Experimental and ARX model-based prediction of concrete strength with waste marble powder as replacement of aggregates

In this study, waste marble powder (WMP) was employed replacing aggregates in certain quantities. To achieve this objective, fine aggregates were substituted with WMP at different percentages of 0%, 10%, 20%, 30%, and 40%. This was done to compare the effects of various tests on the compressive strength (CS), splitting tensile strength (STS), and flexural strength (FS). First, the CS test was performed on the specimens. It was measured that CS of MA10% (10% aggregates were replaced with WMP) was 2.2% larger than that of reference concrete (REF). Furthermore, a noticeable decrease in the strength were observed for 40% of WMP when compared with REF. Then, the STS test was conducted to compare the test results with those of CS. As expected, it was found that the results of STS ordinarily followed the similar trend as CS. For the 10% WMP utilized as stepped replacement for fine aggregates, the significant enhancement in STS of concrete was up to 4.1%. However, while the quantity of WMP was improved to 30% and 40%, it resulted that the values of STS were affected in a downward trend. The FS test was also done. It was revealed that the maximum FS was achieved with 10% adding of WMP, and the minimum FS was obtained with 40% adding of WMP. Additionally, microstructure analysis was performed. It was found that WMP is a good binder with cement and the use of 10% WMP, as replacement of aggregates, is recommended to increase the capacity. Moreover, 20% and 30% of WMP can be utilized without the loss of the strength, however, the use of 40% WMP should be avoided. Moreover, an equation was derived to predict the capacity of concrete with WMP employing the experimental results and findings presented in the literature. Additionally, an estimation method of ARX (autoregressive with external input) models was used to more accurately predict the mechanical behavior of concrete with WMP.