Hostname: page-component-7b7567c8c-wxvwp Total loading time: 0 Render date: 2026-11-19T01:34:37.932Z Has data issue: false hasContentIssue false

Interfacial dynamics and energy cascade in immiscible Rayleigh–Taylor turbulence

Published online by Cambridge University Press:  10 August 2026

Dongxiao Zhao
Affiliation:
School of Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, PR China
Xiaoxue Huang
Affiliation:
Department of Physics, National University of Defense Technology, Changsha, Hunan 410000, PR China
Gaojin Li*
Affiliation:
School of Ocean and Civil Engineering, Shanghai Jiao Tong University, Shanghai 200240, PR China
*
Corresponding author: Gaojin Li, gaojinli@sjtu.edu.cn

Abstract

Content of image described in text.

We investigate interfacial dynamics and multiscale energy transfer in immiscible Rayleigh–Taylor turbulence using numerical simulations with varying surface-tension coefficients $\sigma$. Capillarity is shown to control characteristic length scales, interfacial area, and global energy and enstrophy budgets. The flow exhibits self-similar evolution with respect to surface tension, with the maximum kinetic energy scaling as $\sigma ^{1/2}$ and the flow duration as $\sigma ^{-1/4}$. A scale-by-scale budget shows that surface tension removes kinetic energy at large scales while injecting it at small scales, with the crossover occurring near the Hinze scale. We further recast and verify a local kinematic relationship between surface-tension power and interface stretching, up to conservative transport, $ \boldsymbol{f}^\sigma \boldsymbol{\cdot }\boldsymbol{u} = - \sigma \mathcal{S}\,|\boldsymbol{\nabla }c| + \mathrm{transport}$, where $\boldsymbol{f}^{\sigma }$ is the surface-tension force, $\boldsymbol{u}$ is the velocity, $c$ is the heavy-fluid volume fraction, and $\mathcal{S}$ is the interface stretch rate. This relation links kinetic energy transfer to the scalar-variance cascade, and shows that energy transfer to the interface is governed by local strain. Statistics of individual bubbles and droplets reveal vertically elongated filaments with diameters of approximately three capillary scales, yielding a linear volume–area relation. Their vertical velocities scale with the square root of equivalent diameter, consistent with drag–buoyancy balance. These findings, particularly the direct link between surface-tension power and resolved interface stretching, provide a rigorous physical framework for developing subgrid-scale closures for large-eddy simulations of immiscible turbulent flows.

Information

Type
JFM Papers
Copyright
© The Author(s), 2026. Published by Cambridge University Press
Figure 0

Table 1.Parameters of the numerical simulations. For all cases, the densities of the light and heavy fluids are fixed at ρl=0.9, ρh=1.0$\rho _l=0.9,\ \rho _h=1.0$, the density difference is Δρ=ρh−ρl$\Delta \rho =\rho _h-\rho _l$, the mean density is ρm=(ρl+ρh)/2$\rho _m=(\rho _l+\rho _h)/2$, and the kinematic viscosity is spatially constant and fixed at 2.55×10−4$2.55\times 10^{-4}$. The Taylor-scale (λ$\lambda$, defined in (3.4)) Reynolds number is Reλ=u′λ/ν$\textit{Re}_\lambda =u'\lambda /\nu$, the grid Weber number is WeΔ=((ρgL1/3Δx5/3)/σ)(Δρ/(2ρm))$\textit{We}_\varDelta = (({\rho g L^{1/3}\,\Delta x^{5/3}})/{\sigma }) ({\Delta \rho }/({2\rho _m}))$, and the integral scale Weber number is (ρlu′2l)/σ$(\rho_l u^{\prime 2} l)/\sigma$, where u′$u'$ is the root mean square turbulent velocity, and l$l$ is the integral length scale. The Reynolds and Weber numbers are evaluated at the instants of maximum dissipation: approximately t^=5.8$\widehat {t}=5.8$-6.6$6.6$ for cases X1-X150, t^=11.8$\widehat {t}=11.8$ for case X1Ratio4, and t^≈8$\widehat {t}\approx 8$ for cases X1Φ05$\varPhi 05$ and X1Φ95$\varPhi 95$. The column T^max$\widehat {T}_{max }$ denotes the maximum non-dimensional time reached in each simulation. The column Φ$\varPhi$ denotes the volume fraction of the heavy fluid.Table 1 long description.

