Geotechnical investigation of nanomaterial-based black cotton soil stabilization: an experimental and predictive modeling studies

Abstract

Black Cotton soil causes damage to the lightweight structures and pavements built on it due to its volumetric behavior corresponding to the moisture variation. Therefore, many researchers tried chemical stabilizers for the stabilization of expansive soil, but these stabilizers are not environmentally friendly and not durable as well. Therefore, nanomaterials are used by various researchers and stabilize the expansive soil, but these stabilizers are very expensive and being nano in size can harm the geo-environment. Therefore, in this study nano rice husk ash produced from agricultural waste is used for the stabilization of expansive soil. Two different types of nanomaterial had produced one synthesized through 60 hours of dry milling and the other through 2 hours of dry milling followed by 5 hours of wet milling using ethanol. The main objectives of this research include the use of both types of nRHA to improve the index and engineering properties of Black Cotton soil. The second objective is to assess the efficiency and durability of nRHA-stabilized Black Cotton soil through a cyclic wet-dry study. The third objective is to study the physico-chemical impact of nRHA on Black Cotton soil stabilization. The fourth and final objective is to develop a predictive model to predict the swelling potential of expansive soil. To achieve these objectives, first, the index properties like Atterberg’s limit, free swell index, engineering properties like swelling potential, swell pressure, hydraulic conductivity and unconfined compressive test were studied by using both types of nRHA. The microstructural interaction between soil and nRHA was also studied through a Scanning Electron Microscope and X-ray diffraction study. Second, the efficiency and durability of 7 hours nRHA were studied using swelling and shrinkage study in terms of lateral, vertical and volumetric deformation through alternate wet-dry cycles. The 7 hours nRHA effect was studied through SEM, XRD, cation exchange capacity and pore structure study along with specific surface area by conducting methylene blue test and Brunner Emmit teller (BET) test. Third, the environmental impact study was conducted through pH, Total residue, volatile residue, biochemical oxygen demand and chemical oxygen demand in the leachate collected through a centrifuge machine. Fourth, the predictive model is developed by using Polynomial ridge regression and the performance of the model is measured through the coefficient of correlation, root mean square error and mean absolute error. Finally, the sensitivity analyses were done to find the impact of each independent variable on the predicted output of swelling potential. The results reveal that 0.4% of the 7-hour nRHA yields the greatest improvement in geotechnical properties and swell-shrink potential. XRD and SEM analysis confirms the formation of calcium aluminum oxide (CAO) and calcium aluminum silicate hydrate (CASH) gel in the 0.4% of 7-hour nRHA treated soil, whereas the 60-hour nRHA treatment at 0.4% produces only CAO gel, resulting in comparatively lesser improvement. This concludes that 0.4% is the optimum percentage of stabilizer and 7 hour nRHA performed better than 60-hour nRHA treated soil. Therefore, the further study was only conducted through 7-hour nRHA. Cyclic wet-dry tests further demonstrate that nRHA-treated soils exhibit reductions in cation exchange capacity, specific surface area, and pore volume, thereby minimizing swell-shrink behavior. Adsorption isotherms (Type IV) indicate mesoporous frameworks dominated by capillary condensation. By the fifth cycle, soils treated with 0.4% nRHA show notably lower mesopore and macropore volumes, explaining the overall decrease in swell-shrink potential compared to untreated soils. The environmental impact of the nRHA-treated BC soil leachate treatment was also evaluated, with results showing that all tested parameters pH, Total Residue, Volatile Residue, Biochemical Oxygen Demand and Chemical Oxygen Demand were within permissible limits and there was no threat to the geo-environment due to stabilization with nRHA. Additionally, this research develops a Polynomial Ridge Regression model to predict soil swelling potential based on ten initial parameters, subsequently refined to seven key variables through sensitivity analysis. Training on 70% of the data and testing on the remaining 30%, the model achieves an R² value above 0.85 and maintains RMSE and MAE below 2.2 and 1.65, respectively on both datasets. K-fold cross-validation and residual analysis validate the model’s reliability. Overall, the study underscores the effectiveness of nRHA as an efficient, durable, environmentally friendly, economic stabilizer for Black Cotton soil and highlights the utility of advanced statistical modeling in optimizing soil management practices and construction planning to reduce the risk of damage from expansive soils.