-a according to the USCS and the AASHTO classification systems, respectively. The maximum dry density as determined by the modified Proctor test is 129 pcf (2,066 kg/m³) at an optimum moisture content of 9.4%. The resilient modulus test was performed in the laboratory using the Material Testing System (MTS) machine on base material. Figure 2 depicts the variation of resilient modulus with stress conditions.

Figure 2.

Resilient modulus of base material.

HMA Concrete

The HMA used in the construction is a wearing course. It is a 0.5 in. (12.5 mm) design level 1 Superpave mixture. The asphalt binder was classified as PG 76-22M according to the Performance Grade (PG) specification. The optimum asphalt binder content is 4.1%. The theoretical maximum density of HMA is 154.8 pcf (2,480 kg/m³).

Geosynthetics

Two types of geosynthetics were used in this research, a triaxial geogrid and a high strength woven geotextile. The triaxial geogrid was made by means of punching and drawing polypropylene sheets. The geotextile was made from high-tenacity polypropylene filaments that are formed into weaves. The physical and mechanical properties of these geosynthetics were reported by the manufacturer as follows. The tensile strength of the geogrid with aperture size of 1.57 × 1.57 × 1.57 in. (40 × 40 × 40 mm) at 0.5% strain is 7.71 lb/in. (1.35 kN/m). The geotextile shows tensile strength at 2% strain of 39.9 lb/in. and 150 lb/in. (7 kN/m and 26.3 kN/m) at machine direction and cross direction, respectively ( 10 , 11 ). Index properties of the nonwoven geotextile separator used in Sections 1 and 4 were not tested. However, the particular product was chosen to meet the requirements of Class C geotextile according to the Louisiana Department of Transportation and Development’s standard specifications for this type of application ( 26 ). Class C geotextile has the following minimum values of index properties: grab tensile strength of 3.3 lb/in. (580 N/m), elongation of 50% at failure, and burst strength of 5.3 lb/in. (930 N/m). The Class C geotextile’s strength is much lower than the geogrid’s and its primary functions were separation and filtration; whether the nonwoven geotextile had any reinforcing effects on the pavement’s performance is not known.

In-Situ Pavement Layer Properties

The in-situ material properties of base and subgrade were measured during construction with several devices such as nuclear density gauge, GeoGauge, light falling weight deflectometer (LFWD), and dynamic cone penetrometer (DCP). Asphalt cores were obtained to measure the in-place properties of the HMA. At least five measurements were performed for each property. Tables 2 and 3 show some of the in-situ test measurements for the subgrade and base course layers, respectively.

Table 2.

Subgrade In-Situ Properties

Test section	Moisture content (%)		Dry density (pcf)		Light falling weight deflectometer (tsf)		GeoGauge (tsf)		DCPI (in./blow)
	Avg.	Cov.	Avg.	Cov.	Avg.	Cov.	Avg.	Cov.	Avg.	Cov.
1	49.3	0.05	69.4	0.04	37.6	0.32	145.1	0.14	7.3	0.28
2	46.0	0.07	71.3	0.02	47.0	0.16	214.0	0.07	7.2	0.15
3	49.5	0.11	68.4	0.07	42.8	0.30	215.1	0.10	6.6	0.17
4	49.5	0.12	69.3	0.04	41.8	0.26	190.0	0.15	6.5	0.12
5	51.2	0.05	68.1	0.03	41.8	0.18	219.2	0.13	7.1	0.07
6	46.5	0.04	70.8	0.02	51.2	0.17	239.1	0.06	5.7	0.35
7	48.3	0.22	69.8	0.04	46.0	0.18	217.2	0.08	5.5	0.25

Note: Avg. = average; Cov. = coefficient of variation; DCPI = dynamic cone penetration index; pcf = pounds per cubic foot; tsf = ton per square foot.

Table 3.

Base Layer In-Situ Properties

Test section	Moisture content (%)		Dry density (pcf)		Light falling weight deflectometer (tsf)		GeoGauge (tsf)		DCPI (in./blow)
	Avg.	Cov.	Avg.	Cov.	Avg.	Cov.	Avg.	Cov.	Avg.	Cov.
1	8.8	0.03	122.9	0.01	1200.6	0.17	1261.2	0.12	0.38	0.11
2	7.6	0.07	122.6	0.01	1306.0	0.14	1325.9	0.16	0.32	0.10
3	6.9	0.04	125.5	0.02	1273.7	0.14	1331.1	0.13	0.24	0.17
4	8.2	0.06	122.7	0.02	1279.9	0.12	1712.2	0.18	0.25	0.13
5	5.0	0.09	123.0	0.01	1299.8	0.20	1784.2	0.13	0.27	0.21
6	9.7	0.05	125.6	0.01	786.1	0.14	1320.7	0.22	0.28	0.24
7	9.2	0.03	125.3	0.01	823.9	0.13	1417.1	0.18	0.31	0.24

Note: Avg. = average; Cov. = coefficient of variation; DCPI = dynamic cone penetration index; pcf = pounds per cubic foot; tsf = ton per square foot.

