KAIST Researchers Enhance 3D Printing Strength and Economy Using Light, AI, and Smart Materials

Photocurable 3D printing, widely used in applications ranging from dental prosthetics to industrial prototyping, is known for its speed and precision but has long suffered from a major drawback—fragility under impact or stress. A research team led by Professor Miso Kim from the Department of Mechanical Engineering at KAIST has developed a breakthrough technology that overcomes this limitation, making 3D-printed structures stronger and more cost-effective through the integration of light control and artificial intelligence. The new method is based on digital light processing (DLP) 3D printing, which uses light to cure liquid resin into solid objects. While this technique enables rapid and accurate fabrication of complex shapes, traditional photocurable resins produce parts that are prone to cracking or breaking under mechanical stress. In contrast, conventional manufacturing methods like injection molding offer superior durability but require expensive and time-consuming mold development. To address this, Professor Kim’s team introduced a dual innovation. First, they engineered a novel polyurethane acrylate (PUA) resin featuring dynamic bonds. These bonds allow the material to absorb shocks and vibrations more effectively, significantly enhancing toughness and resilience compared to standard resins. Second, the team implemented grayscale DLP technology, which varies light intensity during printing to control the degree of curing across different regions of a structure. This enables the creation of parts with tailored mechanical strength in specific areas—similar to how bones and cartilage in the human body serve different structural roles. By adjusting light exposure, the same resin can produce zones of high and low stiffness within a single printed object. An AI-driven machine learning algorithm further optimizes the process by analyzing the object’s geometry and expected load conditions to determine the ideal strength distribution. This intelligent design approach seamlessly integrates material science with structural engineering, eliminating the need for trial-and-error and accelerating product development. The economic benefits are substantial. Previously, achieving varied material properties required multi-material printing systems—complex, costly, and difficult to manage. The new method achieves the same results using a single resin and a single printing process, drastically reducing equipment needs, material handling, and production costs. It also cuts down on R&D time and design iterations. Professor Kim emphasized the broad potential of the technology: “This approach expands the possibilities in both material properties and structural design. We expect to see more durable and personalized medical implants that are not only stronger but also more comfortable for patients. In industries like aerospace and robotics, it enables the production of lightweight yet robust components.” The study, published in Advanced Materials, marks a significant leap forward in making 3D printing not just faster and more flexible, but also stronger, smarter, and more economical.