Optimisation of strength development of bentonite and palm bunch ash concrete using fuzzy logic

Abstract

ABSTRACT This study presents green-concrete’s mechanical-properties modelling with cementitious portion replaced by PBA/BN ranging from 0-50% using fuzzy-logic. Incorporation of these admixtures in concrete encourages solid-waste derivatives utilization to achieve cost-effective and eco-efficient construction-materials which also mitigates environmental-degradation challenges due to CO2-emission during cement production and hydration. Assessment of the elemental-oxides inherent in the admixtures showed a good pozzolanic-activity with Al2O3+SiO2+Fe2O3 results of 74.75% and 72.03% for PBA and BN respectively. The maximum and minimum strength-responses were derived at 5%-replacement and 50%-replacement with experimental-results of 19.48-4.17MPa, 21.85-4.84MPa, 26.30-8.637MPa, 32.77-12.713MPa and 35.53-18.46MPa for 3-d,7-d,28-d,60-d and 90-d respectively. Fuzzy-logic technique was utilized to evaluate the strength development property of the green-concrete as it takes precisely into account the issue of vagueness of well-defined parameters to develop expert intelligent-model using Mamdani-FIS. The model’s performance was evaluated by using MLR-models, MAPE and RMSE. R2 of 84.905% and 99.1% was calculated for MLR-model and fuzzy-logic model respectively. RMSE and MAPE score of 2.406-2.14%, 6.5935-7.29%, 0.4183-4.62%, 0.2774-1.54%, 0.7544-2.06%, 0.4197-1.48% and 0.7912-1.66% were calculated for the initial and final setting-time, 3-d,7-d,28-d,60-d and 90-d concrete’s compressive-strength target responses respectively. The results obtained from the indicates better performance for the fuzzy-logic model in predicting the setting-time and mechanical-properties of PBA/BN-concrete.