Thank you for visiting Nature. The browser version you are using has limited support for CSS. For the best experience, we recommend that you use a newer version of the browser (or turn off the compatibility mode in Internet Explorer). At the same time, to ensure continued support, we will display sites without styles and JavaScript.
Nature Volume 599, pages 229–233 (2021) Cite this article
Inspired by living organisms, soft robots are developed from inherently compliant materials, capable of simulating the continuous movement of animals and plants1. In soft robots, standard hinges and bolts are replaced by elastomers assembled into actuators that are programmed to change shape after stimuli are applied, such as inflation 2, 3, 4, and 5. Deformation information is usually directly embedded in the shape of these actuators, and their assembly benefits from the latest developments in rapid prototyping technology6,7,8,9,10,11. However, these manufacturing processes have limitations in scalability, design flexibility, and robustness. Here, we show a new all-in-one method for the manufacturing and programming of soft machines. Our method does not rely on the assembly of individual parts, but uses the interfacial flow in elastomers, which are gradually solidified to produce a solid pneumatic actuator. The shape can be easily customized to suit everything from artificial muscles to grippers. application. We rationalized the fluid mechanics in the assembly of the actuator and modeled its subsequent deformation. We use this quantitative knowledge to program these soft machines and generate complex functions, such as sequential movements obtained from monotonous stimuli. We expect the flexibility, robustness, and predictability of our method to accelerate the diffusion of soft robots by assembling complex actuators (such as long, tortuous, or vascular structures), which is derived from geometric and material nonlinearities. New features pave the way.
Get full access to the journal for 1 year
All prices are net prices. VAT will be added at checkout later. Tax calculation will be done at checkout.
Get time-limited or full article access on ReadCube.
All prices are net prices.
The data used in this study can be obtained from the corresponding authors upon reasonable request. This article provides source data.
The corresponding author can provide the code supporting the results of this research upon reasonable request.
Whitesides, GM Soft robot. Angie. Chemistry International. Ed. 57, 4258–4273 (2018).
Polygerinos, P., Wang, Z., Galloway, KC, Wood, RJ & Walsh, CJ Soft robotic gloves for joint assistance and home rehabilitation. grab. automated system. 73, 135–143 (2015).
Shepherd, RF, etc. Multi-gait soft robot. Process National Academy of Sciences. science. United States 108, 20400-20403 (2011).
ADS CAS PubMed PubMed Central Google Scholar
Martinez, RV, etc. Robot tentacles with three-dimensional mobility based on flexible elastomers. Senior alma mater. 25, 205–212 (2013).
Roche, ET, etc. A bionic soft drive material. Senior alma mater. 26, 1200–1206 (2014).
Mosadegh, B. et al. Pneumatic network of soft robots for quick start. Advanced Features. alma mater. 24, 2163–2170 (2014).
Yap, HK, Ng, HY & Yeow, C.-H. High-strength soft-printable pneumatic device for soft robotics applications. Soft robots. 3, 144–158 (2016).
Wehner, M. etc. Integrated design and manufacturing strategy for completely soft autonomous robots. Nature 536, 451–455 (2016).
ADS CAS PubMed Google Scholar
Zhao, H., Li, Y., Elsamadisi, A. & Shepherd, R. Scalable manufacturing of high-power wearable soft actuators. Extreme Mecha. Wright. 3, 89–104 (2015).
Paek, J., Cho, I. & Kim, J. Micro-robot tentacles with spiral bending capability based on shape-engineered elastic microtubes. science. Representative 5, 10768 (2015).
ADS CAS PubMed PubMed Central Google Scholar
Inflatable toys with programmable shapes inspired by Jin, L., Forte, AE, Deng, B., Rafsanjani, A. & Bertoldi, K. Kirigami. Senior alma mater. 32, 2001863 (2020).
Ilievski, F., Mazzeo, AD, Shepherd, RF, Chen, X. & Whitesides, GM Soft Robotics for Chemists. Angie. Chemistry International. Ed. 50, 1890–1895 (2011).
Marchese, AD, Onal, CD & Rus, D. An autonomous soft robotic fish that can use fluid elastomer actuators for escape maneuvers. Soft robots. 1, 75–87 (2014).
PubMed PubMed Central Google Scholar
Yang, D. etc. Flexion pneumatic linear actuator inspired by muscles. Senior alma mater. technology. 1. 1600055 (2016).
Overvelde, JTB, Kloek, T., D'haen, JJA & Bertoldi, K. Amplify the response of soft actuators by taking advantage of rapid pass instability. Process National Academy of Sciences. science. United States 112, 10863–10868 (2015).