Figure 1

Figure 1.Visualisations of the density fields. (a–c) Density visualisations for case X1 at non-dimensional times t^=tAtg/Lx=2.7,6,13.6$\widehat {t}=t\,\sqrt {{A_t g}/{L_x}}=2.7, 6, 13.6$, respectively, with corresponding Taylor-scale Reynolds numbers Reλ=17.4,54.0,17.1$\textit{Re}_\lambda =17.4, 54.0, 17.1$. (d–f) Density visualisations at non-dimensional time t^=6$\widehat {t}=6$ for cases X2, X4 and X8, respectively, with corresponding Reλ=58.2,62.1,62.2$\textit{Re}_\lambda =58.2, 62.1, 62.2$.

Figure 2

Figure 2.Figure 2 long description.(a) Linear RT growth rate from (3.1) as a function of perturbation wavenumber k$k$, including viscous and surface-tension effects. (b) Time evolution of the mixing width h(t)$h(t)$; the green, red and blue shaded intervals denote the RT growth, turbulent emulsion and segregation stages, respectively. Dashed lines indicate linear fits to h(t)$\sqrt {h(t)}$ over t^∈[2,5.4]$\widehat {t}\in [2,5.4]$, yielding α=0.034$\alpha =0.034$, 0.043$0.043$ and 0.048$0.048$ for cases X1, X2 and X4 (blue, black and red), respectively.

Figure 3

Figure 3.Temporal evolution of (a) Kolmogorov scale η$\eta$, (b) Taylor microscale λ$\lambda$, (c) Hinze scale lH$l_H$, and (d) interfacial area A$A$. In (c), the horizontal dashed lines corresponds to the capillary scale lc=σ/((ρh−ρl)g)$l_c = \sqrt {{\sigma }/({(\rho _h-\rho _l)g})}$. In (d), solid lines represent areas calculated from simulations X1–X30, while dashed lines indicate the estimated area A^lH$\widehat {A}_{l_H}$ in (3.5) based on the respective Hinze scale.

Figure 4

Figure 4.Figure 4 long description.Kinetic energy budgets for case X1. (a) Mean instantaneous budget terms from (3.6), normalised by ε0=ρluRT3/Lz$\varepsilon _0=\rho _l u_{\textit{RT}}^3/L_z$, with uRT=AtgLz$u_{\textit{RT}}=\sqrt {A_t g L_z}$. (b) Normalised mean surface-tension power and time derivative of the total interfacial area multiplied by σ$\sigma$. (c) Cumulative energy balance, normalised by the total potential energy release; solid and dashed lines denote cases X1 and X30, respectively. (d) The PDFs of the three right-hand-side terms in (3.6) at t^=6$\widehat {t}=6$.

Figure 5

Figure 5.Temporal evolution of mean instantaneous budgets for simulations X1–X30: (a) time derivative of kinetic energy, (b) dissipation rate, (c) energy injection rate, and (d) surface-tension power. Normalisation is identical to that in figure 4(a).

Figure 6

Figure 6.Figure 6 long description.Temporal evolution of cumulative budgets for simulations X1–X30: increments of (a) kinetic energy, (b) internal energy, (c) potential energy, and (d) excess surface energy. Normalisation is identical to that in figure 4(c).

