Ideal Edge Loads: Edge-Governed Mechanisms

• Ideal edge loads are idealized loading configurations where the edge, rather than the bulk, governs the system response across diverse physical contexts.
• In tribological studies, edge pulling in finite incommensurate contacts triggers localized commensuration that significantly boosts static friction beyond square-root scaling.
• In structural and hydrodynamic analyses, edge loads concentrate instabilities and singular behaviors—such as beam flutter and sharp-corner velocity spikes—necessitating advanced numerical methods.

Searching arXiv for relevant papers on “ideal edge loads” and associated edge-loading contexts. I’m going to look up arXiv entries related to “ideal edge loads,” edge-driven friction, edge flutter, sharp-edge hydrodynamic loads, and edge ideals to ensure the article is grounded in the cited literature. Ideal edge loads denote idealized loading configurations in which the dominant response is organized by an edge, corner, free end, or wedge line rather than by a spatially uniform bulk forcing. In tribology, the term is associated with boundary-applied lateral driving in finite incommensurate contacts, where edge pulling can nucleate localized commensuration and sharply increase static friction (Mandelli et al., 2018). In structural stability, it describes a long-beam asymptotic regime in which a follower-load instability becomes localized near the free lower end and effectively independent of total length (Langre et al., 2015). In hydrodynamics and elasticity, it denotes either benchmark sharp-corner loads in ideal potential flow or point-force Green’s functions for forces applied parallel to a straight edge in a wedge (Wang et al., 2021, Daddi-Moussa-Ider et al., 20 Mar 2025). This suggests a unifying interpretation: “ideal” refers to analytically or numerically purified edge-governed loading, not to the absence of strong local effects.

1. Edge-driven superlubric contacts

In the tribological literature, ideal edge loads arise in finite incommensurate contacts driven laterally from the boundary. A representative setting is a finite circular island in a 2D Frenkel–Kontorova model, with a 2D elastic monolayer of spring constants $k_1$ and $k_2$, a periodic substrate potential of amplitude $U_0$, a lattice mismatch, and a finite circular geometry. The main simulations use

$$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$

and the chosen substrate strength lies below the Aubry-like transition threshold for the infinite 2D incommensurate system. Under periodic boundary conditions, the energetically preferred orientation is the Novaco–McTague angle $\theta_{NM}\approx 2.5^\circ$, nearest-neighbor distances remain narrowly distributed around the incommensurate equilibrium value, and the system slides under an arbitrarily small force. The bulk is therefore genuinely superlubric (Mandelli et al., 2018).

The key comparison is between three loading protocols: uniform driving, pulling, and pushing. Static friction is analyzed through

$F_s \propto N_p^\gamma,$

with the expected upper bound for an incommensurate superlubric contact given by

$\gamma_{\max}=\frac12,$

that is, square-root scaling. Edge loading breaks the equivalence of driving protocols in a finite contact because the boundary acts as a nucleation site for local commensuration. Pulling is the strongest enhancement mechanism: as system size increases, the leading edge is stretched, and above a critical particle number

$N_p^* \simeq 1\times 10^4$

a localized commensurate dislocation appears near the boundary. The reported finite-size comparison is $N_p=3.8\times 10^3$ below threshold and $N_p=2.4\times 10^4$ above threshold. Once the commensurate patch forms, static friction rises sharply because the localized region acts as a strong pinning center. Pushing and uniform loading distort the contact more weakly and produce much less pronounced local commensuration.

The mechanism is controlled by the competition between elastic cost and substrate corrugation. The dislocation core width is estimated as

$k_2$0

which gives

$k_2$1

Atomistic simulations of gold islands sliding over graphite support the same picture; the estimated dislocation core width is $k_2$2, and the extracted exponents remain sublinear: $k_2$3 A common misconception is that localized commensuration would necessarily contradict superlubric scaling. The numerical results show the opposite: sublinear scaling can persist even while a boundary-induced commensurate region substantially increases the prefactor and the static-friction threshold.

2. Edge-localized flutter in long beams under follower loads

In structural mechanics, ideal edge loads appear as an effective description of long hanging beams under follower forces. The system is an Euler–Bernoulli beam of length $k_2$4, hanging vertically and tensioned by its own weight, with a partially follower force $k_2$5 applied at the lower end. The transverse deflection $k_2$6 satisfies

$k_2$7

with lower-end conditions

$k_2$8

and upper-end clamping

$k_2$9

Because gravity generates a tension profile that increases with height, bending stiffness dominates near the lower edge while tension suppresses motion farther up. This creates a localized unstable edge region and a stable upper region (Langre et al., 2015).

The characteristic length is the gravity length

$U_0$0

and the long-beam regime is $U_0$1. In that regime, the critical load and frequency approach limiting values, the mode shape becomes localized near the lower edge, and flutter prevails for all values of the follower fraction $U_0$2. This sharply contrasts with the short-beam regime, where the problem reduces to the generalized Beck column and the instability type depends on $U_0$3: divergence for $U_0$4 and flutter for $U_0$5.

