Enhancing the Bearing Capacity of Circular Footings Through Optimal Geometrical Design of Geogrid‐Reinforced Waste Foundry Sand Beds

This study explores waste materials, such as fly ash, bottom ash, slag waste, and construction and demolition waste, as sustainable alternatives to depleting natural granular materials for fill applications. Specifically, waste foundry sand (WFS), a by‐product of the aluminum metal‐casting industry, is evaluated as a viable filler material. The research focuses on optimizing the geometric parameters of geogrid reinforcement to develop the load‐bearing capacity and minimize settlement of circular footings placed on WFS beds, using a Sparse Spectra Graph Convolutional Neural Network (SSGCNN) approach. The primary objective is to predict footing pressure for layered soils reinforced with geogrids accurately. Experimental results identified optimal reinforcement parameters at u/B = 0.25, D/B = 3, h/B = 0.25, and l/B = 0.3. Numerical simulations indicated bearing pressures 10%–13% higher for reinforced cases compared to experiments, attributed to modeling assumptions. The proposed SSGCNN model achieved superior predictive performance with an RMSE of 9.15 kPa, MSE of 70.5, MAE of 6.96 kPa, and R 2 of 0.98, outperforming existing methods such as Escaping Bird Search (EBS), deep neural networks (DNN) and artificial neural networks (ANN). The findings demonstrate the potential of integrating WFS and geo‐grid reinforcement for sustainable foundation design with improved bearing capacity.