Optimization of Mechanical Interlocking Joints at Additively Manufactured Bi-Material Composite Interfaces

Continuing advances in the field of additive manufacturing (AM) have expanded its material possibilities, including the introduction of multi-material 3D printing, where a single structure can be made from several different materials arranged specifically to tune the material requirements for optimized function. In contrast to current industrial manufacturing methods, multi-material 3D printing enables small-scale, customized production with minimal waste. It is also typically the only feasible fabrication method for the complicated, multi-material architectures developed by computational design tools. To fully realize the opportunities AM has to offer for multi-material parts, the interfacial behavior and performance of 3D printed multi-material structures needs to be better understood. Recent research has shown that material mixing at the interface of Polyjet multi-material (PMP) printed parts has a noticeable negative impact on mechanical properties, which can be remedied by including a mechanical interlocking joint at the interface to improve strength and toughness of the part. Herein lies an opportunity to optimize the interlocking joint designs for improved properties. In this work we explore the impact of varying interfacial interlocking joint geometries on the overall mechanical behavior of PMP printed parts. We implement a mechanics-driven approach to establish a relationship between geometric parameters and mechanical properties, using first-principles to predict the strength of different interface geometries. By combining this with the computational and predictive power of machine learning, the design space for interlocking interface geometries can be more comprehensively mapped to provide unparalleled insight into the design principles of interlocking joints at multi-material interfaces. This work thus has important implications for the continued improvement of multi-material printing and the realization of computationally-predicted mechanical performance of meta-materials.