**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

**Suggested Citation:**"Chapter 5. Interpretations, Appraisal, and Applications." National Academies of Sciences, Engineering, and Medicine. 2017.

*Quantifying the Influence of Geosynthetics on Pavement Performance*. Washington, DC: The National Academies Press. doi: 10.17226/24841.

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

124 CHAPTER 5. INTERPRETATIONS, APPRAISAL, AND APPLICATIONS Introduction This project developed a methodology for quantifying the influence of geosynthetics on pavement performance. This methodology included the development of the two laboratory test protocols for evaluating the impact of geosynthetics on cross-anisotropy and permanent deformation of UGMs, analytical models for quantifying the impact of geosynthetics, finite element models for computing the critical stresses and strains that control the performance of geosynthetic-reinforced pavements, and ANN models for predicting the performance of geosynthetic-reinforced pavements. One LST testing program with extensive instrumentation was conducted, and the measurements were used to validate the developed finite element models by comparing the measured pavement responses with those predicted by the models. LST Testing Program An LST testing program was conducted on flexible and rigid pavements using an 8-ft-diameter by 6-ft-high circular steel tank. Tables 5.1 and 5.2 summarize the specifics of the LST experiments on flexible pavements and rigid pavements, respectively. A database of pertinent pavement responses with and without reinforcement of the base layer collected under realistic pavement loading conditions was assembled. The established database was then used in the numerical investigation (model and input parameters, etc.) of the LST to assess the validity and applicability of the finite element numerical modeling of pavement structures with geosynthetic-reinforced bases. In particular, data critical to the validation of the numerical modeling of the interaction of a geosynthetic layer with the surrounding medium were examined and included: ï· The stress distributions across the geosynthetic under dynamic loading in both the AC and PCC pavements. ï· The strain measurements in the geosynthetic and at the bottom of the surface layer (AC or PCC) under dynamic loading. ï· The deformed shape of the geosynthetic and the slippage at the interface between the geosynthetic and the unbound material under dynamic loading. ï· The slippage between the bottom of the PCC slab and the supporting unbound granular material base layer at the edge of the loaded slab. The LST testing program confirmed that a careful representation of the geosynthetic material was necessary for the overall numerical modeling of reinforced pavement structures. An appropriate modeling of the geosynthetic material should be able to capture the mechanism of the behavior of the reinforcement in the base layer under dynamic loading. This mechanism was observed to be different between the selected geogrid and geotextile materials evaluated as part of this study. These differences are explained and illustrated in Chapter 4. The effects of these mechanism differences were incorporated in the ANN models that were developed in this project. The extensive LST database assembled in this study would serve as a valuable resource for the verification of future numerical modeling of reinforced base layers in pavement structures.

