Dimana ℎ𝑜 adalah koefisien perpindahan panas antara besi dengan air, 𝑙 adalah ketebalan

dinding tabung dan ℎ𝑖 adalah koefisien perpindahan panas antara permukaan besi dan udara.

Metode analitik yang digunakan untuk menyelesaikan koefisien perpindahan panas secara
keseluruhan melibatkan hukum pendinginan Newton untuk menyelesaikan koefisien
perpindahan panas keseluruhan 𝑈;

𝑄̇ = 𝑈 𝐴𝑎𝑣𝑔 ∆𝑇𝑙𝑚

Dimana 𝑄̇ adalah total laju aliran panas melewati dinding tabung menuju air, 𝐴𝑎𝑣𝑔 adalah
perpindahan panas rata-rata pada tabung dan ∆𝑇𝑙𝑚 log mean temperature difference yang
dicari dengan persamaan ;

∆𝑇1 − ∆𝑇2
∆𝑇𝑙𝑚 =
𝑙𝑛(∆𝑇1 /∆𝑇2 )

Penelitian yang dilakukan oleh W.A.Khan et al memberikan koefisien yang berguna untuk
memodifikasi persamaan bilangan Nusselt standar yang ditunjukan pada persamaan ;

ℎ𝑜 𝑑𝑜 1 1
𝑁𝑢𝑜 = = 𝐶 𝑅𝑒 2 𝑃𝑟 3
𝑘𝑤

Dimana 𝑁𝑢𝑜 adalah bilangan Nusselt pada bagian luar tabung, ℎ𝑜 koefisien perpindahan
panas bagian luar, 𝑑𝑜 diameter luar tabung, 𝑘𝑤 konduktivitastermal air, 𝑅𝑒 bilangan Reynold
pada air, 𝑃𝑟 bilangan Prandtl pada air

0,61 𝑆𝑡 0,091 𝑆𝑡 0,053

𝐶=
[1 − 2exp(−1,09 𝑆𝑡)]

An oscilatorry flow can be classified as either a pulsating flow or reciprocating flow. In the former case
the amplitude of the oscillating velocity is less than the time mean flow velocity, for which the flow
direction never reverses. In the latter case the amplitude of the oscillations s greater than time mean
flow, and flow direction reverses periodically. Most biological flows in peristatic pumps are pulsating.
Thus, pulsating pipe flows have been studied axtensively in medical or biological fields. Pulsating flos
also occur in many engineering applications such as the discharge from a piston pump, the flow in an
intake or exhaust manifold of an internal combustion engine, the flow in hydraulic or pneumatic lines
and control systems. On other hand, reciprocating lows occur in many reciprocating motion machines
such as internal combustion engines. In recent years, a great deal of efforts have been devoted to the
development of Stirling engine and pulse tube cryocoolers for military and space applications. For an
optimum design of Stirling cycle machines, it is important to predict accuratelyheat transfer and
frictional loss in their heat exchangers (heaters, regenerators, and coolers). At the present time the
flow frictional lossand heat transfer correlations used in the design of these heat exchangers are based
on undirectional steady flow conditions, which are unrealistic since Stirling machines operate under
reciprocating flow conditions. Oscillatory flows can alsofind applications in other engineering problem
such as electronic coolings and pulse combustor, etc. In this chapter, we shall present a review of
previous studies on oscillatory flows and associated heat transfer, with emphases on frictional loss
and heat transfer characteristics in oscillatory internal and external flows . The objectives are two fold
to gain better insight into oscillatory flow behaviors and oscillatory heat transfer characteristics and
to suggest future research directions in oscillatingflows and associated heat transfer problems.

For the sake of clarity this article will be divided into a number of sections which will be discussed
separately. It will start with the examination of the governing similarity parameters for oscillatory
flows and the associated heat transfer problems. Then, some major findings concerning the
characteristics of osccilatory fluid mechanics, such as the onset of turbulence, velocity distribution,
and frictional loss will be presented, followed by discussion on the associated heat transfer
characteristics including axial heat transfer enhancement and convection heat transfer. Finally salient
findings and major conclusions will be summarized. A similar review of literature on early work on
oscillatory flows and heat/ mass transfer, including chemical species diffusion and chemical species
separation, was presented by Cooper et al

Reciprocating flow

Consider first a reciprocating incompressible flow in a pipe (with diameter D=2a) where the
reciprocating motion of the fluid is driven by a sinusoidal displacer such that the fluid displacement
xm varies according to