ADS CAS PubMed PubMed Central Google Scholar
Hawkes, EW, Blumenschein, LH, Greer, JD and Okamura, AM A soft robot that grows and navigates in the environment. science. robot. 2. aan3028 (2017).
Majidi, C. Soft Robots: A Perspective-Current Trends and Future Prospects. Soft robots. 1, 5-11 (2014).
Gladman, AS, Matsumoto, EA, Nuzzo, RG, Mahadevan, L. & Lewis, JA Biomimetic 4D printing. Nat. alma mater. 15, 413–418 (2016).
Carnick, M. et al. Artificial muscle based on strain programmable fiber. Science 365, 145–150 (2019).
ADS CAS PubMed PubMed Central Google Scholar
William Boley, JW et al. Deformed structural lattice through multi-material 4D printing. Process National Academy of Sciences. science. United States 116, 20856-20862 (2019).
PubMed PubMed Central Google Scholar
Guseinov, R., McMahan, C., Pérez, J., Darario, C. & Bickel, B. The programming time deformation of the self-driving shell. Nat. Community. 11, 237 (2020).
ADS CAS PubMed PubMed Central Google Scholar
Kim, Y., Yuk, H., Zhao, R., Chester, SA & Zhao, X. Print ferromagnetic domains for unconstrained fast switching soft materials. Nature 558, 274–279 (2018).
ADS CAS PubMed Google Scholar
Hu, W., Lum, GZ, Mastrangeli, M. & Sitti, M. Small soft robot with multi-modal motion. Nature 554, 81–85 (2018).
ADS CAS PubMed Google Scholar
Acombe, E. et al. A hydraulically amplified self-healing electrostatic actuator with muscle-like performance. Science 359, 61–65 (2018).
ADS CAS PubMed Google Scholar
Gorison, B. et al. Resilient inflatable actuator for soft robotics applications. Senior alma mater. 29, 1604977 (2017).
Polygerinos, P. etc. Soft Robots: An overview of fluid-driven essential soft devices; manufacturing, sensing, control, and human-computer interaction applications. Senior English. alma mater. 19, 1700016 (2017).
Siéfert, E., Reyssat, E., Bico, J. & Roman, B. Bionic aerodynamic deformation elastomer. Nat. alma mater. 18, 24–28 (2019).
Vasios, N., Gross, AJ, Soifer, S., Overvelde, JTB & Bertoldi, K. Use viscous flow to simplify the driving of fluid soft robots. Soft robots. 7, 1-9 (2020).
Jambon-Puillet, E., Piéchaud, MR & Brun, P.-T. Elastic amplification of Rayleigh-Taylor instability in solidified melts. Process National Academy of Sciences. science. United States 118, e2020701118 (2021).
MathSciNet CAS PubMed PubMed Central Google Scholar
Bretherton, the movement of long bubbles in the FP tube. J. Fluid machinery. 10, 166–188 (1961).
ADS MathSciNet MATH Google Scholar
Aussillous, P. & Quéré, D. The rapid deposition of fluid on the pipe wall. physics. Fluid 12, 2367–2371 (2000).
ADS CAS MATH Google Scholar
Klaseboer, E., Gupta, R. & Manica, R. Extended Bretherton model of long Taylor bubbles with medium capillary numbers. physics. Fluid 26, 032107 (2014).
Takagi, D. & Huppert, HE Flow and instability of films on cylinders and spheres. J. Fluid machinery. 647, 221–238 (2010).
ADS MathSciNet MATH Google Scholar
Lee, A. etc. The elongated elastic shell is manufactured by curved surface coating. Nat. Community. 7, 11155 (2016).
ADS CAS PubMed PubMed Central Google Scholar
Audoly, B. & Pomeau, Y. Elasticity and geometry: the nonlinear response from curls to shells (Oxford University Press, 2010).
Gerbode, SJ, Puzey, JR, McCormick, AG & Mahadevan, L. How to coil and over-wind cucumber tendrils. Science 337, 1087–1091 (2012).
ADS CAS PubMed Google Scholar
Miller, JT, Lazarus, A., Audoly, B. & Reis, PM The shape of suspended curly hair. physics. Pastor Wright. 112, 068103 (2014).
ADS CAS PubMed Google Scholar
Yang, HD, Greczek, BT & Asbeck, AT Modeling and analysis of high-displacement pneumatic artificial muscles with integrated sensing functions. front. robot. Artificial Intelligence 5, 136 (2019).
PubMed PubMed Central Google Scholar
Kaufman, JJ et al. Structured spheres produced by fluid instability within fibers. Nature 487, 463–467 (2012).
ADS CAS PubMed Google Scholar
Marthelot, J., Strong, EF, Reis, PM & Brun, P.-T. Design soft materials with interface instability in liquid membranes. Nat. Community. 9, 4477 (2018).
