Development of strongly nonlinear structures at the charged boundary of a non-conducting liquid in an electric field
N.M. Zubarev, E.A. Kochurin · 2026
Read Development of strongly nonlinear structures at the charged boundary of a non-conducting liquid in an electric field by N.M. Zubarev, E.A. Kochurin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-19468.txt.
Introduction
N. M. Zubarev and E. A. Kochurin simulate what a charged liquid surface does when an electric field pushes it hard. A non conducting liquid with charge on its free surface, set in a field along the surface normal, grows unstable. The team follows the strongly nonlinear stages of that growth by direct simulation.
Two stages stand out: first, dimples form on the surface; then the pattern develops into sharper structures. Knowing the stages helps read real systems, from inks and sprays to liquid metal devices, where charged surfaces meet strong fields.
Catalog record
| CREATOR | N.M. Zubarev, E.A. Kochurin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 23 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-19468
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Development of strongly nonlinear structures at the charged boundary of a non-conducting liquid in an electric field N.M. Zubarev, E.A. Kochurin Abstract Direct numerical simulation of the strongly nonlinear stages of instability development for a non-conducting liquid with a charged free surface in a normal electric field is performed. It is demonstrated that two main stages of the instability can be distinguished: an initial stage, during which dimples appear on the surface, and a developed stage, during which these dimples transform into expanding bubbles. The bubble size increases with increasing applied field, despite the fact that the scale corresponding to the dominant mode of the instability decreases. An electric field directed perpendicular to the boundary of a liquid with a surface charge has a destabilizing effect on it. The Tonks-Frenkel instability z1 ; z2 , which arises at the surface of a perfectly conducting liquid (e.g., molten metal) in a sufficiently strong external electric field, has been studied most thoroughly. In this case, the field does not penetrate the liquid due to screening by the free electric charge induced on its surface. The opposite situation is also possible, which is the subject of this paper, when an electric field is present in the liquid but absent outside it. This could be a non-conducting liquid with a free surface electric charge, such as liquid helium with electrons localized above its surface z3 ; zz3 , or the interface between a non-conducting liquid and a significantly lighter conducting medium. An example of the latter situation is the boundary between a dielectric liquid and a plasma during the ignition of partial electric discharges in gas bubbles rising in the liquid z4 . The linear stages of instability development of a flat liquid boundary in both situations (when the field is present only outside the liquid or only inside the liquid) are identical: they are described by the same dispersion law \omega^{2}=\left(\alpha/\rho\right)k^{3}-\left(\varepsilon_{0}\varepsilon E^{2}/\rho\right)k^{2}, (1) where \omega is the frequency, k is the wavenumber, \alpha is the surface tension coefficient, \rho is the liquid density, \varepsilon_{0} is the electric constant, E and \varepsilon are the field strength and permittivity outside the liquid (in the first case) or in the liquid (in the second case). It is interesting to compare the nonlinear stages of instability development, i.e., the stages at which the surface inclination angles cease to be small. At these stages, the difference between the situations under discussion will become evident, namely that the electric field is present on different sides of the boundary. Since the instability of the liquid boundary results from the interaction of the applied field and the surface charge (the charge is drawn in by the field), surface disturbances directed outward in the first case and inward in the second will grow more rapidly. This determines the tendency for protrusions (bumps) to form on the surface of a conducting liquid and depressions (dimples) for a non-conducting liquid. What will be the behavior of the fluid at strongly nonlinear stages of instability? Some clues can be given by the general properties of the original equations of motion, which are also reflected in the dispersion law ( 1 ). The equations are invariant with respect to the dilations E\to cE , \lambda\to c^{-2}\lambda , \tau\to c^{-3}\tau (here c is an arbitrary dimensionless constant), i.e., the