In Eq. (1) w is the oscillation angular frequency, is time and xm is the fluid displacement which is
defined by assuming the fluid moves as a plug flow at a cross-sectional mean velocity (u)(t) with the
maximum fluid displacement Xmax. Differentiating Eq.(1) with respect to time gives the cross sectional
mean velocity

Where the maximum cross-sectional mean velocity Umax is related to the maximum fluid
displacement Xmax by

And the crank phase angle O s related to the dimensionless time with i being the number of cycles.
We know define the dimensionless coordinates, time, velocity, pressure and temperature

Where x,r,t,u,v,p and T are the corresponding dimensional quantities, and ro and kf are the density
and thermal conductivity of fluid. The governing dimensionless conservation equations of mass,
momentum and energy for an incompressible reciprocating flow are given by

Where Rew is the kinetic Reynold number with v being the kinematic viscosity of fluid, Ao is the
dimensionless oscillation amplitude of fluid, and Pr is the Prandtl number with af being the termal
diffusitivity of fluid. Equations 4-6 show that Ao, Rew and Pr are the similarity parameters for an
incompressible reciprocating flow and associated heat transfer. For a pipe of finite length, the length
ratio L/D will be an additional parameter which enters the mathematical proble through the inlet and
outlet boundary conditions. It should be noted that some investigators used Rew and the Reynolds
number Re max, as the similarity parameters. The latter is defined as

To show that Remax is also dependent on the oscillation frequency, we rewrite Eq (7) with the aid of
Eq. (3), to give

Therefore, the choice of Rew and Remax as the similarity parameters will not be able to isolate the
effect of the oscillation frequency for the correlation of experimental data. It is noteworthy tha some
previous investigators used the Womersley number

Instead of the kinematic number Rew to describe the effect of the oscillation frequency. In summary
Ao,Rew (or ) Pr and L/D are the similarity parameters for an incompressible reciprocating flow and
associated heat transfer

Flow Regime
The transition from laminar to turbulent oscillatory flows must be understood and characterized
before correlation equations for frictional loss and heat transfer are obtained. In this section we
shall first review previous experimental studies on the onset of turbulence in reciprocating and
pulsating pipe flows. Correlation equations in terms of the appropriate similarity parameters
for the prediction of the onset of turbulence will be presented.

Reciprocating Flow
The onset of turbulence in a reciprocating flow has been given a great deal of attention in the
past [3-15]. In general measurements show that the onset of turbulence in a reciprocating flow
is different from that of a steady flow. For a reciprocating flow, for example, a laminar like
flow exists during the acceleration phase of the half cycle where as a turbulent like flow exists
during certain period in the decelaration phase [3,4]. Figures 2a and 2b show two typical
example of the temporal axial velocity variations of a reciprocating flow at the centerline of
the pipe and near the wall at the
onset of turbulence, measured by a hot wire anemometer [4]. It was found when the values of
Ao and Rew were increased to certain values, periodic turbulent busts occured near the wall
during the deceleration phase of the cycle. It should be noted from these figures that the velocity
fluctuations near the wall are much stronger than in the centerline of the pipe. This finding
implies that the instabilities were generated near the wall and the radial momentum transfer
caused a lower level of velocity fluctuations at the centerline is approached. This is because at
high kinetic Reynolds number the axial velocity near the wall is higher than the centerline
velocity and there exists an inflexion point in the velocity profile near the wall as we shall see
in the subsequent discussion. Thus, if the kinetic Reynold number exceeds a critical value at a
given dimensionless fluid displacement, the fluid flow near the wall may first become unstable
and eddies occur near the wall. These eddies are transferred to the core flow, which cause small
fluctuations.
It is generally agreed in the existing literature that a reciprocating flow becomes unstable with
increasing either the dimensionless fluid displacement Ao or the kinetic Reynolds number Rew.
The critical dimensionless parameter, for which flow transition occurs [4], is

As will be discussed in the next section, the oscillatory boundary layer or the so called Stokes
layer has a thickness of . Equation can be expressed as

