Abstract

• The impact of mix proportions on the electrical resistivity and compressive strength of normal strength concrete incorporated GGBS and waste steel slag. • Multi-scale models were developed to predict the electrical resistivity and compressive strength of normal strength concrete. • A difference-based method for constructing target electrical resistivity and compressive strength is proposed. • Models of use to improve the prediction accuracy of normal strength concrete modified with GGBS and SSA learning are proposed. Steel slag (SS) and Ground-granulated blast-furnace slag (GGBS) were widely utilized in concrete as a portion replacement for normal and crushed stone aggregate to improve mechanical qualities, namely compressive strength (CS), and electrical resistivity (ER). Mechanical attributes obtained early in construction are now crucial in construction design. Electrical resistivity (ER) is a frequently used non-destructive approach for evaluating the microstructure and quality of concrete. Moreover, (ER) is a key concrete feature since it allows engineers to monitor the concrete easily. The effect of steel slag as a partial substitute on concrete electrical resistivity and compressive strength requires a mathematical model. In this way, 134 literature data were gathered and examined. During the modeling process, the most significant elements that impact the CS and ER of concrete with steel slag substitute were examined, including curing time (1–90 days), cement content (92–469) kg/m 3 , water to cement ratio (0.3–0.75), fine aggregate (620–773.3 kg/m 3 ), and steel slag content (0–365 kg/m 3 ). The effect of steel slag as a partial substitute on concrete electrical resistivity and compressive strength requires a computational formula. Designers and concrete producers may use mathematical models to combine concrete with steel slag to obtain specified compressive strength without any experiments. In this study, the Multi Logistic Regression (MLR), an Artificial Neural Network (ANN), and a Full Quadratic (FQ) model were employed to predict the CS and ER of concrete with steel slag aggregate substitution. According to statistical analysis, the artificial neural network model predicted the CS and ER of concrete, including steel slag replacement, better than the other models (MLR, PQ). It has a higher coefficient of determination of 0.992 for ER and 0.993 for compressive strength (CS). It has a small root mean square error (RMSE) of 1.39 MPa and resistance of 28.82 Ω.m for CS and ER of concrete.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call