125 Table 5.1. Summary of Specifics of LST Experiments on Flexible Pavements Experiment Surface Layer Thick. (inch) CAB Layer Thick. (inch) Reinforcement Instrumentation ID No Type Location Transducer type Location Quantity AC- Contr- B06 1 6 6 None (Control) N/A LVDT Surface 5 Accelerometer (1D) Surface 3 Accelerometer (2D) Middle of the base 3 (V and H) Pressure cell Middle of the base, 2 in. and 6 in. below the subgrade 5 (V), 2 (H) Asphalt strain gauge At the bottom of the HMA 1 AC- Contr- B10 2 6 10 None (Control) N/A LVDT Surface 5 Accelerometer (1D) Surface 3 Accelerometer (2D) Middle of the base 3 Pressure cell Middle of top and bottom half of the base, 2 in. and 6 in. below the subgrade 5 (V), 2 (H) Asphalt strain gauge At the bottom of the HMA 1 AC- Grid- B06 3 6 6 Geogrid Base- Subgrade Interface LVDT Surface 6 Accelerometer (1D) Surface 3 Accelerometer (2D) Middle of the base (on geogrid and in the base) 3 (G), 3 (B) Pressure cell Middle of the base, 2 in. and 6 in. below the subgrade 5 (V), 2 (H) Asphalt strain gauge At the bottom of the HMA 1 Strain gauge On geogrid (X and Y directions) 3 AC- Grid- B10 4 6 10 Geogrid Middle of the Base LVDT Surface 6 Accelerometer (1D) Surface 3 Accelerometer (2D) Middle of the base (on geogrid & in the base) 3 (G), 3 (B) Pressure cell Middle of top and bottom half of the base, 2 in. and 6 in. below the subgrade 5 (V), 2 (H) Asphalt strain gauge At the bottom of the HMA 1 Strain gauge On geogrid (X and Y directions) 3 AC- Textile- B06 5 6 6 Geotextile Base- Subgrade Interface LVDT Surface 6 Accelerometer (1D) Surface 3 Accelerometer (2D) Middle of the base (on geogrid and in the base) 3 (G), 3 (B) Pressure cell Middle of the base, 2 in. and 6 in. below the subgrade 5 (V), 2 (H) Asphalt strain gauge At the bottom of the HMA 1 Strain gauge On geotextile (X and Y directions) 3 AC- Textile- B10 6 6 10 Geotextile Middle of the Base LVDT Surface 6 Accelerometer (1D) Surface 3 Accelerometer (2D) Middle of the base (on geogrid and in the base) 3 (G), 3 (B) Pressure cell Middle of top and bottom half of the base, 2 in. and 6 in. below the subgrade 5 (V), 2 (H) Asphalt strain gauge At the bottom of the HMA 1 Strain gauge On geotextile (X and Y directions) 3 Note: V = vertical, H = horizontal, B = base, and G = geogrid/geotextile.

126 Table 5.2. Summary of Specifics of LST Experiments on Rigid Pavements Experiment Surf. Layer Thick. (inch) CAB Layer Thick. (inch) Reinforcement Instrumentation ID No Type Location Transducer type Location Quantity PCC- Contr-IS 7 6 8 None (Control) N/A LVDT Surface and top of the base (H) 5(V), 2(H) Accelerometer (1D) Surface 7 Accelerometer (2D) Middle of the base, embedded in concrete at the interface, top of the base 8(B), 2 (C) Accelerometer (3D) Embedded in concrete at the interface, top of the base 1(B), 1(C) Pressure cell Middle of top and bottom half of the base, 2 in. and 6 in. below the subgrade 10 (V), 1 (H) Concrete strain gauge At the bottom of the PCC 1 PCC- Grid-IS 9 6 8 Geogrid Middle of the Base LVDT Surface and top of the base (H) 5(V), 2(H) Accelerometer (1D) Surface 7 Accelerometer (2D) Middle of the base on geogrid and in the base, embedded in concrete at the interface and top of the base 8(B), 2(C), 4 (G) Accelerometer (3D) Embedded in concrete at the interface, top of the base 2(B), 1(C), 1(G) Pressure cell Middle of top and bottom half of the base, 2 in. and 6 in. below the subgrade 10 (V), 1 (H) Concrete strain gauge At the bottom of the PCC 1 Strain gauge On geogrid (X and Y directions) 5 PCC- Textile- IS 10 6 8 Geotextile Middle of the Base LVDT Surface and top of the base (H) 5(V), 2(H) Accelerometer (1D) Surface 7 Accelerometer (2D) Middle of the base on geotextile and in the base, embedded in concrete at the interface and top of the base 8(B), 2(C), 4 (G) Accelerometer (3D) Embedded in concrete at the interface, top of the base 2(B), 1(C), 1(G) Pressure cell Middle of top and bottom half of the base, 2 in. and 6 in. below the subgrade 10 (V), 1 (H) Concrete strain gauge At the bottom of the PCC 1 Strain gauge On geotextile (X and Y directions) 5 LVDT Surface and top of the base (H) 5(V), 2(H) Note: V = vertical, H = horizontal, B = base, G = geogrid/geotextile, and C = concrete.

