Stabilization of subgrade soils during Interstate 95 lane widening 2a 2b FIGURES 2a and 2b Original (2a) and final (2b) design cross sections thick MSL. As this might not be adequate to provide support for the highway design ESALs, a second layer was located at the top of the MSL/haul road and the pavement section. The multiaxial geogrid selected had high flexural stiffness to provide the most stable surface possible for placing the initial layer of fill over the extremely soft soils below. To address the potential detrimental effects of groundwater and drainage, it was decided to construct the first (lower) 12 inches (305 mm) of lift using an 30 Geosynthetics | August September 2019 open-graded aggregate in order to provide a capillary break and prevent water that might enter the fill from either above or below from becoming trapped and potentially causing frost heave or weakening of the MSL. The second (upper) lift was constructed using a dense-graded aggregate base to provide a smoother and more stable surface that was suitable for construction. A nonwoven geotextile was placed above the first lift and below the second layer of multiaxial geogrid in order to maintain separation of the aggregate layers above from any fine-grained soil that might migrate into the capillary break layer. To direct groundwater from entering the drainage system, the MDTA standard pavement drain was modified to accommodate the new pavement section and lateral finger drains were added under the pavement. In areas where a third lift was used, the multiaxial geogrid was placed above the second lift, and the layer was completed with dense-graded aggregate base. Figures 2a and 2b show cross sections of the design. The geotextiles specified were 6 ounces per square yard (203 g/m2) nonwoven (Class SD) and 8 ounces per square yard (271 g/m2) nonwoven (Class SE). The specified geogrids were integrally formed multiaxial geogrids with nominal aperture dimensions of either 1.6 inches (41 mm) or 2.24 inches (57 mm) that have been calibrated to the Giroud-Han method based on full-scale traffic testing. Once performance of the design was verified using test strips, the contractor worked with MDTA to put an appropriate change order in place, and the solution was implemented. Figures 1, 3 and 4 present photographs of the construction at the site. Results The original design included 12 inches (305 mm) of asphalt, 12 inches of GAB