Design and Analysis of an Intercooled
`Turbofan Engine
`
`Lei Xu
`e-mail: lei@chalmers.se
`Tomas Grönstedt1
`e-mail: tomas.gronstedt@chalmers.se
`
`Department of Applied Mechanics,
`Division of Fluid Dynamics,
`Chalmers University of Technology,
`Gothenburg 41296, Sweden
`
`The performance of an intercooled turbofan engine is analyzed by
`multidisciplinary optimization. A model for making preliminary
`simplified analysis of the mechanical design of the engine is
`coupled to an aircraft model and an engine performance model. A
`conventional turbofan engine with technology representative for a
`year 2013 entry of service is compared with a corresponding in-
`tercooled engine. A mission fuel burn reduction of 3.4% is ob-
`served. The results are analyzed in terms of the relevant con-
`straints such as compressor exit
`temperature,
`turbine entry
`temperature, turbine rotor blade temperature, and compressor exit
`blade height. It is shown that the gas path of an intercooled en-
`gine for medium range commercial transport applications, having
`an overall pressure ratio greater than 70 in top of climb, may still
`be optimized to fulfill a compressor exit blade height constraint.
`This indicates that a state of the art high pressure compressor
`efficiency can be achieved. Empirical data and a parametric com-
`putational fluid dynamics (CFD) study are used to verify the in-
`tercooler heat transfer and pressure loss characteristics.
`关DOI: 10.1115/1.4000857兴
`
`1
`
`Introduction
`A range of questions must be addressed to establish a high
`performance engine-intercooler system. Apart from selecting op-
`timal cycle parameters, the intercooler flow configuration has to
`be defined. In the process, conflicting requirements on pressure
`losses, heat transfer, and weight have to be met. In particular, the
`high overall pressure ratio 共OPR兲 cycles resulting when optimiz-
`ing intercooled engines may require a novel approach for optimiz-
`ing gas paths. A common tradition for one stage high pressure
`turbine 共HPT兲 turbofans has been to tilt the combustor radially in
`order to create a variation between the high pressure compressor
`共HPC兲 exit hub radius and the HPT entry hub radius. This has
`allowed an HPC with a pressure ratio greater than 10.0 to be
`driven by a single stage HPT. In this work, the radially tilted
`combustor design approach is explored in conjuntion with a two
`stage HPT and a three spool engine architecture to allow full
`optimization of the intercooled engine cycle.
`
`2
`
`Intercooler Design and Modeling
`
`2.1 Intercooler Engine Integration. Several design consid-
`erations of the aeroengine-intercooler relate to its external flow
`conditions. The intercooler is mounted inside the bypass channel,
`as illustrated in the upper half in Fig. 1. The diffuser of the inter-
`
`1Corresponding author.
`Contributed by the Aircraft Engine Committee of ASME for publication in the
`JOURNAL OF ENGINEERING FOR GAS TURBINES AND POWER. Manuscript received: June 22,
`2009; final manuscript revised: November 4, 2009; published online August 10,
`2010. Editor: Dilip R. Ballal.
`
`cooler splits the bypass flow into an internal bypass flow and an
`external bypass flow. The static pressure on the up- and downsides
`of the diffuser trailing edge must match each other. Thus, the
`larger the pressure loss is on the internal flow side, the smaller the
`flow speed and the mass flow rate will become. Here, we intro-
`duce the intercooler bypass ratio ⌽ as the ratio of the mass flow
`on the external side of the diffuser to the mass flow on the internal
`side. A 1% external pressure loss over the heat exchanger may
`lead to a ⌽ of 1.0, whereas a 10% pressure loss may lead to a ⌽
`in the range of 10.0–20.0, depending on other design parameters.
`Thus, care must be taken to avoid excessive pressure losses on the
`intercooler external side, and thus, to guarantee that enough air
`passes through the internal bypass channel.
`
`Intercooler Analysis. At the start of the design process,
`2.2
`several common heat exchanger surface types were considered
`such as fin plate surfaces, tube fin surfaces, and tubular surfaces.
`A tubular heat exchanger was selected in this study since it was
`expected to suffer less from the problem of thermal stresses. Two
`tubular configurations have been analyzed. The analysis of the
`initial configuration has been presented in Ref. 关1兴.