127 Measurement of Geosynthetic-Aggregate/Soil Interfacial Slippage The slippage at the geosynthetic-aggregate/soil interface significantly affected the geosynthetic-UGM interaction. The horizontal slippage was calculated as the difference in horizontal displacements between the geosynthetic and the adjacent UGM using the calibrated double-integration procedure that is explained in Appendix I. Figures 5.1â5.4 show the measured interfacial slippage for the geogrid-reinforced and geotextile-reinforced pavement structures, respectively. Slippage at the geogrid interface only occurred when the geogrid was placed at the bottom of the base course. It increased as the load level increased. Slippage at the geotextile interface occurred when the geotextile was placed at the bottom and in the middle of the base course. This slippage was greater than with the geogrids and also increased with load level when the geotextile was placed in the middle of the base course. Figure 5.1. Measured Horizontal Displacements of Geogrid and UGM When Geogrid Was Placed in the Middle of the Base Course -10 -5 0 5 10 15 20 5 10 15 20 25H or iz on ta l D is pl ac em en t ( m ils ) Load (kips) Geogrid in the Middle of Base Geogrid Base (below) Outward Inward

128 Figure 5.2. Measured Horizontal Displacements of Geogrid and UGM When Geogrid Was Placed at the Bottom of the Base Course Figure 5.3. Measured Horizontal Displacements of Geotextile and UGM When Geotextile Was Placed in the Middle of the Base Course -10 -5 0 5 10 15 20 5 10 15 20 25H or iz on ta l D is pl ac em en t ( m ils ) Load (kips) Geogrid at the Bottom of Base Geogrid Subgrade (below) Outward Inward -10 -5 0 5 10 15 20 5 10 15 20 25H or iz on ta l D is pl ac em en t ( m ils ) Load (kips) Geotextile in the Middle of Base Geotextile Base (below) Outward Inward

129 Figure 5.4. Measured Horizontal Displacements of Geotextile and UGM When Geotextile Was Placed at the Bottom of the Base Course Determination of Geosynthetic-Aggregate/Soil Interfacial Properties The interfacial shear stiffness was an important property for characterizing the geosynthetic-aggregate/soil interaction behavior. This property depended on the slippage condition between the geosynthetic and the surrounding aggregates. An analytical solution was derived to determine the interfacial shear stiffness under various slippage conditions using the pullout test data. The details of this solution are found in Appendix B. The measured slippage from the LST tests showed that the maximum horizontal relative displacement between the geosynthetic and aggregates was less than 0.04 inch. This finding demonstrated that the interface slippage normally occurred in the geosynthetic-reinforced aggregates when the relative displacement was in the linear stage of the pullout test. Impact of Geosynthetics on Cross-Anisotropy and Permanent Deformation of UGMs The impact of geosynthetics on cross-anisotropy and permanent deformation of UGMs was evaluated using the rapid triaxial test. The geosynthetic reinforcement was influenced by the geosynthetic type, the sheet stiffness, and the geosynthetic location. In general, the geogrid increased both the vertical and horizontal modulus but not the anisotropic ratio of the UGM, while the geotextile only increased the horizontal modulus, which resulted in an increase in the anisotropic ratio of the UGM. Figure 5.5 shows the horizontal and vertical resilient moduli of the unreinforced UGM at each stress state shown in Table 4.1 in Chapter 4. Figures 5.6â5.8 show the effects of geogrids and geotextiles on the cross-anisotropy of UGMs. -10 -5 0 5 10 15 20 5 10 15 20 25H or iz on ta l D is pl ac em en t ( m ils ) Load (kips) Geotextile at the Bottom of Base Geotextile Subgrade (below) Outward Inward

130 Figure 5.5. Horizontal and Vertical Moduli of Unreinforced UGM at Each Stress State Figure 5.6. Effect of Geosynthetics on Horizontal Modulus of UGM 0 20 40 60 80 100 1 2 3 4 5 6 7 8 9 10 R es ili en t M od ul us (k si ) Stress State No. Horizontal Moduli Vertical Moduli 80 90 100 110 120 130 140 150 160 1 2 3 4 5 6 7 8 9 10N or m al iz ed H or iz on ta l M od ul us R at io (% ) Stress State No. Geotextile Geogrid

