Fluid Mechanics and Machines (BMEE204L)

Module 6 : Boundary layer flow

Dr. Deepakkumar R
Assistant Professor (Sr.)
Department of Thermal and Energy Engineering
School of Mechanical Engineering
Vellore Institute of Technology, Vellore - 632014,Tamilnadu, India.
Contact: 9003693120, [email protected]
 Contents

• Boundary layers, Laminar flow and turbulent flow,

• Boundary layer thickness,
• Momentum integral equation,
• Drag and lift,
• Separation of boundary layer,
• Methods of preventing the boundary layer separation.
 Boundary layer flow

• The boundary layer phenomenon plays a crucial role in fluid-structure interaction

problems.
• It has wide applications in aerodynamic design of vehicles, aircrafts, rockets,
missiles, submarines, turbomachinery blades, etc..
 CONCEPT OF BOUNDARY LAYER
• The concept of boundary layer was introduced by a German scientist Ludwig
Prandtl (1875–1953 AD) in 1904.
• When a solid body moves on a solid surface, there exists a frictional force between
them, which tries to act as a barrier to the motion.
• A similar phenomenon takes place when a fluid flows over a solid surface; the fluid
motion near the surface is hindered due to the presence of viscous shear forces.
• The fluid velocity at the solid surface is zero, which is known as no slip condition.
• Fluid velocity increases with distance from solid surface and approaches the
magnitude of free stream velocity.
• This variation in velocity is limited in a thin region near the solid surface, known
as boundary layer.
 CONCEPT OF BOUNDARY LAYER
• Figure shows the shape of boundary layer formed on an infinitely long flat plate.
• The plate is kept in a flow stream having free stream velocity u∞.
• Due to no slip condition, the velocity at the solid surface is zero.
• The velocity increases from zero at the surface to the free stream velocity at an infinite distance from the plate.
• In fact, the upper bound for the boundary layer is marked at a distance where the velocity becomes 99% of
the free stream velocity.

Fig. 1 Boundary layer

 CONCEPT OF BOUNDARY LAYER
• It is because of the viscous effects that the velocity inside the boundary layer does
not remain constant.
• The viscous effects are related to the velocity variations by the Newton’s law of
viscosity. (1)
• The dynamic viscosity, μ, is constant for the fluid at a given temperature whereas the
velocity gradient, du/dy, varies within the boundary layer.
• This gradient becomes negligibly small as the velocity reaches the free stream
velocity, that is, du/dy→0 making the flow essentially non-viscous (inviscid) outside
the boundary layer region.
 CONCEPT OF BOUNDARY LAYER

• It can further be seen in Fig. the thickness of the boundary layer increases in the
direction of flow, that is, zero at the leading edge and maximum at the trailing edge.
• In fact, the boundary layer undergoes transition from laminar flow to turbulent flow
if the plate is sufficiently long.
• There are inherent disturbances in the flow, which get amplified due to the
weakening of viscous forces in the increasing thickness of boundary layer in the
downstream direction.
• This transition is quantified by the Reynolds number. For the flow over flat plate it is
defined as (2)
where, x is the distance from the leading edge
 CONCEPT OF BOUNDARY LAYER

• With increase in x, Rex increases and a point is reached in the downstream direction,
where the value of the Reynolds number is equal to the critical Reynolds number,
Recrit (Reynolds number at which the flow turns turbulent).
• To determine whether the boundary layer is laminar or turbulent one has to
compute the value of Rex.

• If it is less than Recrit , the boundary layer is laminar, otherwise it is turbulent.

