Purdue University
Purdue e-Pubs
International Refrigeration and Air Conditioning
Conference
School of Mechanical Engineering
2012
New Generation of Air Cooled Heat Exchanger 1
kW Module Design Optimization
Khaled Hassan Saleh
Khaled.h.saleh@gmail.com
Omar Abdelaziz
Vikrant Aute
Reinhard Radermacher
Shapour Azarm
Follow this and additional works at: http://docs.lib.purdue.edu/iracc
Saleh, Khaled Hassan; Abdelaziz, Omar; Aute, Vikrant; Radermacher, Reinhard; and Azarm, Shapour, "New Generation of Air Cooled
Heat Exchanger 1 kW Module Design Optimization" (2012). International Refrigeration and Air Conditioning Conference. Paper 1203.
http://docs.lib.purdue.edu/iracc/1203
This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact epubs@purdue.edu for
additional information.
Complete proceedings may be acquired in print and on CD-ROM directly from the Ray W. Herrick Laboratories at https://engineering.purdue.edu/
Herrick/Events/orderlit.html
2187, Page 1
New Generation of Air Cooled Heat Exchanger 1 kW Module Design Optimization
Khaled SALEH1*, Omar ABDELAZIZ2, Vikrant AUTE3, Reinhard RADERMACHER4,
Shapour AZARM5
1,3,4,5
Center for Environmental Energy Engineering
Department of Mechanical Engineering, University of Maryland
College Park, MD 20742 USA
1
3
Tel: 301-405-5285, Tel: 301-405-8726, 4Tel: 301-405-5286, 5Tel: 301-405-5250, Fax: 301-405-2025
3
Email: vikrant@ umd.edu, 4Email: raderm@umd.edu, 5Email: azarm@umd.edu
2
Oak Ridge National Laboratory
Oak Ridge, TN 37831 USA
2
Tel: 865-574-2089, 2Email: abdelazizoa@ornl.gov
* Corresponding Author Email: ksaleh@umd.edu
ABSTRACT
The objective of this paper is to evaluate and optimize the performance of 1 kW integrated heat exchanger module
for new generations of air cooled heat exchangers. The first objective is to minimize the ratio of the header frontal
area to the entire heat exchanger frontal which will help to reduce the header size. The second objective is to
minimize the pressure drop for the entire heat exchanger, i.e., inside the inlet and outlet headers in addition to
pressure drop inside the tubes. A three step approach is proposed. First step involves selecting the header design
based on previous header optimization studies and then simulating the header using a new 3D CFD simulation
approach. Second step includes solving the heat exchanger using information from the header simulation that
accounts for the variation in refrigerant mass flow rate inside the tubes and obtain the performance for the entire heat
exchanger. Finally, a solver is used to evaluate the overall module performance. Three different headers are
investigated with different header height and size ratio. Then parametric studies are conducted to explore the effect
of header size ratio on the optimum designs. Lastly, design guidelines to optimize the integrated heat exchange
module are provided based on the study results.
1. INTRODUCTION
Modern developments in the area of air cooled heat exchangers resulted in using channels in the range of micro or
mini scale. Consequently, reducing the refrigerant maldistribution inside different channels is an important objective
in order to reduce both the deterioration in heat transfer as well as total pressure drop. In order to achieve both goals,
the size of the heat exchanger headers should be increased. However in many applications there is a need to reduce
the total volume of the heat exchanger such as aeronautics and marine applications. It is well known that in
microchannel heat exchangers the header may occupy up to 20% of the overall heat exchanger packaging volume
(Shah and London, 1978 and Shah, 2006). This implies that smaller flow channels tend to require larger header size
in order to reduce both the pressure drop and to improve the flow uniformity. Obviously there is a tradeoff between
better flow distribution uniformity, total pressure drop in the heat exchanger, and header volume (Abdelaziz et al.,
2010).
In order to optimize novel heat exchanger, there is a need to simulate the entire heat exchanger module – including
headers. Most of the CFD based optimization studies in the area of heat exchanger simulation and design
optimization focus on segment level optimization (Lee et al., 2001). Some existing methods have used curve fitting
to correlate the response from CFD runs inside the optimization step. Other methods used Design of Experiments
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012
2187, Page 2
(DOE), metamodeling, and optimization in heat exchanger design applications (Jing et al., 2005; Park and Moon,
2005; Park et al., 2006). Recently, approximation assisted optimization approach was used to optimize the entire
heat exchanger based on previously built metamodel (Abdelaziz et al., 2010) or based on adaptively updated
metamodel (Saleh et al., 2011). However, most of the previous work focused on the heat exchanger core without
considering the effect of the header which is significantly important for new generations of air cooled heat
exchangers.