131 Figure 5.7. Effect of Geosynthetics on Vertical Modulus of UGM Figure 5.8. Effect of Geosynthetics on Anisotropic Ratio of UGM Compared to the increase in the resilient modulus, the reduction of the permanent deformation of the UGM was a more important benefit of the geosynthetic reinforcement. Figure 5.9 shows the effect of geosynthetic reinforcement on the permanent deformation of the UGM when it was subjected to the different deviatoric stress levels. A description of the selected geosynthetics and the relevant geosynthetic material is found in Appendix C. Researchers found that the geogrid was much more effective than the geotextile at reducing the permanent 80 90 100 110 120 130 140 1 2 3 4 5 6 7 8 9 10 N or m al iz ed V er tic al M od ul us R at io (% ) Stress State No. Geotextile Geogrid 80 90 100 110 120 130 140 150 160 170 1 2 3 4 5 6 7 8 9 10 N or m al iz ed A ni so tr op y R at io (% ) Stress State No. Geotextile Geogrid

132 deformation of the UGM. The effect of geogrid reinforcement was not significant in reducing the permanent deformation until the deviatoric shear stress reached a threshold level (e.g., Ïd = 19 psi in this study). The accurate and efficient laboratory characterization of a geosynthetic-reinforced UGM provided a sound basis for including the geosynthetic material in the Pavement ME Design software. The impact of geosynthetics on cross-anisotropy and permanent deformation of UGMs would further influence the performance of geosynthetic-reinforced pavement structures. Figure 5.9. Effect of Geosynthetic Reinforcement on Reducing Permanent Strain of UGMs Mechanistic-Empirical Permanent Deformation Model for Unreinforced and Geosynthetic- Reinforced UGMs Based on the Drucker-Prager plastic yield criterion, a new mechanistic-empirical rutting model was developed to evaluate the stress-dependent permanent deformation characteristics of geosynthetic-reinforced and unreinforced UGMs. The model is found in Equation 4.7 in Chapter 4. Figures 5.10â5.12 compare the model-predicted permanent strain curves with the laboratory-measured ones at different stress states for both unreinforced and geosynthetic- reinforced UGMs. The stress states used in this testing protocol are tabulated in Table 4.3 in Chapter 4. Figures 5.10â5.12 illustrate that all of the determined RMSE values were relatively small, which indicates that the developed model accurately captured the influence of stress level on the permanent strain of the geosynthetic-reinforced and unreinforced UGMs. Figures 5.10â 5.12 also present the determined coefficients of the developed rutting model. These model coefficients can be used to predict the permanent deformation of UGMs at any stress levels and numbers of load repetitions. 0.0 10.0 20.0 30.0 40.0 50.0 Ïd=10 psi Ïd= 19 psi Ïd=25 psi Ïd=28 psi R ed uc tio n of P er m an en t S tr ai n (% ) TX-1 Mid TX-2 Mid BX-3 Mid GT-4 Mid

Figure Figur 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 A cc um ul at ed P la st ic S tr ai n (% ) 0.0 0.2 0.4 0.6 0.8 1.0 A cc um ul at ed P la st ic S tr ai n (% ) 5.10. Com e 5.11. Com 0 S1: RM S3: RM S5: RM S7: RM 0 S1: RM S3: RM S5: RM S7: RM parison of L Stra parison of Strain C 2000 N SE=0.012 S SE=0.002 S SE=0.012 S SE=0.018 2000 Nu SE=0.008 SE=0.011 SE=0.013 SE=0.012 ab-Measu in Curves fo Lab-Measu urves for G 4000 umber of Lo 2: RMSE=0 4: RMSE=0 6: RMSE=0 4000 mber of Loa S2: RMSE=0 S4: RMSE=0 S6: RMSE=0 133 red and Pro r Unreinfo red and Pro eogrid-Rei 6000 ad Cycles .002 .004 .016 6000 d Cycles .007 .004 .009 posed-Mod rced UGM posed Mod nforced UG 8000 8000 el-Predicte s el-Predicte Ms 10000 10000 d Permane d Permane Contr Contr Contr Contr Contr Contr Contr Predic Predic Predic Predic Predic Predic Predic TX-1-Mi TX-1-Mi TX-1-Mi TX-1-Mi TX-1-Mi TX-1-Mi TX-1-Mi Predict S Predict S Predict S Predict S Predict S Predict S Predict S nt nt ol S1 ol S2 ol S3 ol S4 ol S5 ol S6 ol S7 t S1 t S2 t S3 t S4 t S5 t S6 t S7 d S1 d S2 d S3 d S4 d S5 d S6 d S7 1 2 3 4 5 6 7