• The value of Recrit for the flow over flat plate has been experimentally established
as 5×105.
• It should be noted that there is no abrupt transition from laminar to turbulent
flow within the boundary layer.
 CONCEPT OF BOUNDARY LAYER
• The transition region is of finite length. For the sake of convenience, the transition region
is generally clubbed with the turbulent boundary layer region.
• In the downstream of transition region, the turbulent fluctuations cause the mixing of
different fluid layers and as a result boundary layer in the turbulent region further swells.
• Like laminar boundary layer, the thickness of turbulent boundary layer also increases in
the direction of flow.
• It has been observed that within the turbulent boundary layer, there exists an extremely
thin viscous region close to the surface known as viscous (or laminar) sublayer.
• In this layer, the viscous effects are even stronger than those of laminar boundary
layer. The region is so narrow that the velocity profile may be assumed ‘linear’.
 Boundary layer thickness

• The thickness of boundary layer is the perpendicular distance from the solid surface where
the fluid velocity is 99% of free stream velocity, as given in Eq. (3)
(3)
• One of the consequences of boundary layer is flow retardation near the solid surface,
which leads to the reduction in mass, momentum, and energy rates.
• These deficits are quantified by comparing the mass, momentum, and energy rates with the
condition when there is no boundary layer (inviscid flow).
• The deficits in mass, momentum, and energy rates are expressed in terms of displacement,
momentum, and energy thicknesses.
• To find out these thicknesses, an infinitesimal element of thickness dy is considered in the
deficit region as shown in Fig.
 Boundary layer thickness

• The mass, momentum, and energy deficits are calculated through the elemental strip
and then integrated for the whole region

Fig. 2 Flow deficit in boundary layer

 Boundary layer _Displacement Thickness
• The mass flow rate within the boundary layer is less than the mass flow rate in
the absence of a boundary layer. The mass flow rate is retarded by the amount
shown in Eq. (4) (4)
where width is unity in z-direction.
• The displacement thickness is the measure of this mass flow rate deficit, and
physically can be understood as the distance by which a surface is to be displaced
to compensate this deficit in mass flow rate, as shown in Fig. 3.

Fig. 3 Displacement
thickness
Boundary layer _Displacement Thickness

(5)

(6)
Boundary layer _Momentum Thickness
Boundary layer _Momentum Thickness
 Boundary layer_ Energy Thickness

The energy thickness is, therefore, the measure of the deficit in kinetic energy due
to the formation of a boundary layer.
 BOUNDARY LAYER EQUATIONS
• To develop the governing equations for the boundary layer flow over a flat plate
shown in Fig. the order of magnitude technique has been applied to the Navier–
Stokes (N–S) equation.
• According to this technique, the insignificant terms (i.e., the terms with negligible
order of magnitude) are dropped and the governing equations are evolved.

Fig. Boundary layer

over a flat plate
 BOUNDARY LAYER EQUATIONS
• The governing equations for the boundary layer flow have been developed for steady
and incompressible two-dimensional (2-D) flow.
• The continuity equation for steady incompressible 2-D flow in differential form is
(1)

• The N–S equation for steady incompressible 2-D flow:

(2)

(3)

(4)
 BOUNDARY LAYER EQUATIONS

• Equation (4) does not contain gravity term because the flow is taking place in
horizontal direction (or the potential head is same all along the x-axis).
• The first step in the order of magnitude technique is to non-dimensionalize the
parameters in the following manner:

(5)

• Rewriting continuity equation in non-dimensional form:

(6)
 BOUNDARY LAYER EQUATIONS
• Similarly, rewriting N–S equations in non-dimensional form:
(7)

(8)

• As a second step, the order of magnitude for non-dimensional variables is to be

computed. The order of magnitude of non-dimensional parameters, as defined in
Eq. (5), has been shown in Table.
 BOUNDARY LAYER EQUATIONS
• For the other unknown terms the order of magnitude is obtained as follows:
• Applying the order of magnitude approach to the non-dimensional governing
equations:

(9)

• The order of magnitude for y-direction velocity component, v, is insignificant as

compared to x-direction velocity component, u.
• Applying the order of magnitude approach to x-direction N–S equations:

(10)
 BOUNDARY LAYER EQUATIONS
• Following are the conclusions:

• The terms in Eq. (10) must have the same order on LHS and RHS, that is, O(1).

