Ring Origami: Algebra & Mechanics

• Ring origami is a concept that unifies algebraic constructions—using iterated line intersections in ℂ to form subrings—and geometric mechanisms in closed-loop elastic fold designs.
• The algebraic framework yields cyclotomic rings and rings of integers in imaginary quadratic fields through precise formulas and symmetry-induced closure.
• In mechanics, ring origami principles govern the behavior of annular, cylindrical, and conical folds, leading to buckling, multistability, and spontaneous 2D-to-3D transformations.

Ring origami denotes several distinct but technically connected research traditions. In one established mathematical sense, it is the study of point sets in $\mathbb{C}$ generated from the seed points $S=\{0,1\}$ by iterated intersections of lines in prescribed directions; under suitable hypotheses the constructible set is a subring of $\mathbb{C}$, and in particular can realize rings of integers of imaginary quadratic fields (Buhler et al., 2010, Kritschgau et al., 2016). In mechanics and geometry, the same phrase points to annular, cylindrical, conical, and closed-loop elastic origami systems in which closure around a ring enforces buckling, multistability, or snap-folding, as in curved-crease annuli, concentric folds, origami bellows, Kresling-type conical structures, and rod-based elastic rings (Dias et al., 2012, Nawratil, 2021, Lu et al., 2 Sep 2025). This suggests a useful synthesis: “ring origami” names both an algebraic origami of rings and a geometric origami of closed folded loops.

1. Restricted planar origami and the emergence of rings

In the algebraic-combinatorial model, the sheet is identified with the complex plane $\mathbb C$, the seed points are $S=\{0,1\}$, and a set $U$ of allowed directions is fixed. Directions are represented by unit complex numbers, with $u$ and $-u$ identified because they determine the same geometric line. Given $p\in\mathbb C$ and $u\in U$, the fold through $S=\{0,1\}$0 with direction $S=\{0,1\}$1 is

$S=\{0,1\}$2

and if $S=\{0,1\}$3 are distinct, the corresponding folds intersect in a unique point

$S=\{0,1\}$4

The origami set $S=\{0,1\}$5 is the smallest subset of $S=\{0,1\}$6 containing $S=\{0,1\}$7 and $S=\{0,1\}$8 and closed under all such intersections (Kritschgau et al., 2016).

A central explicit formula is

$S=\{0,1\}$9

from which symmetry, reduction, linearity, projection, and rotation identities follow. In particular,

$\mathbb{C}$0

and

$\mathbb{C}$1

for $\mathbb{C}$2. These identities convert the geometric recursion into algebraic manipulation (Kritschgau et al., 2016).

The foundational theorem of Buhler–Butler–de Launey–Graham states that if $\mathbb{C}$3 is a subgroup of $\mathbb{C}$4 and $\mathbb{C}$5, then $\mathbb{C}$6 is a subring of $\mathbb{C}$7 (Buhler et al., 2010). In the finite equally spaced case $\mathbb{C}$8, the resulting origami rings are cyclotomic: $\mathbb{C}$9 and

$\mathbb C$0

where $\mathbb C$1 (Buhler et al., 2010). The elementary multiplicative generators are quotients of the form

$\mathbb C$2

with $\mathbb C$3.

2. Rings of integers of imaginary quadratic fields

A major arithmetic development is the realization of the ring of integers of every imaginary quadratic field as an origami set. For

$\mathbb C$4

with $\mathbb C$5 squarefree, the ring of integers is

$\mathbb C$6

and

$\mathbb C$7

The theorem proved in “Origami Constructions of Rings of Integers of Imaginary Quadratic Fields” is that if $\mathbb C$8, then

$\mathbb C$9

where

$S=\{0,1\}$0

and

$S=\{0,1\}$1

(Kritschgau et al., 2016).

The proof has two parts in each congruence class: first, every intersection stays inside the target arithmetic lattice; second, every lattice point is constructible from $S=\{0,1\}$2 by a double-induction or lattice-propagation argument. For $S=\{0,1\}$3, the target lattice is $S=\{0,1\}$4, and the construction propagates from adjacent points

$S=\{0,1\}$5

to neighboring horizontal and vertical-diagonal points. For $S=\{0,1\}$6, the same idea is adapted to the shifted lattice

$S=\{0,1\}$7

using the symmetric slanted directions $S=\{0,1\}$8 and $S=\{0,1\}$9 (Kritschgau et al., 2016).

The standard examples are the Gaussian integers $U$0, obtained from $U$1 with

$U$2

the rectangular lattice $U$3, and the shifted lattice

$U$4

for $U$5 (Kritschgau et al., 2016). In this sense, ring origami is not merely a closure theorem for a recursive construction: it is a constructive realization of classical arithmetic orders.

