A concurrent two-scale FC-SCA method for mechanical characterization of additively manufactured components

In the field of metal additive manufacturing (AM), inevitable process-induced geometry defects and heterogeneous grain structure can significantly deviate the mechanical performance of the as-built components from their as-designed counterparts, leading to the multiscale characteristic. This poses a pressing challenge for numerical simulations to predict their mechanical performance with high accuracy and high computational efficiency. To address this challenge, we propose a novel concurrent two-scale finite cell-self-consistent clustering analysis (FC-SCA) method in this study. In this method, the macroscale behavior of the component is modeled using the image-based finite cell method (FCM), while the microscale material behavior is captured via the crystal plasticity SCA method at the integration points of the FCM model. A two-scale coupling algorithm is developed to enable concurrent calculations between these two methods. The FC-SCA method offers two key advantages: efficient meshing of the component with complex geometry directly from its X-ray computed tomography image data, and fast yet accurate evaluation of local material responses governed by the as-built material microstructure. Both advantages are crucial for the mechanical characterization of the AM-fabricated component, particularly lattice structures. The computational accuracy and efficiency of the proposed method are validated by two representative case studies: tensile test of a dog-bone specimen and compression test of an octet lattice cell, both fabricated by metal AM. The numerical predictions show good agreement with experimental data. Notably, the proposed method achieves over 500 times the computational efficiency of the conventional FE² (finite element square) method, with a relative error of below 3% in the tensile test. The lattice compression test is completed within an acceptable runtime, which is computationally prohibitive for the FE² method. In addition, the proposed method has the efficiency much higher than that of FE-SCA for the studied problems. The demonstrated accuracy and efficiency position the FC-SCA method as a powerful tool for digital-twin-compatible simulations of AM components, effectively bridging the gap between as-designed and as-built mechanical performance prediction.