6.

6 Heat and Momentum Transfer in Turbulent Flow

Turbulent flow is called so by its random and rapid fluctuations of groups of fluid particles,
called eddies, throughout the boundary layer. It allows additional mechanism for momentum and
heat transfer. The transverse motion of eddies affects great momentum and heat transfer.
The eddying motion in turbulent flow affects significant fluctuations in the value of
velocity, temperature, pressure, and density (compressible flow). This instantaneous velocity, u
can be measured with hot-wire anemometer probe or other sensitive device. The value of
instantaneous velocity can be expressed as the sum of a mean value of u and fluctuating velocity,
u’:

Mean value of u is determined by averaging it over a time interval which is sufficiently

large so that the net effect of fluctuations is zero. The value of u’ is usually few percent of mean
value of u, but the high frequencies of eddies (about thousand per second) makes a good chance
for momentum and thermal energy transport. For steady turbulent flow, the mean values of
properties are time independent.
Consider a case of upward eddy motion of a fluid furing flow on a surface. The mass flow
rate of fluid per unit area normal to flow is ρv’. With note: h = CpT as the energy of the fluid, T’
is eddy temperature relative to the mean value, q dott = ρCpv’T’. From momentum transfer
equation, turbulent shear stress can be expressed as:

Note that the mean of u’v’ value does not equal to 0 even though mean value of u’ = 0 and
v’ = 0, and experimental results show that the mean of u’v’ is negative quantity. The formula above
are called Reynold stresses.
Turbulent wall shear stress and turbulent heat transfer can be expressed:

where μt is turbulent viscosity and kt is turbulent thermal conductivity. Total shear stress and total
heat flux can be expressed as:
6.7 Derivation of Differential Convection Equations
Conservation of Mass Equation
In steady flow, the mass amount and control volume remains constant, so the conservation
of mass can be expressed as:

Consider that mass flow rate is equal to the product of density, mean velocity, and cross-
sectional area normal to flow, thus the rate of fluid enters the control volume from left surface is
ρu(dy · 1). Therefore the rate of fluid leaves control volume from right surface can be expressed
as:

By substituting to conservation of mass we obtain:

and we can simplify it by dividing dx · dy · 1 become:

Conservation of Momentum Equations

Newton’s second law of motion for the control volume is:

or

where the mass of fluid element within the control volume is:

Consider the flow is steady and two-dimensional and u = u(x,y), the total differential of u is:
Thus the acceleration of fluid element in x-direction becomes:

The net surface force acting in x-direction becomes:

The simplifications below are known as the boundary layer approximations. These
approximations are used and greatly simplify the analysis usually with little loss in accuracy and
helping to obtain analytical solutions for certain types of flow problems.

When gravity effects and other body forces are negligible and the boundary layer
approximations are vaild, using Newton’s second law of motion on volume element for y-direction
thus y-momentum equation can be expressed as:

Noting that variation of pressure in the direction normal to the surface is negligible and thus P =
P(x) and doP/dox = dP/dx.

Conservation of Energy Equation

Consider that energy can be transferred by heat, work, and mass only, the energy balance
for a steady-flow control volume can be expressed as:
 Since energy is a scalar quantity, energy interactions in all directions can be combined in
one equation. Noting the mass flow rate of fluid enters the control volume from left surface is
ρu(dy·1), the energy transfer rate to the control volume by mass in x-direction is:

Repeating this equation for y-direction and combine with x-direction, the net rate of energy
transfer to the control volume becomes:

The net rate of heat conduction in x-direction is:

Repeating this equation for y-direction and combine with x-direction, the net rate of energy
transfer to the control volume by heat conduction becomes:

Then energy equation for steady two-dimensional flow of fluid with constant properties
and negligible shear stressed by substituting equations 6-32 and 6-34 onto 6-30 becomes:

If the shear stresses are not negligible, then we count on 6-35 becomes:

where the viscous dissipation function, Φ, expressed by:

6.8 Solutions of Convection Equations for A Flat Plate

 Consider laminar flow fluid over a flat plate, which surfaces are slightly contoured such
turbine blades can be assumed as flat plates with reasonable accuracy. The fluid approaches the
plate in the x-direction with a uniform upstream velocity, which is equivalent to the free stream
velocity uambient.
When viscous dissipation is negligible, the continuity, momentum, and energy equations
(Eqs. 6-21, 6-28, and 6-35) reduce for steady, incompressible, laminar flow of a fluid with constant
properties over a flat plate to:

with boundary conditions of:

6.9 Nondimensionalized Convection Equations and Similarity

When viscous dissipation is negligible, the continuity, momentum, and energy equations
for steady, incompressible, laminar flow of a fluid with constant properties are written on equations
6-21, 6-28, and 6-35. We can nondimensionalize these equations by dividing all dependent and
independent variables by relevant and meaningful constant quantities: all lengths by L which is
plate length, all velocities by reference velocity which is free stream velocity for a plate, pressure
by twice of free stream dynamic pressure for a plate, and temperature by temperature difference
of Tambient – Ts for a plate. Therefore:

We can simplify to:

with boundary conditions:

 The continuity and momentum equations were first solved in 1908 by the German engineer
H. Blasius, a student of L. Prandtl. This was done by transforming the two partial differential
equations into a single ordinary differential equation by introducing a new independent variable,
called the similarity variable. He defined a dimensionless similarity variable as:

He then introduced a stream function ψ(x, y) as

He then defined a function f(η) as the dependent variable as

Then the velocity components become:

By differentiating these u and v relations, the derivatives of the velocity components can be shown
to be:

Substituting these relations into the momentum equation and simplifying, we obtain:

Using the definitions of f and η, the boundary conditions in terms of the similarity variables can
be expressed as

6.10 Functional Forms of Friction and Convection Coefficients

Solution of u* can be expressed as:
Then the shear stress on surface becomes:

Thus the local friction coefficient becomes:

T* can be expressed as:

Using T* equation, the convection heat transfer coefficient becomes:

We substitute this to the Nusselt number relation becomes:

The average friction and heat transfer coefficients are determined by integrating Cf,x and
Nux over the surface of the given body with respect to x* from 0 to 1. Integration will remove the
dependence on x*, and the average friction coefficient and Nusselt number can be expressed as

6-11 Analogies Between Momentum and Heat Transfer

Reconsider the nondimensionalized momentum and energy equations for steady,
incompressible, laminar flow of a fluid with constant properties and negligible viscous dissipation.
We can simplify momentum and energy equations become:

which are exactly of the same form for the dimensionless velocity u* and temperature T*. The
boundary conditions for u* and T* are also identical. Therefore, the functions u* and T* must be
identical, and thus the first derivatives of u* and T* at the surface must be equal to each other,
Then from equations 6-69, 6-70, and 6-73 we have:

which is known as Reynolds analogy. Or also can be expressed as:

where:

is the Stanton number, which is also a dimensionless heat transfer coefficient.

Reynolds analogy is of limited use because of the restrictions Pr = 1 and doP*/dox* = 0 on
it, and it is desirable to have an analogy that is applicable over a wide range of Pr. This is done by
adding a Prandtl number correction. The friction coefficient and Nusselt number for a flat plate
are determined in Section 6-8 to be:

Taking their ratio and rearranging give the desired relation, known as the modified Reynolds
analogy or Chilton–Colburn analogy,

for 0.6 < Pr < 60. Here jH is called the Colburn j-factor.

Ayu Monica D. D. S.
003201400003
Chapter 6 Summary
Heat and Mass Transfer II