Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Experimental and Numerical Methodology For The Aero-Thermal Analysis of A

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

Applied Thermal Engineering 219 (2023) 119409

Contents lists available at ScienceDirect

Applied Thermal Engineering


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

Research Paper

Experimental and numerical methodology for the aero-thermal analysis of a


SACOC heat exchanger
Miguel Chávez-Modena a ,∗,1 , Leo Miguel González a ,2 , Eusebio Valero a ,3 , Alberto Broatch b ,4 ,
Jorge García-Tíscar b ,5 , Andrés Felgueroso b ,6
a
ETSIAE – School of Aeronautics, Universidad Politécnica de Madrid, Plaza Cardenal Cisneros 3, 28040 Madrid, Spain
b CMT – Motores Térmicos, Universitat Politècnica de València, Camino de Vera, 46022 Valencia, Spain

ARTICLE INFO ABSTRACT

Keywords: Surface air-cooled oil coolers (SACOC) are being used in the aerospace industry. This device is a passive heat
Heat exchangers exchanger used to dissipate a large amount of heat from the oil circuit of a turbofan engine to its secondary
SACOC flow with minimal impact on turbofan efficiency.
CFD
Typically, heat exchangers are formed by parallel fins and it is important to evaluate the interaction
Experimental aerodynamics
between the three-dimensional high-speed flow and the heat exchanger. The characterization of these devices
Thermal management
is usually performed in custom wind tunnels. In addition, it can be done through numerical simulations that
include the solution of the fluid equations in the flow field and a conjugate thermal problem. The benefit
of using numerical methods, once validated, is that it allows a complete and computationally affordable
analysis of the aerothermal performance of the SACOC. This work includes a comparison between numerical
and experimental results in terms of pressures and temperatures to achieve a complete characterization,
thus demonstrating the capability of the numerical model to obtain accurate results and allowing a better
understanding of the physical phenomenon.

1. Introduction be an interesting quantitative study. The goal of such designs is to


use the available surface as a heat exchanger between air and oil. To
Advanced oil cooling strategies for more efficient engine archi- increase the thermal performance, the wet area is increased by adding
tectures are one of the technological challenges that requires more longitudinal fins, reaching the required heat dissipation power. Such
intensive research. New engine designs based on the inlet air flow and a design implies a strong compromise between aerodynamic penalties,
the fuel circuit are approaching their limits as typical cold sources. quantified by the drag increase or pressure drop, and the improvement
Complex coupled problems are found in this kind of mechanical system, in the thermal performance of the heat exchanger. The developed
where the thermal part must be treated adequately. It should be taken research presents an innovative aero-thermal study by testing the new
into account that the heat removal by the aircraft structure will be lim- heat exchanger concept in a 3D shaped transonic wind tunnel capable
ited by using composite materials with lower operational temperature
of reproducing the flow condition within the bypass of an engine.
and thermal conductivity properties. Furthermore, the limitation of the
This surface air/oil heat exchanger (SACOC) can be located on the
maximum fuel temperature decreases the viability of the fuel tank as a
lip that separates the bypass flow of the engine from the core flow
cold source. Therefore, a novel engine cooling strategy seems to be a
between the fan and the downstream outlet guide vane, but other lo-
new industrial demand. In this context, the evaluation of the thermal
cations can be proposed. The SACOC is used to evacuate, in the bypass
performance of an Air Cooled Oil Cooler (ACOC) heat exchanger as-
sembled on the inner wall of the secondary duct of a turbofan could air stream, a large amount of heat contained in the oil (which therefore
contributes to the thrust of the engine) with minimal perturbation, and

∗ Corresponding author.
E-mail addresses: m.chavez@upm.es (M. Chávez-Modena), leo.gonzalez@upm.es (L.M. González), eusebio.valero@upm.es (E. Valero), abroatch@mot.upv.es
(A. Broatch), jorgarti@mot.upv.es (J. García-Tíscar), anfelrod@mot.upv.es (A. Felgueroso).
1
Assistant Professor, Dept. of Applied Mathematics, UPM; Center for Computational Simulation, UPM.
2
Associate Professor, Dept. of Fluid Mechanics and Propulsion, UPM; Center for Computational Simulation, UPM.
3
Full Professor, Dept. of Applied Mathematics, UPM; Center for Computational Simulation, UPM.
4
Full Professor, CMT–Motores Térmicos, UPV.
5
Assistant Professor, CMT–Motores Térmicos, UPV.
6
Graduate Research Assistant, CMT–Motores Térmicos, UPV.

https://doi.org/10.1016/j.applthermaleng.2022.119409
Received 4 April 2022; Received in revised form 15 September 2022; Accepted 26 September 2022
Available online 13 October 2022
1359-4311/© 2022 Elsevier Ltd. All rights reserved.
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

