Inclined flow of a second-gradient incompressible fluid with pressure-dependent viscosity
C. Balitactac and C. Rodriguez · 2025
Read Inclined flow of a second-gradient incompressible fluid with pressure-dependent viscosity by C. Balitactac and C. Rodriguez (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-2507-01986.txt.
Introduction
Balitactac and Rodriguez present a 2025 study of a thick liquid flowing down a sloped plane, in a model where the liquid's thickness itself depends on pressure. Many real liquids barely compress under pressure yet flow much more stiffly, and the classic flow equations cannot capture both traits at once. The authors study a newer extended model that adds higher-order terms, an exponential rule for how stiffness rises with pressure, and matching boundary rules.
They focus on steady flow down the slope, with weak grip at the lower surface and fixed air pressure at the top. Computer simulations show how the flow profile changes with the slope angle. They also vary the outside pressure, the pressure sensitivity, and an internal length scale. The work is the first test of the extended model on this basic flow.
Catalog record
| CREATOR | C. Balitactac and C. Rodriguez |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-01986
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Inclined flow of a second-gradient incompressible fluid with pressure-dependent viscosity C. Balitactac and C. Rodriguez Abstract . Many viscous liquids behave effectively as incompressible under high pressures but display a pronounced dependence of viscosity on pressure. The classical incompressible Navier–Stokes model cannot account for both features, and a simple pressure-dependent modification introduces questions about the well-posedness of the resulting equations. This paper presents the first study of a second-gradient extension of the incompressible Navier–Stokes model, recently introduced by the authors, which includes higher-order spatial derivatives, pressure-sensitive viscosities, and complementary boundary conditions. Focusing on steady flow down an inclined plane, we adopt Barus’ exponential law and impose weak adherence at the lower boundary and a prescribed ambient pressure at the free surface. Through numerical simulations, we examine how the flow profile varies with the angle of inclination, ambient pressure, viscosity sensitivity to pressure, and internal length scale. 1. Introduction Bridgman’s Nobel Prize–winning investigations revealed that many viscous liquids, when subjected to high pressures, behave essentially as incompressible while simultaneously showing a strong pressure dependence on their viscosity [ 5 , 6 ] . This dual behavior, apparent incompressibility coupled with pressure-sensitive viscous response, is not reflected in the classical incompressible Navier–Stokes equations, which assume a constant viscosity independent of pressure. To better capture such high-pressure phenomena, a number of studies have considered a modified version of the incompressible Navier–Stokes model, formulated within the framework of classical continuum mechanics [ 1 , 25 , 4 , 13 , 8 , 17 , 28 , 22 , 19 , 23 , 24 , 21 , 15 , 26 , 14 , 16 , 7 , 20 , 18 ] . In this model, viscosity is treated as a function of pressure, while the incompressibility constraint is retained: \displaystyle\rho\dot{\boldsymbol{v}}=\mathrm{div}\,\boldsymbol{T}+\rho\boldsymbol{b},\quad\boldsymbol{T}=-p\boldsymbol{I}+2\hat{\mu}(p)\boldsymbol{D},\quad\mathrm{div}\,\boldsymbol{v}=0. (1.1) Here, \rho is the constant mass density, \boldsymbol{v} is the velocity field, \dot{\mbox{}} denotes the material time derivative, p is the pressure, \hat{\mu}(p)>0 is the pressure-dependent viscosity, \boldsymbol{D}=\frac{1}{2}(\nabla\boldsymbol{v}+\nabla\boldsymbol{v}^{T}) is the stretching, and \boldsymbol{b} is a body force per unit mass. A widely used empirical relation for \hat{\mu}(p) is Barus’ exponential law [ 3 ] : \displaystyle\hat{\mu}(p)=\mu_{0}\exp(\beta p), (1.2) where \mu_{0}>0 and \beta>0 characterizes the sensitivity of viscosity to changes in pressure. Although this pressure-dependent formulation introduces important physical effects absent in the classical theory, it also gives rise to significant mathematical complications. In particular, the equation determining the pressure may lose ellipticity, rendering the system undetermined or ill-posed; see the discussion in Section 3 of [ 2 ] . Local well-posedness has only been established under the condition \hat{\mu}(p)/p\to 0 as p\to\infty , a requirement incompatible with Barus’ law and experimental observations [ 27 ] . Additional results are largely restricted to steady, low-speed flows in bounded domains [ 10 ] . In recent work [ 2 ] , we proposed a second-gradient incompressible viscous fluid model that retains the essential pressure-dependent viscosity. By augmenting the classical theory with second-gradient contributions, the resulting pressure equation remains elliptic regardless of the velocity field, avoiding the determinability issues inherent to ( 1.1 ); see Section 3 of [ 2 ] . The model introduced in [ 2 ] is characterized by the standard Cauchy stress \boldsymbol{T} and a third-order hyperstress tensor \mathbb{G} given by: \displaystyle\begin{split}\boldsymbol{T}&=-p\boldsymbol{I}+2\hat{\mu}(p)\boldsymbol{D},\\ \mathbb{G}&=-\frac{\ell_{1}^{2}}{2}\Bigl[\boldsymbol{I}\otimes\nabla p+(\boldsymbol{I}\otimes\nabla p)^{T(2,3)}\Bigr]+\mu_{0}(\ell_{2}^{2}+\ell_{3}^{2})\nabla\nabla\boldsymbol{v}\\ &\quad+\frac{1}{2}\mu_{0}(\ell_{2}^{2}-\ell_{3}^{2})\Bigl[\nabla\nabla\boldsymbol{v}^{T(1,2)}+\nabla\nabla\boldsymbol{v}^{T(1,3)}-\Delta\boldsymbol{v}\otimes\boldsymbol{I}\Bigr]\\ &\quad+\mu_{0}\Bigl(-\frac{1}{16}\ell_{2}^{2}+\frac{1}{4}\ell_{3}^{2}-\frac{1}{2}\ell_{4}^{2}\Bigr)\Bigl[\boldsymbol{I}\otimes\Delta\boldsymbol{v}+(\boldsymbol{I}\otimes\Delta\boldsymbol{v})^{T(2,3)}-4\Delta\boldsymbol{v}\otimes\boldsymbol{I}\Bigr],\end{split} (1.3) with the internal length scales \ell_{1},\dots,\ell_{4} related via \displaystyle\ell_{1}^{2}=\frac{3}{4}\ell_{2}^{2}+\frac{1}{2}\ell_{3}^{2}+2\ell_{4}^{2}. (1.4) By applying the principle of virtual power [ 11 , 12 , 9 ] , the governing equations reduce to the incompressibility condition \mathrm{div}\,\boldsymbol{v}=0 and \rho\left[\dot{\boldsymbol{v}}-\ell_{0}^{2}\mathrm{div}\,(\dot{\overline{\nabla\boldsymbol{v}}})\right]=-\nabla(p-\ell_{1}^{2}\Delta p)+\mathrm{div}\,(2\hat{\mu}(p)\boldsymbol{D})-\ell_{1}^{2}\mu_{0}\Delta^{2}\boldsymbol{v}+\rho\boldsymbol{b}, (1.5) where \ell_{0} is an additional internal inertial length scale. While [ 2 ] focused on cylindrical flows with constant viscosity, this paper presents the first analysis of steady flow for the full second-gradient model with