Dynamics of contact points in 2D Boussinesq flow

Yunrui Zheng · 2025

Read Dynamics of contact points in 2D Boussinesq flow by Yunrui Zheng (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-06127.txt.

Introduction

Yunrui Zheng studies moving contact points in a heated fluid in a 2025 solo maths paper. The flow is a 2D Boussinesq flow in an open-top vessel. Fluid meets air at a free top surface and solid walls elsewhere. Gravity and surface tension both shape the motion.

Zheng builds a global theory for data that starts near rest. Solutions exist for all time in that near-rest regime. They also settle back to rest at an exponential rate. The proof uses energy methods in a nonlinear frame. Key steps solve a convected heat equation and bound the geometry. The work builds on an energy-dissipation view from prior research.

Catalog record

CREATORYunrui Zheng
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE202 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-06127

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Dynamics of contact points in 2D Boussinesq flow Thanks: Y. Zheng was supported by National Natural Science Foundation of China with Grant No. 11901350 and No. 12371172. Yunrui Zheng Abstract . We consider the evolution of contact lines for thermal convection of viscous fluids in a 2D open-top vessel. The domain is bounded above by a free moving boundary and otherwise by the solid wall of a vessel. The dynamics of the fluid are governed by the incompressible Boussinesq approximation under the influence of gravity, and the interface between fluid and air is under the effect of capillary forces. Motivated by energy-dissipation structure in [Guo-Tice, J. Eur. Math. Soc, 2024], we develop global well posedness theory in the framework of nonlinear energy methods for the initial data sufficiently close to equilibrium. Moreover, the solutions decay to equilibrium at an exponential rate. Our methods are mainly based on the construction of solutions to convected heat equation and a priori estimates of a geometric formulation of the Boussinesq equations. 2020 Mathematics Subject Classification Primary 35Q30, 35R35, 76D45; Secondary 35B40, 76E17, 76E99 1. Introduction 1.1. Formulation of the Problem We now consider the dynamical cooling process of hot water or coffee contained in 2D cup with adiabatic boundary. A 2 D open top vessel is treated as a bounded, connected open set \mathcal{V}\subseteq\mathbb{R}^{2} which consists of two disjoint sections, i.e., \mathcal{V}=\mathcal{V}_{top}\cup\mathcal{V}_{bot} . We refer to Figure 1 for an example. \mathcal{V}_{top} \mathcal{V}_{bot} Figure 1. A vessel \mathcal{V} . \Omega(t) \Sigma(t) \Sigma_{s}(t) Figure 2. The domain \Omega(t) . Assume that the “top” part \mathcal{V}_{top} is a rectangle channel \mathcal{V}_{top}:=\mathcal{V}\cap\mathbb{R}^{2}_{+}=\{y\in\mathbb{R}^{2}:-\ell<y_{1}<\ell,0\leq y_{2}<L\} for some \ell , L>0 , where \mathbb{R}^{2}_{+} is the upper half plane \mathbb{R}^{2}_{+}=\{y\in\mathbb{R}^{2}:y_{2}\geq 0\} . Similarly, we write the “bottom” part as \mathcal{V}_{bot}:=\mathcal{V}\cap\mathbb{R}^{2}_{-}=\mathcal{V}\cap\{y\in\mathbb{R}^{2}:y_{2}\leq 0\}. Clearly, I=(-\ell,\ell)\times\{0\} . In addition, we also assume that the boundary \partial\mathcal{V} of \mathcal{V} is C^{2} away from the points (\pm\ell,L) . Now we consider a viscous incompressible fluid filling \mathcal{V}_{bot} entirely and \mathcal{V}_{top} partially. More precisely, we assume that the fluid occupies the domain \Omega(t) , \Omega(t):=\mathcal{V}_{bot}\cup\big\{y\in\mathbb{R}^{2}:-\ell<y_{1}<\ell,\ 0<y_{2}<\zeta(y_{1},t)\big\}, where the free surface \zeta(y_{1},t) is assumed to be a graph of the function \zeta:[-\ell,\ell]\times\mathbb{R}_{+}\rightarrow\mathbb{R} satisfying 0<\zeta(\pm\ell,t)\leq L for all t\in\mathbb{R}_{+} , which means the fluid does not spill out of the top domain. For simplicity, we write the free surface as \Sigma(t)=\{y_{2}=\zeta(y_{1},t)\} and the interface between fluid and solid as \Sigma_{s}(t)=\partial\Omega(t)\backslash\Sigma(t) . We refer to Figure 2 for the description of domain. For each t>0 , the fluid is described by its velocity, pressure and temperature (U,P,\Theta):\Omega(t)\rightarrow\mathbb{R}^{2}\times\mathbb{R} which is governed by the incompressible Boussinesq approximation [ 4 ] : (1.1) \left\{\begin{aligned} &\partial_{t}U+U\cdot\nabla U+\nabla P-\mu\Delta U=-g\Theta e_{2}\quad&\text{in}&\ \Omega(t),\\ &\mathop{\rm div}\nolimits U=0\quad&\text{in}&\ \Omega(t),\\ &\partial_{t}\Theta+U\cdot\nabla\Theta-k\Delta\Theta=0\quad&\text{in}&\ \Omega(t),\end{aligned}\right. where \mu>0 is the coefficient of viscosity, k>0 is the coefficient of heat conduction, g>0 is the acceleration of gravity and P=\bar{P}+gy_{2}-P_{atm} by adjusting from the gravitational force and the constant atmospheric pressure P_{atm} to the actual pressure \bar{P} . From the physical viewpoint, the fluid are supposed to satisfy some boundary conditions. On the free surface, the fluid should be kinematic [ 28 ] : (1.2) \left\{\begin{aligned} &S(P,U)\nu=g\zeta\nu-\sigma\mathcal{H}(\zeta)\nu\quad&\text{on}&\ \Sigma(t),\\ &\partial_{t}\zeta=U\cdot\nu\quad&\text{on}&\ \Sigma(t),\end{aligned}\right. where S(P,U) is the viscous stress tensor S(P,U):=PI-\mu\mathbb{D}U , with I the 2\times 2 identity matrix, and \mathbb{D}U=\nabla U+\nabla^{T}U the symmetric gradient of U . Here, \mathcal{H}(\zeta)=\partial_{1}\left(\frac{\partial_{1}\zeta}{(1+|\partial_{1}\zeta|^{2})^{1/2}}\right) is twice the mean curvature of the free surface. \sigma=\sigma_{1}-\sigma_{2}\Theta is the coefficient of surface tension, with \sigma_{1}>0 and \sigma_{2}\in\mathbb{R} . On the fixed boundary, the fluid satisfies the Navier-slip condition: (1.3) \left\{\begin{aligned} &\big(S(P,U)\nu-\beta U\big)\cdot\tau=0\quad&\text{on}&\ \Sigma_{s}(t),\\ &U\cdot\nu=0\quad&\text{on}&\ \Sigma_{s}(t).\end{aligned}\right. Without loss of generality, we assume that the temperature of atmosphere around the fluid is 0 . Due to the adiabatic solid boundary, we assume that \Theta=0 on \Sigma_{s}(t) . The heat on the free surface \Sigma(t) satisfies the Newton’s cooling law due to the heat conduction between fluid and air: (1.4) k\nu\cdot\nabla\Theta+\Theta=0. The fluid on the contact points satisfies the condition: (1.5) \partial_{t}\zeta(\pm\ell,t)=\mathscr{V}\left([\![\gamma]\!]\mp\sigma\frac{\partial_{1}\zeta}{(1+|\partial_{1}\zeta|^{2})^{1/2}}(\pm\ell,t)\right), where \mathscr{V}:\mathbb{R}\rightarrow\mathbb{R} is the contact point velocity response function which is a C^{2} increasing diffeomorphism satisfying \mathscr{V}(0)=0 . [\![\gamma]\!]:=\gamma_{sv}-\gamma_{sf} for \gamma_{sv} , \gamma_{sf}\in\mathbb{R} , where \gamma_{sv} , \gamma_{sf} are measures of the free-energy per unit length with respect to the solid-vapor and solid-fluid intersection. We also assume that the Young relation |[\![\gamma]\!]