In railways, the contact between wheel and rail undergoes extremely high contact pressures and tangential stresses caused e.g. by traction or braking. The tangential force is limited by the maximal adhesion coefficient (AC) and under certain conditions, e.g., caused by water in the contact, low adhesion occurs. Low adhesion interferes with traction and braking and is overcome by wheel-rail sanding, where sand grains are blasted in the wheel-rail contact to raise the AC. Research in this area has been almost exclusively experimental. Using today’s experimental possibilities, the underlying physical effects regarding the interaction of sand grains with wheel-rail surfaces are still not fully understood. In an ongoing project, experimental investigations are carried out to develop and calibrate a DEM model, which considers local effects within the contact region introduced by the sand particles. Sand grains will partially crush when entering the contact, causing plastic deformations on wheel and rail surfaces. ACs could be increased via form closure effects (grains penetrate metal surfaces), or solidified sand powder increases the effective contact area between sand and steel. As a first step to develop such a DEM model, the breakage behaviour of single sand grains under the high contact pressures in wheel-rail contact is considered. Experimental investigations of sand breakage under pure normal load were conducted for two types of rail sand called AT and GB, compare [1]. Fig. 1(a), (b) shows typical results of the crushing tests: after repeated breakage the sand powder solidifies forming one large cluster for AT sand and several small clusters for GB sand. Challenges in the DEM modelling are the unconstrained breakage of single grains and the material modelling of solidified sand powder under the applied high loads. A first simulation result reproducing the behaviour of AT sand is shown in Fig. 1(c).