Oscillations of a stratified liquid partially covered with ice (general case) (Q2191945)

The author here extends previous results he obtained concerning the existence of small motions for an ideal and stratified liquid which fills a bounded domain \(\Omega \) of \(\mathbb{R}^{3}\) and whose density at rest is a function \(\rho _{0}\) of \(x_{3}\) only. The domain \(\Omega \) is divided in three regions: clear water, elastic ice and crumbled ice. The author starts writing the equation of small motions in \(\Omega \) for the velocity field \(\overrightarrow{u}\) of the liquid as \(\frac{\partial \overrightarrow{u}}{\partial t}=\rho _{0}^{-1}(x_{3})(-\nabla p-g\rho \overrightarrow{e}_{3})+f(t,x)\), \(\operatorname{div}(\overrightarrow{u})=0\), \(\frac{\partial \rho }{\partial t}+\nabla \rho \cdot \overrightarrow{u}=0\), where \(p\) is the deviation of the pressure from the equilibrium pressure and \(\rho \) is the deviation from the density \(\rho _{0}\). Different boundary conditions are imposed on the different pieces of the boundary. He then introduces \(\overrightarrow{v}\) through \(\frac{\partial \overrightarrow{v}}{\partial t}=\overrightarrow{u}\), \(div(\overrightarrow{v})=0\), in \(\Omega \), and he writes the problem satisfied by \(\overrightarrow{v}\) and \(p\). He then introduces \(\overrightarrow{w}=\overrightarrow{v}-\rho _{0}^{-1}\nabla \Phi \) with \(\overrightarrow{w}\in \overrightarrow{J}_{0}(\Omega ,\rho _{0})=\{ \overrightarrow{v}\in \overrightarrow{L}_{2}(\Omega ,\rho _{0}):\operatorname{div}\overrightarrow{v}=0\) in \(\Omega \), \(v_{n}=0\) on \(\partial \Omega \}\) and \(\rho _{0}^{-1}\nabla \Phi \in \overrightarrow{G}_{h,S}(\Omega ,\rho _{0})=\{\overrightarrow{v}\in \overrightarrow{L}_{2}(\Omega ,\rho _{0}):\overrightarrow{v}=\rho _{0}^{-1}\nabla p\) in \(\Omega \), \(\operatorname{div}\overrightarrow{v}=0\) in \(\Omega \), \(v_{n}=0\) on \(S\), \(\int_{\Gamma}pd\Gamma =0\}\) and he writes the problem satisfied by \(\overrightarrow{w}\) and \(\Phi \) or \(\rho _{0}^{-1}\nabla \Phi \). He finally writes the last problem in an abstract formulation \(\frac{d^{2}}{dt^{2}}\left(\begin{array}{cc}I_{0} & 0 \\ 0 & A_{C}\end{array}\right) \left(\begin{array}{c}\overrightarrow{w} \\ u\end{array}\right) +\left(\begin{array}{cc}B_{11} & B_{12} \\ B_{21} & B_{22}+N\end{array}\right) \left(\begin{array}{c} \overrightarrow{w} \\ u\end{array}\right) =\left(\begin{array}{c}P_{0}\psi _{0} \\ f\end{array}\right) \), where \(u=(u_{0}, \widehat{u}_{1},u_{2})\) is deduced from \(\rho_{0}^{-1}\nabla \Phi \) and from initial conditions, \(I_{0}\) is the identity operator on \(\overrightarrow{J}_{0}(\Omega ,\rho _{0})\), \(A_{C}\), \(B_{ij}\) and \(N\) are linear operators which satisfy appropriate hypotheses. Initial conditions are added. The author first proves that the initial boundary value problem is equivalent to the preceding Cauchy problem posed in the Hilbert space \(\mathcal{H}=\overrightarrow{J}_{0}(\Omega ,\rho _{0})\oplus H\), where \(H\) is a Hilbert space associated to \(u\). The main result of the paper proves the existence of a unique solution to the Cauchy problem which is strong with respect to the time. For the proof, the author establishes properties of the different operators involved in the preceding abstract formulation and he quotes tools from his previous paper [Vestn. Udmurt. Univ., Mat. Mekh. Komp'yut. Nauki 28, No. 3, 328--347 (2018; Zbl 1421.35293)].