Diffusive kinks turn kirigami into machines

The study addresses whether a synthetic mechanical analog of biological diffusive kinks can be engineered in metamaterials. Kinks—boundaries between distinct configurations—occur across scales, from ferro-electrics and shape-memory alloys to flexible metamaterials. In mechanical systems, kink motion typically relies on inertia or continuous external loading; in overdamped regimes, traveling kinks without sustained loading are generally not observed. In nature, however, stimuli-responsive systems such as Mimosa Pudica display sequential folding that propagates as a slow wave describable by reaction–diffusion dynamics. The authors hypothesize that by combining kirigami architectures (capable of geometry-governed multistability) with viscoelastic materials (enabling strain-rate dependence and relaxation), it is possible to realize non-linear, diffusive kinks in an overdamped mechanical medium. The purpose is to design, model, and experimentally validate viscoelastic kirigami that exhibit sequential snap-back events forming a traveling diffusive kink, and to demonstrate machine-like functions powered by such kinks.

Prior work has shown kinks and transition waves in multistable metamaterials and structures (e.g., beams in series, kirigami, hinged mechanisms, inflatable structures) enabling logic, locomotion, and shape-changing via inertia-driven or load-driven propagation. Reaction–diffusion models are classical in chemistry and biology for morphogenesis and pattern formation; Mimosa Pudica exhibits sequential leaflet folding as a reaction–diffusion-like mechanical wave between stable states. Recent advances in viscoelastic metamaterials reveal strong strain-rate-dependent responses and tunable dynamics, suggesting a pathway to non-inertial wave phenomena. Control of buckling mode in single-material kirigami is often sensitive to imperfections and boundaries; multimaterial designs and strain-rate effects can bias buckling pathways. The present work builds on these insights by leveraging viscoelastic relaxation coupled with geometric instability to produce diffusive kinks in kirigami and to translate them into functional devices.

Design and materials: The kirigami consists of a thick sheet with parallel cut-lines forming periodic unit cells that can buckle in two modes (symmetric and anti-symmetric), corresponding to two textures. To control mode selection, the unit cell is patterned through thickness with two photopolymers (Agilus and TangoPlus, Stratasys) having distinct viscoelastic properties. Geometry was optimized via a hybrid experimental–computational protocol to ensure manufacturability and robustness against inevitable printing imperfections.

Experimental setup: Specimens were fabricated using PolyJet multi-material 3D printing (Objet500 Connex3). A custom test bench with a low-friction Teflon substrate enabled controlled stretch rates from 1 to 5000 mm/s. Kirigami were rapidly stretched to induce anti-symmetric (high-speed) buckling, then unloaded or held to observe viscoelastic relaxation and snap-back into symmetric (low-speed) mode. 3D motion tracking of embedded surface markers (black dots and patterns) quantified unit-cell kinematics; MATLAB and Python codes were used for data analysis and 3D reconstruction. Lateral strain jumps in half-cells were tracked to time snapping events and kink propagation speed.

Reduced-order analytical model: A viscoelastic von Mises truss was constructed to represent switching between anti-symmetric and symmetric modes. The truss comprises a pair of linear springs (K/2) and a Kelvin–Voigt-like branch (spring k in parallel with dashpot c), with geometric nonlinearity and prestrain. Using a Lagrangian formulation with elastic potential and Rayleigh dissipation, the non-dimensional overdamped equation of motion for a unit cell is: ∂U1/∂t = U1 − β U1^3 − 1, where U1 and t are dimensionless displacement and time. The parameter β (a function of stiffnesses, geometry, and prestretch) governs mono-/metastability and snap-back amplitude and delay, while initial condition/imperfection sets the snapping delay. A 1D chain of such trusses coupled by linear springs R yields, in the continuum and overdamped limit, a reaction–diffusion equation: ∂U/∂t = U − β U^3 + ∂^2U/∂X^2, which supports diffusive kinks.

Computational modeling: Nonlinear finite element analysis (Abaqus 2020, standard solver) with visco-hyperelastic material models (first-term Prony series) for Agilus and TangoPlus, and elastic models for rigid parts. Materials assumed incompressible; parameters were calibrated to experimental stress–strain data. 3D C3D8H elements with mesh convergence checks. Boundary conditions included clamped–clamped ends (one end fixed; the other axially driven), with occasional reference-point tie constraints and periodic constraints at unit-cell free boundaries. Geometrical imperfections—motivated by measured PolyJet imperfections—were introduced to nucleate snap-back and traveling kinks. Constant force boundary conditions were used in some simulations to assess kink speed. The models captured fast-rate buckling to anti-symmetric mode, viscoelastic relaxation, imperfection-driven snap-back, and sequential propagation along strips.

Protocols for demonstrations: Additional kirigami plates were designed with unit-cell variations and material patterning to realize a dynamic 2D plus-to-minus texture transformation. Strips were equipped with arm-links, guiders, grippers, and compliant spatial four-bar linkages (HSSR) to demonstrate transport, grasp-and-release, and cyclic paddle motions driven by the diffusive kink. Experimental observations were complemented with simulations of multi-unit assemblies carrying paddles to illustrate sequential rowing-like cycles.