In this paper, an integrated 1 kW heat exchanger module is developed. The effect of changing header parameters
such as header height, header height ratio, and the effect of fluid maldistribution in the heat exchanger tubes is
considered. The remainder of this paper is organized as follows: Section 2 provides details about the heat exchanger
header CFD model and offers a brief overview of the integration method to link the entire heat exchanger solver
with the header output results. Section 3 summarizes the results obtained and presents parametric studies on the
effect of different design variables. Finally, conclusions and design guidelines based on the results are drawn.
2. INTEGRATED HEADER WITH HEAT EXCHANGER
In this section we will introduce the efforts to integrate 1 kW heat exchanger module with headers. The schematic of
the integrated module is presented in Figure 1. The number of tubes on the flow direction is called N tube; however
the number of tubes in the perpendicular direction is called N port as shown in Figure 1. The header total height (LH,i +
LH,o) and the header size ratio (LH,i / LH,o) are two important variables that can affect the refrigerant distribution
inside the tubes.
Figure 1: Schematic of the integrated heat exchanger with header module (Abdelaziz, 2009).
2.1 Header 3D-CFD Simulation
In the current header simulation, blocked geometry technique is used with hexahedral mesh to simulate the flow
distribution inside the inlet and outlet headers as shown in Figure 2. With respect to the pressure drop inside the
tubes and in order to reduce the computational domain, Fluent ® artificially creates a pressure jump across the faces
representing the tubes (Fluent User’s Guide 6.3, 2007). This modeling approach will not be able to capture the vena
contracta and the flow dynamics in the tubes; however, it will be able to account for the effect of overall pressure
drop in each flow channel on the mass flow rate distribution. In addition, hexahedral mesh type helps to simulate
headers with large number of tubes. The main advantage of using the blocked geometry with hexahedral mesh is to
reduce the number of cells significantly. This reduction enables solving the header simulation problem in shorter
time compared to the conventional meshing strategies. Some results for header simulations with different header
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012
2187, Page 3
size ratios (SR) as defined in Equation 1 ranged from 1 to 8 as shown in Figure 3. The results show improved
performance with SR of 4 and 5.
(1)
SR=(LH,i / LH,o)
Porous Jump
interior boudaries
Symmetry
Planes
Symmetry
Planes
Figure 2: Mass Computational domain simplification with porous jump interior boundaries (Abdelaziz, 2009).
3.50E-05
3.00E-05
MFR (Kg/s)
2.50E-05
2.00E-05
H2SR1
H2SR2
H2SR4
H2SR5
H2SR8
1.50E-05
1.00E-05
5.00E-06
0.00E+00
0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 63 66 69
Tube Number
Figure 3: Mass flow rate (MFR) distribution for different header configurations.
2.2 Heat Exchanger Simulation
The segmented ε-NTU solver, CoilDesigner (Jiang et al., 2006), is used to simulate the overall heat exchanger
performance. The heat exchanger solver (CoilDesigner) accounts for the variation in refrigerant mass flow rate
(MFR) inside the tubes based on the header CFD simulation results.
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012
2187, Page 4
2.3 Integrated Heat Exchanger Header Simulation
Three main steps are required to obtain the final results of the integrated module. First step contains selecting the
header design based on the previous conclusions from header optimization study (Abdelaziz, 2009) and then
simulating the new header using the 3D-CFD simulation for the computational domain shown in Figure 2. Second
step includes the heat exchanger solver using information from the header simulation that accounts for the variation
in refrigerant mass flow rate inside the tubes. Finally, one solver accounts for the overall module performance, i.e.,
total pressure drop, total heat exchanger volume, ratio of header frontal area to total heat exchanger frontal area and
total heat transfer.
3. RESULTS
Three different headers are evaluated based on the previous header optimization study (Abdelaziz, 2009) with
different header total height (LH,i + LH,o) and header size ratio (SR) as defined in Equation 1. The heat exchanger
design is selected with the design variables as given in Table 1 (Abdelaziz et al., 2010). Parametric studies are
conducted to explore the effect of header size ratio on the optimum designs.
Table 1: Heat exchanger design data.
Design parameter
value
Number of tubes (Ntube)
69
Number of ports ( Nport)
17
Horizontal spacing (H.S.)
0.875 mm
Vertical spacing (V.S.)
