Professional Documents
Culture Documents
I Facoltà di Ingegneria
Corso di Laurea Specialistica in Ingegneria Aerospaziale
Relatori:
Professore Sergio Chiesa
Professore Rodrigo Martínez-Val
Candidato
Giovanni Medici
Luglio 2009
alla mia famiglia
Sommario
N
elle pagine che seguono è riportato un breve riassunto in italiano del lavoro
effettuato e l’indice tradotto. Il lavoro di tesi è stato svolto presso l’ Univer-
sidad Politécnica de Madrid, sotto la supervisione del Professor Rodrigo
Martínez-Val e con relatore il Professore Sergio Chiesa del Politecnico di Torino. Lo
studio elaborato in questa tesi ha come obiettivo la verifica delle possibili configurazio-
ni dell’impianto propulsivo adottabili in un aeroplano di tipo ala volante. Particolare
interesse desta la configurazione semi sommersa, con un condotto ad S dalla geometria
frontale a semicerchio. Lo studio è stato condotto a partire da un progetto sviluppato
presso la E.T.S.I.A. (Escuela Técnica Superior de Ingenieros Aeronáuticos de Madrid)
da Rodrigo Martínez–Val, Professore ordinario di Cálculo de Aviones presso detto Poli-
tecnico. Il progetto riguarda un aeroplano di tipo ala volante, di medie dimensioni, adi-
bito a trasporto passeggeri. Pur trattandosi di un progetto in stato preliminare, alcuni
studi dettagliati erano già stati condotti, in particolare: struttura principale, bilanciamen-
to di massima, aerodinamica linearizzata, stabilità statica e dinamica del velivolo, studio
dei vortici di scia, studio dell’inquinamento sonoro a terra ed in volo, caratteristiche di
evacuazione, dislocamento dei servizi a terra.
L’inconsueta configurazione dell’aeroplano di tipo ala volante ha richiesto una anali-
si delle caratteristiche peculiari di tale architettura, paragonata a quella più tradizionale
con superfici alari, fusoliera e impennaggio verticale ed orizzontale.
Giovanni Medici ~ Sommario
A tale studio semplificato viene affiancato uno studio di dettaglio del profilo singolo.
Tale analisi bidimensionale è effettuata attraverso il preprocessor Gambit ed il processor
Fluent. Tali softwares richiedono un’accurata messa a punto e la definizione della mesh
prevede l’uso di teorie approssimate per valutare la distanza della prima cella. Grazie al
software sviluppato, la geometria viene elaborata a partire dai punti noti attraverso le
coordinate scritte nei file di testo, attraverso le splines matematiche proprie di Gambit.
All’utente viene dunque richiesto solo di verificare la correttezza della mesh. Una
volta salvato il file utilizzando la terza parte del programma si effettua il setup del pro-
cessor, Fluent, procedura che richiederebbe circa 150 operazioni, ma all’utente vengono
richiesti solo tre dati: altitudine, velocità ed angolo di attacco.
L’ultima parte del programma si prende in carico di scrivere una macro che, letta dal
software CAD CATIA, aggiorna la geometria del modello tridimensionale in modo da
permettere all’utente di effettuare degli studi di compatibilità ed integrazione con il pro-
getto precedente: struttura primaria, carrello principale, volumi ed ingombri in cabina.
Si è sviluppato inoltre un ulteriore programma, in modo tale da permettere all’utente
di modificare ed adattare ai requisiti di progetto i profili rintracciabili in letteratura. An-
che questo programma è stato creato nell’ottica di fornire all’utente un software sempli-
ce da usare e con una interfaccia grafica che permetta di verificare passo passo le modi-
fiche effettuate. Tale programma, Airfoil Editor, permette automaticamente di leggere
diverse tipologie di files di testo contenenti coordinate dei profili (bidimensionali, X Y,
tridimensionali X Y Z, e bidimensionali X Y estradosso Y intradosso). Una volta carica-
to il file permette poi di modificare lo spessore massimo, ruotare il profilo in modo da
centrarlo sugli estremi 0 ed 1 rispetto alla corda, ed aggiungere una linea media incurva-
ta di tipo reflex (caratteristica necessaria al fine di abbassare il valore del coefficiente di
momento). In uscita da tale programma è possibile ottenere il profilo salvato sotto for-
ma di coordinate in un file di testo, oppure un file “.pdf” che racchiuda le informazioni
dettagliate di profilo, linea media e spessore lungo la corda. È inoltre possibile stampare
il risultato oppure salvare un grafico dettagliato racchiudente tutte le operazioni effet-
tuate dall’utente a partire dall’apertura del file originale.
Nel primo caso si tratta di una soluzione geometrica che minimizza le perdite dovute
alla resistenza, deviando il flusso lateralmente. Una presa dinamica invece prevede di
inghiottire il flusso instabile, ma mantenendolo al di fuori del contatto con la palettatura
del motore.
Si noti infine che, pur escludendo lo strato limite dalla presa dinamica del motore, il
particolare condotto ad S, con area frontale quasi costante, ma geometria che passa da
semi–circolare a circolare ed con uno sbalzo di due metri, mette già a dura prova la sta-
bilità del flusso.
Della presa d’aria dedicata allo strato limite si sono calcolate le caratteristiche prin-
cipali, come portata d’aria pressione, temperatura e densità medie. Si è rilevato che la
massa d’aria a disposizione è considerevole, per tale motivo è stato effettuato uno studio
con il fine di valutare la possibilità di utilizzarne parte di essa per alimentare il sistema
di condizionamento dell’aeroplano, ed eventualmente raffreddare lo scambiatore di ca-
lore.
Vari testi disponibili in letteratura si sono utilizzati per determinare le potenze in gio-
co ed i valori tipici delle varie architetture dei sistemi di condizionamento. Il modello è
stato messo a punto rispettando i limiti segnati dalle principali normative. Grazie ai cal-
coli effettuati è apparso chiaro come la tecnologia proposta sia molto promettente se
applicata al progetto dell’aeroplano di tipo ala volante, in quanto ci sono altre potenziali
applicazioni dell’aria inghiottita (ad esempio il sistema antighiaccio delle gondole e pale
del fan motore).
L’ultima analisi effettuata è stata di carattere geometrico, in particolare si sono valu-
tate la compatibilità tra il condotto ad S progettato ed alcuni aspetti del precedente pro-
getto. In particolare si è verificata, creando un modello CATIA dettagliato, la assenza di
intersezioni tra la geometria del condotto ed il longherone posteriore, le centine ed il
carrello principale.
I risultati delle analisi sono stati presentati in alcune tavole tecniche, affiancati da
altri studi, come la verifica di clearance laterale in atterraggio e verifica dello spazio in
cabina.
Al termine dello studio si sono riscontrati dei risultati promettenti ed in linea con
molti testi presenti in letteratura riguardo alla possibilità di adottare una configurazione
semi sommersa con condotto ad S per il sistema propulsivo. Alcune scelte obbligate,
come la deviazione dello strato limite, permettono di adottare alcune architetture inno-
vative molto promettenti.
Lo studio condotto fino ad ora ha un carattere molto condizionato da alcune ipotesi
semplificative che si sono rese necessarie, prima tra tutte la natura bidimensionale del
dominio studiato. Al fine di verificare le effettive possibilità della configurazione semi
sommersa sarebbe necessario condurre uno studio tridimensionale della geometria. Par-
ticolare attenzione dovrebbe essere posta sugli effetti sulla resistenza totale dell’aero-
plano se munito di deviatori dello strato limite con delle prese statiche a tendina per for-
nire la massa d’aria al sistema di condizionamento di cabina.
Indice
Indice delle figure ...................................................................................................................................VII
Introduzione ................................................................................................................................................1
Legenda.......................................................................................................................................................5
Preprocessing ......................................................................................................................................47
Condizioni al contorno........................................................................................................................50
Processing ...........................................................................................................................................51
Risultati ...............................................................................................................................................53
Dettagli................................................................................................................................................60
Gambit ...........................................................................................................................................75
Fluent ............................................................................................................................................79
CATIA ...........................................................................................................................................89
Descrizione..........................................................................................................................................94
TsAGI B Series..............................................................................................................................95
SC 165 r 100..................................................................................................................................98
NASA R 75....................................................................................................................................99
N00165 R 50 .................................................................................................................................99
SC 165 R 50 ................................................................................................................................100
Crociera .......................................................................................................................................103
Manovra ......................................................................................................................................106
Decollo ........................................................................................................................................108
Mesh..................................................................................................................................................127
Condizioni al contorno......................................................................................................................129
Risultati .............................................................................................................................................133
Risultati .............................................................................................................................................136
Programma Matlab™........................................................................................................................151
Matlab™ ...........................................................................................................................................168
Tavole ................................................................................................................................................216
Bibliografia .............................................................................................................................................247
T
he main goal of this thesis is the analysis of the integration of the propulsive
system in a Flying Wing, using a semi sinked configuration. The Flying Wing
FW-300 project has been developed in the E.T.S.I.A. (Escuela Técnica Supe-
rior de Ingenieros Aerónauticos de Madrid). During the study compressibility effects
were taken in count, so that an extensive research on a new profile for the central sec-
tion was needed. In order to perform an analysis as standardized as possible, a suite was
developed using Matlab™: CADCFD Integrator. This suite allows the user to perform a
complete analysis of the linearized aerodynamic characteristics. Another key feature of
the suite is that integrates different softwares: VLM Tornado (linearized aerodynamics
Vortex Lattice Method), Gambit CFD preprocessor, Fluent CFD processor and CATIA
CAD software. Thirteen airfoils have been studied and compared. Once the new airfoil
has been found an analysis of the boundary layer thickness and intake geometries has
been made. A CATIA model was also made in order to check for the integration between
primary structure and main gear.
