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.
L-1 APPENDIX L. CHARACTERIZATION OF MATERIALS USED IN LARGE-SCALE TANK TEST Materials for the flexible and rigid pavements included typically used materials for dense-graded hot-mix asphalt (HMA) and Portland cement concrete (PCC), respectively. Both systems used the same crushed aggregate base and subgrade materials. The subgrade thickness was kept the same in all flexible and rigid experiments. Subgrade Layer The subgrade consisted of a high plasticity clay soil that was sampled from a local source. Before placing the subgrade material in the LST, a series of conventional characterizations were performed to determine swelling potential, moisture-density relation, and particle size distribution. To determine swelling potential of the clay soil, AASHTO T89 and T90 were followed. The soil was sampled, manually pulverized, and then washed over a standard number 40 sieve and allowed to dry for 24 hours at 100°C. Once the soil was ready for testing, standard methods were applied to find the Atterberg limits. The liquid limit curve for the clayey subgrade material is shown in Figure L-1a. Table L-1 summarizes the determined Atterberg limits. AASHTO T11 and T88 were followed to plot a particle size distribution curve for the subgrade material. A laboratory compaction test using modified efforts was also performed in accordance with AASHTO T180 to determine the maximum dry density and the optimum moisture content of the subgrade material. Figure L-1b shows the proctor curve for the clay material. (a) (b) Figure L-1. (a) Liquid limit curve for the subgrade material; (b) modified proctor curve for the subgrade material Table L-1. Atterberg Limits for Subgrade Material Liquid Limit Plastic Limit Plasticity Index Adjusted Plasticity Index 68.1 28.4 39.7 35.3 R² = 0.9541 60 62 64 66 68 70 1 10 100 1000 W at er C on te nt (% ) N drops 90 95 100 105 110 115 120 0 5 10 15 20 25 30 Dr y D en sit y ( pc f) Water Content (%)
L-2 The shear strength parameters of the subgrade material were also determined. Direct shear testing was conducted on three sample replicates subjected to different normal stresses in accordance with ASTM D3080. The achieved density for the direct shear specimens was similar to the in-density of the subgrade material in the LST experiments (95 percent of the maximum dry density at 16 percent water content). The samples were not flooded during testing (i.e., tested unsaturated). Each sample was subjected to a different normal stress that was applied for 24 hours before the shearing phase to ensure proper consolidation. The applied normal stresses were 7.4, 14.8, and 29.7 psi. After consolidation, the samples were sheared at a very slow rate. Figure L-2 presents the normalâshear stress relationship. It was concluded that the subgrade had a peak friction angle of 22.9 degrees and an associated peak cohesion of 6.3 psi. Figure L-2. Direct shear test results for the subgrade material In order to place the required amount of clay in the LST, the task was divided into three days using a team of five people. The goal was to place the soil at 16 percent water content and 95 percent of the maximum dry density to a depth of 5.5 ft. The final thickness of the subgrade was kept at 4.5 ft for the various experiments. The additional 1 ft was placed to protect the clay from contamination and to reduce moisture evaporation. The top 1 ft of the subgrade was removed right before placing the crushed aggregate base in the LST. The process of placing the subgrade material was fairly straightforward. The material was shoveled from the stockpile into 5-gal buckets. Four buckets, weighing 30 lb each, were then placed in a concrete mixer with enough water to reach the desired 16 percent moisture content. This subgrade-water mixture was mixed for approximately 30â40 seconds to maintain an even blend. The moist subgrade was then placed into the LST. To assure the proper amount of water was added to the sample, an average moisture content reading was taken at specified times two days before the filling began to analyze how the water content of the clay stockpile changed throughout the day. The weight of the water mixed with the subgrade was adjusted according to the change in water content of the stockpile. To achieve the required compaction, a gasoline- powered rammer proved to be the best option. Ten to 12 passes lasting approximately 5 to 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 17.9 19.9 0.0 3.1 6.3 9.4 12 .5 15 .6 18 .8 21 .9 25 .0 28 .2 31 .3 Sh ea r S tre ss (p si) Normal Stress (psi) Peak(s) Peak Tangent Residual Peak(s) Peak: Phi = 22.9 C = 6.3 psi
