The Bounded Diameter Conjecture and Sharp Geometric Estimates for Mean Curvature Flow
Yiqi Huang and Wenshuai Jiang · 2025
Read The Bounded Diameter Conjecture and Sharp Geometric Estimates for Mean Curvature Flow by Yiqi Huang and Wenshuai Jiang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-17060.txt.
Introduction
Yiqi Huang and Wenshuai Jiang prove a long open claim about moving surfaces in this 2025 paper. The surfaces move by mean curvature flow, a rule that shrinks each point toward its center of curve. As time nears the first bad point, shapes can stretch or tear. The authors show the inner size of the shape stays bounded.
Their proof also gives sharp bounds on how much the surface bends. They bound the size of the set of bad points as well. The bounds are best possible, as a known ring example shows. No extra shape limits are needed for the results. The same ideas also work for flows that pass through bad points.
Catalog record
| CREATOR | Yiqi Huang and Wenshuai Jiang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 78 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-17060
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Bounded Diameter Conjecture and Sharp Geometric Estimates for Mean Curvature Flow Yiqi Huang and Wenshuai Jiang Abstract . We show that the intrinsic diameter of mean curvature flow in ℝ 3 \mathbb{R}^{3} is uniformly bounded as one approaches the first singular time T T . This confirms the bounded diameter conjecture of Haslhofer [ Has21 , Has25 ] . In addition, we establish several sharp quantitative estimates: the second fundamental form A A has uniformly bounded L 1 L^{1} -norm on each time slice, A A belongs to the weak L 3 L^{3} space on the space-time region, and the singular set 𝒮 \mathcal{S} has finite ℋ 1 \mathcal{H}^{1} -Hausdorff measure. All of the results are optimal due to the marriage ring example and our results do not require any convexity assumptions on the surfaces. Furthermore, our arguments extend naturally to flows through singularities, yielding the same sharp estimates. 1. Introduction A smooth family of closed embedded surfaces \mathcal{M}\subset\mathds{R}^{3}\times[0,T) evolves by mean curvature flow ( [ Bra78 , Hu84 ] ) if the normal velocity at each point is given by the mean curvature vector. Finite-time singularities are inevitable, and a central problem in the study of mean curvature flow is to understand the structure of the singular set and the geometry of the flow near singularities. In this paper, our first main result establishes a uniform bound for the intrinsic diameter of the flow, thereby confirming a conjecture of Haslhofer: Conjecture 1.1 (Haslhofer [ Has21 , Has25 ] ) . The intrinsic diameter of the mean curvature flow \mathcal{M} in \mathds{R}^{3} remains uniformly bounded as t\nearrow T , the first singular time. This conjecture was inspired by an analogous conjecture of Perelman in three-dimensional Ricci flow [ Pe02 ] . For mean curvature flow, the extrinsic diameter is trivially bounded by the avoidance principle, but the intrinsic diameter is far subtler: near singularities, the evolving surface may develop long, spiraling, or fractal-like regions whose intrinsic geometry could, a priori, degenerate. By Huisken’s monotonicity formula [ Hu90 ] , singularities of mean curvature flow are modeled on self-shrinkers. Previous attempts to address Conjecture 1.1 have relied on precise classifications of possible singularity models. The first such result was obtained by Gianniotis and Haslhofer [ GH20 ] , who proved the conjecture in the mean convex case. Thanks to the deep work of White [ Wh00 , Wh03 ] , all tangent flows in this setting are multiplicity-one round cylinders, see [ Hu93 , HS99a , HS99b , SW09 , An12 , CM12 , HK17a ] . Combining this classification with the Łojasiewicz–Simon inequality for cylindrical shrinkers established in Colding and Minicozzi’s seminal work [ CM15 ] (also [ CIM15 ] ) and the canonical neighborhood theorem from Haslhofer-Kleiner [ HK17a , HK17b ] , they ruled out the formation of “fractal tubes” near cylindrical singularities and obtained a uniform diameter bound. For general flows without mean convexity, however, the singularities are considerably more intricate. By the classification theorems of Wang [ Wa16b ] and Bamler–Kleiner [ BK23 ] , every non-compact shrinker in \mathbb{R}^{3} has either conical or cylindrical ends. Du [ Du21 ] recently verified Conjecture 1.1 under the additional assumption that the flow