Reduction of nonlinear PDEs to linear equations (Q2569627)

This interesting note concerns linear and quasilinear elliptic second order differential equations in divergence form. One of the results is for example the following. Let \(\Omega\) be a bounded open subset of \({\mathbb R}^2\), \(A=A(x,\xi)\) such that \[ |\xi|^2+|A(x,\xi)|^2\leq(K+1/K)\langle A(x,\xi),\xi\rangle \] for a.e. \(x\in\Omega\) and any \(\xi\in{\mathbb R}^2\). Let \(u\in W^{1,2}_{loc}(\Omega)\) satisfy the quasilinear equation \[ \text{div}A(x,\nabla u)=0. \] Then there exists a unique symmetric matrix \({\mathcal A}\) with \(\det\mathcal A(x)=1\) a.e. and \(|\xi|^2/K\leq\langle A(x,\xi),\xi\rangle\leq K|\xi|^2\) for a.e \(x\in \Omega\) and every \(\xi\in\mathbb R\) such that \[ \text{div}(\mathcal A(x)\nabla u)=0. \] The paper has also connections with \(G\)-convergence and variational integrals.