Printed bistable structures that deploy and reconfigure

Flat-printed assemblies that snap into stable, load-bearing 3D shapes

Dates
2017โ€“2021
Collaborators
With Kristina Shea (ETH Zurich)
A flat multi-material 3D print that deploys into a stable doubly-curved 3D structure
Flat-printed assemblies of bistable actuators that deploy into stable, load-bearing 3D shapes.

Deployable structures are valued in aerospace, architecture, and medicine because they ship flat or compact and expand on site. The persistent difficulties are precision and integrity: passive deployment tends to be imprecise, prone to jamming, and to leave a structure that cannot carry load once open. This line of work builds reconfigurable structures from a single multi-material 3D print whose deployed states are geometrically prescribed, mechanically stable, and load-bearing. It develops across three stages: the actuator, its autonomous version, and the inverse-design method that targets arbitrary shapes.

A bistable unit actuator

The foundation is a monolithically printed unit actuator based on a Von Mises truss, a shallow two-bar linkage that snaps between two stable equilibria. Both states are stable with no holding force, and the geometry is tuned to maximize stroke, so the expansion ratio approaches unity when actuators are connected in series.

The printed bistable unit actuator in its two states, and the same actuator assembled hierarchically into chevron strips and tetrahedral space-frame structures
The bistable unit actuator (left) and the hierarchical assemblies it builds: serial chevron strips and deployable tetrahedral space frames.

The actuator is a multi-material print combining a stiff structural material, a bistable element, and a compliant hinge. The triggering force is set by the compliant joint's material and length, and we showed it can be tuned over an order of magnitude (roughly 0.5 to 5 N), verified against simulation. The same parameterization also determines whether an assembled sheet activates into positive or negative Gaussian curvature.

Parameterization of the actuator with its stiff, bistable, and compliant materials, and unit configurations that deploy to dome-like and saddle-like curvature
Designing the unit. The actuator combines stiff, bistable, and compliant materials; tuning its geometry sets the trigger force and selects whether the assembled cell deploys to synclastic (dome) or anticlastic (saddle) curvature.

Connecting units in series multiplies the stroke, and a modified dynamic-relaxation solver predicts the deployed geometries to within about 5% of measurement.

Two serially connected actuators deploying, shown in experiment and matching schematic, expanding the structure as both units snap
Two serially connected units. As each snaps, their strokes add, expanding the assembly; experiment and the simulated kinematics agree.

Tiling the cells produces flat precursors that deploy into stable, doubly curved structures, and the approach scales from small modules up to large sheets.

Flat multi-cell tessellations and the positive- and negative-curvature 3D structures they deploy into, with matching simulations
Multi-cell structures. Flat tessellations of the unit cell (left of each pair) deploy into stable domes and saddles (right), reproduced by simulation.
A large, roughly 400 mm flat printed array deploying into a doubly-curved structure
Scaling up. A large (~400 mm) flat array of cells deploys into a self-supporting curved structure, showing the approach is not limited to small modules.

Adding autonomy with shape-memory polymers

The first actuator still requires an external trigger. To make deployment autonomous, we paired the bistable mechanism with a shape-memory polymer: the polymer acts as a temperature-controlled energy source while the bistable element serves as a force amplifier and linear actuator. Above a programmed temperature the polymer recovers and trips the snap, so the structure deploys in response to ambient heat alone, with no motor or external power. Because the recovery temperature can be tuned per actuator, deployment can be sequenced in time, and the deployed structure carries quantified load both during and after activation.

A shape-memory-polymer-driven structure deploying from flat on heating, with expansion and contraction units, and load-bearing tests at 50 g and 200 g in the deployed state
Autonomous, load-bearing deployment. Heating drives the structure from flat to deployed through paired expansion and contraction units; the deployed form sustains load (tested at 50 g and 200 g), with no motor or external power.

Inverse design of multi-stable surfaces

The final stage addresses the design problem directly: given a set of target 3D surfaces, compute one flat sheet that reconfigures into each of them, with every deployed state mechanically stable. We discretize each target as a Chebyshev net, a quadrilateral mesh that captures the surface through the angles of its cells, and map those cells to the planar fabrication sheet.

Discretizing a target surface as a Chebyshev net: sampling points and equal-length edges build a quadrilateral mesh that conforms to the doubly-curved target
Chebyshev-net discretization. The target surface is covered by a mesh of equal-length edges whose quadrilateral cells capture the geometry through their corner angles, providing the map from flat sheet to 3D shape.

The angular defects between the flat and target states are absorbed by bistable actuators along the diagonals, which lengthen or shorten to realize each shape. Unlike purely geometric inverse methods, this guarantees that the reconfigured states are statically stable rather than merely kinematically possible. We fabricated a single multi-material printed sheet that reconfigures into two distinct, stable, doubly curved surfaces.

Why it matters

Taken together, these results establish a route to reconfigurable structures whose deployment is precise, repeatable, autonomous, and load-bearing, all encoded in a single flat print rather than assembled or externally actuated. The targets span temporary shelters and architectural elements, deployable aerospace structures, and biomedical devices that must change shape after placement. The work was published in Scientific Reports, 3D Printing and Additive Manufacturing, and Materials & Design.

Advised by Kristina Shea at ETH Zurich.

Related publications

  1. Chen T, Mueller J, Shea K. Integrated design and simulation of tunable, multi-state structures fabricated monolithically with multi-material 3D printing. Scientific Reports 7, 45671 (2017). PDF
  2. Chen T, Shea K. An autonomous programmable actuator and shape reconfigurable structures using bistability and shape memory polymers. 3DP+ 5(2), 91โ€“101 (2018). PDF
  3. Chen T, Shea K. Computational design of multi-stable, reconfigurable surfaces. Materials & Design 205, 109688 (2021). PDF