Figure 7

Figure 7.Scaling of the temporal minimum Hinze scale and the maxima of interfacial area, kinetic energy (KE), and dissipation rate with the surface-tension coefficient. The kinetic energy is equal to its increment ΔK$\Delta K$, since the flow starts from rest. In (c,d), the normalisations are KE0=(1/2)ρluRT2$\mathrm{KE}_0=( {1}/{2})\rho _l u_{\textit{RT}}^2$ and ϵ0=ρluRT3/Lz$\epsilon _0 = \rho _l u_{\textit{RT}}^3/L_z$, respectively.

Figure 8

Figure 8.Figure 8 long description.Rescaled cumulative kinetic energy budgets for cases X1–X30. Contributions from (a) kinetic energy, (b) internal energy change, (c) potential energy release, and (d) excess surface energy. Both time and energy are non-dimensionalised following (3.13).

Figure 9

Figure 9.Mean enstrophy budget evolution for (a) case X1, (b) case X2, and (c) case X8. The budgets are defined in (3.14).

Figure 10

Figure 10.Figure 10 long description.Filtering spectra of (a,b) the scalar field and (c,d) the kinetic energy, for (a,c) calculated at t^=6$\widehat {t}=6$, and (b,d) calculated at t^=13.6$\widehat {t}=13.6$. The kinetic energy spectra in (c,d) are normalised to ensure equal total energy across all cases. The filtering wavenumber kℓ$k_\ell$ is defined as kℓ=Lx/ℓ$k_\ell =L_x/\ell$. The dashed lines indicate reference slopes only; owing to the limited scale separation, the spectra are not intended to establish classical inertial-range scaling (Zhou 2007).

Figure 11

Figure 11.Compensated scalar filtering spectra E¯ckℓ2/A$\overline {E}_c k_\ell ^2/A$ at non-dimensional times (a) t^=6$\widehat {t}=6$ and (b) t^=13.6$\widehat {t}=13.6$, where A$A$ is the interfacial area of the corresponding case.

Figure 12

Figure 12.Figure 12 long description.(a) Structure function of the volume-fraction field at t^=6$\widehat {t}=6$ versus increment r$r$. (b) Scale-dependent interfacial dimension inferred from S2c(r)$S_2^c(r)$, DsS2=3−ζ≡3−dlog⁡S2c/dlog⁡r$D_s^{S_2}=3-\zeta \equiv 3-{\rm d}\log S_2^c/{\rm d}\log r$, plotted as a function of k=Lx/r$k=L_x/r$ (solid lines), compared with the box-counting estimate Ds$D_s$ for the c=0.5$c=0.5$ isosurface using boxes of size r$r$ (dashed lines).

Figure 13

Figure 13.Spatial average of the scalar variance transfer term Θℓ$\varTheta _\ell$ versus the filtering wavenumber kℓ=Lx/ℓ$k_\ell =L_x/\ell$ at different time instants indicated by the colour bar. (a–c) Simulation cases X1, X2 and X8, respectively.

Figure 14

Figure 14.Joint PDFs of the scalar variance transfer term Θℓ$\varTheta _\ell$ and the nonlinear model approximation ΘℓNL=−(1/12)ℓ2|∇c¯ℓ|2S¯ℓ$\varTheta ^{\textit{NL}}_\ell =-({1}/{12})\ell ^2\, |\boldsymbol{\nabla }\overline {c}_\ell |^2\, \overline {\mathcal{S}}_\ell$. The data correspond to simulation case X1 at time t^=6.3$\widehat {t}=6.3$, for filtering wavenumbers (a) kℓ=8$k_\ell =8$, (b) kℓ=32$k_\ell =32$ and (c) kℓ=64$k_\ell =64$. The correlation coefficients for these scales are 0.81, 0.94 and 0.97, respectively.

Figure 15

Figure 15.Filtered kinetic energy budget terms from (4.9) for case X1, plotted as functions of the filtering wavenumber kℓ$k_\ell$. The curves correspond to different time instants, colour-coded according to the colour bar. All terms are normalised by the maximum unfiltered injection rate observed over the entire simulation duration.

