Hostname: page-component-7b7567c8c-2n9c9 Total loading time: 0 Render date: 2026-11-17T09:20:02.511Z Has data issue: false hasContentIssue false

Unstable miscible displacements in radial flow with chemical reactions

Published online by Cambridge University Press:  26 April 2021

Min Chan Kim
Affiliation:
Department of Chemical Engineering, Jeju National University, Jeju 63243, Republic of Korea
Satyajit Pramanik*
Affiliation:
Department of Mathematics, Indian Institute of Technology Gandhinagar, Palaj, Gandhinagar 382355, India
Vandita Sharma
Affiliation:
Department of Mathematics, Indian Institute of Technology Ropar, Rupnagar 140001, India
Manoranjan Mishra
Affiliation:
Department of Mathematics, Indian Institute of Technology Ropar, Rupnagar 140001, India
*
†Email address for correspondence: satyajit.pramanik@iitgn.ac.in

Abstract

The effects of the $A + B \rightarrow C$ chemical reaction on miscible viscous fingering in a radial source flow are analysed using linear stability theory and numerical simulations. This flow and transport problem is described by a system of nonlinear partial differential equations consisting of Darcy's law for an incompressible fluid coupled with nonlinear advection–diffusion–reaction equations. For an infinitely large Péclet number ($Pe$), the linear stability equations are solved using spectral analysis. Further, the numerical shooting method is used to solve the linearized equations for various values of $Pe$ including the limit $Pe \rightarrow \infty$. In the linear analysis, we aim to capture various critical parameters for the instability using the concept of asymptotic instability, i.e. in the limit $\tau \rightarrow \infty$, where $\tau$ represents the dimensionless time. We restrict our analysis to the asymptotic limit $Da^{\ast }$ $(= Da \tau ) \rightarrow \infty$ and compare the results with the non-reactive case ($Da = 0$) for which $Da^{\ast } = 0$, where $Da$ is the Damköhler number. In the latter case, the dynamics is controlled by the dimensionless parameter $R_{Phys} = -(R_{A} - \beta R_{B})$. In the former case, for a fixed value of $R_{Phys}$, the dynamics is determined by the dimensionless parameter $R_{Chem} = -(R_{C} - R_{B} - R_{A})$. Here, $\beta$ is the ratio of reactants’ initial concentration and $R_{A}$, $R_{B}$ and $R_{C}$ are the log-viscosity ratios. We perform numerical simulations of the coupled nonlinear partial differential equations for large values of $Da$. The critical values $R_{Phys, c}$ and $R_{Chem, c}$ for instability decrease with $Pe$ and they exhibit power laws in $Pe$. In the asymptotic limit of infinitely large $Pe$ they exhibit a power-law dependence on $Pe$ ($R_{Chem, c} \sim Pe^{-1/2}$ as $Pe \rightarrow \infty$) in both the linear and nonlinear regimes.

Information

Type
JFM Papers
Copyright
© The Author(s), 2021. Published by Cambridge University Press
Figure 0

Figure 1.Schematic diagram of a reactive system $A + B \rightarrow C$ considered here.

Figure 1

Figure 2.Base-state concentrations $a_0(\zeta ), b_0(\zeta )$ and $c_0(\zeta )$ along with $\phi _0(\zeta )$ are plotted in the limiting case of $Da^{\ast } \rightarrow \infty$ for different values of $\beta < 1$, $\beta = 1$ and $\beta > 1$: (a) $\beta = 0.5$, (b) $\beta = 1$ and (c) $\beta = 1.5$.

Figure 2

Figure 3.Neutral stability curves obtained using spectral analysis (7-term and 9-term approximations) and numerical shooting method for $R_{Phys}^{\ast } = 0$, $Pe \rightarrow \infty$, $Da^{\ast } \rightarrow \infty$, (a) $\beta = 1/2$, (b) $\beta = 1$ and (c) $\beta = 2$. The region above (below) each curve corresponds to the region of instability (stability). The values of $|R_{Chem}^{\ast }|$ and $k$ corresponding to the lowest point of the curve representing the numerical shooting method denote the critical values for the instability. These are denoted by $|R_{Chem, c}^{\ast }|$ and $k_c$ and are shown by horizontal and vertical dashed-dotted lines, respectively.

Figure 3

Figure 4.Variation of the critical parameter $R_{Chem, c}^{\ast }$ with $R_{Phys}^{\ast }$ obtained from spectral analysis for $Da^{\ast } \rightarrow \infty$, $Pe \rightarrow \infty$ and different values of $\beta$. For $\beta = 1$, positive and negative $R_{Chem, c}^{\ast }$ are identical. For $\beta \neq 1$, positive and negative $R_{Chem, c}^{\ast }$ are not only different, but also have a non-trivial dependence on the sign of $R_{Phys}^{\ast }$.

Figure 4

Figure 5.Linear stability results obtained for purely chemical systems ($R_{Phys} = 0$) with $\beta = 1$ and $Da^{\ast } \rightarrow \infty$ by solving (3.20) and (3.24) (for $Pe \gg 1$) using the numerical shooting method. Dependence of the critical parameters (a) $R_{Chem, c}$ and (b) $n_c$ on $Pe$. The straight lines correspond to the asymptotic behaviour of these critical parameters in the limit of $Pe \rightarrow \infty$ obtained from the spectral analysis discussed in § 3.2.2.

