Granular avalanches are prone to segregate their particles by their sizes, as a result of shearinduced dilatation. This mechanical response of the granular media facilitates small particle percolation through gaps (kinetic sieving), after which large particles are squeezed upwards by surrounding small particles (squeeze expulsion). The development of simple continuum models for particle-size segregation has provided a better understanding, nevertheless determining the functional dependance of the segregation rates remains as a challenge [1]. In this work we improve the comprehension of the segregation rates, by running simple shear experiments of bidisperse granular materials. The experimental shear cells consists of two parallel plates that rotate over pivot points to generate an oscillatory shear flow within the granular media. Shear cell configuration has been used extensively to study granular segregation [2, 3]. For the present work, two different shear cells have been used: a 2D and a 3D cell. The 2D cell width ranges from 55 to 280 mm, in 15 mm steps. The cell is filled with Polyoxymethylene (POM) disks of different diameters but of the same thickness as the cell gap (around 5 mm). The 3D shear cell consists in two parallel plates that pivot in an upper point located at 80 mm from the movable bottom of the load cell. The separation between the plates is set to 45 mm while the plate thickness is 68 mm. For the 3D shear cell, we used the refractive index match (RIM) technique for visualizing the intruder. Poly(methyl methacrylate) (PMMA) spheres, transparent for the substrate and opaque for the intruders, are used in the experiments along with Triton X-100 to achieve the refractive index matched mixture. We ran experiments using single particle intruders immersed in a differently sized medium in both shear cells. Trajectories and local velocities are measured. An estimation of the segregation rates as a function of the size ratio is also obtained for both cells. Size segregation in the 2D shear cell is only observed for large intruders because of the constraints imposed by the planar configuration. In contrast, in the 3D shear cell small particle percolation occurs and it may stop at middle positions in specific cases, while large intruders didn’t segregate under certain conditions. These observations raise the question: what is initiating and restraining segregation? Within the presented framework, we suggest the interstitial fluid and size ratio may prevent and restrain segregation.