Giovanni Medici ~ Engine Integration on a Flying Wing
Index
Introduction.................................................................................................................................................1
Nomenclature ..............................................................................................................................................5
Engine..................................................................................................................................................30
Airfoils Analysis..................................................................................................................................44
Preprocessing ......................................................................................................................................47
Processing............................................................................................................................................51
Results .................................................................................................................................................53
Main Purpose.......................................................................................................................................58
Details..................................................................................................................................................60
Airfoil Editor..................................................................................................................................62
Gambit ...........................................................................................................................................75
Fluent .............................................................................................................................................79
CATIA............................................................................................................................................89
Main Purpose.......................................................................................................................................94
Description ..........................................................................................................................................94
SC 165 R 200.................................................................................................................................97
NASA R 75 ....................................................................................................................................99
N00165 R 50..................................................................................................................................99
SC 165 sym..................................................................................................................................100
SC 165 R 50.................................................................................................................................100
Comparison .......................................................................................................................................102
Cruise ...........................................................................................................................................103
Maneuver .....................................................................................................................................106
Takeoff..........................................................................................................................................108
Summary......................................................................................................................................110
S–Duct Geometry....................................................................................................................................113
Main Purpose.....................................................................................................................................114
Boundary Layer.................................................................................................................................114
Mesh ..................................................................................................................................................127
Results ...............................................................................................................................................133
Main Purpose.....................................................................................................................................136
Results ...............................................................................................................................................136
Compatibility ....................................................................................................................................156
Matlab™............................................................................................................................................168
Drawings ...........................................................................................................................................216
References ..............................................................................................................................................247
Index of Figures
Figure 1.4 – Comparison between usual configuration, BWB and FW. ..................................................13
Figure 1.6 – Comparison between A320, BWB and FW fuselage structures. .........................................16
Figure 1.7 – Architectural solutions of the FW: double skin vaulted shell and flat sandwich shell. .......16
Figure 2.3 – Different Flying Wings projects (VELA 1 and VELA 2 and Tu 404)..................................32
Figure 2.5 – Comparison of wet surfaces between two different engine configurations .........................34
Figure 2.6 – Active Flow Control and Vortex Generating vanes as in [2]................................................36
Figure 3.9 – Progression of shock waves with increasing Mach number. ...............................................55
Figure 4.2 – Phase A communication between software after CAD CFD Integrator ...............................59
Figure 4.6 – Comparison between fourth degree polynomial equation and reflex camber......................64
Figure 4.12 – Example of different operations on a supercritical airfoil and “.pdf” printed version .......68
Figure 4.20 – Help disclaimer and load geometry for the linearized aerodynamics button .....................74
Figure 4.39 – 3D View of the entire CAD model of the Full Wing .........................................................92
Figure 5.17 – for the takeoff, maneuver and cruise case ........................................................................110
Figure 5.19 – Comparison in the Maneuver case between the original and new profiles. .....................112
Figure 6.1 – Different shapes proposed by [13], image originally of [13]. ............................................115
Figure 6.10 – Different file formats, before and after the conversion. ...................................................124
Figure 6.11 – Gambit screen after using CAD CFD Integrator. .............................................................125
Figure 6.17 – Static Temperature and Mach Number contours of the S–Duct. .....................................134
Figure 7.8 – ECS overall system weight penalties comparison in N [39]. .............................................149
Figure 7.11 – Variable heat coefficient and heat coefficients ratio. .......................................................153
Figure I.1 – Sub calls between the routines of CADCFD Integrator .....................................................169
Figure II.1 – 2D Geometry representing the S–Duct added to the RAE airfoil. ....................................216
Figure II.2 – Three view of the CATIA model, fundamental dimensions and angles highlighted . .......218
Figure II.3 – Primary structure, engines, S–Duct and main gear of the model . ....................................220
Figure II.4 – Primary structure, spars and ribs of the CATIA model. ....................................................222
Figure II.6 – Main gear lateral clearance and main gear detail ..............................................................226
Index of Tables
Table 1.1 – Estimation of rolling and vertical accelerations on outermost seated passengers .................20
Introduction
T
his thesis relates the work developed in order to analyze and verify the inte-
gration of a peculiar engine configuration on a Flying Wing project. The FW-
300 is a medium size transport Flying Wing, designed by Professor Rodrigo
Martínez Vals in Madrid. The project is now in an alpha phase, so that the principal
structure has been designed, so was the aerodynamics and other key elements. The inte-
gration in this work of a semi sinked turbofan engine was the main concern of this the-
sis, so that the fluid dynamics around the center of the aircraft and inside the intake
could be investigated.
During the development of this work many papers have been studied in particular on
the Boundary Layer Ingestion BLI and Active Flow Control AFC technologies. When
analyzing deeply the original geometry, since compressibility effects were taken in
count, a mayor drawback on the central airfoil was found. The study could not over-
come the problem, an extended airfoil research was needed.
Since the project was still in an alpha phase, a fast and reliable tool was needed to
perform the research as accurate as possible. In order to comply this design philosophy,
a complete tool, CAD CFD Integrator, was developed using Matlab™.
In order to help the user managing airfoil coordinates text files, another software,
Airfoil Editor, was built. Those tools allow the user to perform a complete study on the
effects of changing an airfoil in the central part of the Flying Wing, through the integra-
tion of different softwares:
Through the developed tool was then possible to perform an analysis of thirteen air-
foils (most of them modified through Airfoil Editor). At the end of the research two best
candidates were found. Using the CADCFD Integrator software some 36 analysis were
performed in order to compare the two airfoils throughout the entire flight envelope.
From the analysis the best candidate was chosen.
Once the new airfoil was found, the analysis with the CFD in cruise condition re-
vealed the thickness of the boundary layer, so that a study of the intakes geometry was
started.
Using data found in [13] [14] [15] a reversed engineered S–Duct shape was de-
signed. And applied to the airfoil of the Flying Wing. Using the method reported in [38]
the boundary conditions of the S–Duct were calculated and a 2D CFD simulation was
performed. Using some refinements of the CADCFD Integrator software multiple geo-
metries were developed. The best candidate was used to perform an extended study.
Using the results found was possible to make a preliminary design of the Environ-
mental Control System or ECS of the airplane, both in terms of air supply and of power
requirements. The compressor was also preliminary designed as well as the heat ex-
changer.
Finally a more detailed CAD CATIA model was made in order to verify (with the
help of the 3D Computer Graphics) compatibility between the original design (and
some of its systems) and the new one. An overall compatibility was found even if fur-
ther study might prove the nacelles drag and engine pylon must be designed.
Nomenclature
I
n this pages are reported the different symbols, acronyms and subscripts used
throughout this thesis. Each acronym or symbol is described through extent name
and the measurement unit. Each unit used is related to the International Standard
Units, if not the change is due to an exception.
N o m e n c l a- Description I.S.
ture
STP Step 3D file [−]
! "
Thrust Specific Fuel Consump- kg
TSFC
tion kN · s
TXT Text File [−]
VLM Vortex Lattice Method [−]
Re Reynolds Number [−]
y +
Non–Dimensional wall distance [−]
CL Lift coefficient [−]
CD Drag coefficient [−]
Cm Moment coefficient [−]
! "
kg
µ Dynamic Viscosity m·s
!m"
uτ Friction Velocity near the Wall s
τw Wall shear stress [Pa]
! "
kg
ṁ Flow rate s
! "
kg
ρ Density m3
δ Boundary Layer Thickness [m]
γ [−]
Isoentropical Exponent
η Efficiency [−]
T Temperature [K]
h Enthalpy [K]
p Pressure [Pa]
! "
J
s Entropy kg · K
! "
J
cp Specific heat kg · K
! 2"
A Area m
! "
W
k Heat transfer coefficient 2
m ·K
P Power [W]
!m"
u Flow velocity parallel to the wall s
! 2"
m
ν Local kinematic viscosity s
Usual Configuration
W
hen thinking about an airplane, each one can recall different shapes and
architectures. The shape of actual airplanes is the result of an optimiza-
tion that lasted almost five decades. Since the first turbojets where in-
troduced a new benchmark of performances was raised. The chance to fly faster and
higher dramatically modified the shape of the propeller era’s airplanes. The multi disci-
plinary character of airplane design lead to a widespread range of possible configura-
tions. As reported in [1], [2], [3], [4], the engineering design process is the tool to pro-
ject a feasible and efficient solution to the given requirements. As reported in [1] “As
any other science, aircraft design is a set of co–ordinates concepts (the knowledge) and
rules (the method)”. Since the concepts do not depend from the shape, virtually any
configuration could be adopted. The reason why in last decades just optimization (and
enhancement) has been made lies probably in the life–cycle of a complex system such
as an airplane.
Nowadays the design and development of an airplane is a huge effort. The require-
ment, such as high speed, high altitude, pressurization, light weight and though structure
are counteractive. Developing a product with such requirements would be a deal, but
airplanes design does not stop here. The growing attention to the environmental ques-
tion builds up a totally new set of constrains such as a lower thrust specific fuel con-
sumption (TSFC), lower CO2 , COx , N Ox emissions, higher reliability and more re-
cyclable materials, lower noise, lower wake turbulence and many more.
Another key feature of new designs is less parts and an optimized interchangeable
use of common parts inside a fleet or family of airplanes.
The great interest on the features just presented lies not only in a smarter use of the
fuel, whose contribute inside the direct operating cost (DOC) is growing due to the fuel
price, but also to design a product that could be sustainable.
Deriving from the same design philosophy is the disposal of the product. Since the
early steps of the design, a combined approach must provide solutions for each part of
the life–cycle of the airplane. The airplane industry during the last 5 decades has seen a
supremacy of two major enterprises: Airbus and Boeing. In order to comprehend the
complexity of designing an aircraft, the product life–cycle is presented in Fig. 1.1.
Customer service
40 – 50 + years
Marketing
30 – 40 + years
The life–cycle of the product starting with its definition and ending with the last cus-
tomer service offered might last some 50 – 60 years. In order to develop a completely
new product an enterprise must figure a great investment whose break even point usu-
ally came 20 years later than the first conceptual design. When considering that if a
company fails to build a successful project will probably bankrupt, appears evident that
the effort to start a totally new design involves huge risks.
By the way when looking to the development of new designs during the last century,
it’s stunning to see the incredible innovation of the first 50 years, when compared to the
stagnation (optimization) of the last 50. This is well described in Fig 1.2.
When thinking on the enormous innovation lead by the change from the propeller to
the turbojet, from the tube fuselage to the frames and ribs pressurized structures, that
occurred during the first fifty years (and almost without any computer), seems strange
that in the last years, with incredible new chance offered by the using of computational
fluid dynamics, finite elements and the brute power of computers, just refinements in
shape have been made.
50 years
50 years
What at first look one might not see is the continuous growth of the efficiency of air-
planes, the better TSFC and the fewer parts, weight and higher safety. This continuous
process is well described by the Fig 1.3. In the same figure the first key feature of the
Flying Wing configuration is presented. Indeed if the Flying Wing is enough capacious,
Pas · km
it can provide a breakthrough in terms of . In the last decades an asymptotic
kg
performance has been identified, the Flying Wing can give (in its first attempt, without
optimization) a benefit that can be compared with the one of the last thirty years.
