Investigation of mix design variables on the performance of Kenaf fibre-reinforced concrete through grey relational analysis/ANN-LM model
摘要
This work develops Kenaf fibre-reinforced concrete by changing the ratios of Kenaf fibres together with different amounts of Ordinary Portland Cement (OC), Fine Aggregate (FA), and Coarse Aggregate (CA) and evaluate the impact of the input variables on output resposnes specifically Workability Slump Flow (WSF), Compressive Strength (COM), and Flexural Strength (FLE). The experimental results were modelled using an ANN trained with the LM algorithm to improve predicted accuracy. From the result, ANN model exhibited a minimum validation MSE of 30.9438 at epoch 1 and a high correlation coefficient (R = 0.97627). Additionally, multi-response optimization using Grey Relational Analysis (GRA) identified the optimal parameter combination as OC1-FE3-CA3-KR2, corresponding to 380 kg/m3 of cement, 660 kg/m3 of fine aggregate, 1185 kg/m3 of coarse aggregate, and 15% Kenaf fibre content.