Abstract: The arching effect is widely observed in geotechnical engineering and is typically studied using trapdoor tests. Traditional trapdoor tests ignore the conditions of passive trapdoors, such as tunnel uplift and anchor plate pullout. To address this gap, a servo-controlled lifting array-type trapdoor test apparatus was developed. A plane strain passive trapdoor test was conducted using dense sand as the backfill material. During the lifting process of trapdoor, the vertical load variations across the entire array and sectional displacements were collected, resulting in load-displacement response curve for the evolution of passive arching effect. The test results show that as the backfill height increases, load redistribution becomes more pronounced. This is evidenced by increases in the maximum arching ratio, ultimate arching ratio, and corresponding normalized displacement. Sectional displacements and sliding surfaces develop towards the backfill surface in a “funnel” shape. Shear strain concentrates on both sides of the trapdoor, forming a significant curved sliding surface. The angle between the sliding surface and the vertical direction is close to 0º at the base and increases with the embankment elevation. When the fill height is greater, new sliding surfaces develop inside the outer sliding surface as the trapdoor continues to rise. A modified limit equilibrium method calculation formula was derived based on the measured angles between the shear band and the vertical direction. The calculated maximum arching ratio closely matches the experimental results.

Key words: trapdoor test, tunnel uplift resistance, soil arching effect, shear band, soil arching ratio