Kelvin-Helmholtz

Kelvin-Helmholtz
Instability
Instability
NTCC Term Paper 2018 (STTP 100)
Amity University, Noida
-Aditya Sengupta (A4455717017)
Abstract
 This paper seeks to explore, theoretically, the phenomenon known as Kelvin-Helmholtz Instability.
The study involves a theoretical outlook towards the above-mentioned instability and a detailed
comparison between K-H Instability and other types of Hydrostatic stabilities. A brief mathematical
derivation of the conditions of Kevin- Helmholtz Instability is also included. The report also includes
real word occurrences and certain subtle applications of the said stability (such as in the field of
electromagnetics, plasma and space physics), and also possible future developments in the field of
Hydrostatic stabilities in general and particularly for K-H Instability.

 The objective of the paper is to provide an elementary approach of understanding hydrostatic

stabilities (with a more detailed study of K-H Instability) without ignoring the mathematical treatment
of the K-H Instability.

 This approach is easy to understand and also provides the reader with the ability to test out such a
physical system experimentally.
What is K-H Instability ?
 Kelvin Helmholtz Instability ( or Velocity-Shear Instability) can be simply defined as the instability
produced at the interface of two fluids, of different densities flowing at a different velocities relative to
one another, when a small perturbation (small in amplitude and wavelength) is introduced at the
interface assuming that the gravitational field is non existent in that region ( i.e. g=0) .

 The study of this and other such instabilities is central to the understanding of turbulent and then
eventually provide a suitable universal theory for it.

 K-H instability can be truly understood only by first understanding the meaning of stability and
instability from the perspective of fluid mechanics which becomes the subject matter of linear stability
analysis.
Linear-Stability Analysis
 To understand stability and instability from the perspective of fluid systems we need to first understand them through a
mechanical analogue.

 Fig (a) represents a system of unstable equilibrium because when the ball is subjected to small perturbation dθ, the
perturbation grows with time whereas in fig (b) the perturbation on the ball decays with time. Therefore, any mechanical
system in equilibrium exists in two conditions-

 Stable equilibrium – is one in which the mechanical characteristics of the physical system remain unchanged on being
subjected to small perturbations,

 Unstable Equilibrium – is one in which the characteristics of the system change in time and/or space on being
subjected to small perturbations.
 This is how to stability is pictured in a purely mechanical system and it provides a good reference for visualizing
stability in fluids. In fluids stability of a system is not absolute unlike the mechanical system in fig1.1 where only
one of two possibilities exist – either stable or unstable equilibrium.

 Here, let us assume a sinusoidal perturbation of small interfacial amplitude and wave number is imposed on the
two systems (i.e. at the interface separating the two fluids in the two systems, take the interface to be situated at y
= 0).

 In the system shown in fig (a) the perturbation will decay and eventually die out and the interfacial amplitude will
decrease with time until it reverts to y = 0. This is because the system prefers to stay in a configuration of low
energy where the less dense fluid i.e. oil is located above the high density fluid i.e. water.

 However, in the system shown in fig (b) the perturbation will grow as the interfacial amplitude will grow with time
until every single layer of oil and water exchange their positions.
 Note that the two systems are in equilibrium, but their stability is determined by how the interfacial
amplitude grows/decays over time upon the interface being subjected to a small perturbation. If
the sinusoidal perturbation, here, grows over time then the system is in an unstable equilibrium
and if it decays over time then it is a stable equilibrium. Depending on the nature of the
growth/decay and that of the perturbation, different types of stabilities exist in nature. The one
mentioned in fig.1.2(b) is known as a Rayleigh-Taylor instability.

 The growth rate of the interfacial amplitude depends on the wavelength λ of the sinusoidal
perturbation. In the case above of the Rayleigh-Taylor Instability, it is assumed that the meniscus
can sample al the possible values of λ in the interval [0, ∞), but there exists at least one value of λ
in that interval which gives maximum growth rate.

 Some systems are often stable for some perturbations say λ 1 and λ2 but maybe unstable for other

wavelengths say λ3, λ4, _ _ _ λn, etc. A system can be considered completely stable if and only if it

is stable for all possible values of λ, for all possible kinds of perturbations in nature. However, such
an idealized system is only a figment of imagination and far from possible. Such a system will
qualify as the absolute frame of reference that Newton and Galileo were searching for.

 Hence, the study of behavior (growth/decay) of the interfacial amplitude of a system, against the
wavelength of a small perturbation imposed on the system, to determine whether the system in
question, is stable or unstable is known as Linear Stability Analysis.
Instabilities – A Comparative Study
 Rayleigh-Taylor Instability –

As discussed in Linear Stability Analysis, Rayleigh-Taylor Instability arises due to a form of Adverse
Density Stratification. In other words, a heavier fluid placed on a comparatively lighter fluid. If the interface
is subjected to a small perturbation, say a tap on the beaker containing the two fluids, then the fluids tend
to the more favorable configuration of the lighter fluid placed over the heavier fluid.