L-3 7 minutes each were made to yield a 3-inch compacted lift. To assure the required 95 percent compaction was being reached, nuclear density gauge readings were taken at the 1-, 2-, and 3-ft levels in the LST. Figure L-3 shows the various phases of the placement and compaction of the subgrade material in the tank. (a) (b) (c) Figure L-3. Placement of the subgrade material in the LST: (a) compaction of the first lift; (b) nuclear density gauge testing; (c) completed placement for the 4.5 ft of subgrade material While a nuclear density gauge was used to ensure achieving target density during the installation of the subgrade, limited dynamic cone penetrometer (DCP) testing was used to assess the density of the subgrade layer before and after the testing experiments. Thus, the DCP test was conducted after the placement of the subgrade layer for Experiment No. 1 to have a baseline reading. Then, before placing the new crushed aggregate base and asphalt layer for Experiment No. 2, another DCP test was conducted on the subgrade layer to determine if any changes to the density had occurred as a result of the pressures generated by the base, asphalt, and loading during placement and testing. In general, the results showed no significant difference between the densities of the pre-tested subgrade compared to the post-tested subgrade (see Figure L-4). Figure L-4. DCP test results for subgrade layer in LST -60 -50 -40 -30 -20 -10 0 0 5 10 15 20 De pt h (in ch ) Number of Blows Pre-Test DCP Post-Test DCP
L-4 Base Layer A typical dense-graded crushed aggregate base was used in the LST experiments. The same material was used for all flexible and rigid pavement testing. Standard specifications for dense-graded crushed aggregate base (CAB) were reviewed for 19 different states throughout the country, and the requirements for liquid limit, plasticity index, Los Angeles abrasion loss, and resistance R-value are summarized in Table L-2. The minimum, median, and maximum for all reported values are also shown at the bottom of Table L-2. Overall, the specifications were very similar among the various surveyed states, with the Nevada Department of Transportation (NDOT) specification being very close to the median values of all the examined statesâ specifications. Accordingly, the selected crushed aggregate base material following the NDOT materialsâ specification was considered to be representative of a typical dense-graded crushed aggregate base. Table L-2. Summary of Selected State Specifications for Crushed Aggregate Base Material State Maximum Liquid Limit (LL) Maximum Plasticity Index (PI) Maximum Los Angeles Abrasion Loss (%) Minimum R-Value Alabama 25 6 60 70 Arizona 25 5 40 70 California 25 6 45 78 Colorado 35 6 45 70 Florida 25 6 45 70 Illinois 25 6 40 72 Indiana 25 5 40 70 Kansas 25 6 50 70 Massachusetts 25 6 45 75 Minnesota 25 6 40 75 Mississippi 25 6 45 70 Nevada 35 6 45 70 New York 25 6 35 70 North Dakota 25 5 50 70 Oklahoma 25 6 50 70 Texas 35 10 45 75 Pennsylvania 35 6 45 72 Virginia 25 6 45 70 Washington 25 6 35 72 Minimum 25 5 35 70 Median 25 6 45 70 Maximum 35 10 60 78 Aggregate base material from a local quarry in northern Nevada was sampled according to the AASHTO T2 protocol and brought back to the UNR facility for testing. Using AASHTO T248 splitting methods, the sample was reduced in size and blended until an adequate sample size and mix were achieved. From the blended sample, the AASHTO T27 and T180 protocols were followed to determine the gradation, maximum dry density, and optimum moisture content. Table L-3 and Figure L-5 show the gradation and moisture-density relationships, respectively.