presents high levels and concentrations of mechanical stresses. The


Nomenclature superposition of the structural function of this SACOC, its de-icing
ability, and its cooling function is a completely new combination.
List of symbols Clearly, the performance of the SACOC is measured in terms of the
𝐶𝑝 Isobaric specific heat capacity (J/(kg⋅K)) maximum heat release capacity with minimal perturbation of the flow
in terms of pressure loss and acoustic emission.
𝜌 Density (kg/m3 )
Surface Air Cooled Oil Coolers are normally composed of parallel
𝜇 Dynamic viscosity (Pa⋅s)
fins integrated into the inner wall of the secondary duct of a turbofan.
𝐻 Height of the test section (m) The geometrical characteristics and location of the fins within the
𝐻𝐹 𝐼𝑁 Fin height (m) engine are designed to minimize aerodynamic effects (drag and noise
𝑘 Conductivity (W/(m⋅K)) increments) while maximizing thermal exchange. This heat exchanger
𝐿 Length of the test section (m) configuration introduces fewer aerodynamic perturbations than current
𝐿𝐹 𝐼𝑁 Fin length (m) plate/tube heat exchanger modules. From an aerodynamic point of
𝑚̇ Mass flow rate (kg/s) view, the measured averaged total pressure losses increased by approx-
𝑀 Mach number (–) imately 1% compared to the clean configuration (without the presence
𝑛 Normal direction (m) of the fin array). These losses were estimated for the entire annular
section of a turbofan with a bypass ratio equal to 6 [1]. The analysis
𝑝 Static pressure (Pa)
of the interaction between the three-dimensional high-velocity bypass
𝑝0 Outflow static pressure (Pa)
flow and the heat exchangers is essential to evaluate and optimize
𝑃𝑟 Prandtl number (–)
the aerothermodynamic performance, and to provide data for engine
𝑄̇ Heat transfer (W) modeling.
𝑅𝑒 Reynolds number (–) Computational fluid dynamics (CFD) approaches have been exten-
𝑆𝐹 𝐼𝑁 Separation between fins (m) sively applied to the study of flow and heat transfer in heat sinks, as
𝑇 Temperature (K) can be seen in the work by Jonsson and Moshfegh [2] and Biber and
𝑇𝑑𝑖𝑠𝑡 Experimental base temperature distribution Belady [3]. Outirba and Hendrick [4] conducted an experimental study
(K) on surface-air-cooled oil coolers (SACOCs), describing a new test rig
𝑢 Velocity (m/s) that allowed complete tests of SACOC breadboards. Ko et al. [5] studied
𝑈∞ Maximum inflow velocity (m/s) the effect of SACOC installation using numerical methods. In this paper,
an efficient numerical procedure was proposed and successfully demon-
𝑈 Overall heat transfer coefficient (W/(m2 ⋅K))
strated for the study of cooler installation involving bypass ducts, and
𝑐𝑠 Speed of sound (m/s)
important design variables were clearly identified. Kim et al. [6] con-
𝑊 𝐹 𝐼𝑁 Fin thickness (m)
ducted a detailed experimental study from which the results validated
𝑦+ Non-dimensional boundary layer distance this numerical work. To assess the performance of a heat exchanger,
(–) an appropriate performance metric should be used. Adams [7] and Doo
𝑋, 𝑌 , 𝑍 Spatial coordinates (m) et al. [8] used these factors to examine the aerothermal performance
ℎ Local heat transfer coefficient (W/(m2 ⋅K)) of variously shaped, primary surface heat exchangers. Although the
𝐶𝑓 Skin friction coefficient (–) loss of energy inside a heat exchanger usually arises near solid-wall
𝜏𝑤 Wall shear stress (Pa) regions, such as the fins, losses due to complex flow mixing inside the
𝐴 Fins wet area (m2 ) flow passage are also important. Doo et al. [9] evaluated the pressure
loss mechanism inside a cross-corrugated plate-type heat exchanger
List of subscripts using the volumetric entropy generation rate. This arrangement results
∞ Referred to inflow in a strong mixing layer between the plates. Although there have
been various fundamental studies on pin-fin shaped surfaces for heat
𝑓 Referred to the fluid
exchangers, most of these are done under the condition of idealized
𝑠 Referred to the solid
heat transfer, without considering the bypass effect, which makes the
𝑡 Total/stagnation value problem complex [10,11]. It is necessary to carry out studies on the
List of acronyms high-speed bypass effect in relation to the fin shapes of various heat
exchangers to understand the advantages of enhanced heat transfer
CFD Computational Fluid Dynamics characteristics for application to aero engines.
CHT Conjugate Heat Transfer Recent experiments in a particular rig to test air/oil heat exchanger
DMP Downstream Measurement Profile have been performed by UPV in collaboration with the von Karman
DMS Downstream Measurement Section Institute (VKI) and Purdue University [1,12]. The design consisted of
LDA Laser Doppler Anemometry fins aligned with the flow direction and integrated into the surface
RANS Reynolds-Averaged Navier Sotkes of the core/bypass flow splitter, downstream of the flow bifurcation,
RTD Resistance Temperature Detector following the typical swirl present just downstream of the fan.
As implementing this multi-fin approach in our case, including the
SACOC Surface Air-Cooled Oil Cooler
whole height of the bypass, would render the mass flow requirement
UMS Upstream Measurement Section
excessive, an alternative approach was implemented consisting of the
reduced-height wind tunnel depicted in Fig. 1.
This work, based on previous work presented in the AIAA AVIA-
TION 2021 FORUM [13], provides a novel analysis of a heat exchanger
therefore a minimal impact on Specific Fuel Consumption (SFC). The
performance by studying it at very high Re (∼ 106 ) and considerable
coolant source is the bypass mass flow delivered by the fan. In addition, large Mach numbers (∼0.4), replicating even a realistic velocity dis-
this device can also contribute to the de-icing of the separation lip tribution of a typical aeronautical case. We consider the configuration
between the core flow and the bypass flow, making efficient use of shown in Fig. 1, where the heat exchanger is placed downstream of the
the extracted heat. The SACOC is finally installed in a location that Outlet Guide Vane (OGV), as in the setup of Kim et al. [6,14]. Such a

2
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 1. SACOC configuration downstream of the turbofan OGV [15].

configuration can be found in modern high-bypass-ratio turbofans, such


as the CFM LEAP-1 A and the Rolls-Royce Trent XWB.
This study is quite original in its context as it includes a realistic 3D
geometry and the complexity given by compressibility and turbulent
regimes and is not comparable to other heat exchanger applications
such as cooling electronic devices where these complex regimes are
not so relevant. In addition, by analyzing a series of fins, the effects
of the outermost fin can be captured, unlike other investigations. A
numerical and experimental methodology is presented that is able to
solve the coupled problem and evaluate the thermal and aerodynamic
performance of the SACOC, together with a detailed comparison be-
tween both approaches aiming to prove the numerical methodology
as a useful and accurate design tool. The comparison between the
numerical and the experimental results is very complete and contains
both fluid dynamic comparisons in terms of pressure drops, local Mach
Fig. 2. Geometry used for the symmetric case with 8 fins and location of the
number and pressure profiles, but also thermal comparisons such as
measurement profiles.
temperature profiles and total heat transfer.

