Sugarcane bagasse ash as a partial replacement in concrete: a performance analysis

Bagasse is a byproduct of sugarcane juice processing. A cogeneration boiler burns bagasse to produce energy, and the resulting ash is known as sugarcane bagasse ash. The research's goal is to determine the strength and durability properties of concrete containing sugarcane bagasse ash as a partial replacement for cement. The high silica concentration in sugarcane bagasse ash makes it a potential pozzolanic material. Cement replaces sugarcane bagasse ash in six different percentages: 0%, 5%, 10%, 15%, 20%, and 25%. Compression strength is carried out for all the mixtures by utilising compression testing machine, also ultrasonic pulse velocity tests are conducted on all the mixtures to evaluate their quality. The assessments of tensile strength, flexural strength, and durability are done on the mixture that achieved the optimum compression strength. The samples are also tested for micro-structural analysis and element composition by utilising a scanning electron microscope and the energy diffraction spectrum, respectively. The experimental work reveals a decrease in workability as sugarcane bagasse ash percentages increase, but the compression strength values still surpass the required mean strength up to a 20% replacement. It is observed that the compression strength increased by 18.99%, 27.21%, 13.73%, and 6.62% for sugarcane bagasse ash blended concrete containing 5%, 10%, 15%, and 20%, respectively. A 10% addition of sugarcane bagasse improves compressive strength, which is considered the optimum percentage. Furthermore, the split tensile strength and flexural strength of concrete containing 10% sugarcane bagasse ash shows better performance, which is increased by 11.50% and 46.62%, respectively. It is also found that the weight loss and the compression strength loss were 8.44% and 21.44%, respectively. Additionally, this paper seeks to forecast compression strength using artificial neural networks. All the mixes are tested for strength after 7, 14, and 28 days of curing. The artificial neural networks is trained using four input parameters and one output parameter and the projected values from the artificial neural network model have a high correlation with the experimental data.