Structural Performance of Finely Dispersed Wastes on Recycled Reactive Powder Concrete and Strength Prediction Using Neural Network Approach for Sustainable construction

Abstract

Today, the main emphasis in sustainable civil engineering world is on enhancing the performance and functionality of conventional materials using recent technologies and local materials. To address this issue, innovations on sustainable materials have been encouraged as a result of the continual accumulation of different wastes in Africa and their consequent environmental complications. Moreover, it is also a challenge to produce high-performance concrete for structural applications from locally produced materials since less costly components of conventional concrete are eliminated by more expensive elements (such as silica fume) to produce newly emerging concretes such as reactive powder concrete. The objectives of this study were to investigate the physical and chemical properties of selected finely dispersed local wastes; to identify the effect of finely dispersed local wastes through optimum mix design criteria on hardened properties of recycled reactive powder concrete (RRPC); to determine the microstructure of RRPC structure; to predict strength properties of RRPC using neural network approach; and to assess the potential of produced concrete for use in sustainable railway sleepers. In this study, two research programs were conducted for the development of RRPC and the entire tests. First, full replacement of silica fume by fly ash with glass powder and rice husk ash with glass powder were designed. Secondly, partial replacement of cement by teff straw ash with brick powder and limestone powder with animal blood were designed. Experimental design method was proposed as a research design. The performances of RRPC mixes were appraised in terms of compressive, split tensile, flexural strengths and microstructure investigation using XRD analysis. Artificial neural network (ANN) models were developed for strength prediction. Moreover, RRPC products were evaluated under static loading. The experimental results indicated that development of RRPC using hand mixing at standard curing from local wastes in this study was an interesting approach to solve raw material shortage for the current generation structural concrete, to fulfil the market demand of sustainable concrete products, to reduce waste disposal cost and related environmental issues. For full replacements of silica fume, 62.87 MPa maximum mean compressive strength were observed using 50% fly ash and 50% glass powder combinations after 28 days standard curing. Whereas, for partial replacement of cement, 54.32 MPa maximum mean compressive strength were observed using 5% replacements of cement by teff straw ash and brick powder combinations after 28 days standard curing. Furthermore, maximum mineralogical compositions (% mass) for Albite, Quartz, Portlandite and Halloysite, which are good indicators for improved quality, were observed from the XRD analysis. For RRPC strength prediction models developed in this study, the best coefficient of xx determination for the ANN prediction models was attained by 16-32-1 network architecture with LOGSIG transfer function for compressive strength and 14-14-1 network architecture with LOGSIG transfer function for flexural strength. As a potential RRPC product in this study, the tested sustainable railway sleepers have satisfactory compressive strength, 57.29 MPa after 28 days as compared to the minimum requirements by International Union of Railways of 50-60 MPa. Moreover, a flexural toughness of 93.49 kNm was observed after static loading of the test specimens. The artificial neural network models were shown to be an appropriate predictive tool for estimating the strengths using mix design proportions as well as early strength values that are obtained from the experimental investigations. Hence, these prediction models help to minimize concrete quality control associated costs and to avoid delay in related works. By far, recycling local waste materials will help to save the environment by avoiding utilization of biomass raw materials for construction and to reduce waste management and related disposal costs.