Abstract: Human shoulder joints exhibit stable but highly active characteristics due to a large amount of soft tissues. Finite Element (FE) modelling plays an important role in enhancing our understanding of the mechanism of shoulder disorders. However, the previous FE shoulder models largely neglected the Three-Dimensional (3D) volume of soft tissues and their sophisticated interactions with the skeletons. This study develops a 3D model of the rotator cuff and deltoid muscles and tendons. It also includes cartilage and, for the first time, main ligaments around the joint to provide a better computational representation of the delicate interaction of the soft tissues. This model has potential value for studying the force transfer mechanism and overall joint stability variation caused by 3D pathological changes of rotator cuff tendons.  Motion analysis systems and Magnetic Resonance (MR) scans were used to collect shoulder movement and geometric data from a young healthy subject, respectively. Based on MR images, a FE model with detailed representations of the musculoskeletal components was constructed. A multi-body model and the measured motion data were utilised to estimate the loading and boundary conditions. Quasi-static FE analyses simulated four instants of the measured scapular abduction. Simultaneously determined glenohumeral motion, stress/strain distribution in soft tissues, contact area, and mean/peak contact pressure were found to increase monotonically from 0? to 30? of abduction. The results of muscle forces, bone-on-bone contact force, and superior-inferior movement of the humeral centre during motion were consistent with previous experimental and numerical results. It is concluded that the constructed FE shoulder model can accurately estimate the biomechanics in the investigated range of motion and may be further used for the comprehensive study of shoulder musculoskeletal disorders.

Key words: shoulder complex, biomechanics, finite element, tglenohumeral joint, subject-specific