Project description
The constructability and performance of non-proprietary thixotropic UHPC for thin bonded bridge deck overlay construction was investigated. Task 1 reviewed literature on UHPC for thin bonded overlays and non-proprietary UHPC. Crack monitoring of experimental slabs cast with UHPC thin bonded overlay (MoDOT project TR201704) and pull-off bond strength and flexural behavior for 16 composite overlay slabs cast with UHPC were determined. The overlay materials included a conventional concrete, a latex-modified concrete, and five non-proprietary UHPC mixtures made with different combinations of saturated lightweight sand (LWS) and expansive agent, fiber volumes of 2% and 3.25%, and overlay thicknesses of 1, 1.5, and 2 in. Flexural testing of the 16 overlay slabs made with substrate concrete used for bridge construction and thin bonded overlays showed that the use of UHPC overlay delayed crack opening and propagation in slab specimens. Task 2 involved fine tuning of two non-proprietary UHPC mixtures to enhance thixotropy. The effect of nanoclay, diutan and welan gums, cellulose-based viscosity-modifying admixture, anti-bleeding and segregation (ABS) admixture, styrene-butadiene rubber, and acrylic ester on key UHPC properties was evaluated. The use of 0.5% of the ABS admixture, by mass of binder, exhibited the highest thixotropy without mitigating mechanical properties of the UHPC. The effect of fiber volume was evaluated, and 3.25% volume was selected to enhance hardened properties. The robustness of the optimized UHPC was evaluated by varying the w/cm by ±10% and the sand moisture content by ±1%. These variations did not have a significant influence on rheology, shrinkage, and compressive strength. Task 3 proposed provisional performance-based specifications for the non-proprietary thixotropic UHPC for bridge overlay construction. The specifications included two proven mixture designs, batching and mixing protocols, test methods, and performance criteria for workability, mechanical properties, as well as autogenous and drying shrinkage. Recommendations were provided for minimum moist curing and compressive strength before bridge opening. Task 4 involved the implementation of the LWS17 and SRA1-3.25% mixture for the rehabilitation of two MoDOT bridge decks. A mockup slab was prepared to demonstrate the constructability of the UHPC. After 8 months of outdoor exposure, the slab did not exhibit any cracking. A field mockup test consisting of two slabs measuring 12 x 12 ft was conducted by the contractor using 1 in. twisted steel fiber instead of ½ in. microfiber. The temperature of the fresh UHPC ranged from 90 to 99 °F, which is higher than the specified value of 86°F. A curing compound, wet burlap, and plastic sheet were stipulated for 7 days before bridge opening. In-situ temperature of the UHPC dropped from 78°F at 1 day to 40°F after 2 days in the Route M overlay and to 20° F at 7 days. All compressive strength results were lower than the minimum specified values of 17,400, and 18,700 psi at 7 days and 28 days, respectively. The 2-, 7-, and 28-day compressive strengths of samples taken from the North bound lane of the Route Z Bridge are 38%, 7%, and 24%, respectively, lower than those from the South bound lane. Furthermore, samples taken from two batches along the same lane of Route Z Bridge exhibited some variation in compressive strength. The high spread in compressive strength between samples taken from different batches of the Z Bridge can be attributed to material quality control issues, especially in the case of excess water that was encountered with the saturated LWS. High strain values of -3000 and -2400 με (in compression) were recorded in the overlay at the midspan of the bridges in Route Z and M overlays, respectively.