50
40
30
Pas · km
kg
20
10
0
1950 1960 1970 1980 1990 2000
year
In order to understand the reason why no enterprise has yet developed this kind of
configuration a close look to the benefit and to the shortcomings of the two different
shapes is now presented.
Flying Wing:
Lower drag.
Higher efficiency.
Increased range.
Lower wake.
Less power required.
More cabin space.
Lower noise.
When engines are not podded under the wing, better ditching capabilities.
Lighter.
Pressurization is more difficult.
Might be difficult to equilibrate.
Due to cabin layout, flight might be difficult for the most external passengers.
Huge design/engineering effort.
In order to better comprehend the bullet list just presented, each entry will be shortly
described. The Flying Wing design presents lower drag, this feature derive from its
shape. The usual configuration of an aircraft has the wing whose function, while in air,
is to hang and keep the fuselage from falling. The fuselage is just a tube that does not
provide any lift, it weights a lot and thus provide the most relevant part of the drag. In a
Flying Wing configuration the fuselage does not exist, since the payload is stored inside
the wings. Lift is provided smoothly (some abrupt changes due to engines if podded),
throughout the span. In the central section, where in a usual configuration one could
find the fuselage, the Flying Wing present the maximum chord, and thus it might pro-
duce the highest amount of lift. The crossover version between a pure Flying Wing and
a common configuration is the so called Blended Wing Body or BWB. In the Fig. 1.4 is
presented a comparison between the different configurations.
The Blended Wing Body presented is a NASA project that is in study since many
years, but at the end of 2008 proved to be an interesting design through a scaled model
(1:7) that has taken the first flight.
A A
A A
L
LL A A
Internal layout features double deck seating with video monitors in the back of each seat. Shown here is the 529-seat
upper deck.
The lack of the fuselage gives the Flying Wing an higher efficiency since the wetted
surface is fewer. Another key feature of the Flying Wing is that it can be inherently
lighter, this mean that can fly longer routes and has an extended range (if calculated re-
spect the number of passengers). In Fig. 1.5 is presented a comparison found in [4]: C-
type FW (thick line), the V-layout with LFC (thin line) and U-layout with LFC and the
original MTOW (dashed line ).
25000
20000
PL [kg]
15000
10000
5000
0
0 5000 10000 15000 20000
R [km]
Another great feature of the Flying Wing is that its shape gives it the chance to have
a lower intensity turbulent wake. This is the result of different components: the most
evident is that since the airplane has fewer wetted surfaces, and thus a lower drag, it
provides less turbulence. The constant sweep angle wing gives the Flying Wing another
input to lower the turbulent wake, so does the C Wing profile. In a scenario of growing
revenue passenger kilometer (RPK), inevitably the most congested airspace will be the
one of the most busy airports. The most limiting value to raise or lower the aircraft
takeoff (and landing) per hour is the space between each aircraft. This time distance be-
tween an aircraft and the one that follows is linked to the aircrafts reciprocal dimen-
sions. Some studies confirms that the Flying Wings could provide in busy airports such
as Chicago an increment of the 40% of aircraft takeoff/landing per hour. More takeoff
and landings mean higher efficiency and lower airport rates. In the same study a com-
parison between a Boeing 777 and a Flying Wing noise is presented. The FW produces
some 10 dB less than the Boeing 777 (that lies around 80 dB). Lower noise means a
more interesting product, also if thinking on the market direction provided by a more
sustainable aircraft. When comparing the Flying Wing with a more conventional con-
figuration, it needs less power (in terms of ratio per passenger). This mean that intrinsi-
cally lowers emissions, and fuel consumption. The thick wing provide an huge amount
of cabin (and storing) space. Space has always been a precious quantity in aviation, with
this new configuration a whole new standard can be set. This room can provide opti-
mized position in term of center of gravity and mass balance. The conventional configu-
ration with podded engines, while providing an unbeatable maintainability has some
drawbacks. First of all, when landing without gear inevitably the engines will scratch
runway or ground, this could result in catastrophic fire or damage to the structure. In the
event of ditching things might even get worse. The frontal surface of turbofan engines is
big, the shape of the intake behaves like an anchor respect to the airplane, causing, in
event of non perfect horizontal attitude (or in case of waves), the airplane to spin and
thus to dissipate huge amount of energy within meters.
A Flying Wing is not suitable for the podded under wing engine configuration. This
characteristic gives the Flying Wing intrinsically better performance in case of emer-
gency landing or ditching.
Those are just few examples of the advantages that this kind of aircraft might give
when compared with a more usual architecture.
By the way the Flying Wing present some major drawbacks. Probably the most limit-
ing one is that since the cabin shape is contained inside the wing, the typical cylindrical
(single, double or triple bubble) approach can not be used.
Pressurization loads are probably the most severe an high altitude airplane reachs.
The cyclic nature of this load highly worst the mechanical characteristic of the structure,
and induce fatigue. The solution adopted in common airplanes is to reduce the pressur-
ized cabin to a cylinder–shaped cage structure (that actually breath as altitude increase).
As known the weakness of the cylinder configuration are the outer bounds. In this area
is placed usually a reinforced bulkhead. The cylindrical cage not only provide enough
strength to sustain the pressurization loads, but also gives the starting point to place the
outer panels, with a major reduction in structure weight. It appears clear that a similar
light structure can not be used in a Flying Wing, this is because even if the structure
might well sustain the pressurization loads, it can not sustain external panels. In order to
design a structure that keeps almost the same shape of the external wing, an heavy over
dimensioned structure must be used. Probably this feature is the reason why no aircraft
industry has ever tried to develop a transport plane in a Flying Wing configuration. In
the next section a few examples of existing Flying Wings will be reported. In the Fig.
1.6 a comparison between an Airbus A-320 and possible Blended Wing Body and Fly-
ing Wing configurations are presented.
There are two principal solutions for the internal pressurized structure of a Flying
Wing. The first is a flat sandwich shell (first presented in Fig. 1.6).
The second is a double skin vaulted shell. While the flat sandwich shell provide a
continuous structure (between the inner and the outer part of the fuselage), it seems to
be heavier.
As reported in [4] “In both cases the outer resisting shell is helped by large ribs, to
carry on concentrated shear forces, to maintain the external aerodynamic shape and to
contribute to torsion stiffness. How both concepts deal with cabin arrangement is shown
in figure below.
Figure 1.7 – Architectural solutions of the FW: double skin vaulted shell and flat sandwich shell.
From the point of view of passengers no major difference exists for headroom and
passages are optimized to become almost equal. Structurally , however, things are quite
apart”. The two different structural architectures are presented in Fig. 1.7.
The flat sandwich shell architecture leads both for pressurization and for torque and
aerodynamic loads to the use of quasi planar or planar panels. This kind of structure has
proved to be inefficient when dealing with internal pressure, this mean that a thickened
structure can not be avoided.
On the other hand the double skin vaulted shell is a more usual quasi cylindrical
shape, that takes count of the pressure loads, while the external skin care of the torque
and bending aerodynamics moments.
This second kind of structure has been proved to resist more to the loads, it appears
lighter and above all has a nice fatigue behavior (it prevents fatigue crack propagation),
due to its load diffusion capability. Some authors (Bradley 2004 , Mukhopadhyay et al
1966 – 2002) reported the double skin vaulted shell structure as the fail–safe.
This point is matter of discussion, since as reported by other authors [2], (Liebeck
2004) argue that if the outer skin has to resist also to a pressure load should be well over
dimensioned, thus the weight gap between the two different architectures would close.
Other authors in spite of using a flat sandwich shell suggest the use of usual fuselage
assembled together in order to obtain a tough, well experimented shape.
One of the greatest deals when designing a new airplane is linked with the amount of
new technology that the project requires. In this case starting from the fuselage, going
through engines and structure in general, the amount of new technology, and thus of
know–how required, is enormous.
Since the fuselage layout is so different from the usual architecture, even if respect-
ing the current normative (FAR-25 and the new CS-25), a study of emergency safe
evacuation of the cabin should be performed.One of the most demanding features that
must be certified for every new airplane, or for any of its derivative, is that the airplane
can be evacuated in less than 90 seconds.
This evacuation requirement was first introduced in 1965 with a value of 120 sec-
onds, then through crew training and a renewed seat and aisle designs, has been re-
duced. Even if the simulation does not represent a real accident (where probably some
passengers might loose consciousness) it sets a benchmark in evacuation capabilities of
both crew and airplane. This test has been introduced since in many of aircraft acci-
dents, the impact with the ground is survivable, then the major threat for the life on
board is propagation of smoke and fire inside the cabin. A well trained crew, and a
smartly designed airplane can provide an easy to evacuate scenario. As reported in [4] a
comparative study, (both with a Boeing 777 and a L-1011 Tristar) has been made. While
the test on a real airplane check the effective evacuation capability of an airplane de-
sign, it does sets some limitation. First of all the airplane must be designed and built. So
that an evacuation on a conceptual design can not be proved. Another drawback is that
during this simulation some of the participants might suffer of light or even moderate
injuries. Even if those limitation are present very few models of evacuations have been
developed. Two of them have been studied at the Cranfield and Greenwich University.
Those model adopts a geometrical perspective, and are related and similar in results
to a network type model (Watts 1987). Even if this perspective gives some limitation, in
a preliminary study it can give interesting results that do not depend on some very sub-
jective factors such as cabin ambiance, demographic and psychological features of pas-
sengers, crew training and personality. The algorithm is a seat–to–exit assignment that is
very versatile.
It can be modified in order to take care of some rules (such as a preferred exit, or a
closed one). Results are presented in terms of total distance ran for each passenger. This
feature has been proved to be very adaptable. Since the model is a geometrical one, the
first requirement is to set up a detailed geometrical model. In Fig. 1.8 the cabin layout
of the Flying Wing is presented. This is the high–capacity configuration, thus the most
difficult to evacuate. In the figure are clearly visible the emergency exits (A type), the
seats, the galleys lavatory and wardrobe.
A A
A A
W W
L L
G G
G G
W W
L L
L
A A L
Result of the study presented in [4] report that with the FW unusual cabin configura-
tion, it is possible to achieve the same standard as the one reach from common configu-
ration such as the one of the Boeing 777 and of the L-1011 Tristar.
Another key feature of the cabin in a Flying Wing is that it is not concentrated
around the center of the airplane (almost coinciding with the roll axis of the plane). In
the design just proposed the cabin is some 22 m wide and thus the most external seats
are no more close to the roll axis (and close to the center of gravity of the airplane).
This condition might induce unprecedented values of acceleration and roll rate in
case of gust. Needless to say that an extent analysis should be carried out throughout the
whole flight envelope of the aircraft.