• Equation (7) can be written as

(11)

• Application of order of magnitude approach to y-direction N–S equations gives:

(12)
 BOUNDARY LAYER EQUATIONS

• All the terms in y-direction momentum equation are insignificant as they are of the
order of ∈, which is very small.
• It can be further concluded that the order of pressure gradient in y-direction is ∈,
that is, . This means is negligibly small.

• Further, as in the inviscid region above the boundary layer, one

can apply the Bernoulli’s equation comfortably, that is,

(13)
 BOUNDARY LAYER EQUATIONS
• For the flow over flat plate, the free stream velocity will be constant throughout the
length of the plate.
• Hence, the pressure gradient in x-direction is zero.
• Equation (11) can be reduced to

The Prandtl boundary layer equations for curved as well as flat surfaces are summarized in Table.
 NEAR-EXACT SOLUTION—BLASIUS SOLUTION FOR LAMINAR
BOUNDARY LAYER FLOW OVER FLAT PLATE

• The mass, momentum, and energy deficits are calculated through the elemental strip
and then integrated for the whole region
• boundary layer thickness is given by

• The local coefficient of skin friction is

• An average value of skin friction coefficient is used, which is obtained as

• The drag force experienced by the plate is given by

 APPROXIMATE SOLUTION—VON KARMAN MOMENTUM INTEGRAL EQUATION
• Another method for obtaining an approximate solution for the flat plate laminar
boundary layer problem was proposed by Hungarian–American aerospace engineer
Theodore von Karman (1881–1963 AD), another doctoral student of Ludwig Prandtl.

• In this approach, the differential boundary layer

momentum equation is first converted to
momentum integral equation.
• The velocity profile is assumed as some function
of y and then boundary layer thickness is
calculated by substituting assumed velocity profile
expression in the integral momentum equation.
• The accuracy of this approximate solution is
compared with that of Blasius solution (see Fig.).
 APPROXIMATE SOLUTION—VON KARMAN MOMENTUM INTEGRAL EQUATION

• The velocity profile is,

• The above equations are the approximate solutions for the boundary layer flow over
the flat plate.
• The percentage deviation in the boundary layer thickness and local skin friction
coefficient are −7.2% and −2.71%, respectively, compared to the Blasius solution.
• The choice of velocity profile has a bearing on the accuracy of the solution obtained.
However, the von Karman approach is easier to apply.
 TURBULENT BOUNDARY LAYER
• As soon as the flow turns turbulent on the surface within the boundary layer, an
extremely thin viscous region appears at the surface where turbulent fluctuations
(disturbances) are damped, shown in Fig.1.
• This region is called viscous (or laminar) sub-layer. The region is so thin that the
velocity profile is assumed linear, as shown in Fig.

• From the experiments, it has been found that the thickness of

the viscous sublayer can be expressed as

Fig. Viscous or laminar sub-layer

 TURBULENT BOUNDARY LAYER
• The thickness of the sub-layer depends on the free stream velocity.
• Higher the free stream velocity, stronger will be the wall shear stress and thinner will be
the laminar sub-layer.
• For the sake of better understanding, the turbulent boundary layer may be divided into two
distinct regions - 1 and 2, as shown in Fig.
• The velocity profile in a turbulent boundary layer is
generally 1/7-th of the power law:

• Using von Karman momentum integral equation, we get,

the boundary layer thickness and local skin friction
coefficient
 TURBULENT BOUNDARY LAYER
• Region 1 extends from the surface up to 20% of the total boundary layer thickness
and remaining is the region 2.
• Region 1 has the velocity profile represented by viscous sub-layer near the surface,
the profile above it, that is, in region 2 may be represented by the power law given
by.

• To model the turbulent velocity profile, n = 7 is usually taken. In fact, this law can be
comfortably extended to region 1.
• Generally, the viscous sub-layer region is very thin, the velocity gradient may be
assumed to be infinite. Using Eq., the solution to the turbulent boundary layer can be
obtained
 LIFT AND DRAG
• Figure shows the various forces acting on an airfoil making an angle α (known as
angle of attack) with the uniform flow stream having velocity u∞.

