betaFeedback

WorksPhysics and Astronomy

Effects of strong fringing magnetic fields on turbulent thermal convection

Abstract

We study the influence of fringing magnetic fields on turbulent thermal convection in a horizontally extended rectangular domain. The magnetic field is created in the gap between two semi-infinite planar magnetic poles, with the convection layer located near the edge of the gap. We employ direct numerical simulations in this setup for fixed Rayleigh and small Prandtl numbers, but vary the fringe-width by controlling the gap between the magnetic poles and the convection cell. The magnetic field generated by the magnets is strong enough to cease the flow in high magnetic flux region of the convection cell. We observe that as the local vertical magnetic field strength increases, the large scale structures become thinner and align themselves perpendicular to the longitudinal sidewalls. We determine the local Nusselt and Reynolds numbers as functions of the local Hartmann number (based on the vertical component of the magnetic field) and estimate the global heat and momentum transport. We show that the global heat transport decreases with increasing fringe-width for strong magnetic fields but increases with increasing fringe-width for weak magnetic fields. In the regions of large vertical magnetic fields, the convective motion becomes confined to the vicinity of the sidewalls. The amplitudes of these wall modes show a non-monotonic dependence on the fringe-width.

Introduction

Flows that are driven by buoyancy forces are a common occurrence in nature as well as in technological applications. The driving mechanism is the temperature dependence of the fluid density, which leads to density variations when heat is transported through the fluid. A simplified paradigm for such flows is the Rayleigh-Bénard convection (RBC), which consists of a fluid layer that is heated from below and cooled from above [13]. While the understanding of RBC has increased substantially in the past decades (see, for example, [1, 14, 52]), it must be noted that a multitude of additional forces, such as those generated by rotation and magnetic fields, can affect buoyancy-driven convection in nature or in industrial applications. The effects of these forces have been relatively less explored. The present study deals with magnetoconvection, that is, thermal convection of electrically conducting fluids under the effect of magnetic fields.

In magnetoconvection, flows are acted upon by buoyancy as well as by Lorentz forces generated due to a magnetic field [59]. Convection in the Sun and planetary dynamos are examples of magnetoconvection occurring in nature where the magnetic field is maintained by the flow. In technological applications, the magnetic field is usually not primarily caused by the flow but externally imposed. Examples are liquid metal batteries for renewable energy storage, growth of semiconductor monocrystals, flow control of hot metal melts by electromagnetic brakes in metallurgy, and heat transfer in blankets in nuclear fusion reactors.

Magnetoconvection is governed by the equations for conservation of mass, momentum and energy as well as Maxwell’s equations and Ohm’s law. The governing nondimensional parameters of magnetoconvection are i) Rayleigh number RaRa – the ratio of buoyancy to dissipative forces, ii) Prandtl number PrPr – the ratio of kinematic viscosity to thermal diffusivity, iii) Hartmann number HaHa – the ratio of Lorentz to viscous force, and iv) the magnetic Prandtl number PmPm – the ratio of kinematic viscosity to magnetic diffusivity. Instead of Hartmann number, the free-fall interaction parameter Nf=Ha2Pr/RaN_f = Ha^2\sqrt{Pr/Ra}, also called the Stuart number, can be used as a governing parameter [34, 67] (Xu et al. 2022). The free-fall interaction parameter represents the ratio of magnetic and buoyant forcing in the fluid. The important nondimensional output parameters of magnetoconvection are i) the Nusselt number NuNu, which quantifies the global heat transport, ii) the Reynolds number ReRe – the ratio of inertial forces to viscous forces, and iii) the magnetic Reynolds number RmRm – the ratio of induction to diffusion of the magnetic field. For sufficiently small magnetic Reynolds numbers, the induced magnetic field is negligible compared to the applied magnetic field and is therefore neglected in the expressions of Lorentz force and Ohm’s law [42, 17, 51, 53]. In such cases, referred to as quasi-static magnetoconvection, the induced magnetic field adjusts instantaneously to the changes in velocity. In the quasi-static approximation, there exists a one-way influence of the magnetic field on the flow only.

Magnetoconvection has been studied theoretically in the past (for example, [13, 25, 11]) as well as with the help of experiments (for example, [38, 19, 15, 5, 9, 28, 55, 56, 67, 22]) and numerical simulations (for example, Liu et al. 2018; Yan et al. 2019; Akhmedagaev et al. 2020a,b; Nicoski et al. 2022). A horizontal magnetic field has been found to cause the large-scale rolls to become quasi two-dimensional and align in the direction of the field [19, 12, 10, 62, 49, 55, 56]. These self-organized flow structures reach an optimal state characterised by a significant increase in heat transport and convective velocities [56]. In contrast, strong vertical magnetic fields suppress convection [13, 15, 67, 2 b], with the flow ceasing above a critical Hartmann number. At this threshold, the flow in the center is fully suppressed with convective motion only in the vicinity of sidewalls [11, 34, 67, 2 b], if existent. These so-called wall modes, which are also present in confined rotating convection [64, 18, 20, 21, 35, 29], are particularly relevant in technical applications in closed vessels. Hurlburt et al. [26] and Nicoski et al. [39] studied convection with tilted magnetic fields. The results of Hurlburt et al. [26] suggest that the mean flows tend to travel in the direction of the tilt. Nicoski et al. [39] reported qualitative similarities between convection with tilted magnetic field and that with vertical magnetic field in terms of the behavior of convection patterns, heat transport, and flow speed.

It must be noted that all the aforementioned works on magnetoconvection have been restricted to uniformly imposed magnetic field, which is an idealized approximation tion. However, in engineering applications such liquid metal batteries [27], cooling blankets in fusion reactors [36], electromagnetic stirring [16], electromagnetic brakes [17], and non-contact flow measurements involving Lorentz force velocimetry [50], the imposed magnetic fields are localized and thus vary in space. Further, strong homogeneous fields in large regions of space can only be generated by magnets of large size which are difficult to design and very costly to build and operate [6]. Thus, it is important to understand the impact of spatially varying magnetic fields on magnetohydrodynamic flows. Although there are several studies on the channel flow of liquid metals under the influence of nonhomogeneous magnetic fields (see, for example, [46]; [57]; [37]; [4]; [30]), no similar work on magnetoconvection has been conducted so far to the best of our knowledge.

Convection flows in horizontally extended domains are often organized into prominent and coherent large-scale patterns that persist for very long times and can extend over scales in the lateral direction that are much larger than the domain height. These so-called superstructures have a strong influence on the turbulent transport properties of the flow. A prominent example of such superstructures is the granulation at the surface of the Sun [45]. While the understanding of the process of formation and evolution of these superstructures and their characteristic scales in the absence of magnetic fields has improved significantly over the recent years (see, for example, [40]; [47]; [32]), the case with inhomogeneous magnetic fields still leaves several questions open as the physical complexity is increased.

In the present work, we attempt to fill some of these gaps and study the effect of strong spatially varying magnetic fields in turbulent convection under the quasi-static approximation. The effects of localized magnetic fields on turbulent superstructures and their impact on the local and global turbulent heat and momentum transport are explored. We consider a Rayleigh-Bénard cell with fringing magnetic fields generated by semi-infinite magnetic poles. The strength of the magnets is such that convection in the regions of strong magnetic fields is fully suppressed in the bulk. Thus, in a single convection cell, we obtain different local regimes of magnetoconvection, ranging from the turbulent regime in the regions of weak magnetic flux to wall-attached convection in the regions of strong magnetic flux region. As mentioned earlier, the superstructures are formed in horizontally-extended domains; hence, the convection cell employed in our present study has a large aspect ratio. We study the spatial variation of size and orientation of turbulent superstructures along with the turbulent transport inside the convection cell. We also study wall-attached convection in the regions of strong magnetic flux regions, the wall modes. Although the setup is a simple way to study the influence of spatially varying magnetic fields on convection, it permits us to carry out such a parametric study systematically. This is the main motivation of the present work. Future studies can be conducted in a similar manner for more complex arrangements.

The outline of the paper is as follows. § 2 describes the magnetoconvection setup along with the governing equations of magnetoconvection and the spatial distribution of the magnetic field. The details of our numerical simulations are also discussed in § 2. In § 3, we present the numerical results, detailing the behavior of large-scale structures, the spatial distribution of heat and momentum transport along with their variations with the magnetic field, the dependence of the global heat and momentum transport on the fringe-width, and wall-attached convection. We conclude in § 4 with a summary and an outlook.

Numerical model

Problem setup and equations

We consider a horizontally extended convection cell of size lx×ly×Hl_x \times l_y \times H which is under the influence of a magnetic field generated by two semi-infinite permanent magnets. The north pole of one magnet faces the bottom of the convection cell and the south pole of the second magnet faces the top of the convection cell. These magnets extend from −∞-\infty to ∞\infty in the xx-direction, ly/2l_y/2 to ∞\infty in the yy-direction, from near the top wall to ∞\infty in the positive zz-direction, and from near the bottom wall to −∞-\infty in the negative zz direction. A schematic of a vertical section of the arrangement is provided in Figure 1(a). Using the model of Votyakov et al. [58] for permanent magnets, we obtain the following relations for the non-dimensional imposed magnetic field B0={Bx,By,Bz}\mathbf{B}_0 = \{B_x, B_y, B_z\} generated in the above configuration:

Figure 1. (a) Schematic of the vertical section at x=lx/2x = l_x/2 of the proposed magnetoconvection arrangement. The semi-infinite permanent magnets (colored in grey) generate the localized magnetic field represented by the vector plots drawn in the convection cell. The magnetic field distribution is described by Eqs. (1) to (3). (b) A three-dimensional view of the convection cell used in our study with the density contours of the vertical component of the normalized imposed magnetic field for δ/H=0.01\delta/H = 0.01. Here, OO refers to the point of the origin, and δ\delta is the gap between the magnetic pole and the convection cell.

Bx=0,(1)B_x = 0, \tag*{(1)}
By=−14πBz,max⁡ln⁡[(y−yc)2−{(z−zc)−(H/2+δ)}2(y−yc)2+{(z−zc)+(H/2+δ)}2],(2)B_y = -\frac{1}{4\pi}B_{z,\max}\ln\left[\frac{(y-y_c)^2-\{(z-z_c)-(H/2+\delta)\}^2}{(y-y_c)^2+\{(z-z_c)+(H/2+\delta)\}^2}\right], \tag*{(2)}
Bz=−12πBz,max⁡arctan⁡[2(H/2+δ)(y−yc)(y−yc)2+(z−zc)2−(H/2+δ)2],(3)B_z = -\frac{1}{2\pi}B_{z,\max}\arctan\left[\frac{2(H/2+\delta)(y-y_c)}{(y-y_c)^2+(z-z_c)^2-(H/2+\delta)^2}\right], \tag*{(3)}

where ycy_c and zcz_c are the yy and zz coordinates, respectively, of the cell-center, Bz,max⁡B_{z,\max} is the maximum limiting value of the vertical component of the magnetic field at y→∞y \to\infty, and δ\delta is the gap between the permanent magnet and the top/bottom wall (see Figure 1). The aforementioned magnetic field distribution satisfies ∇×B0=0\nabla\times\mathbf{B}_0 = 0 and ∇⋅B0=0\nabla\cdot\mathbf{B}_0 = 0. In the above configuration, the magnetic field is almost absent in nearly half of the RBC cell (for y≲ly/2y \lesssim l_y/2), increases steeply about the midplane (y≈ly/2y \approx l_y/2), and then becomes strong in the other half of the cell (for y≳ly/2y \gtrsim l_y/2). A three-dimensional view of the convection cell used in our study with the density contours of the vertical component of the imposed magnetic field is shown in Figure 1(b). The region where the gradient of the magnetic field is large is called the fringe zone; we provide a mathematical definition of the fringe zone later in this subsection. The region of the convection cell that is outside the gap between the magnets will be henceforth referred to as weak magnetic flux region and the one inside the gap will be referred to as strong magnetic flux region.