Before the cyclic plate testing, three DCP tests were conducted around the cyclic plate test location for each test section. A core sample was taken from each section to determine the thickness, air voids, density, and dynamic modulus of the HMA. The values of the properties, which are related to the AASHTOWare Pavement ME design inputs, are summarized in Table 4 for all pavement test sections. The resilient moduli of base and subgrade, evaluated using the relationship of California bearing ratio (CBR) to dynamic cone penetration index (DCPI) suggested by Webster et al. ( 27 ), and that of resilient modulus (M_r) to CBR suggested by Powell et al. ( 28 ) (Equation 1), are also presented in Table 4.

\begin{matrix} CBR = 292 / {(DCPI)}^{1.12} \\ M_{r} (MPa) = 17.58 \times CB R^{0.64} \end{matrix}} \to^{yields} M_{r} (MPa) = 665.06 / {(DCPI)}^{0.7168}

(1)

Table 4.

In-Situ Properties of Subgrade, Base, and Asphalt Layers

Test section		Subgrade	Base		Asphalt
		M_r (psi)	Thickness (in.)	M_r (psi)	Thickness (in.)	Density^a (pcf)	Air voids^a (%)	E* (ksi)
Section 1	Mean	2286	10.3	19,162	3.45	149.9	3.7	555.7
	CV (%)	NA	6.5	NA	NA	NA	NA	NA
Section 2	Mean	2311	18.95	21,694	3.21	146.67	5.75	461.2
	CV (%)	NA	7.2	NA	NA	NA	NA	NA
Section 3	Mean	2453	19.4	26,258	2.82	145.47	6.53	452.8
	CV (%)	NA	5.9	NA	NA	NA	NA	NA
Section 4	Mean	2480	19.3	26,046	3.01	145.77	6.33	461.7
	CV (%)	NA	7.7	NA	NA	NA	NA	NA
Section 5	Mean	2345	20.7	24,207	3.25	148.90	4.34	445.8
	CV (%)	NA	6.8	NA	NA	NA	NA	NA
Section 6	Mean	2716	11	23,480	3.30	145.90	6.23	475.2
	CV (%)	NA	7.2	NA	NA	NA	NA	NA
Section 7	Mean	2793	10.5	21,171	3.23	145.90	6.23	475.2
	CV (%)	NA	7.3	NA	NA	NA	NA	NA

Note: CV = ; psi = pounds per square inch; ksi = kips per square inch; pcf = pounds per cubic foot; M_r = Resilient modulus, estimated from the dynamic cone penetration index (DCPI); Thickness = measured thickness; E* = dynamic complex modulus at a temperature of 30^oC and a loading frequency of 10 Hz; NA = not available.

Determined from the core samples taken from each test lane.

Test Results

Measured Permanent Deformation Behavior of Pavement Sections

Figure 3 illustrates the development of rut depth (permanent deformation) with number of equivalent single axle loads (ESALs) for the seven pavement test sections. The numbers of loading cylces were coverted to ESALs using the fourth power rule as follows.

ESALs = N \times {(\frac{Load}{40 kN})}^{4}

(2)

where N is the number of loading cycles. The results obtained for the different control and reinforced pavement test sections are also summarized in Table 5. The surface permanent deformation was calculated by averaging the readings of two linear variable differential transformers (LVDTs) resting on top of the loading plate. The results show that the surface permanent deformation accumulated with the number of ESALs; sections constructed with geosynthetics experienced less rut depth as compared with the control section; and more reduction in the pavement surface deformation was observed for the double layer reinforcement section (Section 2). The results of in-situ cyclic plate load testing showed that Section 1, which is a common practice in Southern Louisiana, was outperformed by Section 4 (the control section); therefore, the Section 1 configuration is not recommended for future practice. Sections 2, 3, and 5 can sustain, respectively, 2,212,143 ESALs, 1,577,043 ESALs, and 1,784,762 ESALs at a rut depth of 0.75 in. (19.1 mm), which results in traffic benefit ratios (TBR) of 1.56, 1.29, and 1.46, respectively (Table 5). Meanwhile, Section 6 can sustain 92,971 ESALs at a rut depth of 0.75 in. (19.1 mm), which results in TBR of 2.37. Section 5, with a single layer of geotextile placed at the base–subgrade interface, performed better than Section 3 with a single layer of geogrid placed at the base–subgrade interface. However, it should be noted here that the geotextile used in this study has much higher tensile modulus than the geogrid used. The results of recent finite element analysis indicated that geogrids placed at the base–subgrade interface usually perform better than geotextiles of the same tensile modulus ( 29 ).

Table 5.

