Published November 15, 2026 | Version 0.1.0-candidate

The fractional Hardy constant on convex domains: slabs are extremal, small-order limits, and attained constants

Authors/Creators

Description

Unrefereed research candidate. In one sentence: among all convex domains the thin slab has the smallest fractional Hardy constant, so at p = 1 every convex domain keeps at least 2−s of the half-space constant (sharp, improving the earlier factor 1/2); the constants have a common small-order limit; and constants strictly below the half-space value are attained, with explicit examples on balls and planar convex domains.

Context. A Hardy inequality says that a function vanishing outside a region cannot stay large near the region's edge without paying for it in roughness; the best constant measures how strongly the edge forces this. In the classical (derivative) version every convex domain has the same constant as a half-space. In the fractional version, for 0 < s < 1 and p ≥ 1, the roughness is the double integral of |u(x) − u(y)|p|x − y|−N−sp over all pairs of points, and the shape matters. Bianchi, Brasco and Zagati (Math. Ann. 2024) showed half-space universality on convex domains when sp ≥ 1 or p = 2 and left the regime sp < 1, p ≠ 2 open. Bianchi, Stefani and Zagati (Nonlinear Anal. 2026) treated p = 1: the interval does strictly better than the half-line, every convex domain keeps at least half of the half-space constant, and they asked about the small-order limit and about attainment in dimension two or more.

Results. (A) For every p ≥ 1 the smallest constant over convex domains is that of the slab RN−1 × (0,1): the interval constant times an explicit angular factor; long boxes approach it. (B) At p = 1 every convex domain has constant at least 2−s times the half-space constant, and this factor is sharp; the gain over 1/2 is close to 2 for small s and vanishes as s → 1. Consequently s times the constant tends to 2NωN/p as s → 0, uniformly over convex domains and for each fixed p ≥ 1, answering the small-order question. (C) For every p ≥ 1, on a bounded domain with C1 boundary a constant strictly below the half-space value is attained. At p = 1 this gives attained constants on every ball and every bounded planar convex C1 domain; for p > 1 and sp < 1 it gives attained constants on balls, bounded planar convex C1 domains, the interval and rounded long boxes for p in explicit ranges above 1 (Table 1).

Limits. The constant of a particular domain is not computed. The interval constant for 1 < p ≠ 2, sp < 1 is unknown. Whether every convex body in dimension three or more has a strict gap (and hence attains) is open. Domains with corners, such as polygons, are not covered by the attainment theorem. The p-ranges are sufficient conditions, not exact transitions.

Contents. The paper (LaTeX source and PDF); a Python script (mpmath) with 67 high-precision numerical checks of the constants, normalisations, inequalities on parameter grids and the thresholds of Table 1, and its recorded output; the figure script and figure; a dated prior-art record; the bibliography from registry records; a machine-readable index; a research-gate record; a review summary; a replay receipt and SHA-256 sums.

Assurance. Written proofs; the numerical checks use high-precision but not interval-certified quadrature and are not part of any proof. Internal AI-assisted review only (five reviewer roles, an adversarial review and an audit of new proofs by a second AI model, clean replays, one revision round, a re-review and a proof audit): not peer review, independent reproduction or formal verification. The open questions and the half-space and interval constants are due to Bianchi, Brasco, Stefani and Zagati; the slicing method follows Loss and Sloane and Bianchi, Brasco and Zagati. Creator: Anonymous. Text and data CC0-1.0; code MIT.

Notes

Subtitle: Every convex domain keeps at least 2^-s of the half-space constant at p = 1, and balls attain their constants. Code (verification/*.py and the CI workflows) is MIT; prose, manuscript, figure and data are CC0-1.0.

Files

convex-fractional-hardy-constants-0.1.0-candidate.pdf

Files (1.1 MB)

Name Size
md5:a284b24e09ecb3d574c0c4a3c7d2472b
430.0 kB Preview Download
md5:e3ce2e23d361e2fbd05e036215176995
636.7 kB Preview Download
md5:9e60126fb1798084af8e800a0ffc51e2
240 Bytes Download

Additional details