Similarity solutions of shock formation for first-order strictly hyperbolic systems

Jun Eshima, Luc Deike, Howard A. Stone · 2025

Read Similarity solutions of shock formation for first-order strictly hyperbolic systems by Jun Eshima, Luc Deike, Howard A. Stone (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-2511-00672.txt.

Introduction

Jun Eshima, Luc Deike, and Howard Stone study the moment a smooth flow breaks into a shock, a sudden jump that hyperbolic equations produce across physics and engineering. The textbook case is the inviscid Burgers equation, where earlier work showed the dynamics near the newborn shock are self-similar and universal: the same shape appears whatever the starting conditions were.

The authors prove that generosity extends far beyond the textbook case. For general first-order strictly hyperbolic systems in one spatial dimension, shock formation is again self-similar and universal, with the same character as the Burgers case. They derive an analytical formula for the universal self-similar solution. Forecasters of breaking waves and traffic-like shocks gain a single local law to work with.

Catalog record

CREATORJun Eshima, Luc Deike, Howard A. Stone
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-00672

Browse the full library · About the Tower of Babel

Full text

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

Similarity solutions of shock formation for first-order strictly hyperbolic systems Thanks: Submitted to the editors DATE. Jun Eshima, Luc Deike, Howard A. Stone Abstract Shocks due to hyperbolic partial differential equations (PDEs) appear throughout mathematics and science. The canonical example is shock formation in the inviscid Burgers’ equation ∂ u / ∂ t + u ​ ∂ u / ∂ x = 0 \partial u/\partial t+u\partial u/\partial x=0 . Previous studies have shown that when shocks form for the inviscid Burgers’ equation, for positions and times close to the shock singularity, the dynamics are locally self-similar and universal, i.e., dynamics are equivalent regardless of the initial conditions. In this paper, we show that, in fact, shock formation is self-similar and universal for general first-order strictly hyperbolic PDEs in one spatial dimension, and the self-similarity is like that of the inviscid Burgers’ equation. An analytical formula is derived and verified for the self-similar universal solution. keywords Hyperbolic PDEs, Similarity solutions, Burgers’ equations Funding. This research was conducted under the NSF Grant No. 2246791 to H.A.S. and NSF grant No. 2242512 to L.D., and the Eli and Britt Harari Fellowship in the Department of Mechanical and Aerospace Engineering, Princeton University to J.E. † † runningheads: Singularities of hyperbolic PDEs / J. Eshima, L. Deike, and H. A. Stone AMS XXX 1 Introduction Shocks due to hyperbolic partial differential equations (PDEs) have long been studied in mathematics and science with seminal works given by Challis [ 2 ] , Stokes [ 18 ] , and Riemann [ 16 ] . For a summary, see [ 3 , 12 ] . The sustained interest arises due to the prevalence of such equations, with applications spanning many fields of physics, such as fluid dynamics, including the shallow water and compressible gas equations, soft matter physics, such as nonlinear elasticity, magnetohydrodynamics, as occurs in astrophysics and plasma physics, as well as more generic phenomena such as traffic flow. Explicitly, we consider a system of PDEs given by \frac{\partial\boldsymbol{f}}{\partial t}=\boldsymbol{M}(\boldsymbol{f})\cdot\frac{\partial\boldsymbol{f}}{\partial x}, (1) where \boldsymbol{f}(x,t)=(f_{1}(x,t),\cdots,f_{N}(x,t)) are the N dependent variables, x is the spatial variable, t denotes time, and \boldsymbol{M}(\boldsymbol{f}) is an arbitrary function of \boldsymbol{f} (an N by N matrix). Such a system of PDEs given by ( 1 ) is said to be hyperbolic [ 3 ] if \boldsymbol{M} has a basis of eigenvectors, i.e., is diagonalizable, with only real eigenvalues. Furthermore, if the eigenvalues are distinct, then the system is said to be strictly hyperbolic . In this work, we only consider one space dimension. Mathematical studies of shocks often analyze fundamental questions such as uniqueness, existence, and well-posedness. A less commonly asked question is how the shocks are formed, which is the main focus of this paper. Understanding how shocks form is of practical value, as it naturally leads to the understanding of their subsequent regularization. To answer such a question, a novel approach was presented by Pomeau [ 15 ] and