Mesostructure optimization in multi-material additive manufacturing: a theoretical perspective

Abstract

As multi-material additive manufacturing technologies mature, a new opportunity for materials science and engineering emerges between the scale of the microstructure and the scale of an engineering component. Here we explore the problem of "mesostructure optimization", the computational identification of preferred point-to-point distributions of materials structure and properties. We illustrate the opportunity with two simple example problems for 1D and 2D mesostructure optimization respectively, namely (1) a functionally graded cylinder that is computationally optimized to redistribute the Hertzian stress fields and (2) a thin plate made of digital materials computationally designed to simultaneously maximize bending resistance and minimize total weight. The mechanical performance of materials in these two problems is significantly improved as compared to any monolithic-material counterpart, including a topology-optimized monolith in case (2). These results point to new opportunities for multiobjective performance enhancement in materials.

Functionally Graded AISI 316L and AISI H13 Manufactured by L-DED for Die and Mould Applications

“…This possibility, therefore, enables integrating different functionalities in one part and enhancing its performance as it permits proper material design in one single component [16]. In this manner, a new material-centric fabrication paradigm is made available, characterised by pointwise control of the composition and material structure [17].…”

3D printed structures for modeling the Young’s modulus of bamboo parenchyma

“…With more advanced additive manufacturing techniques, this method could better capture the complex nature of biological cell walls. Multi-material techniques, [61], could allow for the construction of more complex models. Recently, fiber reinforced 3D printing inks for ink-writing have been developed [39,40].…”

Mesoscale design of heterogeneous material systems in multi-material additive manufacturing

“…While deformation can be simulated using finite element analysis, the mutation of the mesostructure design between neighboring iterations and the acceptance criteria of the new design are defined using global optimization algorithms. Because this process involves calling a 'black box-like' FEM solver, metaheuristic global optimization algorithms such as genetic algorithms, 54 particle swarm optimization, 55 and simulated annealing 39,56 are used to identify the optimal mesostructure.…”

“…39,[57][58][59][60][61] However, this has not been characterized in situ or confirmed experimentally. DIC enables the tracking of deformation for surface material voxels upon loading and allows for visualization of the interactions between the mesoscale material domains and the resultant strain or stress redistribution.…”

“…Such a material system can be fabricated by multi-nozzle extrusion, 35,36 inkjet 3D printing, 37,38 or directed energy deposition. [31][32][33][34] Since multi-material additive manufacturing provides a unique pathway to implementing an arbitrary mesoscale material distribution, Yu, Cross, and Schuh 39 proposed to use computational optimization schemes for mesostructure design. By investigating the resistance to contact loading and bending as examples, they showed that significant improvement in material efficiency, load bearing capacity, and weight saving can be achieved by properly distributing the mechanical properties across the component.…”