Summary of Cyclic Plate Load Test Results

Reinforcement configuration	Hot mix asphalt thickness^a (in.)	Base thickness^a (in)	Base modulus^b (psi)	Subgrade modulus^b (psi)	Rut depth = 0.5 in.		Rut depth = 0.75 in.
					ESALs	TBR	ESALs	TBR
Section 1	3.45	10.3	19,162	2286	NA	NA	NA	NA
Section 2	3.21	19.0	21,694	2311	1,077,865	2.08	2,212,143	1.81
Section 3	2.82	19.4	26,258	2453	723,723	1.40	1,577,043	1.29
Section 4	3.01	19.3	26,046	2480	518,653	NA	1,225,115	NA
Section 5	3.25	20.7	24,207	2345	881,797	1.70	1,784,762	1.46
Section 6	3.30	11	23,480	2716	25,110	2.41	90,971	2.37
Section 7	3.23	10.5	22,171	2793	10,416	NA	38,357	NA

Note: ESAL = equivalent single axle loads; psi = pounds per square inch; TBR = traffic benefit ratio; NA = not available.

Measured actual thickness.

From dynamic cone penetrometer.

Figure 3.

Accumulated total permanent deformation.

Effect of Variations in Pavement Layer Properties on Surface Permanent Deformation

As with the case of any construction, there are always some variations in the constructed layer thicknesses and properties of the pavement test lane sections (Tables 2 –4). To evaluate the effect of variation in the pavement layer properties on the performance of pavement test sections, the measured TBR for each geosynthetic reinforced section was adjusted through multiplying the TBR value by a verification factor (VF) that takes into consideration the construction variations of layer materials for the same reinforced section as follows:

TBR = VF \overset{'}{TB} R_{adj}

(3)

where TBR is the measured TBR; VF is the factor associated with variation in properties of pavement layer materials resulting from construction; TBR_adj is the adjusted TBR for the geosynthetic reinforced section. An analysis, using AASHTOWare Pavement ME Design software, was performed on the corresponding unreinforced condition of the reinforced test section (using actual layer properties) and the control section to quantify the impact of variation in pavement layer properties on the performance of pavement test sections (i.e., evaluate VF). The traffic load was applied using a single wheel through special axle configuration available in the AASHTOWare Pavement ME Design software. The load is 9,000 lbf (40 kN) with a tire pressure of 105 psi (724 kPa). Since each test section was loaded in a very short time, the resilient modulus of the base and subgrade of each test section was assumed to be constant through the testing period and the corresponding values are presented in Table 5. By considering the measured/estimated in-situ pavement layer’s thickness and properties, the number of ESALs needed to reach 0.5 in. (12.7 mm) and 0.75 in. (19.1 mm) rut depth was calculated for each test section. VF is evaluated here as the number of ESALs at a specific rut depth carried by the corresponding unreinforced condition of the reinforced test section (using actual layers’ properties) divided by that of the control test section. The results of the analyses are presented in Table 6. By dividing the measured TBR values by the VF factors, the TBR associated with geosynthetic reinforcement (TBR_adj) was obtained for each reinforced test section and is also presented in Table 6. It should be pointed out here that the TBR_adj values can be affected by the accuracy of modulus test results for HMA, the variation in DCP test profiles for subgrade layer, accuracy in measuring thicknesses of HMA and base layer, and reliability of the correlation equations used.

Table 6.

Effect of Differences in Constructed Layer Properties and Thickness

Test section	Rut depth = 0.5 in.			Rut depth = 0.75 in
	ESALs	VF	TBR_adj ^a	ESALs	VF	TBR_adj ^a
Section 2	32,400	0.83	2.52	169,800	0.85	2.12
Section 3	33,840	0.86	1.62	170,400	0.86	1.51
Section 4	39,240	NA	NA	199,200	NA	NA
Section 5	40,320	1.03	1.65	203,520	1.02	1.43
Section 6	34,380	1.15	2.09	174,600	1.16	2.83
Section 7	29,880	NA	NA	150,840	NA	NA

Note: ESALs = equivalent single axle loads; TBR = traffic benefit ratio; VF = verification factor; NA = not available.

TBR_adj = TBR/VF.

Structural Contribution of Geosynthetic Reinforcement

AASHTOWare Pavement ME Design software was used in this study to analyze and evaluate the performance of the pavement testing sections constructed in the field. The benefits of geosynthetic reinforcement were demonstrated through the input parameter of the mechanistic part (i.e., resilient modulus), empirical part (i.e., local calibration factor), or both, of the design method. The best match of the entire rut depth–load cycle curve, which was evaluated in relation to least square of errors, was used in this study. Geosynthetics have two primary applications in pavements: base reinforcement and mechanical subgrade stabilization. When geosynthetics are placed at the base–subgrade interface, they can function both as base reinforcement and mechanical subgrade stabilization. In this paper, three methods are proposed to incorporate the effects of geosynthetics in pavement design. In the first method, all the benefits of geosynthetic reinforcement were evaluated as base reinforcement only in relation to equivalent resilient modulus (Figure 4, a and b ). This method is easy to incorporate into both AASHTOWare Pavement ME Design software and 1993 AASHTO Pavement Design ( 30 ) and there is no need to perform extra testing or make assumptions for design. The second proposed method considers the effect of base reinforcement in the influence zone of geosynthetic reinforcement layer; and the rest of the benefit of the geosynthetics at the base–subgrade interface was evaluated as subgrade stabilization (Figure 4, a and c ). In the third method, the service life of the reinforced section is assumed to be similar to the unreinforced section, but the thickness of the base layer decreased by the base course reduction (BCR) factor. More details in relation to the proposed methods are explained below.