In order to better understand the orders of magnitude that play an important role in
this analysis, a simplified study has been carried out in [4]. in this study an estimation of
rolling and vertical accelerations on outermost seated passengers has been made both
for the low altitude, low speed (maneuver) and high altitude high speed (cruise) condi-
tion. In the Table 1.1 results are reported. The study models an asymmetric gust of 1 m/s
only on one side of the wing. In this way an unbalanced lift and drag is produced on the
airplane that reacts with a roll. This study wanted to prove how uncomfortable would be
for the outer passenger this ride.
Table 1.1 – Estimation of rolling and vertical accelerations on outermost seated passengers
In Fig 1.9 is reported a more detailed top view of the Flying Wing.
A A
A A
W W
L L
G G
G G
W W
L L
L
A A L
As the study proves, one flight on a Flying Wing would be a bit more bumpy (in par-
ticular for the most port or starboard passengers), but within common transport flight
experiences.
The Flying Wing is for its nature the most efficient compromise in both structural
and aerodynamics terms. The lack of fuselage does provide a minimum wet surface and
highly reduced control surfaces (such as horizontal and vertical stabilizers). The gain in
terms of weight and efficiency can be reduced (and almost neglected) due to stability
and control issues.
The Flying Wing needs meticulous analysis in order to achieve a stable (or almost
stable configuration). A key feature to look at is the variation of the momentum coeffi-
cient throughout the wingspan. A null momentum coefficient would be desirable, but
keeping this coefficient to values close to zero might be a nice compromise. If the mo-
ment coefficient is not null, an equilibrator angle (trim angle) should be designed. This
angle does give stability, but lowers the drag and efficiency performance of the whole
airplane.
Even if it is true that no Flying Wing transport passenger aircraft has been ever built,
the history of airplanes does provide many examples of this configurations both in mili-
tary, tourism and unmanned roles.
This airplane was a double chevron wing tail less plane and its unconventional de-
sign the claimed attention of a Royal Engineer test team. During a test it was damaged
and then rebuilt as Dunne.4.
The evolution of the studies carried out by Dunne gives probably as result the Dun-
ne–Burges, in Fig. 1.11 there is a reproduction of the airplane.
The Dunne Burges airplane was a “joint venture” between Dunne (the English
Mathematican) and the boat designer Burges. The skill and experience of Burges allow
the airplane to be modified in order to acquire a pontoon. This airplane, built in 1915
was a seaplane.
Some fifteen years later a new project was developed. This project was named Arup
2 and the growing interest for a low budget, low maintenance, easy to fly airplane where
the leading features for this design. In 1932 at the prelude of World War II one of the
easiest kit was the Arup 2. This airplane, even if presenting a vertical stabilizer, and a
fuselage clearly visible, presented a quasi elliptic wing that runs throughout the fuse-
lage. The figures that are presented next derive directly from the airplane kit user guide
of 1932.
With the fast involution caused by the proximity of the war, an huge interest in avia-
tion was proven, in Germany two engineers, Reimar and Walter Horten, and later on
their industry, the Gothaer Waggonfabrik, was particularly specialized with the flying
wings. Nowadays the Horten are one of the most famous examples of flying wings, and
their smooth and elegant shapes gave a considerable help to the development of this
kind of aircraft.
In Fig. 1.13 the Horten Ho I is presented, this design is of 1933.
During World War II the Horten brothers were very active, and provided some of the
most elegnat aircraft ever. The great interest on aircraft development led the Horten to
build some of the most interesting airplanes, the Horten III was the first attempt (1938),
this airplane soared 7000 m in one of its flight, but its peripheral visibility (a tight angle
left and right, and almost horizon in frontal view), gave this plane few chances in a war
scenario. The Horten V was then, in 1942, built. This aircraft was a two seater bomber.
This airplane present incredible similarity with the Northrop. In Fig. 1.14 the Horten V
is presented.
The new turbojet engine and pulse jet engine that have been developed in the last
years of war gave the chance to build one of the most incredible airplanes of its time,
unfortunately the Go 229 aircraft was never operative, but it was almost ready to be
mass developed.
This fighter aircraft provides a futuristic approach, its turbojets and light weight give
it high speed and maneuverability. At the end of the WWII german aircraft engineers
were one of the most trained and advanced, some of them “migrated” to United States,
while others to Russia. In the Fig. 1.15 a schematic view of this flying wing is reported.
The World War gave the energy and power (in economical terms) to the aircraft in-
dustry, after that it left ruins, almost bankrupt countries and fifty five millions of deaths.
Soon after the war was ended another growing tension emerged. The Cold War was the
input of another competition between the two blocs, both in the aircraft and space fields.
Some of the most brilliant German engineers went on their studies in the United
States, the Northrop YRB-49A Flying Wing Bomber is an example of their work.
This aircraft was meant to be a medium bomber aircraft but the unreliability of the
engine and of the auxiliary power unit, united with the incapability of this airplane to
store the first nuclear bombs made this airplane almost unsuccessful. This study as it
will be told shortly gave Northrop the know how some decades later, to build one of the
most innovative bombers ever, the B-2.
Many other prototypes and project were developed but since the most common fea-
tures have been described we will now report briefly only two other examples.
In 1947 took place the first flight of the AW 52, Armstrong Whitworth flying wing.
This british experimental aircraft project lasted 7 years. It was meant to be a jet powered
(four to six of them) passenger aircraft. This prototype was a scaled glider, and its lami-
nar flow could not be preserved. One of the two prototypes was lost but pilot bailed out
successfully. The project was then abandoned since it presented poor flight characteris-
tics. In order to avoid the laminar flow to separate, the wing structure was very peculiar.
It was not a classic ribs, spar and skin, it is made from a shell and then joined together.
This approach was necessary in order to obtain a low roughness surface.
Probably the most famous Flying Wing nowadays is the Northrop B-2 Spirit heavy
bomber. This airplane through its shape is also a “stealth” or low observable aircraft.
This extraordinary airplane is a record beating plane not only in the stealth and bombing
fields, but also in the cost per airplane and cost per maintenance.
By far this was the most expensive airplane ever built, this lack is the direct conse-
quence of a reduced interest by the United States Army in this bomber that cut the initial
order of 132 to only 21 airplanes. The radar absorber external coating is very suscepti-
ble to temperature, this made necessary that each plane when not in flight should be
stored in an hangar with controlled temperatures. The airplane does show brilliant han-
dling and efficient characteristics so that it can both be used in high altitude and low
altitude, terrain following sorties.
The project itself lacks in maintainability and reliability concepts, and this is a factor
that can not be forgiven, even in the military sector.
In this brief summary a few examples of Flying Wing have been presented. The most
interesting feature is the low drag and high efficiency, while the most (almost unre-
solved) drawback is the pressurization problem. This is probably the biggest design is-
sue, when dealing with this kind of configuration.
When taking away the common configuration and thinking to a completely new one,
a wide range of chance opens up and many solutions can be proposed.
One of the most interesting one is the propulsive system. Since this configuration is
dramatically different from the other, also a time must be taken to consider all the
chances and drawbacks of the engine placement. The next chapter will deal with this
analysis.
Engine
T
here are many concerns when dealing with the propulsive system. In modern
airplanes the turbofans give a considerable amount of thrust, this force, ex-
pressed in tons, must be sustained by the structure, in order to pull the air-
plane rigidly. As exposed in the last chapter, aircraft design is an extraordinary multidis-
ciplinary engineering effort. Each system must be integrated in order to respect and
support other parts and disciplines.
Being one of the principal, the propulsive system obey at the just exposed design trend.
When dealing with this, system interaction between different fields must be taken in
count. For example in an engine layout design each of the following field must be
evaluated:
Propulsion
Structure
Maintenance
Weight
Weight Balance
Static and dynamic stability
Aerodynamics
Safety
Zonal Analysis
The exposed disciplines are such different that usually are counteracting. This feature
leads to an election that is always a compromise between the different configurations.
In usual configuration the low wing, reverse T tail and under wing pylons engines
has been proved to be an optimum choice. This position of the engines allow ground
inspection in the preflight checks, speed up maintenance and give a sound fluid dynam-
ics for the compressor. It does not result in any big loss in terms of balance (in case of
one engine failure), even if a particular care on the tail dimensioning should be posed.
Some other configurations are less used and takes the engines near the end of the fu-
selage, this kind of election does give lighter structure, and lower noise but pose some
problems in terms of fuel pumps and engine failure or fire.
In transport airplanes history there are also other solutions, the L 1011 Tristar was a
three engines aircraft, two of them were in the usual underwing configuration, while the
third was characterized by an S-Duct. It started from the vertical stabilizer went down to
the center of the fuselage. Few examples, in particular in the russian airplanes, of pylon
over the wing are present. Those cases were due to a design feature common of many
russian airplanes, that provided the capability to operate from rough surfaces or runway
in bad conditions, snow and dirt, so that engine dirt ingestion was to be avoided.
"
45."*67"8(9.0+,.":7,;+<=)(*+7,9"9576,"8.076"5(>."8..,"?.9+<,.?"*7".9*(80+95"*67"87=,?()+.9"7;"
The usual configuration has proven to be the best trade off between the different *5."?.9+<,"9/(:."+,"/0(:.@.,*"(,?"80.,?+,<"7;"*5."7=*87()?"6+,<"A@79*";7)6()?&"@79*").()6()?BC""
"
"
fields (exposed in the bullet list). But when dealing with a completely new design, each
previous calculated configuration must be analyzed again.
In the various medium size transport flying wings designed (almost each of them was
never built) there are many different engines configurations. This prove that this par-
ticular system is really flexible to the interest of the designer, and can be adjusted on his "
"
" DE$F"G"H(9.0+,."I7,;+<=)(*+7,C"
needs. 45."*67"8(9.0+,.":7,;+<=)(*+7,9"9576,"8.076"5(>."8..,"?.9+<,.?"*7".9*(80+95"*67"87=,?()+.9"7;"
*5."?.9+<,"9/(:."+,"/0(:.@.,*"(,?"80.,?+,<"7;"*5."7=*87()?"6+,<"A@79*";7)6()?&"@79*").()6()?BC""
"
"
" "
" "
" "
" "
DE$F"G"H(9.0+,."I7,;+<=)(*+7,C" DE$F"1"H(9.0+,."I7,;+<=)(*+7,C"
"
"
" !"#$%&"'()*+,"-.//.)0."1223"
"
Figure 2.3 – Different Flying Wings projects (VELA 1 and VELA 2 and Tu 404)
"
"
DE$F"1"H(9.0+,."I7,;+<=)(*+7,C"
Usually the engines in the Flying wing are designed to be near the aft spar, on the
wing’s extrados. This configuration can either be pylon mounted or ducted. In the
NASA Blended Wing Body for example, the first design propose the engines to be in-
stalled with the pylon–mounted architecture (Fig. 2.4).