Fig. 1 Flow over an airfoil

 LIFT AND DRAG

• In general, when a fluid flows over a body, it is acted upon by a resultant force R
known as aerodynamic force whose horizontal component parallel to the flow
stream is known as drag D, and vertical component perpendicular to the flow stream
is known as lift L.
• The component of the aerodynamic force acting along the chordline (straight line
joining the leading edge and trailing edge) of the airfoil is known as axial force A
whereas the component of the aerodynamic force perpendicular to the chordline is
termed as normal force N.
 LIFT AND DRAG
• The lift and the drag can be expressed in terms of axial and normal forces, respectively, as

• It has been seen earlier that the pressure does not remain constant when the flow is taking
place over a curved surface.
• The forces, shown in Fig. 1, arise due to the variation of pressure and shear stress along
the surface of the airfoil.
• For the sake of clarity,
pressure and shear
stress distribution
along the surface of an u – upper
l - lower
airfoil is shown in Fig. 2. Fig. 2. Pressure and shear stress variation along the surface of an airfoil
 LIFT AND DRAG
• The following are the normal and axial forces on the upper and lower surfaces of the airfoil:

• Another practical way to find out lift and drag forces is to express them in terms of free
stream velocity, that is,

• where, CL and CD are known as lift and drag coefficients, respectively, and Ap is the projected
area. The lift and drag coefficients are determined experimentally using a wind tunnel.
 Friction Drag and Form Drag
• The drag is of two types, namely, skin friction drag and pressure (or form) drag.
• The skin friction drag is the drag due to the viscous shear acting on the body
within the boundary layer.
• The drag force due to skin friction is obtained by multiplying the wall shear
stress to the surface area of the body up to the point of boundary layer
separation.
• The pressure drag, on the other hand, is the drag induced due to the non-recovery
of pressure in the rear side of the bluff body due to the boundary layer
separation.
• This results in a difference of pressure across the body in the direction of flow.
 Friction Drag and Form Drag
• For better understanding of these two types of drags, Fig. 1 has been drawn.
• The horizontal thin flat plate will be subjected to only skin friction drag (due to the
boundary layer formation on either side of the plate) whereas the vertical thin plate will have
drag only due to pressure difference, that is, the pressure drag, across it.
• Consequently, an inclined plate will experience both skin friction drag and pressure drag.

Fig. 1 Flow over (a) Horizontal plate (b) Vertical plate

 LIFT AND DRAG
• the free stream Reynolds number affects the separation
of boundary layer and the subsequent formation of low
pressure wake, which is responsible for the drag.
• The variation of drag coefficient with Reynolds
number for the flow over cylinder and sphere is shown
in Fig. 2.
• The drag is higher at low Reynolds number due to
higher viscous shear and vice versa.
• With the onset of turbulent boundary layer there is a
sharp drop in the drag due to the delay in Fig. 2 Variation of drag coefficients for the flow
over cylinder and sphere
separation resulting in higher pressure recovery (i.e.,
low pressure drag).
 LIFT AND DRAG
• Unlike the drag coefficient, the lift
coefficient is largely independent of
Reynolds number. However, both lift and
drag coefficients are affected by the angle
of attack and this variation has been
shown for an airfoil in Fig. 5.
• The lift coefficient varies linearly with
the angle of attack whereas the drag
coefficient varies non-linearly with the
increase in angle of attack. For angles
Fig. 5 Variation of lift and drag coefficients with angle
above 15°, the lift coefficient reaches of attack for an airfoil
maximum.
 LIFT AND DRAG
• The landing and take-off of an airplane is governed by changing the angle of attack of
the flaps and slats provided on its wings.
• An important phenomenon associated with airfoils is stalling, which is characterized by a
sudden reduction in lift coefficient and large increase in drag as the angle of attack
attains a large value.
• This happens due to the flow separation from the upper surface of the airfoil leading to the
formation of large low-pressure wake in the rear portion of the airfoil.
• It can also be noticed that the lift is positive non-zero at zero angle of attack.
• In fact, lift becomes zero at some negative angle of attack termed as zero-lift angle of attack,
α0. For higher negative angle of attacks, negative lift coefficient is obtained, which is useful
during the landing of an aircraft.
 Exercise_6.1.a.