pressure-dependent viscosity, examining gravity-driven motion down an inclined plane of angle \alpha and depth h (see Figure 1 ). The fluid adheres weakly to the lower boundary, while the upper surface is exposed to a prescribed ambient pressure; see ( 2.7 ). We adopt Barus’ exponential law ( 1.2 ) and establish the existence and uniqueness of solutions to the resulting boundary value problem. In Section 3, we then numerically explore how the flow profile is influenced by variations in the angle of inclination, and the nondimensionalized ambient pressure p_{1} , the Barus parameter \beta , and the internal length scale \ell_{1} . 2. Planar inclined Flow In this work we examine the steady flow of a second-gradient fluid with pressure-dependent viscosity of depth h under the force of gravity down an inclined plane making an angle \alpha with the horizontal (see Figure 1 ). We rotate the axes so that the inclined plane corresponds to the xz -plane, and the body force is then given by \rho\boldsymbol{b}=\rho g(\sin\alpha\,\boldsymbol{e}_{x}-\cos\alpha\,\boldsymbol{e}_{y}). (2.1) The field equations ( 1.5 ) are complemented by the boundary conditions expressing weak adherence along the plane y=0 and constant pressure p_{1} (e.g., atmospheric pressure) at y=h : \begin{cases}\boldsymbol{v}=\boldsymbol{0},\quad\boldsymbol{m}=\boldsymbol{0},\quad\frac{\partial p}{\partial\boldsymbol{n}}=0,&\mbox{at }y=0\\ \boldsymbol{t}=-p_{1}\boldsymbol{e}_{y},\quad\boldsymbol{m}=\boldsymbol{0},\quad\frac{\partial p}{\partial\boldsymbol{n}}=0,&\mbox{at }y=h.\end{cases} (2.2) Here \boldsymbol{t} and \boldsymbol{m} are the traction and hypertraction (see [ 9 , 2 ] ), given, respectively, by \displaystyle\boldsymbol{t}=\boldsymbol{T}\boldsymbol{n}-(\mathrm{div}\,\mathbb{G})\boldsymbol{n}-\mathrm{div}\,_{s}(\mathbb{G}\boldsymbol{n})-2K\mathbb{G}[\boldsymbol{n}\otimes\boldsymbol{n}],\quad\boldsymbol{m}=\mathbb{G}[\boldsymbol{n}\otimes\boldsymbol{n}], (2.3) where \boldsymbol{T} and \mathbb{G} are described in ( 1.3 ), \mathrm{div}\,_{s} denotes the surface divergence and K is the mean curvature. We assume the shear flow ansatz \displaystyle\boldsymbol{v}=v(y)\boldsymbol{e}_{x},\quad p=p(y),\quad y\in[0,h], (2.4) and thus, [\mathrm{div}\,2\mu\boldsymbol{D}]=\begin{bmatrix}(\mu v^{\prime})^{\prime}\\ 0\\ 0\end{bmatrix}, (2.5) where \mu=\hat{\mu}(p)=\mu_{0}\exp(\beta p) . Figure 1 . The set-up of inclined flow. The governing field equations ( 1.5 ) reduce to the following ordinary differential equations with \ell:=\ell_{1} : \begin{cases}0=(\mu v^{\prime})^{\prime}-\mu_{0}\ell^{2}v^{\prime\prime\prime\prime}+\rho g\sin\alpha\\ 0=-(p^{\prime}-\ell^{2}p^{\prime\prime\prime})-\rho g\cos\alpha.\end{cases} (2.6) The boundary conditions ( 2.2 ) are equivalent to \begin{cases}v(0)=0,v^{\prime\prime}(0)=0,p^{\prime}(0)=0\\ v^{\prime\prime}(h)=0,p^{\prime}(h)=0\\ \mu v^{\prime}(h)-\mu_{0}\ell^{2}v^{\prime\prime\prime}(h)=0,\\ \ell^{2}p^{\prime\prime}(h)-p(h)=-p_{1}\end{cases}. (2.7) Integrating the second equation in ( 2.6 ) from y to h and using the boundary conditions ( 2.7 ) gives the following: \displaystyle 0 \displaystyle=-p(h)+\ell^{2}p^{\prime\prime}(h)-\rho gh\cos\alpha+p(y)-\ell^{2}p^{\prime\prime}(y)+\rho gy\cos\alpha (2.8) \displaystyle=-p_{1}-\rho gh\cos\alpha+p(y)-\ell^{2}p^{\prime\prime}(y)+\rho gy\cos\alpha. (2.9) Thus, the equation that the pressure satisfies is \begin{cases}\ell^{2}p^{\prime\prime}(y)-p(y)=-p_{1}-(h-y)\rho g\cos\alpha,\quad y\in[0,h],\\ p^{\prime}(0)=p^{\prime}(h)=0.