Equations (2.2) and (2.3) imply that the spatial distribution of the magnetic field is strongly dependent on the gap width (δ\delta) between the magnetic poles and the convection cell. In Figure 2(a) and (Figure 2b), we plot, respectively, the profiles of the normalized mean vertical magnetic field (⟨Bz/Bz,max⁡⟩x,z\langle B_z/B_{z,\max}\rangle_{x,z}) and the normalized mean horizontal magnetic field (⟨∣By∣/Bz,max⁡⟩x,z\langle|B_y|/B_{z,\max}\rangle_{x,z}) versus the normalized lengthwise coordinate y/Hy/H. In the above, ∣⋅∣|\cdot| and ⟨⋅⟩x,z\langle\cdot\rangle_{x,z}, respectively, represent the absolute value and averaging over xx and zz. As evident in Figure 2(a), the gradient of the vertical magnetic field decreases as δ\delta is increased. In the present setup, we define the fringe zone as the region in which 0.1<B^z<0.90.1 < \widehat{B}_z < 0.9, where

Distribution of the profiles of vertical and horizontal magnetic fields

Figure 2. Distribution of the profiles of (a) vertical magnetic field and (b) absolute value of the horizontal magnetic field, both averaged over xx and zz, along the lengthwise direction (normalized by the height HH of the cell) for different values of the normalized gap (δ′=δ/H\delta' = \delta/H) between the magnetic poles and the thermal plates. The magnetic fields are normalized by the maximum value of the vertical magnetic field (Bz′=Bz/Bz,max⁡B'_z = B_z/B_{z,\max} and By′=By/Bz,max⁡B'_y = B_y/B_{z,\max}). The curves are represented as dashed lines in the fringe zone and as solid lines outside the fringe zone.

B^z=⟨Bz⟩x,z−Bz,min⁡,DBz,max⁡,D−Bz,min⁡,D.(4)\widehat{B}_z = \frac{\langle B_z\rangle_{x,z}-B_{z,\min,D}}{B_{z,\max,D}-B_{z,\min,D}}. \tag*{(4)}

In the above, Bz,max⁡,DB_{z,\max,D} and Bz,min⁡,DB_{z,\min,D} are, respectively, the maximum and the minimum values of the vertical magnetic field in the convection cell 1. In Figure 2(a) and (Figure 2b), the curves describing the magnetic field profiles are represented as dashed lines in the fringe zone and as solid lines outside the fringe zone. The figures clearly show that the width of the fringe zone increases with an increase in δ\delta. Further, the horizontal component of the magnetic field on the lateral vertical midplane at y=ly/2y = l_y/2 increases with a decrease in δ\delta. These variations in the magnetic field profile affect the convection patterns along with the associated global heat transport and will be studied in detail in the later sections.

The study will be conducted under the quasi-static approximation, in which the induced magnetic field is neglected as it is very small compared to the applied magnetic field. This approximation is fairly accurate for magnetoconvection in liquid metals [17]. The nondimensionalized governing equations are as follows:

∇⋅u=0(5)\nabla\cdot\mathbf{u} = 0 \tag*{(5)}
∂u∂t+u⋅∇u=−∇p+Tz^+PrRa∇2u+Haz,max⁡2PrRa(j×B~),(6)\frac{\partial\mathbf{u}}{\partial t}+\mathbf{u}\cdot\nabla\mathbf{u}=-\nabla p+T\hat{z}+\sqrt{\frac{\mathit{Pr}}{\mathit{Ra}}}\nabla^2\mathbf{u}+\mathit{Ha}_{z,\max}^{2}\sqrt{\frac{\mathit{Pr}}{\mathit{Ra}}}\left(\mathbf{j}\times\widetilde{\mathbf{B}}\right), \tag*{(6)}
∂T∂t+u⋅∇T=1RaPr∇2T,(7)\frac{\partial T}{\partial t}+\mathbf{u}\cdot\nabla T=\frac{1}{\sqrt{\mathit{Ra}\mathit{Pr}}}\nabla^2T, \tag*{(7)}
j=−∇ϕ+(u×B~),(8)\mathbf{j}=-\nabla\phi+\left(\mathbf{u}\times\widetilde{\mathbf{B}}\right), \tag*{(8)}
∇2ϕ=∇⋅(u×B~),(9)\nabla^2\phi=\nabla\cdot\left(\mathbf{u}\times\widetilde{\mathbf{B}}\right), \tag*{(9)}

where u\boldsymbol{u}, j\boldsymbol{j}, pp, TT, and ϕ\phi are the fields of velocity, current density, pressure, temperature, and electrical potential respectively, and B~\widetilde{\boldsymbol{B}} is the applied magnetic field normalized by Bz,max⁡B_{z,\max}. The governing equations are made dimensionless by using the cell height HH, the imposed temperature difference Δ\Delta, and the free-fall velocity U=αgΔHU = \sqrt{\alpha g \Delta H} (where gg and α\alpha are, respectively, the gravitational acceleration and the volumetric coefficient of thermal expansion of the fluid). The nondimensional governing parameters are the Rayleigh number (RaRa), the Prandtl number (PrPr), the Hartmann number (Haz,max⁡Ha_{z,\max}) based on Bz,max⁡B_{z,\max}, and the normalized gap (δ′\delta') between the thermal plates and the magnetic poles. These parameters are defined as follows:

Ra=αgΔH3νκ,Pr=νκ,Haz,max⁡=Bz,max⁡Hσρ,δ′=δH,(10)Ra = \frac{\alpha g \Delta H^3}{\nu\kappa}, \qquad Pr = \frac{\nu}{\kappa}, \qquad Ha_{z,\max} = B_{z,\max} H \sqrt{\frac{\sigma}{\rho}}, \qquad\delta' = \frac{\delta}{H}, \tag*{(10)}

where ν\nu is the kinematic viscosity, κ\kappa is the thermal diffusivity, ρ\rho is the density, and σ\sigma is the electrical conductivity of the fluid. For the sake of brevity, we henceforth omit the prime from δ′\delta'. In the next section, we describe the simulation details.

Numerical method

We conduct direct numerical simulations of our magnetoconvection setup described in § 2.1. The spatial distribution of the magnetic field is given by equations (2.1) to (2.3). We use a second-order finite difference code developed by Krasnov et al. [31] to numerically solve equations (2.5) to (2.9). The Prandtl number PrPr is chosen to be 0.021, which is the same as that of mercury. The Rayleigh number is fixed at Ra=105Ra = 10^5. For our chosen PrPr, the aforementioned value of RaRa causes significant turbulence in the regions of weak magnetic flux. For Ra=105Ra = 10^5, the critical Hartmann number (above which the bulk convection is suppressed) computed using the linear stability analysis of Chandrasekhar [13] for uniform vertical magnetic field is

Haz,c=88.7.(11)Ha_{z,c} = 88.7. \tag*{(11)}

In order to include the wall-attached convection regime in our analysis, we choose the maximum Hartmann number to be Haz,max⁡=120Ha_{z,\max} = 120, which is slightly higher than the critical Hartmann number.

As mentioned in § 1, we intend to study the behavior of turbulent superstructures in our present study. For our chosen values of RaRa and PrPr, the typical length scales of the superstructures are 3 to 4 times the height of the domain [40]. Therefore, in order to obtain good statistics, we choose the horizontal extent of our domain to be around 5 times the expected length-scale of the superstructures. Furthermore, in order to capture the variation of the superstructures and the global heat transport with magnetic field strength more effectively, it is preferred that the domain is elongated along the yy-direction, that is, along the gradient of the magnetic field. Therefore, we choose the domain-size to be lx×ly×H=16×32×1l_x \times l_y \times H = 16 \times32 \times1.

We employ a grid-resolution of 4800×9600×3004800 \times9600 \times300 points. The mesh is non-uniform in the zz-direction with stronger clustering of the grid points near the top and bottom boundaries. The elliptic equations for pressure, electric potential, and the temperature are solved based on applying cosine transforms in xx and yy-directions and using a tridiagonal solver in the zz-direction. The diffusive term in the temperature transport equation is treated implicitly. The time discretization of the momentum equation uses the fully explicit Adams-Bashforth/Backward-Differentiation method of second order [41]. A constant time step size of 1×10−41 \times10^{-4} free fall time units was chosen for our simulations, which satisfied the Courant–Friedrichs–Lewy (CFL) condition for all runs.

All the walls are rigid and electrically insulated such that the electric current density j\mathbf{j} forms closed field lines inside the cell. The top and bottom walls are held fixed at T=−0.5T = -0.5 and T=0.5T = 0.5 respectively, and the sidewalls are adiabatic with ∂T/∂η=0\partial T/\partial\eta= 0 (where η\eta is the component normal to sidewall). All the simulations are initialized with the linear conduction profile for temperature (which is a function of the zz-coordinate only) and a random noise of amplitude A=0.001A = 0.001 along the zz-direction for velocity. We run the simulations initially on a coarse grid of 480×960×30480 \times960 \times30 points for 100 free-fall time units in which they converge to a statistically steady state. Following this, we successively refine the mesh to the required resolution of 4800×9600×3004800 \times9600 \times300 grid points and allow the simulations to converge after each refinement. Once the simulations reach the statistically steady state at the highest resolution, they are run for a further 20–21 free-fall time units and a snapshot of the flow field is saved after every free-fall time unit.

Table 1 lists the important parameters of our simulation runs. In this table, we also report the turbulent momentum transfer quantified by the global Reynolds number (ReglobalRe_{\mathrm{global}}) and the turbulent heat transport quantified by the global Nusselt number (NuglobalNu_{global}). These quantities are given by

Reglobal=urmsRa/Pr,(12)Re_{global}=u_{rms}\sqrt{Ra/Pr}, \tag*{(12)}
Nuglobal=1+RaPr⟨uzT⟩V,tN,(13)Nu_{global}=1+\sqrt{RaPr}\langle u_zT\rangle_{V,t_N}, \tag*{(13)}

where ⟨⋅⟩\langle\cdot\rangle represents averaging, urms=⟨ux2+uy2+uz2⟩V,tNu_{rms}=\sqrt{\langle u_x^2+u_y^2+u_z^2\rangle_{V,t_N}} with VV being the volume of the convection cell, and tNt_N (also reported in Table 1) is the number of free-fall times run by the solver after attaining a statistically steady state.

Runsδ\deltaReglobalRe_{\mathrm{global}}NuglobalNu_{\mathrm{global}}tNt_N
10.01763±2763 \pm21.80±0.011.80 \pm0.0120
20.3767±5767 \pm51.82±0.011.82 \pm0.0120
31715±4715 \pm41.75±0.011.75 \pm0.0121
43644±3644 \pm31.72±0.011.72 \pm0.0120
59474±3474 \pm31.60±0.011.60 \pm0.0120

Table 1. Parameters of the simulations: the gap (δ\delta) between the magnetic poles and the conducting walls, the global Reynolds number (ReglobalRe_{\mathrm{global}}), the global Nusselt number (NuglobalNu_{\mathrm{global}}), and the number of free-fall time units (tNt_N) run by the solver after attaining statistically steady state. ReglobalRe_{\mathrm{global}} and NuglobalNu_{\mathrm{global}} are averaged over tNt_N timeframes and the errors are the standard deviations of the above quantities. The Rayleigh number, Prandtl number, and the maximum Hartmann number are fixed to Ra=105Ra = 10^5, Pr=0.021Pr = 0.021, and Haz,max⁡=120Ha_{z,\max} = 120 respectively.

