Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/233556520

Shape Optimization of a Dimpled Channel to Enhance Turbulent


Heat Transfer

Article  in  Numerical Heat Transfer Applications · December 2005


DOI: 10.1080/10407780500226571

CITATIONS READS

60 335

2 authors, including:

Kwang-Yong Kim
Inha University
689 PUBLICATIONS   7,270 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Double-layer Transverse-flow microchannel heat sink View project

Grooved microchannel heat sink View project

All content following this page was uploaded by Kwang-Yong Kim on 21 November 2014.

The user has requested enhancement of the downloaded file.


Numerical Heat Transfer, Part A, 48: 901–915, 2005
Copyright # Taylor & Francis Inc.
ISSN: 1040-7782 print=1521-0634 online
DOI: 10.1080/10407780500226571

SHAPE OPTIMIZATION OF A DIMPLED CHANNEL TO


ENHANCE TURBULENT HEAT TRANSFER

Kwang-Yong Kim and Ji-Yong Choi


Department of Mechanical Engineering, Inha University, Incheon,
Republic of Korea

Numerical optimization of a dimpled channel has been carried out to enhance the turbulent
heat transfer. The response surface-based optimization is used as an optimization technique
with three-dimensional Reynolds-averaged Navier–Stokes analysis. Computational results
for heat transfer rate show good agreement with the experimental data. The objective func-
tion is defined as a linear combination of heat transfer- and friction loss-related terms with a
weighting factor. Twenty-seven training points obtained by full factorial designs for three
design variables construct a reliable response surface. In the sensitivity analysis, it is found
that the objective function is most sensitive to the ratio of dimple depth to dimple print dia-
meter. Optimal values of the design variables have been obtained in a range of the weighting
factor.

1. INTRODUCTION
To enhance convective heat transfer in internal flow passages, for example, of
turbine blades, a number of techniques, such as ribs, pin-fins, dimples, etc., are used,
as reviewed by Ligrani et al. [1]. Among these, dimples have low-pressure drop pen-
alties since there is no protrusion into the flow to produce significant pressure drag.
The physical mechanisms for heat transfer augmentation with an array of dimples [2]
are the reattachment of the shear layer that forms across the top of each dimple, the
vortex structures and vertical fluid shedding from each individual dimple inden-
tation, and the periodic unsteadiness which is produced as flow is ejected and then
inrushes to each dimple.
Many parametric studies have been performed experimentally for dimpled sur-
faces. Mahmood and Ligrani [3] investigated the heat transfer performance for chan-
nel height-to-dimple print diameter ratios 0.2, 0.25, 0.5, and 1.0, with the dimple
depth-to-dimple print diameter ratio 0.2 and also with the dimple print diameter-
to-distance between dimples ratio 0.874. Mahmood and Ligrani reported that the
heat transfer is augmented as the ratio of channel height to dimple print diameter
increases. On the other hand, it was found in the experiments of Burgess and Ligrani
[4] that the heat transfer rate also increases as the ratio of dimple depth to dimple

Received 18 March 2005; accepted 13 May 2005.


This work was supported by an INHA University Research Grant.
Address correspondence to Kwang-Yong Kim, Department of Mechanical Engineering, Inha Uni-
versity, 253 Yonghyun-Dong, Nam-Gu, Incheon, 402-751, Republic of Korea. E-mail: kykim@inha.ac.kr

901
902 K.-Y. KIM AND J.-Y. CHOI

NOMENCLATURE

Ad, Ain areas of heat transfer surface and inlet S distance between dimples
plane, respectively T local mean temperature
D dimple diameter b
T periodic component of temperature
Dh channel hydraulic diameter Ui mean velocity components (i ¼ 1, 2, 3)
f friction factor Ub averaged axial velocity at inlet
F objective function x, y, z streamwise, spanwise, and cross-
H channel height streamwise coordinates, respectively
Nu local Nusselt number xi coordinate (i ¼ 1, 2, 3)
Nua average Nusselt number b weighting factor in objective function
p,Dp pressure and pressure drop in a c pressure gradient in streamwise
channel, respectively direction
b
p periodic component of pressure n kinematic viscosity
Pi dimple pitch q fluid density
q0 wall heat flux r increasing rate of bulk temperature in
Re Reynolds number (¼UbDh=n) axial direction

