Experimental and machine learning approaches to investigate the properties of foamed concrete made with surkhi and dolomitic limestone as cement replacements

Sustainable dolomitic limestone (DL) and surkhi (SK) powders replace cement in lightweight foam concrete (LFC) for the first time. A total of fifteen LFC mixes were produced, which included reference mixtures. DL was used as a partial substitute for cement, with varying proportions ranging from 0 to 20% by weight. SK was employed as a partial replacement for cement, with weight fractions of 0, 15, and 30%. DL and SK were analyzed to determine how their integration affected slump flow, j-ring, oven-dry density, setting times, compressive strength (fc) and flexural strength (ff), modulus of elasticity (ME), ultrasonic pulse velocity (ν), water absorption (WA), apparent porosity (AP), air permeability (ko), microstructure, and pore distributions. In this investigation, Artificial Neural Network (ANN) is used to predict the compressive strength of LFC with help of K-fold cross validation. The assessment techniques are coefficient of correlation (R), coefficient of determination (R2), root mean square error (RMSE) and mean square error (MSE) metrics on training and testing datasets. All the LFC’s aspects were considerably improved by adding 10% DL and 15% SK. While replacing cement with them reduced the LFC’s density, it enhanced its workability, mechanical characteristics, and pore properties. Adding 10% DL and 15% SK at 28 days improved fc, ff, ME, ν, AP, WA, and ko. fc, ff, ME, and ko increased 2.5, 19, 6, and 4.3%, respectively. When LFC incorporated with 10% DL and 15% SK at 28 days, AP dropped 8.5%, WA reduced 8.6%, and ko rose 20%. The regression analysis utilizing the ANN method for K = 4 yielded the prediction accuracy (R2 = 0.94) of fc. The scanning electron microscopy revealed that the inclusion of a small quantity of DL and SK improved the microstructure of cement paste and accelerated the process of hydration. Overall, replacing 25% cement with DL and SK enhanced all evaluated qualities and reduced carbonation pollution, which will be contributing to the sustainability in the construction.