Figure 4.

Structural contribution of geosynthetic reinforcement: (a) geosynthetic reinforced section, (b) geosynthetic as base reinforcement, and (c) geosynthetic as both base reinforcement and subgrade stabilization.

Geosynthetic as Base Reinforcement

For design purposes, the reinforcement benefit from geosynthetics can be incorporated into the pavement’s mechanistic–empirical design by adjusting the resilient modulus (M_r) of the base layer to account for the improved stiffness of the base course. In this approach, it is assumed that the geosynthetic has no stabilization effect on the subgrade layer, that is, the subgrade resilient modulus of the reinforced sections is kept the same as the corresponding unreinforced condition. Pavement ME Design software or the 1993 AASHTO Pavement Design Guide can then be used to back-calculate the adjusting resilient modulus (M_r) of the reinforced base layer. The resilient modulus is calculated by assuming that geosynthetic reinforced and unreinforced pavements have similar service life. As shown in Figure 4b, in this approach, it is assumed that the adjusting factor to the resilient modulus of the base layer applies to the whole layer.

There are always differences between the number of load cycles obtained from the experimental program and those estimated from the Pavement ME Design for a specific rut depth. This is caused by many factors, such as differences in loading conditions. The AASHTOWare Pavement ME was calibrated using the Long Term Pavement Performance (LTPP) database which was collected from several pavement sections in the U.S.A. and Canada. The LTPP database includes a wide variety of pavement materials, layer thicknesses, subgrade conditions, climatic conditions, and traffic loads of the LTPP sites. The distress measurement methods also vary for these pavement sections ( 31 ). Therefore, there is a need for a local calibration to the AASHTOWare Pavement ME that reflects the local specific conditions (e.g., 31 –34). The same logic is also applied to the differences between the results of the cyclic plate loading tests on field test lane sections and the results of AASHTOWare Pavement ME. As such, correction factors are needed to account for these differences. The calibration factors in this study were obtained using the control Sections 4 and 7, by dividing the number of load cycles to achieve a specific rut depth obtained from field cyclic plate load tests to those obtained by the Pavement ME Design. For example, the correction factor for the control Section 4 at rut depth of 0.75 in. is 6.15. The number of load cycles used in Pavement ME Design is then determined for reinforced sections by multiplying the number of load cycles obtained from the experimental study by the relevant correction factor. Tamrakar et al. ( 31 ) calibrated the results of full-scale accelerated tests from studies by the U.S. Army Corps of Engineers and found calibration factors ranging from 3.26 to 6.32.

The number of load cycles obtained after applying the calibration factor was then used to back-calculate the effective resilient modulus of the base for the reinforced sections using the Pavement ME Design software to achieve the best fit with the entire rut depth–load cycle curves. The results of the back-calculation are summarized in Table 7. For each reinforced section, the resilient modulus of the base layer’s corresponding unreinforced section was obtained using the CBR–DCPI relationship, as presented above in Equation 4. As can be seen in Table 7, the resilient modulus of the base layer was increased by 28% to 210% by adding geosynthetic reinforcement, with the highest increase observed for Section 6 with the thinnest base. This increase in the base resilient modulus will result in extending the service life of a pavement.

Table 7.

Effective Base Resilient Modulus

Test sections	Base resilient modulus^a (psi)	Effective base resilient modulus^b (psi)	Percentage improvement (%)
Section 2	21694	46450	114
Section 3	26258	33600	28
Section 5	24207	36300	50
Section 6	23480	72790	210

Note: psi = pounds per square inch.

From dynamic cone penetrometer.

Back-calculated from Pavement ME Design software.

Geosynthetics as Both Base Reinforcement and Mechanical Subgrade Stabilization

The more reasonable approach to incorporate the benefits of placing geosynthetics at the base–subgrade interface within the context of the Pavement ME Design is to consider both their reinforcement effect on base layer and their stabilization effect on subgrade. This approach consists of two steps: calculating the increase in stiffness of the base course in the vicinity of the geosynthetics and estimating the subgrade permanent deformation reduction factor. The improved resilient modulus of the base layer is mostly from the increased confining pressure in the influence zone of the geosynthetics provided by the confinement effect of reinforcement. The increased confining pressure results in higher resilient modulus of the base layer. Therefore, improved base resilient modulus can be estimated by measuring the resilient modulus of base materials under the increased confining pressure. After estimating the adjusted resilient modulus of the base in the vicinity of the reinforced layer, PavementME Design software can be used to back-calculate the subgrade permanent deformation reduction factor for the reinforced section. The base course layer at the influence zone of reinforcement should be modeled as a separate layer with improved stiffness. The subgrade permanent deformation reduction factor can be considered in the permanent deformation properties part within the framework of the MEPDG, that is, in the model for predicting permanent deformation in the subgrade as follows:

