The Dirichlet eigenvalue problems for some concave elliptic Hessian operators

Jiaogen Zhang · 2025

Read The Dirichlet eigenvalue problems for some concave elliptic Hessian operators by Jiaogen Zhang (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-2510-16748.txt.

Introduction

Jiaogen Zhang works on a family of hard equations in a 2025 mathematics paper. The equations involve Hessian operators, which are built from second derivatives of a function. The problem is an eigenvalue problem, asking for special numbers and matching solutions. The solution must also vanish at the edge of its region.

Zhang proves early estimates that any solution must obey. He then shows a first nonzero eigenvalue exists, with a smooth matching solution. The class covered includes several famous operators used in geometry. As a bonus, every operator of one broad invariant family gets a unique first eigenvalue. A theory of splitting solution branches completes the work.

Catalog record

CREATORJiaogen Zhang
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE65 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-16748

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The Dirichlet eigenvalue problems for some concave elliptic Hessian operators Jiaogen Zhang Abstract In this manuscript, we investigate a priori estimates for the solution to the Dirichlet eigenvalue problem for a broad class of concave elliptic Hessian operators of the form F ⁡ ( D 2 ​ u ) = − Λ ​ u in ​ Ω , u = 0 on ​ ∂ Ω . F(D^{2}u)=-\Lambda u\quad\textrm{in}\,\Omega,\qquad u=0\quad\textrm{on}\,\partial\Omega. These operators encompass the Monge-Ampère operator, the k k -Hessian operators, and the p p -Monge-Ampère operators. We impose a fairly mild constraint on the operator F F , allowing us to demonstrate the existence of the first nonzero eigenvalue and its corresponding Γ \Gamma -admissible eigenfunction on the smooth, strictly Γ \Gamma -convex domain Ω ⊂ ℝ n \Omega\subset\mathbb{R}^{n} . Furthermore, we prove that the eigenfunction u 1 u_{1} belongs to C ∞ ​ ( Ω ) ∩ C 1 , 1 ​ ( Ω ¯ ) C^{\infty}(\Omega)\cap C^{1,1}(\overline{\Omega}) . As an application, we prove that every invariant Gårding-Dirichlet operator admits a unique first nonzero eigenvalue. Finally, a bifurcation-type theory for these operators is also established. † † J. Zhang: School of Mathematics, Hefei University of Technology, Hefei, Anhui, 230009, P. R. China; e-mail: [email protected] † † Mathematics Subject Classification (2020): 53C20; 53C21; 58A10; 58A14 Keywords. Concave elliptic Hessian operators; Dirichlet eigenvalue problem; A priori estimates 1 Introduction Let \Gamma\subset\mathbb{R}^{n} (for n\geq 2 ) be an open convex invariant cone , and let f:\Gamma\rightarrow\mathbb{R} be a concave elliptic Hessian operator (as defined in Definitions 2.1 and 2.3 ). We define the corresponding set of real symmetric n\times n matrices characterizing as \mathscr{M}(\Gamma)=\{A\in\mathrm{Sym}^{2}(\mathbb{R}^{n})\,|\,\lambda(A)\in\Gamma\} where \lambda(A) denotes the eigenvalue vector of A . For an arbitrary bounded domain \Omega\subset\mathbb{R}^{n} , a smooth function u in \Omega for which D^{2}u(x)\in\mathscr{M}(\Gamma) for all x\in\Omega is said to be \Gamma -admissible. For notational convenience, we set F(A)=f(\lambda(A))\,,\qquad A\in\mathscr{M}(\Gamma). Our main objective is to establish the existence and regularity of a \Gamma -admissible eigenpair (u_{1},\Lambda_{1}) satisfying the following Dirichlet eigenvalue problem \left\{\begin{array}[]{ll}\begin{matrix}\begin{aligned} F(D^{2}u)=&\,-\Lambda