Figure Analytic A permanen state. In t was equi of a UGM influence Figures 5 horizonta 0.0 0.2 0.4 0.6 0.8 1.0 A cc um ul at ed P la st ic S tr ai n (% ) 5.12. Com al Model fo n analytical t deformati his model, t valent to an specimen. of the tensi .13 and 5.14 l and vertic 0 S1: RM S3: RM S6: RM parison of L Strain Cu r Quantifyi model was on of the ge he effect of additional c The develo le sheet stiff show the e al moduli of 2000 N SE=0.016 S SE=0.018 S SE=0.023 S ab-Measu rves for G ng Influenc developed to osynthetic-r lateral confi onfining stre ped analytic ness and the ffect of the the UGM. 4000 umber of L 2: RMSE=0. 5: RMSE=0. 7: RMSE=0. 134 red and Pro eotextile-Re e of Geosyn predict the einforced U nement due ss, which w al model wa location of tensile sheet 6000 oad Cycles 011 032 010 posed-Mod inforced U thetics vertical and GMs when s to the shear as triangula s also capab the geosynt stiffness of 8000 el-Predicte GMs horizontal ubjected to restraint of rly distribut le of quanti hetic within the geosynt 10000 d Permane moduli and a triaxial str the geosynt ed along the fying the the base co hetic on the GT-4 GT-4 GT-4 GT-4 GT-4 GT-4 Predic Predic Predic Predic Predic Predic nt the ess hetic side urse. S1 S2 S3 S5 S6 S7 t S1 t S2 t S3 t S5 t S6 t S7

135 Figure 5.13. Effect of Geosynthetic Sheet Stiffness on Predicted Horizontal Modulus of UGM Figure 5.14. Effect of Geosynthetic Sheet Stiffness on Predicted Vertical Modulus of UGM Development of Finite Element Model for Geosynthetic-Reinforced Pavement A geosynthetic reinforced the granular material through two major mechanisms: (a) lateral confinement to reinforce the UGM near the geosynthetic, and (b) the membrane effect to reduce the vertical stresses in the base and subgrade. The finite element models were developed to simulate the geosynthetic-reinforced pavement structures by taking into account these two mechanisms, and to evaluate the effect of material and geometric factors on the performance of geosynthetic-reinforced pavements. In the finite element model, the lateral confinement was equivalent to an increase in the horizontal and vertical moduli of the UGM in the geosynthetic influence zone. The membrane 0 10 20 30 40 1 2 3 4 5 6 7 8 9 10 H or iz on ta l M od ul us (k si ) Stress State No. Sheet Stiffness=3000 lb/in Sheet Stiffness=6000 lb/in Sheet Stiffness=18000 lb/in 0 20 40 60 80 100 1 2 3 4 5 6 7 8 9 10 V er tic al M od ul us (k si ) Stress State No. Sheet Stiffness=3000 lb/in Sheet Stiffness=6000 lb/in Sheet Stiffness=18000 lb/in

