ASYMPTOTICS OF LARGE SOLUTIONS OF p-LAPLACE EQUATION ON CYLINDERS BECOMING UNBOUNDED
N. N. Dattatreya · 2025
Read ASYMPTOTICS OF LARGE SOLUTIONS OF p-LAPLACE EQUATION ON CYLINDERS BECOMING UNBOUNDED by N. N. Dattatreya (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-2505-23600.txt.
Introduction
Some equations have solutions that blow up at the boundary of their domain, and mathematicians study how those large solutions behave. Dattatreya examines such solutions of the p-Laplace equation on cylinder-shaped domains. The cylinders grow longer and longer, approaching an endless cylinder. The question is whether the solutions settle into the pattern of a simpler problem posed on the cylinder's cross-section.
The answer is yes: the solutions converge, in the right function-space sense, to the cross-section solution, and the paper pins down how fast. Earlier work had rates only for the case where the power p is at least two and for infinite boundary values. Dattatreya extends the rates to all p above one and handles finite boundary values in the same framework. The rate, notably, does not depend on the boundary data chosen.
Catalog record
| CREATOR | N. N. Dattatreya |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-23600
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
ASYMPTOTICS OF LARGE SOLUTIONS OF p-LAPLACE EQUATION ON CYLINDERS BECOMING UNBOUNDED N. N. Dattatreya Abstract . In this article, we study the asymptotic behavior of large solutions for a quasi-linear equation involving the p-Laplacian, defined on a sequence of finite cylindrical domains converging to an infinite cylinder. We demonstrate that the sequence of solutions converges locally, in the Sobolev norm, to a solution of the corresponding cross-sectional problem. Moreover, we establish a convergence rate. As part of our analysis, we extend existing convergence results for the case p ≥ 2 p\geq 2 , which previously lacked explicit convergence rates, to the range 1 < p < 2 1<p<2 . We additionally address solutions with finite Dirichlet boundary data within a unified framework and exhibit that this rate of convergence is independent of the boundary data. 1991 Mathematics Subject Classification 35B44, 35A01, 35J92, 35J62, 35J25. 1. Introduction The problems motivated by hadron colliders and other parts of physics entail studying the asymptotic behaviour of the sequence of solutions to PDEs defined on finite cylinders becoming unbounded in a particular direction or directions. The goal is to capture the essence of the solution or solutions to a similar PDE on the corresponding infinite cylinder. Such types of asymptotics are studied for a broad class of equations, including linear, semilinear and quasilinear. While much of the study focuses on the equation involving a forcing term, the nonlinearity case is scarcely addressed. In this context, the present article investigates similar asymptotic results for quasilinear problems involving p-Laplacian, wich is defined as follows: Let (1) \omega_{1} be a convex domain in \mathbb{R}^{m} such that 0\in\omega_{1} . (2) \omega_{2} be an open bounded domain in \mathbb{R}^{n-m} . (3) S_{\ell}=S_{\ell}(\omega_{1},\omega_{2}):=\ell\omega_{1}\times\omega_{2} and S=S(\omega_{2}):=\mathbb{R}^{m}\times\omega_{2} (4) f:[0,\infty)\to[0,\infty) be a non-decreasing continuous function such that f(0)=0 . (5) g:S\to\mathbb{R} be a W_{loc}^{1,p}(S) function such that g(x)=g(X_{2}) and F(g(x))\in L_{loc}^{1}(S) , where x=(X_{1},X_{2})\in\partial S , X_{2}\in\partial\omega_{2} and (1.1) F(x)=\int_{0}^{x}f(s)\ ds. We consider (1.2) \text{div}\left(|\nabla u_{\ell}|^{p-2}\nabla u_{\ell}\right)=f(u_{\ell})\quad\text{in}\ S_{\ell}, with either of the following boundary conditions: (1.3) u_{\ell}(x)=g(x)\quad\text{on}\ \partial S_{\ell}, (1.4) u_{\ell}(x)\to+\infty\quad\text{as}\ \text{dist}(x,\partial S_{\ell})\to 0, And the corresponding cross-sectional problem (1.5) \text{div}\left(|\nabla u_{\infty}|^{p-2}\nabla u_{\infty}\right)=f(u_{\infty})\quad\text{in}\ \omega_{2}, where, the gradient is in \mathbb{R}^{n-m} . With the notation x=(X_{1},X_{2})\in S_{\ell} , where X_{1}\in\ell\omega_{1} and X_{2}\in\omega_{2} , we consider either of the following boundary conditions: (1.6) u_{\infty}(X_{2})=g(X_{2})\quad\text{on}\ \partial\omega_{2} (1.7) u_{\infty}(X_{2})\to+\infty\quad\text{as}\ \text{dist}(X_{2},\partial\omega_{2})\to 0. Solutions that blow up at the boundary, as in ( 1.4 ) and ( 1.7 ) are called Boundary blow-up solutions or Large solutions , we refer to [ 21 , 9 ] and the references therein for more details. The necessary and sufficient condition on the nonlinearity f that ensures the existence of a large solution