Please use this identifier to cite or link to this item:
Title: Experimental Design of Concrete Mix Proportioning Using Some Industrial By-Products
Authors: Bansal, Varinder Kumar
Supervisor: Bansal, Prem Pal
Kumar, Maneek
Batish, Ajay
Keywords: Industrial by-products;Compressive Strength;Taguchi Method;GRA
Issue Date: 21-Jul-2020
Abstract: With ever-increasing quantity of industrial by-products and waste materials being produced regularly, solid waste management has become the principal environmental concern in the world. The use of these materials makes concrete production economical and in addition, helps in resolving their disposal problems. Using varied industrial by-products individually or in combination as a partial replacement material in concrete has been accepted by many codes worldwide. These industrial by-products are known to have pozzolanic and cementitious properties. Several studies have been carried out to investigate the synergic effect of using ternary blends containing different supplementary cementitious materials (SCMs) in concrete. The shortcomings of individual SCMs in meeting multiple and often conflicting demands of concrete properties can be overcome by using combinations of two or more SCMs. However, the current research work proposes to use a secondary as well as a tertiary combination of these industrial by-products/waste materials to take advantage of their synergic effect. The use of the design of experiments in utilizing the by-products in proportioning concrete mixes is certainly need of the hour. This study is an innovative step towards scientifically applying the Taguchi’s experimental design along with Grey Relational Analysis (GRA) for concrete mix proportioning incorporating various industrial by-products in ternary combinations both as a partial replacement of cement as well as fine aggregate. The study aims at finding the optimal use percentage of some industrial by-products such as fly ash (FA), ladle furnace slag (LFS), copper slag (CS), electric arc furnace slag (EAFS), iron slag (IS) and glass powder (GP) as substitutes to cement and fine aggregates in ternary combinations as well as investigating the effect of these industrial by-products on the strength and durability properties of concrete. The results obtained from the experiments for strength and durability properties were analysed with ANOVA and Grey Relational Analysis (GRA) method to find the optimal industrial by-products and their percentage replacement for cement and sand. Finally the models were developed using ANN and ANFIS to predict the strength and durability properties. The results of ANOVA study for strength properties show that the highest compressive strength was achieved by replacing 10% of cement and 30% of sand with by-products at all curing periods. Fly ash was the optimum by-product among the chosen by-products for the replacement of cement to achieve the highest compressive strength and split tensile strength. For all water to binder ratios, the percentage of by-product to be used as binder is the most significant factor that affects compressive and split tensile strengths. The results of ANOVA for durability properties show that for lowest depth of water penetration in concrete, the optimal combination of by-products are 10% of fly ash as cement replacement and 30% of iron slag as sand replacement at w/b ratio of 0.40 and at 90 days of curing. For lowest depth of wear in concrete, the optimal combination of by-products are 10% of fly ash and 20% of electric arc furnace slag at w/b ratio of 0.40 and applicable for 28 & 90 days of curing age. Fly-ash and glass powder are the most suitable replacements for cement and sand respectively, at 28 days curing for achieving the lowest depth of water penetration. GRA for strength and durability properties taken together show that 10% of the fly ash is the optimal parameter for cement replacement for achieving desirable strength and durability properties at all the curing ages and water to binder ratios, whereas, 20% is the optimal replacement level of sand replacement for all the water to binder ratios and at all the curing ages. The confirmation experiment conducted to verify the optimum mix proportions of concrete made with industrial by-products resulted in higher compressive and split tensile strengths and lesser depth of water penetration and wear than all other mixes including the control mixes at 28 days of curing. This improvement in strength and durability properties is attributed to the synergic effect of optimal by-products due to their pozzolanic activity and cementitious properties. Additionally, the SEM images of the concrete containing the optimal parameters of the industrial by-products were compared with the corresponding control concrete SEM image. The microstructure of concrete made with combination of industrial by-products show denser and more uniform structure than the control concrete structure and justifies the improved strength and durability properties. The XRD spectra shows the existence of hydrated phases such as quartz, calcium hydroxide, calcium silicate hydrates, calcium silicate, ettringites and aluminium oxide silicate. XRD analysis show that the concrete mixture containing industrial by-products has less quantity of calcium hydroxide [Ca(OH)2] than the control mix at 90 days of curing. The Ca(OH)¬2 present in the mix containing industrial by-products is consumed through the pozzolanic reaction of different by-products, and thus, converted this into calcium silicate hydrate gel (CSH gel) which lends strength and durability to the concrete. This fact helps in improvement of the strength and durability of concrete containing partial replacement materials of cement and fine aggregate than the control concrete. The neural network and ANFIS models could predict the strength and durability properties of concrete containing industrial by-products with satisfactory performance owing to their distributed and parallel computing nature. The predicted values of the compressive strength of concrete from ANFIS model were found to highly accurate. Moreover, a comparison of the performance indices showed that the ANFIS model provided better results than the ANN model.
Appears in Collections:Doctoral Theses@CED

Files in This Item:
File Description SizeFormat 
Thesis for Libraray CD.pdf10.18 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.