With the nondimensionalization

$U_0$6

the governing equation becomes

$U_0$7

For the semi-infinite beam, matched asymptotics yield an implicit flutter condition, and at threshold one obtains

$U_0$8

The corresponding flutter frequency satisfies

$U_0$9

and the threshold mode shape is

$$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$0

The critical load is independent of $$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$1 in this long-beam asymptotic flutter solution, and the threshold is fixed by an Airy-function edge condition near the neutral point where $$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$2.

A local wave-stability argument uses

$$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$3

For $$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$4, the local tension $$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$5 is negative enough to support unstable waves; for $$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$6, tension stabilizes the beam. The neutral point corresponds dimensionally to

$$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$7

This supports the interpretation that the instability is governed by a lower edge region of length $$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$8, while the rest of the beam is dynamically passive. In this sense, the follower force behaves as an ideal edge load: the unstable dynamics are generated locally at the boundary and do not require the entire finite-length structure to participate.

3. Sharp-edge hydrodynamic loads in potential flow

In hydrodynamic analysis, ideal edge loads refer to the benchmark loads obtained in an inviscid, incompressible, irrotational setting when the sharp-corner singularity is represented accurately rather than smeared numerically. The governing equation is Laplace’s equation,

$$a=1,\qquad k_1=10,\qquad k_2=5,\qquad \lambda_{sub}\approx \frac{1+\sqrt5}{3},\qquad U_0=0.075,$$9

and the singularity enters through the corner expansion

$\theta_{NM}\approx 2.5^\circ$0

For a rectangle corner with interior angle $\theta_{NM}\approx 2.5^\circ$1, the leading term is

$\theta_{NM}\approx 2.5^\circ$2

The velocity potential may remain regular while velocity components diverge algebraically; this is why edge loads involving velocity squares or velocity derivatives become numerically delicate (Wang et al., 2021).

The frequency-domain body–fluid problem is formulated with harmonic motion $\theta_{NM}\approx 2.5^\circ$3, body Neumann data, a free-surface Robin condition, and a radiation condition on a truncated matching boundary. Conventional FEM interpolates

$\theta_{NM}\approx 2.5^\circ$4

whereas XFEM enriches the approximation by shifted singular corner functions,

$\theta_{NM}\approx 2.5^\circ$5

with enrichment functions taken from the corner-flow solution, typically

$\theta_{NM}\approx 2.5^\circ$6

Three enrichment strategies are compared: point enrichment, patch enrichment, and radius enrichment. Radius enrichment is emphasized as mesh-independent, unlike point and patch enrichment. Because enriched bases generate singular integrands, the implementation uses DECUHR adaptive quadrature for 2D singular element integrals and adaptive Gaussian subdivision for 1D singular integrals.

Two benchmarks organize the analysis. For a thin flat plate in an infinite fluid domain, the computed quantities are the velocity potential, the horizontal velocity along the plate, and the added mass. Radius enrichment gives the best convergence: for the potential, the rate improves from about $\theta_{NM}\approx 2.5^\circ$7 to $\theta_{NM}\approx 2.5^\circ$8 for linear XFEM and from about $\theta_{NM}\approx 2.5^\circ$9 to $F_s \propto N_p^\gamma,$0 for quadratic XFEM. For a forced heaving rectangle at the free surface, the linear hydrodynamic coefficients $F_s \propto N_p^\gamma,$1 and $F_s \propto N_p^\gamma,$2 are relatively insensitive to the edge singularity, but the second-order mean vertical load obtained by direct pressure integration is not. The time-averaged force depends on $F_s \propto N_p^\gamma,$3 and on

$F_s \propto N_p^\gamma,$4

so it is directly sensitive to the singular velocity field. Conventional linear and quadratic FEM struggle to converge for this load even on very fine meshes, whereas XFEM converges rapidly even on coarse meshes; quadratic XFEM performs best overall. The reported implementation guidance is correspondingly specific: use XFEM for sharp-edge problems when loads involve local velocity singularities, prefer radius enrichment, use the first singular term for quadratic XFEM, and use about three enrichment functions for linear XFEM.

4. Green’s functions for forces parallel to a straight edge in wedges

A more literal idealized edge loading problem is the derivation of Green’s functions for a point force applied parallel to the straight edge of a wedge. The geometry consists of two planar boundaries meeting at a straight edge, with wedge opening $F_s \propto N_p^\gamma,$5, $F_s \propto N_p^\gamma,$6, and the force located at

$F_s \propto N_p^\gamma,$7

and oriented along the edge direction $F_s \propto N_p^\gamma,$8. The elastic displacement field satisfies the Navier–Cauchy equation

$F_s \propto N_p^\gamma,$9

and the solution is represented through the Imai harmonic potentials,

$\gamma_{\max}=\frac12,$0

Assuming incompressibility, the same formulas apply to low-Reynolds-number viscous flow after the identifications $\gamma_{\max}=\frac12,$1, $\gamma_{\max}=\frac12,$2, and $\gamma_{\max}=\frac12,$3 so that $\gamma_{\max}=\frac12,$4 (Daddi-Moussa-Ider et al., 20 Mar 2025).

