On a local solvability of the contact Muskat problem

Nataliya Vasylyeva
2024-11-22
Abstract:In the paper, we discuss the two-dimensional contact Muskat problem with zero surface tension of a free boundary. The initial shape of the unknown interface is a smooth simple curve which forms acute corners $\delta_{0}$ and $\delta_{1}$ with fixed boundaries. Under suitable assumptions on the given data, the one-to-one local classical solvability of this problem is proved. We also describe the sufficient conditions on the data in the model which provide the existence of the "waiting time" phenomenon.
Analysis of PDEs
What problem does this paper attempt to address?
This paper attempts to solve the local solvability of the Contact Muskat Problem under the condition of zero surface tension. Specifically, the two - dimensional Contact Muskat Problem is studied, where the initial shape of the unknown interface is a smooth simple curve that forms acute angles \(\delta_0\) and \(\delta_1\) with the fixed boundary. The goal of the paper is to prove that under appropriate assumptions of the given data, the problem is solvable in the one - dimensional local classical sense and to describe the sufficient conditions for the model data that provide the existence of the "waiting time" phenomenon. ### Overview of the Main Problem #### 1. Research Background The Muskat problem was proposed by Morris Muskat in 1934 to describe the viscous displacement of a two - phase fluid system in a two - dimensional porous medium. This problem describes the process of water invading oil sands and involves the phenomenon of increasing the reservoir pressure by water injection and driving the oil to the production well during the secondary oil recovery process. The fluid motion follows Darcy's law, and the boundary between the two liquid phases is an unknown interface to be determined, namely the so - called free boundary. #### 2. Specific Problem This paper focuses on the two - dimensional Contact Muskat Problem under the condition of zero surface tension. Let \(\Omega\subset\mathbb{R}^2\) be a bounded region, and its boundary \(\partial\Omega\in C^{2 +\beta}\), \(\beta\in(0,1)\). For each \(\tau\in[0,T]\), a simple curve \(\Gamma(\tau)\subseteq\overline{\Omega}\) divides the region \(\Omega\) into two sub - domains \(\Omega_1(\tau)\) and \(\Omega_2(\tau)\) such that: \[ \Omega=\Omega_1(\tau)\cup\Omega_2(\tau)\cup\Gamma(\tau),\quad\Omega_1(\tau)\cap\Omega_2(\tau)=\varnothing,\quad\partial\Omega_1(\tau)\cap\partial\Omega_2(\tau)=\Gamma(\tau) \] Furthermore, \(\Gamma(\tau)\) intersects with two points \(A_0\) and \(A_1\) on \(\partial\Omega\), that is: \[ \partial\Omega\cap\Gamma(\tau)=\{A_0,A_1\} \] #### 3. Mathematical Model Given positive numbers \(T,k_1,k_2\) and \(k_1\neq k_2\), we need to solve the unknown liquid pressures \(P_1 = P_1(y_1,y_2,\tau)\) and \(P_2 = P_2(y_1,y_2,\tau)\) and the unknown interface \(\Gamma(\tau)\), satisfying the following system: \[ \begin{cases} \Delta_y P_i = 0&\text{in }\Omega_i(\tau),\tau\in(0,T),i = 1,2,\\ P_1 - P_2 = 0&\text{on }\Gamma(\tau),\tau\in[0,T],\\ V_n=-k_1\frac{\partial P_1}{\partial n_\tau}=-k_2\frac{\partial P_2}{\partial n_\tau}&\text{on }\Gamma(\tau),\tau\in[0,T],\\ P_1 = p_1(y)&\text{on } \]