Figure 5

Figure 6.Effects of the domain boundary on concentration distribution of the product $C$ at $\tau = 1$ for $Pe = 3000$, $Da = 100$, $\beta = 1$, $R_{Phys} = 0$ ($R_{A} = R_{B} = 0$), and $R_{Chem} (= -R_{C}) = -7$ (a) and $R_{Chem} (= -R_{C}) = 7$ (b).

Figure 6

Figure 7.Temporal evolution of the concentration distribution of the product $C$ for $Da = 100$, $Pe = 3000$, $\beta = 1, R_{A} = R_{B} = 0$, and $R_{C} = 7$ (a) and $R_{C} = -7$ (b).

Figure 7

Figure 8.Effects of Péclet number $Pe$ on the concentration distribution of the product $C$ at $\tau = 1$ for $Da = 100$, $\beta = 1, R_{A} = R_{B} = 0$, and $R_{C} = 7$ (a) and $R_{C} = -7$ (b).

Figure 8

Figure 9.Concentration of product $C$ at $\tau = 1$ for $Pe = 3000$, $Da = 100$, $\beta < 1$, $\beta = 1$, $\beta >1$, and different values of $R_{Phys} = - (R_{A} - \beta R_{B})$ with positive and negative $R_{Chem} = -(R_{C} - R_{B} - R_{A})$. (a) For $R_{Chem} = -7$, we choose $(R_{A}, R_{B}, R_{C}) = (2, 0, 9), (0, 0, 7)$ and $(-2, 0, 5)$ for $R_{Phys} = -2$, $R_{Phys} = 0$ and $R_{Phys} = 2$, respectively. (b) For $R_{Chem} = 7$, we choose $(R_{A}, R_{B}, R_{C})=$$(2, 0, -5)$, $(0, 0, -7)$ and $(-2, 0, -9)$ for $R_{Phys} = -2$, $R_{Phys} = 0$ and $R_{Phys} = 2$, respectively.

Figure 9

Figure 10.Evolution of interfacial length for $\beta = 0.5, 1$ and $1.5$, $Da^{\ast } = 100$, $Pe = 3000$, and (a) $R_{Phys} = 2$, $R_{Chem} = 7$, (b) $R_{Phys} = -2$, $R_{Chem} = 7$, (c) $R_{Phys} = 2$, $R_{Chem} = -7$ and (d) $R_{Phys} = -2$, $R_{Chem} = -7$. The dash-dotted lines represent the corresponding stable displacement for $R_{Phys} = 0$, $R_{Chem} = 0$. The maiden deviation of each continuous line from the corresponding dash-dotted line denotes the onset of instability for the respective values of $\beta$.

Figure 10

Figure 11.Dependence of the critical parameter (a) $R_{Phys, c}$ on $Pe$ for the non-reactive system ($Da = 0$) and (b) $|R_{Chem, c}|$ on $Pe$ for pure reactive systems ($R_{Phys} = 0$) with $\beta = 1$. In the reactive cases, LSA is performed for $Da^{\ast } \rightarrow \infty$; NLS are performed with $Da = 100$. For the reactive cases, the straight lines correspond to $\sim Pe^{-0.50}$.

Figure 11

Figure 12.Dependence of the critical parameter $R_{Chem, c}$ on $\beta$ captured from LSA for $Pe \rightarrow \infty$ and $Da^{\ast } \rightarrow \infty$, as well as NLS for fixed $Pe = 3000$ and $Da = 100$. For both LSA and NLS, we choose $R_{Phys} = 0$.

Figure 12

Figure 13.Dependence of the critical parameter $|R_{Chem, c}|$ on $Pe$, calculated for $Da^{\ast } \rightarrow \infty$ from LSA for different values of $\beta = 0.5$ (circle), $1.0$ (down-pointing triangle), $1.5$ (square), $2.0$ (up-pointing triangle). Open symbols correspond to positive $R_{Chem, c}$, while filled symbols correspond to negative $R_{Chem, c}$. The inset shows the collapse of the data on a single master curve that follows a power law: $|R_{Chem, c}| \sim P\textrm {e}^{-0.50}$.

Figure 13

Figure 14.Dependence of the critical parameter $|R_{Chem, c}|$ on $Pe$, calculated for $Da = 100$ from NLS for different values of $\beta = 0.5$ (circle), $1.0$ (down-pointing triangle), $1.5$ (square), $2.0$ (up-pointing triangle). Open symbols correspond to positive $R_{Chem, c}$, while filled symbols correspond to negative $R_{Chem, c}$. The inset shows the collapse of the data on a single master curve that follows a power law: $|R_{Chem, c}| \sim Pe^{-0.55 \pm 0.01}$.

Figure 14

Figure 15.Concentration fields of the product $C$ for the parameters $Pe = 3000, Da = 100, R_A = R_B = 0$. In the upper panel, $R_C = 7, \tau = 10$; in the lower panel, $R_C = -7, \tau = 5$. The variable time step backward differentiation formula was employed and relative tolerance was set to $10^{-4}$.

Figure 15

Figure 16.Concentration fields of the product $C$ for the parameters $Pe = 3000, Da = 100, R_A = R_B = 0$. In the upper panel, $R_C = 7, \tau = 10$; in the lower panel, $R_C = -7, \tau = 5$. The number of elements used in these simulations is 375 484.

Figure 16

Table 1.Detailed explanation of mesh structures employed for the case of $Pe=3000, Da=100, R_A=R_B=0$ and $R_C=7$.

Figure 17

Table 2.Time steps used in the case of $Pe=3000, Da=100, R_A=R_B=0$ and $R_C=7$. The number of element used is 375 484 (fine).