Where Re = umax. The physical significance of Eq [17b] is clear. It indicates that transition
to turbulent flow begins when theReynolds number Reg based on the Stokes layerthickness
exceeds a critical value. Critical values B obatained by various investigators are listed in Table
1. The discrepancies of critical values may be attributed to different criteria or measurement
methods employed by various researchers. Some researchers used flow visualization to
determine the onset of turbulencewhile others relied on velocity measurement by hot wire
anemometry or LDA
The related problem of oscillatory heat transfer in a heated pipe subjected to a perriodically
reciprocating flow has important applications to the design of heat exchangers and pulse tubes
in Stirling machines and cryocoolers. Simon and Seume presented a survey of literature on this
subject, and pointed out important difference in fluid flow and heat transfer between steady
and reciprocating flows. Iwabuchi and Kanzaka experimentally investigated heat transfer of
reciprocating flow in a test facility which was designed to obtain data for a specific prototype
engine; they presented their heat transfer data in terms of the piston speed (rpm), mean pressure
and phase difference between the opposing pistons, but did not attempt to correlate their heat
transfer data in terms of dimensionless parameters.
Hwang and Dybbs graphically presented their oscillatory heat transfer data in terms of Nusselt
number versus Reynolds number at selected oscillating amplitudes. Roach and Bell designed
and constructed a test facility for the study of heat transfer in a packed tube under conditions
of rapidly reversing flows. Since their experimental results did not show frequency dependence
in either pressure drop or heat transfer, they correlated the cycle averaged Nusselt number in
terms of the Reynold number as in the case of a steady flow. Wu et al investigated oscillating
flow resistance and heat transfer in a gap heat exchanger of a cryocooler. They found that the
value of the cycle averaged Nusselt number is constant at a value about 9, and that it is
independent of the osccilatory frequency over a wide range of the Reynolds number. They
found that the heat transfer rate increases of the oscillatory frequency and the tidal
displacement. Lee et al presented the results of their analysis on the heat transfer subjected to
a reciprocating laminar flow in a circular pipe by assuming the time averaged wall temperature
being a sinusoidal function of axial distance. Liao at al analyzed the conjugated heat transfer
characteristics of a reciprocating laminar flow between parallel plates with one plate hetaed at
constant heat flux and the other one insulated.
In the following we shall present some studies on forced convection of laminar and turbulent
reciprocating flows in a pipe or duct. The purposes of these studies are to gain physical insight
on heat transfer characteristics in a reciprocating flow and to asses the effects of the oscillation
frequency and the fluid displacement.

Heat Transfer in a Laminar reciprocating Pipe Flow

Constant Wall Temperature
We shall now present a numerical solution for forced convection of a laminar reciprocating
flow in a heated pipe at constant wall temperature, which was obtained by Zhao and Cheng.
Transient temperature profiles near the entrance of the pipe (X=4.5) for Ao =15 Rew = 64
presented in Fig.19. At this kinetic Reynods number, annular effects exist in the velocity
profiles (not shown). It is interesting to note that annular effects also exist in the temperature
profiles of an oscillatory flow as shown Fig.19. This annular effect becomes more pronounced
as the kinetic Reynolds number is increased.

The variations of the local instantaneous Nusselt number Nu along different dimensionless
axial locations of the pipe at Ao =15, Rew =64 and L/D = 40 for a complete cycle are presented
in Fig 20. The solid lines represent locations in the entrance region while the dashed lines
represent locations in the exit region. It should be noted that the phase difference between the
entrance location and the coresponding exit location is O=180. Near the inlet at X=2 for
example, the instantaneous Nusselt number increases with until it reaches a maximum value
around O=90. This is because the colder fluid enters the entrance region with
the cross sectional mean velocity according to Eq. (2) which has a maximum velocity around
O=90. Since the colder fluid enters with a decreasing velocity after O>90 the heat transfer rate
begins to decrease after O>90. The heat transfer rate continues to decrease as the velocity of
the entering fluid decreases to zero at about 180. Subsequently, the fluid reverses its direction
and the warmer fluid passes through the location at X=2; consequently, the heat transfer
between the fluid and the pipe continues to decrease. The value of Nu decreases as the value
of X is increased from the inlet to the mid section of the pipe (X=20)> toward the middle of
the pipe, the instantaneous Nusselt number becomes vanishingly small. Its value is almost
symmetric with respect to O with the maximum value occurring near O = 180

The effects of Ao and Rew on the cycle averaged local Nusselt Number Nux of air along the
axial location are presented in Fig.21. Generally, the cycle averaged local Nusselt number Nux
is symmetrical with respect of the mid section of the pipe because of the symmetrical boundary
conditions for both velocity and temperature for the problem under consideration. A
comparison of case 1 (Ao=20 and Rew =64) and Case 2 (Ao=20 and Rew = 250), shows that
the value of Nux increases with the increase of Rew at a fixed value of Ao, Which implies that
the heat transfer rate increases with the increase of frequency at a fixed oscillation amplitude
of fluid. This is because the thermal boundary layer thickness becomes thinner with the increase
of Re, Consequently, the heat transfer rate increases at higher Rew. Similarly, a comparison of
Case 1 (Ao=20 and Rew=64) and Case 2 (Ao=20 and Rew=250), shows that the value of Nux
increases with the increase of Rew at a fixed value of Ao, which implies that the heat transfer
rate increases with the increase of frequency at a fixed oscillation amplitude of fluid.

