Mathematical analysis of a thermodynamically consistent reduced model for iron corrosion (Q6046150)

The authors propose a reduced model accounting for corrosion of iron, in which ferric cations and electrons evolve in a fixed oxide layer subject to a self-consistent electrostatic potential (a diffusion Poisson coupled model) written as: \(\partial _{t}u_{1}+\partial _{x}J_{1}=0\), \(\partial _{t}u_{2}+\partial _{x}J_{2}=0\), \(-\lambda ^{2}\partial _{xx}^{2}v_{0}=z_{1}u_{1}+z_{2}u_{2}+\rho _{hl}\) in \((0,1)\), where \(u_{1}\) is the density of ferric cations, \(u_{2}\) the density of electrons, \(v_{0}\) the electric potential, \(J_{1}=-d_{1}(\partial _{x}u_{1}+z_{1}u_{1}\partial _{x}v_{0})\), \(J_{2}=-d_{2}(\partial _{x}u_{2}+z_{2}u_{2}\partial _{x}v_{0})\), \(\lambda ^{2}\) the scaled Debye length, and \(\rho _{hl}\) the net charge density of the ionic species in the host lattice (which is constant). Boundary conditions are added for \(J_{1}\), \(J_{2}\) and \(v_{0}\)\ at 0 and 1, and initial conditions are given for \(u_{1}\) and \(u_{2}\). After some change of variables, the authors consider the problem \(\partial _{t}u_{i}+\partial _{x}J_{i}=0\), \(i=1,2\), \(-\lambda ^{2}\partial _{xx}^{2}v_{0}=u_{0}\), in \( (0,+\infty )\times (0,1)\), with \(J_{i}=-\sigma _{i}(v_{i})\partial _{x}\xi _{i}\), \(\xi _{i}=z_{i}v_{0}+v_{i}\), \(u0=z_{1}u_{1}+z_{2}u_{2}+\rho _{hl}\). Boundary conditions are imposed on \((0,+\infty )\times \{0,1\}\) and initial conditions are imposed on \(u_{0}\), \(u_{1}\) and \(u_{2}\). The authors introduce an abstract formulation of this problem as: Find \((u,v)\) such that \(u\in H_{loc}^{1}(\mathbb{R}_{+};V^{\ast })\), \(v\in L_{loc}^{2}(\mathbb{R} _{+};V^{\ast })\cap L_{loc}^{\infty }(\mathbb{R}_{+}\times \lbrack 0,1])\), \( u^{\prime }(t)+A(v(t),v(t))=0\), \(u(t)=Ev(t)\), for a.e. \(t\in \mathbb{R}_{+}\) , and \(u(0)=u^{in}\), where \(V=H^{1}(0,1)\times H^{1}(0,1)\times H^{1}(0,1)\), \(A\) is an operator: \(H\times V\rightarrow V^{\ast }\) and \(E\) an operator: \( W\rightarrow H^{\ast }\), with \(H=H^{1}(0,1)\times L^{2}(0,1)\times L^{2}(0,1) \), \(W=H^{1}(0,1)\times L^{2}(0,1)\times L^{\infty }(0,1)\). The main result of the paper proves the existence of a solution to this last problem, under appropriate hypotheses on the data. For the proof, the authors first introduce a regularized problem, cutting off the nonlinearities applied to the chemical potentials \(v_{1},v_{2}\) at a certain level \(M\), through the introduction of a truncation operator. They also modify the boundary conditions. They finally decompose \((0,T)\), \(T<+\infty \), in successive intervals of length \(T/n\). Using results concerning monotone operators, they prove the existence of a solution to the modified problem for any \(M>0\) and for every \(n\geq 1\). They prove uniform estimates on this solution which allow to pass to the limit when \(n\rightarrow \infty \). Applying the Moser-Alikakos iteration technique, they finally prove estimates on the limit determined in the previous step, which are independent on \(M\). The paper ends with some remarks concerning possible extensions of the problem.