u\\ u=&\,0\end{aligned}&\begin{aligned} \quad&\mbox{in }\Omega,\\ \quad&\mbox{on }\partial\Omega,\end{aligned}\end{matrix}\end{array}\right. (1.1) where D^{2}u denotes the Hessian matrix of u . A notable characteristic of this equation is its degeneracy near the boundary, as the right-hand side tends to zero. Fully nonlinear Dirichlet eigenvalue problems form a deep and interdisciplinary area, bridging fully nonlinear PDE theory, spectral geometry, and the calculus of variations. Their complexity stems fundamentally from the nonlinearity of the operator F —a phenomenon with no analogue in the classical spectral theory of the Laplacian. Furthermore, the Dirichlet eigenvalue problem for such operators reveals deep geometric content through its connection to the Brunn-Minkowski inequality in convex geometry. For further reading, we refer to [ 4 , 23 , 28 ] , among other works. The bibliography related to the Dirichlet eigenvalue problem for fully nonlinear second order operators is very wide. A canonical example is the Dirichlet eigenvalue problem for the Monge-Ampère operator. This problem consists in finding a nontrivial pair (u_{1},\Lambda_{1})\in C(\overline{\Omega})\times\mathbb{R} such that \left\{\begin{array}[]{ll}\begin{matrix}\begin{aligned} \det(D^{2}u)=&\,(-\Lambda u)^{n}\\ u=&\,0\end{aligned}&\begin{aligned} \quad&\mbox{in }\Omega,\\ \quad&\mbox{on }\partial\Omega.\end{aligned}\end{matrix}\end{array}\right. A landmark systematic study of this problem was undertaken by Lions [ 22 ] , who established the existence of the first eigenvalue \Lambda_{1} , proved the regularity u_{1}\in C^{\infty}(\Omega)\cap C^{1,1}(\overline{\Omega}) for the corresponding eigenfunction, and provided a variational characterization of \Lambda_{1} . Subsequent contributions by Tso [ 30 ] , Hong–Huang–Wang [ 17 ] , Le–Savin [ 26 ] , and Le [ 24 , 25 ] established existence, uniqueness, and geometric regularity properties under appropriate convexity constraints. Wang’s seminal work [ 36 ] extended the eigenvalue framework to the k -Hessian operators \sigma_{k}(D^{2}u)=\sum_{1\leq i_{1}<i_{2}<\cdots<i_{k}\leq n}\lambda_{i_{1}}\lambda_{i_{2}}\cdots\lambda_{i_{k}}\,,\qquad\lambda=\lambda(D^{2}u)\,, unifying Monge-Ampère theory with a broader class of fully nonlinear elliptic operators. Recent progress includes the study of the p -Monge-Ampère operators \mathcal{M}_{p}(D^{2}u)=\prod_{1\leq i_{1}<i_{2}<\cdots<i_{p}\leq n}(\lambda_{i_{1}}+\lambda_{i_{2}}+\cdots+\lambda_{i_{p}})\,,\qquad\lambda=\lambda(D^{2}u)\,, with the critical case p=n-1 resolved in [ 15 , 27 ] , and completion of all integer cases 1\leq p\leq n by the author and Zheng [ 38 ] . We now describe the several fundamental structures for f . Let \Gamma be a convex invariant cone with vertex at the origin satisfying \Gamma_{n}=\{\lambda\in\mathbb{R}^{n}\,:\,\lambda_{i}>0,\,i=1,\cdots,n\}\subset\Gamma\subset\{\lambda\in\mathbb{R}^{n}\,:\,\sum_{i=1}^{n}\lambda_{i}>0\}=\Gamma_{1}\,. We shall assume further that f satisfies the following condition on \Gamma . Definition 1.1 . We say an elliptic operator f:\Gamma\rightarrow\mathbb{R} satisfies the Condition (T) if there exists a uniform constant \mathscr{T}>0 such that \prod_{i=1}^{n}\frac{\partial f}{\partial\lambda_{i}}\geq{\mathscr{T}}\,. (1.2) Heuristically, Condition (T) allows us exploit an established upper bound for \{\frac{\partial f}{\partial\lambda_{i}}\}_{i=1}^{n} to derive a lower bound, thereby providing two-sided control. This type of argument is common in a priori estimates when the maximum principle is applied