Predicting Bearing Capacity of Strip Footing on Slope Using a Soft Computing Technique

The paper proposes a new hybrid soft computing technique to predict the bearing capacity of a strip footing on c–ϕ (cohesive-frictional) slope. The finite element analysis based on the Plaxis code is adopted to simulate the strip footing on c–ϕ slope. The influences of design parameters (i.e., the slope angle, distance from the foundation to the slope, and soil strength) on the bearing capacity are investigated. The obtained finite element analysis (FEA) results are compared to the previous study. A good agreement between the prediction from FEA results and those of the prior research is received. The FEA results are prepared in the design table for practical design. Based on FEA results, a machine learning approach, Multivariate adaptive regression spline (MARS), is adopted for proposing an empirical equation and considering the impact of each investigated parameter on the bearing capacity. These results can be a good reference for practical engineering in designing strip footing on c–ϕ slope.