2. Methodology
measurement are shown in Fig. 2: UMS1 (𝑋∕𝐻 = 0), DMS1 (𝑋∕𝐻 =
2.1. Description of the problem 2.5) and DMS2 (𝑋∕𝐻 = 4), which stand for Upstream Measurement
Section 1 and Downstream Measurement Sections 1 and 2, respectively,
A simplified model is proposed to reduce the intrinsic complexity while the Downstream Measurement Profiles are defined as DMP1
found in the problem described above. First, a reduced part of the (𝑋∕𝐻 = 2.4), DMP2 (𝑋∕𝐻 = 2.8), DMP3 (𝑋∕𝐻 = 3.2) and DMP4
global geometry is used to examine the aerothermal performance of (𝑋∕𝐻 = 3.6).
a finned surface (see Fig. 2). To simplify the flow field, the location
of the SACOC geometry inside the engine is defined as matching the 2.2. Experimental setup
more recent setups of Kim et al. [6,14] and Gelain et al. [16], this
is, downstream of the outlet guide vanes (OGV). This position of The High Mass-Flow and High Temperature (HMHT) flow bench,
the SACOC is found in modern turbofan aero engines. As this work installed in the 8P laboratory of UPV, is a multipurpose experimen-
compares computational results with those obtained in a wind tunnel tal facility for aerodynamic studies in both automotive [17–19] and
without curvature, the slight curvature of the model normally present aeronautical engineering [20], as it provides a fairly high flow rate,
in industrial designs will be neglected in the rectangular computational more than 2 kg/s with cold flow and hot flows at temperatures that
domain. Geometric dimensions are normalized to the height of the wind can reach 1000 ◦ C with a flow rate of 1 kg/s. This facility is powered
tunnel, 𝐻. The dimensions of the wind tunnel are width, 𝑊 ∕𝐻 = by a 500 kW two-stage centrifugal compressor in series with a 1 MW
1, and length, 𝐿∕𝐻 = 5.35. The fins are characterized with length, diesel combustor. This setup allows performing tests in which both the
𝐿𝐹 𝐼𝑁∕𝐻 = 1.25, height, 𝐻𝐹 𝐼𝑁∕𝐻 = 0.2, with the leading and trailing Mach and Reynolds numbers of the flow can be controlled.
edge at 63.43 degrees, thickness, 𝑊 𝐹 𝐼𝑁∕𝐻 = 10−2 , and a separation In this facility, a continuous mass flow rate of 2.15 kg/s at 3 bar(A)
between fins, 𝑆𝐹 𝐼𝑁∕𝐻 = 2.175 ⋅ 10−2 . and 30 ◦ C can be available in the measurement room. To regulate the
The geometry used to examine the heat exchange performance total mass flow delivered to the test room, a series of venting valves are
contains 16 fins. The fins lie on a base with length 𝐿𝐵∕𝐻 = 1.33, height placed in the main pipe between two settling tanks. These reservoirs of
𝐻𝐵∕𝐻 = 1.5⋅10−2 and width 𝑊 𝐵∕𝐻 = 0.58. The base is surrounded by 3 m3 are capable of damping sudden changes in operating conditions,
thermoplastic material, with a distance of 𝐷𝑈 ∕𝐻 = 0.16 and a thickness helping to protect different parts of the facility and deliver a more
of 𝑊 𝑈 ∕𝐻 = 5 ⋅ 10−2 . Finally, the bottom wall is a thick steel plate uniform and controlled mass flow to the test room. A schematic of the
with thickness 𝑊 𝑆∕𝐻 = 5 ⋅ 10−2 . A symmetry condition is used for facility is shown in Fig. 3.
the mid plane 𝑍 = 0 and only half domain, 8 fins, is finally computed
(see Fig. 2). The fins are numbered so that fin 1 is the closest to the 2.2.1. Wind tunnel
symmetry plane, increasing the fin number in the Z direction, where In the test room of the HMHT stand, a custom modular rig is
the fin number 8 is the most external one. The leading edge of the fin installed. It is mainly formed by a settling chamber, a transition device,
is at 𝑋∕𝐻 = 0.75. Additionally, the locations of the velocity profile a distortion panel holder, and the test section where the SACOC is fixed.

3
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 3. Detail of the oil conditioning system (left) and schematic of the facility and wind tunnel (right).

The wind tunnel, depicted in Fig. 4, discharges with a large silencer that Table 1
Properties of the materials used in the numerical setup.
is then routed upward outside the building to avoid noise pollution. The
wind tunnel walls are made of steel AISI 304 [21]. Al AW7075 ULTEM Steel AISI
T651 [23] 1010 [24] 304 [21]
The flow is delivered from the HMHT stand through an S-shaped
Density, 𝜌 [kg/m3 ] 2810 1270 8055–7930
pipe. At the top of this pipe, a tap was set to characterize the temper-
Specific heat, 𝐶𝑝 [J/kg ⋅ K] 960 2000 480–500
ature of the incoming air. This pipe drives the flow into the settling Thermal cond., 𝑘𝑠 [W/m ⋅ K] 130 0.21 15.1–16.2
chamber via a pierced end designed with such flow properties that it
enhances a smooth transition to the test chamber. The settling chamber
is a reservoir intended to stagnate the fluid, decrease the level of
turbulence of the flow, and straighten the streamlines upstream of The SACOC in the experiments is a heat exchanger made of alu-
the test section. Several CFD studies were performed to determine the minum AW7075 T651 [23]. On the oil side, 18 fins with a height of
best configuration of the settling chamber and the best performance 30% 𝐻𝐹 𝐼𝑁 and also the same angle at the leading and trailing edges
was found with a configuration based on Purdue’s PETAL settling with the same thickness are symmetrically distributed. To enhance heat
chamber [22]. To achieve better flow conditions in the test section, transfer, the oil circuit is in counterflow, i.e. the oil flows beneath
two honeycomb straighteners are placed inside the settling chamber. the SACOC in the opposite direction to the air. Since the bottom wall
Afterwards, to gently drive the flow into the test section, a 3D- where the SACOC is mounted is made of steel, and to maintain the
printed transition connects the circular settling chamber with a heat exchanger as adiabatic as possible (extracting heat through the
squared-pipe. The interior surface of the transition has been treated to SACOC solely), a 3D printed oil pan is attached to the wall, whereas the
be as smooth as possible, and the geometry does not present abrupt exchanger is attached to the oil pan. The material used for the oil pan is
changes to avoid flow detachment. Then the flow develops in the ULTEM 1010 [24], a thermoplastic ideal for long-term heat resistance.
straight section and reaches a distortion screen capable of transform- The extra aluminum surface compared to the numerical case needed
ing a uniform flow profile into the required pressure profile, which to attach the exchanger to the oil pan can be considered negligible.
must mimic the profile measured in the actual turbofan bypass [16]. Regarding solid properties, Table 1 shows the aluminum, ULTEM 1010
Basically, the distortion screen is a 3D-printed honeycomb panel with and steel properties.
variable porosity that induces a determined pressure drop depending Finally, a straight section discharges the flow to the ambient. This
on that porosity (detail in Fig. 4). Pores have a hexagonal shape as pipe faces the ventilation system, so the air leaves the test room
they are structurally sound and produce less pressure drop across the smoothly.
panel. As the test section represents only a close region around the
actual exchanger, the profile is then reproduced with high accuracy 2.2.2. Oil conditioner
until close to the top wall of the tunnel, since the real engine is larger, Oil that responds to SAE 5w-30 rheological and thermal specifica-
and for the experiments a boundary layer will necessarily be generated tions [25,26] is used, although almost any type of lubricant can be
by the upper wall. Between the distortion screen and the test section, used in this circuit. The oil is driven throughout the circuit by means
a straight pipe is placed, so the first measurement plane, UMS1, is the of a pump. Upstream of the pump, a water-cooled heat exchanger
2.3 H downstream panel. reduces the temperature below 110 ◦ C, with a typical drop of 40 ◦ C.
The test section is a square duct with the lateral and top walls made The pump works at a constant speed of 1450 rpm with a power of
of borosilicate, specially treated for laser wavelength transparency. 370 W, so to regulate the mass flow in the circuit, a bypass with a
There are three apertures on the top wall of the test section to allow manually controlled valve is used. The oil then passes through another
intrusive measurements. In case non-intrusive techniques are required, heat exchanger devoted to keeping the flow meter temperature at the
the top wall can be replaced by a continuous borosilicate glass. To nominal value of operation, with a temperature typically under 60 ◦ C.
allow thermal measurements with infrared cameras, additional walls Between the heat exchanger and the flow meter, a filter eliminates
can be installed with specially treated windows in the SACOC region possible impurities that could be created along the circuit, using a series
for infrared transparency in the spectrum from 0.3 to 13 μm. These of paper walls with micrometric porosity. Once filtered, the fluid passes
walls are made of aluminum and have a hole in which a window can through a high-precision Coriolis KROHNE Optimass 3300C S04 mass
be attached. This attachment is a rectangular 3D printed part with the flow meter with a typical accuracy of 1% of the reading.
window off-center to allow a wider field of view when rotating the Subsequently, the temperature of the oil is increased by means of
attachment. an electric heater. This element is made up of two in-series electrical