L-5 Table L-3. Gradation for Crushed Aggregate Base Material Sieve Size Percent Passing Lower Limit Specification Upper Limit Specification 1 inch 100 100 100 3/4 inch 97 90 100 1/2 inch 86 - - 3/8 inch 68 - - No. 4 46 35 65 No. 8 30 - - No. 10 26 25 53 No. 16 20 15 40 No. 30 15 - - No. 40 12 12 28 No. 50 10 - - No. 100 8 - - No. 200 3.1 2 10 Figure L-5. Modified proctor curve for crushed aggregate base material The rapid triaxial test (RaTT) was employed to determine the cross-anisotropic properties of granular base material used in the Large-Scale Tank test. A total of 10 stress states associated with three test modes (i.e., compression, shear, and extension modes) were applied in the test protocol. Table L-4 presents the results of the anisotropic properties of the base material used in the Large-Scale Tank test. 126 128 130 132 134 136 138 0 2 4 6 8 10 12 Dr y D en sit y ( pc f) Water Content (%)
L-6 Table L-4. Results of Rapid Triaxial Tests for Base Materials Used in Large-Scale Tank Test Ï1 (kPa) Ï3 (kPa) Ex (MPa) Ey (MPa) Gxy (MPa) νxy νxx Ex/Ey Gxy/Ey 40 25 70.3 129.6 49.0 0.15 0.47 0.54 0.38 50 25 79.8 141.8 60.3 0.2 0.42 0.56 0.43 70 40 93.0 200.0 80.6 0.16 0.35 0.47 0.40 130 60 121.6 312.5 98.4 0.17 0.43 0.39 0.31 150 70 153.1 374.6 119.7 0.15 0.42 0.41 0.32 170 100 166.3 388.1 133.3 0.22 0.47 0.43 0.34 220 120 205.2 442.9 145.5 0.16 0.42 0.46 0.33 250 140 224.1 519.7 166.8 0.13 0.42 0.43 0.32 250 120 200.3 495.2 152.5 0.2 0.42 0.40 0.31 250 105 180.7 456.7 151.2 0.16 0.45 0.40 0.31 Average 0.17 0.43 0.45 0.35 The constitutive models of the base material used in this study are shown in Equations L-1 to L-3. 321 1 ( ) ( 1)kk octy a a a IE k P P P Ï = + (L-1) x y En E = (L-2) xy y G m E = (L-3) where 1I is the first invariant of the stress tensor; octÏ is the octahedral shear stress; aP is the atmospheric pressure; 1k , 2k , and 3k are regression constants; xE is the horizontal resilient modulus; yE is the vertical resilient modulus; and xyG is the shear modulus in the x yâ plane. According to the results presented in Table L-4, the parameters in the constitutive models were determined by using the Solver function, and shown in Table L-5. Table L-5. Determination of the Cross-Anisotropic Properties of the Base Material Used in Large-Scale Tank Test Parameters k1 k2 k3 n m νxy νxx Determined Values 1545 0.75 -0.1 0.45 0.35 0.17 0.43 Geosynthetic Layer A geogrid and a geotextile that are typical of such products and are currently being used in crushed aggregate base courses were selected for the LST testing. Both materials were installed according to the manufacturer specifications (i.e., no wrinkles or pretension). For experiments where the geosynthetic was to be placed at the interface between the base and the subgrade, the top 1-inch layer of the subgrade was replaced to allow for a level,
L-7 even surface for the geosynthetic to contact. A 9-ft by 9-ft square section of geosynthetic was cut from the roll and then laid over the top of the tank. The edges were trimmed around the perimeter of the Large-Scale Tank to ensure a proper fit. Once the geosynthetic was in place, U- shaped tacks were placed around the edges to maintain activation tension. This approach worked extremely well and kept the geosynthetic firmly in place. A similar procedure was used for experiments where the geosynthetic was to be placed in the middle of the base. The direct tension tests were conducted to determine the sheet modulus of geosynthetic products used in the Large-Scale Tank tests (see Figure L-6). Figure L-7 shows the relationships between the tensile force and the tensile strain for the tested geogrid and geotextile. âMDâ is the abbreviation for machine direction. âXMDâ is the abbreviation for cross-machine direction. Both the geogrid and geotextile in the machine direction had a smaller sheet modulus than those in the cross-machine direction. The ductility of geosynthetics in the machine direction was much higher than that in the cross-machine direction. (a) Tensile Test Setup for Geogrid (b) Tensile Test Setup for Geotextile Figure L-6. Direct Tension Test for Determining Sheet Modulus of Geosynthetics