`One problem identified during the evaluation of the initial con-
`figuration was the large pressure losses occurring on the external
`surfaces. From the test results of Kays and London 关2兴, it can be
`seen that the staggered tube bank used for the initial configuration
`has a larger friction factor than most other configurations. Com-
`bined with a relatively high flow speed in the bypass channel, this
`configuration generates high pressure losses on its external sur-
`faces. As mentioned in Sec. 2.1, this will greatly reduce the
`amount of air passing over the external intercooler surface, and
`hence, limit the heat exchange. To improve the design and resolve
`some of the problems mentioned above, a refined configuration
`was conceived. First, the tube shape is changed from the circular
`cross section to an elliptic one. Due to the streamlined shape, the
`total drag force on the elliptic tubes is greatly reduced compared
`with tubes with circular cross section. As explained earlier, this
`will allow more air to pass through the internal bypass channel for
`heat exchange. A second advantage of the elliptic tubes is the
`larger wetted surface it has compared with the round tubes having
`the same cross-sectional area. Additionally, the elliptic shape re-
`duces vortex shedding tremendously. This alleviates problems as-
`sociated with flow induced vibrations, which could be a serious
`problem in real applications. Also, the overall arrangement of the
`refined configurations is more regular in its shape, which enables
`a more accurate assessment of the performance using CFD tools.
`The refined configuration features a standard shape multi pass
`overall counterflow heat exchanger. Considerably more useful
`validation information could therefore be found in the literature.
`Details of the design configuration can be found in Ref. 关3兴.
`2.3
`Intercooler Performance Modeling. Before the impact
`of the intercooler on the engine and the flight mission can be
`studied, a model based on correlations is created to predict the
`pressure loss and heat transfer of the flow through the intercooler.
`This model is then incorporated into the engine system evaluation
`tool GESTPAN 关4兴.
`relatively accurately
`that
`Correlation based methods exist
`model the internal side performance. The selected methods have
`been described previously in Ref. 关1兴. In contrast to the internal
`side, no correlation or test data with the suitable range of Rey-
`nolds number for the external side could be found in open litera-
`ture. To have a reasonably accurate assessment, a CFD study was
`carried out to evaluate the pressure loss and heat transfer.
`The commercial softwares ICEM-CFD and CFX are used to gen-
`erate the mesh and compute the flow on the external side of the
`heat exchanger. Preliminary computations show that the flow pat-
`tern does not change much after passing through three to four tube
`rows; hence, five tube rows are meshed. The whole domain con-
`sisted of about 700,000 hexagonal cells, and the mesh was con-
`centrated around the tubes to resolve the boundary layers. The
`
`Journal of Engineering for Gas Turbines and Power
`Copyright © 2010 by ASME
`
`NOVEMBER 2010, Vol. 132 / 114503-1
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`GE v. UTC
`Trial IPR2016-00952
`
`UTC-2022.001
`
`

`

`Fig. 1 Cross sectional drawings of optimal engines
`
`efficiency compressors with blade heights less than the given
`value. An intercooled engine is then set up, subject to the same
`constraints and is optimized for comparison with the conventional
`cycle.
`
`3.1.1 Conventional Engine. The optimal conventional turbo-
`fan is determined by carrying out a search in fan pressure ratio
`共FPR兲, bypass ratio 共BPR兲, and OPR, subject to the constraints
`imposed in Table 2. The minimum mission fuel burn was found to
`be 3633 kg. The optimal cycle parameters are found in Table 3.
`
`3.1.2 Intercooled Engine. As for the conventional engine,
`FPR, BPR, and OPR are key optimization variables. Due to the
`introduction of the intercooler into the engine, additional optimi-
`zation parameters are available for the designer. For example, the
`pressure ratio split between the intermediate pressure compressor
`共IPC兲 and HPC has a strong influence on the engine performance.