4
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 4. Experimental test rig with distortion screen and SACOC details. (1) Settling chamber; (2) Honeycombs; (3) Distortion screen; (4) Traverse system; (5) SACOC.

resistors with 2.6 and 2.5 kW, respectively. A thermocouple is set after Table 2
Summary of the estimated experimental uncertainties.
the electric heater to feed a PID control system that regulates the power
Parameter Uncertainties
needed in the resistors to reach the desired fluid temperature. Finally,
the oil goes to the SACOC oil pan. To ensure that the pan is filled with 𝛥𝑝 1.7%
𝑝𝑡 ∕𝑝𝑜 1.2%
oil and that air bubbles are not retained below the aluminum, two oil
𝑇𝑡 ∕𝑇∞ 0.5%
reservoirs are placed before and after the exchanger, at higher altitude. 𝑇𝑏𝑎𝑠𝑒 ∕𝑇∞ 0.7%
Once the lubricant leaves the SACOC and the outlet reservoir, it 𝑄̇ 1.56%
returns to the oil storage tank. The pump suctions oil from this tank,
passing it through a filter that avoids the entry of dirt into the circuit.
In addition, a branch connects the tank to the main circuit downstream
SACOC can be calculated taking into account the mass flow rate of oil
of the pump, so if the pressure exceeds a certain value, the valve opens
circulating across the pan, assuming a constant heat capacity:
and returns the fluid to the tank. A scheme showing the operation of ( 4 )
the oil conditioner is shown in Fig. 3. ̇ 𝑝 ∑
𝑚𝑐 ∑4
𝑄̇ = 𝑚𝑐
̇ 𝑝 𝛥𝑇 = 𝑇in𝑖 − 𝑇out𝑖 (1)
4 𝑖=1 𝑖=1
2.2.3. Measurements Once the heat transfer is obtained, the next step is to derive the heat
For aerodynamic measurements, a set of Kiel probes of 3.2 mm transfer coefficient. To do so, Eq. (2) expresses the heat transfer as a
diameter was used. Kiel probes have a surrounding shield around function of the overall heat transfer coefficient, 𝑈 , and the logarithmic
the pressure port, so they are not influenced by swirling, being also mean temperature difference, 𝛥𝑇𝑙𝑚 :
less sensitive to the flow direction. In this way, Kiel probes are the ( ) ( )
𝑇2ℎ − 𝑇2𝑐 − 𝑇1ℎ − 𝑇1𝑐
best option to determine the total pressure in the section. To perform 𝑄̇ = 𝑈 𝐴𝛥𝑇𝑙𝑚 = 𝑈 𝐴 𝑇 −𝑇
, (2)
the measurements in an automated way, traverse systems have been ln 𝑇2ℎ −𝑇2𝑐
1ℎ 1𝑐
designed and constructed. In total, 3 biaxial traverses, one for each
where the subindex 2 refers to the trailing edge region of the fins,
measurement plane, are used. Each system is composed of two stepper
which in this case is the oil inlet and the air outlet, and the subindex
motors, one per each axis, controlled by an Arduino microcontroller 1 refers to the leading edge region of the fins, which corresponds to
via a MATLAB routine. GE UNIK 5000 sensors with a range up to 6 bar the air inlet and the oil outlet. Subindices ℎ and 𝑐 refer to hot and cold
and an accuracy of 0.2% F.S. are used to determine the pressure values. fluid, respectively; here, oil and air. Therefore, the overall heat transfer
The acquisition process is carried out by a NI PXI system, which is also coefficient is derived from Eq. (2).
operated through MATLAB. A FLIR A400 thermographic camera with an accuracy of ±2 K is
Furthermore, a three-axis Dantec Dynamics Laser Doppler Anemom- used to measure the wall temperature distribution of the specimen,
etry (LDA) system with a typical accuracy of 0.11% can be used to from which the heat transfer efficiency of the prototype can be es-
obtain velocity measurements in a non-intrusive way. This enables the timated. To validate the results obtained with the IR camera, K-type
analysis of other planes of the test section, beyond the three planes surface thermocouples can be introduced through the bottom screw
analyzed through the Kiel probes. Seeding was provided upstream of holes to measure the aluminum temperature at different points. For a
the test section with a TSI 9306 six-jet atomizer. better determination of the temperature distribution, the SACOC and its
For thermal measurements, the Kiel probes also incorporate class surroundings are painted in black with a special graphite-based paint
1 T-type thermocouples, so both total pressure and temperature can with high conductivity, which does not modify the temperature on the
be acquired at the same time. However, because of the gap between surface. The result obtained with the thermographic camera is shown
the location of the thermocouple and the top of the head, the bottom- in Fig. 5. The evolution of the temperatures shown on the right-hand
side of the figure evidences a considerable increase between the leading
most region cannot be reached by this thermocouple. To perform a
and the trailing edge. It is also interesting to observe the effect of the
more precise analysis of the temperature variation, 4 class 1 K-type
insulation material on the sudden decrease in temperature after the
thermocouples are centered downstream of the SACOC fins on the
fins. However, although temperatures follow a similar trend along the
lower wall to see the evolution of the thermal boundary layer at
SACOC span, they decrease as they approach the lateral borders.
DMP1, DMP2, DMP3, and DMP4. These thermocouples can be moved
A summary of the estimated uncertainties associated with typical
to analyze the temperature distribution along their verticals. measurement values is presented in Table 2.
To measure the heat exchanged during the process, both the inlet
and outlet oil temperatures of the oil pan are measured at different 2.3. Numerical
radii and at the center of the pipe: three Pt1000 RTDs are used to
characterize the radial temperature distribution and one is inserted co- Consider the steady turbulent compressible fluid that flows through
axially to the pipe line. By averaging the temperature measurements at a duct that contains a finned surface where a three-dimensional con-
each section and computing the difference, the heat dissipated by the jugate heat transfer problem is also coupled. A schematic description

