Existence of variational solutions to doubly nonlinear systems in nondecreasing domains

Leah Schätzler, Christoph Scheven, Jarkko Siltakoski, Calvin Stanko · 2025

Read Existence of variational solutions to doubly nonlinear systems in nondecreasing domains by Leah Schätzler, Christoph Scheven, Jarkko Siltakoski, Calvin Stanko (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-2505-00148.txt.

Introduction

Leah Schätzler and co-authors prove, in a 2025 mathematics paper, that solutions exist for a class of doubly nonlinear systems in domains that grow over time. Such systems model flows where both the storage term and the spreading term bend away from linearity. The moving boundary makes the problem much harder.

The team tackles the problem on domains that never shrink and whose borders are negligibly thin. They assume the energy function is convex and pushes back growth, with no cap imposed from above. A nonlinear version of the minimizing movements method constructs the solutions. Existence proofs like this are the ground floor for any later simulation or application of these models.

Catalog record

CREATORLeah Schätzler, Christoph Scheven, Jarkko Siltakoski, Calvin Stanko
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE97 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-00148

Browse the full library · About the Tower of Babel

Full text

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

Existence of variational solutions to doubly nonlinear systems in nondecreasing domains Leah Schätzler, Christoph Scheven, Jarkko Siltakoski, Calvin Stanko Abstract . For q ∈ ( 0 , ∞ ) q\in(0,\infty) , we consider the Cauchy–Dirichlet problem to doubly nonlinear systems of the form ∂ t ( | u | q − 1 ​ u ) − div ⁡ ( D ξ ​ f ​ ( x , u , D ​ u ) ) = − D u ​ f ​ ( x , u , D ​ u ) \displaystyle\partial_{t}\big(|u|^{q-1}u\big)-\operatorname{div}\big(D_{\xi}f(x,u,Du)\big)=-D_{u}f(x,u,Du) in a bounded noncylindrical domain E ⊂ ℝ n + 1 E\subset\mathds{R}^{n+1} . We assume that x ↦ f ⁡ ( x , u , ξ ) x\mapsto f(x,u,\xi) is integrable, that ( u , ξ ) ↦ f ⁡ ( x , u , ξ ) (u,\xi)\mapsto f(x,u,\xi) is convex, and that f f satisfies a p p -coercivity condition for some p ∈ ( 1 , ∞ ) p\in(1,\infty) . However, we do not impose any specific growth condition from above on f f . For nondecreasing domains that merely satisfy ℒ n + 1 ​ ( ∂ E ) = 0 \mathcal{L}^{n+1}(\partial E)=0 , we prove the existence of variational solutions u ∈ C 0 ​ ( [ 0 , T ] , L q + 1 ​ ( E , ℝ N ) ) u\in C^{0}([0,T];L^{q+1}(E,\mathds{R}^{N})) via a nonlinear version of the method of minimizing movements. Moreover, under additional assumptions on E E and a p p -growth condition on f f , we show that | u | q − 1 ​ u |u|^{q-1}u admits a weak time derivative in the dual ( V p , 0 ​ ( E ) ) ′ (V^{p,0}(E))^{\prime} of the subspace V p , 0 ​ ( E ) ⊂ L p ​ ( 0 , T , W 1 , p ​ ( Ω , ℝ N ) ) V^{p,0}(E)\subset L^{p}(0,T;W^{1,p}(\Omega,\mathds{R}^{N})) that encodes zero boundary values. 2020 Mathematics Subject Classification 35A01, 35A15, 35K51, 35K55, 49J40 1. Introduction The main objective of the present paper is to prove the existence of variational solutions u\colon E\to\mathds{R}^{N} , N\geq 1 , of doubly nonlinear systems, where E\subset\mathds{R}^{n}\times[0,T] for n\geq 2 and some T>0 is a noncylindrical domain. Here, we will focus on domains that are nondecreasing with respect to time (see ( 2.2 ) for the precise definition), whereas general domains will be treated in an upcoming paper. Before we specify in detail the more general systems considered in this paper, we briefly discuss the prototype equation, which is given by (1.1) \displaystyle\partial_{t}(|u|^{q-1}u)-\Div(|Du|^{p-2}Du)=0 \displaystyle\text{ in }E with parameters q\in(0,\infty) and p\in(1,\infty) . For the special choice of q=1 and p=2 , this reduces to the heat equation. In the case of fixed q=1 and arbitrary p\in(1,\infty) we are dealing with the parabolic p -Laplace equation. Vice versa, for fixed p=2 and arbitrary q\in(0,\infty) , ( 1.1 ) is the porous medium equation (also called fast diffusion equation for q>1 ). More generally, for p-1>q , ( 1.1 ) is called slow diffusion equation, and for p-1<q it is called fast diffusion equation [ 31 ] . This is represented in the behaviour of solutions. Slow diffusion equations allow