Global well-posedness for a time-fractional doubly nonlinear equation
Goro Akagi, Giacomo Enrico Sodini, Ulisse Stefanelli · 2025
Read Global well-posedness for a time-fractional doubly nonlinear equation by Goro Akagi, Giacomo Enrico Sodini, Ulisse Stefanelli (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-2508-13694.txt.
Introduction
Goro Akagi, Giacomo Sodini, and Ulisse Stefanelli prove results about a hard class of equations in a 2025 mathematics paper. Their equation mixes fractional time change with two layers of nonlinearity. Such models suit materials whose response carries memory. The main terms are allowed very general, graph like forms.
The team shows solutions exist for all time, not just a short span. Their method smooths the equation first, then builds approximations on grids. Letting the smoothing fade yields a true weak solution. Uniqueness of the solution is also discussed under extra rules. A mild Lipschitz term rides along as a perturbation. The paper is set in the language of modern analysis and aims at specialists.
Catalog record
| CREATOR | Goro Akagi, Giacomo Enrico Sodini, Ulisse Stefanelli |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-13694
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Global well-posedness for a time-fractional doubly nonlinear equation Goro Akagi, Giacomo Enrico Sodini, Ulisse Stefanelli Abstract . We consider a time-fractional parabolic equation of doubly nonlinear type, featuring nonlinear terms both inside and outside the differential operator in time. The main nonlinearities are maximal monotone graphs, without restrictions on the growth. In addition, a Lipschitz continuous perturbation is considered. The existence of global weak solutions is obtained via a regularization and Galerkin approximation method. Uniqueness is also discussed under some additional assumptions. 2020 Mathematics Subject Classification Primary: 35K55 1. Introduction This paper is concerned with the nonlinear time-fractional parabolic equation (1.1) \partial^{\theta}_{t}(\alpha(u)-\alpha(u_{0}))-\Delta u+\beta(u)\ni g(x,t,u). Here, the symbol \partial^{\theta}_{t} stands for the Riemann–Liouville fractional derivative of order \theta\in(0,1) which is defined by \partial^{\theta}_{t}f(t):=\frac{1}{\Gamma(1{-}\theta)}\frac{{\rm d}}{{\rm d}s}\int_{0}^{t}(t{-}s)^{-\theta}f(s)\,{\rm d}t\quad\forall t>0 where \Gamma is the Euler \Gamma -function. Hence if \alpha(u)|_{t=0}=\alpha(u_{0}) , then the Riemann–Liouville derivative \partial^{\theta}_{t}(\alpha(u)-\alpha(u_{0})) coincides of the Caputo derivative of \alpha(u) of order \theta\in(0,1) in a weak sense. Equation ( 1.1 ) is posed in the space-time cylinder Q:=\Omega\times(0,T) , where T>0 is a final time and \Omega is a nonempty, open, connected, bounded set of {\mathbb{R}}^{n} with Lipschitz boundary \partial\Omega . The nonlinearities \alpha,\,\beta:{\mathbb{R}}\to 2^{\mathbb{R}} are maximal monotone graphs, with no growth restrictions, whereas g(x,t,\cdot) is a Lipschitz continuous function. The initial value \alpha(u_{0}) is given and the equation is complemented by homogeneous Dirichlet boundary conditions. The occurrence of nonlinearities both inside and outside the differential operator in time \partial_{t}^{\theta} qualifies equation ( 1.1 ) as of doubly nonlinear type. Doubly nonlinear problems occur frequently in connection with applications, especially in relation to nonlinear diffusion phenomena. In the nonfractional case, for different choices of the map \alpha , equation ( 1.1 ) may arise in connection to the two-phase Stefan problem (for \alpha={\rm id}+H , where H is the Heaviside graph), the porous-medium equation (for \alpha(u)=|u|^{p-2}u for some p\in(1,2) ) and the Hele–Shaw model (for \alpha=H ), see [ 42 ] for a detailed discussion. Also motivated by their relevance in applications, doubly nonlinear problems are a mainstay in evolution-equation theory. In the classical, nonfractional case, existence results for doubly nonlinear equations can be traced back to Grange–Mignot [ 21 ] . Besides the classical references by Barbu [ 12 ] , DiBenedetto–Showalter [ 17 ] , Alt–Luckhaus [ 10 ] , and Bernis [ 13 ] , one may also refer to [ 2 , 3 , 4 , 5 , 19 , 23 , 24 , 25 , 29 , 34 , 35 , 36 ] , among many others. See also [ 9 , 37 ] for some global variational approaches and [ 8 , 14 , 26 , 27 , 30 ] for a collection of recent results. To the best of our knowledge, existence for time-fractional parabolic doubly nonlinear equations has not been proved