5
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 5. Image composition from the IR camera (left) and temperature evolution in the center line (𝑍∕𝐻 = 0) and the external line (𝑍∕𝐻 = 0.235) (right).

of the model problem is given in Fig. 2. The problem domain can


be decomposed into two parts: the fluid part where the compressible
Reynolds-averaged Navier–Stokes (RANS) equations are solved and a
solid part where a heat transfer problem is also solved. Both problems
are coupled by heat transfer through the flat plate surface, where no-
slip velocity and heat conduction conditions are applied. Below, the
sub-indices 𝑓 and 𝑠 are used to determine the regions of fluid and
solid temperature, respectively. Note that the solid is divided into three
different materials to reproduce the experiment. Aluminum for the
SACOC model, ULTEM 1010 [24] where the SACOC model is located,
and steel for the floor of the wind tunnel. The solid properties used in
the numerical simulations were shown in Table 1.

2.3.1. Governing equations


Considering the air as an ideal gas, the continuity, momentum, and
Fig. 6. Boundary conditions imposed at the different surfaces of half fluid and solid
energy non-dimensional equations for the steady, compressible, and domain (symmetric hypothesis).
turbulent flow are:
𝜕𝜌𝑢𝑖
= 0, (3)
𝜕𝑥𝑖 direction of the wall. For the bottom wall, interface between fins, and
[ ]
𝜕𝜌𝑢𝑖 𝑢𝑗 𝜕𝑝 𝜕 1 𝜕𝑢𝑖 𝜕𝑢𝑗 the fins, no-slip boundary condition is used for the velocity, and heat
= + ( + ) − 𝜌𝑢′𝑖 𝑢′𝑗 , (4)
𝜕𝑥𝑗 𝜕𝑥𝑖 𝜕𝑥𝑗 𝑅𝑒 𝜕𝑥𝑗 𝜕𝑥𝑖 transfer is allowed between the three solids and the air.
( ) 𝜕𝑇
𝜕𝜌𝑢𝑖 𝑇𝑓 𝜕 1 𝜕𝑇𝑓 The boundary conditions 𝜕𝑛𝑠 = 0 and 𝑇𝑠 = 𝑇𝑑𝑖𝑠𝑡 are used for the
= − − + 𝜌𝑢′𝑖 𝑇𝑓 , (5)
𝜕𝑥𝑖 𝜕𝑥𝑖 𝑅𝑒𝑃 𝑟 𝜕𝑥𝑖 surface of the fin and the horizontal base of the fin, respectively, where
𝑇𝑑𝑖𝑠𝑡 is the experimental temperature distribution shown in Fig. 5.
where the equation of state for perfect gases is also added to the system
Additionally, the coupling between the compressible fluid problem and
(3)–(5).
the heat transfer problem is given by the CHT boundary conditions:
The equation that models the heat conduction effects of the conju-
gate heat transfer problem in the solid region inside the fins and floor 𝑇𝑠 = 𝑇𝑓 , (7)
is as follows: 𝜕𝑇𝑓
𝜕𝑇𝑠
𝑘𝑠 = 𝑘𝑓 . (8)
∇2 𝑇𝑠 = 0, (6) 𝜕𝑛 𝜕𝑛
They are applied to the body/fluid interface providing the continu-
where 𝑇𝑠 is the solid temperature.
ity of thermal fields and heat flux between the body and flow at the
Distances are non-dimensionalized with the wind tunnel height 𝐻,
interface.
velocity with the maximum inflow value 𝑈∞ , pressure with the outflow
static pressure 𝑃𝑠 , temperatures and density with the inflow values 𝑇∞
and 𝜌∞ respectively. Consequently, the non-dimensional fluid variables 2.3.3. Simulation parameters
in the equations are listed as: 𝜌 is the density, 𝑢𝑖 is the velocity in the 𝑖 As compressible effects and turbulent effects are taken into account,
direction, 𝑝 is the pressure, 𝑢′𝑖 is the turbulent deviation of the velocity, the energy equation is coupled with the momentum equation, and the
𝑇𝑓 is the temperature. speed of sound is not fixed. A finite-volume code is used to discretize
The dominant nondimensional numbers are: the Reynolds number both the fluid mechanics and the heat transfer coupled problem.
defined as 𝑅𝑒 = 𝜌∞ 𝑈∞ LFIN∕𝜇, the Mach number as 𝑀 = 𝑈 ∕𝑐𝑠 where The fluid problem is solved through the RANS equations for tur-
𝑐𝑠 is speed of sound, and the Prandtl number 𝑃 𝑟 = 𝐶𝑝 𝜇∕𝑘𝑓 where 𝜇 bulent flow and its solution is coupled with the elliptic heat problem,
the fluid viscosity, 𝑘𝑓 = 𝑘𝑎𝑖𝑟 the fluid (air) thermal conductivity and 𝐶𝑝 which solves the temperature field in the solid part. Turbulent effects
specific heat. The viscosity is computed through Sutherland’s law. are added using a 𝑘 − 𝜔 SST model [27]. The computational domain is
discretized with 5.91 ⋅ 106 hexahedron. Fig. 7 shows the grid used for
2.3.2. Boundary conditions the simulations. More details about grid generation and convergence
Eqs. (3)–(5) are complemented with the boundary conditions shown are detailed in a previous work [28]. To correctly solve the velocity
in Fig. 6. As the problem presents a symmetry plane 𝑍 = 0, a symmetry boundary layer, we have ensured that 𝑦+ < 1. As the fluid is a gas,
boundary is included in the simulation (see Fig. 6). For the inflow 𝑃 𝑟 ≈ 1, the velocity and the thickness of the thermal boundary layers
boundary conditions, an experimentally measured inflow total pressure are equivalent 𝛿𝑣 ≈ 𝛿𝑇 .
and temperature profile, turbulence intensity, and viscosity ratio values The coupled system of Eqs. (3)–(5) and (6) are solved using an
are imposed. For the outflow, the pressure outlet boundary condition, implicit second-order discretization method to compute the Navier–
𝑝 = 𝑝0 , is used and the temperature is fixed to the inlet value, 𝑇𝑓 = 𝑇∞ . Stokes equations and to calculate the density, temperature, pressure,
For the top and the other lateral walls no-slip and adiabatic bound- and velocity fields at the steady state. For these steady simulations,
ary conditions are considered, with 𝜕𝑇 ∕𝜕𝑛 = 0 where 𝑛 is the normal an iterative process is used to ensure the steady-state condition to

