Well posedness for multidimensional scalar conservation laws with discontinuous non-degenerate flux

D. Mitrovic · 2025

Read Well posedness for multidimensional scalar conservation laws with discontinuous non-degenerate flux by D. Mitrovic (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-2509-19043.txt.

Introduction

Darko Mitrovic, writing alone, studies when a class of flow equations has solutions that behave well. The equations are scalar conservation laws, which track a quantity that is neither created nor destroyed as it moves. Their flux, the rule for how the quantity flows, can jump abruptly where the medium changes. Such jumps arise at interfaces in layered materials or changing road conditions.

Mitrovic allows the flux to be smooth on each side of a smooth interface, with the two sides tied to space and to the solution itself. The flux also vanishes at the edges of the range the solution may take. For this setting, and for a broader case where the flux varies with bounded variation in space, he builds a stable semigroup of solutions. That construction gives the problem a well-posed framework: solutions exist, are unique, and depend steadily on the data.

Catalog record

CREATORD. Mitrovic
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE43 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-19043

Browse the full library · About the Tower of Babel

Full text

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

Well posedness for multidimensional scalar conservation laws with discontinuous non-degenerate flux D. Mitrovic Abstract . We show well posedness of the Cauchy problem for scalar conservation laws with discontinuous non-degenerate flux which locally have the form ∂ t u + ∑ k = 1 d ∂ x k ( f L k ​ ( 𝒙 , u ) ​ H ​ ( − x j + ζ ⁡ ( x ^ j ) ) + f R k ​ ( 𝒙 , u ) ​ H ​ ( x j − ζ ⁡ ( x ^ j ) ) ) = 0 , u | t = 0 = u 0 , a ≤ u 0 ≤ b , a , b ∈ ℝ , \begin{split}&\partial_{t}u+\sum\limits_{k=1}^{d}\partial_{x_{k}}\left(f^{k}_{L}({\bm{x}},u)H(-x_{j}+\zeta(\hat{x}_{j}))+f^{k}_{R}({\bm{x}},u)H(x_{j}-\zeta(\hat{x}_{j}))\right)=0,\\ &u\big|_{t=0}=u_{0},\ \ a\leq u_{0}\leq b,\ \ a,b\in\mathbb{R},\end{split} where 𝒙 ∈ U ⊂ ℝ d {\bm{x}}\in U\subset\mathbb{R}^{d} and x ^ j = ( x j , … , x j − 1 , x j + 1 , … , x d ) ∈ ℝ d − 1 \hat{x}_{j}=(x_{j},\dots,x_{j-1},x_{j+1},\dots,x_{d})\in\mathbb{R}^{d-1} for some open set U U and the Heaviside function H H . The flux functions are such that OPEN f L k , f R k ∈ C 1 ​ ( ℝ d + 1 ) ) for k = 1 , … , d , and f L , R k ​ ( 𝒙 , a ) = f L , R k ​ ( 𝒙 , b ) = 0 f^{k}_{L},f^{k}_{R}\in C^{1}(\mathbb{R}^{d+1}))\ \ \text{ for $k=1,\dots,d,$ and }\ \ f^{k}_{L,R}({\bm{x}},a)=f^{k}_{L,R}({\bm{x}},b)=0 while the interface function satisfies ζ ∈ C 1 ​ ( ℝ d − 1 ) \zeta\in C^{1}(\mathbb{R}^{d-1}) . We also consider the problem of the general form ∂ t u + div 𝔣 ( 𝒙 , u ) = 0 , 𝔣 ∈ B V ( ℝ d ; C 1 ( ℝ ) ) , 𝔣 ( 𝒙 , a ) = 𝔣 ( 𝒙 , b ) = 0 , u | t = 0 = u 0 , a ≤ u 0 ≤ b , a , b ∈ ℝ , \begin{split}&\partial_{t}u+\Div{\mathfrak{f}}({\bm{x}},u)=0,\ \ {\mathfrak{f}}\in BV(\mathbb{R}^{d};C^{1}(\mathbb{R})),\ \ {\mathfrak{f}}({\bm{x}},a)={\mathfrak{f}}({\bm{x}},b)=0,\\ &u\big|_{t=0}=u_{0},\ \ a\leq u_{0}\leq b,\ \ a,b\in\mathbb{R},\end{split} and construct a stable semigroup of solutions. 2020 Mathematics Subject Classification 35L65; 35L45 1. Introduction We consider a multidimensional scalar conservation law with a spatially discontinuous flux, written here in the general form \partial_{t}u+\Div{\mathfrak{f}}({\bm{x}},u)=0,\qquad u|_{t=0}=u_{0}({\bm{x}}),\quad{\bm{x}}\in\mathbb{R}^{d},\;t>0, (1.1) where u(t,{\bm{x}}) is the unknown scalar function, and the flux {\mathfrak{f}}:\mathbb{R}^{d}\times\mathbb{R}\to\mathbb{R}^{d} may have discontinuities in the spatial variable to be specified below. We assume a\leq u_{0}({\bm{x}})\leq b for some constants a,b\in\mathbb{R} , and that {\mathfrak{f}}({\bm{x}},\lambda) is of class C^{1} in the \lambda -variable and has bounded variation in the {\bm{x}} -variable (more precisely, {\mathfrak{f}}\in BV(\mathbb{R}^{d};C^{1}(\mathbb{R}))^{d} ). Conservation laws of the form ( 1.1 ) model various physical phenomena, such as traffic flow on networks with variable road conditions, two-phase flow in heterogeneous porous media, sedimentation processes, and so on. When the flux function is spatially homogeneous (or sufficiently smooth in {\bm{x}} ), the standard theory of Kružkov [ 23 ] provides a well-posedness framework for the Cauchy problem: there is a unique entropy solution u(t,{\bm{x}}) satisfying the entropy inequalities and depending continuously on the initial data in L^{1} . However, if {\mathfrak{f}} has discontinuities in {\bm{x}} , the classical Kružkov entropy condition is not enough to single out a unique solution. Indeed, in one space dimension it has been shown that many different admissible solution concepts can be defined for the same discontinuous-flux problem, depending on additional conditions imposed at the points of flux discontinuity (see, e.g., [ 4 , 11 ] and references therein). The choice of the “correct” admissibility criterion is often guided by the physical modeling context. One popular selection principle is the vanishing viscosity criterion, wherein one declares the physically relevant solution to be the limit of solutions of ( 1.1 ) with a small viscous term \varepsilon\Delta u added. This approach has the advantage of being tied to a specific modeling assumption (negligible but nonzero diffusion) and has been extensively studied (see [ 4 ] for the one-dimensional case with a single discontinuity where also a general theory of L^{1} -dissipative solvers was considered). In multiple space dimensions, the analytical difficulties are even more pronounced. The flux discontinuities may occur along irregular hypersurfaces in \mathbb{R}^{d} , and there is no canonical “Riemann problem” that isolates the behavior at a single interface as in the one-dimensional case. A general existence result for entropy solutions was proved by E. Yu. Panov [ 27 ] using the vanishing viscosity method and compactness arguments, under quite broad assumptions on {\mathfrak{f}} . However, without further admissibility conditions, uniqueness of solutions in the multidimensional discontinuous-flux setting is not guaranteed. Recently, several works have contributed to establishing well-posedness under additional structural assumptions. In [ 4 ] , the authors introduced the concept of an L^{1} - dissipative germ to characterize interface entropy conditions in one space dimension (allowing a finite number of flux discontinuities on the real line). In [ 6 ] , we extended some of these ideas to multidimensional laws with flux discontinuous across a finite number of smooth hypersurfaces. We proposed multidimensional entropy conditions (generalizing the one-dimensional “crossing condition” of [ 18 ] ) and proved that, for fluxes with finitely many spatial discontinuity surfaces, the vanishing viscosity limit does indeed converge to a unique entropy solution satisfying those conditions (thus yielding well-posedness). On the other hand, [ 11 ] established a uniqueness theory for entropy solutions in several dimensions by developing a kinetic formulation and imposing an adapted entropy condition at flux discontinuities. In a subsequent work, [ 12 ] , the authors further analyzed the structure of entropy solutions, proving that any entropy solution (under a genuine nonlinearity hypothesis similar to ( 1.3 )) admits strong traces on the flux discontinuity set and satisfies a generalized Kato inequality. These properties were used to deduce uniqueness of solutions, given an appropriate condition relating the traces of any two solutions at the discontinuities. Despite these advances, a complete well-posedness theory for multidimensional conservation laws with general discontinuous fluxes has remained elusive. In particular, the question of existence of a solution that satisfies a given entropy admissibility criterion (such as the multidimensional entropy inequality of [ 6 ] ) was open in full generality, since compactness methods alone do not ensure that the limit of vanishing viscosity approximations satisfies the desired entropy condition. In the current contribution, we make a step forward by proving both existence and uniqueness to multidimensional scalar conservation laws with flux discontinuity along a non-intersecting manifolds (see Figure 1). More precisely, beside already introduced conditions on flux, we assume that {\mathfrak{f}} satisfies the boundary conditions {\mathfrak{f}}({\bm{x}},a)={\mathfrak{f}}({\bm{x}},b)=0, (1.2) which provide the maximum principle. Moreover, we assume the non-degeneracy condition: for almost every {\bm{x}}\in\mathbb{R}^{d} and every direction {\bm{\xi}}\in S^{d-1} (the unit sphere in \mathbb{R}^{d} ), the function \lambda\mapsto{\bm{\xi}}\cdot\partial_{\lambda}{\mathfrak{f}}({\bm{x}},\lambda) (1.3) is not identically zero on any non-degenerate interval. Let us now precise geometrical conditions that the flux should satisfy. Roughly speaking, we assume that the flux discontinuities are confined to a finite number of smooth “interfaces” that do not intersect each other. Specifically, we assume: (a) There exist

Read the complete text