Development of a high-capacity apparatus with a novel structural design for large-scale direct shear testing of soils

Testing coarse-grained soils under in-situ stress conditions is limited by the size and capacity of available equipment, often requiring particle size scaling that can alter the shear response. This study presents the design, construction, and performance evaluation of a novel large-scale direct shear apparatus with high load capacity. The load-controlled apparatus comprises reinforced concrete and steel reaction structures that apply normal and shear forces to shear surfaces of 1000 mm or 1500 mm in diameter. So far, normal loads up to 2000 kN (2562 kPa normal stress) have been applied on the 1000-mm-diameter specimen. Benchmark tests using medium-graded sand (no fines, maximum particle size of 2 mm) confirm the equipment’s reliability and repeatability, yielding results comparable to conventional small-scale devices. The developed apparatus combines large-scale loading capability with the ability to test saturated specimens; a feature not found in equipment of this scale. This study introduces several novel design aspects: a reinforced concrete reaction system enabling saturated testing with its pool-like configuration and enhancing operational efficiency through below-ground placement; a stabilization mechanism for the lower box that reduces structural forces on components; a half-open lower box design ensuring shear force transfer without eccentricity; and a gap generation technique eliminating the need to lift the upper box while maintaining uniform gap thickness. The observed non-linearity in failure envelopes at low stress levels aligns with the literature and highlights the importance of testing within actual in-situ stress ranges. Design considerations presented herein provide valuable insights for developing future large-scale direct shear setups. • Direct shear test for 1-m diameter, submerged or dry specimens, up to cobble size • Testing up to 2.6 MPa to avoid unsafe strength parameter selection at high stresses • Frictionless setup with floating upper shear box removes the biggest source of error