During in situ direct shear testing on the Questa mined-rock pile materials, New Mexico, USA, large rock fragments were encountered in the shear box. These large fragments can modify the shear strength and deformational characteristics of the material. To study the role of oversize particles on the results of direct shear tests, laboratory experiments and numerical modeling were conducted. In the experimental study, a shear box 5 cm × 5 cm in size was used. The oversize particle was simulated using stainless steel spheres that were placed along the shear plane. In the numerical investigation, a hybrid discrete-finite element model was utilized where the shear box was simulated by finite elements while the rock pile material was represented by discrete elements. Both experimental and numerical findings suggest that the presence of the oversize particle along the shear plane and at the center of the box causes an increase in the friction angle whereas the cohesion decreases for the material and the range of normal stresses used.
Oversize particles in a direct shear test or a triaxial test can change the shear and deformational characteristics of soil. This subject has been studied by some researchers with emphasis on the role of scalping on the mechanical behavior of rock fill dam material. Hennes  investigated the effect of particle size, shape, and grading on the friction angle of dry, rounded gravel and crushed rock. The shear test apparatus was 152.4 mm × 304.8 mm (6- inch × 12-inch) in size. He concluded that with the increase in the maximum particle size, the friction angle increases. The same conclusion was made as some complementary triaxial tests were conducted by this author. Vallerga et al.  did not observe any change in friction angle of granular material using traixial tests for the situation that the specimen diameter to maximum particle size ratio was larger than about 15. In the series of tests conduced, a fixed void ratio for the specimens was used irrespective of the maximum particle size. To study the effect of specimen size on the shear strength, several drained triaxial tests on saturated river sand and gravel and quarry material were conducted by Holtz and Gibbs . Very small (34.925mm × 76.2 mm), small (82.55 mm × 206.375 mm), medium (152.4 mm × 381 mm), and large (228.6 mm × 571.5 mm) specimens were tested. Test results on sand samples (passing No. 4 sieve size) compacted at 70% relative density showed similar friction coefficients irrespective of the sample size except for the very small samples that showed marked increase in shear strength. The authors attributed this increase in shear strength to the appreciable coarse sand of 4.763 mm (3/16-inch) maximum size that resulted in a maximum particle size to cell diameter ratio of 0.14.