Figure 16

Figure 16.Time–wavenumber evolution of the scale-by-scale surface-tension power contribution d⟨Ψσ(kℓ)⟩/dkℓ$\text{d}\langle \varPsi _{\sigma }(k_\ell )\rangle /\text{d}k_\ell$ for cases X1, X2 and X8, starting at t^=2.7$\widehat {t}=2.7$ (established turbulence). The filtering wavenumber is shown on a logarithmic scale. Solid lines indicate zero-crossings; dashed lines mark the Hinze scale (3.4).

Figure 17

Figure 17.The PDFs of the alignment cosine between the velocity and surface-tension fields at time t^=6$\widehat {t}=6$, evaluated at the interface, for: (a) large scales (ℓ>Lx/8$\ell \gt L_x/8$); (b) intermediate scales (Lx/64<ℓ$L_x/64 \lt \ell \lt L_x/32$), where the band-pass filtered velocity is u~[ℓ1,ℓ2]≡u~ℓ1−u~ℓ2$\widetilde {\boldsymbol{u}}_{[\ell _1,\ell _2]} \equiv \widetilde {\boldsymbol{u}}_{\ell _1} - \widetilde {\boldsymbol{u}}_{\ell _2}$; and (c) small scales (ℓ$\ell \lt L_x/64$), where the residual velocity is uℓ′≡u−u~ℓ$\boldsymbol{u}'_\ell \equiv \boldsymbol{u} - \widetilde {\boldsymbol{u}}_\ell$.

Figure 18

Figure 18.Alignment between the scalar gradient and the eigenvectors of the strain rate tensor for simulation case X1 at non-dimensional times (a–c) t^=6$\widehat {t}=6$ and (d–f) t^=13.6$\widehat {t}=13.6$, for (a,d) unfiltered fields, (b,e) large-scale fields (scales >Lx/8$\gt L_x/8$), and (c, f) small-scale fields (scales $\lt L_x/64$). The high-pass filter is defined as cℓ′≡c−c¯ℓ$c'_\ell \equiv c-\overline {c}_\ell$, and eℓ′$\boldsymbol{e}_\ell'$ is the eigenvector of the high-pass strain rate tensor.

Figure 19

Figure 19.Colour function fields for cases (a,c) X1Φ95$\varPhi 95$ and (b,d) X1Φ05$\varPhi 05$ at non-dimensional times (a,b) t^=6.5$\widehat {t}=6.5$ and (c,d) t^=12$\widehat {t}=12$. The colormap represents the volume fraction field.

Figure 20

Figure 20.Figure 20 long description.(a) Time evolution of the total number of isolated bubbles and droplets for cases X1Φ95$\varPhi 95$ (labelled as bubbles) and X1Φ05$\varPhi 05$ (labelled as drops). (b) Droplet and bubble size distributions at 5.5$5.5\lt \widehat {t}\lt 6.5$ (early time) and 11$11\lt \widehat {t}\lt 12$ (late time). Two dashed lines indicate reference scalings d−10/3$d^{-10/3}$ and d−3/2$d^{-3/2}$, and lc≡σg(ρh−ρl)$l_c\equiv \sqrt {\frac{\sigma}{g(\rho _h-\rho _l)}}$ is the capillary scale.

Figure 21

Figure 21.(a) Scatter plot of normalised area versus normalised volume for isolated bubbles and droplets at early (5.5$5.5\lt \widehat {t}\lt 6.5$) and late (11$11\lt \widehat {t}\lt 12$) times. For clarity, the red, green and cyan points are vertically shifted by −150, 150 and 300 units, respectively. (b) The mean volume-to-surface ratio is ⟨V/A|d⟩$\langle V/A|d\rangle$, and the mean maximum absolute signed distance is ⟨rLS|d⟩$\langle r_{\textit{LS}}|d\rangle$, conditioned on the equivalent diameter d$d$. All quantities are normalised by the capillary scale lc$l_c$.

