| dc.description.abstract | Both unpaved roads and pavement sustainability are dependent upon base course performance and longevity. With the depletion of natural resources and limited funding for necessary pavement rehabilitation, alternative aggregate resources, and addition of geosynthetics must be analyzed as potential solutions. Recycled concrete aggregate (RCA) and reclaimed asphalt pavement (RAP) are potential alternatives to virgin granular base (VGB) typically used. The addition of geosynthetics at the interface of the base course and subgrade can stabilize the base course section through separation, lateral restraint, and the tensioned membrane effect. This large-scale box study focused on granular base options for roadway applications. All base course sections were constructed atop a laboratory-blended subgrade material of pulverized kaolin, ASTM C33 sand, and water using a known moisture content-CBR relationship. Locally sourced RCA, RAP, and VGB were tested with and without geosynthetics (nonwoven geotextile, woven geotextile, triaxial geogrid, and combined nonwoven geotextile with triaxial geogrid) installed at the base course- subgrade interface for cyclic plate load tests on unpaved roads at varying load magnitudes. Based on performance and economy, nonwoven geotextile was selected for analysis in concrete paved test sections under 40 kN cyclic loading both before and after a rainfall event; a VGB control section was compared against two nonwoven-geotextile-stabilized test sections, one with RCA and one with VGB. Earth pressure cells recorded changes in vertical stresses at the base course-subgrade interface at varying lateral distances from the center of the load plate for both unpaved and paved test sections. Vertical displacements were recorded at the plate for all tests; for the paved tests, vertical displacement transducers also recorded along the diagonal of the loaded concrete slab from the loaded corner to a distance of three times the radius from the center of the load plate as well as on the corner of the other non-loaded slab nearest the loaded slab corner. Permanent and resilient deformation as well as interface stress reductions were analyzed for both the unpaved and paved test sections. The unpaved test sections were first analyzed in terms of accumulated permanent deformation, their unique permanent-to-resilient deformation ratios, and interface stress reduction. Replacement of VGB with RAP did not limit permanent deformation, so RCA and VGB were focused on. In both VGB and RCA unpaved test sections, the addition of geosynthetics limited the permanent deformation as compared with control sections, but the resilient deformations were very similar. Interface stress reductions and thus increases in stress distribution angle were achieved through the addition of geosynthetics in both VGB and RCA. The replacement of VGB with RCA had a greater effect than the addition of geosynthetics to the VGB both on decreased permanent deformations and increased stress distribution angles. The stress reduction method was effective at calculating resilient moduli (Mr) of the unpaved sections, but the modified Burmister solution yielded Mr that more closely reflected reductions in permanent deformation through the addition of geosynthetics. AASHTO (1993) design charts and methods were used to estimate the composite subgrade reaction moduli for the unpaved test sections. The three concrete paved test sections were then analyzed in terms of vertical displacements and base course-subgrade interface stress reduction. Permanent deformations in the concrete sections were reduced by the addition of nonwoven geotextile, and the replacement of VGB with RCA further reduced permanent deformation in the nonwoven-geotextile-stabilized sections before rainfall. Rainfall caused an increase in permanent deformation in the paved sections, but its effect disappeared within 1,000 load cycles. Interface stress reductions were observed through the addition of geosynthetics and the replacement of VGB with RCA in both pre- and post-rainfall concrete paved sections. The Westergaard (1926) method and the measured vertical displacements were used to calculate the subgrade reaction moduli and estimate expected tensile stresses in the slabs. Subgrade reaction moduli in the paved sections were approximately 60% to 70% of those calculated for the unpaved sections using the AASHTO (1993) design chart. Both the unpaved and paved sections were simulated in the KENPAVE software to estimate slab tensile stresses and while the vertical displacement reductions were reflected in subgrade reaction moduli, the change in expected tensile slab stresses was very small. | |