In turbulent magnetoconvection with vertical magnetic fields, four boundary layers are formed: 1) viscous boundary layers near all the walls, 2) thermal boundary and 3) Hartmann layers near the top and bottom walls, and 4) Shercliff layers near the sidewalls. Figure 3 exhibits a sketch of the different types of boundary layers. In order to obtain accurate results, it is important that all these layers are adequately resolved. In our cases, although the imposed magnetic field has a horizontal component as well, it is very small compared to the vertical component near the sidewalls. Hence, the sidewalls have been considered to contain only the Shercliff layers and no Hartmann layers. It must also be noted that the thicknesses of all the boundary layers vary along the yy-direction because they depend on the local Hartmann number, which, in turn, affects the Reynolds and Nusselt numbers. Thus, for a conservative resolution analysis, the minimum thickness of these boundary layers has been considered. These are given by

Figure 3. Sketches of the vertical (xx-zz) cross section of a Rayleigh-Bénard convection cell (with aspect ratio similar to ours) under vertical magnetic field showing (a) thermal boundary layers, (b) viscous boundary layers, and (c) Hartmann and Shercliff layers.

δT,min=12Numax,(14)\delta_{T,min}=\frac{1}{2Nu_{max}}, \tag*{(14)}
δu,min=14Remax,(15)\delta_{u,min}=\frac{1}{4\sqrt{Re_{max}}}, \tag*{(15)}
δH,min=1Haz,max,(16)\delta_{H,min}=\frac{1}{Ha_{z,max}}, \tag*{(16)}
δS,min=1Haz,max,(17)\delta_{S,min}=\frac{1}{\sqrt{Ha_{z,max}}}, \tag*{(17)}

where δT,min\delta_{T,min}, δu,min\delta_{u,min}, δH,min\delta_{H,min}, and δS,min\delta_{S,min} are the minimum thicknesses of thermal boundary layers, viscous boundary layers, Hartmann layers, and the Shercliff layers, respectively. Here, we also remark that although the relation for the viscous boundary layer thickness given by δu∼Re−1/2\delta_u\sim Re^{-1/2} is not very accurate (see, for example, Breuer et al. 2004; Scheel et al. 2012; Bhattacharya et al. 2021), it provides a reasonable estimate for small Prandtl number convection. Further, NumaxNu_{max} and RemaxRe_{max}, respectively, denote the Nusselt and Reynolds numbers in the regions where the magnetic fields are weak. They are computed as follows. For all our simulation cases, we identify the regions where B^z<0.1\hat{B}_z<0.1 and compute the time-averaged Nusselt and Reynolds number integrated over these regions for every case. The maximum values of the above Nusselt and Reynolds numbers (which turned out to be for the case with δ=0.3\delta=0.3) are taken to be NumaxNu_{max} and RemaxRe_{max}, respectively, and are observed to be Numax=2.66Nu_{max}=2.66 and Remax=1115Re_{max}=1115. Based on the above values, we have a minimum of 12 points, 76 points, and 12 points, respectively, in the viscous boundary layers, thermal boundary layers, and in the Hartmann layers adjacent to the top and bottom walls; and 28 points in the Shercliff layer adjacent to the sidewalls. Thus, our simulations are well resolved and also satisfy the resolution criterion of Grötzbach [23] and Verzicco & Camussi [54]. In the next section, we discuss our results in detail.

Results

In this section, we discuss the characteristics of the large-scale convection patterns, the spatial profile of the large-scale momentum and the global heat transport, and the wall-attached convection in the regions of strong magnetic flux. For our analysis, we introduce the local vertical Hartmann number Haz(y)Ha_z(y) which is defined as,

Haz(y)=Haz,max⁡⟨Bz⟩x,zBz,max⁡,(18)Ha_z(y) = Ha_{z,\max} \frac{\langle B_z\rangle_{x,z}}{B_{z,\max}}, \tag*{(18)}

where BzB_z is the vertical component of the local magnetic field. Thus, Haz(y)Ha_z(y) quantifies the strength of BzB_z averaged over the corresponding xx-zz-plane. We will discuss the variations of the heat and momentum transport with Haz(y)Ha_z(y) and their resulting impact on the global dynamics.

Large-scale structures

We use our numerical data to analyse the flow structures in the convection cell for different fringe-widths. Figure 4(a)-(Figure 4e) exhibit the contour plots of temperature field on the horizontal midplane for δ=0.01\delta= 0.01 to 9. The corresponding contour plots of the vertical velocity field are shown in Figure 4(f)-(Figure 4g). The data for the above plots are averaged over 20 to 21 timeframes (see Table 1). The figures show that bulk convection is completely suppressed in the regions of strong magnetic fields. The region of suppressed bulk convection becomes smaller as δ\delta is increased. This is expected because for large δ\delta, the gradient of the magnetic field is small and hence the region with Haz(y)Ha_z(y) above the critical Hartmann number is also small.

Figure 4. Contour plots of time-averaged fields on the horizontal midplane for different fringe-widths of the imposed magnetic field. Plots of the temperature field for (a) δ=0.01\delta= 0.01, (b) δ=0.3\delta= 0.3, (c) δ=1\delta= 1, (d) δ=3\delta= 3, and (e) δ=9\delta= 9. Plots of the vertical velocity field for (f) δ=0.01\delta= 0.01, (g) δ=0.3\delta= 0.3, (h) δ=1\delta= 1, (i) δ=3\delta= 3, and (j) δ=9\delta= 9. The flows are organized into large-scale patterns that extend over scales larger than the domain height. The positions corresponding to Haz(y)=Haz,cHa_z(y) = Ha_{z,c}, Haz(y)=0.8Haz,cHa_z(y) = 0.8Ha_{z,c}, and the edge of the magnetic poles are represented by purple, blue, and black horizontal lines, respectively, in decreasing order of thickness. The magnetic poles extend from y=16y = 16 to y=∞y = \infty.

Figure 4(a)–(Figure 4j) also show that in the regions of Haz(y)<Haz,max⁡Ha_z(y) < Ha_{z,\max}, the flow gets organized into superstructures. The superstructures observed in the above figures in the weak magnetic flux regions are qualitatively similar to those observed by Pandey et al. [40] for Pr=0.021Pr = 0.021. Further, the superstructures are relatively isotropic in the regions of weak magnetic flux, that is, they do not show any preference towards specific orientations. However, as the strength of the vertical magnetic field increases, the superstructures become more elongated and align themselves perpendicular to the sidewalls. The change in spatial structure of the flow is more clearly visible for δ=9\delta= 9, in which the vertical magnetic field changes gradually. The transition in the orientation of the superstructures begins to occur for Haz(y)≈0.8Haz,cHa_z(y) \approx0.8Ha_{z,c} where the flow is dominated by ascending and descending planar jets originating from the sidewalls. The flow structures in this regime are very similar to those observed by Akhmedagaev et al. [3] in their simulation data of RBC in a cylindrical cell under strong uniform vertical magnetic field. They can be attributed to quasi-two-dimensional vortex sheets which are often found in magnetohydrodynamic flows with strong magnetic fields [66]. As one moves further towards stronger magnetic flux region, the structures originating from the sidewalls extend less into the bulk, until for Haz(y)≳Haz,cHa_z(y) \gtrsim Ha_{z,c}, the flow is confined only near the sidewalls. These wall-attached flows will be discussed in detail in § 3.3.

It can be visually observed that the size of the convection patterns decreases along the direction of increasing magnetic field. We focus specifically on the results of δ=9\delta= 9 because the increase of magnetic field along yy is gradual in this case, thus enabling us to obtain better statistics for analyzing the variations of the structures with the magnetic field strength. In order to quantitatively analyze the evolution of the size of the structures along yy, we divide the horizontal midplane into 10 subslices along the yy-direction and estimate the characteristic length scale of these structures in each subslice using the vertical velocity data. Each subslice spans the entire width of the convection cell. Since the structures are large in the low magnetic flux region, the first two subslices (corresponding to 0<y<5.340 < y < 5.34 and 5.34<y<10.675.34 < y < 10.67) are larger than the rest of the subslices; the larger subslices have a width of 5.34 units and the smaller subslices are 2.67 units wide. We employ the procedure outlined in Pandey et al. [40] to estimate the length scale as elaborated below. For each subslice, we compute the Fourier transform of the vertical velocity field at each snapshot to obtain u^z(k,ϕk,t)\widehat{u}_z(k,\phi_k,t), where kk is the magnitude of the wavevector and ϕk\phi_k is the corresponding azimuthal angle in the wavenumber space. We calculate the azimuthally averaged kinetic energy spectra, which is obtained by

Eu(k,t)=12π∫02π∣u^z(k,ϕk,t)∣2 dϕk,(19)E_u(k,t) = \frac{1}{2\pi}\int_{0}^{2\pi} \left|\widehat{u}_z(k,\phi_k,t)\right|^2\,\mathrm{d}\phi_k, \tag*{(19)}

The above computed kinetic energy spectra are averaged over 20 snapshots. The wavenumber kmax⁡k_{\max} corresponding to which the spectrum is maximum is the characteristic wavenumber2; and the characteristic length scale λ\lambda is computed as

λ=2πkmax⁡.(20)\lambda= \frac{2\pi}{k_{\max}}. \tag*{(20)}

The contour plot of uzu_z is redrawn in Figure 5(a) and the computed characteristic length scale for each subslice is plotted versus yy in Figure 5(b). Figure 5(b) shows that λ\lambda decreases with the increasing magnetic field strength along yy, which is consistent with the visual observation that the size of the patterns decreases along yy. Note that the characteristic length scale in the weak magnetic flux region is λ=5.5\lambda= 5.5, which is of the same order as the value of λ≈4\lambda\approx4 computed by Pandey et al. [40] using their numerical data for Ra=105Ra = 10^5 and Pr=0.021Pr = 0.021. Figure 5(b) also shows that the characteristic length scales computed using our data are larger than the critical wavelength λc\lambda_c Chandrasekhar [13] at the onset of convection for the corresponding Hartmann number. In the wall-attached convection region where Haz(y)>Haz,max⁡Ha_z(y) > Ha_{z,\max}, the computed length scale ranges from λ≈1.5\lambda\approx1.5 to λ≈2\lambda\approx2 and denotes the characteristic size of the wall-modes. These values are fairly close to the analytically derived value of λc,w=2\lambda_{c,w} = \sqrt{2} for wall modes in magnetoconvection with one sidewall Busse [11].

Pattern analysis for $\delta = 9$

Figure 5. Pattern analysis for δ=9\delta= 9. (a) Contour plot of time-averaged vertical velocity field on the horizontal midplane. Solid lines are replotted from Figure 4. (b) The evolution of the characteristic horizontal length-scale of the superstructures (red circles) and the critical wavelength λc\lambda_c (black curve) at the onset of convection along the horizontal yy-direction. The length scale of the convection patterns becomes smaller along the direction of the increasing magnetic field and is larger than the critical wavelength.

We now examine the convection structures on different vertical planes along yy for δ=9\delta= 9. In Figure 6(a,Figure 6b,Figure 6c), we show the density plots of vertical velocity on the planes at y=3.4y = 3.4, y=11.9y = 11.9 and y=19.7y = 19.7. The local Hartmann number at these locations are Haz(y=3.4)=25Ha_z(y = 3.4) = 25, Haz(y=11.9)=45Ha_z(y = 11.9) = 45, and Haz(y=19.7)=75Ha_z(y = 19.7) = 75. The velocity fluctuations are strong at y=3.4y = 3.4 where the flow is highly turbulent. At y=11.9y = 11.9, the flow structures consist of quasi-two-dimensional upward and downward streams occupying the entire plane. The velocity fluctuations become weak at y=19.7y = 19.7 and the flow structures become elongated along the xx-direction, consistent with the observation from Figure 4(e) and (Figure 4j). We plot the mean horizontal velocity ⟨ux2+uy2⟩x\left\langle\sqrt{u_x^2 + u_y^2}\right\rangle_x, the mean absolute vertical velocity ⟨∣uz∣⟩x\langle|u_z| \rangle_x, and the mean temperature ⟨T⟩x\langle T\rangle_x versus zz on the above three planes in Figure 7(a,Figure 7b,Figure 7c) respectively. These figures further reinstate that the velocity fluctuations decrease along yy. Moreover, the temperature approaches the linear conduction profile as the local magnetic field strength increases.

