At 23:00 Beijing time on September 21, 2022, Professor Ni Xiaoyue of Duke University, Professor John A. Rogers of Northwest University, Professor Huang Yonggang, and Wang Heling of Tsinghua University jointly published a new study entitled “A dynamically reprogrammable surface with self-evolving shape morphing” in the journal Nature.
This research result constructs a flexible artificial surface that can be continuously and rapidly deformed under the drive of electromagnetic force, establishes an inverse problem solving strategy based on mechanical model and three-dimensional imaging feedback control, and achieves the target shape of the flexible surface to accurately reproduce the real-time change for the first time. The corresponding authors are Wang Heling, Huang Yonggang, John A. Rogers, and Ni Xiaoyue; The first authors are Bai Yun and Wang Heling.
Patterning technologies such as lithography and printing can reproduce complex and fine target shapes in physical space, making it have superior performance. Compared with the static structures that these technologies can achieve, the process of continuous change of shape over time is more widely present in nature and biology, such as ripples in water, delicate forms of insects in flight, and complex time-varying shapes are the key to the realization of many biological functions. Man-made structures cannot yet replicate the shape of the target over time, so in application scenarios where time scales are critical, their functions and performance are difficult to compare with living organisms. By introducing drivenable components, artificial flexible substances and structures can change shape under the action of external incentives, but one of the challenges in reproducing complex time-varying shapes is to establish an inverse problem solving strategy that enables structures to be deformed into a variety of different target shapes.
Recently, the team of Professor Ni Xiaoyue of Duke University, Professor John A. Rogers of Northwestern University, Professor Huang Yonggang, and Wang Heling of Tsinghua University constructed a flexible surface composed of a network of metal wires, which generates electromagnetic forces under the action of static magnetic fields and distributed reprogrammable currents, driving the rapid and continuous complex deformation of flexible surfaces. Combining mechanical model, real-time 3D imaging, digital feedback control, and optimization algorithm, an inverse problem solving strategy is established to guide the flexible surface deformation to accurately fork the continuous change of the three-dimensional target shape on the time series, and to offset the influence of internal defects and external interference factors on the deformation.
Reprogrammable electromagnetic drive:This flexible surface consists of a network of serpentine strips in which metals encased in polyimide are used to conduct currents. There are a large number of independently applied voltage ports around the flexible surface to control the current distribution inside the structure, generate programmable electromagnetic forces in the static magnetic field, and achieve structural deformation. The port voltage quickly changes the current distribution under the control of a digital signal, and the drive structure deforms into a large number of different shapes (Figure 1).
Figure 1: Complex deformation of a distributed electromagnetic force drive structure.
Model-based anti-problem solving strategy:Through clever structural design, a flexible structure with unconventional mechanical behavior is constructed, and the approximate linear relationship between deformation and port voltage is realized. The linearization theoretical model is combined with the optimization algorithm to establish an inverse problem solving strategy to achieve flexible surface deformation into a series of abstract designs or naturally existing time-varying target shapes, including the growth of “bubbles”, movement, splitting, vibration process, and the process of water droplets falling on the solid surface and collapsing after dispersion.
Experiment-based inverse problem solving strategy:When there are complex physical mechanisms such as nonlinearity, structural defects, environmental disturbances, etc., the above model-based inverse problem solving strategy is difficult to work. The experiment-based strategy obtains the current shape of the structure in real time through three-dimensional imaging, and feedback the shape to the optimization algorithm in real time through digital circuit control, which optimizes and adjusts the port voltage to form a closed-loop control until the error between the structure shape and the target shape reaches a minimum, realizing the self-starring of the flexible surface to the target shape. This method does not require a physical model, so it is suitable for nonlinear systems and can overcome the effects of the structure’s own defects and environmental disturbances.
Quasi-real-time shape learning:This ability to auto-star flexible surfaces allows it to learn the shape changes of actual objects in quasi-real time, such as the shape of the palm of the hand when the human hand makes various movements.
Multifunctional shape evolution:In addition to evolving into the target shape, the flexible surface is also capable of self-starring to achieve specific functions. Reflective parts are introduced on a flexible surface, and two laser beams are irradiated on the flexible surface and reflected to a receiving plane. Using an inverse problem-solving strategy similar to the one described above, the flexible surface can auto-denormalize, enabling two laser beams to converge on the receiving plane (optical function) while maintaining the height of its center to a specified target value (shape function). (Source: Science Network)
Related Paper Information:https://doi.org/10.1038/s41586-022-05061-w