On the existence of cut points of connected generalized Sierpiński carpets

Huo-Jun Ruan, Yang Wang, Jian-Ci XiaoAnnales 2023 · DOI

TL;DR

This paper tackles the problem of determining whether a specific type of fractal, known as a generalized Sierpiński carpet (GSC), has "cut points" – points whose removal disconnects the fractal. The authors introduce a more efficient criterion for detecting these cut points, reducing the need to analyze an infinite sequence of related graphs. They also demonstrate the construction of GSCs with any desired finite number of cut points.

What Problem Does This Paper Address?

Fractals, like the Sierpiński carpet, are fascinating geometric objects that often exhibit complex structures. A key question in their study is understanding their topological properties, such as connectedness and the existence of "cut points." A cut point is a point whose removal disconnects a connected space. For generalized Sierpiński carpets (GSCs), a previous result established that they have cut points if and only if a sequence of associated "Hata graphs" (which encode connectivity information) all exhibit a "long tail" property. This condition, however, requires checking an infinite number of graphs, making it computationally challenging. This paper aims to simplify this criterion by showing that checking a finite number of these Hata graphs is sufficient to determine the existence of cut points. Additionally, the paper explores the possible number of cut points a GSC can have, constructing examples with any given finite number of cut points.

Methods

The core of the paper's contribution lies in refining the criterion for the existence of cut points in GSCs. The authors build upon the concept of "Hata graphs" ($\Gamma_n$) associated with GSCs. Previously, a GSC was known to have cut points if and only if $\chi(\Gamma_n) \ge |D|^{n-1}-1$ for all $n \ge 2$, where $\chi$ is a measure of the "long tail" property. The key innovation here is proving that it suffices to check this condition for a finite number of $n$ (specifically, up to $M=38+3$ is suggested). This is achieved through a detailed analysis of "essential cut vertices" in Hata graphs and their relationship to the overall structure of the GSC. The paper also employs constructive methods to demonstrate the existence of GSCs with a specific number of cut points, utilizing geometric arguments and properties of self-similar sets.

Datasets

Not discussed in the paper. The paper deals with theoretical constructions of mathematical objects (fractals and graphs) rather than empirical data.

Key Insights and Ideas

  • Finite Criterion for Cut Points: The most significant insight is that the condition for a GSC to have cut points can be reduced from an infinite check on Hata graphs to a finite one. This makes the problem computationally tractable and provides a more "algorithmic" solution.
  • "Fragile" vs. "Non-fragile" GSCs: The paper distinguishes between "fragile" GSCs (where a specific decomposition leads to a singleton intersection) and "non-fragile" ones. The finite criterion applies to non-fragile GSCs, highlighting a structural difference in how cut points arise.
  • Constructing Any Number of Cut Points: The authors demonstrate that it's possible to construct GSCs with exactly $m$ cut points for any positive integer $m$. This is a powerful result for understanding the diversity of topological properties within this class of fractals.
  • Non-fragile GSCs with Multiple Cut Points: A surprising result is that if a non-fragile GSC has two or more cut points, it must have infinitely many. This implies a strong structural constraint on non-fragile GSCs with a finite number of cut points (only zero or one).

Key Results

  • Theorem 1.3: For a non-fragile connected GSC, cut points exist if and only if a condition related to the Hata graph $\Gamma_M$ holds for a fixed $M$ (independent of the GSC's parameters).
  • Theorem 1.4: For any positive integer $m$, there exists a connected GSC with exactly $m$ cut points.
  • Theorem 1.5: If a GSC is non-fragile and has at least two cut points, then it must have infinitely many cut points.

Limitations and Future Work

  • Constant M: While the paper establishes that a finite $M$ exists for the criterion, the specific value derived ($M=38+3$) is quite large. Future work might aim to find a tighter bound for $M$.
  • Fragile GSCs: The finite criterion is established for non-fragile GSCs. Extending this to fragile GSCs or understanding their cut point behavior more generally could be a direction for future research.
  • Characterization of Cut Points: The paper provides conditions for the existence of cut points and constructs GSCs with a specific number of them. Further work could explore more detailed characterizations of the locations and distribution of cut points.
  • Generalizations: The study focuses on generalized Sierpiński carpets. Investigating similar questions for other classes of fractals could be a natural extension.

Summary updated

Abstract

In a previous work joint with Dai and Luo, we show that a connected generalized Sierpiński carpet (or shortly a GSC) has cut points if and only if the associated \(n\)-th Hata graph has a long tail for all \(n\ge 2\). In this paper, we extend the above result by showing that it suffices to check a finite number of those graphs to reach a conclusion. This criterion provides a truly "algorithmic" solution to the cut point problem of connected GSCs. We also construct for each \(m\ge 1\) a connected GSC with exactly \(m\) cut points and demonstrate that when \(m\ge 2\), such a GSC must be of the so-called non-fragile type.
BibTeX
@inproceedings{ruan2023existence,
  author = {Huo-Jun Ruan and Yang Wang and Jian-Ci Xiao},
  title = {{On the existence of cut points of connected generalized Sierpiński carpets}},
  booktitle = {Annales},
  year = {2023},
  doi = {10.54330/afm.127049},
}

Loading related content...

Figure