is called the Keller-Osserman condition: (A1) \Psi_{p}(r):=\biggl(1-\frac{1}{p}\biggl)^{\frac{1}{p}}\int_{r}^{\infty}\frac{dx}{F(s)^{\frac{1}{p}}}<\infty\quad\text{for all}\ r>0, where, F is defined in ( 1.1 ). The Keller-Osserman condition first appeared in 1957 in the independent works of Keller [ 18 ] and Osserman [ 22 ] , which were to aid Electrodynamics and Geometry, respectively. For the existence of finite solutions, as in ( 1.3 ) and ( 1.6 ), assumption ( (A1) ) is not required; the existence and uniqueness of a finite, locally C^{1,\alpha} weak solution are addressed in [ 12 ] and C^{1,\alpha} regularity in [ 13 ] . The existence of solutions can also be obtained by monotonicity methods [ 14 ] . For more regarding p-Laplacian we refer to [ 19 ] and [ 16 ] . A priori estimates for positive solutions can be found in [ 10 ] and references therein. We refer to [ 20 , 11 , 21 ] for the existence of large solutions. The uniqueness of a large solution requires an additional assumption on f , we refer to [ 21 ] , and the condition is as follows (A2) \displaystyle\liminf_{t\to\infty}\frac{\Psi_{p}(t)}{\Psi_{p}(t)}>1\quad\text{for all}\ \in(0,1). For more on the uniqueness we refer to [ 15 ] . With this setup, we derive the following theorems, where, the first one addresses the asymptotics of solutions with finite boundary data Theorem 1.1 . We assume ( (A1) ). Let u_{\ell} solve ( 1.2 ), ( 1.3 ), and u_{\infty} solve ( 1.5 ), ( 1.6 ), then there exists a constant C independent of \ell such that ||\nabla(u_{\ell}-u_{\infty})||_{L^{p}(\tilde{S})}\leq\frac{C}{\ell^{1/p}}, for any \tilde{S}\subset\subset S and \ell sufficiently large. And the second one addresses large solutions, Theorem 1.2 . We assume ( (A1) ). Let u_{\ell} be the solution to ( 1.2 ) and ( 1.4 ), then (1) There exist u\in W_{loc}^{1,p}(S) such that u_{\ell}\rightharpoonup u in W^{1,p}_{loc}(S) and for every X_{1}\in\omega_{1} , v\left(X_{2}\right):=u\left(X_{1},X_{2}\right) solves ( 1.5 ) with ( 1.7 ). (2) With the assumption ( (A2) ), u(X_{1},X_{2})=u_{\infty}(X_{2}) for every X_{1}\in\omega_{1} . (3) There exists a positive constant C independent of \ell such that ||\nabla(u_{\ell}-u)||_{L^{p}(\tilde{S})}\leq\frac{C}{\ell^{1/p}}, for any \tilde{S}\subset\subset S and \ell sufficiently large. The theorems are stated separately for clarity, but they will be proved simultaneously. The assumption ( (A1) ) in Theorem 1.1 is a technical requirement; to drop it, one would need to replace Proposition 2.3 with an alternative bound that does not rely on the existence of a large solution in one dimension. Our approach unifies arguments for large solutions and finite solutions. Also, we aim to obtain a convergence rate for large solutions as an addendum to the work [ 9 ] and extend the results to solutions with finite boundary values. Cylindrical domains naturally arise in various applications, such as measuring pressure in water or gas pipelines, modelling gas flow in pipes [ 1 ] , and theoretically in the study of cylindrical gravitational waves, which originate from cylindrical sources and propagate along the axial direction [ 3 ] , They also appear in the context of porous media flow [ 2 ] . The asymptotic behaviour of solutions in the above-mentioned settings has been extensively studied for various operators having forcing terms with different boundary conditions and eigenvalue problems. And we refer to some of the works [ 5 , 8 , 7 , 17 ] . A study of said asymptotics of solutions is carried out in the case of a uniformly elliptic operator with Dirichlet boundary data, and the convergence of eigenvectors is studied in [ 6 ] . Semilinear equations are studied in [ 4 ] . This work extends the results discussed in [ 9 ] in three aspects: While the cited article focuses on infinite boundary data, our focus is on both finite and infinite boundary data. The result in [ 9 ] is established for 2\leq p<\infty . In the current article, we extend them to 1<p<2 . Additionally, this article provides a rate for convergence. Although the proof is a simple bookkeeping, the rate and the extension to 1<p<2 case together justify the article. In section 2, we give definitions and preliminary results. In section 3, we prove Theorem 1.1 , and Theorem 1.2 with an additional assumption ( (A2) ). Section 4 discusses the loss of uniqueness; here, we also give the proof of the case 1<p<2 . 2. Preliminaries We began by defining the notation of a solution , subsolutions and supersolutions , followed by a statement of Comparison Principle . Subsequently, we derive bounds for \{u_{\ell}\}_{\ell} and \left\{||\nabla u_{\ell}||_{L^{p}}\right\}_{\ell} . However, before proceeding, we