print diameter increases. Burgess and Ligrani observed the change in heat transfer
rate for dimple depth-to-dimple print diameter ratios 0.1, 0.2, and 0.3 with the
channel height-to-dimple print diameter ratio 1.0 and the dimple print diameter-
to-distance between dimples ratio 0.874. Ligrani et al. [5] analyzed the effects of
turbulence intensity on heat transfer and friction loss. It was reported that the
increase in friction loss is proportional to turbulence intensity but does not affect
the heat transfer.
Patrick and Tafti [6] calculated computationally the flow structure and heat
transfer in a channel with a staggered array of dimples using direct numerical simula-
tion in the laminar to early turbulent regime. They found that the thermal efficiency
using dimples is significantly higher than with other heat transfer augmentation tech-
niques being employed currently. Isaev and Leont’ev [7] used Reynolds-averaged
Navier–Stokes analysis to investigate the vortex structure over the dimple surface.
Park et al. [8] predicted turbulent flows in a channel with dimples. Their steady-state
results show centrally located vortex pairs and vortex pairs located near the spanwise
edges of individual dimples. These numerical studies, however, do not report on the
effect of the geometric parameters on thermal performance of dimpled channels.
Also, in the experimental works mentioned earlier, researchers tried to find the opti-
mal case for better thermal performance among only a few tested cases. Thus, the
optimal shape of dimpled channels, considering wide ranges of geometric variables,
has not yet been suggested. However, for the case of rib-roughened channels, a series
of investigations has been carried out by Kim and Kim [9–11] to find the optimum
shapes of the channel by optimizing the geometric design variables with numerical
optimization techniques.
Shape optimization of dimpled surfaces for enhancement of turbulent heat
transfer should be based on precise analysis of the flow structure. With the aid of
high-performance computers, numerical optimization techniques based on
Reynolds-averaged Navier–Stokes analysis have been developed rapidly in the last
decades. Among the methods of numerical optimization, the response surface-based
optimization method [12], being a global optimization method, has many advantages
SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 903

over the gradient-based methods [9, 13]. Recently, response surface-based optimiza-
tions have been applied to many single- and multidisciplinary optimization problems
[14, 15].
In this work, a numerical optimization procedure is presented for the design of
a dimpled channel with a single surface roughened. Turbulent convective heat trans-
fer is analyzed with Reynolds-averaged Navier–Stokes analysis. The response sur-
face method is employed to maximize thermal performance of the surface with
three geometric design variables.

2. NUMERICAL ANALYSIS
For the calculation of fluid flow and convective heat transfer in a dimpled
channel, CFX-5.7 [16], which employs unstructured grids, is used in this work.
To adopt periodic boundary conditions, modifications of source terms in
streamwise momentum and energy equations have been made to calibrate the grad-
ual decrease and increase of pressure and temperature, respectively. Finally, for
three-dimensional steady incompressible flows, mass, momentum, and energy con-
servation equations in tensor form can be written as follows:

qUi
¼0 ð1Þ
qxi
 
qUi q qUi 1 qb
p
Uj ¼ n  þ cd1i ð2Þ
qxj qxj qxj q qxi

!
q  
b ¼ q
b
qT
qcp Uj T k  rUj d1j ð3Þ
qxj qxj qxj

where b b ðx; y; zÞ are the pressure and temperature transformed as


pðx; y; zÞ and T
follows in order to use the periodic boundary conditions [17] in the streamwise
direction, x.
b
pðx; y; zÞ ¼ pðx; y; zÞ þ cx ð4Þ

b ðx; y; zÞ ¼ Tðx; y; zÞ  rx
T ð5Þ

Here, c is the pressure gradient along the streamwise direction, and r is the rate of
bulk temperature increase due to wall heat flux, q0.

q0 A d
r¼ ð6Þ
PiUb Ain

The shear stress transport (SST) turbulence model with automatic wall treat-
ment [18] is used as a turbulence closure. Basically, the SST model combines the
advantages of the k–E and k–x models with a blending function. The k–x model is
904 K.-Y. KIM AND J.-Y. CHOI

activated in the near-wall region, and the k–e model is used in the rest of the region.
Bardina et al. [19] showed that the SST model captures separation under adverse
pressure gradient well compared to other eddy viscosity models, and thus predicts
well the near-wall turbulence that plays a vital role in the accurate prediction of tur-
bulent heat transfer. The numerical model of Lai et al. [20] is adopted for modeling
of turbulent heat flux.
The geometric parameters and computational domain are shown in Figure 1.
The computational domain is composed of a single dimple with periodic boundary
conditions on the surfaces normal to the streamwise direction, and also with sym-
metric conditions on the surfaces normal to the cross-streamwise direction. In
Figure 2, an example of the grid system is shown for the whole computational
domain and also near the dimple wall. An unstructured tetrahedral grid system is
used with the hexahedral at the wall region to resolve the high-velocity gradient.
In the present calculation, uniform heat flux is specified on the dimpled surface
and channel wall between the dimples. Bulk velocity corresponding to Reynolds

Figure 1. Geometry and design variables.


SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 905

Figure 2. Example of the grid system.

number based on hydraulic diameter, 5,000, and constant temperature are set at all
computational nodes as initial values to help the faster convergence of the iterative
calculation.
With the periodic conditions, it is difficult to assign specific flow rate to the cal-
culated flow. Thus, an iterative procedure is inevitably employed. At the initial stage of
the calculation, the pressure gradient in the streamwise direction is assumed through
the source term in the streamwise momentum equation, and is continually updated
until the Reynolds number reaches within 1% of the target Reynolds number.

3. OPTIMIZATION TECHNIQUES
The response surface method (RSM) [12] involves the collective use of design of
experiment techniques, regression analysis, and analysis of variance. In the present
work, the RSM is used to obtain the optimal shape of dimpled channels. The opti-
mization problem is defined as minimization of an objective function, F(x) with
xil  xi  xiu, where x is a vector of design variables, and xil and xiu are lower and
upper bounds of each design variable, respectively. The RSM performs a series of
numerical analyses for a prescribed set of design points, and constructs a response
surface using the calculated quantities over the design space.
A second-order polynomial has been used to represent the response surface,
and the coefficients of the polynomial have been determined by the least-squares
method. A prescribed set of design points, so-called training points, is selected by full
factorial design [12].
The dimpled channel with a single surface roughened is shown in Figure 1.
With five geometric parameters, i.e., channel height (H), dimple print diameter
906 K.-Y. KIM AND J.-Y. CHOI

(D), dimple depth (d ), distance between dimples (S), and dimple pitch (Pi), there exist
four dimensionless variables; H=D, d=D, D=S, and S=Pi. In the present optimization,
S=Pi was set to 1.0, to reduce the number of design variables. Therefore, three design
variables, such as H=D, d=D, and D=S, were selected as design variables in the
optimization.
To maximize the performance of dimples, the optimal shape should be determ-
ined by a compromise between the enhancement of heat transfer and reduction in the
friction loss. After the definition proposed by Gee and Webb [21], average Nusselt
number and 1=3 power of friction factor became indexes for representing the thermal
performance, as has been used in many experimental works [22], and also in the present
optimization. These two objectives are combined with a weighting factor which is
frequently adopted in multiobjective optimizations [9–11]. Therefore, the present
optimization problem is defined as minimization of the following objective function:

F ¼ FNu þ bFf ð7Þ

where b is the weighting factor. This factor should be determined by the designer con-
sidering the overall energy economy of the system. The first term on the right-hand side
is defined as the inverse of the average Nusselt number,
1
FNu ¼ ð8Þ
Nua
where
R
A ðNu=Nus Þ dA
Nua ¼
A

Nus ¼ 0:023 Re0:8 Pr0:4

Nus is the Nusselt number obtained from the Dittus-Boelter correlation, which is for
fully developed turbulent flow in a smooth pipe, and the integration is performed
over the heated surface (A).
The second term, which is related to friction loss, is defined as
 1=3
f
Ff ¼ ð9Þ
f0

where

Dp Dh
f ¼
2qUb2 Pi

f0 ¼ 2ð2:236 ln Re  4:639Þ2

f0 is a friction factor for fully developed flow in a smooth pipe, and is obtained from
Petukhov’s empirical correlation [23], which is modified from the Karman-Nikuradse
correlation for the best fit in the range 104 < Re < 106.
SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 907

4. RESULTS AND DISCUSSION


For validation of the present numerical solution, results of the average Nusselt
number on a dimpled surface are compared with the experimental data of Bunker
et al. [24] at several different Reynolds numbers with H=D ¼ 0.9, d=D ¼ 0.227,
and D=S ¼ D=Pi ¼ 0.825, in Figure 3, where the computational results agree well
with the experimental data. Grid dependency of the numerical solution has been
tested to determine the optimum number of grids in the earlier stage of this research.
The number of grids varies with the shape of dimples from 150 to 180,000.
In the present optimization, Reynolds number based on hydraulic diameter
with 25C air is 5,000. Uniform heat flux is imposed on the dimpled surface, and
the flat wall opposite the dimpled surface is assumed to be adiabatic. For the opti-
mization, response surface-based optimization is used and the 27 training points are
selected by full factorial design to construct the response surface. Ranges of design
variables are listed in Table 1.
Numerical optimization has been performed in the range 0.0–0.1 of the weight-
ing factor. To measure uncertainty in the set of coefficients in the polynomial, analy-
sis of variance (ANOVA) and regression analysis provided by t-statistic [12] are
implemented.
Figures 4 and 5 show results of sensitivity analyses for two components of the
objective function. Here, the percent change of each design variable, dv, is varied
within 10% of the optimal value, and the subscript ‘‘opt’’ represents the value at
optimal shape for b ¼ 0.06. From these sensitive analyses, it is evident that the heat

Figure 3. Comparison between predicted and measured Nusselt number distributions.


908 K.-Y. KIM AND J.-Y. CHOI

Table 1. Design variables and ranges

Design variable Lower bound Upper bound

H=D 0.2 1.0


d=D 0.1 0.4
D=S 0.5 0.9

transfer and friction coefficients are most sensitive to d=D, and insensitive to H=D,
which is confirmed by the experiment performed by Burgess and Ligrani [4]. The
results of sensitivity analysis for the objective function with b ¼ 0.06 are shown in
Figure 6. It is found that the objective function is most sensitive to the ratio of dim-
ple depth to dimple print diameter, as expected from Figures 4 and 5. Figure 5 shows
that the friction loss-related component of the objective function is far from the
minimum at the optimum shape, unlike the case of the heat transfer-related term
shown in Figure 4.
Results of optimization for b ¼ 0.06 are shown in Table 2. The reference shape
is the same as the shape tested by Bunker et al. [24]. The average Nusselt number for
the optimal shape is more than twice that of the reference shape. Also, the value of
the friction loss-related term (Ff) is smaller than that of the reference shape. It is
noted that the decreases in H=D and D=S are remarkable in comparison with the ref-
erence shape. This is consistent with the experimental results obtained by Mahmood

Figure 4. Sensitivity analysis of FNu for optimal shape (b ¼ 0.06).


SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 909

Figure 5. Sensitivity analysis of Ff for optimal shape (b ¼ 0.06).

Figure 6. Sensitivity analysis of objective function for optimal shape (b ¼ 0.06).


910 K.-Y. KIM AND J.-Y. CHOI

Table 2. Results of optimization for b ¼ 0.06

Design variable

H=D d=D D=S Nua Ff F

Reference 0.90 0.23 0.83 2.03 5.46 0.82


Optimum 0.41 0.24 0.55 4.38 4.23 0.48

and Ligrani [3], who reported that heat transfer rate increases as H=D ratio
decreases. Finally, the objective function is reduced to 58.5% of the value for
the reference shape.
Local Nusselt number contours on dimpled walls are shown in Figure 7. In
both cases, heat transfer rate decreases abruptly just behind the front rim of the dim-
ple, due to flow separation, and increases downstream to reach a maximum near the
rear rim of the dimple. The distributions are not symmetric due to the asymmetric
flow structure. Isaev and Leont’ev [7] reported that the turbulent flow past deep dim-
ples (d=D ¼ 0.22–0.24) becomes asymmetric under conditions of the absence of a
performed symmetric vortex structure. Figure 8 shows the flow structure in the opti-
mal dimple. Flow separation occurs at the front rim of the dimple, and a recircula-
tion zone appears in the upstream portion of the dimple. In most of the recirculation
region, heat transfer rate is lower than that at the upstream of the dimple, as shown
in Figure 7. It is found in Figures 7 and 8 that the heat transfer rate increases rapidly
on the rear part of the dimple surface because of the flow reattachment, and this con-
tinues until the rear rim of the dimple is reached. This is also confirmed by Figure 9,
where local Nusselt number distributions along the dimple diagonal in the stream-
wise direction are compared for reference and optimum shapes. The short vertical
lines on the curves in Figure 9 indicate the dimple locations. It is obvious that the
heat transfer rate reaches a maximum near the rear rim of the dimple, and decreases
rapidly downstream. Although the maximum level of local Nusselt number of the
optimum shape is lower than that of the reference shape, the optimum shape gives
the higher value in most of the region except downstream of the dimple. On the other
hand, the distributions of local friction factor are compared in Figure 10. Upstream
and downstream of the optimum dimple, the friction factor is reduced largely with
the optimum shape. Friction factor also reaches a maximum at the rear rim of the
dimple, and the level depends significantly on the dimple shape.
Optimal values of three design variables are plotted for different weighting fac-
tors in Figure 11. With the increase of weighting factor, in other words, as the
designer’s purpose is shifted to the reduction of pressure drop, d=D and D=S
decrease, but H=D increases. Since the friction loss, i.e., the pressure drop across
the dimple decreases as the dimple depth is reduced, the optimal value of d=D moves
toward the smaller value with the increasing of the weighting factor. The optimum
value of H=D increases with the weighting factor, because friction loss decreases
as the channel height is increased.
The computations were carried out using a personal computer, Intel Pentium
IV CPU, 3.0 GHz. The computational time for a single flow analysis is in the range
of 6–8 h, depending on the geometry considered and the convergence rate.
SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 911

Figure 7. Nusselt number contours on dimpled surfaces (b ¼ 0.06).


912 K.-Y. KIM AND J.-Y. CHOI

Figure 8. Streamlines on cross section including streamwise dimple diagonal for optimal shape (b ¼ 0.06).

5. CONCLUSIONS
Shape of a dimpled channel has been optimized to maximize the performance
of heat transfer by the response surface-based optimization method coupled with
Reynolds-averaged Navier–Stokes analyses of fluid flow and heat transfer. Calcu-
lated averaged Nusselt numbers show reasonable agreement with the experimental
data for several different Reynolds numbers. The objective function was defined
in order to maximize the performance of the dimples by compromising between aug-
mentation of heat transfer and reduction in friction loss with a weighting factor.
Twenty-seven training points selected by full factorial design for three design

Figure 9. Nusselt number distributions along the streamwise dimple diagonal for reference and optimum
shapes (b ¼ 0.06) (short vertical lines indicate the dimple locations).
SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 913

Figure 10. Friction factor distributions along the streamwise dimple diagonal for reference and optimum
shapes (b ¼ 0.06) (short vertical lines indicate the dimple locations).

Figure 11. Variations of optimal design variables with weighting factor.


914 K.-Y. KIM AND J.-Y. CHOI

variables construct a reliable response surface. It is found that both the heat transfer-
and friction loss-related components of the objective function are most sensitive to
the ratio of dimple depth to dimple print diameter. The optimal values of design vari-
ables were obtained by varying the weighting factor. As the weighting factor
increases, in other words, as the design emphasis is shifted toward reduction in fric-
tion loss, optimal values of the ratio of dimple depth to dimple print diameter and
the ratio of dimple print diameter to distance between dimples decrease, but the ratio
of channel height to dimple print diameter increases.

REFERENCES
1. P. M. Ligrani, M. M. Oliveira, and T. Blaskovich, Comparison of Heat Transfer Augmen-
tation Techniques, AIAA J., vol. 41, no. 3, pp. 337–362, 2003.
2. G. I. Mahmood, M. L. Hill, D. L. Nelson, P. M. Ligrani, H.-K. Moon, and B. Glezer,
Local Heat Transfer and Flow Structure on and above a Dimpled Surface in a Channel,
J. Turbomachinery, vol. 123, no. 1, pp. 115–123, 2001.
3. G. I. Mahmood and P. M. Ligrani, Heat Transfer in a Dimpled Channel: Combined
Infuences of Aspect Ratio, Temperature Ratio, Reynolds Number, and Flow Structure,
Int. J. Heat Mass Transfer, vol. 45, no. 10, pp. 2011–2020, 2002.
4. N. K. Burgess and P. M. Ligrani, Effects of Dimple Depth on Nusselt Numbers and Fric-
tion Factors for Internal Cooling Channel, ASME Turbo Expo 2004, Vienna, Austria,
ASME Paper GT2004-5432, 2004.
5. P. M. Ligrani, N. K. Burgess, and S. Y. Won, Nusselt Number and Flow Structure on and
above a Shallow Dimpled Surface within a Channel Including Effects of Inlet Turbulence
Intensity Level, ASME Turbo Expo 2004, Vienna, Austria, ASME Paper GT2004-54231,
2004.
6. W. V. Patrick and D. K. Tafti, Computations of Flow Structure and Heat Transfer in a
Dimpled Channel at Low to Moderate Reynolds Number, 2004 ASME Heat Transfer=
Fluids Engineering Summer Conf., Charlotte, NC, ASME Paper HT-FED04-56171, 2004.
7. S. A. Isaev and A. I. Leont’ev, Numerical Simulation of Vortex Enhancement of Heat
Transfer under Conditions of Turbulent Flow past a Spherical Dimple on the Wall of a
Narrow Channel, High Temperature, vol. 41, no. 5, pp. 665–679, 2003.
8. J. Park, P. R. Desam, and P. M. Ligrani, Numerical Predictions of Flow Structure above a
Dimpled Surface in a Channel, Numer. Heat Transfer A, vol. 45, pp. 1–20, 2004.
9. K. Y. Kim and S. S. Kim, Shape Optimization of Rib-Roughened Surface to Enhance
Turbulent Heat Transfer, Int. J. Heat Mass Transfer, vol. 45, pp. 2719–2727, 2002.
10. H. M. Kim and K. Y. Kim, Design Optimization of Rib-Roughened Channel to Enhance
Turbulent Heat Transfer, Int. J. Heat Mass Transfer, vol. 47, pp. 5159–5168, 2004.
11. H. M. Kim and K. Y. Kim, Optimization of Three-Dimensional Angled Ribs with RANS
Analysis of Turbulent Heat Transfer, ASME Turbo Expo 2004, Vienna, Austria, ASME
Paper GT2004-53346, 2004.
12. R. H. Myers and C. C. Montgomery, Response Surface Methodology: Progress and
Product Optimization Using Designed Experiments, Wiley, New York, 1995.
13. S. Y. Lee and K. Y. Kim, Design Optimization of Axial Flow Compressor Blades with
Three-Dimensional Navier–Stokes Solver, Korean Soc. Mech. Eng. Int. J., vol. 14,
pp. 1005–1012, 2000.
14. C. S. Ahn and K. Y. Kim, Aerodynamic Design Optimization of a Compressor Rotor
with Navier–Stokes Analysis, Proc. Inst. Mech. Eng. A, J. Power Energy, vol. 217,
pp. 179–184, 2003.
SHAPE OPTIMIZATION OF A DIMPLED CHANNEL 915

15. W. Shyy, N. Papila, R. Vaidyanathan, and K. Tucker, Global Design Optimization Aero-
dynamics and Rocket Propulsion Components, Prog. Aerospace Sci., vol. 37, pp. 59–118,
2001.
16. CFX-5.7 Solver Theory, Ansys, 2004.
17. B. W. Webb and S. Ramadhyani, Conjugate Heat Transfer in a Channel with Staggered
Ribs, Int. J. Heat Mass Transfer, vol. 28, pp. 1679–1687, 1985.
18. F. Menter and T. Esch, Elements of Industrial Heat Transfer Predictions, 16th Brazilian
Congress of Mechanical Engineering (COBEM), Uberlandia, Brazil, 2001.
19. J. E. Bardina, P. G. Huang, and T. Coakley, Turbulence Modeling Validation, AIAA
Paper 97-2121, 1997.
20. Y. G. Lai and R. M. C. So, Near-Wall Modeling of Turbulent Heat Fluxes, Int. J. Heat
Mass Transfer, vol. 33, pp. 1429–1440, 1990.
21. D. L. Gee and R. L. Webb, Forced Convection Heat Transfer in Helically Rib-
Roughened Tubes, Int. J. Heat Mass Transfer, vol. 23, pp. 1127–1136, 1980.
22. J. C. Han, J. S. Park, and C. K. Lei, Heat Transfer Enhancement in Channels with
Turbulence Promoters, J. Eng. Turbines Power, vol. 107, pp. 628–635, 1985.
23. B. S. Petukhov, Advances in Heat Transfer, vol. 6, pp. 503–504, Academic Press,
New York, 1970.
24. R. S. Bunker, M. Gotovskii, M. Belen’kiy, and B. Fokin, Heat Transfer and Pressure Loss
for Flows inside Converging and Diverging Channels with Surface Concavity Shape
Effects,4th Int. Conf. Compact Heat Exchangers and Enhancement Technology, Crete
Island, Greece, No. 2003GRC016, 2003.

View publication stats

You might also like