136 effect was simulated by defining the geosynthetic as a membrane element and characterizing the geosynthetic-aggregate/soil interface interaction using the Goodman model (43). The developed geosynthetic-reinforced pavement models were successfully validated by comparing the predicted pavement responses to the LST measurements. These finite element models were also able to quantify the effect of layer thickness, layer modulus, geosynthetic sheet stiffness, and geosynthetic location on pavement performance. The finite element modeling technique provided a sound basis for predicting the critical stresses and strains that control the performance of geosynthetic-reinforced pavements. Using this approach, a large database of critical pavement responses was established for a wide range of geosynthetic-reinforced pavement structures. This database was used in developing ANN models of the critical strains in pavement structures. Predictions of Geosynthetic-Reinforced Pavement Performance The ANN models were used to predict the responses of geosynthetic-reinforced pavement structures when they were subjected to a standard single-axle load (i.e., 18,000-lb single-axle load). The established database of geosynthetic-reinforced pavement responses was used to train and validate the ANN models with the following critical strains: ï· Horizontal tensile strain at the bottom of the asphalt layer for fatigue cracking. ï· Vertical compressive strains within the asphalt layer, base course, and subgrade for rutting. The developed ANN models were accurate and efficient in predicting these critical responses of arbitrary user-inputted geosynthetic-reinforced pavement structures. Compared to the finite element method, the major advantage of the ANN approach was that it was compatible with the Pavement ME Design software and greatly reduced computer run times. Figures 5.15 and 5.16 compare the effect of the base modulus on the predicted critical responses of unreinforced and geogrid-reinforced pavements as shown in Figure 4.83 in Chapter 4. The ANN models were sensitive to the variation of the base modulus. The compressive strain in the base course and subgrade decreased with an increasing base modulus. Figures 5.17 and 5.18 show the sensitivity of the subgrade modulus to the critical responses of the unreinforced and geogrid- reinforced pavements as predicted by the ANN models. The compressive strain at the top of the subgrade decreased, while the average compressive strain in the base layer slightly increased with an increasing subgrade modulus. The effects of the tensile sheet stiffness of the geogrid on the predicted critical pavement responses are shown in Figures 5.19 and 5.20. The geogrid with a higher tensile sheet stiffness achieved more beneficial effects in reducing the compressive strain in the base layer and subgrade.

137 Figure 5.15. Effect of Base Modulus on Average Compressive Strain in Base Layer Figure 5.16. Effect of Base Modulus on Compressive Strain at the Top of Subgrade 0 300 600 900 1200 0 20 40 60 80 A ve ra ge C om pr es si ve S tr ai n in B as e L ay er (Î¼ Îµ) Base Course Modulus (ksi) Unreinforced Pavement Geogrid-Middle Reinforced Pavement 600 800 1000 1200 0 20 40 60 80C om pr es si ve S tr ai n at th e T op o f Su bg ra de (Î¼ Îµ) Base Course Modulus (ksi) Unreinforced Pavement Geogrid-Middle Reinforced Pavement

138 Figure 5.17. Effect of Subgrade Modulus on Average Compressive Strain in Base Layer Figure 5.18. Effect of Subgrade Modulus on Compressive Strain at the Top of Subgrade 0 300 600 900 0 5 10 15 20 25 30A ve ra ge C om pr es si ve S tr ai n in B as e L ay er (Î¼ Îµ) Subgrade Modulus (ksi) Unreinforced Pavement Geogrid-Middle Reinforced Pavement 0 400 800 1200 1600 0 5 10 15 20 25 30 C om pr es si ve S tr ai n at th e T op o f Su bg ra de (Î¼ Îµ) Subgrade Modulus (ksi) Unreinforced Pavement Geogrid-Middle Reinforced Pavement

139 Figure 5.19. Effect of Tensile Sheet Stiffness of Geogrid on Average Compressive Strain in Base Layer Figure 5.20. Effect of Tensile Sheet Stiffness of Geogrid on Compressive Strain at the Top of Subgrade The performance of geosynthetic-reinforced flexible pavements includes fatigue cracking, permanent deformation, and IRI. These pavement performance measures are significantly influenced by traffic and climate factors, which have been taken into account in the current Pavement ME Design software. In this study, the geosynthetic-reinforced pavement structure with the input material properties was first made equivalent to an unreinforced pavement structure with the modified input material properties to achieve the identical critical responses. The modified material properties were then input into the Pavement ME Design software to predict the performance of the geosynthetic-reinforced pavement. The calibration of this approach relied on the validity of the calibration that was done in the existing versions of the 400 500 600 700 800 Unreinforced 1200 2400 3600 A ve ra ge C om pr es si ve S tr ai n in B as e L ay er (Âµ Îµ) Tensile Sheet Stiffness of Geogrid (lb/in) 600 700 800 900 1000 1100 1200 Unreinforced 1200 2400 3600 C om pr es si ve S tr ai n at th e T op o f Su bg ra de (Âµ Îµ) Tensile Sheet Stiffness of Geogrid (lb/in)

