Rod-Sphere Model: A Cross-Disciplinary Framework

• Rod-Sphere Model is a family of representations that couple anisotropic rod-like elements with spherical components through mechanisms like confinement, composition, and mechanics.
• It includes variants such as rods on spherical surfaces, sphere-rod composite particles, and mechanically linked mixtures, each defined by unique geometric constraints and interaction laws.
• The model provides practical insights for applications in packing, rheology, locomotion, and computational simulations, underscoring its cross-disciplinary relevance in soft matter and continuum mechanics.

The literature surveyed here suggests that “Rod-Sphere Model” is not a single standardized model but a family of representations in which rod-like and spherical elements are coupled through confinement, composition, or mechanics. Concrete realizations include hard or soft spherocylinders on a spherical surface, mixed suspensions of rods and spheres, T-shaped sphere-rod amphiphiles, a low-Reynolds swimmer built from two spheres connected by perpendicular rods, continuum Lennard-Jones sphere-rod interactions, and a rod-like intruder percolating through a static bed of spheres (Smallenburg et al., 2016, Rajendra et al., 2022, Anzivino et al., 2023, Haputhanthrige et al., 10 Feb 2025, Najafi et al., 2010, Wang et al., 2024, Petit et al., 12 Jun 2026). A plausible implication is that the term functions as a cross-disciplinary umbrella rather than a unique model class.

1. Scope and principal usages

Across the cited literature, the phrase is attached to several distinct geometric meanings. In some papers, the “sphere” is a confining substrate and the “rod” is the mobile anisotropic particle. In others, the sphere and rod are separate constituents of a composite particle, a mixture, or a mechanical assembly. This terminological breadth is essential, because the governing variables, admissible degrees of freedom, and relevant observables differ substantially between these usages.

Usage	Defining construction	Representative source
Rods on a spherical surface	Spherocylinders constrained by a sphere	(Smallenburg et al., 2016, Rajendra et al., 2022)
Sphere-rod composite particle	Hydrophilic sphere attached to a rod midpoint	(Haputhanthrige et al., 10 Feb 2025)
Rod-sphere assembly or mixture	Rods and spheres linked or mixed mechanically	(Najafi et al., 2010, Anzivino et al., 2023, Petit et al., 12 Jun 2026, Wang et al., 2024)

The hard-particle close-packing model of short spherocylinders confined tangentially to a sphere and the soft-repulsive monolayer model of spherocylinders whose centers lie on a sphere are the most direct realizations of rods on spherical geometry (Smallenburg et al., 2016, Rajendra et al., 2022). By contrast, the KTOF4 system is a chemically specific “sphere-rod” amphiphile with a T-shaped geometry, while the two-sphere propeller, dense rod-sphere suspensions, and granular percolation problem use rods and spheres as distinct mechanical entities (Haputhanthrige et al., 10 Feb 2025, Najafi et al., 2010, Anzivino et al., 2023, Petit et al., 12 Jun 2026).

2. Geometric definitions and kinematic degrees of freedom

One major class consists of spherocylinders constrained by spherical geometry. In the hard-particle packing problem, each rod has center-of-mass position $\mathbf{r}_i$ constrained by $|\mathbf{r}_i|=R$ and orientation $\hat{\mathbf{n}}_i$ constrained to the local tangent plane, $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$, with aspect ratio in the range $0 \le L/D \le 1$ (Smallenburg et al., 2016). In the soft-repulsive monolayer, by contrast, the center is constrained to $|\mathbf{r}_i|=R$ and the translational velocity satisfies $\mathbf{v}_i\cdot \mathbf{r}_i=0$, but the rod is free to rotate about any axis that passes through its center of mass (Rajendra et al., 2022). That distinction is decisive: one model enforces tangential directors, whereas the other permits full 3D orientation above a spherical manifold.

A second class uses genuine sphere-rod composite particles. The KTOF4 amphiphile comprises a hydrophilic Keggin-type cluster treated geometrically as a rigid sphere and a hydrophobic oligodialkylfluorene segment treated as a rigid rod, joined by a flexible linker synthesized through azide-alkyne Huisgen cycloaddition (Haputhanthrige et al., 10 Feb 2025). The nontrivial geometric point is that the sphere is attached to the center of the rod-like segment, not to one end, so the resulting amphiphile is T-shaped rather than head-tail linear. The paper’s schematics show the Keggin cluster as a roughly 1.0 nm diameter sphere and the OF segment as a rod of approximate length 3.4 nm and width 1.3 nm.