ADS CAS PubMed PubMed Central Google Scholar
Zhao, C. Wait. The shear flow of the nanosheets induces the arrangement to form a layered nanocomposite material. Nature 580, 210–215 (2020).
ADS CAS PubMed Google Scholar
Nerger, BA, Brun, P.-T. & Nelson, CM Marangoni Flow drives the arrangement of fibrous cell-loaded hydrogels. science. Advanced 6. eaaz7748 (2020).
ADS CAS PubMed PubMed Central Google Scholar
Cybulski, O. et al. Concentric liquid reactor for chemical synthesis and separation. Nature 586, 57–63 (2020).
ADS CAS PubMed Google Scholar
This work was supported by the NSF CAREER Award (CBET 2042930) and the Princeton University Materials Research Science and Engineering Center (NSF Grant DMR-1420541), the Princeton Young family and the David T. Wilkinson Innovation Fund.
Department of Chemistry and Biological Engineering, Princeton University, Princeton, New Jersey, USA
Trevor J. Jones, Etienne Jambon-Puillet, Joel Marthelot & P.-T. Buren
Aix-Marseille University, CNRS, IUSTI, Marseille, France
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
You can also search for this author in PubMed Google Scholar
TJJ, JM and P.-TB conceived the project. TJJ, EJ-P. Conducted experiments with P.-TB and analyzed the data. TJJ performed a Kirchhoff stick simulation. EJ-P. A finite element simulation was performed. All authors have written manuscripts.
Corresponds to P.-T. Bulen.
The author declares no competing interests.
Peer review information Nature thanks Detlef Lohse, Orlin Velev and other anonymous reviewers for their contributions to the peer review of this work.
The publisher states that Springer Nature remains neutral on the jurisdiction claims of published maps and agency affiliates.
a, VPS-08, 16 and 32 (strain 1%, frequency {1, 2.5, 2.5} Hz) oscillating shear rheological measurement. The cross represents τc. b, when t <τc, readjust the time-varying viscosity. The dashed line is the fit of equation (5). c, Dogbone uniaxial tensile test data of multiple VPS-32, VPS-16 and VPS-08 samples. The black curve is the fitting of various constitutive models. d, the table shows the rheological and elastic material constants of VPS
a, a schematic diagram of calculating the shape of the liquid bath. b. For various values of \(R/{{\mathscr{l}}}}_{{\rm{c}}}\) and hi/R, the meniscus shape obtained by solving equation (9) contour. c. The contour of the model shape obtained from the matching of the bath solution (Equation (9)) and the film thickness solution (Equation (2)), R = 1.6 mm. The solid line represents the membrane, and the dashed line represents the bath. Color-coded average film thickness hf. d, e, the porosity ϕ = (1 − hi/R)2 (d) and the average film thickness hf (e) as a function of the axial position of the one-meter long actuator (sample 1). Error bars represent the spread of measurement uncertainty. f, the porosity of three different samples. The uncertainty is displayed as a box plot (middle 50%) and a whisker plot (full range) superimposed on the marker. g, the film thickness of three different samples. The upward (respectively downward) triangle represents the maximum thickness \(h(0)\) (the minimum thickness \(h({\psi }_{0})\) ), and the diamond represents the average thickness hf. Uncertainty is displayed as a box plot (middle 50%) and a whisker plot (full range) superimposed on the marker
The deformation of the analog actuator shown in a, b is \(P\approx 23\) kPa (experimental shape, G = 0.36 MPa, \({J}_{m}=14\) ). Inset: Deformed cross-sectional view. The color coded von Mises stress, and the black line shows the edge of the undeformed configuration. b, Curvature κ as a function of inflation pressure P. The circle represents the experimental data (VPS-32, R = 1.57 mm, \(h(0)=52\,{\rm{\mu }}{\rm{m }}\), hi/R = 0.34), real The line is the experimental cross-sectional shape extracted by image analysis for finite element simulation, and the dashed line is the use of \ (h(0)\). Illustration: Comparison between experimental and model shapes. c. When we change hf and R, the curvature κ is a function of the applied pressure P of the shape of our model (see the legend in d). d, the same data as c is rescaled according to equation (17), the dashed line is a power law fit
a, a schematic diagram of pneumatic inflation of the cross section of the bubble casting actuator. b, the bending diagram of the bubble casting actuator. c, Actuator curvature \(\kappa \) as the pressure applied by different length L (i), thickness hf (ii), radius R (iii) and shear modulus G (iv; see the legend in Figure 3a) Function of P, b). d. The re-adjusted curvature \(\kappa R\) is a three-dimensional graph as a function of the re-adjusted pressure \(P/G\) and the re-adjusted film thickness R/hf. The mark shows the experimental data, and the blue surface is our model (Equation (16)), using Ghent's constitutive law \({\psi }_{0}={\rm{\pi }}/4,\, \varphi =0.4225,\,\chi =0.45,\,{J}_{m}=14\)
a. A series of images (i) and calculated elastic curve (ii) for foam casting actuators that are inflated when one end is clamped and the other end is blocked by a wall (see expanded data Figure 7c; scale, 1 centimeter). b. The blocking force F as a function of the inflation pressure P of various actuators (R, G, hf and L have been changed)
a, c, The two pipes are held tightly together by a connector (not shown here) that allows one pipe to rotate relative to the other. Applying rotation at the end of the drainage step, that is, around the gel point, the polymer is still deformable, but no longer flows significantly under the action of gravity, resulting in actuators as shown in b and d. Color is a guide for the eyes; only one polymer is used. The rotating cylinder (shown in a (c respectively)) rotates the membrane, thereby rotating the driving direction under pressure. In this example, a 90° (180°) rotation produces two curvatures of equal magnitude but in orthogonal planes (respectively with opposite signs).
