Cantor-direction Kakeya sets and a Hölder graph obstruction

Definitions and mathematical statement

N. Katz asks whether, for a triadic Cantor slope set $C$, there is a planar set $E$ that is a union of a "one-dimensional family" of unit segments with slopes in $C$, has Lebesgue measure zero, and whose individually doubled segments have union of positive measure. The family parameterization, multiplicity per slope, measurability and meaning of doubling are not specified. Doubling individual segments along their axes is different from the global homothety $E\mapsto2E$, which preserves nullity.

The aid attributes an affirmative standard finite-scale Kakeya-type result to Bateman--Katz: there are arrangements of $N=3^n$ Cantor-slope tubes for which the ratio of doubled-tube union area to original union area is at least a constant times $\log\log N$. Here axial doubling extends each tube along its direction. The snapshot gives no complete tube parameters or coherent limiting construction. That growing area ratio by itself does not prove the literal existence of one fixed null union with a positive-measure doubled union, and the aid explicitly retains that distinction.

Its additional obstruction uses one exact segment per slope in the standard middle-thirds Cantor set \[ C=\left\{\sum_{j\geq1}a_j3^{-j}:a_j\in\{0,2\}\right\}, \qquad s=\frac{\log2}{\log3}. \] Let $0<\beta\leq1$ and let the placement map $a:C\to{\mathbb{R}}^2$ satisfy $|a(m)-a(m')|\leq L|m-m'|^\beta$. Put $u(m)=(1,m)/\sqrt{1+m^2}$. For any fixed $\lambda>0$, define the centered axial dilation union \[ E_\lambda=\{a(m)+t u(m):m\in C,\ -\lambda/2\leq t\leq\lambda/2\}. \] The aid reports the upper box-dimension bound \[ \overline{\dim}_{\mathrm B}E_\lambda\leq1+\frac{s}{\beta}. \] The same statement holds for an endpoint-anchored placement with $0\leq t\leq\lambda$. Thus if $\beta>s$, every such fixed dilation is planar-null, including the doubled union. The sets in this regular graph model are compact, so measurability is not an added obstacle there.

This theorem rules out the desired measure jump within this regular one-segment-per-slope class. It does not rule it out for irregular placements, multiple segments per slope or a differently defined one-dimensional family. The conventional range $\beta\leq1$ is explicit: for a higher exponent the direction map itself would require a separate regularity audit, and the displayed formula is not automatically extended. Neither a fixed-set affirmative construction nor a complete negative resolution of the original wording is supplied.

Short English statement

The aid separates a known finite-scale tube phenomenon from the source's under-specified fixed-set question. It also proves an obstruction for regular slope-indexed placements: if the placement is sufficiently Hölder continuous, doubling the individual segments still gives a null set. Irregular or differently parameterized families are not excluded.

Sources

• Local UnsolvedMath registry, permanent ID $20001136$, source code catalogue entry, supplied complete source record and research summary.
• American Institute of Mathematics, Recent trends in additive combinatorics, Problem 4.5, N. Katz, supplied URL https://aimath.org/WWN/additivecomb/additivecomb.pdf; canonical local record aim-combinatorics-notes.json, notes[260]. The supplied verification describes an official-PDF check; none was performed here.
• Bateman and Katz, as named by the summary for the finite-scale Cantor-slope tube result. No title, year or full bibliographic reference is supplied, nor an authoritative citation for the Hölder graph obstruction. The original literature field is empty.
• Research attempt 1: partial result; problem partially solved at the recorded run; supplied rigor label "proved", novelty confidence "low". The machine-generated aid is not peer reviewed and requires independent verification before citation.