Many studies have been conducted to evaluate the pattern of glenoid erosion from CT scans e.g. (Mullaji et al., 1994). The basic postulate was that the knowledge of the eroded geometry compared to the normal ones enabled the choice of the most appropriate design a fixation of glenoid component in presence of osteoarthrisis. The presence of mechanical loadings in soft tissues and bones probably requires the additional knowledge of the stresses within the bones. All these data are then expected to allow the surgery to be planned to obtain the most secure fixation in the available bone. Different complex shoulder models have been introduced in the literature combining the rigid body motion e.g. (Engin and Turner, 1989) — or beam modelling for bones — with the passive and active truss elements for modelling ligaments and muscles e.g. (Van der Helm et al., 1992). Although providing very useful information on forces and kinematics, most of these models are not interested in the bone stresses particularly at the epiphysis and at the articular surfaces.