δ_{SG} (N) = k_{GS} [β_{SG} (\frac{ε_{0}}{ε_{r}}) e^{- {(\frac{ρ}{N})}^{β}} ε_{v} h]

(4)

where $δ_{SG}$ is the permanent deformation for the subgrade layer at N repetitions of load; N is the number of load repetitions; $ε_{0}$ , β, and ρ are material parameters; $ε_{r}$ is the resilient strain imposed in laboratory tests to obtain material properties; $ε_{v}$ is the vertical strain of the asphalt material; h is the thickness of the subgrade layer; $β_{SG}$ is the national model calibration factor for subgrade material and k_GS is the reduction factor because of geosynthetic reinforcement. The following equation can be used to evaluate the increased confining pressure at the influence zone of geosynthetic layers:

Δ σ_{3} = \frac{T}{S}

(5)

where T is the tensile force in geosynthetics (lb/in. or kN/m), which is taken at 2% strain in this study, a common value used in design; S is the influence zone of geosynthetics, which is assumed to be 4 in. (102 mm) in this study ( 35 ). The estimated increased confining pressure is about 7 psi (50 kPa) for geogrid used in this study, which results in a confining pressure of about 10 psi (70 kPa) for base material within the influence zone of geosynthetics. The increase in confining pressure results in about a two-thirds increase of resilient modulus for the base materials used in this study, as shown in Figure 2. Again, the PavementME Design software was used here to back-calculate the subgrade permanent deformation reduction factor for the reinforced sections. The results of back-calculation show the subgrade permanent deformation reduction factor of 0.63, 0.65, 0.57, and 0.35 for Sections 2, 3, 5, and 6, respectively. The highest impact of reinforcement layer on subgrade permanent deformation was observed in Section 6 with the thinnest base layer thickness. The permanent deformation of the subgrade layer in Section 6 was 0.35 times the permanent deformation of subgrade in an unreinforced section with similar properties.

Base Course Reduction

If the service life of the reinforced section is assumed to be the same as that of the unreinforced section, then the base layer thickness can be reduced with the increase of base resilient modulus. This benefit of geosynthetic reinforcement is usually evaluated in relation to the BCR factor, which is defined as the base thickness of the reinforced section divided by the base thickness of the unreinforced section for a given traffic level. As can be seen from Table 8, the values of BCR range from 36% to 52% for pavement sections with geosynthetic layer. It seems that the thickness of the base layer can be reduced by about half or more with the inclusion of geosynthetic layers. However, for the pavement section with 11 in. (279 mm) base layer (Section 6), the value of BCR was 36% in this study, which means that an 11 in. (279 mm) thick unreinforced base layer can be reduced to 4 in. thick with the same performance. These BCR values seem to be not realistic in engineering practice. The suitability of the AASHTOWare PavementME for evaluation of BCR might need further investigation through sensitivity analysis.

Table 8.

Reduction in Base Layer Thickness

Test sections	Base thickness (in.)	Reduced thickness (in.)	Base course reduction (%)
Section 2	18.95	6.8	36
Section 3	19.4	9.27	48
Section 5	20.69	10.83	52
Section 6	11	4	36

Proposed Generalized Design Method for Geogrid Reinforcement

In the design, for simplicity, the benefit of geogrid reinforcement is assumed to be completely taken by the base layer as base reinforcement, that is, the effective base resilient modulus (M_{r_eff}) alone is used to account for the benefit of geosynthetic reinforcement, as shown in Figure 5. The effective base resilient modulus (M_{r_eff}) can be expressed as:

M_{r_eff} = (1 + α) M_{r}

(6)

where α is the percentage increase in base layer modulus, and M_r is the resilient modulus of the base course without reinforcement. Based on the results of analysis presented in the subsection “Geosynthetics as Base Reinforcement” (Table 7) and the results of rolling wheel accelerated load testing of the same sections (with improvement of 82% and 36% for Sections 2 and 3, respectively) ( 14 ), a conservative value of α = 30% was selected for a single layer of geogrid reinforcement placed at the base–subgrade interface, and a value of α = 90% was selected for double layers of geogrid reinforcement.

Figure 5.

Effect of geogrid on resilient modulus of base material: (a) single layer of geogrid reinforcement, and (b) double layers of geogrid reinforcement.

Theoretically, the effect of geogrid only increases the resilient modulus of the base within the vicinity of the geogrid (i.e., within the influence zone of the geogrid). The influence zone is assumed to be 4 in. above/below the geogrid in this study, as shown in Figure 5. The α₁ and α₂ are related to the α by the following equations:

(1 + α) M_{r} = \frac{(1 + α_{1}) M_{r} \times 4 + M_{r} \times (D - 4)}{D}

(7)

(1 + α) M_{r} = \frac{(1 + α_{1}) M_{r} \times 4 + (1 + α_{2}) M_{r} \times 8 + M_{r} \times (D - 12)}{D}

(8)

By substituting an α value of 30% for a single layer of geogrid reinforcement into Equation 7, the α₁ is back-calculated as about 135%. By substituting an α value of 90% for a double layer of geogrid reinforcement into Equation 8, the α₂ is back-calculated as about 135% too. The α₁ and α₂ values are almost identical. These values (i.e., α₁ = α₂ = 135%) are then used to calculate the α values for pavement sections with different base thickness, which are presented in Table 9.

