Experimental Evaluation of Iron Ore Tailings Sand as Infill for Geocell-Reinforced Railway Track Sub-Ballast Using a Uniaxial Cyclic Loading Box Test

Abstract Iron ore tailings sand (ore-sand) is a potential alternative for railway sub-ballast, although as a purely frictional granular material its direct application is limited. Geocell reinforcement may overcome this limitation, though its effectiveness under cyclic loading and with non-conventional infills remains insufficiently understood. This study evaluated the cyclic response of geocell-reinforced railway sub-ballast using a uniaxial rigid-walled box test across 900,000 cycles, assessing the effectiveness of geocell reinforcement for ore-sand and the ability of the test configuration to distinguish reinforced from unreinforced behavior as a screening tool. Six test sections using ore-sand and a conventional reference aggregate were subjected to long-term cyclic loading under three sequential stages: unreinforced and reinforced with polypropylene (PP) and high-density polyethylene (HDPE) geocells. Key indicators evaluated were cyclic recovery modulus, permanent deformation, and stress attenuation and distribution at the layer base. Geocell reinforcement increased cyclic stiffness by 16.0–26.2% and reduced vertical stress transmission by up to 10.4% in ore-sand sections, whereas it reduced stiffness in the reference aggregate, indicating a material-dependent response. Benefits were less consistent for permanent deformation, and neither material reached a stable deformation rate within 900,000 cycles. No conclusive performance difference was identified between PP and HDPE geocells. For both materials, stress attenuation was the most sensitive indicator of reinforcement contribution. Overall, results suggest the uniaxial box test can serve as a preliminary screening tool for evaluating geocell-reinforced granular layers, highlighting the need for plane-strain validation and further long-term testing.