Contact in fluid-plate interaction: formation and detachment
Srđan Trifunović 1 · 2025
Read Contact in fluid-plate interaction: formation and detachment by Srđan Trifunović 1 (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-2505-10508.txt.
Introduction
Srdan Trifunovic, writing alone, studies what happens when an elastic plate floating on fluid touches the floor beneath it, in a 2025 mathematics paper. The setup models structures that ride on a layer of compressible, viscous fluid. His results show new contact can form only on negligible spots. All contact must lift within finite time unless the load pushes down too hard. A strong local upward force can also peel any contact point free on demand.
Catalog record
| CREATOR | Srđan Trifunović 1 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 112 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-10508
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Contact in fluid-plate interaction: formation and detachment Thanks: 1 Department of Mathematics and Informatics, Faculty of Sciences, University of Novi Sad, Serbia. e-mail: [email protected] Srđan Trifunović 1 Abstract . In this paper, we study an interaction problem between an elastic plate and a compressible viscous fluid located between the rigid bottom z = 0 z=0 and the plate. First, by utilizing the vertical fluid dissipation, we show that ln η ( t ) ∈ L 1 \ln\eta(t)\in L^{1} for any t > 0 t>0 provided that ln η 0 ∈ L 1 \ln\eta_{0}\in L^{1} , ensuring that additional plate contact can form only on a set of a measure zero. Then, by utilizing the expanding capability of compressible fluid pressure, we show that all contact has to detach in finite time provided that the source force acting onto the plate is not pushing down excessively. Finally, we show that contact at any point can be detached in any given time with a strong enough source force localized around that point which is pulling the plate up. This is the first result where detachment of contact is proven in fluid-structure interaction. Dedicated to Šárka Nečasová for her 60th birthday Keywords and phrases: fluid-structure interaction, contact problem, post-contact dynamics, compressible Navier-Stokes AMS Mathematical Subject classification (2020): 74M15 (Primary), 74F10, 76N06 (Secondary) 1. The model Here, we study an interaction problem between a compressible viscous fluid and an elastic plate. The plate moves only vertically and it is located above the rigid bottom \{z=0\} . Therefore, it is described by non-negative scalar displacement \eta:G_{T}\to\mathbb{R}_{0}^{+} , where G_{T}:=(0,T)\times\Gamma and \Gamma is a 2D torus. The displacement \eta satisfies the linear plate equation on G_{T} : (1) \displaystyle\eta_{tt}+\Delta^{2}\eta=-S^{\eta}\mathbf{f}_{fl}\cdot\mathbf{e}_{z}+F, where \mathbf{f}_{fl} is