Contour plots of the time-averaged vertical velocity field

Figure 6. Contour plots of the time-averaged vertical velocity field for δ=9\delta= 9 in vertical planes at (a) y=3.4y = 3.4 (Haz(y)=25Ha_z(y) = 25), (b) y=11.9y = 11.9 (Haz(y)=45Ha_z(y) = 45), and (c) y=19.7y = 19.7 (Haz(y)=75Ha_z(y) = 75). The velocity fluctuations decrease with the increasing magnetic field strength along yy.

Profiles for $\delta = 9$ of the mean horizontal velocity, mean absolute vertical velocity, and mean temperature

Figure 7. Profiles for δ=9\delta= 9 of (a) the mean horizontal velocity ⟨ux2+uy2⟩x\langle\sqrt{u_x^2+u_y^2}\rangle_x, (b) the mean absolute vertical velocity ⟨∣uz∣⟩x\langle|u_z|\rangle_x, and (c) the mean temperature along the vertical zz-direction at y=3.9y = 3.9 (black dotted curves), y=11.9y = 11.9 (red dashed curves), and y=19.7y = 19.7 (blue solid curves). As the magnetic field strength along yy increases, the velocity fluctuations decrease and the temperature approaches the linear conduction profile.

We now quantitatively analyse the isotropy of the flow using our numerical data. Towards this objective, we compute the following ratios: the local vertical anisotropy parameter 2Ez(y)/Eh(y)2E_z(y)/E_h(y), the local horizontal anisotropy parameter Ey(y)/Ex(y)E_y(y)/E_x(y), and the ratio of the kinetic energies along the zz and yy directions. In the above, Ez(y)=0.5⟨uz2⟩x,zE_z(y) = 0.5\langle u_z^2\rangle_{x,z}, is the vertical kinetic energy, Eh(y)=0.5⟨ux2+uy2⟩x,zE_h(y) = 0.5\langle u_x^2 + u_y^2\rangle_{x,z} is the horizontal kinetic energy, Ex(y)=0.5⟨ux2⟩x,zE_x(y) = 0.5\langle u_x^2\rangle_{x,z} is the kinetic energy along xx direction, and Ey(y)=0.5⟨uy2⟩x,zE_y(y) = 0.5\langle u_y^2\rangle_{x,z} is the kinetic energy along the yy direction, all averaged over the xx-zz plane. To understand the variation of flow anisotropy with magnetic field strength, the aforementioned anisotropy factors are plotted versus Haz(y)/Haz,cHa_z(y)/Ha_{z,c} for different δ\delta in Figure 8(a)–(Figure 8e). In these plots, we focus on the regions where 0.25Haz,c<Haz(y)<Haz,c0.25Ha_{z,c} < Ha_z(y) < Ha_{z,c} because, as discussed later in § 3.2, the flow undergoes significant changes in this parameter interval. Further, we add horizontal tick labels with the coordinates of yy on the top of each plot to better understand the evolution of anisotropy along the horizontal direction.

Mean profiles of local anisotropy parameters

Figure 8. Mean profiles of the local vertical anisotropy parameter 2Ez(y)/Eh(y)2E_z(y)/E_h(y) (solid green curves), the local horizontal anisotropy parameter Ey(y)/Ex(y)E_y(y)/E_x(y) (dashed red curves), and the ratio of the kinetic energy along zz and yy directions – Ez(y)/Ey(y)E_z(y)/E_y(y) (dotted blue curves) – versus the normalized Hartmann number, Haz(y)/Haz,cHa_z(y)/Ha_{z,c}, based on the mean vertical magnetic field Bz(y)B_z(y), for (a) δ=0.01\delta= 0.01, (b) δ=0.3\delta= 0.3, (c) δ=1\delta= 1, (d) δ=3\delta= 3, and (e) δ=9\delta= 9. Both the vertical and the horizontal anisotropy parameters increase with the local vertical magnetic field strength.

The above plots show that the flow is roughly isotropic in the regions of weak vertical magnetic field corresponding to Haz(y)<0.3Haz,cHa_z(y) < 0.3Ha_{z,c}. However, with increasing Haz(y)Ha_z(y), the vertical velocity fluctuations become more dominant compared to the horizontal ones. This property is consistent with the anisotropy observed by Yan et al. [61] and Akhmedagaev et al. [3] for strong Hartmann number convection with uniform magnetic fields, and is reminiscent of the stable cellular regime described by Zürner et al. [67]. The increase of anisotropy with Haz(y)Ha_z(y) becomes more pronounced as the fringe-width, governed by δ\delta, increases. This is because as the fringe-width increases, the span of yy falling in the regime of 0.25Haz,c<Haz(y)<Haz,c0.25Ha_{z,c} < Ha_z(y) < Ha_{z,c} also increases as exhibited in Figure 8(a)–(Figure 8e). Thus, we get better statistics, and hence, clearer trends in this regime for large δ\delta.

Figure 8(a)–(Figure 8e) also show that for strong magnetic fields, there is anisotropy among the horizontal components of velocity as well. For Haz(y)>0.65Haz,cHa_z(y) > 0.65Ha_{z,c}, Ey(y)E_y(y) begins to dominate Ex(y)E_x(y), implying that the velocity fluctuations in the yy direction are stronger compared to that in the xx direction. This is also consistent with Figure 4(a)–(Figure 4j) which show that for moderately strong Hartmann numbers, the superstructure rolls orient themselves perpendicular to the longitudinal sidewalls. In the next subsection, we discuss the effects of the fringing magnetic fields on the local as well as global momentum and heat transport.

Heat and momentum transport

We analyze the spatial variation of the large-scale heat and momentum transport in our numerical setup. Towards this objective, we first compute the local planar Reynolds number Re(y)Re(y) and Nusselt number Nu(y)Nu(y) for every δ\delta. These are given by

Re(y)=RaPr⟨ux2+uy2+uz2⟩x,z,tN1/2,Nu(y)=1+RaPr⟨uzT⟩x,z,tN.(21)Re(y) = \sqrt{\frac{Ra}{Pr}}\langle u_x^2+u_y^2+u_z^2\rangle_{x,z,t_N}^{1/2}, \qquad Nu(y) = 1+\sqrt{RaPr}\langle u_zT\rangle_{x,z,t_N}. \tag*{(21)}

It must be noted that in order to avoid clutter, Nu(y)Nu(y) and Re(y)Re(y) are computed for every 120th xx-zz-plane. Moreover, Nu(y)Nu(y) exhibits strong spatial fluctuations even after averaging over all the timeframes; these fluctuations are smoothened by computing the moving average of Nu(y)Nu(y) using a window-size of 5 xx-zz-planes.

We plot Nu(y)Nu(y) and Re(y)Re(y) versus yy in figures Figure 9(a) and (Figure 9b) respectively. Figure 9(a) shows that Nu(y)Nu(y) fluctuates between 2.0 and 4.0 in the region of weak magnetic flux, drops steeply in the fringe-zone and becomes close to unity in the strong magnetic flux region. Figure 9(b) shows that Re(y)Re(y) follows a similar trend; it is large in the weak magnetic flux region, drops steeply in the fringe-zone, and becomes negligibly small in the strong magnetic flux region. The gradients of both Re(y)Re(y) and Nu(y)Nu(y) curves in the fringe-zone decrease with an increase of δ\delta. This is expected because BzB_z, which suppresses kinetic energy and heat transport in the fluid, has steeper gradients for small δ\delta.

Variations of the local planar Nusselt and Reynolds numbers with the longitudinal direction $y$

Figure 9. Variations of the local planar Nusselt and Reynolds numbers with the longitudinal direction yy. For δ=0.01\delta= 0.01 to 9: (a) local Nusselt number, Nu(y)Nu(y), and (b) local Reynolds number, Re(y)Re(y), versus yy.

We now look for a universal dependence of the local heat and momentum transport on the local magnetic field strength. Taking inspiration from the recent experimental work of Zürner et al. [67] on thermal convection under uniform vertical magnetic fields, we plot the normalized local Nusselt number Nu~(y)\widetilde{Nu}(y) versus the normalized local vertical Hartmann number Haz(y)/Haz,cHa_z(y)/Ha_{z,c}. The normalized local Nusselt number is given by

Nu~(y)=Nu(y)−1Numax⁡−1,(22)\widetilde{Nu}(y) = \frac{Nu(y)-1}{Nu_{\max}-1}, \tag*{(22)}

where Numax⁡=2.66Nu_{\max}=2.66 (see § 2.2). Figure 10(a) shows that for all δ\delta, the points collapse well into a single curve, thus showing a universal dependence of Nu(y)Nu(y) on Haz(y)Ha_z(y). The plot shows that for Haz(y)≲0.3Haz,cHa_z(y) \lesssim0.3Ha_{z,c}, the Nusselt number does not change significantly and remains close to its maximum value. This regime corresponds to the turbulent and isotropic regime. For Haz(y)≳0.3Haz,cHa_z(y) \gtrsim0.3Ha_{z,c}, the flow transitions into the cellular regime (as per the nomenclature of Zürner et al. [67]) and the local Nusselt number starts to drop sharply as Haz(y)Ha_z(y) increases. For Haz(y)<0.85Haz,cHa_z(y) < 0.85Ha_{z,c}, the best fit curve for our data is

Variations of normalized local Nusselt and Reynolds numbers

Figure 10. Variations of the normalized local Nusselt and Reynolds numbers with the normalized local vertical magnetic field strengths. For δ=0.01\delta= 0.01 to 9: (a) normalized local Nusselt number, Nu~(y)\widetilde{Nu}(y), versus the normalized local Hartmann number, Haz(y)/HacHa_z(y)/Ha_c, based on vertical magnetic field strength; and (b) normalized local Reynolds number, Re~(y)\widetilde{Re}(y), versus the local free-fall interaction parameter, Nf(y)N_f(y). The best-fit curves for the above data are also shown.

Nu~(y)=[1+χ1{Haz(y)Haz,c}γ1]−1,(23)\widetilde{Nu}(y) = \left[1+\chi_1\left\{\frac{Ha_z(y)}{Ha_{z,c}}\right\}^{\gamma_1}\right]^{-1}, \tag*{(23)}

where γ1=2.1±0.3\gamma_1=2.1\pm0.3 and χ1=4.0±0.7\chi_1=4.0\pm0.7. Interestingly, the values of χ1\chi_1 and especially γ1\gamma_1 are close to those observed by Zürner et al. [67], who reported γ1=2.03±0.06\gamma_1=2.03\pm0.06 and χ1=5.9±0.3\chi_1=5.9\pm0.3 for uniform magnetic fields. For Haz(y)>0.85Haz,cHa_z(y)>0.85Ha_{z,c}, Nu~(y)\widetilde{Nu}(y) decreases more steeply than (3.6) and is described by the following power law,

Nu~(y)=0.028{Haz(y)Haz,c}−10.5±0.9.(24)\widetilde{Nu}(y)=0.028\left\{\frac{Ha_z(y)}{Ha_{z,c}}\right\}^{-10.5\pm0.9}. \tag*{(24)}

This regime corresponds to the wall-mode regime in which the flow gets further suppressed and begins to confine itself near the sidewalls. It must be noted that the Zürner

et al. (2020) could not obtain a best-fit expression for NuNu in the wall-mode regime due to lack of data close to the wall.

