Incompressible limit for a fluid mixture

Weak solutions to PDEs (35D30) Navier-Stokes equations (35Q30) Existence, uniqueness, and regularity theory for compressible fluids and gas dynamics (76N10)

Abstract: In this paper we discuss the incompressible limit for multicomponent fluids in the isothermal ideal case. Both a direct limit-passage in the equation of state and the low Mach-number limit in rescaled PDEs are investigated. Using the relative energy inequality, we obtain convergence results for the densities and the velocity-field under the condition that the incompressible model possesses a sufficiently smooth solution, which is granted at least for a short time. Moreover, in comparison to single-component flows, uniform estimates and the convergence of the pressure are needed in the multicomponent case because the incompressible velocity field is not divergence-free. We show that certain constellations of the mobility tensor allow to control gradients of the entropic variables and yield the convergence of the pressure in L1.

Review text

In fluid mechanics, the equation of state of a pure fluid involves the density \(\rho =\hat{\rho}(T,p)\) or the specific volume \(1/\rho =\hat{\nu}(T, p)\), where the constitutive functions \(\hat{\rho},\hat{\nu}\) depend on the temperature \(T\) and pressure \(p\), respectively. The incompressibility is a property defined via \(\partial_p\hat{\varrho}=0\) or \(\partial_p\hat{\nu}=0\). The author considers the isothermal viscous case for balance equations. Actually, this is the compressible Navier-Stokes (NSt) system which in the pressure-velocity formulation has the form: \[ \partial_p\hat{\rho}(T,p) (\partial_tp+v\cdot\nabla p)= -\hat{\rho}(T,p)\,\operatorname{div}\, v ,\tag{1} \] \[ \hat{\rho}(T,p) (\partial_tv+(v\cdot\nabla )v)+ \nabla p = \,\operatorname{div}\,\mathbb{S} +\hat{\varrho}(T,p)b. \] Here, \(\mathbb{S}\) is the viscous stress tensor. Note that \(\partial_p\hat{\varrho}\) yields the incompressible NSt equations with constant density for the variables \(p\), \((v_1, v_2, v_3)\). The incompressible limit for multicomponent fluids in the isothermal ideal case is an interesting topic here. The incompressibility is related to the Mach-number \(\mathrm{Ma}= \sqrt{\rho}|v|/\sqrt{p}\) or \(\mathrm{Ma}=|v|/ \sqrt{\partial_{\varrho}p}\). A flow is called a low Mach-number flow if, after rescaling of position, time, and all variables and data occurring in the NSt equations, it obeys \[ \partial_tp\varrho + \operatorname{div}(\varrho v)=0, \tag{2} \] \[ \varrho (\partial_tv+(v\cdot\nabla )v)+ (\epsilon^{-2})\nabla p = \,\operatorname{div}\,\mathbb{S} +\varrho b, \] with a global Mach-number \(\mbox{Ma} = \epsilon \ll 1\). Next, the author considers the incompressible limit for a fluid mixture of \(N > 1\) components. Here, the fluid mixture is assumed homogeneous, Newtonian and ideal. In particular, it is volume-additive and therefore it obeys the equation of state \(\sum\limits_{i=1}^{N}\rho_i \hat{\nu}(T,p)=1\), where \(\rho_1, \dots,\rho_N\) are the partial mass densities of the components and \(\hat{\nu}_i(T,p)= (1/\hat{\rho}_i)(T,p)\) denotes the specific volumes of the \(i^{th}\) component as pure substance. The evolution of the partial mass densities \(\rho_1 , \dots,\rho_N\) and of the velocity field \(v = (v_1 , v_2 , v_3 )\) is described by the system of PDE: \[ \partial_t\rho_i + \operatorname{div}(\rho_iv+J^i)=0 \ \ (i=1,\dots,N), \tag{3} \] \[ \partial_t(\varrho v)+ \operatorname{div}(\varrho v\otimes v- \mathbb{S})+\nabla p = \varrho b, \] where \(J^i\) (\(i=1,2,\dots,N\)) are the diffusion fluxes, \(\mathbb{S}\) is the viscous stress tensor, \(p\) is the thermodynamic pressure, and \(b\) is the gravitational acceleration. Further, under the condition that the incompressible model possesses a sufficiently smooth solution the author considers the relative energy inequality. Thus, it is obtained convergence results for the densities and the velocity-field. In comparison to single-component flows, uniform estimates and the convergence of the pressure are needed in the multicomponent case. The reason is that the incompressible velocity field is not divergence-free. An important consequence is that ``certain constellations of the mobility tensor allow to control gradients of the entropic variables and yield the convergence of the pressure in \(L^1\) class.

Recommendations

Cites work

Cited in

(2)

This page was built for publication: Incompressible limit for a fluid mixture

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6158271)