only encounters multiplicity-one neck singularity or singularity with conical end, using the mean convex neighborhood theorem by Choi-Haslhofer-Hershkovits [ CHH22 ] and the uniqueness of tangent flows at conical singularities proved by Chodosh and Schulze [ CS21 ] . A recent breakthrough of Bamler and Kleiner [ BK23 ] resolved the higher-multiplicity issue in mean curvature flow in \mathds{R}^{3} . Nevertheless, the structure of singularities with cylindrical ends remains largely mysterious. This connects to a long-standing question posed by Ilmanen [ Il95 ] : Is the round cylinder the only complete embedded shrinker in \mathbb{R}^{3} with a cylindrical end? Wang [ Wa16a ] proves the rigidity under the assumption of the convergence of infinite order on the cylindrical end, and constructed non-trivial incomplete examples with such ends. Apart from this, however, very little is known. As a consequence, previous approaches relying on establishing the canonical neighborhood theorems cannot address the general setting. In this paper, we confirm the bounded diameter conjecture 1.1 in full generality. In fact, our main theorem provides much stronger information: we establish optimal curvature estimates and quantitative regularity of the singular set. Theorem 1.2 . Let \{\mathcal{M}_{t}\subset\mathds{R}^{3}\}_{t\in[0,T)} be a smooth closed embedded mean curvature flow, then there exists some constant C=C(\mathcal{M}_{0})<\infty such that for all 0\leq t<T (1) The intrinsic diameter of \mathcal{M}_{t} is uniformly bounded: D_{int}(\mathcal{M}_{t})<C; (2) The L^{1} -norm of the second fundamental form A is uniformly bounded: \int_{\mathcal{M}_{t}}|A|<C; (3) The second fundamental form A has bounded weak L^{3} -norm: \mu_{\mathcal{M}}\big(\{X\in\mathcal{M}:|A|(X)\geq s^{-1}\}\big)\leq Cs^{3}; (4) The singular set \mathcal{S}_{T} of the final time slice \mathcal{M}_{T} is contained in a countable union of embedded 1-dimensional Lipschitz manifolds together with countably many points. Moreover, \mathcal{S}_{T} has finite 1-dimensional Hausdorff measure: \mathcal{H}^{1}(\mathcal{S}_{T})\leq C. Remark 1.3 . The estimates in Theorem 1.2 are sharp. For example, the marriage ring solution, which shrinks to a round circle at time T , realizes equality in the curvature and singular set bounds. Remark 1.4 . Our argument in fact yields the stronger Minkowski estimate: Area(B_{r}(\mathcal{S}_{T}))\leq Cr , which refines the Hausdorff estimate above. Moreover, our argument provides an integral estimate for regularity scale, which is much stronger than the curvature estimate. The uniform L^{1} -bound on the second fundamental form A is in fact stronger than the intrinsic diameter bound. Indeed, Topping [ To08 ] established the inequality D_{int}(\mathcal{M}_{t})\leq C\int_{\mathcal{M}_{t}}|H| , which shows that controlling the mean curvature in L^{1} already implies a diameter bound. However, we do not need to use Topping’s result in our proof. In fact, we prove the diameter bound and the L^{1} bound of A simultaneously. Under the mean convexity assumption, Gianniotis and Haslhofer [ GH20 ] proved an L^{1} -bound on A ; see also related work of Head [ He13 ] , Cheeger–Haslhofer–Naber [ CHN13 ] , and Du [ Du21 ] . Our result removes all convexity assumptions and establishes the same conclusion in complete generality. For the space–time estimates, Cheeger–Haslhofer–Naber [ CHN13 ] previously obtained the L^{p} bound for A with any p<3 in the mean convex setting. Without assuming convexity, we improve this to an optimal weak L^{3} bound, which is sharp in view of the cylindrical self-shrinkers. It is instructive to compare this with the corresponding elliptic estimates. In [ NV20 ] , Naber-Valtorta obtained the optimal weak L^{7} estimate for minimizing hypersurfaces, refining the earlier L^{p} bound with p<7 by Cheeger-Naber [ CN13b ] . Similar optimal estimates have since been obtained for minimizing harmonic maps [ NV17 ] and related work for Yang-Mills and Einstein manifolds [ NV19 , JN21 ] . Our results further extend beyond the first singular time, to weak formulations of mean curvature flow. There are several notions of weak mean curvature flow in the literature, including the Brakke flow and level-set formulations. Very recently, Bamler and Kleiner [ BK23 ] introduced the notion of almost regular flows : these are well-behaved Brakke flows whose local scale functions satisfy quantitative integrability conditions. They proved that this class of flows