Alexandra Ion, Patrick Baudisch
3D printers are celebrated for their ability to arrange matter freely in space. They convert digital models to physical objects, thereby making fabrication available for the masses. While many early adopters and researchers focused on using them to design the outer shape of objects, 3D printers also enable designing objects' internal structure.
The ability to design and fabricate internal structures is powerful, as it enables users to engineer new types of materials, with properties that are not found in nature. Such materials are known as mechanical metamaterials.
Metamaterials have thousands of degrees of freedom, allowing them to cover a large design space. Since metamaterials are solely defined by their structure (i.e., their geometry), they can become a tool for HCI. As such, we can enable novice users to participate in and drive this material revolution.
In this article, we outline traditional metamaterials and their programmable properties, discussing how metamaterials can be actuated for interactivity. We then lay out our vision for future metamaterials. We believe that metamaterials will emerge from being materials to being complete devices. We envision cell-based metamaterial devices that sense, process, and output information—implemented without electronics, purely in the material structure itself.
Traditional Metamaterials Enable Programmable Properties
Mechanical metamaterials are artificial structures with unusual properties that originate in their geometry, rather than in the specific material from which they are made .
Researchers in physics and related fields have discovered cell designs that enable materials to change their volume, absorb shocks, propagate waves, or enable localized elasticity. These extreme properties result from simple building blocks: The material consists of cells containing parts specifically designed to deform in certain ways. Figure 1 illustrates the structure of these cells, which typically include bending or rotating parts, buckling beams, or constrained beams; it also shows how the macroscopic material properties are affected.
The capabilities of metamaterials can be seen in a powerful example: the "unfeelability cloak" . Researchers designed the material's geometry such that an object hidden within the block of material cannot be felt; touching the material that hides an object feels the same as a plain block of material. They achieve this by computing the cell geometry (e.g., beam thicknesses) around the hidden object to deform differently in order to feel the same to the outside.
Making Metamaterials Interactive
Since metamaterials originate in engineering disciplines, most research is concerned with testing the mechanical response of the structure. Therefore, these materials are mainly actuated by a mechanical force.
However, we think HCI researchers can push toward interactive metamaterials by drawing inspiration from areas within HCI, such as shape-changing interfaces or self-assembly, to create metamaterials that can be controlled interactively. Possible forms of actuation include embedding mechanisms actuated by motors, pneumatics, shape-memory polymers, electromagnets, or thermally responsive, light-absorbing, or hydrophilic materials.
Metamaterials are designed to deform in specific locations—the cells' flexures; therefore, that is what needs to be actuated. For inspiration, in Figure 2 we illustrate one of the few examples of actuated metamaterials. The material consists of 3D shearing cells and is pneumatically actuated . Inflating one of the integrated air pockets at an edge flattens the incident faces, which causes the cell to shear. By selectively inflating combinations of air pockets, users can control the overall tilting direction(s) of the material.
|Figure 2. One of the few examples of actuated metamaterials. Inflating the air pockets (blue) selectively moves the material into different configurations, e.g., to shrink uniformly or tilt.|
We think that interactive metamaterials will be a trend in the near future, as interest from the HCI community is growing , but so far there is still much to explore.
Future Metamaterials are the Machine
Traditionally, metamaterials were understood as materials—we think of them as devices.
We argue that viewing metamaterials as devices allows us to push the boundaries of metamaterials further. In our research, we propose unifying material and device to develop metamaterial devices. Such metamaterial devices can receive input and process the information to produce output (Figure 3). To instantiate these new types of devices, we investigated three key elements (Figure 4):
|Figure 3. We don't think of metamaterials as materials—but rather as devices. We envision that future metamaterials can sense, process that input, and output information or reconfigure themselves.|
- Materials that process analog inputs by implementing mechanisms based on their microstructure 
- Materials that process digital signals by embedding mechanical computation into the object's microstructure 
- Interactive metamaterial objects that output information to the user by changing their outside textures .
