3D printing technology, also known as additive manufacturing, has brought a true revolution in the way we design and create objects, from simple prototypes to complex components for the aerospace and medical industries. Yet, despite incredible progress, there is a significant gap between what a computer designs and what a 3D printer can actually produce. This discrepancy between the digital blueprint and the physical product represents one of the biggest obstacles to the wider application of this technology in critical sectors where precision and reliability are of crucial importance. Designs created by advanced algorithms often exceed the actual capabilities of manufacturing devices, resulting in parts whose actual performance deviates from expectations. Scientists at the Massachusetts Institute of Technology (MIT) have now developed an innovative method that could bridge this gap, enabling design software to take into account the physical limitations of 3D printers in advance.

The problem at the interface of the digital and physical worlds

Over the past decade, the fusion of new design and manufacturing technologies has reshaped industries such as aerospace, automotive, and biomedical engineering. In these fields, materials must meet extremely strict criteria, such as specific strength-to-weight ratios and other performance thresholds. 3D printing has emerged as a technology that allows for the creation of parts with previously unimaginable internal structures, opening the door to lighter, stronger, and more functional components. One of the most advanced computational design techniques used for this purpose is topology optimization.

Topology optimization is an algorithmic process that, within a given space, optimally distributes material to achieve desired characteristics – for example, maximum stiffness with minimum mass. The results are often organic, seemingly random structures that resemble natural forms like bones or cellular networks, and which are significantly more efficient than those designed by traditional methods. It is used to design materials with optimized stiffness, strength, maximum energy absorption, fluid permeability, and many other properties. However, it is precisely this complexity and fineness of detail that topology optimization generates that poses a challenge for 3D printers. The problem lies in the physical limitations of the printing process itself. One of the key limitations is the size of the print head, i.e., the nozzle that extrudes the material. If the algorithm, for example, specifies a layer thickness of 0.5 millimeters, and the printer's nozzle can physically extrude only a layer of a minimum of 1 millimeter, the final product will be deformed and imprecise. This difference between the digital instruction and the physical realization leads to unexpected variations in mass and density, which directly affects the mechanical properties of the part.

Anisotropy: The hidden weakness of 3D printing

Another fundamental problem arises from the very way 3D printers build objects – layer by layer. The print head moves across the work surface, extruding a thin filament of molten material. Each new layer is applied to the previous one, which has already begun to cool in the meantime. Because of this, the bond between individual layers is not as perfect as the material within the layer itself. This phenomenon, known as anisotropy, means that the mechanical properties of the object depend on the direction of the applied force. The part will be significantly stronger in the direction in which the material was printed (along the print lines), but noticeably weaker perpendicular to the layers, at the points where the layers joined. This creates potential points of weakness where delamination or fracture can occur under load, even if the load is considerably less than what the material should theoretically withstand. It was this mismatch between the expected and actual properties of the material that was the focus of the research team from MIT.

“If these constraints are not taken into account, printers can deposit either too much or too little material, so your part becomes heavier or lighter than intended. This can also lead to a significant overestimation or underestimation of the material's performance,” explains Josephine Carstensen, an associate professor in the Department of Civil and Environmental Engineering and the lead researcher. “With our technique, you know what you are getting in terms of performance because the numerical model and the experimental results match very well.” The research is described in detail in the scientific journal Materials and Design, in a paper authored by Carstensen and doctoral student Hajin Kim-Tackowiak.

An innovative approach: Incorporating imperfections into the design itself

Instead of trying to modify the printer's hardware, the researchers opted for a more elegant solution: "teaching" the design software about the imperfections of the manufacturing process. “We realized that we know these constraints from the very beginning, and the scientific field has gotten better at quantifying them. So we figured we could design with them in mind from the start,” says Kim-Tackowiak. In previous work, Professor Carstensen developed an algorithm that incorporated information about the printer's nozzle size into the design process of beam structures. In this research, they upgraded this approach to include the direction of the print head's movement and the consequent impact of weak bonding between layers. They also adapted the method to work with more complex, porous structures that can have highly elastic properties.

Their approach allows users to add variables to the design algorithms that precisely account for the center of the material filament being extruded from the nozzle and the exact location of the weaker bonds between layers. Crucially, the approach also automatically dictates the optimal path the print head should follow during production to minimize negative effects. In this way, the software does not create an idealized design, but an optimized blueprint that is already adapted to the real capabilities and flaws of a specific 3D printer.

Confirmation through experiments and real results

To test their technique, the researchers used it to create a series of repeating 2D designs with different sizes of hollow pores, i.e., different densities. They then compared these samples with materials of the same densities but made using traditional topology optimization methods that do not take printer limitations into account. The test results were unequivocal. Materials designed with traditional methods deviated significantly from their predicted mechanical performance, especially at material densities below 70%. On the other hand, materials designed with the new MIT team's technique showed performance that was much closer to that predicted in the computational model. The researchers also found that conventional designs consistently deposited excess material during manufacturing, making the parts heavier and less efficient. Overall, the MIT scientists' approach resulted in parts with more reliable and predictable performance at most of the tested densities.

“One of the challenges with topology optimization has been that it takes a lot of expertise to get good results, so that when you transfer the design from the computer, the materials behave as you intended,” Carstensen points out. “We are trying to make it easier to obtain these high-fidelity products.” This advancement could democratize the use of advanced design techniques, reducing the reliance on experienced 3D printing experts who have so far had to manually intervene and adjust designs to compensate for machine limitations.

The future of design and manufacturing

The researchers believe this is the first time a design technique has simultaneously considered both the print head size and the problem of weak inter-layer bonding. “When you are designing something, you should use as much context as possible,” emphasizes Kim-Tackowiak. “It was satisfying to see that bringing more context into the design process makes your final materials more accurate. This means there are fewer surprises. Especially when we are investing more and more computational resources into these designs, it’s nice to see that we can connect what comes out of the computer with what comes out of the manufacturing process.”

In future work, the team hopes to improve their method for higher-density materials and for different types of materials such as cement and ceramics, which bring their own specific printing challenges. Nevertheless, they point out that their approach already offers a significant improvement over existing techniques. The scientists say this work opens the way for designing with a greater number of materials. “We would like this to enable the use of materials that people have neglected because printing with them led to problems,” concludes Kim-Tackowiak. “Now we can leverage these properties or work with these ‘quirks’ instead of simply not using all the material options available to us.” This innovation not only improves the reliability of 3D printing but also promises to unlock the full potential of additive manufacturing, enabling the creation of a new generation of materials and products with performance that was previously only achievable in theory.