Inflection point instability in Hartmann channel flow with variable electric conductivity
Roman Okatev 1, Oleg Zikanov 2, Dmitry Krasnov 3, Peter Frick 1 · 2025
Read Inflection point instability in Hartmann channel flow with variable electric conductivity by Roman Okatev 1, Oleg Zikanov 2, Dmitry Krasnov 3, Peter Frick 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-2509-02086.txt.
Introduction
Roman Okatev, Oleg Zikanov, Dmitry Krasnov, and Peter Frick study a new instability in Hartmann channel flow with varying conductivity. Hartmann flow moves conducting fluid through a channel under a crosswise magnetic field. In liquid metal applications, conductivity and viscosity can vary across the channel width. That variation, absent in the classic setup, reshapes the base velocity profile.
Linear conductivity change makes the profile asymmetric and creates inflection points. Strong enough change triggers instability at far lower flow rates than the classic threshold. The instability shows long axial waves and growth centered in the channel core. Its traits echo the classic inviscid inflection point mechanism. Direct simulations confirm the resulting transition to turbulence.
Catalog record
| CREATOR | Roman Okatev 1, Oleg Zikanov 2, Dmitry Krasnov 3, Peter Frick 1 |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 26 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2509-02086
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Inflection point instability in Hartmann channel flow with variable electric conductivity Roman Okatev 1, Oleg Zikanov 2, Dmitry Krasnov 3, Peter Frick 1 Abstract The new type of instability of the flow of an electrically conducting fluid in a channel with imposed wall-normal magnetic field is studied using linear stability analysis and direct numerical simulations. The instability is triggered by the inflection points that develop in the base velocity profile when the electric conductivity of the fluid varies across the channel. † † affiliation: 1 Institute of Continuous Media Mechanics, Korolyov str.1, Perm, 614018, Russian Federation 2 University of Michigan - Dearborn, Dearborn, MI 48128-1491, USA 3 Technische Universitat Ilmenau, PF 100565, 98684 Ilmenau, Germany 1 Introduction In this paper, we revisit the question of the instability of a wall-bounded parallel shear flow of an electrically conducting fluid affected by an imposed steady magnetic field. The novelty of the work is that the physical properties of the fluid, such as its electric conductivity or viscosity, are not required to be constant. In addition to purely theoretical interest in how the variation of physical properties affects instability, the study is motivated by technological applications of liquid metals, such as breeding blankets of nuclear fusion reactors (see, e.g., Smolentsev 2021 and Mistrangelo et al. 2021 , development of MHD (magnetohydrodynamic) stirring systems ( Denisov et al. 2014 ; Stiller et al. 2013 ) , and MHD separation of impurities ( Kolesnichenko 2013 ; Zhang et al. 2014 ) . In all of these flows, the properties of the fluid may vary because of strong gradients of temperature or impurity concentration. Typically, there is a technological need in either intensification or suppression of mixing. Better understanding and quantitative assessment of the instability and transition to turbulence is, therefore, desirable. Technological and laboratory flows of liquid metals are typically characterized by small values of the magnetic Reynolds and Prandtl numbers. Accurate analysis can be performed within the framework of the inductionless (quasi-static) approximation, considering only the effect of the magnetic field on the flow and not the reverse effect of the flow on the field ( Davidson 2017 ) . Two nondimensional parameters determine the state of such a flow: the hydrodynamic Reynolds number \Rey\equiv Uh/\nu and the Hartmann number {\textit{Ha}}=Bh\sqrt{\sigma/\rho\nu} , where U and h are the typical velocity and length scales, \nu is the kinematic viscosity, B is the induction of the applied magnetic field, \sigma is the electric conductivity, and \rho is the density. A classical example of the MHD effect is the Hartmann flow – a pressure-driven flow along a channel with an imposed wall-normal magnetic field and perfectly electrically insulating walls ( Hartmann 1937 ) . The magnetic field flattens the laminar velocity profile in the core of the channel and concentrates the viscous stresses within the Hartmann boundary layers, whose typical thickness scales as \delta\sim{\textit{Ha}}^{-1} . The magnetic field also shifts the instability threshold to higher \Rey , as was first found in the experimental study by Hartmann & Lazarus 1937 . A detailed review of the instability and transition to turbulence in the Hartmann channel flow is provided by Zikanov et al. 2014 . At high Ha , the transition first occurs within the Hartmann layers. The process is largely determined by the value of the Reynolds number based on \delta