`
`Table 2 Engine design parameters and constraints
`
`Parameter
`
`␩FAN
`␩IPC
`␩HPC
`␩HPT
`␩IPT
`␩LPT
`hblade exit,HPC 共m兲
`Tblade,rotor 共K兲
`T30 共K兲
`T40 共K兲
`FNet,TO 共lbf兲
`
`Value
`
`91.0
`91.0
`91.5
`90.0
`91.0
`91.5
`0.012
`1196
`900
`1850
`27,300
`
`Table 3 Conventional engine parameters: take-off „TO…, top-
`of-climb „TOC…, and mid-cruise „MC…
`Parameter
`TO
`
`TOC
`
`MC
`
`冋共1 + ␶2兲冉 ␷o
`
`− 1冊 + f
`
`⌬P =
`
`SST k− ␻ turbulence model was used in the computation since it
`is generally believed that it can predict the separation better than
`the k⑀model. The resulting friction factor f and Colburn J-factor
`StPr2/3 are given in Table 1. The friction factor f is calculated
`from the equation
`共␳Vmax兲2␷i
`␷m
`␷i
`␷i
`2
`c
`where ␷i and ␷o are the fluid inlet and outlet specific volumes, ␷m
`is the average of the above two quantities, ␶is the ratio between
`the minimum flow cross-sectional area Ac and the heat exchanger
`frontal area, and A is the total heat transfer area.
`
`共1兲
`
`册
`
`A A
`
`3 Simulation Results and Discussion
`Details about engine performance modeling have been pub-
`lished in previous work 关4兴. Preliminary design data for the inter-
`cooled and conventional turbofan engines will be given below.
`The aircraft model has been described in Ref. 关5兴. Although the
`aircraft design code is general and may be used to model any
`subsonic aircraft of conventional type, the particular Boeing 737-
`800 model definition analyzed in Ref. 关5兴 is re-used here. The only
`difference being that the cruise phase length is shortened so that
`the total mission length is now 925 km.
`
`3.1 Optimal Engines. A baseline engine for a medium range
`commercial transport application is established and optimized for
`the given mission. The engine design parameters are set up to
`reflect year 2013 technology. The component efficiencies and op-
`timization constraints are given in Table 2. The blade material
`temperature has been selected to be representative for a fourth
`generation single crystal alloy. The constraint defining the com-
`pressor exit blade height is imposed to limit the deteriorating ef-
`fect of high pressure compressor tip leakage. Thermal expansion
`during engine transients makes it very difficult to design high
`
`Table 1 Friction factor and J factor for elliptic tube banks
`
`Re
`
`3000.0
`16,000.0
`24,000.0
`28,000.0
`32,500.0
`37,500.0
`42,500.0
`53,500.0
`
`StPr2/3
`
`0.00300
`0.00229
`0.00209
`0.00205
`0.00200
`0.00194
`0.00190
`0.00185
`
`f
`
`0.0120
`0.0079
`0.0064
`0.0061
`0.0060
`0.0059
`0.0058
`0.0056
`
`OPR
`BPR
`FPR
`␲IPC
`␲HPC
`G2 共kg/s兲
`T30 共K兲
`T40 共K兲
`SFC 共mg/N s兲
`
`37.25
`11.83
`1.463
`5.000
`5.091
`460.8
`900.0
`1850.0
`7.444
`
`47.01
`11.50
`1.575
`5.421
`5.507
`212.4
`798.3
`1641.1
`14.72
`
`43.32
`11.97
`1.529
`5.231
`5.416
`206.6
`779.0
`1580.4
`14.58
`
`114503-2 / Vol. 132, NOVEMBER 2010
`
`Transactions of the ASME
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`UTC-2022.002
`
`

`

`Table 4 Intercooled engine parameters: take-off „TO…, top-of-
`climb „TOC…, and mid-cruise „MC…
`Parameter
`TO
`
`TOC
`
`MC
`
`Table 5 Comparison of key module weight breakdowns: na-
`celle „Na…; thrust reverser „TR…
`Parameter
`Conventional
`
`Intercooled
`
`OPR
`BPR
`FPR
`␲IPC
`␲HPC
`G2 共kg/s兲
`T30 共K兲
`T40 共K兲
`SFC 共mg/N s兲
`
`57.15
`12.17
`1.478
`3.301
`11.71
`457.7
`900.0
`1850.0
`7.191
`
`71.99
`11.69
`1.591
`3.478
`13.01
`210.3
`796.2
`1631.5
`14.23
`
`66.48
`12.15
`1.544
`3.389
`12.70
`204.65
`776.5
`1573.0
`14.11
`
`FAN
`IPC
`HPC
`HPT
`IPT
`LPT
`Na+TR
`IC
`Total
`
`841.56
`89.86
`30.19
`62.59
`19.20
`412.23
`906.44
`-
`2932.9
`
`841.58
`72.15
`38.42
`75.51
`13.85
`358.47
`909.62
`114.76
`2969.3
`
`Furthermore, the design parameters of the intercooler can also be
`tuned to reduce the engine fuel consumption. As was done for the
`conventional engine optimization, a multidimensional search is
`carried out to minimize the fuel burn for the given mission. The
`optimal engine and intercooler parameters determined here are
`summarized in Table 4.
