Computational Modeling of Desiccation Properties (CW, LS, and VS) of Waste-Based Activated Ash-Treated Black Cotton Soil for Sustainable Subgrade Using Artificial Neural Network, Gray-Wolf, and Moth-Flame Optimization Techniques

Abstract

Artificial neural network (ANN), gray-wolf, and moth-flame optimization (GWO and MFO) techniques have been used in this research work to predict the effect of activated sawdust ash (ASDA) on the crack width (CW), linear shrinkage (LS), and volumetric shrinkage (VS) of a black cotton soil utilized as a subgrade material. Problematic soils or black cotton soils are not good pavement foundation materials except that they are pretreated in order to meet the basic strength characteristics required for roads in Nigeria. Due to this reason, there has been ongoing research to evaluate the best practices in which black cotton soils can be favorably utilized in earthwork construction. On the other hand, there is a huge concern on the solid waste management system in the wood processing environment and the recycling of sawdust into ash and its reuse as an alternative binder has offered a sustainable disposal system. The work tries to use AI-based techniques to predict the crack and shrinkage behaviors of BCS treated with saw dust ash activated with alkali materials. There was appreciable improvement in the shrinkage and crack parameters over the 30-day drying period due to the addition of ASDA. The intelligent model results showed that the three techniques successfully predicted the CW, LS, and VS with a performance accuracy above 90%, while ANN produced the minimal error in performance outperforming the other techniques. Sensitivity study showed that the drying time (T) was the most influential of the studied parameter. Hence, soil stabilization has shown its potential system of waste management in the wood processing industry.