the force with which fluid acts onto the plate, \mathbf{e}_{z}:=(0,0,1) , S^{\eta}:=\sqrt{1+|\nabla\eta|^{2}} is the Jacobian of the transformation from the Eulerian to the Lagrangian coordinates of the plate, and F is the outer force. The graph of \eta(t) is denoted as \displaystyle\Gamma^{\eta}(t):=\{(x,y,\eta(t,x,y)):(x,y)\in\Gamma\},\quad t\in[0,T]. The fluid fills the domain at time t (see figure 1 ) \displaystyle\Omega^{\eta}(t):=\{(x,y,z):(x,y)\in\Gamma,0<z<\eta(t,x,y)\},\quad t\in[0,T], while the time-space cylinder is denoted as \displaystyle Q_{T}^{\eta}:=\bigcup_{t\in(0,T)}\{t\}\times\Omega^{\eta}(t). The fluid is described by the density \rho:Q_{T}^{\eta}\to\mathbb{R}_{0}^{+} and velocity \mathbf{u}=(u_{1},u_{2},u_{3}):Q_{T}^{\eta}\to\mathbb{R}^{3} and it is governed by the compressible Navier-Stokes equations on Q_{T}^{\eta} : (2) \begin{split}\partial_{t}\rho+\nabla\cdot(\rho\mathbf{u})&=0,\\ \partial_{t}(\rho\mathbf{u})+\nabla\cdot(\rho\mathbf{u}\otimes\mathbf{u})&=-\nabla p(\rho)+\nabla\cdot\mathbb{S}(\nabla\mathbf{u}),\end{split} where \displaystyle\mathbb{S}(\nabla\mathbf{u}):=\mu\left(\nabla\mathbf{u}+\nabla^{\tau}\mathbf{u}-\frac{2}{3}\nabla\cdot\mathbf{u}\mathbb{I}\right)+\lambda\nabla\cdot\mathbf{u}\mathbb{I},\quad\mu>0,~\lambda\geq 0, \mathbb{I} is the 3\times 3 identity matrix, and the pressure p is set to be p(\rho)=\rho^{\gamma}. The fluid velocity satisfies the no-slip condition on the rigid bottom \{z=0\} : (3) \displaystyle\mathbf{u}(t,x,y,0)=0,\quad t\in(0,T),~(x,y)\in\Gamma. Finally, the following kinematic and dynamic coupling conditions are assumed on G_{T} : (4) \displaystyle\partial_{t}\eta(t,x,y)\mathbf{e}_{z} \displaystyle= \displaystyle\mathbf{u}(t,x,y,\eta(t,x,y)), (5) \displaystyle\mathbf{f}_{fl}(t,x,y) \displaystyle= \displaystyle\big[(-p(\rho)\mathbb{I}+\mathbb{S}(\nabla\mathbf{u}))\nu^{\eta}\big](t,x,y,\eta(t,x,y)), where \nu^{\eta}=(-\nabla\eta,1)/\sqrt{1+{|\nabla\eta|^{2}}} is the unit outward normal vector on the graph \Gamma^{\eta} . Figure 1. An example of the fluid domain \Omega^{\eta}(t) located between the rigid bottom z=0 and elastic plate z=\eta(t,x,y) . 2. Weak formulation and main results First, the problem at hand allows contact which a priori induces irregular shapes of the fluid domain and consequently we cannot expect solutions to be regular. Indeed, this has been observed in various models (see for example [ 43 ] ). Moreover, any sufficiently regular solution cannot have contact as it was shown in [ 6 ] . Therefore, the only reasonable framework is the one of weak solutions. We start with introducing some concepts well-known in fluid-structure interaction literature. First, since the fluid domain can degenerate into multiple disconnected regions at any time (even possibly into continuum many of them as it was observed in [ 28 , Section 3.1] ), some notions which are standard in fluid-structure interaction analysis have to be carefully redefined. Therefore, for a given \eta(t)\in H^{2}(\Gamma) such that \eta(t)>0 a.e. on \Gamma , we start by introducing: \displaystyle H_{0z}^{1}(\Omega^{\eta}(t)):=\{f\in H^{1}(\Omega^{\eta}(t)):\gamma_{|O\times\{0\}}f=0\text{ for every closed set }O\subset\{\eta(t)>0\}\}, where H^{1}(\Omega^{\eta}(t))\equiv W^{1,1}(\Omega^{\eta}(t)) is the standard Sobolev space, \gamma_{|O\times\{0\}}f is the trace of f on O\times\{0\} , and \{\eta(t)>0\}:=\{(x,y)\in\Gamma:\eta(t,x,y)>0\} . Next, given f\in H_{0z}^{1}(\Omega^{\eta}(t)) , the Lagrangian trace operator is defined as \displaystyle\gamma_{|\Gamma^{\eta}(t)}f:=\hat{\gamma}_{|\Gamma^{\eta}(t)}e_{0}[f], where e_{0}[f] is the extension by zero to \Gamma\times(-1,0) which preserves the norm \|f\|_{H^{1}(\Omega^{\eta}(t))}=\|e_{0}[f]\|_{H^{1}(\Omega^{\eta}(t)\cup\Gamma\times(-1,0))} by Lemma A.2 , while \hat{\gamma}_{|\Gamma^{\eta}(t)} is the standard Lagrangian trace defined as \displaystyle\hat{\gamma}_{|\Gamma^{\eta}(t)}g(x,y):=g(x,y,\eta(t,x,y)),\quad(x,y)\in\Gamma, for any g\in C^{1}(\overline{\Omega^{\eta}(t)\cup\Gamma\times(-1,0)}) and then extended to a continuous linear operator from H^{1}(\Omega^{\eta}(t)\cup\Gamma\times(-1,0)) to H^{s}(\Gamma) for any s\in(0,1/2) (see [ 38 ] ). For time-dependent functions, trace will be denoted as \displaystyle\left(\gamma_{|\Gamma^{\eta}}f\right)(t)=\gamma_{|\Gamma^{\eta}(t)}f(t). Finally, we define the standard Bochner spaces corresponding to the moving domains: \displaystyle L^{p}(0,T;L^{q}(\Omega^{\eta}(t))):= \displaystyle\{f\in L^{1}(Q_{T}^{\eta}): \displaystyle f(t)\in L^{q}(\Omega^{\eta}(t))\text{ for a.a }t\in(0,T),~\|f(t)\|_{L^{q}(\Omega^{\eta}(t))}\in L^{p}(0,T)\}, with the norms (note that " (t) " is omitted in the norm for simplicity) \displaystyle\|f\|_{L^{p}(0,T;L^{q}(\Omega^{\eta}))} \displaystyle:= \displaystyle\left(\int_{0}^{T}\|f(t)\|_{L^{q}(\Omega^{\eta}(t))}^{p}\right)^{\frac{1}{p}},\quad p\in[1,\infty), \displaystyle\|f\|_{L^{\infty}(0,T;L^{q}(\Omega^{\eta}))} \displaystyle:= \displaystyle\esssup_{t\in(0,T)}\|f(t)\|_{L^{q}(\Omega^{\eta}(t))}, and \displaystyle L^{p}(0,T;W^{1,q}(\Omega^{\eta}(t))):=\{f\in L^{p}(0,T;L^{q}(\Omega^{\eta}(t))):\nabla f\in L^{p}(0,T;L^{q}(\Omega^{\eta}(t)))\}, with the norms \displaystyle\|f\|_{L^{p}(0,T;W^{1,q}(\Omega^{\eta}))} \displaystyle:= \displaystyle\left(\int_{0}^{T}\|f(t)\|_{W^{1,q}(\Omega^{\eta}(t))}^{p}\right)^{\frac{1}{p}},\quad p\in[1,\infty), \displaystyle\|f\|_{L^{\infty}(0,T;W^{1,q}(\Omega^{\eta}))} \displaystyle:= \displaystyle\text{ess}\sup_{t\in(0,T)}\|f(t)\|_{W^{1,q}(\Omega^{\eta}(t))}. Specially, the following space is defined \displaystyle L^{p}(0,T;H_{0z}^{1}(\Omega^{\eta}(t))):=\{f\in L^{p}(0,T;H^{1}(\Omega^{\eta})):f(t)\in