`The number of intercooler tubes were determined to be 5621,
`and the intercooler weight was estimated at 114 kg. Temperature
`change on the hot side was 53 K, and the inside pressure loss was
`3.35%. The external pressure loss was determined to be around
`5.3% before mixing. The optimal intercooler bypass ratio ⌽ was
`established to be around 17.0.
`The mission fuel burn for the intercooled engine was deter-
`mined to 3522 kg corresponding to a reduction of 3.1% compared
`with the conventional engine. Since the aircraft is considered fixed
`the carousel effect will here be quite limited. As described in Ref.
`关1兴, repeated mission analysis may be undertaken to take into
`account that the reduced fuel will allow a reduced take-off weight,
`which, in turn, will save additional fuel. For the relatively short
`mission studied here, this effect will increase the saving to 3.4%,
`whereas for the aircraft design range mission, this will correspond
`to a reduction in fuel burn equal to 4.0%.
`
`3.1.3 Gas Paths—Optimal Engines. The gas paths of the op-
`timal intercooled engine and the optimal conventional engine are
`displayed in Fig. 1. Since the optimal IPC pressure ratio for the
`intercooled engine is quite low, an HPC pressure ratio of 13.0 is
`needed in top-of–climb 共TOC兲. This would require a considerable
`aerodynamic load to be taken on a single stage turbine. In con-
`junction with having to simultaneously meet the HPC exit blade
`height constraint, a single stage HPT was not found feasible. A
`reduced OPR engine could have been designed with a single stage
`HPT, but this would have incurred a fuel burn penalty, even when
`taking the lower weight of the single stage core into account.
`One of the key features of the intercooled engine design dis-
`played in Fig. 1 is the gas path definition. Radially tilted combus-
`tors are typical for single stage HPT designs, allowing high tur-
`bine blade speeds and still limiting the rotational speed. Simplified
`turbine stress and aerodynamic analysis can be used to show 共see,
`for instance, Ref. 关6兴兲, that
`␴blade root ⬀ N2
`
`共2兲
`
`共3兲
`⌬h ⬀ U2
`where ␴blade root is the centrifugal stress in the turbine blade root
`and ⌬h is the enthalpy drop over a turbine stage. From these
`relations, it is clear that a radially tilted combustor may allow an
`increase in turbine power without compromising centrifugal root
`stresses. The other element that has to be considered to establish
`the suggested engine gas path definition is to realize that it is
`necessary to locate the HPC entrance as close to the low pressure
`shaft as possible. This will maximize the blade height for a given
`area.
`
`3.1.4 Weight Assessment. The weights of the key modules of
`the two engines described above are summarized in Table 5. The
`total weight of the two powerplants is quite similar but a consid-
`erable variation between the different modules can be noted. The
`top-of-climb pressure ratio distribution of the IPC and HPC of the
`conventional engine is 5.4 and 5.5, respectively, whereas for the
`intercooled engine, the corresponding numbers are 3.5 and 13.0.
`This leads to a relatively larger proportion of weight distributed
`on the HPC for the intercooled engine. The intercooler also re-
`duces the power requirement of the HPC, which, in turn, reduces
`the weight penalty that normally would arise from the consider-
`able pressure ratio of the HPC. The major weight benefits of the
`intercooled engine stem from the fact that the overall pressure
`ratio is much higher, making the turbines more compact. This is
`illustrated through the comparative cross-sectional drawing, i.e.,
`Fig. 1. This has the effect that the two stage HPT of the inter-
`cooled engine is only slightly heavier than the one stage HPT of
`the conventional engine. The weight benefits of compression be-
`come less marked for the intermediate pressure turbine 共IPT兲, and
`the effect is further reduced for the low pressure turbine 共LPT兲.