289.0 ft
40.9 ft
160.8 ft
Usual turbofan engines can sustain random turbulence on the disk face, but as re-
ferred in [10] [11] [12] [13] [15] [16] can not withstands cyclic loads due to a disturbed
non axial symmetric flow. Since the fan is spinning with an high speed, a cyclic load on
the blades could result in bad fatigue characteristics. This is probably the reason that
explain the choice of the NASA engineers when developing the Blended Wing Body.
Why should one design the engines closer to the wing? This question has multiple
answers. Almost hiding the engines behind the thick wing means that the wet surface,
thus the resistance, lowers considerably. As can be seen in Fig. 2.5
289.0 ft
289.0 ft 40.9 ft
160.8 ft
Figure 2.5 – Comparison of wet surfaces between two different engine configurations
40.9 ft
Figure 2.6 – Active Flow Control and Vortex Generating vanes as in [2]
The holes, both of the AFC and of the BLI (but particularly of the Boundary Layer
Ingestion) needs to be cleared from and dirt or insect in each flight, above all insect
pose a severe threat to many instruments (like pitot probes) but the little holes provide
an ideal environment, hard to check and clear. Other studies propose some passive con-
trol devices, most of all Vortex Generating, VG vanes. This solution appears sound
enough but provide lower values of pressure recovery, since basically turns the whole
flow to turbulent. The gain in terms of drag is explained in Fig. 2.7, as reported in [11].
Momentum excess to
balance the momentum
deficit: engine thrust
u1
Podded uw
engines Momentum deficit
airframe drag
BLI
u!
The reduction of thrust moments with the BLI configuration can provide even lighter
structure, and among all an easier to manage and equilibrate aircraft.
A A
A A
W W
L L
G G
G G
W W
L L
L
A A L
The Flying Wing design resulted with those values. Starting with the values in the
Table 2.1 and using some of the figures in [1] [4] geometrical data was rebuilt.
Variable Value
Overall length 46 [m]
Overall width 77 [m]
Height above ground 16 [m]
! "
Wing area 893 m2
Wing span 75 [m]
Aspect ratio 6.3
Primary structure of the Flying Wing consist of two spars at 11% and 67% of the
chord, the 56% in between is used to store passenger, cargo and fuel. The torque box is
strengthened by the ribs, starting from the central symmetry axis the first rib is designed
to be at some 3.9 m. As reported in normative the minimum aisle dimension is some 0.5
m wide, while for the seats a 0.5 per column is a standard, this give a minimum of 3.5
width for the six seat columns configuration.
This value has been raised to 3.9 in order to allow a more comfortable aisle and to
leave some centimeters on the left and the right of the most outside column, in order to
avoid the “tunnel effect”. Cabin layout is described in Fig. 2.9.
0.1 m 0.6 m
1.55 m
3.9 m
There are four sector dedicated to the passengers (as presented in Fig. 2.10) and two
cargo bays. The standard medium size cargo container is some 2 – 2.5 m wide, so that
the third rib must be at least at 5 m. The particular configuration of a Flying Wing gives
the designer an unusual amount of internal space to store system and fuel. Thrust spe-
cific fuel consumption and weight are even lower in this aircraft, so that the amount of
required fuel is drastically reduced.
A A
W W
L L
G G
G G
W W
L L
A A
In last figure are described also the location of the emergency exits, as described in
normative, 2 type A exit must be designed in front of the aircraft, and two type A exit in
the aft part of it. Since the cabin is some 67% of the chord, the exit must run for the last
33% of the chord. In order to grant the passenger a wide enough aisle (1.1 m wide), a
mobile beak must be designed.
In the Fig. 2.11 a three view of the basic geometry and structure of the Flying Wing
is presented. As described before, the primary structure consist of two principal spars
and 7 ribs (other ribs are present but in this study only the ribs that contain the payload
are designed).
The Flying Wing is supposed to be twin engine, the position of the engines usually
should be as close as possible. This position gives stability to the balance, but has some
mayor drawbacks: if both engines share the same intake (or present intakes really close
to each other), in case of failure of one engine, maybe the second engine can present
failure too. Similar situation of risk is posed by an engine fire failure or engine explo-
sion. Another drawback is posed in terms of maintainability and zonal analysis. Engine
bays are usually one of the most dense and critical in terms of temperature, vibration
and noise. This severe environment reduces the overall reliability of the components
close to the engines.
A close to the center, but split configuration is desirable. Since in the first passenger
bay there is an emergency exit on the back, engine can not be designed to be in the first
bay. In the second, on the back there is the main gear bay, even if it pose some integra-
tion concerns, the main gear does not cross the turbofan engine, so that in terms of zonal
analysis it does not pose an high risk.
The study made by Stephane D’Urso (Università degli Studi di Napoli), was based
on the analysis of the static and dynamic stability characteristics of the Flying Wing, it
was a linearized study so that compressibility effects where not taken in count. In the
Fig. 2.12 there is the airfoil layout of the airplane. Three airfoils are used: NACA 64-
215 modified (0 – 50% wing span), NACA 63-018 modified (50% – 65% wing span)
and TsAgi Series II (65% – 100% wing span).
A A
A A
W W
L L
G G
G G
W W
L L
L
A A L
As referred by Stephane D’Urso the Flying Wing does prove to have noticeable static
and dynamic stability characteristics.
In order to verify which is the best configuration for the engines, the purpose of this
study is to analyze the semi sinked engines configuration.
Airfoils Analysis
T
he FW300 has three principal airfoils, that smoothly blend each other from
the root to the tip of the wing. Those airfoils are meant to reduce, and almost
elide, the moment coefficient. As is presented (in fig. 3.1) going from left to
right we have GWA-1, NACA 64-215, and a NACA 63-018 modified (we will call it
NACA 63-017). The Flying-Wing need to carry his payload inside the central part of the
wing, for this reason even if NACA 63-017 works well in transonic, is still a pretty thick
profile, and is almost symmetric.
%6.7(48.+41.9+)-(-(04.:+20;-
<'=
%"
)5,.90(4-
'()*)+,-./012.2!*00)1(4+-5
!" *>8>
&"
"
!&"
!!"
!%"
Those airfoils were chosen in a previous work and grant nice flight characteristic and
a noticeable static and dynamic stability (reference napoli).
We want to investigate which is the best solution for the engines placement both with
a fluid dynamic, structural, system and acoustic approach.
In accord with system and maintenance point of view, a central and split configura-
tion seems to offer a nice balance. In a pure engine fluid dynamic approach, the podded
configuration is still unbeatable, but if we consider the entire airplane (and its specific
configuration, the flying wing) a semi-sinked configuration seems the best. It is evident
the better aerodynamic, and the minor drag coefficient, but there are also other attrib-
utes. The semi-sinked configuration gives the chance of a pure thrust, without any para-
site pitching moment, in fact the thrust's vector passes through the center of gravity of
the plane. Another important point is that this configuration needs less structure, that
means less weight.
The first step of the analysis is then given by the interest of knowledge over the flow
around the engine's inlet area. Even if cruise's flight angle is modest (1.1895 degrees),
the cruise's velocity is high (Mach 0.8). This, coupled with the big chord, give rise to a
boundary layer that can not be neglected.
We need to investigate the central zone of the wing. Since we are close to the plane
of symmetry the flow is almost parallel to the wing's profile. This allows us to consider
a pure bidimensional CFD (computational fluid dynamics) problem.
2
Y [m]
−2
−4
0 2 4 6 8 10 12 14 16 18 20
X [m]
Since the 63-018 modified airfoil is not available in literature, a computer program
was needed to acquire vertex information of the profile's geometry. A program devel-
oped by Ladson for a NASA program [Ref. 19] seemed to fit well. This program is a
Fortran based script that reads from a text input and writes a text output with 160 vertex
that describe the airfoil.
As cross check a first attempt was made with a well known airfoil, the NACA 63018,
whose vertex are reported in literature [Ref. 18].
The Ladson method gave us a smoother and precise airfoil and so we obtained the
vertex of our NACA 63017 airfoil.
Detail
0.6
0.5
0.4
Y [m]
0.3
0.2
0.1
Abbott
Ladson
Preprocessing
Mesh definition
Since the flow will present a turbulent boundary layer (and its thickness is our target
result), an analysis of the grid spacing must be made.
In particular is important to evaluate y+ value. A simplified completely turbulent ap-
proach has been used to find a conservative value of this quantity. Given the flow speed,
static pressure and characteristic length, this approach returns the value of y+ in milli-
meters.
The importance of boundary layer around the profile is crucial for the correct opera-
tion of turbulent models.
Using the law of the wall, once an estimated value of y+ as been taken, the distance
between the wall and the first row of our mesh can be calculated.
y+ · ν
yp = ( 3.1 )
uτ
The upper bound of the log-layer depends on, among others, pressure gradients and
Reynolds number. As the Reynolds number increases, the upper bound increase too. Big
y+ values are not desirable, because the wake component grows above the log-layer.
Using excessive stretching in the normal direction to the wall should be avoided.
It is important to have at least a few cells inside the boundary layer. The mesh should
have at least 10 cells within the viscosity-affected near-wall region (Rey < 200) to be
able to resolve the mean velocity and turbulent quantities in that region.
It should be remembered that y+, y*, and Rey are not fixed geometrical quantities.
They are all solution-dependent. For example, when doubling the mesh (thereby halving
the wall distance), the new y+ does not necessarily become half of the y+ for the original
mesh.
Through the problem definition the density ρ and the kinematic viscosity of the flow
µ have been calculated, so that the dynamic viscosity can be calculated using the simply
relation.
ρ
ν= (3.2)
µ
We can estimate the friction velocity thanks to eq (3.3) and some conservative as-
sumption for estimate Cf like consider the flow fully turbulent and taking the skin fric-
tion coefficient of a flat plate (empirical correlation), Ue is an inlet flow velocity and it
is a flow parameters give by the problem.
! "
C¯f 0.0296
uτ = Ue · ≈ Ue · (3.3)
2 Rec
1
5
This choice is conservative because the turbulent boundary layer depth close to the
wall is bigger in comparison to the laminar boundary layer depth for the same problem.