• Rate-programmable buckling: Multimaterial viscoelastic kirigami exhibit mode selection controlled by stretch rate: low rates yield symmetric buckling; high rates yield anti-symmetric buckling.
• Viscoelastic snap-back: After a fast stretch to the anti-symmetric mode, unit cells relax; within certain strain ranges, they undergo viscoelastic snap-back into the symmetric mode. 3D tracking shows initial slow creep followed by rapid snap-back over O(100 s). Snap-back amplitude increases as β decreases, while snap-back delay shortens with larger geometric imperfections.
• Reduced model behavior: The unit-cell model ∂U1/∂t = U1 − β U1^3 − 1 captures metastability and sensitivity to imperfections. In a chain, the continuum reaction–diffusion equation ∂U/∂t = U − β U^3 + ∂^2U/∂X^2 supports traveling diffusive kinks. For β = 0.05, numerical solutions show transition between an up-state and down-state (e.g., roots around U ≈ +3.843 and U ≈ −4.907) with a kink moving at constant speed. Kink velocity increases with β (i.e., with reduced energy difference between states in the model).
• Imperfection-mediated nucleation and propagation: In FE models and experiments, significant PolyJet-induced asymmetries and clamping imperfections are essential to nucleate the kink. The boundary rotation from a snapping cell transfers a geometric imperfection to its neighbor, triggering sequential snap-back and a propagating front.
• Constant-speed, overdamped kinks: Measured sequences of lateral-strain jumps across half-cells reveal a kink propagating at approximately constant velocity across the strip. Experiments report a speed on the order of 1 mm/s, much slower than elastic wave speeds, confirming diffusive (overdamped) dynamics.
• Functional demonstrations: Diffusive kinks enable sensing-like behavior (touch-triggered propagation), dynamic 2D shape morphing (plus-to-minus texture), transport of a ping-pong ball via sequential lifting of arm-links/guiders, sequential grasp-and-release of objects, and cyclic paddle motions (rower-like) that can potentially perform work and support soft robotic functions.
• Practical considerations: Due to limited viscoelastic contrast between Agilus and TangoPlus, imperfections (from fabrication and clamping) and prestretch are exploited to realize snap-back and traveling kinks.

The work demonstrates a synthetic analog of biological diffusive kinks in an overdamped mechanical system. By coupling viscoelastic relaxation with geometrically multistable kirigami unit cells, the authors realize metastable-to-stable transitions that, when spatially coupled, produce a traveling reaction–diffusion front. The findings answer the posed question by showing that a slow, sequential folding wave—akin to Mimosa Pudica—can be engineered in a metamaterial without relying on inertia or continuous external driving, aside from initial rate-controlled loading and local imperfection or touch to nucleate the wave.

Significance includes: (i) introducing diffusive kinks as a new mode of information/mechanical energy transport in mechanical metamaterials; (ii) enabling arbitrarily slow, controlled wave propagation suitable for time-dependent functions, machine logic elements, and plant-inspired actuation; and (iii) providing a minimal reduced-order framework (unit-cell viscoelastic truss and reaction–diffusion continuum) that links material/geometry parameters and imperfections to macroscopic wave behavior.

Relevance to the field spans soft robotics and programmable matter, where controlled, slow, and spatially coordinated motions are valuable. The demonstrations of sensing, morphing, transport, manipulation, and cyclic actuation showcase how diffusive kinks can be harnessed to perform mechanical tasks. The strong role of imperfections reframes them from nuisances into design features that can seed and steer sequential instabilities.

This study shows that viscoelastic kirigami can be engineered to exhibit bistability and viscoelastic snap-back that collectively give rise to traveling diffusive kinks. These slow, overdamped fronts are initiated by imperfections or touch and propagate at near-constant speed, enabling functions such as sensing, dynamic shape morphing, transport, manipulation, and cyclic motions. Diffusive kinks complement inertial transition waves by allowing arbitrarily slow propagation, expanding the toolkit for plant-mimicking motions and soft robotic functionalities.

Future directions include: enhancing robustness and force output via polymers with greater contrast between instantaneous and long-term stiffness; exploiting advances in additive manufacturing to tailor through-thickness material distributions and controlled imperfections; miniaturizing for dynamic optical and physical property modulation (e.g., dynamic holography); integrating feedback and control to program kink initiation, speed, and routing; and extending the approach to other overdamped media (poroelastic structures, microfluidic controllers, hydraulically powered soft robots).

• Material property contrast: The viscoelastic contrast between Agilus and TangoPlus is limited; by itself, it is insufficient to consistently trigger snap-back in prestretched cells, necessitating reliance on geometric imperfections and boundary conditions.
• Sensitivity to imperfections: Snap-back thresholds and delays are highly sensitive to small geometric asymmetries and clamping conditions; while used constructively here, this sensitivity may hinder reproducibility and robustness.
• Friction and substrate effects: Nonuniform friction between the kirigami and the substrate influences kink propagation speed and consistency.
• Force/payload limits: Forces and payload capabilities are bounded by the achievable viscoelastic contrasts and stiffness levels of available photopolymers.
• Model simplifications: The reduced truss and reaction–diffusion models capture qualitative behavior but idealize complex 3D deformations and material nonlinearities; mapping parameters (e.g., β) to real geometries/materials is qualitative.