6
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 7. Global perspective of grid and details of the fin and base zone.

Fig. 8. Streamlines colored by Mach number (top) and line integral convolution colored with wall shear stress (bottom) at the interior side of fin 1 and outer side of fin 8.

Fig. 9. Skin friction coefficient on fin surface.

Fig. 10. Temperature distribution over the surface of fin 1 (closest to the symmetry plane) and fin 8 (most external).

be satisfied with low residuals. To accelerate solver convergence, a The coupled heat transfer within a fluid and an adjoining solid, known
coupled algebraic multigrid method based on fixed V-cycles is used. as conjugate heat transfer, is computed during the simulation.
For a CHT analysis, the energy equation is solved throughout the
fluid and solid solution domain with an efficient implicit thermal cou- 3. Results
pling at the fluid/solid interface. In the body/fluid interface continuity
of the thermal fields is provided by specifying the equality of tempera- An experimental campaign was carried out to characterize how
ture and heat fluxes of a body and a flow in the vicinity of the interface. the SACOC performed under different circumstances, and one of the

7
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 11. Heat transfer coefficient on fin surface.

Table 3
Comparison of the non-dimensionalized pressure drop 𝛥𝑝∕𝑝0 and the overall heat
transfer coefficient 𝑈 and relative error (in parentheses).
𝛥𝑝∕𝑝0 [–] 𝑈 [W∕m2 K]
Exp. 6.72 × 10−3 896.0
Num. 6.51 × 10−3 (3.1%) 885.3 (1.2%)

main objectives of this paper is to compare the experimental results


with those obtained by numerical computations carried out in parallel. Fig. 12. Mach number contour comparison between experimental (top) and numerical
Turbulent and compressible effects should be taken into account, as the (bottom) results at the Z/H=0 section.
Reynolds, Mach, and Prandtl numbers are 𝑅𝑒 = 1.05 × 106 , 𝑀∞ = 0.36,
and 𝑃 𝑟 = 0.707 respectively. In this work, experimental validation
and flow analysis are offered, including aerodynamic and thermal
measurements.

3.1. Pressure drop and heat transfer

The general results for pressure drop and heat transfer coefficient
(see Table 3) show fairly good agreement in both cases. The non-
dimensionalized pressure drop by the static pressure is computed as the
Fig. 13. Mach number profile comparison between experimental and numerical results
total pressure difference between UMS1 and DMS2 in a region laterally in section Z/H=0 section at X/H=3 (left) and X/H=4 (right).
limited by the planes 𝑍∕𝐻 = ±0.3, so the fins with some margins
are included, and the uncertainties induced by the impossibility of
characterizing the flow very close to the lateral walls do not affect
The thermal behavior of fins 1 and 8 is also studied. The distribution
the calculations. In the experiments, a slightly higher pressure drop
of the temperature field on the two fin surfaces is represented in
is observed, probably induced by the irregularities and gaps intrinsic
Fig. 10. It can be seen that for both fins, the heat exchanged increases
to the tunnel that the CFD obviates. In the heat exchanged, a similar
the fin temperature in the stream-wise direction, and the temperature
result of slightly higher values in the experiments can be observed. In
distribution is very similar. Note the upward temperature distribution
this case, the extra heat measured can be explained by the imperfect
deflection, enhanced by the streamlines deflection shown in Fig. 8.
adiabatic insulation of the oil pan, which releases part of the heat as ̇
In Fig. 11 the local heat transfer coefficient (ℎ = 𝑄∕(𝐴(𝑇 𝑠 − 𝑇∞ ))) for
free convection to the test room through its base.
both sides of the same fin is shown, where 𝐴 is the cell wet area (the
area of the solid in contact with the fluid) and 𝑇𝑠 is the temperature
3.2. Nearby fin flow
of the cell, which varies from one cell to another. Despite a similar
temperature distribution when fins 1 and 8 are compared, the local heat
In Fig. 8 the streamlines on the inner side of the closest fin to
transfer coefficient contains some differences (see Fig. 11), showing
the symmetry plane (fin 1) and the outer side of the external one
(fin 8) are colored with the Mach number. In the same figure, the higher heat transfer values on the outer side of the exterior one (fin
integral line convolution is shown, which is a postprocessing technique 8) compared to the rest of the internal fins. Note that the values of the
to visualize the flow vector field, highlighting better separated areas heat transfer in higher area of the fins are related to the increase of the
of the flow. Focusing on the flow between the fins, we can observe an skin friction produced by the separation flow, being higher on the outer
acceleration at the leading edge of the fin and how the streamlines are side of fin 8 that is affected by the free stream. The rest of the internal
slightly deflected upward (see Fig. 8 (top)). Besides, the flow pattern fins contribute similarly to the global heat transfer and the flat plate
and the blue area depicts a separation region located along the leading adds 6% to the total heat transfer of the SACOC. With these figures,
edge (see Fig. 8 (bottom)). The deflection and separation regions are the fin efficiency can be calculated [29], having an average value of
produced by the inclination angle of the leading edge. The size of 0.939.
the separation region observed at the leading edge depends on the fin
position, being larger on the outer side of fin 8 as the flow is not so 3.3. Fin wake
constrained as it is in the inner fins.
Additionally, the skin friction coefficient (𝐶𝑓 = |𝜏𝑤 |∕( 12 𝜌∞ 𝑢2∞ )) In Fig. 12 Mach number contours are compared between the exper-
between the most centered (fin 1) and the most external (fin 8) is imental results measured with the LDA technique, and the numerical
compared in Fig. 9. The outer side of fin 8, where the contact flow results in the fin wake region in the 𝑍∕𝐻 = 0 plane. The wake
is less confined than in the other cases, presents a clearly larger shear generated behind the fins, 𝑋∕𝐻 > 2.45, characterizes a recirculation
range compared to the other three because of the separation region area and thermal diffusion behind the fins. We can observe some
shown in Fig. 8. differences between the experimental and the numerical wakes, with

8
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 14. Non-dimensional total pressure profile comparison between experimental and numerical results at UMS1, DMS1 and DMS2.

