Infill optimization in 3D printing is something that is constantly being tinkered with by designers, mathematicians and researchers around the world. The reason is ultimately to find ways to produce more with less.
In fact, the majority of 3D print processes allow for some sort of optimization of internal density when it comes to producing a 3D printed part.With stereolithography, producing hollowed out 3D prints ensures a high level of accuracy with minimal material use and efficient speeds. Polyjet 3D printing technology lets you fill your parts with a much more affordable support material and FDM 3D printing gives users the option to slice your 3D model file at varying degrees of infill to speed things up while using less material to boot.The drawback to most of these cost and time saving measures often comes in the form of diminished strength in the produced part. It does after all make sense that a part that contains 5% internal material and 95% air would be weaker to withstand external forces than a solid part.Well, it seems that a group of researchers from leading European technical universities have been hard at work to optimize internal infill structures for 3D printing with a little inspiration from the natural world.
The research paper (recently published here) was put together by Dr. Jun Wu (now assistant professor at TU Delft, the Netherlands), Dr. Niels Aage (associate professor at TU Denmark), Dr. Rüdiger Westermann (professor at TU Munich, Germany), and Dr. Ole Sigmund (professor at TU Denmark), and goes into great detail on how porous formations found in bone structures are key to the implementation of efficient yet incredibly durable infill structures.The paper suggests a way to produce optimal lightweight parts, by mimicking the compact cortical bone forming outer shell and spongy trabecular bone occupying its interior as is seen in the cross section of a human femur below next to its 3D printed counter-part.Their algorithm optimizes the distribution of material in a way similar to how bone reacts to external body forces. As a result, the structures mimic those found in nature. The optimized pattern varies in the shape’s interior. This is in contrast to the many repeating patterns used in 3D print slicing software.From a technical perspectve, the researcher’s algorithm is built upon Topology Optimization, an engineering approach for designing lightweight structures. Topology Optimization has been widely used in aviation and automotive industries where the weight of parts and equipment plays a significant role. It puts material where it is needed, in order to maximize the stiffness of structures, under the prescribed external forces.However, as the researchers results show, the structures suggested by topology optimization are sensitive to material damage and to variations in the external forces. After all, efficiency and robustness are a pair of contradicting factors.By enforcing the micro-structures to spread across the shape interior, the new so-called bone-infill achieves a good balance between efficiency and robustness. Not surprisingly, the optimized structures from both the standard and the bone-like optimization algorithms outperform the repeating pattern in current slicing software. This is explained in a test on designing the infill for a 2D femur-shaped object. On the left, we have the standard topology optimized structure. In the middle is the bone-infill, while the repeating pattern is shown on the right. The compliance value (c), that is the inverse of stiffness, shall be as small as possible.While their work remains in the research phase for now, they have tested out their method by fabricating structures using SLS and FDM printers.So, will you be able to create bone-like internal density in your 3D prints with the next update of your favorite slicing software? Jun Wu suggests that for now the algorithm is most likely to be integrated in the design chain (CAD/CAE software) instead of slicing software. This should be a big relief for designers around the world that are tasked with producing parts as cost-effective and efficiently as possible.Going forward, Jun notes that the (research) software package for their infill optimization formulas will be released soon, and that a test version is already available upon request. At the end of the day, as Jun tells, “we welcome good collaborators on bringing the research results to end users.”