yet. This note is intended to fill this gap. Our main result is the well-posedness of a variational formulation of equation ( 1.1 ), where the elliptic operator is considered in the classical weak sense. Note that, in case \alpha is linear, existence of solutions to equation ( 1.1 ) has already been obtained in [ 6 ] , see also [ 28 , 7 ] . In fact, the specific case of porous media has already been considered in [ 31 , 32 , 33 ] . In addition, a number of results on long-time behavior and decay is available, see [ 1 , 18 , 39 , 40 ] , where nonetheless existence of solutions is just assumed. Furthermore, in [ 15 ] , the asymptotic behavior of solutions for such time-fractional porous medium equations has also been studied in detail. The reader is also referred to [ 20 , 38 ] for some alternative approaches based on viscosity solutions. To prove existence of a solution to ( 1.1 ) we proceed by regularization and approximation. The maximal monotone graphs \alpha and \beta are replaced by their Yosida approximations and \alpha is made strongly monotone by adding a multiple of the identity to it. Such a regularized problem is then tackled by a Galerkin discretization method. A suite of a-priori estimates allows the passage to the limit, first in the Galerkin approximation and then in the regularization, bringing to the existence proof. In the classical nonfractional case, a pivotal tool in the analysis of doubly nonlinear problems driven by subdifferential operators is the availability of a local chain-rule equality, which eventually allows for obtaining a-priori estimates. In the time-fractional setting, the local chain-rule equality is replaced by a nonlocal chain-rule inequality, which makes the analysis more challenging. Uniqueness is generally not to be expected for doubly nonlinear problems, see [ 17 ] . We are nonetheless in the position of providing a uniqueness result under some stronger assumptions, namely, for \alpha strongly monotone and \beta Lipschitz continuous. The plan of the paper is as follows. We present our setting and state the well-posedness result, i.e., Theorem 2.1 , in Section 2 . The uniqueness of solutions is proved in Section 3 . The existence proof is then presented in Section 4 . A regularization and Galerkin approximation of the problem is introduced in Sections 4.1 – 4.2 . We then prove a-priori estimates and pass to the limit in the Galerkin approximation in Sections 4.3 – 4.4 . Additional estimates for the regularized problem are obtained in Section 4.5 and we eventually remove the regularization by a limit passage in Sections 4.6 – 4.7 . 2. Main result We devote this section to the statement of the main well-posedness result, namely Theorem 2.1 . Let us start by introducing some notation. 2.1. Functional analytic framework Recall that we ask for (A0): Let \Omega\subset{\mathbb{R}}^{n} be a nonempty bounded Lipschitz domain, and let T\in(0,\infty) . In the following, we set Q:=\Omega\times(0,T) and we use the spaces H:=L^{2}(\Omega),\quad V:=H^{1}_{0}(\Omega),\quad V^{*}:=H^{-1}(\Omega) where V^{*} is the dual of V , so that (V,H,V^{*}) is a classical Hilbert triplet. Note that the choice of the space V corresponds to the case of homogeneous Dirichlet boundary conditions. Other boundary conditions can also be treated, by redefining V to be a closed subset of H^{1}(\Omega) containing H^{1}_{0}(\Omega) . The symbol (\cdot,\cdot) stands for the standard scalar product in H whereas \langle\cdot,\cdot\rangle is the duality pairing between V^{*} and V . We denote by \|\cdot\| the norm in H , by \|\cdot\|_{E} the norm in the generic Banach space E , and by c_{V}>0 the norm of the injection V\subset H , namely, \|u\|\leq c_{V}\|u\|_{V} for all u\in V . We note that c_{V} is given as c_{V}=\lambda_{1}(\Omega)^{-1/2} , where \lambda_{1}(\Omega) stands for the principal eigenvalue of the Dirichlet Laplacian in \Omega . We set A:V\to V^{*} to be the Riesz mapping given by \langle Au,v\rangle:=\int_{\Omega}\nabla u(x)\cdot\nabla v(x)\,{\rm d}x\quad\forall u,v\in V. In particular, it holds \|u\|_{V}^{2}=\langle Au,u\rangle for all u\in V . 2.2. Maximal monotone operators Given any maximal monotone graph \gamma:{\mathbb{R}}\to 2^{\mathbb{R}} with 0\in\gamma(0) , for all r\in D(\gamma):=\{r\in{\mathbb{R}}\ :\ \gamma(r)\not=\emptyset\} we define by \gamma^{\circ}(r) the point in the closed interval \gamma(r) which is closest to 0 . One can uniquely find (see e.g. [ 16 , Example 2.8.1, p. 43] ) a proper, lower semicontinuous and convex