Fig. 15. Vorticity isocontour at different YZ sections.

the experimental wake slightly thinner and shorter than the numerical also explains the reason behind the fact that the deviation compared
one (see Fig. 13). with the experiments decreases along the streamwise direction.
Next, a deeper understanding of the different sections along the
wake region is presented in Fig. 14. These figures compare the ex- 4. Conclusion
perimental and numerical total pressure profiles in sections UMS1,
DMS1, and DMS2. These results are obtained by averaging the non- Surface Air-Cooled Oil Coolers are gaining more and more impor-
dimensional pressure values between 𝑍∕𝐻 = ±0.22, which are consid- tance in the aerospace sector to refrigerate the lubricant. They use the
ered regions of interest. cold air from the bypass as cold sink in the exchange process, taking
The total pressure at UMS1 obtained by CFD is the result of using also advantage of releasing the enthalpy present in the hot oil into the
propulsive flow. However, it is crucial to maintain the aerodynamic
the experimental velocity data as the inflow boundary condition in that
impact of the heat exchanger as low as possible for propulsive efficiency
section. In general, all sections present very similar values and trends,
purposes.
with the numerical results on most occasions within the standard devi-
This work presents a comparison between the experimental charac-
ation of the experimental data. It seems noteworthy that the dispersion
terization of a SACOC and the corresponding numerical computations
in the Kiel measurements increases closer to the top wall, which can be
of the equivalent fluid–thermal problem. The numerical methodology
related to the openings in the wall where the probes were inserted.
is based on a finite-volume code that is able to compute the turbulent
Finally, Fig. 15 shows the contours of the vorticity field in different and compressible flow that occurs in this heat exchanger coupling the
cross-sections of the SACOC obtained computationally. It can be easily result with the associated thermal problem. Regarding the experimental
observed in this representation how the two main vortices present on validation, the total pressure and temperature profiles, together with
the outer side of the external fin are captured. A detailed view of the heat release and the pressure drop are compared with their equiv-
the vorticity field between the fins of a particular cross-section is also alent numerical counterparts and are found to present consistent and
presented. matching results.
For a complete wake characterization, the thermal analysis is com- In summary, a robust aerothermal simulation methodology for this
pared with the experimental and numerical temperature profiles in four type of surface heat exchanger working under realistic conditions of
different sections at 𝑍∕𝐻 = 0. Fig. 16 shows a consistent match in the velocity distribution, Mach and Reynolds has been validated. Moreover,
temperature slope and the location of the maximum value. The main a group of several fins has been considered, capturing the effect of the
differences found can be explained by a velocity under-prediction in outermost fin. The resulting numerical data, supported by the experi-
the simulations, which influences the values of total temperature. This mental campaign, can therefore be used to perform an analysis of the

9
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

Fig. 16. Non-dimensional total temperature profile comparison between experimental and numerical results at DMP1, DMP2, DMP3 and DMP4.