This is because the thermal boundary layer thickness becomes thinner with the increase of
Rew. Consequently, the heat transfer rate increases at higher Rew, Similarly a comparison of
Case 2 (Ao=20 and Rew=250) and Case 3 (Ao=35 and Rew=250) shows that the value of Nux
increases with the increase of Ao for a fixed value of Rew, which implies that the heat transfer
rate increases with the increase of oscillation amplitude of fluid at a fixed value of frequency.
This can be explained based on the enrgy equation given in Eq6. With fixed values of Rew and
Pr, the convection term in Eq (6) becomes more significant with the increasing value of Ao,
Physically a higher value of Ao means a larger amount of fluid is heated by the pipe during
each cycle. It can also be observed from Fig.21 that the heat transfer rate becomes vanishingly
small at the middle of the heated pipe (X=20) for smaller Ao because most of the fluid near the
middle of the pipe never exits from the heated pipe.

The effects of Ao and Rew on the instantaneous space averaged Nusselt number Nux for three
different cases are illustrated in Fig.22. It is observed that the space averaged instantaneous
heat transfer rate increases with the increase of either the dimensionless fluid displacement or
the kinetic Reynolds number. It is interesting to compare the temporal variation of the
instantaneous space averaged Nusselt number with the specified sinusoidal variation of the
cross sectional mean flow velocity given by Eq (2). It is found that the phase difference between
the cross sectional mean velocity and the instantaneous space averaged Nusselt number Nu is
about 18 for all cases.

8.12 Uniform Wall Heat Flux

In this section, the experimental and numerical investigations by Zhao and Cheng on oscillatory
transfer in a heated pipe with uniform heat flux subjected to a laminar periodically
reciprocating flow will be discussed. The test setup consisting of two velocity straighteners,
two coolers, two mixing chambers and a heated test section is shown in Fig.23. The test section
was made of a long copper tube, with length L=60,5 cm, inside diameter Di=1,35 cm and
outside diameter Do=1,57 cm. The test section was heated uniformly by a wrapped around
insulated flexible heater. Each end of the heated tube was connected to two identical coolers of
the same inside diameter. Two mixing chambers were inserted between the test section and the
coolers. The reciprocating motion of the fluid in the test section was created by a sinusoidal
motion generator which was designed such that the fluid displacement varies according to Eq
(1).
Figure 24 illustrates typical termporal temperature variations measured at the left and right
mixing chambers during one cycle for Ao=34,9 and Rew=120,1. As shown in this figure, when
the fluid flow moves from the left to right, the fluid temperature in the left mixing chamber
continues to drop untill the fluid low reverses its direcftion at the end of the first half cycle as
shown in the figure. For small values of Ao, however, a different behavior occurs, which will
be be discussed next. The temperature history curves in the left mixing chamber for Ao =15,3
and 34,9 at Rew=120,1 are presented in Fig 25 for comparison. The square symbols represent
the case for Ao =34,9 while the circle symbols represent those for Ao =15,3. For the case of
Ao =34,9, when the fluid reverses its direction at the beginning of the cycle, the fluid
temperature drops immediately due to the cooling effect of the fluid in the cooler which is the
same curve shown in Fig.24. For the case of A=15,3, the fluid temperature increases at the
beginning of the cycle and starts to drop at about 45 of the crank angle.

