A multiscale method bio-inspired by bone architecture to enhance stiffness to weight ratio of mechanical parts

• This paper proposes a bone bio-inspired multiscale method for mass reduction • The multiscale method integrates two bone scales: macroscale and mesoscale • The mechanical performance was evaluated through the stiffness to weight ratio • A three-point bending test was performed using both experimental and FEA • The proposed method increases stiffness by 24% compared to topological optimization A current challenge in mechanical engineering is reducing the mass of mechanisms while maintaining performance, in order to decrease energy and resource consumption. This challenge also exists in Nature, and the architecture of long bones provides an excellent example of how Evolution has solved this problem. Bone architecture is complex, highly adapted to mechanical stresses, and yet extremely resource efficient. The use of such complex structures is now feasible in industry thanks to the advent of additive manufacturing. In this paper, a bone bio-inspired, multiscale mass reduction method for mechanical parts is presented, aiming to enhance the stiffness to weight ratio. A two-scale approach is developed: the macroscale (bone shape, topological level) and the mesoscale (trabecular bone structure, structural level). This multiscale method is then compared with each method applied separately. For this comparison, the three-point bending test (representative of industrial demands) is used. Tests were performed both experimentally and through finite element analysis. The experiments show the efficiency of the multiscale method, with improvements of 24% compared to the topological level method and 7% compared to the structural level method. The finite element analysis results support and explain this trend, indicating that the multiscale structure achieves a better distribution of internal stresses. Thus, this study demonstrates that combining two mass reduction methods operating at different scales yields better results than either method alone. The results obtained for this multiscale method are highly encouraging and suggest potential applications for industrial parts in the automotive and aerospace sectors.