The boundary configurations are 2NOS, 2FRS, and NOS–FRS. No-slip imposes $\gamma_{\max}=\frac12,$5 at $\gamma_{\max}=\frac12,$6. Free-slip imposes

$\gamma_{\max}=\frac12,$7

The derivation applies a Fourier transform in $\gamma_{\max}=\frac12,$8,

$\gamma_{\max}=\frac12,$9

and a Kontorovich–Lebedev transform in $N_p^* \simeq 1\times 10^4$0,

$N_p^* \simeq 1\times 10^4$1

which reduces the Laplace equation to

$N_p^* \simeq 1\times 10^4$2

The general transformed solution is therefore

$N_p^* \simeq 1\times 10^4$3

with coefficients fixed by the wedge boundaries.

For a force along $N_p^* \simeq 1\times 10^4$4, the bulk scalar potential is

$N_p^* \simeq 1\times 10^4$5

where

$N_p^* \simeq 1\times 10^4$6

and

$N_p^* \simeq 1\times 10^4$7

The wedge solution is written as a free-space part plus a boundary correction. After inversion, the potentials are expressed as one-dimensional integrals over $N_p^* \simeq 1\times 10^4$8, with kernels involving Legendre functions. The limit $N_p^* \simeq 1\times 10^4$9 recovers the planar half-space image-system results, providing a consistency check. The physical significance is direct: the edge-parallel point-force singularity is the fundamental solution for edge-guided microswimmers in low-Reynolds-number flow and for inclusions or actuators in wedge-shaped elastic environments.

5. Common structural themes and misconceptions

Across these literatures, a recurring theme is that the edge is not merely a geometric boundary but the location where the dominant mechanism is selected. In superlubric contacts, the bulk remains incommensurate and weakly corrugated while the leading edge nucleates a localized commensurate dislocation; in hanging beams, the upper region becomes a passive reservoir of stable tension while the lower edge supports unstable waves; in sharp-corner hydrodynamics, the potential can be well behaved even though the corner velocity field is singular; and in wedge Green’s functions, the bulk fundamental solution must be augmented by a boundary correction because the edge and intersecting walls alter the singular response (Mandelli et al., 2018, Langre et al., 2015, Wang et al., 2021, Daddi-Moussa-Ider et al., 20 Mar 2025).

Several misconceptions follow from ignoring this localization. One is that sublinear friction scaling should exclude a substantial static-friction threshold; the edge-driven FK results show instead that sublinear scaling and a strong friction boost can coexist. Another is that mesh refinement with smooth finite elements is always sufficient for sharp-edge loads; the XFEM study shows that this is not practical for second-order mean loads computed by direct pressure integration, because the relevant integrand depends on singular velocity components. A further misconception is that boundary sensitivity disappears in long systems; the beam problem shows the opposite asymptotic trend, namely that sufficiently long structures can become more edge-controlled, not less.

A plausible cross-domain implication is that ideal edge loads are typically governed by a characteristic localization scale. The tribological problem uses the critical particle number $N_p=3.8\times 10^3$0 and the dislocation core width $N_p=3.8\times 10^3$1; the beam problem uses the gravity length $N_p=3.8\times 10^3$2 and the neutral point $N_p=3.8\times 10^3$3; the XFEM study uses the enriched corner neighborhood; and the wedge problem is organized by the opening angle $N_p=3.8\times 10^3$4 and by transform-space decay. In each case, the relevant asymptotics depend on when the bulk ceases to control the response and the edge-governed mechanism takes over.

6. Distinction from edge ideals in commutative algebra

The phrase should not be confused with the algebraic notion of an edge ideal. In commutative algebra, for a graph $N_p=3.8\times 10^3$5 with edge set $N_p=3.8\times 10^3$6, the edge ideal is

$N_p=3.8\times 10^3$7

where $N_p=3.8\times 10^3$8, and one studies graded Betti numbers, maximal shifts

$N_p=3.8\times 10^3$9

and subadditivity inequalities such as

$N_p=2.4\times 10^4$0

for the minimal graded free resolution of $N_p=2.4\times 10^4$1 (Abedelfatah, 2023).

This terminological distinction matters because “ideal edge loads” belongs to the mechanics, tribology, hydrodynamics, and continuum-response vocabulary summarized above, whereas “edge ideal” refers to a squarefree quadratic monomial ideal associated with a graph. The overlap is lexical rather than conceptual.