H_{0z}^{1}(\Omega^{\eta}(t))\text{ for a.a. }t\in(0,T)\}. The regularity and compatibility conditions for initial data are as follows (6) \displaystyle\begin{cases}&\eta_{0}>0\text{ a.e. on }\Gamma,\quad\eta_{0}\in H^{2}(\Gamma),\quad\ln\eta_{0}\in L^{1}(\Gamma),\quad v_{0}\in L^{2}(\Gamma),\\ &\rho_{0}\in L^{\gamma}(\Omega^{\eta_{0}}),\quad(\rho\mathbf{u})_{0}\in L^{\frac{2\gamma}{\gamma+1}}(\Omega^{\eta_{0}}),\quad\frac{(\rho\mathbf{u})_{0}^{2}}{\rho_{0}}\in L^{1}(\Omega^{\eta_{0}}),\\ &\rho_{0}>0\text{ in }\{(x,y,z)\in\Omega^{\eta_{0}}:(\rho\mathbf{u})_{0}(x,y,z)>0\},\quad\int_{\Omega^{\eta_{0}}}\rho_{0}:=m>0,\quad\rho_{0}\geq 0,\end{cases} where \Omega^{\eta_{0}}:=\Omega^{\eta}(0) and m is the total mass. Note that there is a new condition on \ln\eta_{0} which allows contact initially. Let us now introduce a concept of weak solution which will be studied: Definition 2.1 ( Weak solution ) . We say that (\rho,\mathbf{u},\eta) is a weak solution to ( 1 )-( 5 ) if the initial data satisfies ( 6 ) , \gamma>3 and: (1) \int_{\Omega^{\eta}(t)}\rho(t)=m and \eta(t)>0 a.e. on \Gamma for all t\in[0,T] , and the solution is in the following regularity class: \displaystyle\rho\in L^{\infty}(0,T;L^{\gamma}(\Omega^{\eta}(t))),\quad\mathbf{u}\in L^{2}(0,T;H_{0z}^{1}(\Omega^{\eta}(t))), \displaystyle\rho|\mathbf{u}|^{2}\in L^{\infty}(0,T;L^{1}(\Omega^{\eta}(t))),\quad\frac{\rho^{\gamma}}{\eta},~\frac{\rho|u_{3}|^{2}}{\eta}\in L^{1}(Q_{T}^{\eta}), \displaystyle\eta\in W^{1,\infty}(0,T;L^{2}(\Gamma))\cap L^{\infty}(0,T;H^{2}(\Gamma)),\quad\ln\eta\in C([0,T];L^{1}(\Gamma)); (2) The kinematic coupling \gamma_{|{\Gamma}^{\eta}}\mathbf{u}=\partial_{t}\eta\mathbf{e}_{z} holds a.e. on G_{T} ; (3) The renormalized continuity equation (7) \int_{Q_{T}^{\eta}}\rho B(\rho)(\partial_{t}\varphi+\mathbf{u}\cdot\nabla\varphi)=\int_{Q_{T}^{\eta}}b(\rho)(\nabla\cdot\mathbf{u})\varphi-\int_{\Omega^{\eta_{0}}}\rho_{0}B(\rho_{0})\varphi(0) holds for all functions \varphi\in C_{c}^{\infty}([0,T)\times\mathbb{R}^{3}) and any b\in L^{\infty}(0,\infty)\cap C[0,\infty) such that b(0)=0 with B(\rho) being any primitive function to \frac{b(\rho)}{\rho^{2}} ; (4) The coupled momentum equation \displaystyle\int_{Q_{T}^{\eta}}\rho\mathbf{u}\cdot\partial_{t}\boldsymbol{\varphi}+\int_{Q_{T}^{\eta}}(\rho\mathbf{u}\otimes\mathbf{u}):\nabla\boldsymbol{\varphi}+\int_{Q_{T}^{\eta}}\rho^{\gamma}(\nabla\cdot\boldsymbol{\varphi})-\int_{Q_{T}^{\eta}}\mathbb{S}(\nabla\mathbf{u}):\nabla\boldsymbol{\varphi} \displaystyle+\int_{G_{T}}\partial_{t}\eta\partial_{t}\psi-\int_{G_{T}}\Delta\eta\Delta\psi+\int_{G_{T}}F\psi (8) \displaystyle=-\int_{\Omega^{\eta_{0}}}(\rho\mathbf{u})_{0}\cdot\boldsymbol{\varphi}(0)-\int_{\Gamma}v_{0}\psi holds for all \boldsymbol{\varphi}\in C_{c}^{\infty}([0,T)\times\mathbb{R}^{3}) and all \psi\in