characteristic spatial ( \lambda ) and temporal ( \tau ) scales of the problem change with the field following the power laws E^{-2} and E^{-3} , respectively. Indeed, a key role in the development of instability is played by the so-called dominant mode corresponding to the maximum of its growth rate, i.e., the extremum \mbox{Im}(d\omega/dk)=0 in the dispersion law. As can be easily found from ( 1 ), we have k_{d}=2\varepsilon\varepsilon_{0}E^{2}/3\alpha and \gamma_{d}=\mbox{Im}\,\omega(k_{d})\propto E^{3} for the wavenumber and the growth rate of the dominant instability mode. Thus, the characteristic scales of the problem are related to the field E as \lambda\propto k_{d}^{-1}\propto E^{-2} and \tau\propto\gamma_{d}^{-1}\propto E^{-3} . Moreover, it was shown in z5 ; z6 ; zz6 ; zzz6 for the Tonks-Frenkel instability that the process of its development is scale-invariant: self-similar solutions corresponding to the scaling \lambda\propto\tau^{2/3} are realized. They are responsible for the formation of conical cusps on the liquid surface. The dynamics of the liquid near the forming cones, i.e., at strongly nonlinear stages of instability, will be determined not by the external but by the local field E_{loc} . The relations for the scales \lambda\propto E_{loc}^{-2} and \tau\propto E_{loc}^{-3} are then valid linking the unlimited sharpening of the surface in a finite time (the quantities \lambda and \tau become zero) and the growth of the local field. As will be shown below, these considerations, which allow us to adequately describe the behavior of a conducting liquid in an electric field, are not applicable to the case of a non-conducting liquid, which is of interest to us. Contrary to the estimate for the dominant mode scale \lambda_{d}=2\pi/k_{d}\propto E^{-2} , the characteristic size of the forming structures increases with increasing E , i.e., they are determined by other factors. This qualitatively distinguishes the compared situations despite the identical descriptions in the linear approximation and the same scaling. Let us consider a two-dimensional plane potential flow of an ideal incompressible fluid with a free boundary y=\eta(x,t) in a uniform electric field of strength E . In the unperturbed state \eta=0 , i.e., the fluid occupies the half-plane y\leq 0 . Here, a Cartesian coordinate system \{x,y\} is used, for which the y -axis is codirectional with the applied electric field, and the x -axis coincides with the unperturbed boundary. The fluid is assumed to be non-conducting (dielectric) with permittivity \varepsilon . A free surface charge is distributed along its boundary ensuring its equipotentiality. We assume that the surface charge density in the unperturbed state is \sigma=-\varepsilon_{0}\varepsilon E , which guarantees complete screening of the electric field above the fluid. The potentials of fluid velocity \phi and electric field \varphi in the region y<\eta(x,t) satisfy the Laplace equations \nabla^{2}\phi=0 and \nabla^{2}\varphi=0 . On the liquid surface y=\eta(x,t) , the dynamic and kinematic boundary conditions are satisfied: \phi_{t}+\frac{1}{2}(\nabla\phi)^{2}=\frac{\alpha}{\rho}K-\frac{\varepsilon_{0}\varepsilon}{2\rho}(\nabla\varphi)^{2}, \eta_{t}+\eta_{x}\phi_{x}=\phi_{y}, where K=\eta_{xx}(1+\eta_{x}^{2})^{-3/2} is the local curvature of the boundary. The equations of motion are closed by the conditions \nabla\phi\to 0 and \nabla\varphi\to\{0,-E\} at the depth y\to-\infty , and by the condition \varphi=0 at the boundary y=\eta(x,t) . In the case of harmonic perturbations of the surface \eta\propto\exp(ikx-i\omega t) of small amplitude (i.e. \left|\eta_{x}\right|\ll 1 ), these equations lead to the dispersion law ( 1 ). We will assume that the functions \phi , \varphi , and \eta are periodic along the x -axis with some given period \lambda_{0} . We will switch to dimensionless notation, using the corresponding wavenumber k_{0}=2\pi/\lambda_{0} and frequency \omega_{0}=(\alpha k_{0}^{3}/\rho)^{1/2} : \tilde{y}=yk_{0},\quad\tilde{x}=xk_{0},\quad\tilde{\eta}=\eta k_{0}, \tilde{t}=t\omega_{0},\quad\tilde{\phi}=\phi k_{0}^{2}/\omega_{0},\quad\tilde{\varphi}=\varphi k_{0}/E. In dimensionless form, the dispersion relation ( 1 ) takes the form (here and below, the “ \sim ” signs are omitted for convenience): \omega^{2}=k^{3}-\beta^{2}k^{2}, where the dimensionless parameter \beta^{2}=\varepsilon_{0}\varepsilon E^{2}/\alpha k_{0} is introduced. Since \beta\propto E , it can be considered to characterize the strength of the applied electric field. The parameter \beta can also be interpreted in terms of the characteristic scales of the problem. Two scales were