3. Classification, ring criteria, density, and symmetry

Beyond the subgroup case, the structure of origami rings depends sharply on the number and arrangement of directions. For three directions, Nedrenco proved that, after normalization,

$U$6

This set is a ring if and only if

$U$7

equivalently if the trace and norm

$U$8

are integers. In particular, for $U$9, ringness is equivalent to $u$0 being an integer, and then

$u$1

for a positive squarefree integer $u$2 (Nedrenco, 2015). By contrast, if $u$3 contains at least four different directions, then $u$4 is dense in $u$5 (Nedrenco, 2015).

A second explicit classification is given in “When is an origami set a ring?”. There the constructible set is written

$u$6

with

$u$7

Thus every origami set is a rank-$u$8 module over its real part, and ringness is equivalent to a quadratic integrality condition for the basic nonreal generator $u$9. The paper gives several equivalent criteria, including

$-u$0

equivalently

$-u$1

(Möller, 2018).

The symmetry theory of such origami structures has also been made explicit. For three-angle sets $-u$2, the resulting line pattern is periodic and yields only three wallpaper groups: $-u$3 according as the associated triangles are scalene, isosceles, or equilateral (Chari et al., 23 Jun 2025). For larger angle sets, the point set becomes dense and the relevant finite point groups are cyclic or dihedral. More precisely, a finite group occurs as a point group $-u$4 for some angle set with at least three angles if and only if it is

$-u$5

for an even integer $-u$6 (Chari et al., 23 Jun 2025). This gives ring origami a second, group-theoretic sense: cyclic closure can organize not only arithmetic closure but also rotational and reflectional symmetry.

4. Curved-crease annuli and concentric ring folds

In the mechanics literature, ring origami often means a folded annulus or a family of concentric curved folds. The simplest model is an annular strip of thickness $-u$7, width $-u$8, and crease radius $-u$9, with

$p\in\mathbb C$0

folded along a central circular crease. If deformation is isometric away from the crease, the two sides become developable surfaces, and the folded crease is a space curve with curvature $p\in\mathbb C$1, torsion $p\in\mathbb C$2, and dihedral angle $p\in\mathbb C$3. The key geometric relations are

$p\in\mathbb C$4

where $p\in\mathbb C$5 are generator angles on the two sides (Dias et al., 2012).

These equations encode geometric frustration. A cut annulus can remain planar by overlapping, but a closed annulus folded along a circular crease cannot generally remain planar without stretching. The reason is that for any actual fold with $p\in\mathbb C$6,

$p\in\mathbb C$7

so closure must be achieved by out-of-plane buckling and torsion rather than by a planar circle (Dias et al., 2012). In the narrow, stiff, weakly folded regime, the paper derives the asymptotic energy

$p\in\mathbb C$8

with

$p\in\mathbb C$9

and shows that stiff folds have nearly constant curvature with oscillatory torsion, whereas softer folds exhibit oscillatory curvature and torsion (Dias et al., 2012).

For multiple concentric folds, the geometry propagates recursively. In “The shape and mechanics of curved fold origami structures”, the $u\in U$0-th crease has geodesic curvature

$u\in U$1

and the generator angle satisfies

$u\in U$2

Once one folded ring is specified, developability gives recursion relations for neighboring rings. In the continuum limit of vanishing fold spacing, the smooth surface intersecting all mountain folds has Gaussian curvature

$u\in U$3

and a family of open folds with constant fold angle generates a helicoid (Dias et al., 2012). This identifies dense ring pleating with a discrete realization of negatively curved geometry.

5. Cylindrical and conical closed-loop origami

A different line of work studies closed folded cylinders and conical rings assembled from repeated cells. In origami bellows, a periodic crease tessellation is wrapped into a closed tube with $u\in U$4 identical unit cells around the circumference. For rotationally symmetric rigid-face states, closure imposes

$u\in U$5

For both Miura-ori-based cylindrical bellows and triangular tessellation bellows, the rigid-face geometry admits at most two admissible cylindrical states, and the existence of one or two such states is determined by the design angles $u\in U$6 and $u\in U$7 (Reid et al., 2016). The maximally deployable point is

$u\in U$8

where one state is flat-folded and the other maximally extended, giving

$u\in U$9

(Reid et al., 2016).

The same paper emphasizes that geometric bistability and mechanical bistability are not identical. Real bellows balance hinge energy and plate-bending energy, so a mathematically bistable cylinder may self-deploy if the collapsed state cannot be stably maintained. For the tested prototypes, a critical design angle appears near

$S=\{0,1\}$00

above which stronger locking occurs (Reid et al., 2016).

