Abstract
Marine coral sand-clay mixtures (MCCM) are widely used in marine engineering, with their mechanical behavior strongly influenced by clay content. This study investigates the effects of 3D-printed triaxial geogrid reinforcement on MCCM through triaxial testing. Based on the experimental results, a dataset was established, while a novel machine learning model named GP-BPNN was proposed, integrating genetic algorithm (GA), particle swarm optimization (PSO), and backpropagation neural network (BPNN). This model was applied for the first time to predict the strength of MCCM. Results show that lower clay content, more reinforcement layers, and higher confining pressure significantly enhance the strength and cohesion of MCCM, with little effect on the internal friction angle. The strength first decreases, then increases, and finally decreases again with increasing water content. Particle breakage is influenced by clay content and water content; moreover, fractal analysis reveals a linear relationship between the breakage rate and the fractal dimension. SEM images reveal the interaction between MCCM and the geogrid. Additional stress and matrix suction analyses highlight the effects of reinforcement layers and water content on the strength. These findings offer insight into triaxial geogrid-reinforced MCCM behavior and provide guidance for marine engineering construction.