Table 9.

Percentage Improvement of Base Resilient Modulus

Base thickness (in.)	α
	Single layer of geogrid (%)	Double layers of geogrid (%)
12	45	135
14	40	115
16	35	100
18	30	90

Note: α = percentage increase in base layer modulus.

Once the effective base resilient modulus (M_{r_eff}) is determined, it can then be input into the AASHTOWare Pavement ME Design software to estimate the service life of the reinforced pavement section. The corresponding TBR is then evaluated, as presented in Table 10.

Table 10.

Traffic Benefit Ratio (Based on AASHTOWare PavementME)

Base thickness (in.)	Traffic benefit ratio
	Single layer of geogrid	Double layers of geogrid
12	1.55	2.25
14	1.50	2.15
16	1.47	2.10
18	1.45	2.00

As indicated in Table 8, and in the analysis of accelerated rolling wheel tests results on the same geogrid-reinforced pavement sections by Abu-Farsakh et al. ( 36 ), the current version of PavementME gives an unreasonable reduction in base layer thickness, so it is not recommended to use the PavementME to determine the thickness of the base layer of a reinforced pavement section. Instead, the AASHTO 1993 Pavement Design Guide is used here to determined the BCRs, which are presented in Table 11.

Table 11.

Base Course Reduction Factors (Based on 1993 AASHTO Design Method)

Base thickness (in.)	Base course reduction
	Single layer of geogrid (%)	Double layers of geogrid (%)
12	77	50^a
14	79	55^a
16	80	55
18	81	60

Note: ^aNot realistic.

Based on the analyses, the α values presented in Table 9, the TBR values presented in Table 10, and the BCR factors presented in Table 11 are recommended for the design of geogrid-reinforced flexible pavement built over a weak subgrade (CBR = 0.5–3).

Conclusion

In this research, cyclic plate load tests were conducted on seven unreinforced and geosynthetic reinforced test sections. Results from experimental study were used to develop a framework to quantify the benefits of geosynthetic reinforcement in current pavement design methods. The proposed methods were capable of quantitatively representing the benefits of geosynthetic reinforcement observed from large-scale cyclic plate tests. The experimental study and analyses described in this paper allow the following conclusions to be made:

The test results demonstrate that both geogrid and geotextile significantly improved the performance of the pavement sections by reducing the surface permanent deformation and extending the service life of pavement sections. With a single layer of geosynthetic placed at the base–subgrade interface, the adjusted TBR (TBR_adj) can be increased up to 1.5 at a rut depth of 0.75 in. for pavement constructed using 18 in. thick base layer on top of weak subgrade soil. The inclusion of a second layer of geosynthetic reinforcement (Section 2) significantly increases TBR_adj to 2.83 at a rut depth of 0.75 in.

The objective of this project is to provide methodologies that objectively quantify the functions and benefits of geosynthetics in pavements. Those functions include base reinforcement and subgrade stabilization. To account for these benefits within the context of the Pavement ME Design, three methods are proposed in this paper. The first method considers the benefits of geosynthetic reinforcement as a base improvement only. In this method, the resilient modulus of the base layer in a reinforced section increased by an adjusting factor. For the experimental study explained in this paper, the resilient modulus of the base layer was increased by from 28% to 210% by adding single/double geosynthetic reinforcement layers.

The second method considers the benefits of geosynthetic reinforcement for both base reinforcement and subgrade stabilization. In this method, an increase in stiffness of the base layer in the influence zone of geosynthetic layers was calculated based on increased confining pressure. Then the subgrade permanent deformation reduction factor was back-calculated using PavementME Design software. The calculated subgrade permanent deformation reduction factor shows that permanent deformation in the subgrade layer for reinforced sections can be reduced by 0.35 times that of an unreinforced section with similar properties, while a two-thirds increase of the base resilient modulus occurred at the vicinity of the geosynthetic layer.

Incorporating the benefits of geosynthetics for BCR in PavementME design results in unpractical design values for base course thickness. The values of BCR in this study range from 36% to 52%. This means that the thickness of the base layer can be reduced by about half or more with the inclusion of geosynthetic reinforcement layers, which seems to be not realistic in engineering practice (e.g., an 11 in. thick unreinforced base layer can be reduced to 4 in. with same performance in this study). The PavementME may be not suitable for evaluation of BCR and this should be further investigated.

A generalized design method was proposed to quantify the benefits of geogrid reinforcement for any pavement section within the context of the mechanistic–empirical concept through evaluating the effective base resilient modulus (M_{r_eff}) within the influence zone of the geogrid.