L-8 Figure L-7. Relationships between Tensile Force and Tensile Strain for Geosynthetics Asphalt Concrete Layer A typical dense-graded HMA with a PG 64-22 unmodified asphalt binder was used in all flexible pavement experiments. Before the final placement of the asphalt layer on top of the base layer in the LST, a trial asphalt placement was conducted to select the most appropriate compaction technique. Three methods of compaction were attempted to see which would yield the best in-place compaction. The first method used a vibro-plate with the HMA placed and compacted in two 3-inch lifts. The second used a mechanical rammer for compaction and placed the asphalt in two 3-inch lifts. The third was also with the rammer, but with three 2-inch lifts. It was decided after finishing that the mechanical rammer was not suitable for HMA compaction and would not be used due to excessive difficulty in achieving an even surface. As a result of the trial compaction, it was decided to place the asphalt mixture using three 2-inch lifts and compacting for a longer period of time using the vibro-plate in order to achieve a target in-place density of 92 to 96 percent. The asphalt mixture was delivered with a dump truck from a local hot-mix plant supplier. The plant mix was dumped directly in front of the LST and shoveled in until 2.25 inches of uncompacted material was in place. A vibro-plate was then used for compaction of the lift by driving it around the perimeter of the tank from the outside edge to the inside for best compaction. This was repeated for a total of three 2-inch lifts after compaction. A thin lift nuclear density gauge was used at several locations around the surface of the tank to measure the in-place density of the compacted asphalt concrete surface layer. Figure L-8 shows pictures from the various steps during placement of the asphalt concrete layer. 0 1000 2000 3000 4000 5000 6000 7000 0 5 10 15 20 25 30 Te ns ile Fo rce (lb s/f t) Tensile Strain (%) Geogrid MD Geogrid XMD Geotextile MD Geotextile XMD
L-9 (a) (b) (c) Figure L-8. (a) Compaction of the HMA top lift (from outside edge to the center of the tank); (b) thin lift nuclear density gauge measurement; (c) finished HMA surface layer Plant-produced loose mixtures were sampled during placement of the material in the LST and were tested for theoretical maximum specific gravity (Gmm) in accordance with AASHTO T209. Table L-7 summarizes the test results for the maximum theoretical specific gravities. The data show that consistent Gmm values were obtained for the various materials and similar to the Gmm value from the mix design. Loading of the pavement structure was conducted seven days after the placement of the HMA layer in all experimental cases. This was done to eliminate additional uncertainties due to variability in the asphalt material properties as a result of oxidative aging. Cores were taken immediately after the completion of testing for each experiment (see Figure L-9) to measure the thickness and verify the in-place asphalt layer density. A total of 10 cores were taken at various locations in the asphalt concrete layer. Five of the cores were extracted from pre-determined locations matching a nuclear gauge density test. Table L-8 summarizes the asphalt layer thicknesses and core densities for all six flexible pavement experiments. Table L-7. Theoretical Maximum Specific Gravity Results for Plant-Produced Mixture Experiment Theoretical Maximum Specific Gravity, Gmm ID No. Measured Mix Design AC-Contr-B06 1 2.497 2.491 AC-Contr-B10 2 2.491 AC-Grid-B06 3 2.502 AC-Grid-B10 4 2.499 AC-Textile-B06 5 2.499 AC-Textile-B10 6 2.491