Find out the values of constants if velocity within the boundary layer is
represented by the following functions using the appropriate boundary
conditions

In addition, determine the displacement, momentum, and energy

thicknesses
Exercise_6.1.a._Solution
Exercise_6.1.a._Solution
Exercise_6.1.a._Solution
Exercise_6.1.a._Solution
Exercise_6.1.b
Exercise_6.1.b._Solution
Exercise_6.1.b._Solution
Exercise_6.1.b._Solution
Exercise_6.2.
Exercise_6.2_Solution
 Separation of boundary layer
• Figure shows the flow of an ideal fluid over a circular

cylinder.
• When an ideal fluid flows over a curved body (here

cylinder) there exists two stagnation points, that is,

one on the front (S1) and another on the rear (S2).

• At stagnation points, the velocity is zero and the

pressure is maximum.
Fig. Flow past a cylinder (Ideal fluid)
• The area of flow decreases from maximum at point

S1 to minimum at P or Q from any arbitrary

reference plane (shown by dashed line).

 Separation of boundary layer

• The area of flow increases in the same way from point P or Q to stagnation point S2.

• From continuity, the velocity will increase from S1 to P/Q and decrease from P/Q to

S2.
• In accordance to Bernoulli’s principle, the pressure will fall in the front half of the

cylinder and the pressure in the rear half of the cylinder rises by an equal amount.
• Thus, there is a 100% recovery in pressure in the rear half of cylinder.

• The pressure becomes maximum at point S2 (same as the pressure at S1).

• It should be remembered that there will not be any boundary layer formation when

an ideal fluid flows over solid body.

 Separation of boundary layer
• Figure shows the flow of real fluid over the same cylinder

for a considerably high free stream Reynolds number.

• Unlike ideal fluid flow, the flow of real fluid flow will not

result in two stagnation points.

• In fact, there is only one stagnation point at the front, that

is, S1.
• Due to the formation of boundary layer and its subsequent

detachment in the rear half of the cylinder due to the Fig. Flow past a cylinder (Real fluid)

presence of adverse pressure gradient.

• The recovery of pressure is not 100% in the rear half of the

cylinder.
 Separation of boundary layer
• The pressure recovery is only partial and takes place only in the regions between point P

and point of separation S2 in the upper half and between point Q and point of separation

S′2 in the lower half of the cylinder.

• The region between points S2 and S′2 is thus a low pressure region known as wake.

• The wake or separated region consists of eddies or vortices, which are characterized by

swirling of fluid and also by the existence of reverse current generated as the fluid flows

over an obstacle.
• Actually, a long chain of vortices known as von Karman vortex street is formed in the

downstream side of the cylinder.

• Larger the wake, larger is the pressure difference across the cylinder.
 Separation of boundary layer

• If a body is designed in a way that the wake is almost eliminated (i.e., no flow

separation), it is termed as streamlined body.

• The bodies such as cylinder, sphere, or any other body contour where the wake is

formed as a result of flow separation are termed as bluff bodies.

• Such bodies are always acted upon by an additional drag due to the difference in

pressures across them. This type of drag is known as pressure (or form) drag.
• The pressure distribution of flow of an ideal and real fluid past a cylinder is shown

in Fig., (plus sign indicates rise in pressure and minus indicates fall in pressure).
 Separation of boundary layer
• The pressure distribution for the flow of an ideal fluid over a cylinder is symmetric, whereas

the pressure distribution is asymmetric if real fluid flows over a cylinder at relatively higher

Reynolds number.
• Due to the boundary layer separation, a low pressure wake is formed on the downstream side

of the cylinder.
• This inconsistency in the flow of ideal and real fluids is known as d’ Alembert’s paradox.