`3.2 Parametric Studies and Analysis of the Search Space.
`To analyze the optimal solution obtained, a number of parametric
`variations have been carried out. The FPR/BPR search space is
`illustrated in Fig. 2. The circular marker represent the optimal
`SFC point and the white curves represent iso curves for SFC. The
`SFC optimum is, as expected, located at higher BPR and lower
`FPR than the point of optimal fuel burn. The black lines represent
`the weight of a single engine, including both nacelle and thrust
`reverser. The somewhat noisy region running across the search
`space, indicated by the dashed line, originates from a jump in the
`number of stages in the LPT 共eight stages above the line and
`seven stages below兲. The noise arises due to truncation error in the
`rapidly changing region of the stage jump. The colors and the
`color bar indicate mission fuel burn. The location of the optimum,
`which is the square marker, is thus the design point, where a seven
`stage LPT is sufficient and an optimal combination of FPR and
`BPR is selected.
`
`4 Conclusions
`A fuel burn benefit of 3.4% was observed from introducing an
`intercooler into a turbofan engine. A large part of the fuel saving
`had been lost by meeting the HPC exit blade height restriction if
`the engine gas path would not have been re-optimized. The CFD
`analysis has increased the confidence in the pressure loss and heat
`transfer models used to predict the intercooler performance, as
`well as in its aerodynamic design. Engine weight for the conven-
`tional and the intercooled engines were predicted to be quite simi-
`lar, despite the considerable differences in OPR. Module weight
`breakdown support the idea that weight penalty added by the in-
`creased OPR and the added intercooler is to a large extent can-
`celled by the increased power density of the intercooler cycle. The
`intercooler concept stands out as one of the more promising ways
`of increasing the efficiency of the turbofan engine, and is expected
`
`Journal of Engineering for Gas Turbines and Power
`
`NOVEMBER 2010, Vol. 132 / 114503-3
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`UTC-2022.003
`
`

`

`Fig. 2 Fuel burn variation in FPR/BPR search space for intercooled engine.
`OPR optimal. Black lines are iso curves for engine weight and white lines
`are for SFC.
`
`to be a leading candidate among the more radical concepts being
`considered to provide a way forward toward greener air transport.
`
`Acknowledgment
`This study has been performed under the European Commis-
`sion Contract No. AIP5-CT-2006-030876. The authors gratefully
`acknowledge this funding. The authors also acknowledge the
`valuable discussions provided by Anders Lundbladh at Volvo
`Aero. We also acknowledge the general discussions on intercool-
`ing provided by Andrew Rolt at Rolls-Royce.
`
`References
`关1兴 Xu, L., Gustafsson, B., and Grönstedt, T., 2007, “Mission Optimization of an
`Intercooled Engine,” Paper No. ISABE-1157.
`关2兴 Kays, W. M., and London, A. L., 1964, Compact Heat Exchanger, 2nd ed.,
`McGraw-Hill, New York.
`关3兴 Xu, L., 2008, “Innovative Core Concepts for Future Aero Engines,”Licentiate
`thesis, Chalmers University of Technology, Gothenburg, Sweden.
`关4兴 Grönstedt, T., 2001, “Development of Methods for Analysis and Optimization
`of Complex Jet Engine Systems,” Ph.D. thesis, Division of Thermo and Fluid
`Dynamics, Chalmers University of Technology, Gothenburg, Sweden.
`关5兴 Avellan, R., and Grönstedt, T., 2007, “Preliminary Design of Subsonic Trans-
`port Aircraft/Engines,” Paper No. ISABE-1195.
`关6兴 Saravanamuttoo, H. I. H., Rogers, G. F. C., Cohen, G. F. C., and Straznicky, P.
`V., 2008, Gas Turbine Theory, Prentice-Hall, Englewood Cliffs, NJ.
`
`114503-4 / Vol. 132, NOVEMBER 2010
`
`Transactions of the ASME
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`UTC-2022.004
`
`