1.24 mm
Refrigerant MFR
0.025 kg /s
Tube length (L)
1120.073 mm
Tube inner diameter (Din)
0.467 mm
Air Pressure Drop(∆Pair)
52.88 Pa
3.1 Integrated Heat Exchanger Module Results
Two objectives are considered while optimizing the integrated heat exchanger module as shown in Equation 2. An
integrated heat exchanger module describes the entire heat exchanger body along with the inlet and the outlet
header. The first objective is to minimize refrigerant mass flow rate standard deviation (σMFR) inside the tubes.
Second objective is to minimize the header frontal area divided by the total heat exchanger frontal area (Area Ratio).
Applying the first objective satisfies the reduction of pressure drop and better heat transfer distribution along the
tubes. The second objective reduces the obstacles in the air flow direction.
Objectives:
Minimize σMFR
Minimize Area Ratio
Subject to:
∆P total < 1000 Pa
Area Ratio < 0.12
Where:
Area Ratio = header frontal area/total heat exchanger frontal area
σMFR
= refrigerant mass flow rate standard deviation
(2)
Three header designs, viz. Header#1, Header#2, and Header#3, as described in Table 2, are evaluated using the
aforementioned steps. Integrated module results for the three headers are presented in Table 2. As observed from
Figure 4 and Figure 5, there is a tradeoff between the total volume and the refrigerant pressure drop. In Header#1 the
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012
2187, Page 5
refrigerant pressure drop is minimum however the total volume is maximum. The main reason behind the reduction
of the pressure drop is the large header which also causes the increase in the total volume. On the other hand,
Header#3 has the smallest header height with the minimum total however the refrigerant pressure drop is
maximized. As for the air side pressure drop and heat transfer coefficient, it is the same for all cases as the heat
exchanger configuration is fixed as given in Table 1.
Design
∆Pref (Pa)
Header#1
Header#2
Header#3
Heat
Load
(W)
1000.08
999.56
1001.21
Total
Volume
(cc)
185.87
184.01
174.72
500
490
480
470
460
450
440
430
420
410
400
Table 2: Integrated module results.
Material
Header
∆Pref
Volume
Height
SR
(Pa)
(cc)
(mm)
23.02
422.43
16.21
5.00
22.89
432.37
14.86
10.00
22.2
481.91
8.11
5.00
Area
Ratio
σMFR%
0.119
0.110
0.063
12.2
14.4
7.06
Header#3
Header#2
Header#1
174
176
178
180
182
Total Volume (cc)
184
186
188
∆Pref (Pa)
Figure 4: Refrigerant pressure drop versus total module volume.
500
490
480
470
460
450
440
430
420
410
400
0.06
Header#3
Header#2
Header#1
0.07
0.08
0.09
0.1
0.11
Area Ratio
Figure 5: Refrigerant pressure drop versus area ratio.
0.12
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012
2187, Page 6
3.2 Header Size Ratio Parametric Study
Additional investigation for header size ratio effects is performed for Header#1 as shown in Figure 6 and Figure 7.
3D-CFD simulations are performed for different header size ratio. Then heat exchanger solver for each case was run.
Finally overall integrated module solver used to obtain the integrated heat exchanger module performance. The
results show the impact of header size ratio on the module performance. For Header#1 the results are shown in
Table 3. It can be concluded that an optimum header size ratio of 2 is obtained for Header#1 design. At this
optimum value the refrigerant pressure drop is minimum at 400.266 Pa with total module volume of 185.874 cc.
Design
H1SR1
H1SR2
H1SR4
H1SR5
H1SR8
16.00
435.00
14.00
430.00
12.00
425.00
10.00
420.00
∆Pref (Pa)
σMFR%
Table 3: Header#1 Parametric study results.
Total
Header
Heat Load
∆Pref
Volume
Height
(W)
(Pa)
(cc)
(mm)
1001.202
185.874
411.350
16.210
1001.624
185.874
400.266
16.210
1000.480
185.874
411.350
16.210
1000.080
185.874
422.430
16.210
999.682
185.874
430.045
16.210
Header
Size
Ratio
1
2
4
5
8
8.00
6.00
σMFR%
0.119
0.119
0.119
0.119
0.119
6.260
3.680
11.060
12.200
13.751
415.00
410.00
4.00
405.00
2.00
400.00
0.00
Area Ratio
395.00
1
2
3
4
5
6
Header Size Ratio
7
8
Figure 6: Refrigerant MFR standard deviation versus
header size ratio for Header#1.