as the length scale {\rm R}\equiv U\delta/\nu=\Rey/{\textit{Ha}} . The linear stability limit is {\rm R}_{cr,lin}=48250 ( Lingwood & Alboussiere 1999 ) . The actual transition to turbulence occurs at much lower values. For example, the compilation of earlier experiments reported by Branover 1978 provides {\rm R}_{cr}\approx 215 . Further experiments and numerical simulations showed that at {\textit{Ha}}\gtrsim 20 the transition occurs in the range 200<{\rm R}<400 (see Zikanov et al. 2014 , for a review) . The inconsistency with the predictions of the linear stability analysis is explained by the nonlinear bypass transition mechanism, also found in hydrodynamic wall-bounded parallel shear flows ( Krasnov et al. 2004 ) . If the magnetic field is strong, the instability of a liquid metal flow in a channel is shifted to very large values of \Rey . As an example, {\textit{Ha}}\approx 10^{4} is expected in liquid-metal components of Tokamak fusion reactors ( Smolentsev 2021 ; Mistrangelo et al. 2021 ) , which means \Rey_{cr}\sim 10^{6} and higher. It should be noted that large temperature gradients that are inevitably present in reactor blankets may cause transition to spatially complex unsteady flows at much lower \Rey . The reason is the instability to magnetoconvective fluctuations (see Zikanov et al. 2021 , for a review) . As we demonstrate later in the paper, variation of physical properties may lead to a significant deformation of the laminar Hartmann velocity profile. Recently, Okatev et al. 2023 found a generalization of the Hartmann solution for the case of a fluid whose properties (electrical conductivity and/or viscosity) vary across the layer. They have shown that velocity profiles can be significantly distorted and may develop inflection points. In this paper, we analyze the consequences of this transformation for the stability of the flow. This is the first study of the system, in which we focus on linear instability. The role of the nonlinear bypass mechanism and the effects related to the density change and the appearance of buoyancy forces are left to future studies. 2 Problem setting and base flow state (a) (b) Figure 1: (a) , Schematics of the flow. (b) , Base velocity profiles at {\textit{Ha}}=50 , \varsigma=0 (red dashed line corresponds to the ordinary Hartmann flow profile). See text for further explanation. We consider the wall-bounded parallel shear flow of an electrically conducting viscous fluid in a channel with non-conducting walls. A wall-normal magnetic field of induction B is imposed (see figure 1 a). The kinematic viscosity \nu and the electrical conductivity \sigma of the fluid may vary. The governing equations and boundary conditions in the non-dimensional form are \displaystyle\frac{\partial\mathbf{u}}{\partial t}+\mathbf{u}\cdot\nabla\mathbf{u} \displaystyle= \displaystyle-\nabla p+\frac{1}{\Rey}\nabla\cdot 2\nu\mathbf{S}+\frac{{\textit{Ha}}^{2}}{\Rey}\mathbf{j}\times\mathbf{e}_{z}, (1) \displaystyle\nabla\cdot\mathbf{u} \displaystyle= \displaystyle 0, (2) \displaystyle\mathbf{j} \displaystyle= \displaystyle\sigma\left(-\nabla\varphi+\mathbf{u}\times\mathbf{e}_{z}\right), (3) \displaystyle\nabla\cdot\left(\sigma\nabla\varphi\right) \displaystyle= \displaystyle\nabla\cdot\left(\sigma\mathbf{u}\times\mathbf{e}_{z}\right), (4) \displaystyle\mathbf{u}=\mathbf{0},~\frac{\partial\varphi}{\partial n} \displaystyle= \displaystyle 0\quad\mathrm{at}\quad z=\pm 1, (5) where \mathbf{u} is the velocity, p is the pressure, \mathbf{S}=\left(\nabla\mathbf{u}+\mathbf{u}\nabla\right)/2 is the rate of strain tensor, \mathbf{j} is the electric current density, \varphi is the electric potential. The Hartmann number {\textit{Ha}}=Bh\sqrt{\sigma_{0}/\rho\nu_{0}} and the Reynolds number \Rey=U_{0}h/\nu_{0} are defined using the mean values of viscosity \nu_{0} and conductivity \sigma_{0} . In this paper, we do not consider the transport of the variable coefficients \nu and \sigma by flow. Fixed stationary distributions \nu(z) and \sigma(z) corresponding to the base state are assumed. This simplification is consistent with the main focus of this paper on the hydrodynamic mechanisms of linear instability. Using ( 1 ) and ( 3 ) we can obtain the equation for the stationary base velocity profile \frac{d^{2}U}{dz^{2}}=-\frac{\Rey}{\nu}\frac{dp}{dx}+{\textit{Ha}}^{2}\left(\frac{\sigma}{\nu}\right)U-\frac{1}{\nu}\frac{d\nu}{dz}\frac{dU}{dz}. (6) The pressure gradient is determined by the normalization condition \frac{1}{2}\int_{-1}^{1}U(z)dz=1. (7) We consider linear distributions of \nu and \sigma with values at z=0 taken as reference values: \sigma(z)=1+\varkappa z,\quad\nu(z)=1+\varsigma z. (8) The coefficients \varkappa and \varsigma account for deviations from the classic case of constant \sigma and \nu and are the focus of this work. Thus ( 6 ) becomes \frac{d^{2}U}{dz^{2}}=-\frac{\Rey}{1+\varsigma z}\frac{dp}{dx}+{\textit{Ha}}^{2}\left(\frac{1+\varkappa z}{1+\varsigma z}\right)U-\frac{\varsigma}{1+\varsigma z}\frac{dU}{dz}. (9) (a) (b)