We conduct a similar analysis for Re(y)Re(y) as well. We compute the normalized Reynolds number given by

Re~(y)=Re(y)Remax⁡,(25)\widetilde{Re}(y) = \frac{Re(y)}{Re_{\max}}, \tag*{(25)}

where Remax⁡=1115Re_{\max} = 1115 (see § 2.2). The results of Zürner et al. [67] suggest that Re~(y)\widetilde{Re}(y) should scale as Haz(y)/Haz,cHa_z(y)/\sqrt{Ha_{z,c}} instead of Haz(y)/Haz,cHa_z(y)/Ha_{z,c}. Let us now consider the local free-fall interaction parameter Nf(y)N_f(y), which is given by

Nf(y)=Haz2(y)PrRa.(26)N_f(y) = Ha_z^2(y)\sqrt{\frac{Pr}{Ra}}. \tag*{(26)}

For large Rayleigh numbers, the critical Hartmann number can be approximated as Haz,c≈Ra/πHa_{z,c} \approx\sqrt{Ra}/\pi (Chandrasekhar [13]). Using this relation, we get

Nf(y)=Haz2(y)πRaPrπ≈Haz2(y)PrHaz,cπ∼Haz2(y)Haz,c.(27)N_f(y) = Ha_z^2(y)\frac{\pi}{\sqrt{Ra}}\frac{\sqrt{Pr}}{\pi} \approx\frac{Ha_z^2(y)\sqrt{Pr}}{Ha_{z,c}\pi} \sim\frac{Ha_z^2(y)}{Ha_{z,c}}. \tag*{(27)}

The above relation shows that Nf(y)N_f(y) is proportional to Haz2(y)/Haz,cHa_z^2(y)/Ha_{z,c}, implying that Re~(y)\widetilde{Re}(y) scales with Nf(y)N_f(y). Thus, we plot Re~(y)\widetilde{Re}(y) versus Nf(y)N_f(y). The plot in Figure 10(b) shows that, similar to the Nusselt number, the data points for all δ\delta collapse into a single curve. For Nf(y)<0.17N_f(y) < 0.17, which corresponds to Haz(y)<0.22Haz,cHa_z(y) < 0.22Ha_{z,c}, the Reynolds number retains its maximum value (Remax⁡Re_{\max}). At higher Nf(y)N_f(y), the Reynolds number starts to fall sharply with the interaction parameter. For Nf(y)<3.2N_f(y) < 3.2, which corresponds to Haz(y)<0.95Haz,cHa_z(y) < 0.95Ha_{z,c}, the best fit curve for our data is given by

Re~(y)=[1+χ2{Nf(y)}γ2]−1,(28)\widetilde{Re}(y) = [1 + \chi_2\{N_f(y)\}^{\gamma_2}]^{-1}, \tag*{(28)}

with χ2=1.40±0.03\chi_2 = 1.40 \pm0.03 and γ2=1.01±0.03\gamma_2 = 1.01 \pm0.03. The value of γ2\gamma_2 is again close to that observed by Zürner et al. [67], who reported γ2=0.87±0.03\gamma_2 = 0.87 \pm0.03 for uniform magnetic fields. For Haz(y)>0.95Haz,cHa_z(y) > 0.95Ha_{z,c}, which corresponds to the wall-attached convection regime, Re~(y)\widetilde{Re}(y) falls even more steeply than (28) and is described by the following power law,

Re~(y)=18{Nf(y)}−4.0±0.2.(29)\widetilde{Re}(y) = 18\{N_f(y)\}^{-4.0\pm0.2}. \tag*{(29)}

We now analyze the impact of the local heat and momentum transport on the evolution of the viscous and thermal boundary layers near the top and bottom walls of our convection cell. Again, we concentrate on δ=9\delta= 9 to obtain better statistics and a clearer picture of the evolution. The thermal boundary layer thickness δT\delta_T is the depth where a linear fit of the mean temperature profile near the wall intersects with the mean bulk temperature (T=0.5T = 0.5). The viscous boundary layer thickness is the depth where a linear fit of the horizontal velocity (uh=⟨(ux2+uy2)1/2⟩x)(u_h = \langle(u_x^2 + u_y^2)^{1/2}\rangle_x) profile near the wall intersects with its local maximum. See Breuer et al. [8], Ahlers et al. [1] and Scheel et al. [43] for a detailed procedure to compute the boundary layer thicknesses. We compute the thermal and viscous boundary layer thicknesses at different yy-coordinates for δ=9\delta= 9 and plot them versus the corresponding local Hartmann number Haz(y)Ha_z(y) in Figure 11(a,Figure 11b), respectively.

Boundary analysis plots for $\delta = 9$

Figure 11. Boundary analysis for δ=9\delta= 9. (a) Local thermal boundary layer thickness δT(y)\delta_T(y) and (b) local viscous boundary layer thickness δu(y)\delta_u(y) near the top and bottom walls versus the local Hartmann number Haz(y)Ha_z(y). The thermal boundary layer thickness increases with the increase of Haz(y)Ha_z(y). The viscous boundary layer thickness decreases with increasing Haz(y)Ha_z(y) and approaches Hartmann type scaling of δu∼Haz−1\delta_u \sim Ha_z^{-1} at large Hartmann numbers.

Figure 11(a) shows that the thermal boundary layer thickness increases with Haz(y)Ha_z(y). This is expected because the thermal boundary layer thickness is inversely proportional to the local Nusselt number Nu(y)Nu(y) which decreases with increasing Haz(y)Ha_z(y). The viscous boundary layer, on the other hand, exhibits an interesting behaviour. As evident in Figure 11(b), the boundary layer thickness δu\delta_u decreases with Haz(y)Ha_z(y), with the decrease becoming steeper at large Haz(y)Ha_z(y). To gain insight into this behavior, we take inspiration from the work of Lim et al. [33] and using dimensional analysis on the momentum equation, we obtain the following scaling for δu\delta_u:

δu∼1Re(y)+Haz2(y).(30)\delta_u \sim\frac{1}{\sqrt{Re(y) + Ha_z^2(y)}}. \tag*{(30)}

The above scaling suggests that for Haz(y)≪Re(y)Ha_z(y) \ll\sqrt{Re(y)}, δu\delta_u follows the Prandtl-Blasius type scaling of δu∼Re−1/2\delta_u \sim Re^{-1/2} in which the boundary layer thickness should increase with the decrease of Re(y)Re(y), and hence, with the increase of Haz(y)Ha_z(y). However, we do not observe such regime in our convection cell. When Haz(y)∼Re(y)Ha_z(y) \sim\sqrt{Re(y)}, the boundary layer thickness varies marginally with yy, as observed for Haz(y)<50Ha_z(y) < 50 from our data. For Haz(y)≫Re(y)Ha_z(y) \gg\sqrt{Re(y)}, the boundary follows the Hartmann type scaling of δu∼Haz−1\delta_u \sim Ha_z^{-1} and hence, δu\delta_u decreases sharply with Haz(y)Ha_z(y). From Figure 11(b), it can be seen that δu\delta_u approaches the Hartmann type scaling at Haz(y)≈70Ha_z(y) \approx70.

Having studied the local variations of heat and momentum transport along with their associated boundary layers, we now analyze the global Reynolds and Nusselt numbers (ReglobalRe_{global} and NuglobalNu_{global} respectively) and their dependence on δ\delta. These global quantities are computed numerically using equations (12) and (13) from our simulation data and are plotted versus δ\delta in Figure 12(a) and (Figure 12b) as black squares. The figures show that both ReglobalRe_{global} and NuglobalNu_{global} decrease as δ\delta increases. This result is counter intuitive because for large δ\delta, bulk convection is completely ceased only in a small region (as seen in Figure 4) and hence one would expect the overall heat and momentum transport to be stronger. However, a careful look at Figure 2(a) reveals that the vertical magnetic field is stronger in the weak magnetic flux region for large fringe-widths, which, in turn, weakens convection in that region. The increase in convection in the strong magnetic flux region is unable to compensate the suppression of convection in the weak magnetic flux region, resulting in a decrease of overall heat and mass transport for large fringe-widths.

Plots of the global Reynolds and Nusselt numbers

Figure 12. Plots of (a) the global Reynolds number ReglobalRe_{\mathrm{global}} (black squares for Haz,max⁡=120Ha_{z,\max}=120 and purple circles for Haz,max⁡=30Ha_{z,\max}=30), computed from our numerical data using equation (12), versus δ\delta, and (b) the global Nusselt number NuglobalNu_{\mathrm{global}} (black squares for Haz,max⁡=120Ha_{z,\max}=120 and purple circles for Haz,max⁡=30Ha_{z,\max}=30), computed from our numerical data using equation (13), versus δ\delta. Also shown in the plots are the estimates (solid curves) of ReglobalRe_{\mathrm{global}} and NuglobalNu_{\mathrm{global}} computed using equations (31) and (32) for different values of Haz,max⁡Ha_{z,\max}.

The above results, however, do not indicate whether such a variation of global Reynolds and Nusselt numbers with δ\delta holds for all Haz,max⁡Ha_{z,\max}. To explore this aspect, we need to estimate ReglobalRe_{global} and NuglobalNu_{global} for a given value of Haz,maxHa_{z,max} and δ\delta. Towards this objective, we make use of the best fit relations for Nu(y)Nu(y) and Re(y)Re(y) given by equations (23), (24), (28), and (29) obtained earlier, and numerically integrate them over the entire domain to estimate global Reynolds and Nusselt numbers as follows:

Reglobal=1ly∫0lyRe(y) dy=1ly∫0lyRe(Haz(y)) dy,(31)Re_{global} = \frac{1}{l_y}\int_{0}^{l_y} Re(y)\,\mathrm{d}y = \frac{1}{l_y}\int_{0}^{l_y} Re(Ha_z(y))\,\mathrm{d}y, \tag*{(31)}
Nuglobal=1ly∫0lyNu(y) dy=1ly∫0lyNu(Haz(y)) dy,(32)Nu_{global} = \frac{1}{l_y}\int_{0}^{l_y} Nu(y)\,\mathrm{d}y = \frac{1}{l_y}\int_{0}^{l_y} Nu(Ha_z(y))\,\mathrm{d}y, \tag*{(32)}

where we recall that

Haz(y)=Haz,max⟨Bz⟩x,zBz,max,Ha_z(y) = Ha_{z,max}\frac{\langle B_z\rangle_{x,z}}{B_{z,max}},

and Bz=Bz(x,y,z,δ)B_z = B_z(x,y,z,\delta) is given by equation (3). Thus, for given Haz,maxHa_{z,max} and δ\delta, one can estimate ReglobalRe_{global} and NuglobalNu_{global} using equations (31) and (32). We compute these estimates for Haz,max=120Ha_{z,max} = 120, 60, 50, and 30, and for δ\delta ranging from 0.005 to 9.1 in increments of 0.01. The estimated values of ReglobalRe_{global} and NuglobalNu_{global} are plotted versus δ\delta in Figure 12(a) and (Figure 12b) as solid curves. The curves for Haz,max=120Ha_{z,max} = 120 fit the data points of directly computed ReglobalRe_{global} and NuglobalNu_{global} (black squares) very well, implying that equations (31) and (32) provide fairly accurate estimates. In the process of computing the estimates of ReglobalRe_{global} and NuglobalNu_{global} for Haz,max<120Ha_{z,max} < 120, we assume that the equations and coefficients given in relations (3.6), (3.7), (3.11), and (3.12) remain unchanged as the maximum Hartmann number is varied. We also assume that the values of Haz(y)=0.85Haz,cHa_z(y) = 0.85Ha_{z,c} and Nf(y)=3.2N_f(y) = 3.2, which were observed to separate the two components of the piecewise functions for Nu(y)Nu(y) and Re(y)Re(y) for Haz,max=120Ha_{z,max} = 120, remain the same for all values of Haz,maxHa_{z,max}. These assumptions are based on the fact that the dependence of Nu(y)Nu(y) and Re(y)Re(y) on Haz(y)Ha_z(y) is suggested to be universal since the normalized local Nusselt and Reynolds number curves for different δ\delta collapse as shown in Figure 10(a,Figure 10b).