|<\sigma ( [ 29 ] ), which is necessary for the equilibrium state we will consider in the following. For convenience, we introduce the inverse function \mathscr{W} of \mathscr{V} and rewrite ( 1.5 ) as (1.6) \mathscr{W}\big(\partial_{t}\zeta(\pm\ell,t)\big)=[\![\gamma]\!]\mp\sigma\frac{\partial_{1}\zeta}{(1+|\partial_{1}\zeta|^{2})^{1/2}}(\pm\ell,t), which is the dynamic law at contact points. The contact points in 2D or contact lines in 3D fluids, with temperature absent, has been studied for a long time, and we may refer to [ 6 , de Gennes] for an overview. There are so many papers dealing with free boundary problems of contact points or contact lines, that we have to mention some works closely related to our background settings. For the contact lines in inviscid flow, we refer to [ 7 , T. de Poyferré] and [ 20 , Ming-Wang] for the breakthrough of water wave problem. The analysis for Navier-slip conditions with contact angle can be referred to [ 21 , Ren-E] and [ 5 , Cox] . When the contact angle is fixed by \frac{\pi}{2} , the well-posedness of 2D contact points problem in the Navier-Stokes flow is studied in [ 22 , Schweizer] and the 3D contact line problem with an additional periodic direction is studied in [ 2 , Bodea] . For the non-stationary problem, the well-posedness and stability for 2D Navier-stokes flow are studied by [ 11 , Guo-Tice] and [ 12 , Guo-Tice-Wu-Zheng] . There are also many results for the stationary problem of contact points or lines. The solvability of stationary Navier-Stokes problem for dynamic contact angles and moving contact lines is studied by [ 17 , Kröner] and [ 23 , Socolosky] , respectively. The existence results of the Navier-Stokes equations for both static and dynamic contact points and lines in weighted Hölder spaces are proved by [ 24 , 25 , Solonnikov] . The 3D contact lines for stationary Navier-Stokes equations with fixed angle \frac{\pi}{2} is studied by [ 14 , Jin] . The local well-posedness and stability for the 2D contact points in the Stokes flow are studied by [ 10 , Guo-Tice] and [ 27 , Tice-Zheng] The contact angles are also closely related to the droplet problem. The stability problem for droplets governed by the Stokes equations is studied in [ 26 , Tice-Wu] . The well-posedness for droplets governed by Darcy’s law is studied in [ 18 , Knüpfer-Masmoudi] . Recently, the well-posedness of the Stokes equations on a wedge in weighted Sobolve spaces with Navier-slip boundary conditions is studied by [ 3 , Bravin-Gnann-Knüpfer-Masmoudi-Roodenburg-Sauer] . To the authors’ best knowledge, there is still no well-posedness and stability for the dynamic contact points problem with temperature at present. Our aim is to establish the first result for the global well-posedness for the 2D Boussinesq contact line problem. By referring to [ 10 , Guo-Tice] or [ 8 , Finn] , we known that ( 1.1 ), under the condition of Young relation, admits an equilibrium for U=0 , P=P_{0} a constant, \Theta=0 and \zeta(y_{1},t)=\zeta_{0}(y_{1}) satisfies (1.7) P_{0}=g\zeta_{0}-\sigma_{1}\mathcal{H}(\zeta_{0})\quad\text{on}\ (-\ell,\ell),\quad\sigma_{1}\frac{\partial_{1}\zeta_{0}}{\sqrt{1+|\partial_{1}\zeta_{0}|^{2}}}(\pm\ell)=\pm[\![\gamma]\!]. This equilibrium would allow

Read the complete text