Eggers, and Fontelos [ 4 ] , who investigated the inviscid Burgers’ equation \frac{\partial u}{\partial t}+u\frac{\partial u}{\partial x}=0, (2) and showed that the shock formation is locally self-similar through a universal solution (see section 2 ). The Burgers’ equation is often regarded as the prototypical equation for shock formation for various applications [ 1 ] , such as benchmarking numerical codes. From the perspective of shock formation, several authors [ 7 , 17 , 10 , 13 , 8 ] have shown that particular examples of shock formation in hyperbolic PDEs are closely related to the Burgers’ equation. The formation of singularities from smooth conditions, such as those that arise from the Burgers’ equation, is also of interest when investigating the well-posedness of flow equations. The most famous example is whether the Navier-Stokes equations form singularities in finite-time, which is one of the Millenium prize problems [ 9 ] . Notably, the identification of similarity solutions of finite-time singularities have been used in approaching such problems [ 6 , 19 ] . In this paper, we show that the relation to the Burgers’ equation is a generic property of shock formation of hyperbolic PDEs. In particular, shock formation of first-order systems of strictly hyperbolic PDEs in one dimension is locally self-similar and we derive the self-similar solutions. To clarify, in our analysis, we assume that a finite-time shock exists and investigate the local self-similarity of such shock formation. We therefore do not investigate whether shocks form or not. For such studies, see for example [ 14 , 11 ] . In particular, John [ 11 ] proved that a finite-time shock must form when the initial data for strictly hyperbolic PDEs ( 1 ) is sufficiently small, i.e., the smallness leads the waves to behave as simple waves, which therefore form a Burgers-type shock. The method shown in this paper is a generalization of the analysis of the Burgers’ equation by Pomeau [ 15 ] and Eggers and Fontelos [ 4 ] . Additionally, it is a generalization of a result in a recent work by the authors of this paper [ 8 ] , where the same self-similar solutions of shock formation are derived for the specific example of thin film Marangoni flow. Naturally, the formatting and ideas presented in this paper are therefore closely related to these previous works. The structure of the paper is as follows. In section 2 the existing self-similar shock solution for the Burgers’ equation is given and in section 3 we state the main, analogous result of this paper for equation ( 1 ). The result is derived in detail in section 4 , with some corollaries provided in section 5 . Finally, in section 6 , the method of solution is presented and verified through an example using the shallow water equations. 2 Background: Burgers’ Equation The similarity analysis of the inviscid Burgers’ equation forms the basis of the results to follow. Thus, we first remind a reader of the similarity solution of the inviscid Burgers’ Equation ( 2 ), following the approach given in [ 4 ] . Let a shock form at (x_{*},t_{*}) with u(x_{*},t_{*})=u_{*} . Then, as the shock forms, for positions and times close to the shock, i.e., |x-x_{*}|\ll 1 and t\rightarrow t_{*}^{-} , the leading-order expansion of ( 2 ) gives \frac{\partial u}{\partial t}+u_{*}\frac{\partial u}{\partial x}=0, (3) which is a linear advection equation of velocity u_{*} and does not develop a shock by itself. For further analysis, the next-order nonlinearity is considered, where the shock occurs. We also consider variables local to the shock formation. Let \overline{x}:=x-x_{*}+u_{*}\tau be the frame of the shock and let \tau:=t_{*}-t be the local time, where the sign is chosen so that \tau remains positive as the shock is approached t\rightarrow t_{*}^{-} . Finally, we let u^{\prime}:=u-u_{*} . Then, ( 2 ) becomes \frac{\partial u^{\prime}}{\partial\tau}-u^{\prime}\frac{\partial u^{\prime}}{\partial\overline{x}}=0. (4) The scalings for the self-similar ansatz may be derived by balancing \partial u^{\prime}/\partial\tau and u^{\prime}\partial u^{\prime}/\partial\overline{x} to deduce that \overline{x}=\textit{O}(\tau^{\alpha}) and u^{\prime}=\textit{O}(\tau^{\alpha-1}) as \tau\rightarrow 0^{+} for some constant \alpha , to be determined. Explicitly, the self-similar ansatz is then given by u^{\prime}=\tau^{\alpha-1}F(\xi), (5) where \xi:=\overline{x}\tau^{-\alpha} is the similarity coordinate. The ansatz ( 5 ) may be substituted into ( 4 ) to give an ordinary differential equation

Read the complete text