L-10 Figure L-9. Coring of the asphalt concrete layer in the LST Table L-8. Asphalt Layer Thicknesses and Core Densities Experimenta Asphalt Layer Thickness (inch) Core Densities (%) ID No. Average Standard Deviation Average Standard Deviation AC-Contr-B06 1 5.8 0.4 93.7 1.3 AC-Contr-B10 2 5.6 0.1 92.7 2.2 AC-Grid-B06 3 5.7 0.2 93.4 1.3 AC-Grid-B10 4 5.7 0.3 95.3 1.5 AC-Textile-B06 5 5.9 0.1 94.0 1.9 AC-Textile-B10 6 5.9 0.3 96.9 0.9 a Experiment No. 1 (control-thin CAB layer), 2 (control-thick CAB layer), 3 (geogrid-thin CAB layer), 4 (geogrid-thick CAB layer), 5 (geotextile-thin CAB layer), and 6 (geotextile-thick CAB layer). Selected asphalt core specimens (referred to as plant-mixed, field-compacted [PMFC]) from each of the flexible experiments were also tested for dynamic modulus in accordance with AASHTO TP79-13. Plant-mixed, laboratory-compacted (PMLC) samples from each experiment were also prepared at a similar corresponding in-place air void level and tested for dynamic modulus. The developed dynamic modulus master curve for each of the PMLCs was compared to the associated PMFCs, as shown in Figure L-10. It should be noted that the master curves are all reported at the average asphalt layer temperature of 79°F that was recorded during the LST experiment.
L-11 (a) (b) (c) (d) (e) (f) Figure L-10. Asphalt mixture dynamic modulus master curves at 79°F (average of three replicates): (a) Experiment No. 1; (b) Experiment No. 2; (c) Experiment No. 3; (d) Experiment No. 4; (e) Experiment No. 5; and (f) Experiment No. 6 0 1 10 100 1,000 10,000 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Dy na m ic Mo du lu s | E* |, k si Reduced Frequency, Hz PMFC average air voids 6.63% PMLC average air voids 6.37% 0 1 10 100 1,000 10,000 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Dy na m ic Mo du lu s | E* |, k si Reduced Frequency, Hz PMFC average air voids 8.10% PMLC average air voids 7.90% 0 1 10 100 1,000 10,000 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Dy na m ic Mo du lu s | E* |, k si Reduced Frequency, Hz PMFC average air voids 7.07% PMLC average air voids 7.26% 0 1 10 100 1,000 10,000 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Dy na m ic Mo du lu s | E* |, k si Reduced Frequency, Hz PMFC average air voids 4.72% PMLC average air voids 4.75% 0 1 10 100 1,000 10,000 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Dy na m ic Mo du lu s | E* |, k si Reduced Frequency, Hz PMFC average air voids 5.95% PMLC average air voids 6.36% 0 1 10 100 1,000 10,000 1.E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Dy na m ic Mo du lu s | E* |, k si Reduced Frequency, Hz PMFC average air voids 3.10% PMLC average air voids 2.75%
L-12 Portland Cement Concrete Layer The testing for rigid pavement in LST consisted of a half-slab PCC in order to allow for the measurements of slippage at the interface between the bottom of the PCC slab and the supporting base layer, as well as the deflection at the pavement edge. These two measurements were needed for modeling the properties of the composite concrete-base surface layer. Furthermore, an irrigation system was introduced to the rigid pavement experiments to model the effects of a wet base layer on the pavement responses, which is considered critical for modeling the mechanics of erosion in rigid pavements. The irrigation system consisted of a main plastic hose running around the diameter of the LST with holes placed every 6 inches and a set of quarter-inch soaker hoses attached to the main hose (on the side of the half tank) and running along the diameter of the tank. By connecting the hose to a water source, the base was flooded until it was evenly partially saturated. A picture of the irrigation system is shown in Figure L-11. The irrigation system was embedded in the crushed aggregate base layer at 1 inch below the top of the base. In order to provide the necessary vertical pressure and confinement for the base layer during the loading of the pavement, a full PCC slab with a 6-inch gap in the middle of the tank was constructed. The 6-inch gap was selected to allow for the installation of the instrumentations needed for the assessment of slippage at the PCC-base interface. Figure L-12 shows a schematic of the setup for rigid pavement experiments. Figure L-11. Completed irrigation system: soaker hose configuration