A fine mesh around the airfoil is used for a depth that is bigger than the real value where
the flow is still laminar, and correct elsewhere.
It is useful to make an estimation of the boundary layer thickness, just to check if our
mesh is sufficiently fine close to the wall, for this purpose we can use the conservative
estimation showed in eq. (3.4) with the same assumption that we made for eq. (3.3)
c · 0.385
∆y = 1
(3.4)
Rec 5
Starting from the flow parameters of the problem: density, kinematic viscosity, inlet
flow velocity, chord, Reynolds number related to the chord, and from the turbulent
model specification (y+) an approximate value of yp, Δy that will fit for mesh building
process, can be extimated.
The boundaries of the mesh must be huge, since the wake of the 21.3 m chord's air-
foil will take needs space to dissolve.
As referred in Fluent tutorials a parabolic mesh might be a nice choice due to the
shape and the speed of the airfoil.
+
+
Gy Gy
+++++++++++++
+++++++++++
++++++++++++ ++++++++
+ Gz
++
+++ Gx + +++++
+Gz +++++
++++++ Gx
+
+ ++++++++++
++++++++++ ++++++++
+++++++++++
+++++++++++++++
+
+
Y Y
+
Z X Z X
The parabola keeps in the focus the airfoil and the vertex is placed some 15 chord
upstream. The pressure outlet, instead, is placed 25 chords down stream and 30 up and
down respect the intrados and extrados respectively.
Boundary Conditions
In the preprocessing the complete set of boundaries condition's family must be speci-
fied (even if through Fluent is still possible to change them).
The wall, fluid and interior conditions where applied respectively to the airfoil, to the
airstream and to the meshing guides.
As for the most important boundary conditions (inlet and outlet) a pressure far inlet
and pressure outlet was chosen. Those conditions are the one prescribed in the case of a
compressible gas study.
Once the mesh has been created a grid check was made to ensure that the quality of
the grid was satisfactory. The last step of this phase was that of exporting the mesh to let
Fluent handle it.
The same assuntions made for the Cruise case where used to develop the mesh of
Maneuver and Take off case.
Processing
The boundary conditions chosen for the cruise case are:
CRUISE
Re 131344666 [−]
! "
kg
Density 0.330
m3
! "
−5 kg
Dynamic Viscosity 1.4325 · 10
ms
To analyze the mesh a Spalart-Allmaras's one equation model and a k-epsilon two
equation model where used, since the solution converged faster with the Spalart-
Allmaras's model this one was chosen also to analyze the other two cases.
Sutherland law was used to model air dynamic viscosity (3.5), this law works only in
case of ideal gas so this option was chosen in the Fluent GUI.
! "3
T0 + S T 2
µ = µ0 (3.5)
T + S T0
Where:
kg
µ0 = 1.7894 · 10−5
ms
S = 110.56 K (3.6)
T0 = 216.812 K
This technique was recommended in the Fluent tutorial, in fact this approach avoids di-
vergence in the first iterations (the first 50 – 100 iterations).
After that, to achieve a faster convergence the Courant Number was raised:
v̄ · ∆t
Cr = ≈ 20 (3.9)
∆I
Results
Once solution converged results’ analysis was possible. As reported in the figure be-
low, a double shock on the upper and the lower side of the airfoil is present. This phe-
nomenon leads to a drag coefficient rise, and to an unacceptable separation on the upper
side of the airfoil not far from the best position to put engine’s inlet.
1.23e+00
1.10e+00
9.75e-01
8.48e-01
7.22e-01
5.95e-01
4.69e-01
3.42e-01
2.16e-01
8.95e-02
-3.70e-02
-1.63e-01
-2.90e-01
-4.16e-01
-5.43e-01
-6.69e-01
-7.96e-01
-9.22e-01
-1.05e+00
-1.18e+00
-1.30e+00
The field below the shock is still unstable and this condition helps separation. The
large, adverse pressure gradient induced by the shock causes the boundary layer to sepa-
rate.
3.88e+02
3.68e+02
3.49e+02
3.30e+02
3.10e+02
2.91e+02
2.71e+02
2.52e+02
2.33e+02
2.13e+02
1.94e+02
1.74e+02
1.55e+02
1.36e+02
1.16e+02
9.69e+01
7.75e+01
5.82e+01
3.88e+01
1.94e+01
0.00e+00
3.88e+02
3.68e+02
3.49e+02
3.30e+02
3.10e+02
2.91e+02
2.71e+02
2.52e+02
2.33e+02
2.13e+02
1.94e+02
1.74e+02
1.55e+02
1.36e+02
1.16e+02
9.69e+01
7.75e+01
5.82e+01
3.88e+01
1.94e+01
0.00e+00
In order to determine the exact position of the separation another figure is presented.
Since the point of separation is where the wall shear stress vanishes, we present the x
component of the wall shear stress.Flow reversal is indicated by negative values of the x
component of the wall shear stress.
airfoildown
airfoilup
3.00e+01
2.50e+01
2.00e+01
1.50e+01
X-Wall
Shear 1.00e+01
Stress
(pascal)
5.00e+00
0.00e+00
-5.00e+00
0 2.5 5 7.5 10 12.5 15 17.5 20 22.5
Position (m)
In the figure below there is the progression of shock waves with increasing Mach
number, as shown in Aerodynamics for Naval Aviators,2 a classic Navy training man-
ual. Our case of study seems to match the fourth case of the image (Mach 0.82). In fact,
even if the airfoil’s speed is lower (Mach 0.8), the symmetric, thick, non super–critical
configuration leads to a worst behavior.
This result leads to a new target for this thesis that is the research of a better profile
for the central section of the wing. In order to analyze as much airfoil as possible, with a
common procedure of preprocessing and processing, a computer program was devel-
oped CAD CFD Integrator.
Main Purpose
S
ince a widespread airfoil investigation should be carried out, a Matlab™ based
program was written. This program is meant to provide a tool to integrate dif-
ferent commercial softwares under a common application. Phase A is crucial
and has yet few constrains defined; in this phase of the project a versatile and easy to
use program is needed to perform similar iterative tasks.
In our case we need to choose an airfoil that grants:
enough thickness to provide comfortable cabin space
low drag coefficient
non 0 lift coefficient
low moment coefficient
almost laminar flows even at high speed (cruise speed)
Those requirements leads to a though research work, and in order to check each of
the constrain one should use different softwares. This massive work in spite of the fact
that takes a big amount of time, leads inevitably to little errors of setup (for example in
the processor setup the user has to make 150 operations, that even if simple, asks for
precise number like cos(α) or sin(α) in radiants with unfriendly floating numbers).
Another problem is that even if the commercial software is powerful, usually can not
talk to other area’s software, this communication must be made by the user sometimes
leading to frustrating copy paste from one application to another.
This situation is represented in (fig. 4.1) with the dotted double arrows.
CFD
CAD VLM
SYSTEMS
CFD
SYSTEMS
Figure 4.2 – Phase A communication between software after CAD CFD Integrator
Details
This programs highly automate the analysis of the effects of a different airfoil on the
airplane; in particular this program can:
load an airfoil
modify an airfoil
modify a VLM Tornado geometry file
analyze the complete airplane model through VLM Tornado
plot geometry of a VLM Tornado file
plot the results of a VLM Tornado analysis
write a journal file that speeds up mesh generation with Gambit
write a journal file that sets up the environment in Fluent
write a Macro for CATIA that rebuilds the geometry with the new profile
The program was built to be as easy to use as possible (with GUI or Graphic User
Interface), and with help buttons to describe each function.
When calculating something, a wait bar was made to let the user know what is going
on and how much has to wait. Particular care was taken also on an easy (and dynamic)
subfolder management, so that the user always saves in the correct folder, and keep the
program well-ordered.
Airfoil Editor
This program has many functions, is capable of reading airfoil coordinates text files
in various formats: two columns (X, Y) data 2D airfoil, three columns (X, Y, Z) 3D air-
foil, three columns (X, Y upper side, Y lower side). Each profile can be read no matter if
clockwise or counterclockwise, first–last (from 0 to 1) or last–first (from 1 to 0), since
the profile is processed to create an interpolant spline.
Once the user has imported the text file, a column with the read data appears on the
right (this column will be updated each time the user modifies the profile), and a big
plot with markers point will appear on the left side.
Now both the analytical and graphical approach helps the user to modify as desired
the profile.
Some powerful instrument that the program provide are:
• Rotation:
This tool allow to rotate of an angle the profile so that the first point lies on x=0
and y=0 and the last point x=1 and y=0, to perform this operation the program
calculates the angle and builds an Euler rotation matrix.
! "
cosα −sinα
{X}rotated = · {X} (4.1)
sinα cosα
Once rotated it also fits all the data to lie in the range [0 1]. A profile correctly
rotated is necessary to perform successive analysis. The rotation is presented on
both the column vectors on the right and on the plot with a new color and differ-
ent markers.
In (fig 4.5) a profile rotated can be seen.
• Camber:
This tool allows the user to add a reflex camber to the airfoil or to add a percen-
tile of camber to the profile. This reflex camber has been obtained from Liebeck
[Ref. 32]. This represent the mean line of the central profile chosen for the
Blended-Wing-Body designed by NASA and Boeing. The first attempt was to
obtain the curve through five points, as it seem to be matching a fourth degree
polynomial equation. Unfortunately, as presented in (fig. 4.6), even if similar,
the equation has a different shape.
Figure 4.6 – Comparison between fourth degree polynomial equation and reflex camber
To match the correct shape an interpolation spline of the data was used. In order
to obtain an initial data vector from the image in the pdf was needed another
program. The program chosen was Engauge, this program allows, once selected
three known values points on the image, to calculate the position of every other
black pixel.
Since the mean line lied in a low quality image, a smooth of the data was neces-
sary. In order to achieve smooth enough data cftool (a Matlab™ toolbox that al-
lows to create custom fits on data), was used. As presented in (fig. 4.8) the re-
sulted spline is smooth and goes from x=0 y=0 to x=1 y=0. In the same figure is
also present a comparison between the original raw data in red dotted line and the
new interpolated geometry in a solid blue line. The geometry fits well the original.
0.03
0.02
0.01
y [m]
0
−0.01
−0.02
−0.03
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x [m]
In (fig. 4.9) is described the effect of adding 100% camber to a supercritical pro-
file.
• Thickness:
This tool calculate the max thickness of the profile and ask the user if he wants
to redefine this quantity. This tool was essential since the 17% thickness is
really rare in aviation history. As before, once a thickened profile has been
asked, the programs plots with different color and marker the new geometry, and
updates the column on the right.