A third class uses rods and spheres as separate bodies in mechanical models. The low-Reynolds swimmer consists of a large sphere of radius $R$, a small sphere of radius $a$, a long rod of variable length $L(t)$, and a short rod of fixed length $|\mathbf{r}_i|=R$0, with the angle between the rods fixed at $|\mathbf{r}_i|=R$1 (Najafi et al., 2010). Its two internal degrees of freedom are

$|\mathbf{r}_i|=R$2

with the small sphere position in the body frame

$|\mathbf{r}_i|=R$3

In dense-suspension rheology, spheres of diameter $|\mathbf{r}_i|=R$4 are mixed with rods modeled conceptually as spherocylinders but implemented numerically as rigid linear assemblies of glued spheres of diameter $|\mathbf{r}_i|=R$5 (Anzivino et al., 2023). In granular percolation, the rod is explicitly a rigid body made of $|\mathbf{r}_i|=R$6 collinear small spheres of diameter $|\mathbf{r}_i|=R$7, with length

$|\mathbf{r}_i|=R$8

aspect ratio

$|\mathbf{r}_i|=R$9

and scaled length

$\hat{\mathbf{n}}_i$0

(Petit et al., 12 Jun 2026).

3. Interaction laws and governing mechanics

The simplest rod-sphere formulations are hard-core or purely repulsive. In the tangential packing model, excluded volume is purely hard-core: two rods are either non-overlapping or forbidden, and the optimization problem is to minimize $\hat{\mathbf{n}}_i$1, equivalently maximize the number density

$\hat{\mathbf{n}}_i$2

and surface coverage

$\hat{\mathbf{n}}_i$3

(Smallenburg et al., 2016). In the soft monolayer model, rods interact via a generalized Weeks–Chandler–Andersen repulsion written in terms of the minimum distance $\hat{\mathbf{n}}_i$4 between the two rod cores,

$\hat{\mathbf{n}}_i$5

with the interaction force acting along the line of shortest distance between the two spherocylinder cores (Rajendra et al., 2022).

A more explicit continuum interaction law appears in the analytical sphere-thin-rod treatment. There the pairwise potential is the Lennard-Jones 12–6 law

$\hat{\mathbf{n}}_i$6

the sphere is a 3D continuum of LJ sites, and the rod is a material line of uniform line number density $\hat{\mathbf{n}}_i$7 (Wang et al., 2024). The sphere-rod interaction is built by integrating the sphere-point potential along the rod,

$\hat{\mathbf{n}}_i$8

and the paper gives exact closed forms for the finite-rod potential, the infinite-rod limit, the force on the sphere, and the torque on the rod.

Hydrodynamic rod-sphere models introduce force-free and torque-free constraints rather than direct pair potentials. The two-sphere propeller is analyzed in Stokes flow using Oseen’s Green’s function for a point force near a rigid sphere, with the swimmer rigid-body motion obtained from

$\hat{\mathbf{n}}_i$9

(Najafi et al., 2010). Granular rod-sphere transport instead uses DEM with a linear spring-dashpot normal force law,

$\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$0

with tangential forces explicitly set to zero, so contacts are frictionless (Petit et al., 12 Jun 2026).

4. Ordering, packing, and topological structure

When rods are constrained to a sphere, curvature-induced frustration organizes both positional and orientational order. In the hard tangential model, the planar reference area is

$\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$1

and the corresponding planar close-packing benchmark is

$\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$2

(Smallenburg et al., 2016). Small clusters show a rich variety of structures, including chiral threefold-symmetric motifs, while for larger clusters very short rods with $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$3 remain largely disordered and rods with $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$4 or $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$5 prefer a global baseball-like geometry of smectic-like domains. The model also introduces global polar and baseball order parameters and a local alignment parameter $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$6.

The soft monolayer model produces a different sequence because rods are not forced into the tangent plane. For the main case $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$7, $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$8, and $\hat{\mathbf{n}}_i \cdot \hat{\mathbf{r}}_i = 0$9, the system shows a low-density disordered fluid for $0 \le L/D \le 1$0, an intermediate orientationally ordered spherical fluid for $0 \le L/D \le 1$1, and a high-density solid for $0 \le L/D \le 1$2, with the LC-solid transition near $0 \le L/D \le 1$3 and $0 \le L/D \le 1$4 (Rajendra et al., 2022). The ordered fluid is a hedgehog-like monolayer in which rods align approximately along the local outward normal, quantified by

$0 \le L/D \le 1$5

At higher packing fractions the system forms a curvature-frustrated solid with crystalline domains, disclinations, dislocations, and grain-boundary scars.

Sphere-rod amphiphiles add a further variant of spherical ordering. In dilute water/1,4-dioxane solution, KTOF4 forms spherical inclusions with radii roughly 25–90 nm and alternating KTOF4-rich and solvent-rich layers with a repeat spacing of about 4.8 nm (Haputhanthrige et al., 10 Feb 2025). Two internal organizations occur: concentric spherical shells and nearly planar lamellae inside a spherical boundary. The layered structures show no in-plane orientational order and therefore are described as analogous to smectic A. The flat internal layers display elementary edge and screw dislocations, with Burgers vector

$0 \le L/D \le 1$6

This contrast suggests that “rod-sphere” ordering on curved geometry may refer either to tangent-field frustration, to normal-field hedgehog alignment, or to lamellar packing inside finite spherical aggregates.

