Optimization of Mix Proportions for Novel Dry Stack Interlocking Concrete Blocks Using ANN

This paper proposes novel concrete interlocking blocks made of fly ash and GGBS which are an alternative for the conventional concrete blocks. The artificial neural network (ANN) technique is used to estimate the mechanical strength of interlocking blocks and is verified with experimental investigation. The ANN model is based on the Levenberg–Marquardt principle which is executed using MATLAB. The inputs are given in the percentage ratio of cement: fly ash: crushed stone aggregate (FA): coarse aggregate (CA) for the process of learning, testing, and validation. The selected model is subjected to several trials in terms of mean square error, containing 4 input, 2 sets of 10 hidden layers, and one output components. In this study, a total of 2600 blocks of different mixes were tested as per IS 2185-1 (2005) to assess 3, 7, 14, 21, and 28 days’ strength. The experimental investigations were carried out in two phases. In the first phase, experimental investigations to identify the optimum mix proportions of cement, aggregate, fly ash, and ground granulated blast furnace slag to achieve desired compressive strength was carried out. In the second phase, the identified mix proportions were analysed using ANN to predict the compressive strength of interlocking blocks. The results indicate that the proposed ANN model developed to determine the mechanical strength and cost of interlocking blocks has excellent prediction ability.

This creative work has been published under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license which allows copying, and redistribution as well as adaptation of the original work provided appropriate credit is given to the original author and the conditions of the license are met.