Figure 12(a) also shows that the decrease of ReglobalRe_{global} with δ\delta becomes less apparent with a decrease of Haz,maxHa_{z,max}. In fact, for Haz,max=30Ha_{z,max} = 30, ReglobalRe_{global} increases marginally with δ\delta. Figure 12(b) indicates that NuglobalNu_{global} increases with δ\delta for Haz,max=60Ha_{z,max} = 60, 50, and 30, with the above increase becoming more apparent with the decrease of Haz,maxHa_{z,max}. To test the correctness of the above predictions, we conduct two more direct numerical simulations of our magnetoconvection setup for Haz,max⁡=30Ha_{z,\max}=30 and δ=0.01\delta=0.01 and 99. The grid-size and the numerical schemes are the same as those described in § 2.2. We run the simulations for 10 free-fall time units after attaining a statistically steady state. We compute the time-averaged global Reynolds and Nusselt numbers using the above two simulation data and plot them in Figure 12(a) and (Figure 12b) respectively. The figures show that the simulation results are in close agreement with the predictions, thereby reinforcing our theory. It is therefore clear that for smaller maximum magnetic field strengths, the overall heat and momentum transport increases with the increase of fringe-width. This implies that for small values of Haz,max⁡Ha_{z,\max}, the suppression of convection in the weak magnetic flux region is small compared to the increase of convection in the strong magnetic flux region. Thus, we can conclude that the variation of ReglobalRe_{\mathrm{global}} and NuglobalNu_{\mathrm{global}} with the fringe-width depends critically on Haz,max⁡Ha_{z,\max}. Finally, we discuss the dynamics of wall-attached convection in the strong magnetic flux region in the next subsection.

Wall-attached convection

We plot the time-averaged isosurfaces of uz=0.01u_z=0.01 (red) and uz=−0.01u_z=-0.01 (blue) for δ=1\delta=1 and δ=9\delta=9 in Figure 13(a) and (Figure 13b), respectively. The figures show that in the strong magnetic flux region, the flow in the bulk is completely suppressed with alternating up-and downwelling flow regions attached to the sidewalls. The structure of the wall-modes is consistent with that observed by Liu et al. [34] in their simulations of thermal convection in a rectangular domain with uniform vertical magnetic field. However, unlike for rotating convection (Horn & Schmid [24]; Zhang et al. [63]) or convection in cylindrical domains (Akhmedagaev et al. [3]), the wall modes do not oscillate or move along the sidewalls (Grannan et al. [22]; Schumacher [44]).

Isosurfaces of vertical velocity for $\delta=1$ and $\delta=9$

Figure 13. Isosurfaces of uz=0.01u_z = 0.01 (red) and uz=−0.01u_z = -0.01 (blue) for (a) δ=1\delta= 1 and (b) δ=9\delta= 9. In the regions with strong magnetic fields, convective motions are confined near the sidewalls.

The wall-modes are dense and uniformly distributed along the sidewalls for δ=9\delta=9. However, for δ=1\delta=1, they are visibly thinner with isosurfaces clustering more towards the corners instead of being uniformly distributed along the sidewalls. The clustering occurs because of weaker Lorentz force in the corners, which, in turn, results due to the electric current getting further constrained by the two insulated sidewalls in proximity of each other. The above trend in the local differences of the wall modes is also observed for δ=0.01\delta=0.01, 0.30.3, and 33 as shown later in this section.

For each sidewall in the strong magnetic flux region, we compute the mean distance δˉw\bar{\delta}_w of the wall modes from the sidewall. This distance is computed as follows,

δˉw=∫δwuz,max⁡ dl∫uz,max⁡ dl,(33)\bar{\delta}_w = \frac{\int\delta_w u_{z,\max} \, \mathrm{d}l}{\int u_{z,\max} \, \mathrm{d}l}, \tag*{(33)}

where dl=dy\mathrm{d}l = \mathrm{d}y for the longitudinal sidewall and dl=dx\mathrm{d}l = \mathrm{d}x for the lateral sidewall. In the above, δw\delta_w is the shortest distance between the sidewall and the point of maximum absolute velocity uz,max⁡u_{z,\max} adjacent to the sidewall, and the summations are over the regions where Haz(y)>Haz,cHa_z(y) > Ha_{z,c}. Thus, δˉw\bar{\delta}_w is the weighted average of the shortest distance between the point of maximum absolute velocity and the sidewall, with the maximum absolute velocity being the weights. We also compute the mean distance averaged over all the sidewalls. We plot the mean distances for each sidewall as well as the overall mean distance versus δ\delta in Figure 14. The figure shows that the wall modes are in general more attached to the lateral sidewall (y=32y = 32) compared to the longitudinal sidewalls (x=0x = 0 and x=16x = 16). There is, however, no clear trend regarding the variation of δˉw\bar{\delta}_w with the fringe-width.

Figure 14. The average distance between the wall modes and the sidewalls versus δ\delta in the regions of Haz(y)>Haz,cHa_z(y) > Ha_{z,c}. The wall modes are more attached to the lateral sidewalls than the longitudinal sidewalls.

We further analyze the amplitudes of the wall modes along the three sidewalls for our runs. Towards this objective, we measure the vertical velocity field at every point on the horizontal midplane at a distance of δˉw\bar{\delta}_w from the sidewalls. We plot this vertical velocity profile along every sidewall in Figure 15. The figure shows that with the exception of δ=9\delta= 9, the amplitudes of the wall modes are stronger near the corners and become weaker as one moves away from the corners. This behavior in the amplitudes is consistent with Figure 13(a) which shows that the wall-modes are clustered in the corners. As explained earlier, these local differences in amplitudes occur due the current being more constrained in the corners resulting in weaker convection-suppressing Lorentz force in these regions. Figure 15 also shows that near the corners, the amplitudes of the wall modes grow with an increase of δ\delta. However, in the regions away from the corners, the amplitudes decrease as δ\delta is decreased from 9 to 1; and then increases with a further decrease of δ\delta. For δ=0.3\delta= 0.3 and 1, the amplitudes become negligible small away from the corners. However, for δ=0.01\delta= 0.01, although the amplitudes of the modes away from the corners are smaller than those near the corners, they are larger than the amplitudes of wall modes away from the corners for δ=0.3\delta= 0.3 and 1.

Vertical velocity profiles for five values of $\delta$ along the sidewalls

Figure 15. Vertical velocity uz,wu_{z,w} measured on the horizontal midplane at distance of δˉw\bar{\delta}_w from the sidewalls in the regions of Ha(y)>Haz,cHa(y) > Ha_{z,c} (nonshaded regions) for (a) δ=0.01\delta= 0.01 (black curve), (b) δ=0.3\delta= 0.3 (purple curve), (c) δ=1\delta= 1 (blue curve), (d) δ=3\delta= 3 (green curve), and (e) δ=9\delta= 9 (red curve). The shaded regions correspond to Haz(y)<Haz,cHa_z(y) < Ha_{z,c} where the flow is not fully suppressed in the bulk. The amplitudes of the wall modes in the regions away from the corners exhibit a non-monotonic dependence on δ\delta. Note that the above plots are resolved by a total of at least 9000 grid points in the horizontal direction.

The reason for the non-monotonic behavior of wall-modes with δ\delta away from the corners is not yet clear. A possible explanation is that in addition to the vertical component of the magnetic field BzB_z, the horizontal component ByB_y also plays a role in stabilizing convection near the sidewalls. From Figure 2(b), it can be seen that for small δ\delta or fringe-width, although ByB_y is strong in y=ly/2y = l_y/2 midplane, it falls sharply and becomes very small as one moves away from the midplane. Thus, for δ=0.01\delta= 0.01, ByB_y is too weak to stabilize convection near the sidewalls at y≫ly/2y \gg l_y/2, resulting in sustained wall-attached convection.

On the other hand, for large δ\delta, ByB_y in the regions away from y=ly/2y = l_y/2 midplane is not as weak as that for small δ\delta. Thus, as δ\delta is increased up to δ=1\delta= 1, ByB_y becomes stronger and suppresses wall-attached convection. On further increasing δ\delta, although ByB_y grows only marginally, BzB_z becomes too weak to suppress wall-attached convection. Hence, we get prominent wall modes again for δ=3\delta= 3 and 99. The non-monotonic behavior of wall-modes will be explored in detail in a future study.

Summary and conclusions

In this paper, we performed detailed numerical simulations of turbulent convection of a horizontally extended domain under the influence of fringing magnetic fields. The magnetic field is generated in the gap between semi-infinite poles of permanent magnets, with the convection domain located near the edge of the gap. The quasi-static approximation of magnetohydrodynamics is applied. We kept the Rayleigh number, Prandtl number, and the maximum Hartmann number fixed at Ra=105Ra = 10^5, Pr=0.021Pr = 0.021, and Haz,max=120Ha_{z,max} = 120 respectively, and varied the fringe-width by controlling the gap δ\delta between the magnetic poles and the thermal plates. We analysed the spatial distribution of large-scale convection patterns, heat and momentum transport, and wall-attached convection, and their variations with the magnetic field profiles.

The convection patterns are found to comprise of structures whose horizontal dimensions are larger than the domain height. These structures do not exhibit any preferred orientation in regions of weak magnetic field, but become aligned perpendicular to the longitudinal sidewalls in the fringe zone. Further, the flow progressively becomes anisotropic towards the vertical direction in the fringe zone.

We obtained the best-fit relations for local Reynolds and Nusselt number as a function of the local Hartmann number that are valid for any fringe-width. We integrated these relations over the entire domain to estimate the global Reynolds and Nusselt numbers for different values of Haz,maxHa_{z,max}. The variation of the above global quantities with fringe-width was shown to depend on Haz,maxHa_{z,max} and is governed by the balance between the convective motion in the weak magnetic flux region outside the gap between the magnets and suppression of convection in the strong magnetic flux region inside the gap between the magnets. As the fringe-width increases, the magnetic field becomes weaker in the strong magnetic flux region; it becomes stronger in the weak magnetic flux region. For large Haz,maxHa_{z,max}, the increase of convection in the strong magnetic field region is unable to compensate for the suppression of convection in the weak magnetic flux region, resulting in a decrease of the global heat and momentum transport with increasing fringe-width. However, for small Haz,maxHa_{z,max}, the increase of convection in the strong magnetic field region is more than the decrease of convection in the weak magnetic flux region, resulting in an increase of global heat and momentum transport with increasing fringe-width.

We further analyzed wall-attached convection in the regions of strong magnetic flux. We showed that the amplitudes of the wall modes away from the corners decrease when δ\delta is decreased from 99 to 11, but begins to increase on further reduction of δ\delta. We believe that the horizontal components of the magnetic field play a crucial role in the stability of the wall modes. A detailed study on the this behavior will be conducted in a future work.

Here, we also remark that the initial conditions play a minor role in the evolution of the large-scale structures. In Appendix A, we compare therefore some of our results for δ=9\delta= 9 with another simulation of the same setup with identical parameters, but with different initial conditions. We observe that although there is a marginal difference in the alignment of the superstructures, the other results such as the global Nusselt and Reynolds numbers, and the variations of anisotropy and the local heat transport along yy remain largely unchanged.