1
2
3
4
5
6
Header Size Ratio
7
8
Figure 7: Refrigerant pressure drop versus header size
ratio for Header#1.
CONCLUSIONS
An optimized 1 kW air cooled heat exchanger module is presented. Two different designs can be considered; the
first design leads to minimize the header frontal area however the refrigerant mass flow rate maldistribution will
increase. The alternative design has a low refrigerant mass flow rate maldistribution while increasing the header
frontal area. Optimum designs with area ratio between 1 % and 12 % are presented. The corresponding refrigerant
mass flow rate relative standard deviation is between 1 % and 14 %. The heat exchanger solver accounts for the
variation in refrigerant mass flow rates inside the tubes calculated based on the header 3D-CFD simulation. The
results confirmed the importance of header total height and header size ratio on the final design. Using headers with
larger height and low header size ratio improves the refrigerant mass flow rate distribution and reduces the
refrigerant pressure drop while increasing the total module volume. On the other hand, headers with smaller height
need larger header size ratio and lead to increased pressure drop while reducing the module total volume.
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012
2187, Page 7
NOMENCLATURE
English Symbols
CFD
Din
H.S.
L
LH,i
LH,o
MFR
Ntube
Nport
NTU
SR
V.S
Computational Fluid Dynamics
Tube inner diameter
Horizontal spacing
Tube length
Header inlet height
Header outlet height
Mass flow rate
Number of tubes
Number of ports
Number of transfer units
Header size ratio (LH,i / LH,o)
Vertical spacing
(mm)
(mm)
(mm)
(mm)
(mm)
(kg/s)
(mm)
Greek Symbols
ε
σMFR
∆Pref
∆Pair
Heat exchanger effectiveness
Refrigerant mass flow rate relative standard deviation
Total refrigerant pressure drop inside the header and tubes
Total air pressure drop
(Pa)
(Pa)
REFERENCES
Abdelaziz, O., 2009, Development of Multi-Scale, Multi-Physics, Analysis Capability and its Application to Novel
Heat Exchanger Design and Optimization, Ph.D. Dissertation, University of Maryland, College Park, MD,
USA.
Abdelaziz, O., Aute, V., Azarm, S., and Radermacher, R., 2010, Approximation assisted optimization for novel
compact heat exchanger designs, HVAC&R Research, vol.16, no.5: p. 707-728.
Fluent Inc., 2007, Fluent User’s Guide 6.3, Lebanon, NH.
Jiang, H. B., Aute, V. and Radermacher, R., 2006, CoilDesigner: a general-purpose simulation and design tool for
air-to-refrigerant heat exchangers, International Journal of Refrigeration-Revue Internationale Du Froid, vol.
29, no.4: p. 601-610.
Jing W., Jianbing, W., Liuyang, G., Suili, W., and Dayong, H., 2005, CFD optimization application on airside plate
fins of condenser coil of gravity-assisted heat pipe, translated by San Francisco, CA, United states: American
Society of Mechanical Engineers, p. 79-785.
Lee K. S., W. S. Kim and J. M. Si, 2001, Optimal shape and arrangement of staggered pins in the channel of a plate
heat exchanger, International Journal of Heat and Mass Transfer, vol. 44, no.17: p. 3223-3231.
Park K. and Moon, S., 2005, Optimal design of heat exchangers using the progressive quadratic response surface
model, International Journal of Heat and Mass Transfer, vol. 48, no. 11: p. 2126-2139.
Park K., Oh, P. K., and Lim, H. J., 2006, The application of the CFD and Kriging method to an optimization of heat
sink, International Journal of Heat and Mass Transfer, vol. 49, no.19-20: p. 3439-3447.
Saleh, K., Radermacher, R., Aute, V., and Azarm, S., 2011, Online Approximation Assisted Optimization of a
Novel Air-Cooled Heat Exchanger,10th IEA Heat Pump Conference 2011, Tokyo, Japan, paper no.00272.
Shah, R. K., 2006, Advances in science and technology of compact heat exchangers, Heat Transfer Eng., vol. 27,
no. 5: p. 3-22.
Shah, R. K. and London, A. L., 1978, Laminar flow forced convection in ducts, a supplement to Advances in Heat
Transfer, Academic, New York, USA.
ACKNOWLEDGEMENTS
The work presented in this paper was supported in part through a grant from the U.S. Office of Naval Research,
Grant # N000140710468. Such support does not constitute an endorsement by the funding agency of the opinions
expressed in the paper.
International Refrigeration and Air Conditioning Conference at Purdue, July 16-19, 2012