The reason for different behaviors of temporal temperature variations for low and high values
of Ao can be explained as follows. Consider the inlet fluid temperature at the beginning of the
cycle as the fluid is reversing its direction. The fluid in the mixing chamber is cooled by
convection throughthe fluid coming from the cooler while, at the same time, it is heated by the
fluid residing in the heated section by conduction. AT a high value of Ao (Ao=34,9 for
example) where the maximum cross sectional mean velocity is high (as can be deduced from
Eq 2), forced convection effect predominates. Therefore, the fluid temperature in the mixing
chamber drops immediately when the fluid reverses direction as the colder fluid enters from
the cooler. However, for Ao=15,3 where the fluid velocity is small and axial heat conduction
effect predominates at least initially, the fluid temperature at first begins to increase. Later in
the cycle when the fluid velocity is increased, the fluid temperature starts to drop because the
cooling effect of the fluid from the cooler begins to be felt. Experimental and numerical results
of the cycle averaged wall temperature variation along axial locations in the left hand side of
the test section (X<22,4) for various values of Ao and Rew are presented in Fig.26. Generally,
the cycle averaged wall temperature progressively increases with the axial locations, with the
minimum value occuring at the inlet (i.e left mixing chamber) and maximum value at the mid
section of the heated tube. It can also be observed that the cycle averaged wall temperature
decreases with the increase of either the kinetic Reynolds number Rew or the dimendionless
oscillation amplitude of fluid Ao,

In addition, the numerical solution is in good agreement with the experimental data. It may also
be important to realize from Fig.26 that when the pipe is uniformly heated at a constant heat
flux the variation of the time averaged wall temperature along the axial locations takes on an
exponential shape. This fact suggests that the variation of the cross sectional time averaged
temperature of fluid along the axial locations may take a similar shape. Based on the
experimental data for Ao=8,5 to 34,9 and Rew=23 to 46, Zhao and Cheng(109) obtained the
following correlation equation of the space cycle averaged Nusselt number Nu for a
reciprocating laminar flow of air in a long tube with constant heat flux:

With Tw being the space cycle averaged temperatures measured at the outer surface of the
heated tube and Tm being the cycle averaged temperature at the left mixing chamber. Equation
(74) with experimental data is presented in Fig.27. The space cycle averaged Nusselt number
Nu obtained from the numerical simulation is also presented in Fig.27. for comparison. It is
shown that the numerical solution (represented by the dashed line ) is slightly lower or higher
than Eq (74) depending on whether Rew<280 or Rew> 280. Note that Eq (74) indicates that
the heat transfer rate increases with both the dimensionless oscillation amplitude of fluid Ao
and the kinetic Reynolds number Rew; the increase in heat transfer is more sensitive to Ao than
to Rew because the exponent of Ao is greater than that of Rew.

8.2 Heat Transfer in a Turbulent Reciprocating Duct Flow

Cooper at al (110) experimentally investigated convective heat transfer in a rectangular duct
heated from below subjected to a periodically reciprocating flow. Their test setup included two
major components; a test section and a sinusoidal motion generator which had adjustable
speeds and strokes. The test section was made of a channel with a width of 15,24 cm and
variable heights 2a ranging from 1905 to 7595 cm. The bottom of the test section (with an area
of 6,35x15,24 cm2) was heated by an electric heater and the corresponding heat flux q and the
surface temperature Tw were measured by a heat flux transducer and three thermocouples,
respectively. The heated floor was periodically cooled by a reciprocating flow of fresh air (with
a constant temperature Tf) supplied through the ports at both ends of the test section. The
frequency and the stroke of oscillations were varied such that the range of Rew was from 43 to
684 while the range of dimensionless oscillation amplitude of the fluid Ao was from 23 to 60,
such that the parameter (AoRew) ranging from 603 to 5568. In comparison with typical values
of critical parameter B = ranging from 400 to 800, it may be concluded that Coper et al. Heat
transfer experiments (110) were performed in the transitional and turbulent reciprocating flow
regimes. Based on the experimental data by Cooper et al. (110), we now obtain a correlation
equation of the space cycle averaged Nusselt number Nu for a reciprocatingturbulent flow of
air in a rectangular duct with one side heated:
With Tw being the space cycle averaged temperatures measured at the outer surface of the
heated floor and Tf being the ambient temperature. A comparison of Eq.(76) with experimental
data by Cooper at al (110) is presented in Fig 28. It shown taht Eq(76) (represented by the solid
line) fits well with the experimental data with a standard deviation of 11,1. Note that Eq (76)
is valid in the ranges of 23 <Ao<600 and 43<Rew<684.
The major findings about the convective heat transfer in a reciprocating duct or pipe flow can
be summarized as follows; the numerical results reveal that annular effects also exist in the
temperature profiles of a reciprocating flow in a pipe heated. By constant temperature at high
kinetic Reynolds numbers near the entrance and exit locations of the pipe ii) measurements on
heat transfer in a pipe with uniform heat flux subjected to a laminar reciprocating flow show
that the time averaged wall temperature variation along the axial location takes on an
exponential shape and iii) based on the experimental data, correlation equations for the space
cycle averaged Nusselt number in terms of Ao and Rew have been obtained for oscillatory heat
transfer in laminar and turbulent reciprocating pipe and duct flows.

9 Concluding Remarks
This review has brought together a variety of analytical, numerical, and experimental findings
and conclusions on fluid mechanics and heat transfer characteristics of incompressible internal
and external oscilatorry flows. Salient results are summarized below :
A. An examination of the governing equations and boundary conditions shows that the
governing similarity parameters for fluid flow and heat transfer in a pipe of finite length
subjected to a sinusoidally reciprocating flow are; the kinetic Reynolds number Rew
(or the Womersley number ), the dimensionless oscillation amplitude of fluid Ao, the
length to diameter ratio of tyhe pipe L/D, and the Prandtl number Pr. For the problem
of an incompressible pulsating flow and associated heat transfer, the similarity
parameters are Re, Ao, Rew, Pr and L/D
B. It is generally agreed in the literature that a reciprocating flow becomes unstable with
increasing either the dimensionless fluid displacement Ao or the kinetic Reynolds
number Rew. The onset of turbulence can be described by a critical dimensionless
parameter Bcr = (Ao x Rew) with typical values ranging from 400 to 800. On other
hand, the onset of turbulence in a pulsating flow is governed by three parameters,
 namely the mean flow Reynold number Re, the dimensionless fluid amplitude Ao and
the Womersley number . When the mean flow reynold number is higher than Recr, the
pulsating flow is in the turbulent flow regime. When the mean flow Reynold number is
lower than Recr whether the pulsating flow is laminar or turbulent depending on the
dimensionless fluid amplitude Ao and the Womersley number
C. At the same kinetic Reynolds number Rew and the same dimensionless fluid
displacement Ao, the experimental data show that the cycle averaged friction
coefficiens of a turbullent reciprocating flow is substantially higher than that of a
laminar reciprocating flow. The value of these coefficients decrease as the
dimensionless fluid displacement Ao or the kinetic Reynolds number Rew is increased.
For a laminar pulsating duct flow, the predicted friction coefficient is shown to increase
as the oscillation frequency is increased. The seemingly contradictory results are owing
to different characteristic velocities used in the definitions of the friction coefficients
for a reciprocating flow and pulsating flow. However, the shearing stress increases as
the oscillation frequency is increased for the both the reciprocating flow and the
pulsating flow
D. It has been shown that the axial heat transfer in the presence of an axial temperature
gradient can be greatly augmented by a reciprocating flow. This heat transfer mode
involves the periodic interaction of axial convection with radial heat conduction
between the fluid core and the stokes boundary layer formed by the oscillations. The
effective thermal diffusivity is proportional to the kinetic Reynolds number Rew and
the dimensionless displacement amplitude of fluid Ao, and it also depends on the
thermophysical properties of fluid and solid
E. The numerical results reveal that annular effects also exist in the temperature profiles
of a reciprocating flow in a pipe heated by constant temperature at high kinetic
Reynolds number near the entrance and exit location of the pipe. Measurements on heat
transfer in a pipe with uniform heat flux subjected to a laminar reciprocating flow show
that the cycle averaged wall temperature variation along the axials location takes on an
exponential shape. Correlation equation for the space cycle averaged Nusselt number
in terms of Ao and Rew have been obtained for oscillatory heat transfer in laminar and
turbulent reciprocating internal flows.
F. Heat transfer in a laminar pulsating boundary layer flow depends on a frequency
parameter, the Strouhal number Str at a given the velocity ratio. Both analyses and
measurements show that the heat transfer rate of a pulsating flow is not much different
from that of the mean flow in the ranges of smaller and larger Strx, but it becomes more
significant near the region of Strx=1. The Reynolds analogy for the steady boundary
layer flow and pipe flow is not valid for the pulsating boundary layer flow because the
frequency effect on the skin friction and heat transfer behave in opposite directions. For
a turbulent pulsating boundary layer flow over a heated flat plate, measurements show
that heat transfer is augmented by flow pulsations as compared with the corresponding
mean flow
G. The cycle averaged Nusselt number of a laminar pulsating internal flow can be either
higher or lower than the steady flow value, depending on the frequency. The difference
of the hate transfer rate between a pulsating flow and the corresponding mean flow is
increased as the velocity ratio is increased. The applicability of the quasi steady analysis
on heat transfer in a laminar pulsating duct flow can be justified