A coating that can hide objects from view or an implant that acts exactly like bone tissue; these extraordinary objects are already made of “metamaterials”. Researchers at TU Delft have now developed an AI tool that can detect such extraordinary materials, while also making them production-ready and durable. This allows you to create devices with unprecedented functionality. They published their findings Advanced Materials.

Properties of ordinary materials, such as hardness and flexibility, are determined by the molecular composition of the material, but the properties of metamaterials are determined by the geometry of the structure from which they are formed. Researchers digitally design these structures and then 3D print them. The resulting metamaterials can exhibit unnatural and extreme properties. For example, researchers have created metamaterials that are solid but behave like liquids.

“Traditionally, designers use materials at hand to develop a new device or machine. The problem is that the range of material properties available is limited. Some properties we want do not exist in nature. Our approach is: Tell us what properties you want and we will design a material with those properties . Then you will not get a material, but something between the structure and the material, a metamaterial,” says the professor. Amir Zadpour, Department of Biomechanical Engineering.

A new AI tool detects realistic “metamaterials” with unusual properties. Credit: TU Delft

reverse design

This material discovery process requires solving the so-called “inverse problem”: the problem of finding a geometry that produces the desired properties. Inverse problems are notoriously difficult to solve, and that’s where artificial intelligence comes into play. TU Delft researchers have developed deep learning models that solve these inverse problems.

“Even if inverse problems have been solved in the past, they have been limited by the assumption that fine-scale geometry can be created from an infinite number of building blocks. The problem with this assumption is that metamaterials are often made by 3D printing, and real 3D printers have a limited resolution, which limits the number of building blocks that can fit into a given device.” ” says first author Dr. Helda Pakhlawani.

Artificial intelligence models developed by TU Delft researchers break new ground by overcoming such simple assumptions. “We can now simply ask: How many building blocks does your manufacturing technology allow you to pack into your device? “The model then finds the geometry that gives you the properties you want based on the number of building blocks you can make.”

Revealing full potential

An important practical issue that previous studies overlooked was the durability of metamaterials. Most existing structures deteriorate with repeated use. This is because current approaches to the design of metamaterials do not take durability into account.

“Until now it has only been about what properties can be achieved. Our research considers durability and selects the strongest designs from a large number of design candidates. This makes our designs truly practical rather than just theoretical adventures,” says Zadpour.

The paper’s corresponding author, Associate Professor Mohammad J. Mirzaali, says the possibilities of metamaterials seem endless, but the full potential is far from being realized. This is because the search for optimal metamaterial design is still largely intuition-based, trial-and-error, and therefore time-consuming. The use of the reverse engineering process, where desired properties are the starting point of design, is still very rare in the field of metamaterials.

“But we believe that the step we have taken is revolutionary in the field of metamaterials. This could lead to a variety of new applications.” It could be used in orthopedic implants, surgical instruments, soft robots, adaptive mirrors and exosuits.