Dissipative solutions for 2D Onsager's conjecture

Abstract

In 1949, L. Onsager conjectured that weak solutions of the Euler equations in $C_{t,x}^{\gamma }$ are energy conservative when $\gamma >\frac{1}{3}$, whereas nonconservative solutions exist when $\gamma <\frac{1}{3}$. Constantin, E and Titi [13] proved the conservation part, while Isett [26] constructed nontrivial $C_{t,x}^{\gamma }$ weak solutions with compact support in time in three dimensions for each $0<\gamma <\frac{1}{3}$. Building on this breakthrough work, Buckmaster, De Lellis, Székelyhidi Jr. and Vicol [7] obtained strictly dissipative solutions below the Onsager threshold. In two dimensions, Giri and Radu [23] first constructed a nontrivial weak solution in $C^{\frac{1}{3}-}({\mathbb{T}}^2\times [0,T])$ and left an open problem whether it would be possible to construct dissipative solutions (controllable energy) below $\frac{1}{3}$ threshold. In this paper, we reslove this question and thereby obtain infinitely many strictly dissipative weak solutions in $C^{\gamma }({\mathbb{T}}^2\times [0,T])$ for every $0<\gamma <\frac{1}{3}$. Moreover, we prove that the associated wild initial data are dense in the divergence-free subspace of $C^{{\gamma }^{\prime }}({\mathbb{T}}^2)$ for any $0<{\gamma }^{\prime }<\frac{1}{3}$. The key new ingredient is a family of staircase traveling waves, which we combine with a Picard iteration coupled to the Newton--Nash scheme to control both the Reynolds stress and the kinetic energy. The construction also applies in every dimension $d\ge 2$.

arXiv commentWe made the necessary modifications and added acknowledgments