Kresling-type triangulated rings on cones of revolution generalize the classical cylindrical anti-prism strip. The paper “Multi-stable design of triangulated origami structures on cones of revolution” distinguishes an anti-frustum based conical triangulation and a spiral-motion based conical triangulation. In the anti-frustum setting, the closed strip satisfies

$S=\{0,1\}$01

and can snap between realizations $S=\{0,1\}$02 and $S=\{0,1\}$03 on cones with half-apex angles $S=\{0,1\}$04 and $S=\{0,1\}$05 (Nawratil, 2021). In the spiral-motion setting, realizability reduces to three edge equations

$S=\{0,1\}$06

which force

$S=\{0,1\}$07

The resulting bistable designs lie on an algebraic compatibility curve $S=\{0,1\}$08, and multistability is possible only for

$S=\{0,1\}$09

(Nawratil, 2021).

This conical theory also introduces shaky realizations, self-intersection-free intervals, and a normalized area invariant. For spiral-motion realizations, if $S=\{0,1\}$10 is the cross-sectional polygon area and $S=\{0,1\}$11 is the distance from the cut plane to the cone apex, then

$S=\{0,1\}$12

is constant for $S=\{0,1\}$13 (Nawratil, 2021). The same paper defines a snappability index

$S=\{0,1\}$14

evaluated at the shaky realization, thereby linking geometric singularity to a mechanics-based measure of snap-through difficulty.

6. Elastic rod ring origami, inverse design, and adjacent uses of “ring”

A recent explicitly named ring-origami mechanism is the closed-loop rod system with programmed out-of-plane natural curvature. Here square and hexagonal rings are assembled from rod segments of equal length $S=\{0,1\}$15, width $S=\{0,1\}$16, and height $S=\{0,1\}$17, connected by rigid joints or rounded corners, and modeled as multi-segment Kirchhoff rods. The constitutive law is

$S=\{0,1\}$18

with

$S=\{0,1\}$19

and the main control parameter is

$S=\{0,1\}$20

or $S=\{0,1\}$21 for alternating-sign patterns (Lu et al., 2 Sep 2025).

The paper studies type-I rings, in which all edges have the same positive out-of-plane natural curvature, and type-II rings, in which the edge curvatures alternate in sign. The central result is that natural-curvature-induced out-of-plane bending moments destabilize the planar ring, so that a 2D elastic ring spontaneously snaps out of plane and reaches equilibrium in a 3D configuration (Lu et al., 2 Sep 2025). The reported multistable intervals are narrow for square rings,

$S=\{0,1\}$22

for type-I squares, and

$S=\{0,1\}$23

for type-II squares, but significantly wider for type-I hexagons: $S=\{0,1\}$24 (Lu et al., 2 Sep 2025). Multiple 3D stable states occur only when

$S=\{0,1\}$25

and the paper uses $S=\{0,1\}$26 (Lu et al., 2 Sep 2025).

At special discrete curvature values, the ring achieves zero-strain-energy 3D states. Type-I square rings form spherical configurations when

$S=\{0,1\}$27

type-I hexagonal rings form a heart-like state at

$S=\{0,1\}$28

and spherical states at higher half-integers, while type-II square and hexagonal rings form figure-eight configurations at positive integers of $S=\{0,1\}$29 (Lu et al., 2 Sep 2025). These examples make ring origami a concrete design platform for spontaneous 2D-to-3D transformation, multistable 3D transitions, and compact monostable zero-energy 3D configurations.

Two further adjacent literatures delimit the scope of the term. In computational origami design, the circle/river method reduces flap allocation to circle packing, and deciding whether a given set of circles can be packed into an equilateral triangle, a rectangle, or a square is NP-hard (Demaine et al., 2010). In inverse geometric design, spatially modulated generalized Miura-ori tessellations approximate target surfaces under face planarity and vertex developability constraints, with exact one-DOF rigid-foldability for generalized cylinders and constrained optimization for doubly curved surfaces (Dudte et al., 2018). A separate use of “origami” occurs in toric origami manifolds, where the relevant rings are ordinary and equivariant cohomology rings rather than folded annular objects; in that setting one finds face-ring descriptions, torsion-free cohomology in broad classes, and

$S=\{0,1\}$30

for orientable toric origami manifolds with coorientable fold (Ayzenberg et al., 2014, Holm et al., 2014).

Taken together, these strands show that ring origami is not a single theory but a family of closure-driven theories. In algebraic origami, folding generates subrings of $S=\{0,1\}$31, cyclotomic rings, and rings of integers of imaginary quadratic fields. In geometric mechanics, cyclic closure forces annuli, cylinders, cones, and elastic rings into buckled, multistable, or snap-folded 3D states. The common structural feature is the same: a local folding rule becomes nontrivial only when it is required to close globally around a loop.