The design of such intricate microstructures, which enable the functionality of metamaterial devices, is challenging. The complexity of the design arises from the fact that not only is a suitable cell geometry necessary, but also that cells need to play together in a well-defined way. Figure 5 illustrates how cell constraints interact, which makes their design difficult. To support users in designing such microstructures, we create tools that optimize the cell structures and autogenerate cell assemblies from high-level user input .
|Figure 5. Metamaterials are challenging to design; thus they require computational design tools. In this example (a), setting one cell rigid (b) prevents seven cells from shearing, changing the output drastically (c).|
Because metamaterial devices are defined by their geometry, they employ several benefits:
- No assembly. The key benefit of our approach is that the resulting devices can be 3D-printed in one piece and thus do not require assembly. This means that they can be 3D-printed as a part of a larger structure, such as a full door, including latch mechanism and lock.
- Sustainability. Metamaterial devices are printed from a single material. This makes them easy to recycle, in contrast to composite materials. Moreover, metamaterials tend to save material due to their inherently porous structure.
- Printability. Although metamaterials are complex structures, they can be designed to be easy to print. The structure can be optimized to serve as a support structure during manufacturing, as demonstrated recently .
Plenty of Room at the Bottom... of Materials
While traditional metamaterials can save material, create better insulators, redirect light, and do many other things, we believe that extending the functionality of materials by letting them sense and process input, and perform mechanical functions will push the boundaries of metamaterials even further.
A future metamaterial might not only be shock-absorbing but also regulate its own rate of shock-absorption by sensing input, computing the rate, and mechanically actuating itself to that setting—all performed without electronics, but rather within the material structure itself. We believe that solving the following two challenges is necessary for realizing such intelligent materials:
Breaking homogeneity. We need to push toward heterogenous metamaterials that combine topologically distinct cells. As we demonstrated in our previous work, mechanical, digital, and output cells are distinct, which enables their unique behavior. We believe that composing metamaterials from distinct cell types will enlarge the design space significantly.
While many cell designs have been investigated by the research community, they were not combined. These cells were mostly analyzed as repetitive patterns of a single cell type, or recently with varying parameters across the entire grid. Due to the topological differences, such heterogeneous metamaterials are challenging to realize.
Computational design tools for exploration. Some researchers already investigated optimization approaches that return a metamaterial structure for user-defined objectives, such as following a specific deformation path or load conditions. However, we argue that due to the underexplored design space of metamaterials, we don't know what objectives are even possible. Therefore, we are convinced that we need to provide users with efficient tools for the exploration of new cell types and their combinations, since exploration is an important tool in the invention process.
Our previous work has already received attention outside of computer science; for example, it has inspired materials for the thermal control of spacecraft  and volume-filling compliant mechanisms . We see many potential application areas: self-regulating medical devices without hazardous materials, efficient energy-harvesting materials that react to changes in the environment, miniaturized soft robots, and weather-aware facades in architecture.
Understanding the design space of metamaterials, which spans many disciplines, is important to foster future developments. Furthermore, integrating concepts from self-assembly, shape-changing interfaces, programmable matter, and soft robotics will also push metamaterials further. We believe that the possibilities of heterogenous metamaterial devices will be extensive, which emphasizes the importance of a comprehensive framework that can be shared with and built upon by fellow researchers.
We thank our colleagues Ludwig Wall, Pedro Lopes, David Lindlbauer, Philipp Herholz, Marc Alexa, and Robert Kovacs for insightful collaborations.
1. Bertoldi, K., Vitelli, V., Christensen, J., and van Hecke, M. Flexible mechanical metamaterials. Nature Reviews Materials 2, 11 (2017), 17066.
2. Bückmann, T., Thiel, M., Kadic, M., Schittny, R., and Wegener, M. An elasto-mechanical unfeelability cloak made of pentamode metamaterials. Nature Communications 5, 4130 (2014). DOI:10.1038/ncomms5130; https://www.nature.com/articles/ncomms5130
3. Overvelde, J.T.B., de Jong, T.A., Shevchenko, Y., Becerra, S.A., Whitesides, G.M., Weaver, J., Hoberman, C., and Bertoldi, K. A three-dimensional actuated origami-inspired transformable metamaterial with multiple degrees of freedom. Nature Communications 7 (2016), 10929.