• Other Functions:
There are other functions that helps the user deal with the airfoil generation,
those functions allows saving, printing and clear all plots to start again the de-
sign. Each function has been developed paying attention so that each possible
combination of user input gives the correct output (for instance saving or print-
ing a cleared plot will result with a note dialog to the user that there is nothing
to save). Some examples of the application in use are presented below.
Among all the function of this tool, two of them are really useful. Those functions
allow the user to plot the entire geometry with also superimposed the mean line and the
thickness. This kind of graphic is very common in literature, so that might be useful for
comparison or presentation purposes.
This last option gives the user the chance to save his work in a graphical shape that
can be accompanied with the informations stored in the coordinate vertex file.
In order to give to the user the complete history of each operation he has actually
performed another tool was built.
This tool allows the user to generate the original plot and every operation he per-
formed on it, in the order he actually made them. Each operation has a different marker
and color in the plot, and an understandable legend clears every remaining doubt. In Fig
4.12 are presented those two last functions of the Airfoil Editor tool.
0.1
0.05
y/c [−]
−0.05
−0.1
0.2
0.15
y/c [−]
0.1
0.05
Airfoil nplx.dat
0.3
0.2
0.1
Y [m]
−0.1
−0.2
Original Airfoil
−0.3 Thickness 0.165
Rotation 0.89507 °
Camber 40 %
Figure 4.12 – Example of different operations on a supercritical airfoil and “.pdf” printed version
VLM Tornado
The first tool the user finds in the graphical user interface of CAD CFD Integrator is
the button “Linearized Tornado Setup”. This button allow the user to perform many
tasks:
• choose a new airfoil to load
• regenerate the geometry of the entire airplane
• read and update a geometry file of VLM Tornado
• load a custom state and plot geometry and lattice of the linearized simulation
• launch VLM Tornado in batch mode to perform an alpha sweep simulation
• save results of the simulation
• plot results of the simulation
This tool allows the user to modify launch and analyze a linearized aerodynamics
simulation of the entire airplane. This feature can be used to analyze the effects of a new
profile in the central section on the whole geometry .
• Choose a new airfoil to load
Once a profile has been modified with the “Airfoil Editor” tool, is ready to be
managed by other tools of the program. For instance, VLM Tornado needs a par-
ticular format for the airfoil, as is shown in the table below.
So the tool reads a common airfoil profile and saves it to be read by VLM Tor-
nado, asking the user which name he wants to save as.
#" !!"
" "
!#" !"
!!"
!$"
%6.7(48.+41.9+)-(-(04.:+20;-
<'= %'()*+,(-.+/*,0123*.+4(5.1367(2+8()296(:2;.03<
%"
)5,.90(4-
'()*)+,-./012.2!*00)1(4+-5
!" *>8>
&"
" &"!
&!!
!&" %"!
%!!
!!" &!
=.8;(>!-..18
$"!
$! $!!
!%"
! #"!
#!!
!!" " !" #" $" $!
! =.8;(7!-..18
"!
'()*)+,-./012.3!*00)1(4+-5 !$!
!
=.8;(;!-..18
Side $4+5103678+-(66(-12/(0+5(/027+10)+0(.9167:
Front
'()*+;!-((.)/0123
"!
&
Top ISO !
$!
#!
"!
%!
!"!
$!
!#!
#!
!
'()*+,!-((.)/0123
Since the analysis takes time, a wait bar was built. Once analysis is complete,
the user is informed through an information dialog box.
• Plot results
This tool can be accessed directly form the main window (through the first
question dialog box), and allows the user to plot the result of one (or more) dif-
ferent alpha sweep simulations. It gives CL vs. α where α is both presented in
degrees and radiant, this feature was not included in Matlab™, so additional
code was needed.
Angle of Attack α [deg]
0.0 2.9 5.7 8.6 11.5 14.3 17.2 20.1
1.2
1
CL − α for the full wing
0.8
Airspeed: 126.893 [m/s]
Lift Coefficient CL [ − ]
0.6
0.4
0.2
−0.2
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Angle of Attack α [rad]
A complete suite was developed in order to allow the user to analyze through vortex
lattice method the effect of a different profile in the central section on the entire air-
plane, with just few clicks.
Figure 4.20 – Help disclaimer and load geometry for the linearized aerodynamics button
VLM Tornado structured arrays are presented graphically in Fig. 4.21. The structured
array is a feature of Matlab™ that helps the user to store informations that have a com-
mon root.
•geo.nwing
•geo.nelem
•geo
•geo.refpoint
•geo.symetric
Gambit
Gambit is the preprocessor of Fluent. It allows the user to define and export a mesh
file that can be easily read by Fluent. The Gambit’s interface help the user also with
some useful tools such as a graphical mesh quality analysis (of different important val-
ues), automatic boundary layer propagation, boundary conditions setup and many oth-
ers. It’s also useful if somebody wants to keep an easy to read mesh (with names of each
edge and each face). The second tool of CAD CFD Integrator communicate with Gam-
bit. First ask the user to select an airfoil and processes it. Then ask in which folder he
wants to save the “.jou” file with all the commands to setup the geometry. In each step
of the program if the user wants to abort the sequence a note will be sent to him, and the
program will close without any error. The process go through two steps:
+ + + + + ++ + +
Gy + + + +
+ + +
+ ++ + +
++ + +
++
+
+++ +
++
+Gz Gx ++ + +
+
+++ + + + +++
++ + + +
++ + + + +
+ + + +
+ + + + ++
Z X
• Airfoil selection
Since Gambit can manage 3D environment the Matlab™ program will add a
third column of 0 (so that X Y and Z coordinates are present in the file). Gambit
has several different import options but for our case (vertex coordinates), a sim-
ple text file will be enough. Gambit’s format requirements are pretty low, so that
placing a text with two columns of coordinates works just fine.
• Journal file
The preprocessor can read a journal file (“.jou”), that store all the commands the
user would usually do and can read like an old style cassette computer all the
information stored in the text file. The approach used for this tool was to store in
a file (ASCII format), all the informations relative to the boundary conditions
(the external parabola), so that a common block of operation can be easily ac-
cessed. Here are presented the operations the journal do:
Boundary Type
Color Boundary condition
wall
+
pressure inlet pressure inlet: pressure far field
pressure outlet +
internal
+
+ pressure outlet
internal
Gy
+ Gz
+
+++
++
+ Gx +
wall: airfoil up
Gy + + + + + + + ++ + + + + +
++ +
++ + + +
+ + ++
++
+Gz + +++
+
+ Gx
+
+++ + +++ ++ + +
++ + + + +
+ ++ + + + + + +
+ + + + + + ++ + +
Y +
Z X
+ wall: airfoil down
Y
Z X
The same caution was taken here to define the mesh as was described in Central Air-
foil chapter, the big difference now is that the user just need to define what profile he
wants to use, the software will do the rest.
A nice feature of CAD CFD Integrator is that his core can manage profiles with dif-
ferent numbers of points, and will spline in Gambit them all to define the airfoil. This
interactive feature allows the user to manage a great database of airfoil without care if
some are better defined than others.
In the table 4.2 is described the journal file format.
When Gambit finds “\” it reads as command continues through next line
...
undo endgroup
save
Once journal file read in Gambit is complete, the user will face a nicely defined ge-
ometry. The geometry will be divided in 8 faces, and each face and edge has a peculiar
name, so that it can be easily accessed.
That’s not all, indeed each edge is prepared with nodes so that a typical QUAD ele-
ment mesh can be exported in just 4 clicks. The program could manage to do the mesh
of his own, but this is the most sensitive step of the process, and only the user can
evaluate if node spacing is correct in the whole geometry.
Node spacing was thought to be coarse near the airfoil’s boundary and pretty sparse
near the outer boundary (parabola). The journal also provide the user some useful in-
formation about y+ estimation described in central airfoil chapter. In fig. 4.25 result of
the journal file is presented.
+
oo o
oo o
oooooo o
ooooo o
oooooooo o
oo+oooooooo
o o
o o
oo o o
oooooo o
ooooo o
o
+
oo
oooooooo o o
o
+ooooooo o o
oooooo o o
o
oooo o o o o
oo o o o o
o
ooo o o
oo o o o
o
o o o o o
oo o o o o
o
o o o o o
o
o o
oo oo o o
oo o
o
o
o o
ooo o
o o o o
o o o
o
o oo o o
o o o
o
o o
oo oGyo o
o
o oo o
o
o o oo oo o
o
o oo o
o o
o o
o
o
ooo
oGz o
o Gx o
o
+ o o o o o o oooooooooooooooooooooooooo o
++
+
o o
+o
o+o
o o
+ooooooooooooooooo o o o o o o
+o o
+
o
o
o
ooo o
o
o o
o
o o o
o
oo o o
o o
o
o
o oo o oo o
o
o o
o o
o
o o
oo o o o o
o
o o
oo o o o o
o o
oo o o
o
o o o o o
o
oo o o o o
o
oo oo o
o
oo oo o o
o
oo o o o o
o
oo o o o o
o
ooo o o o o
o
ooo o o o o
o
ooo o o o o
oooo o o
oo+
oooooooo o o
oo
ooooooo
+ o
ooooooo o o
oooooo o o
oooo o
Y oo+ o
ooooooooo o
oooooooo o
ooooo o
Z X oooo
ooo o
ooo o
o+
An easy to use tool was made in order to communicate with Gambit preprocessor,
and setup the geometry and mesh to be exported in Fluent with just few clicks. The tool
is easy and flexible, and will tell the user what is going to do in each step.
Fluent
Fluent is a CFD processor software. It analyze a mesh file and then, given boundary
conditions and a model, iterates until convergence (or divergence) occurs.
This software is the calculations’ core of the CFD analysis. Since it can manage jour-
nal files “.jou” too the same integration as shown before with Gambit, has been devel-
oped through CAD CFD Integrator.
Unfortunately, even if Fluent and Gambit are produced by the same software house,
and both use a “.jou” journal file, the formatting code of each application is totally dif-
ferent. It was then needed another read out of the code and “reverse engineering”.
Fluent is an huge application that can manage dozen of different types of flows and
fluids. Its built–in models can manage effects such as: Thermal, Radiating, Turbulence,
Viscous, Compressible and much more.
In our application there were some repetitive actions the user should do to setup the
program, and other that need calculations (even if simple) that were time wasting.
The setup of the model is maybe the most difficult part of the entire analysis. The
user must perform almost 150 operations, some of them require an accurate transcrip-
tion of numbers.