5. Rheology, transport, and locomotion

In dense shear flow, rod-sphere models are used to study how shape anisotropy modifies jamming and viscosity. The non-Brownian suspension model parameterizes mixtures by total packing fraction

$0 \le L/D \le 1$7

and relative sphere fraction

$0 \le L/D \le 1$8

with viscosity fitted by

$0 \le L/D \le 1$9

(Anzivino et al., 2023). For pure spheres the fit gives $|\mathbf{r}_i|=R$0, $|\mathbf{r}_i|=R$1, and $|\mathbf{r}_i|=R$2. The central result is that adding short rods decreases viscosity, whereas adding long rods increases viscosity at fixed total packing fraction, with a practical crossover at roughly $|\mathbf{r}_i|=R$3 and a maximum jamming fraction near $|\mathbf{r}_i|=R$4.

In locomotion, the two-sphere propeller shows how rods and spheres can generate nonreciprocal actuation in Stokes flow. The stroke

$|\mathbf{r}_i|=R$5

traces a loop in cylindrical shape space and bypasses the scallop theorem (Najafi et al., 2010). The resulting swimmer undergoes combined translation and rotation, typically along a helical path, and both speed and direction are tunable through $|\mathbf{r}_i|=R$6, $|\mathbf{r}_i|=R$7, $|\mathbf{r}_i|=R$8, $|\mathbf{r}_i|=R$9, $\mathbf{v}_i\cdot \mathbf{r}_i=0$0, $\mathbf{v}_i\cdot \mathbf{r}_i=0$1, $\mathbf{v}_i\cdot \mathbf{r}_i=0$2, and $\mathbf{v}_i\cdot \mathbf{r}_i=0$3.

In granular transport through sphere beds, the control parameters are the rod length ratio $\mathbf{v}_i\cdot \mathbf{r}_i=0$4 and the size ratio $\mathbf{v}_i\cdot \mathbf{r}_i=0$5 (Petit et al., 12 Jun 2026). The geometrical reference threshold is

$\mathbf{v}_i\cdot \mathbf{r}_i=0$6

which is the trapping threshold for a spherical intruder based on the minimum pore throat defined by three touching large spheres. The model identifies trapping and passing regimes, with a divergence of the penetration length

$\mathbf{v}_i\cdot \mathbf{r}_i=0$7

For the example $\mathbf{v}_i\cdot \mathbf{r}_i=0$8, the fit gives $\mathbf{v}_i\cdot \mathbf{r}_i=0$9 and $R$0. Trapped rods always have

$R$1

contacts, and short rods percolate nearly twice as fast as long rods.

For hybrid hydrodynamic models with a rod component near a boundary, a uniformly valid local drag law for a slender rod parallel to a plane wall provides a wall-corrected resistive-force closure for the rod part, whereas the sphere part must generally be handled separately (Koens et al., 2021).

6. Computational strategies, reductions, and conceptual limits

The computational realization of rod-sphere models is highly model-specific. Hard rods on spheres have been studied by constant-$R$2 Monte Carlo with variable sphere radius, orientational bias fields, and event-driven molecular dynamics adapted to rods constrained on a sphere (Smallenburg et al., 2016). Soft monolayers use NVT molecular dynamics with velocity Verlet, an adaptation of RATTLE to enforce spherical constraints, and a Berendsen thermostat (Rajendra et al., 2022). Dense suspensions are simulated in LAMMPS using DEM with lubrication and drag, while granular passing and trapping in sphere beds are simulated in MercuryDPM (Anzivino et al., 2023, Petit et al., 12 Jun 2026). Analytical models replace numerical pair summations by exact integrations, as in the continuum LJ sphere-rod potential and the Stokeslet-near-sphere swimmer formulation (Wang et al., 2024, Najafi et al., 2010).

The limits of the term are equally important. Some formulations are purely geometric and hard-core; others require solvent selectivity, amphiphilic segregation, electrostatics, anchoring, or hydrodynamic image systems. The KTOF4 study is explicit that purely excluded-volume shape effects are not sufficient, because the reentrant flat–concentric–flat sequence is controlled by solvent properties rather than geometry alone (Haputhanthrige et al., 10 Feb 2025). A plausible implication is that no single reduced “rod-sphere model” can be universal across packing, rheology, amphiphilic self-assembly, microswimming, and granular percolation.

In continuum mechanics and soft robotics, the review of rod theories does not explicitly formulate a rod-sphere model, but it places such constructions conceptually near lumped-mass models, pseudo-rigid chains, and discrete elastic rods, with Cosserat rod theory as the most general continuum starting point (Alessi et al., 2024). This suggests that many rod-sphere models can be read as problem-specific reductions of more general slender-body theories, chosen to preserve the rod anisotropy while simplifying contact detection, confinement, or coupling to spherical geometry.