compactly supported solutions and perturbations propagate with finite speed, while for fast diffusion equations, perturbations propagate with infinite speed, which prevents solutions with compact support. Furthermore, parabolic systems also have a wide range of applications such as fluid dynamics, cf. [ 37 , 53 ] , especially the modeling of glacier formations, solidification and crystal growth, as well as the Stefan problem, cf. [ 56 ] , the Skorokhod problem, cf. [ 55 ] , soil science, filtration, cf. [ 4 , 6 , 7 , 38 , 54 ] , and mathematical biology, where reaction-diffusion equations are used for modeling pattern formations, cf. [ 25 , 45 ] . For parabolic equation in noncylindrical domains exhibiting stochastic properties, we refer to [ 42 , 21 , 22 ] . Other prominent examples of problems on time varying domains arise in fluid and quantum mechanics, cf. [ 27 , 10 , 9 ] . A more detailed overview can be found in [ 36 ] . We also want to give a brief insight on previous results. First, we will have a look on doubly nonlinear equations in cylindrical domains. The first doubly nonlinear equations were analyzed in [ 30 ] by Grange and Mignot, as well as in [ 3 ] by Alt and Luckhaus, who proved the existence of weak solutions for the equation \displaystyle\partial_{t}b(u)-\Div(a(b(u),Du))=f(b(u)). Here, b is the gradient of a convex C^{1} -function satisfying b(0)=0 and a(z,\xi) is a continuous and elliptic function in (z,\xi) satisfying a (p-1) -growth condition in the second variable. The authors split the lateral boundary \Omega_{T} into two complementary parts, where the first satisfies a Neumann condition and the second a Dirichlet condition. Furthermore, in their proofs the time derivative \partial_{t}b(u) was replaced by the difference quotient \Delta_{-h}b(u) , where h is fixed, and a Galerkin method was applied to the elliptic problems. The results in [ 3 ] and [ 30 ] were generalized to higher order doubly nonlinear equations on unbounded domains by Bernis in [ 8 ] . Further, Akagi and Stefanelli considered in [ 2 ] a Cauchy-Dirichlet problem for an equation of the form \displaystyle\partial_{t}b(u)-\Div(a(Du))\ni f and with Dirichlet boundary values, where b\subset\mathds{R}\times\mathds{R} and a\subset\mathds{R}^{n}\times\mathds{R}^{n} are maximal monotone graphs satisfying polynomial growth conditions. They proved the existence for the dual formulation \displaystyle-\Div(a(Db^{-1}(v))\ni f-\partial_{t}v by means of the so called Weighted Energy Dissipation Functional approach, which is also known as Elliptic Regularization. Moreover, several existence results for doubly nonlinear equations have been proven via regularization techniques and a priori estimates by Ivanov, Mkrtychyan and Jäger in [ 31 , 32 , 33 ] . For p\in(1,\infty) the Dirichlet boundary values were considered in the space W^{1,p}(\Omega_{T})\cap L^{\infty}(\Omega_{T}) . Later Bögelein, Duzaar, Marcellini and Scheven showed in [ 15 ] that there exist non-negative variational solutions to the Cauchy–Dirichlet problem with time-independent boundary values to a general doubly nonlinear equation \displaystyle\partial_{t}b(u)-\Div(D_{\xi}f(x,u,Du))=-D_{u}f(x,u,Du), where the nonlinearity b is an increasing, piecewise C^{1} function satisfying certain growth conditions, and f merely satisfies convexity and coercivity assumption, but no specific growth conditions. This allows f to have an exponential or non-standard (p,q) -growth, where 1<p<q<\infty . Furthermore, assuming that b=|u|^{q-1}u , and that f is independent of x and u and fulfills a standard p -growth condition, Schätzler [ 49 ] was able to deal with time-dependent Dirichlet boundary values g\in L^{p}(0,T;W^{1,p}(\Omega,\mathds{R}^{N})) with g(0)\in L^{q+1}(\Omega) and \partial_{t}g\in L^{1}(0,T;L^{q+1}(\Omega)) . For related obstacle problems, see also [ 50 , 51 ] . Moving on to equations in noncylindrical domains, in other words domains which are allowed to grow or shrink with respect to time, we have a first look on linear equations. A significant amount of research has been dedicated to linear equations in noncylindrical domains, cf. [ 1 , 5 , 20 , 24 , 41 , 43 ] . The method of minimizing movements, which we also use in the present article, was first implemented by Gianazza and Savaré [ 28 ] for

Read the complete text