Fig. Pressure distribution on cylinder for the flow of (a) Ideal fluid (b) Real fluid
 Separation of boundary layer
• For the flow past a cylinder or sphere, the characteristic dimension used in the computation of Reynolds

number is the diameter D, that is,

• For extremely low velocity flows of real fluids (i.e., Re < 4), the fluid follows the contour of body with no

boundary layer separation (Creeping Flow). The flow pattern would be similar to the one shown in Fig.(a).
• The effect of Reynolds number on the flow pattern is shown in Fig. With the increase in free stream Re,

the adverse pressure gradient created will start increasing. The boundary layer separation takes place.

The points of separation move towards the top and bottom of the cylinder.
• Up to Re = 2 × 105, the boundary layer flow remains laminar and separation occurs at an angle of about

80°. Increasing the Re beyond 2 × 105 would see a transition from laminar to turbulent boundary flow

within the boundary layer.

 Separation of boundary layer

• The increased inertia causes the flow within the turbulent boundary layer to remain intact

with the contour of the cylinder in the downstream side and the separation is delayed.
• The wake formed is narrower and the separation occurs at an angle of about 140°.

• In some cases, the formation of

turbulent boundary is desirable as it
helps in the reduction of overall drag.
• For instance, dimples are provided on
the golf ball surface, which make the
flow turbulent within the boundary layer.
• As mentioned earlier, the separation is
delayed resulting in a reduced
pressure drag.
• This helps the golf ball to cover longer Fig. Flow patterns over a circular cylinder
(a) Re < 2 × 105 (b) Re > 2 × 105
distances.
 Methods of preventing the boundary layer separation

It has been observed that the boundary layer separation gives rise to an

additional pressure drag. Hence, the separation is not desirable from

aerodynamics point of view. Following are the techniques employed to

avoid boundary layer separation.

• Streamlining

• Fluid injection

• Suction
 Methods of preventing the boundary layer separation

Streamlining
• Requires the body to be elongated to narrow down the wake region and to minimize

the adverse pressure gradient, as shown in Fig.

• It has been observed that the streamlining reduces the drag coefficient by almost

95%.

Fig. Streamlining of cylindrical body

 Methods of preventing the boundary layer separation

Fluid injection
• Requires blowing of the high energy fluid from the surface at the point where

separation is expected to occur, as shown in Fig.

• This will increase turbulence which, in turn, would increase skin friction.

• The magnitude of rise in skin friction drag is negligible in comparison to the

reduction in pressure drag.

Fig. Streamlining of cylindrical body

 Methods of preventing the boundary layer separation

Suction
• Requires creation of low pressure suction slots on the surface of the body.

• This will remove low-momentum fluid particles near the surface delaying the

transition to turbulent boundary layer formation.

suction slot

Fig. Suction
 POINTS TO REMEMBER

• The viscous effects are limited only to a small region near the solid surface known as boundary layer.

Above the boundary layer, the flow may be treated as inviscid.

• Within the turbulent boundary layer, there is an extremely thin region (viscous or laminar sub-layer)

near the solid surface where the viscous effects are tremendously high.

• The boundary layer separation occurs only due to the existence of adverse pressure gradient.

• Since there is no variation in pressure along the solid flat surface, the boundary layer remains intact

with the surface. Hence, there is no boundary layer separation.

• In case of curved bodies, the pressure is not fully recovered and the body experiences a drag due to

the difference in pressure between its front and rear portions.

• Larger the wake size, larger will be the pressure drop across a non-flat body and higher will be the

pressure drag.
 POINTS TO REMEMBER

• Reynolds number governs the flow separation on curved bodies such as

cylinders or spheres.
(a) At very low Reynolds number (Re < 4), there is no flow separation as the adverse pressure

gradient in the rear side is negligibly small due to ultralow free stream velocity.

(b) For Re < 2 × 105, the flow separates at an angle of about 80° for flow over a cylinder.

(c) For Re > 2 × 105, the flow separates at an angle of about 140° for flow over a cylinder.

• Airfoil is said to be stalled when there is a sudden drop in lift coefficient

at a large angle of attack.

• Lift is usually non-zero at zero angle of attack for an airfoil.