L-13 Figure L-12. LST top view for rigid pavement with irrigation system The specification for a typical PCC material used for roadway paving was selected for the LST experiments. Table L-9 shows the requirements for the PCC material. A minimum 28-day flexural strength of 650 psi was required. A private local company was hired to provide the ready mix concrete for the rigid experiments. To verify the quality of the concrete delivered during each LST experiment, samples were taken during the placement of the concrete and tested for fresh and hardened properties. Limited core and slab samples were also collected and tested at the end of each experiment to verify that the concrete was being cured properly. Placement of the concrete was performed using standard industry procedures. The PCC mix was brought in using a concrete truck. Slump, unit weight, and air content tests were conducted, and 21 compressive strength cylinders and seven flexural strength beams were prepared for curing and testing. While the test samples were being prepared, the concrete was poured directly into the LST ring. A standard vibrating rod was used to finalize the placement of the PCC mix. To ensure that the concrete cured properly, a curing blanket was placed over the slab and was hydrated throughout the day. The test results for fresh and hardened PCC properties are shown in Table L-10. The LST testing of the concrete pavement initiated when at least 85 percent of the specified 28-day flexural strength was achieved (i.e., 552 psi). Based on the maturity testing measurements, it was determined that a 13-day curing period of the concrete slab was necessary before the start of the testing. The same curing technique and duration were implemented for all rigid pavement experiments. Table L-11 shows the test results for the PCC samples taken out of the slab right after the completion of each of the rigid experiments.
L-14 Table L-9. Mix Design Requirements for PCC Material Test Test Method Requirements Water to Cementitious Ratio - 0.45 Maximum Flexural Strength at 28 Days (psi) ASTM C 78 650 Minimum Slump (inches) ASTM C 143 Initial 2 Maximum After Addition of HRWRa 4 Maximum Air Content (percent) ASTM C 173 or ASTM C231 No. 467 Aggregate 5.5 a High-range water reducer. Table L-10. Hardened and Fresh Properties of PCC Sampled During Placement Property Curing Duration (Days) Experiment No. a 7 9 10 Hardened Properties Compressive Strength, psi (ASTM C39) 5 4,501 4,438 4,387 7 4,625 4,562 4,505 9 4,785 4,672 4,598 12 4,876 4,904 4,861 14 5,203 5,128 5,024 28 6,086 6,073 5,984 Flexural Strength, psi (ASTM C78) 7 502 494 476 9 545 535 521 12 605 583 567 14 619 601 598 Fresh Properties Air Content, % (ASTM C138) 5.3 5.6 4.9 Slump, inch (ASTM C143) 3.75 4.25 4.00 Unit Weight (ASTM C173) 142.3 142.6 144.2 a Experiment No. 7 (control), 9 (geogrid), and 10 (geotextile). Table L-11. Post-Testing Properties of PCC (Core/Slab Specimens) Property Exp. No.a Test Results Average Standard Deviation Minimum Maximum Core Thickness, inch 7 5.953 0.005 5.949 5.958 Compressive Strength, psi (ASTM C39) 4,679 23 4,661 4,705 Flexural Strength, psi (ASTM C78) 562 N/Ab 557 567 Core Thickness, inch 9 6.075 0.003 6.072 6.077 Compressive Strength, psi (ASTM C39) 4,760 40 4,722 4,802 Flexural Strength, psi (ASTM C78) 566 N/Ab 563 569 Core Thickness, inch 10 5.853 0.004 5.849 5.856 Compressive Strength, psi (ASTM C39) 4,901 79 4,826 4,984 Flexural Strength, psi (ASTM C78) 579 N/Ab 577 581 a Experiment No. 7 (control), 9 (geogrid), and 10 (geotextile). b Only two flexural strength beams were cut from each slab.
