Prediction of slope stability with comparative study between unstable slopes reinforced with breast wall and soil nailing using kNN and least square regression analysis

The paper attempts to analyze the effectiveness of breast wall, retaining wall and soil nailing reinforcements in enhancing the stability of unstable slopes. The considered study area is an ongoing project of highway construction in Nagaland, India. The concerned stretch divided into two main typical cross sections with the designed reinforcement of breast wall yielded a potential stable cut slope with factor of safety of 1.3 imparting 120 kPa shear resistance to the inherent geomaterial whereas the road embankment was stable with 2.615 when supported by retaining wall due to the generation of 100 kPa normal stress. Based on combined investigations in the considered stretch, the cut slopes were analyzed using three parametric variations of soil nailing, to overcome the limitations of designed reinforcement and improve the stability condition of the slopes. The alternative resulted in an average 63.8% enhancement in the slope stability. The staged construction methodology helped in solidifying foundation soil which decreased the lateral pressure and enhanced the shear resistance. Further, to contribute to the existing literature of slope stability and widen the prospect of acquired dataset to the field circumstances similar to the simulating conditions, k-nearest neighbor and least squares regression analysis were utilized to develop generalized mathematical models to predict the slope conditions and factor of safety for the inherent geomaterial properties and for the slopes reinforced with soil nailing. The developed models depicted an average R2 value of 0.80 and accuracy of 81.7%, thereby validating employment of the models for slope stability prediction.