Footnotes

Author Contributions

The authors confirm contribution to the paper as follows: study conception and design: M. Abu-Farsak, S. Hanandeh, M. Saghebfar; data collection: S. Hanandeh, M. Saghebfar; analysis and interpretation of results: M. Abu-Farsak, S. Hanandeh, Q. Chen; draft manuscript preparation: M. Abu-Farsak, Q. Chen. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Murad Y. Abu-Farsakh

Qiming Chen

Shadi Hanandeh

References

Tingle

Jersey

Cyclic Plate Load Testing of Geosynthetic-Reinforced Unbound Aggregate Roads. Transportation Research Record: Journal of the Transportation Research Board, 2005. 1936: 60–69.

Chen

Abu-Farsakh

Tao

Laboratory Evaluation of Geogrid Base Reinforcement and Corresponding Instrumentation Program. Geotechnical Testing Journal, Vol. 32, No. 6. 2009, pp. 516–525.

Henry

Clapp

Davids

Humphrey

Barna

Structural Improvements of Flexible Pavements Using Geosynthetics for Base Course Reinforcement. Report No. ERDC/CRREL TR-09-11. Federal Highway Administration, Washington, D.C., 2009.

Perkins

S. W.

Christopher

B. R.

Cuelho

E. G.

Eiksund

G. R.

Schwartz

C. S.

Svanø

A Mechanistic-Empirical Model for Base-Reinforced Flexible Pavements. International Journal of Pavement Engineering, Vol. 10, No. 2, 2009, pp. 101–114.

Abu-Farsakh

Chen

Evaluation of Geogrid Base Reinforcement in Flexible Pavement Using Cyclic Plate Load Testing. International Journal of Pavement Engineering, Vol. 12, 2011, pp. 275–288.

Jersey

S. R.

Tingle

J. S.

Norwood

G. J.

Kwon

Wayne

Full-Scale Evaluation of Geogrid-Reinforced Thin Flexible Pavements. Transportation Research Record: Journal of the Transportation Research Board, 2012. 2310: 61–71.

Tang

Stoffels

S. M.

Palomino

A. M.

Resilient and Permanent Deformation Characteristics of Unbound Pavement Layers Modified by Geogrids. Transportation Research Record: Journal of the Transportation Research, 2013. 1853: 3–10.

Tanyu

B. F.

Aydilek

A. H.

Lau

A. W.

Edil

T. B.

Benson

C. H.

Laboratory Evaluation of Geocell-Reinforced Gravel Subbase Over Poor Subgrades. Geosynthetics International, Vol. 20, No. 2, 2013, pp. 47–61.

Tang

Abu-Farsakh

Hanandeh

Chen

Performance of Reinforced and Stabilized Unpaved Test Sections Built Over Native Soft Soil Under Full-Scale Moving Wheel Loads. Transportation Research Record: Journal of the Transportation Research Board, 2015. 2511: 81–89.

10.

Abu-Farsakh

Hanandeh

Mohammad

Chen

Performance of Geosynthetic Reinforced/Stabilized Paved Roads Built Over Soft Soil Under Cyclic Plate Loads. Geotextiles and Geomembranes, Vol. 44, No. 6, 2016, pp. 845–853.

11.

Saghebfar

Hossain

Lacina

B. A.

Performance of Geotextile-Reinforced Bases for Paved Roads. Transportation Research Record: Journal of the Transportation Research Board, 2016. 2580: 3–10.

12.

Perkins

S. W.

Geosynthetic Reinforcement of Flexible Pavements. Laboratory Based Pavement Test Sections. Report FHWA/MT-99-001/8138. Federal Highway Administration, Washington, D.C., 1999.

13.

Al-Qadi

I. L.

Dessouky

S. H.

Kwon

Tutumluer

Geogrid in Flexible Pavements: Validated Mechanism. Transportation Research Record: Journal of Transportation Research Board. 2008. 2045: 102–109.

14.

Chen

Abu-Farsakh

Hanandeh

Mohammad

Performance Evaluation of Geosynthetic Reinforced Flexible Pavement Using Full-Scale Accelerated Loading Test. Geosynthetic International Journal, Vol. 25, No. 1, 2018, pp. 26–36.

15.

Luo

Lytton

R. L.

Hajj

E. Y.

Siddharthan

R. V.

Elfass

Quantifying the Influence of Geosynthetics on Pavement Performance. Report for NCHRP Project 01-50. National Academies of Science, Engineering and Medicine, Washington, D.C., 2017.

16.

Perking

Guelho

Large Scale Laboratory Testing of Geosynthetics in Roadway Applications. Report No. FHWA/MT-21-001/9564-602. Montana Department of Transporation, Bozeman, MT, 2021.

17.

Sharbaf

M. R.

Ghafoori

Laboratory Evaluation of Geogrid-Reinforced Flexbile Pavements. Transportation Engineering, Vol. 4, 2021, pp. 1–10.