C_{c}^{\infty}([0,T)\times\Gamma) such that \boldsymbol{\varphi}(t,x,y,\eta(t,x))=\psi(t,x,y)\mathbf{e}_{z} on G_{T} and \boldsymbol{\varphi}=0 on \{z=0\} ; (5) The contact inequality holds for a.a. t\in(0,T) : \displaystyle\int_{0}^{t}\int_{\Gamma}F+\int_{0}^{t}\int_{\Omega^{\eta}}\frac{\rho^{\gamma}}{\eta}+\int_{0}^{t}\int_{\Omega^{\eta}}\frac{\rho|u_{3}|^{2}}{\eta}-\int_{\Gamma}\ln\eta(t) \displaystyle\leq\int_{0}^{t}\int_{\Omega^{\eta}}\rho u_{3}z\frac{\partial_{t}\eta}{\eta^{2}}+\int_{0}^{t}\int_{\Omega^{\eta}}\rho u_{1}u_{3}\frac{z\partial_{x}\eta}{\eta^{2}}+\int_{0}^{t}\int_{\Omega^{\eta}}\rho u_{2}u_{3}\frac{z\partial_{y}\eta}{\eta^{2}} \displaystyle-\frac{1}{2}\int_{0}^{t}\int_{\Omega^{\eta}}(\partial_{z}u_{1}+\partial_{x}u_{3})\frac{z\partial_{x}\eta}{\eta^{2}}-\frac{1}{2}\int_{0}^{t}\int_{\Omega^{\eta}}(\partial_{z}u_{2}+\partial_{y}u_{3})\frac{z\partial_{y}\eta}{\eta^{2}}-\int_{\Gamma}\ln\eta_{0} (9) \displaystyle\quad+\int_{\Omega^{\eta}(t)}\rho(t)u_{3}(t)\frac{z}{\eta(t)}-\int_{\Omega^{\eta_{0}}}(\rho u_{3})_{0}\frac{z}{\eta_{0}}+\int_{\Gamma}\partial_{t}\eta(t)-\int_{\Gamma}v_{0}; (6) The energy inequality holds for a.a. t\in(0,T) \displaystyle\displaystyle{\int_{\Omega^{\eta}(t)}\left(\frac{1}{2}\rho|\mathbf{u}|^{2}+\frac{\rho^{\gamma}}{\gamma-1}\right)(t)+\int_{\Gamma}\left(\frac{1}{2}|\partial_{t}\eta|^{2}+\frac{1}{2}|\Delta\eta|^{2}\right)(t)}+\int_{0}^{t}\int_{\Omega^{\eta}}\mathbb{S}(\nabla\mathbf{u}):\nabla\mathbf{u} (10) \displaystyle\leq\displaystyle{\int_{\Omega^{\eta_{0}}}\left(\frac{1}{2\rho_{0}}|(\rho\mathbf{u})_{0}|^{2}+\frac{\rho_{0}^{\gamma}}{\gamma-1}\right)}+\int_{\Gamma}\left(\frac{1}{2}|v_{0}|^{2}+\frac{1}{2}|\Delta\eta_{0}|^{2}\right)+\int_{0}^{t}\int_{\Gamma}F\partial_{t}\eta. Remark 2.1 . (1) The inequality ( 9 ) in this definition is new. It can formally be obtained from ( 8 ) by choosing \boldsymbol{\varphi}=\chi_{\{0,t\}}\frac{z}{\eta}\mathbf{e}_{z} and \psi=\chi_{\{0,t\}} , at least provided that \eta>0 , as it was done in Section 4.5.1 . It preserves crucial information about total forces acting onto the plate, even if \{\eta=0\}\neq\emptyset , which is in contrast with ( 8 ) where this information is lost. The inequality, instead of equality, comes from possible defects of \frac{\rho^{\gamma}}{\eta} , \partial_{t}(\ln\eta^{-}) and the contact force which is a result of plate not being able to pass through the rigid boundary \{z=0\} . For some applications, it is worth noting that it can be localized in time-space, i.e. it holds in the form of ( 37 ) as an inequality for all non-negative \psi\in C_{c}^{\infty}([0,T)\times\Gamma) such that \partial_{t}\psi\leq 0 . (2) Without contact, weak solutions for similar problems were constructed in [ 8 , 46 ] , strong solutions in [ 35 , 36 , 37 ] and weak-strong uniqueness was shown in [ 44 ] . The first result