Figure 22

Figure 22.Joint PDFs of the middle-to-maximum and minimum-to-maximum moment of inertia ratios in (a,b) the bubble case (X1Φ95$\varPhi 95$), and (c,d) the droplet case (X1Φ05$\varPhi 05$), for (a,c) early time (5.5$5.5\lt \widehat {t}\lt 6.5$), and (b,d) late time (11$11\lt \widehat {t}\lt 12$).

Figure 23

Figure 23.Alignment of principal axes for isolated (a,c) bubbles (case X1Φ95$\varPhi 95$) and (b,d) droplets (case X1Φ05$\varPhi 05$) at late time (11$11\lt \widehat {t}\lt 12$). Here, (a,b) show the alignment with mean vorticity, while (c,d) show the alignment of the major axis with the three coordinate axes. Solid lines show all objects, while dashed lines correspond to long filamentary objects (moment of inertia ratio Imin/Imax<0.15$I_{\textit{min}}/I_{\textit{max}}\lt 0.15$). Here, e^major$\widehat{\boldsymbol e}_{major}$, e^middle$\widehat{\boldsymbol e}_{middle}$, and e^minor$\widehat{\boldsymbol e}_{minor}$ denote the unit vectors along the three principal axes, while i^$\widehat{\boldsymbol i}$, with i=x,y,z$i=x,y,z$, denotes a Cartesian basis vector.

Figure 24

Figure 24.Joint PDFs of the (a,d) mean vertical velocity, (b,e) mean enstrophy, and (c, f) mean squared strain rate averaged over isolated structures, plotted against the equivalent diameter normalised by capillary length, for (a–c) the bubble case (X1Φ95$\varPhi 95$) and (d–f) the droplet case (X1Φ05$\varPhi 05$) at late time (11$11\lt \widehat {t}\lt 12$). Solid black lines denote the conditional means. Purple dashed lines indicate theoretical scalings: the drag–buoyancy balance ∼d1/2$\sim d^{1/2}$ in (a,d), and the dissipation scaling ∼d3/2$\sim d^{3/2}$ in (b,c,e, f).

Figure 25

Figure 25.Figure 25 long description.Scatter plots of surface-tension power versus surface stretching, integrated over individual (a,b) bubbles and (c,d) droplets, for (a,c) early time (5.5$5.5\lt \widehat {t}\lt 6.5$), and (b,d) late time (11$11\lt \widehat {t}\lt 12$).

Figure 26

Figure 26.(a–c) Density visualisations for cases X2, X4 and X8, respectively, at the non-dimensional time t^=13.6$\widehat {t}=13.6$.

Figure 27

Figure 27.Temporal evolution of the Hinze scale for case X1Ratio4 (see table 1). The horizontal dashed line indicates the capillary length scale, lc$l_c$, normalised by the grid spacing Δx$\Delta x$.

Figure 28

Figure 28.Spatial average of the scalar variance transfer density d⟨Θℓ⟩/dlog⁡kℓ$\text{d}\langle \varTheta _\ell \rangle /\text{d} \log k_\ell$ versus the filtering wavenumber kℓ=Lx/ℓ$k_\ell =L_x/\ell$ at different time instants indicated by the colour bar, for (a–c) simulation cases X1, X2 and X8, respectively.

Figure 29

Figure 29.Filtered kinetic energy budgets, differentiated with respect to log⁡kℓ$\log k_\ell$, for case X1 as a function of the filtering wavenumber kℓ$k_\ell$. Curves denote different times, colour-coded by the colour bar. All terms are normalised by the maximum unfiltered injection rate over the simulation.

Figure 30

Figure 30.Compensated coefficient CD$C_D$ as a function of surface-tension coefficient for cases X1–X30.