relevant flow structures as they relate to the aerothermal performance [6] S. Kim, J.K. Min, M.Y. Ha, C. Son, Investigation of high-speed bypass effect on
of the SACOC, and to evaluate the pressure drop caused by the presence the performance of the surface air-oil heat exchanger for an aero engine, Int. J.
Heat Mass Transfer 77 (2014) 321–334.
of the heat exchanger. Analysis of the CFD results allows an accurate
[7] J.C. Adams, Advanced Heat Transfer Surfaces for Gas Turbine Heat
analysis and visualization of the turbulent thermal wake, which may
Exchangers (Ph.D. thesis), 19, University of Oxford, 2004.
have an appreciable impact on the downstream components of the [8] J.H. Doo, M.Y. Ha, J.K. Min, R. Stieger, A. Rolt, C. Son, An investigation
engine. of crosscorrugated heat exchanger primary surfaces for advanced intercooled-
cycle aero engines. Part-I: Novel geometry of primary surface, Int. J. Heat Mass
Declaration of competing interest Transfer 55 (2012) 5256–5267.
[9] J.H. Doo, M.Y. Ha, J.K. Min, R. Stieger, A. Rolt, C. Son, An investigation of
crosscorrugated heat exchanger primary surfaces for advanced intercooled-cycle
The authors declare that they have no known competing finan-
aero engines (Part-II: Design optimization of primary surface), Int. J. Heat Mass
cial interests or personal relationships that could have appeared to Transfer 61 (2013) 138–148.
influence the work reported in this paper. [10] Y.L.H. Z. G. Qu, Three-dimensional numerical simulation on laminar heat transfer
and fluid flow characteristic of strip fin surface with x arrangement of strips,
Data availability Trans. ASME, J. Heat Transfer 126 (2004) 697–707.
[11] T.S.L. W. Q. Tao, The influence of strip location on the pressure drop and heat
The authors do not have permission to share data. transfer performance of a slotted fin, Numer. Heat Transfer A 52 (2007) 463–480.
[12] L. Villafañe, G. Paniagua, Aerodynamic impact of finned heat exchangers on
transonic flows, Exp. Therm Fluid Sci. 97 (2018) 223–236, http://dx.doi.org/10.
Acknowledgments 1016/j.expthermflusci.2018.04.012.
[13] M. Chavez Modena, L.M. González, E. Valero, A. Broatch, J. García-Tíscar,
This project has received funding from Clean Sky 2 Joint Undertak- A. Felgueroso, Numerical and experimental analysis of thermo-aerodynamic
ing under the European Union’s Horizon 2020 research and innovation performance in an aero engine surface heat exchanger, in: AIAA AVIATION 2021
programme under grant agreement No 831977 Aerodynamic upgrade FORUM, 2021, p. 2901.
[14] M. Kim, M.Y. Ha, J.K. Min, A numerical study on various pin.fin shaped surface
of Surface Air-Cooled Oil Coolers (SACOC). Leo M. González acknowl-
air-oil heat exchangers for an aero gas-turbine engine, Int. J. Heat Mass Transfer
edges the financial support from the Spanish Ministry of Science, 93 (2016) 637–652.
Innovation and Universities (MCIU) under grant RTI2018-096791-B- [15] A. Broatch, P. Olmeda, J. García-Tíscar, A. Felgueroso, M. Chávez-Modena, L.
C21 Hidrodinámica de elementos de amortiguamiento del movimiento de González, M. Gelain, A. Couilleaux, Experimental aerothermal characterization
aerogeneradores flotantes. A. Felgueroso is supported through the Pro- of surface air-cooled oil coolers for turbofan engines, Int. J. Heat Mass Transfer
grama de Apoyo para la Investigación 𝑦 Desarrollo of the Universitat 190 (2022) 122775.
[16] M.C. Gelain, A. Couilleaux, M. Errera, R. Vicquelin, O. Gicquel, Conjugate heat
Politècnica de València under grant PAID-01-20 n◦ 21589. The authors
transfer analysis of a surface air-cooled oil cooler (SACOC) installed in a turbofan
gratefully acknowledge the Universidad Politécnica de Madrid (www. by-pass duct, in: AIAA AVIATION 2021 FORUM, 2021, p. 3163.
upm.es) for providing computing resources on Magerit Supercomputer. [17] A. Torregrosa, A. Broatch, J. García-Tíscar, F. Roig, Experimental verification
We also thank Safran Aircraft Engines for their kind permission to share of hydrodynamic similarity in hot flows, Exp. Therm Fluid Sci. (2020) 110220,
the data presented in this publication. http://dx.doi.org/10.1016/j.expthermflusci.2020.110220.
[18] J.R. Serrano, F.J. Arnau, L.M. Gracía-Cuevas, V. Samala, L. Smith, Experimental
References approach for the characterization and performance analysis of twin entry radial-
inflow turbines in a gas stand and with different flow admission conditions, Appl.
Therm. Eng. 159 (2019) 113737, http://dx.doi.org/10.1016/j.applthermaleng.
[1] J. Sousa, L. Villafane, G. Paniagua, Thermal analysis and modeling of surface
2019.113737.
heat exchangers operating in the transonic regime, Energy 64 (2014) 961–969,
[19] J.R. Serrano, F.J. Arnau, L.M. García-Cuevas, P. Soler, R. Cheung, Experimental
http://dx.doi.org/10.1016/j.energy.2013.11.032.
validation of a one-dimensional twin-entry radial turbine model under non-linear
[2] H. Jonsson, B. Moshfegh, Enhancement of the cooling performance of circular
pulse conditions, Int. J. Engine Res. 22 (2) (2019) 390–406, http://dx.doi.org/
pin fin heat sinks under flow by-pass conditions, in: Proceedings of the Eighth
10.1177/1468087419869157.
IEEE Inter Society Conference on Thermal Phenomena (ITHERM), San Diego, CA,
vol. 24, (2) 2002, pp. 425–432. [20] C. Research, D.I. Service, Novel system removes debris from compressor, Res.
[3] C.R. Biber, C.L. Belady, Pressure drop predictions for heat sinks: What is the Results Mag. 50 (2016) 22.
best method? Adv. Electron. Packag. 19 (2) (1997) 1829–1835. [21] Steel AISI 304 properties, 2021, http://www.matweb.com/search/datasheet.
[4] B. Outirba, P. Hendrick, Development and testing of advanced surface oil coolers aspx?matguid=bd20a4281ae3430d97cfbebf6904ec50. (Accessed 10 May 2021).
for aero-engines, ISABE (2013). [22] G. Paniagua, D. Cuadrado, J. Saavedra, V. Andreoli, T. Meyer, J.P. Solano, R.
[5] J.A. Ko, S.K. Kim, M.Y. Ha, J.K. Min, R. Stieger, S. Mason, J.H. Doo, C. Son, A Herrero, S. Meyer, D. Lawrence, Design of the purdue experimental turbine
study on the installation of the surface air-oil heat exchanger for the application aerothermal laboratory for optical and surface aerothermal measurements, J.
to aero gas-turbine engine, ISABE 19 (2013). Eng. Gas Turbines Power 141 (1) (2018) http://dx.doi.org/10.1115/1.4040683.

10
M. Chávez-Modena et al. Applied Thermal Engineering 219 (2023) 119409

[23] A.G.C.F. Products, Characteristics of aluminium 7075, 2021, https://www. [26] S. Wrenick, P. Sutor, H. Pangilinan, E.E. Schwarz, Heat transfer properties of
spacematdb.com/spacemat/manudatasheets/alcoa_alloy_7075.pdf. (Accessed 30 engine oils, in: World Tribology Congress III, Volume 1, ASMEDC, 2005, pp.
May 2021). 595–596, http://dx.doi.org/10.1115/WTC2005-64316.
[24] ULTEM 1010 properties, 2021, https://www.stratasys.com/es/materials/search/ [27] F. Menter, Zonal two equation 𝑘 − 𝜔 turbulence models for aerodynamic flows,
ultem1010. (Accessed 26 May 2021). in: 23rd Fluid Dynamics, Plasmadynamics, and Lasers Conference, 1993, p. 2906.
[25] G.K. Gyimah, Z.-n. Guo, P. Huang, S.-z. Jiang, G.C. Barber, G.-x. Liu, J.W. Liu, [28] M. Chávez-Modena, L.M. González, E. Valero, Numerical optimization of the fin
Improvers of pressure-viscosity coefficients of two-phase liquid-solid lubricants, shape experiments of a heat conjugate problem surface air/oil heat exchanger
DEStech Trans. Eng. Technol. Res. (2016) http://dx.doi.org/10.12783/dtetr/ (SACOC), Int. J. Heat Mass Transfer 182 (2022) 121971.
amita2016/3562. [29] J.E. Hesselgreaves, R. Law, D.A. Reay, Compact Heat Exchangers. Chapter 7
-Thermal Design, second ed., Elsevier, 2017, pp. 275–360, http://dx.doi.org/10.
1016/b978-0-08-100305-3.00007-0.

11

You might also like