\end{cases} (2.10) One readily finds that p has the specific form p(y)=\frac{\ell\rho g\cos\alpha}{2\sinh(h/\ell)}\Bigl((1-e^{-h/\ell})e^{y/\ell}+(1-e^{h/\ell})e^{-y/\ell}\Bigr)+p_{1}+\rho g(h-y)\cos\alpha. (2.11) In the classical continuum setting with \ell=0 , we have \displaystyle p_{c}(y)=p_{1}+\rho g(h-y)\cos\alpha. (2.12) Integrating the first equation of ( 2.6 ) from y to h and using the boundary conditions yields: \displaystyle 0 \displaystyle=\mu(p(h))v^{\prime}(h)-\mu_{0}\ell^{2}v^{\prime\prime\prime}(h)+\rho gh\sin\alpha-\mu(p(y))v^{\prime}(y) (2.13) \displaystyle\quad+\mu_{0}\ell^{2}v^{\prime\prime\prime}(y)-\rho gy\sin\alpha (2.14) \displaystyle=\rho gh\sin\alpha-\mu(p(y))v^{\prime}(y)+\mu_{0}\ell^{2}v^{\prime\prime\prime}(y)-\rho gy\sin\alpha. (2.15) Thus, the velocity is governed by the boundary value problem \begin{cases}\mu_{0}\ell^{2}v^{\prime\prime\prime}(y)-\mu(p(y))v^{\prime}(y)=-\rho g(h-y)\sin\alpha,\quad y\in[0,h],\\ v(0)=0,v^{\prime\prime}(0)=v^{\prime\prime}(h)=0.\end{cases} (2.16) If \beta=0 , we obtain the following solution v_{\beta=0}(\sigma) to ( 2.16 ) corresponding to pressure-independent viscosity: v_{\beta=0}(y)=\frac{g\rho\sin\alpha}{2\mu_{0}}\Bigl(2hy-y^{2}-2\ell^{2}+2\ell^{2}\cosh\Bigl(\frac{h-2y}{2\ell}\Bigr)\mathrm{sech}\Bigl(\frac{h}{2\ell}\Bigr)\Bigr) (2.17) The pressure-dependent viscosity model rooted in classical continuum mechanics corresponds to \begin{cases}-\mu(p(y))v^{\prime}(y)=-\rho g(h-y)\sin\alpha,\quad y\in[0,h],\\ v(0)=0,\end{cases} (2.18) with p(y) given by ( 2.12 ). The solution to ( 2.18 ), v_{c}(y) , is readily found to be \displaystyle v_{c}(y) \displaystyle=\frac{\sec\alpha\tan\alpha\exp(-\beta(gh\rho\cos\alpha+p_{1}))}{g\beta^{2}\rho\mu_{0}} (2.19) \displaystyle\times\biggl[\exp(yg\beta\rho\cos\alpha)(1+g\beta\rho\cos\alpha(h-y))-gh\beta\rho\cos\alpha-1\biggr] (2.20) with the classical solution to ( 2.18 ) corresponding to pressure-independent viscosity ( \beta=0 ) given by \displaystyle v_{c,\beta=0}(y)=\frac{g\rho\sin\alpha}{2\mu_{0}}(2hy-y^{2}). (2.21) We now non-dimensionalize the variables. We define \displaystyle\sigma=\frac{y}{h},\quad\lambda=\frac{\ell}{h},\quad u(\sigma)=\frac{\mu_{0}}{\rho gh^{2}}v(h\sigma), (2.22) \displaystyle\pi(\sigma)=\frac{1}{\rho gh}p(h\sigma),\quad\pi_{1}=\frac{p_{1}}{\rho gh},\quad\gamma=\beta\rho gh. (2.23) The nondimensionalized pressure is given by \pi(\sigma)=\frac{\lambda\cos\alpha}{2\sinh(1/\lambda)}\Bigl((1-e^{-1/\lambda})e^{\sigma/\lambda}+(1-e^{1/\lambda})e^{-\sigma/\lambda}\Bigr)+\pi_{1}+(1-\sigma)\cos\alpha, (2.24) the dimensionless velocity profile u_{c} corresponding to v_{c} is given by \displaystyle u_{c}(\sigma) \displaystyle=\frac{\sec\alpha\tan\alpha\exp(-\gamma(\cos\alpha+\pi_{1}))}{\gamma^{2}} \displaystyle\times\biggl[\exp(\sigma\gamma\cos\alpha)(1+(1-\sigma)\gamma\cos\alpha)-\gamma\cos\alpha-1\biggr], and the dimensionless velocity profile u_{\gamma=0} corresponding to v_{\beta=0} is given by u_{\gamma=0}(\sigma)=\sin\alpha\Bigl(\sigma-\frac{\sigma^{2}}{2}-\lambda^{2}\Bigl(1-\cosh\Bigl(\frac{1-2\sigma}{2\lambda}\Bigr)\mathrm{sech}\Bigl(\frac{1}{2\lambda}\Bigr)\Bigr)\Bigr). (2.25) The dimensionless pressure \pi_{c} corresponding to p_{c} and velocity profile u_{c,\gamma=0} corresponding to v_{c,\beta=0} are