Our present work provides important insights into the dynamics of thermal convection under spatially varying magnetic fields, which may find applications in several industrial and astrophysical flows. It should therefore be considered as a first step towards understanding thermally-driven flows in applications where nonhomogeneous magnetic fields are encountered. For example, our results are expected to help in designing engineering applications such as cooling blankets in fusion reactors. Although we worked on a small set of parameters, we expect our results to hold for higher Rayleigh numbers as well. In the future, we plan to conduct a similar analysis for fluids at different Prandtl numbers.

Acknowledgements

The authors thank A. Pandey, Y. Kolesnikov, M. Brynjell-Rahkola, M. K. Verma, and A. Ranjan for useful discussions. The authors are grateful to the anonymous referees for their suggestions on improving the manuscript. The authors gratefully acknowledge the computing time provided to them on the high-performance computer Noctua2 at the NHR Center PC2 of Paderborn University (Germany). These are funded by the Federal Ministry of Education and Research and the state governments participating on the basis of the resolutions of the GWK for the national high performance computing at universities (www.nhr-verein.de/unsere-partner). The authors further acknowledge the computing time provided by Leibniz Supercomputing Center, Garching, Germany, under the projects “pn49ma” and “pn68ni”. The authors thank the Computing Center of Technische Universität Ilmenau for the resources provided to them for postprocessing and visualization of their simulaion data.

Funding

The work of S.B. is sponsored by a Postdoctoral Fellowship of the Alexander von Humboldt Foundation, Germany.

Declaration of interests

The authors report no conflict of interest.

Author ORCIDs

Shashwat Bhattacharya https://orcid.org/0000-0001-7462-7680 Thomas Boeck https://orcid.org/0000-0002-0814-7432 Dmitry Krasnov https://orcid.org/0000-0002-8339-7749 Jörg Schumacher https://orcid.org/0000-0002-1359-4536

Appendix A

Effect of initial conditions

In this section, we explore the impact of initial conditions on the solutions in our convection setup as described in § 2. In the present work, all the simulations were initialized with the conduction profile for temperature and a small perturbation of amplitude A=0.001A = 0.001 in the zz-direction for velocity. Here, we consider another simulation of our setup for δ=9\delta= 9, Haz,max=120Ha_{z,max} = 120, but with the pure convective state solution (that is, the solution for Haz,max=0Ha_{z,max} = 0) as the initial condition. Like the previous runs, this simulation was initially run on a coarse grid of 480×960×30480 \times960 \times30 points for 100 free-fall time units in which it converged to a statistically steady state. Following this, the mesh was successively refined to the resolution of 4800×9600×3004800 \times9600 \times300 grid points and the simulation was allowed to converge after each refinement. Once the simulation reached the statistically steady state at the highest resolution, it was run for a further 21 free-fall time units and a snapshot was saved after every free-fall time unit. Henceforth, we refer to this simulation as IC 2 and the earlier simulation for δ=9\delta= 9 (with the quiescent state being the initial condition) as IC 1.

In Figure 16(a)–(Figure 16c), we show contour plots of vertical velocity uzu_z on the horizontal midplane for IC 2 at three different snapshots corresponding to t=0t = 0, t=9t = 9, and t=18t = 18, respectively; here, tt is measured from the time-stamp of the first snapshot. In Figure 16(d), we show the density plot of uzu_z averaged over 21 free-fall time units. For comparison, we also exhibit the contour plot of vertical velocity on the horizontal midplane for IC 1. In Figure 16(e)–(Figure 16g), the plots correspond to the snapshots at t=0t = 0, t=9t = 9, and t=18t = 18 respectively, where tt is measured from the time-stamp of the first snapshot. Figure 16(h) exhibits the density plot of vertical velocity averaged over 20 free-fall time units. All the figures show that the large-scale patterns are qualitatively similar in the weak magnetic flux region for both IC 1 and IC 2. In the fringe zone, however, there is a difference in the alignment of the superstructures with the reorientation of the patterns perpendicular to the longitudinal sidewalls being more prominent for IC 1. Further, in the region where Haz≈Haz,cHa_z \approx Ha_{z,c}, convection is completely suppressed in the bulk for IC 1 but some traces of convection are still present for IC 2. We remark that minor effects of initial conditions on magnetohydrodynamic flows have already been observed in the past; see, for example, Sukoriansky et al. [48] and Zikanov et al. [65].

Figure 16. Contour plots of vertical velocity uzu_z for δ=9\delta= 9 in the horizontal midplane for different initial conditions. For the simulation IC 2, contour plot of uzu_z at (a) t=0t = 0, (b) t=9t = 9, and (c) t=18t = 18, where tt is measured from the time-stamp of the first snapshot saved at statistically steady state. For the simulation IC 1, contour plot of uzu_z at (e) t=0t = 0, (f) t=9t = 9, and (g) t=18t = 18, where tt is measured from the time-stamp of the first snapshot saved at statistically steady state. Contour plots of time-averaged vertical velocity for (d) IC 2 and (h) IC 1. The positions corresponding to Haz(y)=Haz,cHa_z(y) = Ha_{z,c}, Haz(y)=0.8Haz,cHa_z(y) = 0.8Ha_{z,c}, and the edge of the magnetic poles are represented by purple, blue, and black horizontal lines, respectively, in decreasing order of thickness. The magnetic poles extend from y=16y = 16 to y=∞y = \infty.

We compute the planar kinetic energy (E(y)=0.5⟨ux2+uy2+uz2⟩x,zE(y) = 0.5\langle u_x^2 + u_y^2 + u_z^2\rangle_{x,z}), the local vertical anisotropy parameter (2Ez(y)/Eh(y)2E_z(y)/E_h(y)), the local horizontal anisotropy parameter (Ey(y)/Ex(y)E_y(y)/E_x(y)), and the ratio of the kinetic energies along the zz and yy directions for IC 1 and IC 2. The aforementioned anisotropy parameters are defined in § 3.1. We plot the above quantities versus the local Hartmann number in Figure 17(a)–(Figure 17d). Figure 17(a) shows that the E(y)E(y)-Haz(y)Ha_z(y) curves for IC 1 and IC 2 collapse, implying that the local variations of kinetic energy do not change with the initial conditions. From Figure 17(b)–(Figure 17d), it is clear that although the curves of the anisotropy factors for the two different initial conditions do not collapse, they still follow the same trend. Similar to IC 1, the vertical velocity fluctuations for IC 2 become stronger compared to the horizontal ones with the increase of Haz(y)Ha_z(y). Further, for both IC 1 and IC 2, the velocity fluctuations in the yy-direction become stronger compared to those in the xx-direction as Haz(y)Ha_z(y) increases.

Four-panel plots showing mean profiles of local kinetic energy and anisotropy parameters for IC 1 and IC 2

Figure 17. For δ=9\delta= 9 and different initial conditions: Mean profiles of (a) the local kinetic energy E(y)E(y), (b) the local vertical anisotropy parameter 2Ez(y)/Eh(y)2E_z(y)/E_h(y), (c) the local horizontal anisotropy parameter Ey(y)/Ex(y)E_y(y)/E_x(y), and (d) the ratio of the kinetic energy along zz and yy-directions versus the Hartmann number, Haz(y)Ha_z(y), based on the mean vertical magnetic field Bz(y)B_z(y). For both sets of initial conditions, the variations of E(y)E(y) with Haz(y)Ha_z(y) are identical and the anisotropy parameters increase with the local vertical magnetic field strength.

We plot the normalized local Nusselt number Nuf(y)Nu_f(y) versus the normalized local vertical Hartmann number Haz(y)/Haz,cHa_z(y)/Ha_{z,c} for both sets of initial conditions in Figure 18(a). For the above data, we also plot the normalized Reynolds number Re~(y)\widetilde{Re}(y) versus the local free-fall interaction parameter Nf(y)N_f(y) in Figure 18(b). The quantities Nu~(y)\widetilde{Nu}(y), Re~(y)\widetilde{Re}(y) and Nf(y)N_f(y) are defined in equations (3.5), (3.8), and (3.10), respectively. Figure 18(a,Figure 18b) show that both Nu~(y)\widetilde{Nu}(y) and Re~(y)\widetilde{Re}(y) for the two different initial conditions collapse well, showing that the dependence of the above quantities on the local magnetic field is independent of the initial conditions. The time-averaged global Nusselt and Reynolds numbers for IC 2 are calculated to be Nuglobal=2.60Nu_{global}=2.60 and Reglobal=479Re_{global}=479. These values are nearly identical to those for IC 1 where Nuglobal=2.60Nu_{global}=2.60 and Reglobal=474Re_{global}=474.

Variations of the normalized local planar Nusselt and Reynolds numbers with normalized local vertical magnetic field strengths

Figure 18. Variations of the normalized local planar Nusselt and Reynolds numbers with the normalized local vertical magnetic field strengths. For IC 1 and IC 2: (a) normalized local Nusselt number, Nu~(y)\widetilde{Nu}(y), versus the normalized local Hartmann number, Ha(y)/HacHa(y)/Ha_c, based on vertical magnetic field strength; and (b) normalized local Reynolds number, Re~(y)\widetilde{Re}(y), versus the local free-fall interaction parameter, Nf(y)N_f(y).

Our results show that the although the initial conditions can have some impact on the size and alignment of the convection patterns in the fringe zone, they do not alter the global heat and momentum transport, the local variations of heat and momentum transport, and the trend in the variations of local anisotropy.