140 Pavement ME Design software. Figures 5.21â5.23 compare the effect of the base modulus on the predicted performance of geosynthetic-reinforced and unreinforced pavements after 10-year service in College Station, Texas. The geogrid placed in the middle of the base course was effective at reducing the rutting damage and IRI of pavement while slightly reducing the fatigue cracking of the pavement. The geogrid had a more beneficial effect on pavement performance when it was placed in the base course with a smaller resilient modulus. Figures 5.24â5.26 show the sensitivity of the subgrade modulus on the predicted performance of geosynthetic-reinforced and unreinforced pavements. The geogrid was more effective at reducing the rutting and IRI when the subgrade had a smaller resilient modulus. Figures 5.27â5.29 present the effect of the tensile sheet stiffness of the geogrid on the predicted pavement performance. The rutting depth, fatigue cracking, and IRI slightly decreased with increasing sheet stiffness of the geogrid. Figure 5.21. Effect of Base Modulus on Rutting Depth of Geosynthetic-Reinforced and Unreinforced Pavements Figure 5.22. Effect of Base Modulus on Fatigue Cracking of Geosynthetic-Reinforced and Unreinforced Pavements 0.4 0.5 0.6 0.7 0.8 Base Modulus=20 ksi Base Modulus=40 ksi Base Modulus=60 ksi R ut tin g D ep th (i nc h) Unreinforced Geogrid-Middle Reinforced Pavement 0 1 2 3 4 5 Base Modulus=20 ksi Base Modulus=40 ksi Base Modulus=60 ksi Fa tig ue C ra ck in g (% L an e A re a) Unreinforced Geogrid-Middle Reinforced Pavement

141 Figure 5.23. Effect of Base Modulus on IRI of Geosynthetic-Reinforced and Unreinforced Pavements Figure 5.24. Effect of Subgrade Modulus on Rutting Depth of Geosynthetic-Reinforced and Unreinforced Pavements 85 90 95 100 105 Base Modulus=20 ksi Base Modulus=40 ksi Base Modulus=60 ksi In te rn at io na l R ou gh ne ss In de x (in /m i) Unreinforced Geogrid-Middle Reinforced Pavement 0 0.2 0.4 0.6 0.8 1 Subgrade Modulus=5 ksi Subgrade Modulus=15 ksi Subgrade Modulus=25 ksi R ut tin g D ep th (i nc h) Unreinforced Geogrid-Middle Reinforced Pavement

142 Figure 5.25. Effect of Subgrade Modulus on Fatigue Cracking of Geosynthetic-Reinforced and Unreinforced Pavements Figure 5.26. Effect of Subgrade Modulus on IRI of Geosynthetic-Reinforced and Unreinforced Pavements 0 1 2 3 4 5 Subgrade Modulus=5 ksi Subgrade Modulus=15 ksi Subgrade Modulus=25 ksi Fa tig ue C ra ck in g (% L an e A re a) Unreinforced Geogrid-Middle Reinforced Pavement 85 90 95 100 105 110 Subgrade Modulus=5 ksi Subgrade Modulus=15 ksi Subgrade Modulus=25 ksi In te rn at io na l R ou gh ne ss In de x (in /m i) Unreinforced Geogrid-Middle Reinforced Pavement

143 Figure 5.27. Effect of Sheet Stiffness of Geogrid on Rutting Depth of Reinforced Pavements Figure 5.28. Effect of Sheet Stiffness of Geogrid on Fatigue Cracking of Reinforced Pavements 0 0.2 0.4 0.6 0.8 Control Sheet Stiffness=1200 lb/in Sheet Stiffness=2400 lb/in Sheet Stiffness=3600 lb/in R ut tin g D ep th (i nc h) 0 0.5 1 1.5 2 2.5 Control Sheet Stiffness=1200 lb/in Sheet Stiffness=2400 lb/in Sheet Stiffness=3600 lb/in Fa tig ue C ra ck in g (% L an e A re a)

144 Figure 5.29. Effect of Sheet Stiffness of Geogrid on IRI of Reinforced Pavements 80 85 90 95 100 Control Sheet Stiffness=1200 lb/in Sheet Stiffness=2400 lb/in Sheet Stiffness=3600 lb/in In te rn at io na l R ou gh ne ss In de x (in /m i)