a. An experimental device similar to the Bresserton flow in the manufacturing process. The camera captures the velocity U before the bubble to determine the capillary number Ca. b. Experimental device for expansion and bending experiment. The pressure sensor records the internal and external pressure difference P, while the camera records the shape of the actuator. c. Experimental device for blocking force-pressure experiment. The pressure sensor records the internal and external pressure difference P, while the load cell measures the force Fd, an experimental device for the force-extension experiment. The pressure sensor records the internal and external pressure difference P, while the Instron measures the force F and displacement \({\ell }\).
Supplementary text, supplementary equations and supplementary references
Bubble casting manufacturing method. The tortuous channel in the acrylic mold is filled with solidified VPS melt. Then an elongated bubble is injected through the channel. Since gravity will cause drainage, the VPS film has time to solidify. The resulting soft actuator bends when inflated, demonstrating the ability to hold the Blackberry.
High aspect ratio pneumatic muscles. According to equation (3) in the main text, the comparison between the inflatable actuator and the Kirchhoff rod simulation with the natural curvature κ under pressure P. The human-scale actuator exhibits a coiling behavior when inflated. The prosthesis mimics a human arm with contractile force. Inflate the actuators with diameters of 12.8 mm and 1.0 mm to lift the water bottle and paper clip. A high-aspect-ratio bending actuator wraps itself around the raspberry and lifts it up without any damage.
The sequence of numbers is driven. The programming logic is represented by schematic diagrams. At different waiting times τw,i, the bubbles are injected through the channel numbers (1,2,3,4) in a sequential manner. As the polymer melt crosslinks and diverges at the curing time τc, the evolving viscosity increases as the melt solidifies into an elastic solid. Injecting air into the resulting actuator will cause the order of the actuator numbers to bend.
Deformed into a complex shape. An actuator with a tortuous path is inflated in the water. The actuator is bent from a plane to a curve on the surface of a sphere. The helical actuator is connected to the circular membrane and converts the local curvature of the rod into a global curvature of the membrane. The actuator is connected to the edge of the film to perform the self-folding of the five-sided box. The two branched actuators are connected to both sides of the membrane in the shape of a tail fin. The fin actuator is alternately inflated in a water bath with tracer particles. During the manufacturing process, the tubular mold is rotated 180 degrees at the gel point. The resulting actuator bends in two different directions. During the manufacturing process, a three-segment tubular mold is rotated 180 degrees and 90 degrees at the gel point. The resulting actuator bends in three directions and two planes.
A soft machine acquired in a limited space. A modular soft machine consisting of two actuators is programmed to use a single pressure source to grab and retrieve a ball from a cylindrical container.
Program curvature changes. The linear actuator made of bubble casting is programmed to have three parts with different curvatures (curvature ratio 4:2:1). The coding is done by injecting bubbles into the tube in a stepwise sequence to predict the final film thickness. The actuator expands on the water bath, forming a spiral with appropriate curvature.
Jones, TJ, Jambon-Puillet, E., Marthelot, J. etc. Bubble casting soft robot. Nature 599, 229–233 (2021). https://doi.org/10.1038/s41586-021-04029-6
DOI: https://doi.org/10.1038/s41586-021-04029-6
Anyone you share the following link with can read this content:
Sorry, there is currently no shareable link in this article.
Provided by Springer Nature SharedIt content sharing program
By submitting a comment, you agree to abide by our terms and community guidelines. If you find content that is abusive or does not comply with our terms or guidelines, please mark it as inappropriate.
Nature ISSN 1476-4687 (online) ISSN 0028-0836 (print)