The work sits at the intersection of two lines of research that developed in parallel. On one hand, digital photopatterning of polymer films already allowed creating regions with different stiffness and shrinkage, inducing bending and wrinkling in a flat sheet. On the other hand, computational origami learned to unfold arbitrary surfaces into crease patterns, and replacing mountain folds with compressed rods and the rest with a stretched membrane gave membrane tensegrity. In the archive, this work is surrounded by studies on light chips, robotic film growth, oxide membranes with memory, as well as methods for inverse design of materials and high-speed X-ray imaging of melting. However, there are almost no neighbors that combine polymer films with origami inverse design to obtain arbitrary dry 3D shapes: the closest work is at a distance of 0.519, and no one stands close to this topic.
First, a parent film is prepared from a copolymer of ethyl acrylate and tert-butyl acrylate with a cross-linker, by casting the solution into the LCD printer's bath and exposing the entire area with a 405 nm UV LED. Then the film is immersed in a solution of the second monomer (methyl methacrylate with cross-linker) until equilibrium swelling. The swollen film is placed between two PET sheets on the LCD screen of the same printer and exposed through a programmed bitmask for 80–90 seconds, polymerizing the second monomer only in the open areas. Afterward, the unreacted monomer is washed out with acetone, and upon drying, the unexposed areas shrink to 64.4% of the swollen size, while the exposed rods retain 96.3%. The inverse design is implemented in Rhinoceros/Grasshopper: the target 3D mesh is converted into an origami tessellation using generalized Resch patterns, developability is optimized in Crane, and then mountain folds are assigned as rods, while the remaining faces and valley folds become the membrane. Additionally, perimeter treatment is introduced: either rod segments are inserted along the boundary so that after shrinkage the length matches the target, or zero fold angles are set on the contour so that the rods form a continuous locking ring.
After drying, the membrane regions shrink to 64.4% of the swollen size, while the rod regions retain 96.3%, giving a difference in planar strain of 50%. The Young's modulus of the rod material is 0.77 ± 0.16 GPa, and that of the membrane is 0.38 ± 0.05 MPa, so the stiffness contrast reaches about 2000-fold. This combination allows the shrinking membrane to stretch over the unconnected rods and hold the final shape without external support. For three target surfaces—a dome with positive Gaussian curvature, a hyperbolic surface, and an elementary cell of a gyroid with negative curvature—the average deviation of the scanned shape from the target was 1.0%, 2.1%, and 1.7% of the bounding box diagonal, respectively. Perimeter optimization proved crucial: for the gyroid, the average deviation dropped from 3.0% without contour treatment to 1.8% with partial length matching and to 1.3% with full closure of the rods into a ring with zero fold angles.
This work differs from neighboring studies by merging two previously separate lines—digital photopatterning of polymers and inverse design of membrane tensegrity—into a single technological pipeline. Unlike hydrogel systems that require an aqueous environment and yield temporary shapes, here we obtain dry, permanent, self-supporting structures. Unlike multilayer self-folding films, where rigid and soft phases are stacked through the thickness and precise two-sided alignment is needed, here rods and membrane lie side by side in a single layer, and all shape information is encoded in a one-sided exposure pattern. This is the first demonstration that origami-based inverse design can be directly transferred to a polymer film using a consumer LCD printer, without specialized equipment.
Two directions for improvement arise from current limitations. First, the local fold angle is not controlled: the membrane shrinkage is uniform, whereas the target surface typically requires a distribution of angles. Grayscale exposure modulation could set the desired distribution and improve accuracy. Second, star vertices may snap through and invert during drying; shifting each vertex to the desired branch would make the process robust for targets with many small protrusions. If the membrane is replaced with a stimulus-responsive polymer, the transformation becomes reversible: flat films could be stored compactly and deployed into load-bearing surfaces upon exposure to humidity or another trigger. The inverse design algorithm naturally extends to biomimetic functional geometries, such as the wing-like surface that Firmiana forms from the flat wall of its fruit.
The method could influence the design of lightweight deployable structures, soft robotics, and biomimetic materials where dry, self-supporting 3D shapes from flat blanks are needed. The possibility of roll-to-roll manufacturing opens the way to inexpensive, large-format surfaces with complex curvature.
Immediate next steps include implementing grayscale exposure modulation to control local fold angles and developing methods to suppress snapping of star vertices. It is also worth testing the reversibility of the transformation when the membrane is replaced with a stimulus-responsive polymer.
In the field of polymer materials, an open question remains: how to precisely control local curvature during self-assembly from a flat sheet without external influences. This work shows that origami-based inverse design combined with one-sided photopatterning solves this problem for moderately complex surfaces, but the issue of controlling fold angles and preventing unwanted bifurcations during drying is not yet closed.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
This work combines digital photopatterning of polymer films with origami-inspired inverse design to turn a flat elastomer sheet into arbitrary 3D surfaces. A film of ethyl acrylate and tert-butyl acrylate is swollen in a second monomer and selectively exposed to UV light through a liquid-crystal photomask, creating stiff rods within a soft, shrinking membrane. An inverse design algorithm computes the rod layout from the target surface, and perimeter optimization reduces the average deviation to 1.0–2.1% of the target size.
Where its strength is
A key strength is the large contrast in shrinkage (50%) and elastic modulus (about 2000-fold) between the phases, which allows the membrane to stretch over unattached rods and hold its shape without external support. This paves the way for roll-to-roll manufacturing of dry, permanent, self-supporting structures without precise two-sided alignment.
Where it could go next
Worth trying
Why this may matter
Further development of this approach could lead to low-cost roll-to-roll production of 3D structures with programmable shape, which are in demand in soft robotics, biomedical devices, and metamaterials.
Nearby works in our archive
The archive is sparse around this work: the closest review is only 0.49, against a usual 0.64. It looks like few have drilled nearby.
Goethe, in 'The Metamorphosis of Plants' (1790), suggested that all plant organs are modified leaves. Modern mechanics confirms: the shape of an organ is determined by inhomogeneous deformation and stiffness patterns, and it is precisely this principle that is reproduced in the polymer film.
The idea of a flat sheet that folds itself into a complex three-dimensional shape according to a programmed pattern resonates with 'smart matter' from science fiction, such as the programmable materials in Neal Stephenson's 'The Diamond Age' or the deployable structures in the series 'The Expanse'.