Improvement of bearing capacity of shallow foundations with cemented sand as replacement layer

Cement-treated sand replacement layers can enhance the settlement-controlled performance of shallow foundations on loose sand; however, the coupled effects of finite treatment geometry and curing-induced hardening on serviceability response have not been fully quantified. This study presents a systematic laboratory model-testing program on a rigid 60 × 60 mm square footing resting on a cemented sand layer overlying loose sand. The experimental variables included cement content (3–9%), improvement depth ratio (D/B = 0.5–2.0), improvement width ratio (W/B = 2.0–5.0), and curing period (3–28 days). To isolate the mechanical contribution of densification from chemical bonding, a control series of mechanically compacted sand layers (Dr​ = 95%) without cement was also tested. For the untreated loose sand, the bearing pressure at the serviceability settlement criterion s/B = 10% was established as the baseline (q10​ = 19.2 kPa). Mechanical densification increased q10​ up to 171.7 kPa at D/B = 2.0, whereas a thinner cemented layer (D/B = 1.0) with 3% cement surpassed this capacity. Increasing W/B enhanced load transfer and altered the observed failure mechanism from localized punching toward a slab-like response; at 9% cement and W/B = 5.0, q10​ reached 967.4 kPa. At high cementation and large W/B, the response became stiffer with reduced ductility, indicating a trade-off between capacity and brittle behavior. These findings provide design-oriented insight for selecting finite layer geometry to satisfy a settlement-based criterion (s/B = 10%) within the limitations of model-scale testing, and future numerical verification is recommended to further interpret stress distribution and failure mechanisms.