Quadratic convergence conjecture for planar diagonal diffusion matrices with constant trace

About 4 years old · traced to

Let Ω\Omega be the domain and let uεu^{\varepsilon} and uu denote, respectively, the solution of the rapidly oscillating elliptic equation and its homogenized limit, with data f,gf,g as in the paper. Assume that A∈C2(T2;S+2)A\in C^2(\mathbb{T}^2;\mathcal{S}^2_+) has the form

A(y)=diag⁡(a1(y),a2(y)),y∈R2,A(y)=\operatorname{diag}(a_1(y),a_2(y)),\qquad y\in\mathbb{R}^2,

where a1,a2∈C2(T2;(0,1))a_1,a_2\in C^2(\mathbb{T}^2;(0,1)) and a1+a2≡1a_1+a_2\equiv 1. Quadratic convergence conjecture. For any choice of f,gf,g,

∥uε−u∥L∞(Ω)=O(ε2)as ε↘0.\left\|u^{\varepsilon}-u\right\|_{L^{\infty}(\Omega)}=\mathcal{O}(\varepsilon^2)\qquad\text{as }\varepsilon\searrow 0.

This conjecture predicts the unexpectedly fast homogenization rate for two-dimensional diagonal diffusion matrices of constant trace; the surrounding discussion notes that the analogous behavior is not observed in dimensions n≥3n\geq 3.

References

Primary source

Xiaoqin Guo, Timo Sprekeler and Hung V. Tran, “Characterizations of diffusion matrices in homogenization of elliptic equations in nondivergence-form”, arXiv:2201.01974 (2022).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.