A thorough study on the effect of red mud, granite, limestone and marble slurry powder on the strengths of steel fibres-reinforced self-consolidation concrete: Experimental and numerical prediction

Abstract

Natural resources protection has become an essential issue for engineers worldwide. Using waste materials helps to reduce natural resources usage and may lead to keeping the environment safe and clean. Therefore, in this study, a comprehensive evaluation was carried out on the impact of waste materials as a fractional substitute of cement on the compressive and splitting tensile strength of self-consolidation concrete (SCC). A total of 145 mixtures were cast, and slump, compressive and splitting tensile strengths were measured at 28, 56 and 90 days of curing. Red mud (RM) was used as a cement substitute at five contents: 0, 2.5, 5, 7.5 and 10% in terms of weight. In addition, granite (G), limestone (LS) and marble slurry powder (MSP) were used separately as a filler to improve the properties of SCC at four contents of 25, 50, 75 and 100 kg/m3. Furthermore, steel fibres (SF) were utilized at two volumetric percentages of 0.1 and 0.2% further to enhance the compressive and splitting tensile strengths of SCC. Results showed that the highest compressive and splitting tensile strength were achieved when 2.5% RM was included. Moreover, the enhancement impact of MSP was greater than G and LS on the strength of SCC. Based on the acquired experimental results, novel correction factors for the evaluation of compressive and splitting tensile strengths of SCC containing RM, G, LS, MSP, SF were proposed employing artificial neural network (ANN), neural networks type group method of data handling (GMDH-NN) and combinatorial algorithm group method of data handling (GMDH-Combi) as the machine learning approaches. In addition to the developed models' performances, the effect of each contributing parameter on the properties of SCC was comprehensively investigated in each model using parametric and sensitivity analysis to inspect the developed models’ generality and robustness. In addition, a new correlation was introduced to anticipate the splitting tensile strength based on compressive strength when RM, G, LS, MSP, SF are incorporated. A spreadsheet was also designated using the developed models to provide a simple calculation tool for the waste and by-product included SCC compressive and splitting tensile strength evaluation.