References

  1. [1]Ahlers, G., Grossmann, S. & Lohse, D. 2009 Heat transfer and large scale dynamics in turbulent Rayleigh-Bénard convection. Rev. Mod. Phys. 81 (2), 503–537.DOI
  2. [2]Akhmedagaev, R., Zikanov, O., Krasnov, D. & Schumacher, J. 2020a Rayleigh-Bénard convection in strong vertical magnetic field: flow structure and verification of numerical method. Magnetohydrodynamics 56, 157–166.arxiv.org/abs/2002.09404
  3. [3]Akhmedagaev, R., Zikanov, O., Krasnov, D. & Schumacher, J. 2020b Turbulent Rayleigh-Bénard convection in a strong vertical magnetic field. J. Fluid Mech. 895, R4.DOI
  4. [4]Albets-Chico, X., Grigoríadis, D. G. E., Votyakov, E. V. & Kassinos, S. 2013 Direct numerical simulation of turbulent liquid metal flow entering a magnetic field. Fusion Eng. Des. 88, 3108–3124.
  5. [5]Aurnou, J. M. & Olson, P. L. 2001 Experiments on Rayleigh-Bénard convection, magnetoconvection and rotating magnetoconvection in liquid gallium. J. Fluid Mech. 430, 283–307.DOI
  6. [6]Barleon, L., Mack, K.-J. & Stieglitz, R. 1996 The MEKKA-facility a Flexible Tool to Investigate MHD-flow Phenomena. Tech. Rep. FZKA 5821.DOI
  7. [8]Breuer, M., Wessling, S., Schmalzl, J. & Hansen, U. 2004 Effect of inertia in Rayleigh-Bénard convection. Phys. Rev. E 69 (2), 026302.pubmed.ncbi.nlm.nih.gov/14995554
  8. [9]Burr, U. & Müller, U. 2001 Rayleigh–Bénard convection in liquid metal layers under the influence of a vertical magnetic field. Phys. Fluids 13, 3247–3257.DOI
  9. [10]Burr, U. & Müller, U. 2002 Rayleigh–Bénard convection in liquid metal layers under the influence of a horizontal magnetic field. J. Fluid Mech. 453, 345–369.DOI
  10. [11]Busse, F. H. 2008 Asymptotic theory of wall-attached convection in a horizontal fluid layer with a vertical magnetic field. Phys. Fluids 20 (2), 024102.DOI
  11. [12]Busse, F. H. & Clever, R. M. 1983 Stability of convection rolls in the presence of a horizontal magnetic field. J. Theor. Appl. Mech. 2, 495–502.DOI
  12. [13]Chandrasekhar, S. 1981 Hydrodynamic and Hydromagnetic Stability. Oxford: Dover publications.
  13. [14]Chillà, F. & Schumacher, J. 2012 New perspectives in turbulent Rayleigh-Bénard convection. Eur. Phys. J. E 35 (7), 58.DOI
  14. [15]Cioni, S., Chaumat, S. & Sommeria, J. 2000 Effect of a vertical magnetic field on turbulent Rayleigh-Bénard convection. Phys. Rev. Lett. 62 (4), R4520–R4523.pubmed.ncbi.nlm.nih.gov/11089073
  15. [16]Davidson, P. A. 1999 Magnetohydrodynamics in materials processing. Annu. Rev. Fluid Mech. 31 (1), 273–300.DOI
  16. [17]Davidson, P. A. 2017 An Introduction to Magnetohydrodynamics, 2nd edn. Cambridge: Cambridge University Press.
  17. [18]Ecke, R. E., Zhong, F. & Knobloch, E. 1992 Hopf bifurcation with broken reflection symmetry in rotating Rayleigh–Bénard convection. Europhys. Lett. 19, 177–182.DOI
  18. [19]Fauve, S., Laroche, C. & Libchaber, A. 1981 Effect of a horizontal magnetic field on convective instabilities in mercury. J. Physique Lett. 42 (21), L455.DOI
  19. [20]Goldstein, H. F., Knobloch, E., Mercader, I. & Net, M. 1993 Convection in a rotating cylinder. Part 1. Linear theory for moderate Prandtl numbers. J. Fluid Mech. 248, 583– 604.DOI
  20. [21]Goldstein, H. F., Knobloch, E., Mercader, I. & Net, M. 1994 Convection in a rotating cylinder. Part 2. Linear theory for low Prandtl numbers. J. Fluid Mech. 262, 293–324.DOI
  21. [22]Grannan, A. M., Cheng, J. S., Aggarwal, A., Hawkins, E. K., Xu, Y., Horn, S., Sánchez-Álvarez, J. & Aurnou, J. M. 2022 Experimental pub crawl from Rayleigh- Bénard to magnetostrophic convection. J. Fluid Mech. 939, R1.DOI
  22. [23]Grötzbach, G. 1983 Spatial resolution requirements for direct numerical simulation of the Rayleigh-Bénard convection. J. Comput. Phys. 49 (2), 241–264.DOI
  23. [24]Horn, S. & Schmid, P. J. 2017 Prograde, retrograde, and oscillatory modes in rotating Rayleigh–Bénard convection. J. Fluid Mech. 831, 182–211.DOI
  24. [25]Houchens, B. C., Witkowski, L. Martin & Walker, J. S. 2002 Rayleigh–Bénard instability in a vertical cylinder with a vertical magnetic field. J. Fluid Mech. 469, 189–207.DOI
  25. [26]Hurlburt, N., Matthews, P. & Proctor, M. 1996 Nonlinear compressible convection in oblique magnetic fields. Astrophys. J. 457, 933.DOI
  26. [27]Kelley, D. & Weier, T. 2018 Fluid Mechanics of Liquid Metal Batteries. Appl. Mech. Rev. 70, 020801.arxiv.org/abs/1710.03150
  27. [28]King, E. M. & Aurnou, J. M. 2015 Magnetostrophic balance as the optimal state for turbulent magnetoconvection. Proc. Natl. Acad. Sci. U.S.A. 112, 990–994.DOI
  28. [29]King, E. M., Stellmach, S. & Aurnou, J. M. 2012 Heat transfer by rapidly rotating Rayleigh–Bénard convection. J. Fluid Mech. 691, 568–582.DOI
  29. [30]Klüber, V., Bühler, L. & Mistrangelo, C. 2020 Numerical simulation of 3D magnetohydrodynamic liquid metal flow in a spatially varying solenoidal magnetic field. Fusion Eng. Des. 156, 111659.DOI
  30. [31]Krasnov, D., Zikanov, O. & Boeck, T. 2011 Comparative study of finite difference approaches in simulation of magnetohydrodynamic turbulence at low magnetic reynolds number. Comput. Fluids 50 (1), 46 – 59.
  31. [32]Krug, D., Lohse, D. & Stevens, R. J. A. M. 2020 Coherence of temperature and velocity superstructures in turbulent Rayleigh-Bénard convection. J. Fluid Mech. 887, A2.DOI
  32. [33]Lim, Z. L., Chong, K. L., Ding, G.-Y. & Xia, K.-Q. 2019 Quasistatic magnetoconvection: heat transport enhancement and boundary layer crossing. J. Fluid Mech. 870, 519–542.arxiv.org/abs/1902.08754
  33. [34]Liu, W., Krasnov, D. & Schumacher, J. 2018 Wall modes in magnetoconvection at high Hartmann numbers. J. Fluid Mech. 849, R2.DOI
  34. [35]Liu, Y. & Ecke, R. E. 1999 Nonlinear travelling waves in rotating Rayleigh–Bénard convection: stability boundaries and phase diffusion. Phys. Rev. E 59, 4091–4105.DOI
  35. [36]Mistralengo, C., Bühler, L., Smolentsev, S., Klüber, V., Maione, I. & Aubert, J. 2021 MHD flow in liquid metal blankets: Major design issues, MHD guidelines and numerical analysis. Fusion Eng. Des. 173, 112795.
  36. [37]Moreau, R., Smolentsev, S. & Cuevas, S. 2010 MHD flow in an insulating rectangular duct under a non-uniform magnetic field. PMC Phys. B 3, 3.DOI
  37. [38]Nakagawa, Y. 1957 Experiments on the inhibition of thermal convection by a magnetic field. Proc. R. Soc. Lond. 240, 108–113.DOI
  38. [39]Nicoski, J. A., Yan, M. & Calkins, M. A. 2022 Quasistatic magnetoconvection with a tilted magnetic field. Phys. Rev. Fluids 7, 043504.DOI
  39. [40]Pandey, A., Scheel, J. D. & Schumacher, J. 2018 Turbulent superstructures in Rayleigh-Bénard convection. Nat. Commun. 9, 2118.
  40. [41]Peyret, R. 2002 Spectral Methods for Incompressible Viscous Flows. New York: Springer.
  41. [42]Roberts, P. H. 1967 An introduction to magnetohydrodynamics. London: Longmans.
  42. [43]Scheel, J. D., K. Em. & White, K. R. 2012 Thermal and viscous boundary layers in turbulent Rayleigh–Bénard convection. J. Fluid Mech. 711, 281–305.DOI
  43. [44]Schumacher, J. 2022 The various facets of liquid metal convection. J. Fluid Mech. 946, F1.DOI
  44. [45]Schumacher, J. & Sreenivasan, K. R. 2020 Colloquium: Unusual dynamics of convection in the Sun. Rev. Mod. Phys. 92, 041001.DOI
  45. [46]Sterl, A. 1990 Numerical simulation of liquid-metal MHD flows in rectangular ducts. J. Fluid Mech. 216, 161–191.DOI
  46. [47]Stevens, R. J. A. M., Blass, Alexander, Zhu, Xiaojue, Verzicco, Roberto & Lohse, Detlef 2018 Turbulent thermal superstructures in Rayleigh-Bénard convection. Phys. Rev. Fluids 3, 041501(R).
  47. [48]Sukoriansky, S., Zilberman, I. & Branover, H. 1986 Experimental studies of turbulence in mercury flows with transverse magnetic fields. Exp. Fluids 4, 11–16.DOI
  48. [49]Tasaka, Y., Igaki, K., Yanagisawa, T., Vogt, T., Zürner, T. & Eckert, S. 2016 Regular flow reversals in Rayleigh-Bénard convection in a horizontal magnetic field. Phys. Rev. E 93, 043109.DOI
  49. [50]Thess, A., Votyakov, E. V. & Kolesnikov, Y. 2006 Lorentz Force Velocimetry. Phys. Rev. Lett. 96, 164501.pubmed.ncbi.nlm.nih.gov/16712237
  50. [51]Verma, M. K. 2017 Anisotropy in Quasi-Static Magnetohydrodynamic Turbulence. Rep. Prog. Phys. 80 (8), 087001.DOI
  51. [52]Verma, M. K. 2018 Physics of Buoyant Flows: From Instabilities to Turbulence. Singapore: World Scientific.DOI
  52. [53]Verma, M. K. 2019 Energy transfers in Fluid Flows: Multiscale and Spectral Perspectives. Cambridge: Cambridge University Press.DOI
  53. [54]Verzicco, R. & Camussi, R. 2003 Numerical experiments on strongly turbulent thermal convection in a slender cylindrical cell. J. Fluid Mech. 477, 19–49.DOI
  54. [55]Vogt, T., Ishimi, W., Yanagisawa, T., Tasaka, Y., Sakuraba, A. & Eckert, S. 2018 Transition between quasi-two-dimensional and three-dimensional Rayleigh-Bénard convection in a horizontal magnetic field. Phys. Rev. Fluids 3, 013503.DOI
  55. [56]Vogt, T., Yang, J.-C., Schindler, F. & Eckert, S. 2021 Free-fall velocities and heat transport enhancement in liquid metal magneto-convection. J. Fluid Mech. 915, A68.DOI
  56. [57]Votyakov, E. V. & Kassinos, S. C. 2009 On the analogy between streamlined magnetic and solid obstacles. Phys. Fluids 21, 097102.DOI
  57. [58]Votyakov, E. V., Kassinos, S. C. & Alberts-Chico, X. 2009 Analytic models of heterogeneous magnetic fields for liquid metal flow simulations. Theor. Comput. Fluid Dyn 23, 571–578.DOI
  58. [59]Weiss, N. O. & Proctor, M. R. E. 2014 Magnetoconvection. Cambridge: Cambridge University Press.
  59. [61]Yan, M., Calkins, M., Maffei, A., Julien, K., Tobias, S. & Marti, P. 2019 Heat transfer and flow regimes in quasi-static magnetoconvection with a vertical magnetic field. J. Fluid Mech. 877, 1186–1206.
  60. [62]Yanagisawa, T., Hamano, Y., Miyagoshi, T., Yamagishi, Y., Tasaka, Y. & Takeda, Y. 2013 Convection patterns in a liquid metal under an imposed horizontal magnetic field. Phys. Rev. E 88 (6), 063020.
  61. [63]ZHANG, X., VAN GILS, D. P. M., HORN, S., WEDI, M., ZWIRNER, L., AHLERS, G., ECKE, R. E., WEISS, S., BODENSCHATZ, E. & SHISHKINA, O. 2020 Boundary Zonal Flow in Rotating Turbulent Rayleigh-Bénard Convection. Phys. Rev. Lett. 124, 084505.DOI
  62. [64]ZHONG, F., ECKE, R. E. & STEINBERG, V. 1991 Asymmetric modes and transition to vortex structures in rotating Rayleigh–Bénard convection. Phys. Rev. Lett. 67, 2473–2476.pubmed.ncbi.nlm.nih.gov/10044435
  63. [65]ZIKANOV, O., KRASNOV, D., BOECK, T. & SUKORIANSKY, S. 2019 Decay of turbulence in a liquid metal duct flow with transverse magnetic field. J. Fluid Mech. 867, 661–690.
  64. [66]ZIKANOV, O. & THESS, A. 1998 Direct numerical simulation of forced MHD turbulence at low magnetic Reynolds number. J. Fluid Mech. 358, 299–333.DOI
  65. [67]ZÜ RNER, T., SCHINDLER, F., VOGT, T., ECKERT, S. & SCHUMACHER, J. 2020 Flow regimes of Rayleigh-Bénard convection in a vertical magnetic field. J. Fluid Mech. 894, A21.DOI

Paper details

vlnc:W_jt2vtpwq
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents