International Communications in Heat and Mass Transfer 57 (2014) 254–263

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ichmt

Optimization of finned-tube heat exchangers for diesel exhaust waste

heat recovery using CFD and CCD techniques☆
M. Hatami a,⁎, M. Jafaryar b, D.D. Ganji a, M. Gorji-Bandpy a
a
Babol University of Technology, Mechanical Engineering Department, Babol, Iran
b
Department of Mechanical Engineering, Mazandaran Institute of Technology, Babol, Iran

a r t i c l e i n f o a b s t r a c t

Available online 15 August 2014 In this paper, response surface methodology (RSM) based on central composite design (CCD) is applied to obtain
an optimization design of finned type heat exchangers (HEX) to recover waste heat from the exhaust of a diesel
Keywords: engine. The design is performed for a single point operation (1600 rpm and 60 N m) of an OM314 diesel engine
Central composite design (CCD) obtained from experimental measurements. Based on the CCD principle, fifteen HEX cases with different fins
Response surface method (RSM) height, thickness and number are modeled numerically and the optimization is done to have the maximum
Heat recovery
heat recovery amount and minimum of pressure drop along the heat exchanger.
Optimization
Finned heat exchanger
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction technique. They reported that fins should be in the exhaust gas passage
for more heat transfer and designed 18 cases with different fin numbers
It is evident that exhaust of the engines is another main source that a and thicknesses and found the most effective case. Zhang et al. [7]
large amount of energy wastes through it. Researchers confirm that modeled a finned tube evaporator heat exchanger for an organic
more than 30–40% of fuel energy wastes from the exhaust and just Rankine cycle (ORC) and concluded that waste-heat recovery efficiency
12–25% of the fuel energy convert to useful work [1–3]. On the other is between 60% and 70% for most of the engine's operating region; also,
hand, statistics show that production of a number of internal combus- they mentioned that heat transfer area for a finned tube evaporator
tion engines grows very fast and concern on increasing the harmful should be selected carefully based on the engine's most typical operating
greenhouse gases (GHG) will appear. So, researchers are motivated to region. Ghazikhani et al. [8] used a simple double pipe heat exchanger in
recover heat from the waste sources in engines by using the applicable diesel engine and performed an exergy analysis for finding the relation
ways. Some researchers attempt to enhance the heat transfer by a spe- between irreversibilities and exhaust sound level. Recently, they [9]
cial design of HEXs in exhaust of diesel engines due to high applications estimated in an experimental work that BSFC could be improved approx-
which were discussed in the previous sections. Recently, Mokkapati and imately 12% in different load and speeds of an OM314 diesel engine by
Lin [4] inserted a twisted tape in the corrugated tube exhaust of a heavy using the recovered exergy from a simple double pipe heat exchanger
duty diesel engine to increase recovered heat and evaluated its impact in exhaust. Also, they showed that exergy recovery increases by increas-
on engine performance and economics. Pandiyarajan et al. [5] designed ing the engine load and speeds especially in high speeds. Recently,
a finned-tube heat exchanger and used a thermal energy storage using Hossain and Bari [10,11] applied a new HEX for a diesel engine experi-
cylindrical phase change material (PCM) capsules and found that nearly mentally and numerically. After that they optimized the working fluid
10–15% of fuel power is stored as heat in the combined storage system pressure and the orientation of heat exchangers and found that the addi-
in different loads. tional power increased from 16% to 23.7%. Also, they investigated the
In another experimental work, Lee and Bae [6] designed a little heat parallel and series configurations of HEXs which obtained additional
exchanger with fins in the exhaust by design of experiment (DOE) power. Mavridou et al. [12] examined two groups of configurations:
(a) a classical shell and tube heat exchanger using staggered cross-flow
and (b) a cross-flow plate heat exchanger, initially with finned surfaces
on the exhaust gas side and then with metal foam material substituting
for the fins and they attempt to minimize the volume and weight of
☆ Communicated by W.J. Minkowycz.
⁎ Corresponding author.
the arrangement, while at the same time maintaining the heat transfer
E-mail addresses: m.hatami2010@gmail.com (M. Hatami), ddg_davood@yahoo.com from the gas side at a maximum range. A complete review of different
(D.D. Ganji). heat exchanger designs is gathered by authors [1].

http://dx.doi.org/10.1016/j.icheatmasstransfer.2014.08.015
0735-1933/© 2014 Elsevier Ltd. All rights reserved.
 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263 255

Fig. 1. Schematic of designed finned heat exchanger.

According to our review [1] on heat exchanger designs for exhaust numbers), we aims to reach the maximum heat recovery in an acceptable
heat recovery, a special design with high heat recovery and less pressure pressure drop which makes back pressure in exhaust.
drop is highly necessary which is a gap in this field. The current study
aims to fill this gap to introduce finned type heat exchangers and 3. Numerical modeling
optimize its geometry for heat recovery application using numerical
techniques based on experimental data. 3.1. Governing equations

The numerical simulation was performed with a three dimensional

2. Problem description steady-state turbulent flow system. To solve the problem, governing
equations for the flow and conjugate heat transfer were modified
In our previous study [9], a simple double pipe HEX was mounted according to the conditions of the simulation setup. As the problem
on exhaust of an OM314 diesel engine for exergy and heat recovery. In was assumed to be steady, the time dependent parameters were
the current study, firstly this HEX is modeled numerically to validate dropped from the equations. The resulting equations were [10,11]:
the results and choose a suitable viscous model. Then, by using the
Continuity equation:
CCD technique, 15 cases of finned-tube HEXs are designed to optimize
the geometry of finned HEX. Fig. 1 shows a schematic of these HEXs.
!
∇: ρ V ¼ 0 ð1Þ
Actually, by optimizing the geometry of fins (height, thickness and

Table 1
Different properties of designed geometry with mesh number details.

Case number Factors Grid numbers

Fin height Fin thickness Fin number Volume 1 (gases) Volume 2 (fins) Volume 3 (water) Total
(H) (mm) (T) (mm)

1 20 1 4 514,321 8892 183,804 707,017

2 20 1 10 609,257 12,870 182,636 804,763
3 20 2.5 7 539,587 10,179 182,629 732,395
4 20 4 4 493,674 8658 183,649 685,981
5 20 4 10 558,865 10,881 183,066 752,812
6 35 1 10 732,090 15,444 181,819 929,353
7 35 2.5 4 547,469 10,296 183,882 741,647
8 35 2.5 7 619,536 12,285 183,256 815,077
9 35 2.5 10 686,567 15,327 183,822 885,716
10 35 4 7 746,870 12,285 182,426 941,581
11 50 1 4 686,938 11,934 182,969 881,841
12 50 1 10 743,601 17,550 182,840 943,991
13 50 2.5 7 618,252 14,274 183,333 815,859
14 50 4 4 672,747 11,115 182,958 866,820
15 50 4 10 809,969 16,848 183,300 1,010,117

Table 2
Temperature-dependent properties of exhaust gases.

Exhaust gas properties A + B × T + C × T2 + D × T3

A B C D

ρ (kg/m3) 2.504012288761e+00 −5.958486188418e−03 5.578942358587e−06 −1.772600918994e−09

Cp (J/kg K) 1.015580935928e+03 −1.512248401853e−01 4.544870294058e−04 −1.785063817137e−07
μ (kg/m s) 1.325186910351e−06 6.740061370040e−08 −3.749043579926e−11 1.110074961972e−14
k (W/m K) −3.182421851331e−03 1.185847825677e−04 −7.706004236629e−08 2.939653967062e−11
256 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263

Table 3 Table 4
Physical properties of coolant and fins material. Level of factors in CCD technique.

ρ (kg/m3) Cp (J/kg K) μ (kg/m s) k (W/m K) Factors Symbol −1 level Central point +1 level

Water 998.2 4182 0.001003 0.6 Fin height (mm) H 20 35 50

Carbon steel 7858 486 – 52 Fin thickness (mm) T 1 2.5 4
Fin numbers N 4 7 10

Momentum equations: and

x-momentum: !
2
∂ ∂ ∂ ∂ε ε ε

! ðρε Þ þ ðρεui Þ ¼ α ε μ eff þ C 1ε ðGk þ C 3ε Gb Þ−C 2ε ρ −Rε þ Sε
∂p ∂τ ∂τ yx ∂τ zx ∂t ∂xi ∂x j ∂x j k k
∇: ρu V ¼ − þ xx þ þ ð2Þ
∂x ∂x ∂y ∂z ð7Þ

where Gk represents the generation of turbulence kinetic energy due to

y-momentum:
the mean velocity gradients and Gb is the generation of turbulence

! kinetic energy due to buoyancy and C1e = 1.42, C2e = 1.68 in RNG k-
∂p ∂τ xy ∂τ yy ∂τzy
∇: ρυ V ¼ − þ þ þ þ ρg ð3Þ ε model.
∂y ∂x ∂y ∂z
Transport equations for SST k-ω are:
!
z-momentum: ∂ ∂ ∂ ∂k
ðρkÞ þ ðρkui Þ ¼ Γk e −Y þ S
þG ð8Þ
k k k

! ∂t ∂xi ∂x j ∂x j
∂p ∂τ ∂τ yz ∂τ zz
∇: ρw V ¼ − þ xz þ þ ð4Þ
∂z ∂x ∂y ∂z
and
!
Energy equation:
∂ ∂ ∂ ∂ω
! ðρωÞ þ ðρωui Þ ¼ Γω þ Gω −Y ω þ Dω þ Sω ð9Þ
  2 2 2 ∂t ∂xi ∂x j ∂x j
∂T ∂T ∂T ∂ T ∂ T ∂ T
ρcp u þυ þw ¼λ þ þ ð5Þ
∂x ∂y ∂z ∂x2 ∂y2 ∂z2
where Gk represents the generation of turbulence kinetic energy due to
mean velocity gradients and Gω represents the generation of ω [13,14].
In this paper, three viscous models are examined. Standard and Transferred heat amount or heat flux is an important parameter in
renormalization-group (RNG) k-ε model and shear-stress transport HEX analysis. The average heat flux is,
(SST) k-ω model. For RNG k-ε model, thermal effect is considered in

the enhanced wall treatment panel. Transport equations for RNG k-ε Q̇ g þQ̇ c
Q̇ ave ¼ ð10Þ
model in general form are: 2

! Q c is the heat transferred to the coolant which is calculated as,

∂ ∂ ∂ ∂k

ðρkÞ þ ðρkui Þ ¼ α k μ eff þ Gk þ Gb −ρε−Y M þ Sk ð6Þ
∂t ∂xi ∂x j ∂x j Q̇ c ¼ ṁ c C pc T c;out −T c;in ð11Þ

(a) (b)
420 296
Experimental Experimental
RNG k-e RNG k-e
Standard k-e Standard k-e
SST k-w SST k-w
294
Exhaust out let temperature (K)

Water out let temperature (K)

400

292
380

290
360

288
340
20 40 60 80 100 20 40 60 80 100
Engine load (%) Engine load (%)

Fig. 2. Comparison among experimental values and three different viscous models.
 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263 257

Fig. 3. A sample generated mesh and boundary layer mesh around fins and walls.

where ṁc is the mass flow rate of coolant and Cp.c is the specific heat of 3.2. Geometry, mesh generation and boundary conditions
coolant. Q g is the heat transferred to exhaust gases which for
temperature-dependent specific heat can be calculated as, In the modeling, all HEX lengths are 70 cm with 12 cm inner diameter
and 14 cm outer diameter while the gas inlet and outlet diameters are
Z T g;out considered 48 mm as shown in Fig. 1. Both inlet and outlet of the
Q̇ g ¼ ṁ g C pg dT ð12Þ water pass are placed in the upper surface of the HEX with a 2 cm diam-
T g;in
eter. As described before 15 cases of HEXs are modeled and their proper-
ties are presented in Table 1. The HEXs are considered well-insulated;
where ṁ g is the exhaust gases mass flow rate which is the sum of the hence, the heat losses to the environment are totally neglected. As an
intake air and fuel consumption mass flow rates [11]. approximation, the properties of air can be used for diesel exhaust gas

10
All cases, water side wall 240,025 Cells
Case 2, gases side wall 25 324,793 Cells
Case 3, gases side wall 424,768 Cells
552,944 Cells
Case 7, gases side wall 641,037 Cells
8 Case 9, gases side wall 780,310 Cells
Case 11, gases side wall 20
Velocity (m/s)

6
15
Y+

4
10

2 5

-0.3 -0.2 -0.1 0 0.1 0.2 0.3 -0.08 -0.06 -0.04 -0.02 0 0.02
Position (m) Radial Position (m)

Fig. 4. Y plus (Y+) on the gases and water side walls for different cases. Fig. 5. Mesh independency examination for velocity profile in middle plane.
258 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263

calculations and the error associated with neglecting the combustion The 3D grid system was established using the SolidWorks and
products is usually no more than about 2%. Due to high temperature in ANSYS-WORBENCH commercial software. In each HEX, three separate
exhaust, temperature dependent properties are considered for exhaust domains or volumes should be considered for gases, solid and liquid
gases and for each property, a fourth order polynomial is plotted and phases. Meshes are tetrahedral and their details are presented in
its equation and coefficients are shown in Table 2. Also, all the solid Table 1. Also, grid independence tests are carried out to ensure that a
phases (fins and walls) are made from carbon steel with 3 mm thick- nearly grid independent solution can be obtained. In the test, six differ-
nesses and water is the coolant and their properties are presented in ent grid systems with approximately 200,000 to 800,000 cells are
Table 3. adopted for calculation of the whole heat exchanger, and the difference

Fig. 6. Temperature contour (K) of different cases in water outlet section for a) case 2, b) case 4, c) case 8, d) case 10, e) case 13 and f) case 15.
 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263 259

in the results of two last cases was negligible. As described before, the 4. Optimization analysis
commercial code ANSYS 14.50 is adopted to simulate the flow and
heat transfer in the computational domain. As reported by Patankar 4.1. Response surface methodology (RSM)
[21], the governing equations are discretized by the finite volume meth-
od [13]. The QUICK scheme is used to discretize the convective terms Response surface methodology (RSM) is a collection of mathematical
and the SIMPLE algorithm is used to deal with the coupling between and statistical. Originally, RSM was developed to model experimental re-
velocity and pressure. sponses and then migrated into the modeling of numerical experiments.

420
410
400
390
380
370
360
350
340
330
320
310
300 (a)
290

420
410
400
390
380
370
360
350
340
330
320
310 (b)
300
290

420
410
400
390
380
370
360
350
340
330
320
310
300 (c)
290

420
410
400
390
380
370
360
350
340
330
320
310
300
290
(d)

Fig. 7. Longitudinal temperature contour (K) for a) case 4, b) case 8, c) case 14 and d) case 15.
260 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263

390

Exhaust outlet temperature (K)

292

Water outlet temperature (K)

385

380
290

375

370
288
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Case number Case number
(a) (b)
Fig. 8. Comparison between the outcomes in 15 different HEX design. a) Exhaust and b) water temperature.

The difference is in the type of error generated by the response. In phys- 4.2. Central composite design (CCD)
ical experiments, inaccuracy can be due, for example, to measurement
errors while, in computer experiments, numerical noise is a result of A Box–Wilson central composite design, commonly called “cen-
incomplete convergence of iterative processes, round-off errors or tral composite design”, contains an imbedded factorial or fractional
the discrete representation of continuous physical phenomena. The factorial design with center points that are augmented with a
application of RSM to design optimization is aimed at reducing the group of ‘star points’ that allow estimation of curvature. If the
cost of expensive analysis methods (e.g. finite element method or CFD distance from the center of the design space to a factorial point is
analysis) and their associated numerical noise. Generally, the structure ± 1 unit for each factor, the distance from the center of the design
of the relationship between the response and the independent variables space to a star point is ± α with |α| N 1. The precise value of α
is unknown. The first step in RSM is to find a suitable approximation to depends on certain properties desired for the design and on the
the true relationship. The most common forms are low-order polyno- number of factors involved [15,16].
mials (first or second-order). Second order model can significantly
improve the optimization process when a first-order model suffers 5. Results and discussions
lack of fit due to interaction between variables and surface curvature.
A general second-order model is defined as [15]: As described before, three viscous models are examined to find the
most accurate viscous model for the current problem. For this aim, stan-
X
n X
n n X
X n 
2  dard k-ε, RNG k-ε and SST k-ω are used for a double pipe heat exchanger
y ¼ a0 þ a i xi þ aii xi þ aij xi x j  ð13Þ
ib j in exhaust of an OM314 diesel engine in different engine loads in
i¼1 i¼1 i¼1 j¼1
1600 rpm according to previous experimental study [9]. As shown in
where xi and xj are the design variables and a are the tuning parameters. Fig. 2 both RNG k-ε and SST k-ω methods have acceptable results in
CCD or central composite design is one of the modules in RSM to obtain HEX modeling compared to experimental data. After that the finned
the points of each factor according to their levels. HEXs are investigated numerically using CCD technique. According to

3000 400

350
2500

300
Recovered Heat (W)

Pressure drop (Pa)

2000
250

1500 200

150
1000

100
500
50
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Case number Case number
(a) (b)
Fig. 9. a) Recovered heat and b) pressure drop for all designed cases.
 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263 261

Fig. 10. Effect of fin dimensions on recovered heat amount. a) Height and thickness and b) height and fin numbers.

CCD, 15 HEX cases should be modeled and their properties are present- In Fig. 6 temperature contour in outlet water section is indicated for
ed in Table 1 in details of grid numbers. six sample cases. In the zoomed area of these figures, increasing the
These different geometries are selected according to defined levels temperature of outlet water is completely evident as well as the effect
(+1 and −1) for each factor which is shown in Table 4. In CCD details, of fin dimensions and numbers in water temperatures. Another view
α is considered to be a unit and six points are considered for central of temperature contours is displayed in Fig. 7 which completely shows
points and two responses are considered for the optimization, pressure the longitudinal fin effect in the heat transfer. Fig. 8(a) and (b) demon-
drop and recovered heat. A sample of mesh generated is shown through strates the effect of all designed cases on the exhaust and water outlet
Fig. 3 and boundary layer mesh is also considered for both gases and temperatures respectively. As seen in this figure case 15 has the maxi-
water pass. Furthermore, to examine the accuracy of RNG k-ε in all 15 mum water outlet temperature and case 11 has the minimum value,
cases Y + on walls is checked during the runs to be in an acceptable so the maximum heat recovery will occur in case 15 and minimum
range and its results are shown in Fig. 4 for five samples. As seen, transferred heat is for case 11 which are shown in bar chart in Fig. 9-a.
these third order polynomial diagrams of Y+ are under 10 in all cases Because pressure drop due to back pressure in exhaust is an important
which confirm the accuracy of the modeling. Also, mesh independency issue, it should be considered in minimum. Fig. 9 reveals that also case
study is performed in all modeling and one case is shown in Fig. 5 for 15 has the best heat transfer, but it also has a large pressure drop and
case 2 and about 600,000 cells are suitable for modeling to reduce the case 2 has the minimum pressure drop. So, an optimization study is
time of divergence. needed which is performed using CCD technique in the present study.

Fig. 11. 3D surfaces for the effects of fins dimensions on recovered heat.
262 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263

Fig. 12. Effect of fin dimensions on pressure drop in HEX. a) Height and thickness and b) thickness and fin numbers.

In RSM, a polynomial model with quadratic order is applied to of these factors on pressure drop are shown in Fig. 12 contours. As an im-
responses (recovered heat and pressure drop). After considering the portant outcome, fin thickness has no significant effect on pressure drop,
power transformation models for analysis the results. P-value Probs while increasing the fin numbers increases it and increasing the fin thick-
was estimated to be 0.0045 and 0.0008 while R-squared was 0.8838 nesses can decrease the pressure drop in high fin numbers. After optimi-
and 0.6394 for Q and Δp respectively. Considering the effects of main zation analysis using CCD technique, Table 5 is obtained and ten suitable
factors and also the interactions between two-factor, Eq. (13) takes HEX designs are proposed. Design number 1 which has a desirability of
the form 0.826 is the best design and its desirability and fin geometry are plotted
in contour of Fig. 13. It can be concluded that the height of the fins has
Q ¼ 1963:54930 þ 7:07974H−65:62457T−109:22023N þ 4:3875H  T more negative effect in pressure drop and fin number and fin thickness
2 2 2
þ2:99478H  N þ 0:23077T  N−0:39827H −1:86245 T þ 4:93677N effect is more positive in heat recovery effect.
ð14Þ
6. Conclusion
and for pressure drop,
In this study, response surface methodology (RSM) based on central
Δp ¼ þ199:59750−1:49262H þ 11:00826T þ 2:04420N−0:28554H  T composite design (CCD) was applied to obtain an optimization design of
2 2 2
þ0:10021H  N−5:17726T  N þ 0:070195H þ 5:91261T þ 0:36926N finned type heat exchangers (HEX) to recover waste heat from the
ð15Þ exhaust of a diesel engine. For this aim, fifteen HEXs with different fin
dimensions and six central points were designed numerically. Results
Contours in Fig. 10 show the effect of fin height, thickness and num- show that height of the fins has more effect on pressure drop than fin
bers on heat recovery amount and their 3D schematic is shown in Fig. 11. numbers and thicknesses while, fin numbers have maximum effect to
An increase in all these factors makes an increase in heat recovery enhance the heat recovery. According to CCD optimization technique,
amount although the effect of fin numbers and heights is more sensible a special design for finned heat exchanger is proposed to fill the gap in
than thickness due to its high slope in the surface in Fig. 11. The effects this area.

Table 5
The best optimized cases obtained from CCD analysis.

Design number Fin height (mm) Fin thickness (mm) Fin number Recovered heat (W) Pressure drop (Pa) Desirability Best design

1 39.27 4.00 10.00 2611.048 232.647 0.826 ✓

2 39.14 4.00 10.00 2608.099 232.158 0.826
3 39.50 4.00 10.00 2616.305 233.530 0.826
4 40.06 4.00 10.00 2629.195 235.756 0.826
5 38.48 4.00 10.00 2592.322 229.614 0.826
6 38.14 4.00 10.00 2584.194 228.350 0.825
7 39.22 3.97 10.00 2607.470 232.590 0.825
8 38.36 4.00 9.98 2587.668 229.288 0.825
9 37.66 4.00 10.00 2572.380 226.563 0.825
10 37.47 4.00 10.00 2567.909 225.903 0.824
 M. Hatami et al. / International Communications in Heat and Mass Transfer 57 (2014) 254–263 263

Fig. 13. Desirability graph for optimized case.

References conference of internal combustion engine, Olympic international hotel, Tehran,

Iran, 2011.
[1] M. Hatami, D.D. Ganji, M. Gorji-Bandpy, A review of different heat exchangers [9] M. Ghazikhani, M. Hatami, D.D. Ganji, M. Gorji-Bandpy, Gh. Shahi, A. Behravan,
designs for increasing the diesel exhaust waste heat recovery, Renew. Sust. Energ. Exergy recovery from the exhaust cooling in a DI diesel engine for BSFC reduction
Rev. 37 (2014) 168–181. purposes, Energy 65 (2014) 44–51.
[2] R. Saidur, N.A. Rahim, H.W. Ping, M.I. Jahirul, S. Mekhilef, H.H. Masjuki, Energy and [10] Shekh Nisar Hossain, Saiful Bari, Waste heat recovery from the exhaust of a diesel
emission analysis for industrial motors in Malaysia, Energy Policy 37 (9) (2009) generator using Rankine Cycle, Energy Conversion and Management 75 (2013)
3650–3658. 141–151.
[3] M. Hasanuzzaman, N.A. Rahim, R. Saidur, S.N. Kazi, Energy savings and emissions re- [11] Saiful Bari, Shekh Nisar Hossain, Waste heat recovery from a diesel engine using
ductions for rewinding and replacement of industrial motor, Energy 36 (1) (2011) shell and tube heat exchanger, Applied Thermal Engineering 61 (2013) 355–363.
233–240. [12] S. Mavridou, G.C. Mavropoulos, D. Bouris, D.T. Hountalas, G. Bergeles, Comparative
[4] Vamsi Mokkapati, Chuen-Sen Lin, Numerical study of an exhaust heat recovery sys- design study of a diesel exhaust gas heat exchanger for truck applications with
tem using corrugated tube heat exchanger with twisted tape inserts, Int. Commun. conventional and state of the art heat transfer enhancements, Appl. Therm. Eng.
Heat Mass Transfer (2014), http://dx.doi.org/10.1016/j.icheatmasstransfer.2014.07. 30 (2010) 935–947.
002. [13] S.V. Patankar (Ed.), Numerical Heat Transfer and Fluid Flow, McGrawHill, NewYork,
[5] V. Pandiyarajan, M. Chinna Pandian, E. Malan, R. Velraj, R.V. Seeniraj, Experimental 1980.
investigation on heat recovery from diesel engine exhaust using finned shell and [14] Harun Bilirgen, Stephen Dunbar, Edward K. Levy, Numerical modeling of finned heat
tube heat exchanger and thermal storage system, Appl. Energy 88 (2011) 77–87. exchangers, Appl. Therm. Eng. 61 (2013) 278–288.
[6] S. Lee, C. Bae, Design of a heat exchanger to reduce the exhaust temperature in a [15] N. Aslan, Application of response surface methodology and central composite
spark-ignition engine, Int. J. Therm. Sci. 47 (2008) 468–478. rotatable design for modeling the influence of some operating variables of a
[7] H.G. Zhang, E.H. Wang, B.Y. Fan, Heat transfer analysis of a finned-tube evaporator Multi-Gravity Separator for coal cleaning, Fuel 86 (2007) 769–776.
for engine exhaust heat recovery, Energy Convers. Manag. 65 (2013) 438–447. [16] Lei Sun, Chun-Lu. Zhang, Evaluation of elliptical finned-tube heat exchanger
[8] M. Ghazikhani, M. Hatami, Gh. Shahi, A. Behravan, Exergy analysis and experimental performance using CFD and response surface methodology, Int. J. Therm. Sci. 75
investigation on exhaust sound level in a DI diesel engine, 7th international (2014) 45–53.