Approximately 2.78 Mt of coal fly ash is produced in the Philippines, with a low utilization rate. Using fly ash-based geopolymer for construction will lessen the load sent to landfills and will result in lower GHG emissions compared to OPC. It is necessary to characterize the fly ash and optimize the geopolymer components to determine if it can replace OPC for in situ applications. The activator-to-precursor ratio, the water-to-solids ratio, and the sodium hydroxide-to-sodium silicate ratio were optimized using a randomized I-optimal design from the experimental results of 21 runs with five replicates, for a total of 105 specimens of 50 mm × 50 mm × 50 mm paste cubes. The engineering properties chosen as the optimization responses were the unconfined compressive strength (UCS), the initial setting time, and the final setting time. The samples were also ambient-cured with the outdoor temperature ranging from 30 °C to 35 °C and relative humidity of 50% ± 10% to simulate the on-site environment. Runs with high unconfined compressive strength (UCS) and short setting times were observed to have a low water-to-solids (W/S) ratio. All runs with a UCS greater than 20 MPa had a W/S ratio of 0.2, and the runs with the lowest UCS had a W/S of 0.4. The initial setting time for design mixes with a W/S ratio of 0.2 ranged from 8 to 105 min. Meanwhile, five out of seven design mixes with a W/S ratio of 0.4 took longer than 1440 min to set. Specimens with an alkali activator ratio (NaOH/WG) of 0.5 (1:2) and 0.4 (1:2.5) also had significantly lower setting times than those with an alkali activator ratio of 1. The RSM model was verified through confirmatory tests. The results of the confirmatory tests are agreeable, with deviations from the expected UCS ranging from 0 to 38.12%. The generated model is a reliable reference to estimate the UCS and setting time of low-calcium FA geopolymer paste for in situ applications.
Keywords: Class F fly ash; RSM; compressive strength; final setting time; geopolymer paste; initial setting time.