 Kelvin Helmholtz Instability –

Also Known as K-H Instability, it arises as a result of a strain generated between two or among more than
two fluids (of different densities) separated by an interface and travelling at different velocities relative to
one another.
The system in, unlike the system in slide 3 represents two fluids with a ‘continuous velocity profile’
where vx = f(y) i.e. the velocity in the x direction is a function of the y-components of the two fluids.
The point at which the velocity profile intersects the X-axis is called the inflexion point. Such a
system is also prone to small interfacial perturbations that result in instabilities of the nature of K-H
Instability.

 Rayleigh Instability –
The Rayleigh Instability, as shown in fig. (a) involves an infinitely long cylindrical column of liquid in
the absence of any external gravitational force. Theoretically in such a system, the system should
remain unchanged in the absence of perturbations, assuming that the internal forces of adhesion
and gravity among the fluid particles are balanced.
However, on being imposed by small perturbations, the different layers of the fluid experience
different surface tension forces leading to Rayleigh Instability as shown in fig. (b).
K-H Instability – A simple Physical Interpretation

The Physics behind the cause of this phenomenon is rather elementary. By virtue of a “velocity
shear/gradient” and also the “equation of continuity” in fluid mechanics the pressure exerted on the
disturbance in region I is more than that in region II and similarly the pressure exerted in region IV is
more than that in region III as shown in Fig. above The equation of continuity states that A 1v1 = A2v2
Therefore, v α 1/A
 
and P α (1/A)
So Pαv
 
Therefore, pressure exerted is more in the region where velocity of fluid is less, as it can be seen in the
figures above.
Mathematical Treatment; Conclusions of the dispersion
relation
The following are the assumptions regarding the system of two fluids shoes in
fig. These assumptions are necessary for using the method of ‘normal mode
linear stability analysis’ to determine the conditions of instability of such a
system.

 Both fluids 1 and 2 are inviscid i.e. γ 1 = γ2 = 0 where, γ1 and γ2 are the
coefficients of viscosity of the fluids 1 and 2 respectively.

 No acceleration due to gravity i.e. g = 0.

 Initially the interface is flat i.e. unperturbed interface at y = 0.

 The fluids are incompressible.

 The flow is two-dimensional flow i.e. our mathematical overview only

involves the use X-axis and Y-axis.
 The equations governing this initial laminar flow are called the Euler equations of laminar flow –

, i =1 (top fluid) and i = 2 (bottom fluid)

The continuity equation is,
= 0, (Incompressibility equation)
Here, represents a velocity field and represents a pressure field in a fluid i. The value i can take
two possible values 1 or 2 in the system shows and both the fluids are governed by the Euler
equation and the continuity equation that represents incompressibility. represents the density of the
fluid i.
In Cartesian form, given = (Ui , Vi), where Ui is the velocity in the x- direction and V i is the velocity in
the y- direction, the Euler equation becomes,
(1)
(2)
(3)
 The equation (1), (2) and (3) represent the Cartesian forms of the Euler equations and these are the
three governing equations of the system depicted in fig. in slide 10 Equation (1) is called the x-
momentum equation, equation (2) is called the y- momentum equation and equation (3) is the
continuity equation following from the incompressibility property of the fluids involved in the system.
It is given that the y- component of the velocity is 0 i.e. = 0 and = constant, so substituting V 1 =V2 = 0
in equations (1) and (2) to check of the flow above is a solution of the Euler equation governing this
system, we get,
from (1)
and from (2)
Both equations above imply that pi = constant

A small sinusoidal interfacial perturbation is imposed and y = 0 is the unperturbed case. This
perturbation causes the following modification to the system variables –
(4a)
and (4b)
i.e. {p = 0}
Where , and are the imposed velocity fields and pressure fields respectively.
  We use the above conditions in equations (1), (2) and (3) and obtain,

(8)
(9)
(10)

 We use the above equations along with the kinematic boundary condition i.e.

i.e.

 And the dynamic condition

Where, Where σ is the interfacial tension and K= - is the curvature.

 And the normal mode assumption to get the following dispersion relation -
  

This equation above is the dispersion relation for the given system. It’s a quadratic in ω for a given
value of k.

If k is real then is complex,

If Re(ω) is greater than 0 then the interfacial amplitude grows and thus the instability persists. But

If Re(ω) is less than 0 then the interface amplitude decays.

 Solving Dispersion Relation-

On solving the dispersion relation in ω using the quadratic formula we get,
 Where the first term represents ωi (imaginary part) and the second term represents ω r (real part). For
the condition of instability as mentioned before, ω r must be greater than 0, such that,

Leaving out k = 0 (trivial case),

 

which is the condition for instability in the given system,

such that,

Or

Where is the cut off value.

CONCLUSION –

 Interchanging fluid 1 and 2 or in other words interchanging 1 and 2 in the above inequation doesn’t
affect the condition for instability. This means that the condition of instability is not governed by a
preference of whether 1 is located over 2 or whether 2 is placed over 1.

 Only the magnitude of relative velocity (U1 – U2) matters and not the absolute velocities of the two
individual fluids taken one at a time.
THANK YOU