4. Qamar, I., Groh, R., Holman, D., and Roudaut, A. Bridging the gap between material science and human-computer interaction. Interactions 26, 5 (Sept.-Oct. 2019), 64–69; https://doi.org/10.1145/3344943
5. Ion, A., Frohnhofen, J., Wall, L., Kovacs, R., Alistar, M., Lindsay, J., Lopes, P., Chen, H.-T., and Baudisch, P. Metamaterial mechanisms. Proc. of the Annual Symposium on User Interface Software and Technology. ACM, New York, 2016.
6. Ion, A., Wall, L., Kovacs, R., and Baudisch, P. Digital mechanical metamaterials. Proc. of the SIGCHI Conference on Human Factors in Computing Systems. ACM, New York, 2017.
7. Ion, A., Kovacs, R., Schneider, O., Lopes, P., and Baudisch, P. Metamaterial textures. Proc. of the SIGCHI Conference on Human Factors in Computing Systems. ACM, New York, 2018.
8. Ion, A., Lindlbauer, D., Herholz, P., Alexa, M., and Baudisch, P. Understanding metamaterial mechanisms. Proc. of the SIGCHI Conference on Human Factors in Computing Systems. ACM, New York, 2019.
9. Martínez, J, Hornus, S., Song, H, and Lefebvre, S. Polyhedral Voronoi diagrams for additive manufacturing. ACM Trans. on Graphics 37, 4 (2018).
10. Phoenix, A.A. and Wilson, E. Variable thermal Conductance metamaterials for passive or active thermal management. Proc. of the ASME Conference on Smart Materials, Adaptive Structures and Intelligent Systems. ASME, 2017.
11. Shaw, L.A., Sun, F., Portela, C.M., Barranco, R.I., Greer, J.R., and Hopkins, J.B. Computationally efficient design of directionally compliant metamaterials. Nature Communications 10, 291 (2019).
Alexandra Ion is a postdoctoral researcher in the Interactive Geometry Lab at ETH Zurich. She received her Ph.D. from the Hasso Plattner Institute, where she worked on metamaterial devices. [email protected]
Patrick Baudisch is a professor of computer science and chair of the HCI Lab at the Hasso Plattner Institute. His recent research focuses on making natural interaction physical, in particular by means of haptics and interactive fabrication. He is a member of the CHI Academy and is an ACM Distinguished Scientist. [email protected]
Sidebar: Getting Started with Metamaterials
You can print your first metamaterial by downloading our metamaterial pliers from thingiverse: http://www.thingiverse.com/HassoPlattnerInstitute_HCI/collections/metamaterial-mechanisms
We printed them on an Ultimaker 2+.
For metamaterials with many cells, you will want to choose a soft material, such as TPU (e.g., Ninjaflex). For stiffer objects, we recommend Nylon (e.g., Taulmann Alloy 910), as it is stiff yet not brittle. Generally, your material should have high "yield strength" and "elongation at break." These can be found in material properties specification sheets.
How to Prototype
If you want to explore your own structures, you can simulate them using common FEA packages (e.g., ABAQUS). You can also use our specialized 3D editor to build your first materials: https://github.com/jfrohnhofen/metamaterial-mechanisms.
For physical prototyping, you can simply laser cut your structures from 3mm rubber foam. This is faster than 3D printing and will give you a good intuition for your new material. To prototype 3D structures, you can use linkage kits, such as GeoMag, or build your own custom linkage kit (e.g., https://www.thingiverse.com/thing:2849789) that fits your metamaterial.
For a Deep Dive into the Topic
Besides the overview in Figure 1 and Bertoldi et al. , you can read the Handbook of Compliant Mechanisms by Larry Howell, Spencer Magleby, and Brian Olsen (John Wiley and Sons, 2013).
Since metamaterials rely on compliance, this book will give you a good background and even features a library of compliant mechanisms that can serve as an inspiration for your metamaterial structures.
Copyright held by authors. Publication rights licensed to ACM.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2020 ACM, Inc.
No Comments Found