to the linearized operator. We note that this hypothesis is also employed in the recent study by Guo-Phong [ 14 ] , which investigates L^{\infty} estimates for fully nonlinear equations on non-Kähler manifolds. Our first primary result of this manuscript is the following existence theorem. Theorem 1.2 . Let \Omega be a smooth bounded and strictly \Gamma -convex domain within \mathbb{R}^{n} . Assume F is a concave elliptic Hessian operator on an open convex cone \mathscr{M}(\Gamma)\subset\mathrm{Sym}^{2}(\mathbb{R}^{n}) that satisfies the Condition (T) . Then, there exists a positive number \Lambda_{1}=\Lambda_{1}(\Omega) such that ( 1.1 ) possesses a \Gamma -admissible solution u_{1}\in C^{\infty}(\Omega)\,\cap\,C^{1,1}(\overline{\Omega}) . Furthermore, the solution is unique up to scalar multiplication. Remark 1.3 . (1) The positive constant \Lambda_{1} will be referred to as the first eigenvalue of the operator F , and u_{1} will be referred to as its corresponding eigenfunction. (2) Examples of f satisfying Condition (T) including (i) (\Gamma_{k},\sigma_{k}^{\frac{1}{k}}) , the k -Hessian operator for 1\leq k\leq n , where \sigma_{k} is the k -th elementary symmetric polynomial \sigma_{k}(\lambda)=\sum_{1\leq i_{1}<i_{2}<\cdots<i_{k}\leq n}\lambda_{i_{1}}\cdots\lambda_{i_{k}} and \Gamma_{k}=\{\lambda\in\mathbb{R}^{n}\,|\,\sigma_{j}(\lambda)>0,\,j=1,\cdots,k\}. The proof of Condition (T) can be found in the work of Wang [ 37 , p. 183] . (ii) (\mathcal{P}_{p},\mathcal{M}^{1/\binom{n}{p}}_{p}) , the p -Monge-Ampère operator for 1\leq p\leq n , where \mathcal{M}_{p} is a symmetric polynomial given by \mathcal{M}_{p}(\lambda)=\prod_{1\leq i_{1}<i_{2}<\cdots<i_{p}\leq n}(\lambda_{i_{1}}+\lambda_{i_{2}}+\cdots+\lambda_{i_{p}})\,. and \mathcal{P}_{p}=\{\lambda\in\mathbb{R}^{n}\,:\,\lambda_{i_{1}}+\cdots+\lambda_{i_{p}}>0,\,1\leq i_{1}<\cdots<i_{p}\leq n\} . The proof of Condition (T) has been provided in [ 8 , 38 ] . (3) Notably for all n\geq k>l\geq 1 , the Hessian quotient equations (\Gamma_{k},(\frac{\sigma_{k}}{\sigma_{l}})^{1/(k-l)}) fail to meet the Condition (T) . In a recent work, Harvey and Lawson [ 18 ] showed that the class of operators satisfying Condition (T) is unexpectedly broad, and actually includes all invariant Gårding-Dirichlet operators. Moreover, they established the following strict ellipticity property on the Gårding cone \Gamma : f(\lambda+\tau)\geq f(\lambda)+\Big(\prod_{i=1}^{n}\tau_{i}\Big)^{\frac{1}{n}}\qquad\text{for all }\lambda\in\Gamma\,\text{ and }\tau\in\Gamma_{n}\,. (1.3) In Lemma 2.4 below, we prove that inequality ( 1.3 ) implies ( 1.2 ). Combining this with Theorem 1.2 yields the following corollary. Corollary 1.4 . Let \Omega be a smooth bounded and strictly \Gamma -convex domain within \mathbb{R}^{n} . Assume F is an invariant Gårding-Dirichlet operator on an open convex cone \mathscr{M}(\Gamma)\subset\mathrm{Sym}^{2}(\mathbb{R}^{n}) . Then, there exists a positive number \Lambda_{1}=\Lambda_{1}(\Omega) such that ( 1.1 ) possesses a \Gamma -admissible solution u_{1}\in C^{\infty}(\Omega)\,\cap\,C^{1,1}(\overline{\Omega}) . Furthermore, the solution is unique up to scalar multiplication. To prove Theorem 1.2 , it is necessary to investigate the following more general Dirichlet problem: F(D^{2}u)=\psi(x,u)\quad\textrm{in}\,\Omega,\qquad u=0\quad\textrm{on}\,\partial\Omega. (1.4) This problem represents a special case of the form: F(D^{2}u)=\psi(x,u,\nabla u)\,. (1.5) The Dirichlet problem for equations of the type ( 1.5 ) has received considerable attention over the

Read the complete text