Contact discontinuities for 2-D isentropic Euler are unique in 1-D but wildly non-unique otherwise

TL;DR

This paper demonstrates that contact discontinuities in 2-D isentropic Euler equations are non-unique in certain conditions, contrasting with their uniqueness in 1-D, by developing a computational framework that uses pressure law flexibility.

Contribution

The authors introduce a novel computational approach leveraging pressure law degrees of freedom to establish non-uniqueness of contact discontinuities in 2-D Euler systems.

Findings

Non-uniqueness of contact discontinuities in 2-D Euler equations.

Existence of a smooth pressure law ensuring non-uniqueness.

Uniqueness of classical 1-D contact discontinuity solutions within bounded weak solutions.

Abstract

We develop a general framework for studying non-uniqueness of the Riemann problem for the isentropic compressible Euler system in two spatial dimensions, and in this paper we present the most delicate result of our method: non-uniqueness of the contact discontinuity. Our approach is computational, and uses the pressure law as an additional degree of freedom. The stability of the contact discontinuities for this system is a major open problem (see Gui-Qiang Chen and Ya-Guang Wang [Nonlinear partial differential equations, volume 7 of Abel Symposia. Springer, Heidelberg, 2012.]). We find a smooth pressure law $p$, verifying the physically relevant condition $p'>0$, such that for the isentropic compressible Euler system with this pressure law, contact discontinuity initial data is wildly non-unique in the class of bounded, admissible weak solutions. This result resolves the question of uniqueness for contact discontinuity solutions in the compressible regime. Moreover, in the same regularity class in which we have non-uniqueness of the contact discontinuity, i.e. $L^\infty$, with no $BV$ regularity or self-similarity, we show that the classical contact discontinuity solution to the two-dimensional isentropic compressible Euler system is in fact unique in the class of bounded, admissible weak solutions if we restrict to 1-D solutions.