18.

Al-Qadi

I. L.

Dessouky

Tutumluer

Kwon

Geogrid Mechanism in Low-Volume Flexible Pavements: Accelerated Testing of Full-Scale Heavily Instrumented Pavement Sections. International Journal of Pavement Engineering, Vol. 12, No. 2, 2011, pp. 121–135.

19.

Kwon

Tutumluer

Konietzky

Aggregate Base Residual Stresses Affecting Geogrid Reinforced Flexible Pavement Response. International Journal of Pavement Engineering, Vol. 9, No. 4, 2008, pp. 275–285.

20.

Kwon

Tutumluer

Al-Qadi

I. L.

. Validated Mechanistic Model for Geogrid Base Reinforced Flexible Pavements. Journal of Transportation Engineering, Vol. 135, No. 12, 2009, pp. 915–926.

21.

Perkins

S. W.

Christopher

B. R.

Eiksund

G. R.

Schwartz

C. W.

Svano

Modeling Effects of Reinforcement on Lateral Confinement of Roadway Aggregate. In Advances in Pavement Engineering, Geotechnical Special Publication 130 ( Schwartz

C. W.

Tutumluer

Tashman

, eds.), ASCE, Reston, VA, 2005, pp. 283–296.

22.

Tang

Stoffels

S. M.

Palomino

A. M.

Mechanistic-Empirical Approach to Characterizing Permanent Deformation of Reinforced Soft Soil Subgrade. Geotextiles and Geomembranes, Vol. 44, No. 3, 2015, pp. 429–441.

23.

Luo

Lytton

R. L.

Hajj

E. Y.

Siddharthan

R. V.

Numerical Modeling of Geogrid-Reinforced Flexible Pavement and Corresponding Validation Using Large-Scale Tank Test. Construction and Building Materials, Vol. 122, 2016, pp. 214–230.

24.

Sun

Han

Fei

Guo

Zhang

Geosynthetic-Stabilized Flexible Pavements: Solution Derivation and Mechanistic–Empirical Analysis. Geotextiles and Geomembranes, Vol. 48, No. 4, 2020, pp. 468–478.

25.

American Association of State Highway and Transportation Officials (AASHTO). AASHTOWare Pavement ME Design Build 2.2 Release Notes. Washington, D.C., 2015.

26.

Louisiana Department of Transportation and Development, Louisiana Standard Specifications for Roads and Bridges Book. Baton Rouge, Louisiana. 2006.

27.

Webster

S. L.

Grau

R. H.

Williams

R. R.

Description and Application of Dual Mass Dynamic Cone Penetrometer. Instruction Report No. GL-92-3. U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS, 1992.

28.

Powell

W. D.

Potter

J. F.

Mayhew

H. C.

Nunn

M. E.

The Structural Design of Bituminous Roads. Report LR1132. Transport and Road Research Laboratory (TRRL), Crowthorne, Berkshire, UK, 1984.

29.

Zadehmohamad

Luo

Abu-Farsakh

Voyiadjis

Evaluating Long-Term Benefits of Geosynthetics in Flexible Pavements Built Over Weak Subgrades by Finite Element and Mechanistic-Empirical Analyses. Geotextile and Geomembrane, 2022 (accepted for publication).

30.

AASHTO. AASHTO Guide for Design of Pavement Structures. American Association of State Highway and Transportation Officials. Washington, D.C., 1993.

31.

Tamrakar

Kwon

Wayne

Lee

Calibration of Pavement ME Rutting Model for Geogrid Stabilized Roadways. Transportation Geotechnics, Vol. 31, No. 2, 2021, p. 100684.

32.

Souliman

M. I.

Mamlouk

M. S.

El-Basyouny

M. M.

Zapata

C. E.

Calibration of the AASHTO MEPDG for Designing Flexible Pavements in Arizona Conditions. International Journal of Pavements, Vol. 9, No. 1-2-3, 2010, pp. 2–13.

33.

Banerjee

Prozzi

J. A.

Freiman

Regional Calibration of Permanent Deformation Performance Models for Rehabilitated Flexible Pavements. In TRB 90th Annual Meeting Compendium of Papers DVD (No. 11-1025), Washington, D.C., 2011.

34.

Hall

Xiao

Wang

Calibration of the MEPDG for Flexible Pavement Design in Arkansas. Transportation Research Record: Journal of the Transportation Research Board, 2011. 2226: 135–141.

35.

Kwon

Tutumluer

Geogrid Base Reinforcement With Aggregate Interlock and Modeling of Associated Stiffness Enhancement in Mechanistic Pavement Analysis. Transportation Research Record: Journal of the Transportation Research Board, 2009. 2116: 85–95.

36.

Abu-Farsakh

M. Y.

Chen

Hanandeh

Accelerated Load Testing of Geosynthetic Base Reinforced/Stabilized Unpaved and Pavement Test Sections. Report No. FHWA/ LA.188/603. Louisiana Transportation Research Center, Baton Rouge, LA, 2019.