The third part of CAD CFD Integrator was developed to communicate with Fluent
so that all the repetitive aspects of the setup were computer–made. A totally user–free
application could be developed, but this kind of analysis always requires adjustments
that only a prepared user can make. For this reason CAD CFD Integrator let the user
modify what he wants and does not launch the iteration.
The software ask the user to locate the mesh file and the profile, then ask him to de-
fine boundary conditions in form of: mach speed, altitude, and angle of attack. It then
calculate every other needed variable and start writing and saving the journal file.
In chapter Central Airfoil was explained that dynamic viscosity was chosen to obey
the Sutherland Law. In this way to define the fluid around the airfoil through the pres-
sure far field inlet and pressure outlet condition, few quantities are needed:
Mach number
Angle of attack
Static pressure
Density
Total pressure
Static Temperature
Total Temperature
Since the airplane flies in the atmosphere many of the previous information are
cross–correlated, in particular given just three quantities, using the International Stan-
dard Atmosphere as a reference all the others are obtained. Those three quantities are:
Mach number
Angle of attack
Altitude
The software developed contain a routine with a chart derived from International
Standard Atmosphere data. It might plot all the data inserted to a figure, but to keep it an
easy to use and fast program, this optional chart was not printed on screen.
Standard Atmosphere ISO
1
!ref=1.225 kg/m3 pref=10130 Pa Tref=288.2 K
0.9
0.8
0.7
!ratio Tratio pratio
0.6
Pressure
0.5 Density
Temperature
0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30
Altitude [km]
Since the user puts also speed (in therms of Mach Speed) and angle of attack the
program can calculate also:
total pressure
total temperature
! "
T0 γ−1 2
= 1+ M (4.2)
T 2
Now that every boundary condition is known the program can perform one last cal-
culation. Fluent can give also informations about: Forces, Moments, and relative coeffi-
cients.
Those information are independent of the analysis, thus they are relative to the so
called reference values. Even if the user can leave those values as they are, and watch
the results in terms of pressure and temperature, it’s easier to understand the results if
those reference values are setup correctly. The user can then work (and compare) with
quantities that are well–known. For example, the Lift Coefficient is an dimensionless
value, since the code calculates pressures (and then Forces), it’s clear that to obtain the
Lift Coefficient it rearrange eq. 4.3:
1
L= · ρ · S · V 2 · CL (4.3)
2
So that in reference value the CADCFD Integrator program calculate all data
needed:
density
speed
reference surface
reference length (for moment coefficient)
To calculate the reference surface the program needs to know which profile has been
used (thus the question to the user). It then calculates the area with a modified trapezi
method.
! b "#
n−1
(fi + fi+1 ) · (xi+1 − xi )
$
S= f (x) ≈ (4.4)
a i=0
2
Trapezi Method
f(x)
(f4 + f5 ) · (x5 − x4 )
2
f4
f5
a = x0 x1 x2 xi b = x7
Now that everything is calculated the program start writing and saving the journal
file that will be placed in the folder the user specifies.
The first two steps do not need any further explanation since the processor needs the
mesh to work and a grid check is always welcome by the user, so that can investigate if
the mesh loaded correctly (or if it is a good mesh either).
sin(α) in radiants. The journal cares that correct values, always with the same
floating point rule, are written inside the box. In the fig 4.34 there is an example
for the α = 2 [ deg ] case.
• Define Solution
The journal also setup the solution, so that the user will start the iteration with
an extreme low Courant Number, so that divergence might be avoided. Other
information is communicated from CADCFD Integrator to Fluent. In fig. 4.35
there is an example of the Solution Control panel.
• Initialize Solution
The last command the journal file gives to Fluent is to initialize the solution
starting from pressure far inlet boundary condition. This command is the last
before an analysis is usually started. The user might still change what he wants
(excluding boundary conditions) without affecting this initialization.
In the table below an example of the journal file formatting sequence is presented
(note that this is just a part of the file written by the program). This code can be read
without any particular difficulty, each command is presented in a “sort of” English that
an user (who manage Fluent) can understand. It helps a lot that each command in this
scripting language is enclosed into two round parenthesis. This feature avoid each line
break problem that was present with other scripting languages.
...
(cx-gui-do cx-set-real-entry-list "Force Monitors*Frame1*Frame3*Frame1(Force Vector)*RealEntry1(X)" '(
21104.28))
(cx-gui-do cx-set-real-entry-list "Force Monitors*Frame1*Frame3*Frame1(Force Vector)*RealEntry1(X)" '(
0.99985))
(cx-gui-do cx-set-real-entry-list "Force Monitors*Frame1*Frame3*Frame1(Force Vector)*RealEntry2(Y)" '(
0.017452))
(cx-gui-do cx-activate-item "Force Monitors*PanelButtons*PushButton1(Apply)")
(cx-gui-do cx-set-list-selections "Force Monitors*Frame1*Frame1*DropDownList2(Coefficient)" '( 1))
(cx-gui-do cx-activate-item "Force Monitors*Frame1*Frame1*DropDownList2(Coefficient)")
(cx-gui-do cx-set-list-selections "Force Monitors*Frame1*Frame2*Frame1*List1(Wall Zones)" '( 0 1))
(cx-gui-do cx-activate-item "Force Monitors*Frame1*Frame2*Frame1*List1(Wall Zones)")
(cx-gui-do cx-activate-item "Force Monitors*PanelButtons*PushButton1(Apply)")
(cx-gui-do cx-activate-item "Force Monitors*PanelButtons*PushButton2(Close)")
(cx-gui-do cx-activate-item "Operating Conditions*PanelButtons*PushButton1(OK)")
(cx-gui-do cx-activate-item "MenuBar*DefineMenu*Boundary Conditions...")
(cx-gui-do cx-set-list-selections "Boundary Conditions*Table1*Frame1*List1(Zone)" '( 0))
(cx-gui-do cx-activate-item "Boundary Conditions*Table1*Frame1*List1(Zone)")
(cx-gui-do cx-activate-item "Boundary Conditions*PanelButtons*PushButton1(OK)")
(cx-gui-do cx-set-list-selections "fluid-2-1*Table3*Table1*DropDownList1(Material Name)" '( 0))
(cx-gui-do cx-activate-item "fluid-2-1*Table3*Table1*DropDownList1(Material Name)")
(cx-gui-do cx-activate-item "fluid-2-1*PanelButtons*PushButton1(OK)")
(cx-gui-do cx-activate-item "MenuBar*ReportMenu*Reference Values...")
(cx-gui-do cx-set-real-entry-list "Reference Values*Frame2(Reference Values)*RealEntry1(Area)" '( 1))
(cx-gui-do cx-set-list-selections "Reference Values*DropDownList1(Compute From)" '( 2))
(cx-gui-do cx-activate-item "Reference Values*DropDownList1(Compute From)")
(cx-gui-do cx-activate-item "MenuBar*SolveMenu*Case Check...")
(cx-gui-do cx-activate-item "Case Check*PanelButtons*PushButton2(Cancel)")
(cx-gui-do cx-activate-item "MenuBar*InitializeSubMenu*Initialize...")
(cx-gui-do cx-set-list-selections "Solution Initialization*DropDownList1(Compute From)" '( 3))
(cx-gui-do cx-activate-item "Solution Initialization*DropDownList1(Compute From)")
(cx-gui-do cx-activate-item "Solution Initialization*PanelButtons*PushButton1(OK)")
(cx-gui-do cx-activate-item "Solution Initialization*PanelButtons*PushButton2(Cancel)")
(cx-gui-do cx-activate-item "MenuBar*ReportMenu*Forces...")
...
An easy to use tool was made in order to communicate with Fluent processor, and
setup the analysis with just few clicks. The tool is easy and flexible, and will tell the
user what is going to do in each step.
CATIA
This CAD software is one of the most complex and complete development program.
It has several different modules that can help the user manage and project using struc-
ture, electrical systems, hydraulic systems and almost every other project related area.
The user could also use a FEM (Finite Element Method) to verify the structure under
different loads. Since CATIA was one of the first to develop a program that track each
and every action the user make, and can usually (if the user did not add any reference on
child of that action) delete and update a modified project, now almost every big enter-
prise uses this software.
In this work CATIA will be used to check volumes and to place avionics with a
black–box approach. In order to achieve this result a program was developed so that
through Matlab™ (and CADCFD Integrator) the user can communicate with CATIA.
As obvious CATIA even if uses a Macro scripting code, has different dictionary and
scripting commands from Fluent and Gambit. It was then necessary to “reverse engi-
neer” a coded file in order to understand what command CATIA needed.
This last tool of CADCFD Integrator asks the user to choose a geometry to plot (the
one saved through the Linearized Aerodynamics tool), and then ask where the user
wants the .IGES file to be saved.
In the table 4.4 there is the formatting sequence required for an .IGES file.
The .IGES file has two different pieces, in the first after an initialization (START
RECORD) there are some comments and then the points are called, but not defined.
116,000000.000,000000.000,000000.000,0,0,0; 1P 1
116,000016.000,000000.000,000129.995,0,0,0; 3P 2
116,000031.999,000000.000,000227.666,0,0,0; 5P 3
116,000047.899,000000.000,000298.425,0,0,0; 7P 4
116,000063.900,000000.000,000348.903,0,0,0; 9P 5
116,000079.901,000000.000,000384.353,0,0,0; 11P 6
116,000095.899,000000.000,000410.549,0,0,0; 13P 7
116,000112.000,000000.000,000433.379,0,0,0; 15P 8
116,000128.000,000000.000,000456.062,0,0,0; 17P 9
116,000149.000,000000.000,000485.901,0,0,0; 19P 10
S 1G 4D 20P 10 T 1
In the second piece the points (and geometry) is defined through coordinates, and
record is closed with a summary of what is inside.
This file format is pretty old and thus managing operations like splines and curves in
general, is really a tough work. In this work the developed program create an .IGES file
with all vertex points, and then through CATIA Macros rebuilds the entire geometry of
the plane, except the two wings.
The two wings are not created simply because the user might add some sort of re-
finements on the multi–section surface. Profiles with their sharp edges result a bit hard
to spline and need to be smoothed before being used. As reported before, the first part of
the Macro reads the .IGES file and the second one adds the new vertex (for the central
part) to a CATIA part file: BaseWing.CATPart.
Further scripting investigation should be made in order to achieve a faster and more
powerful Macro. In fig 4.38 is represented the role of each created file.
generates
CADCFD Integrator
.IGES !
.CATvbs .CATPart
ONCE IN CATIA
copy
.CATvbs .IGES
to
Figure 4.39 – 3D View of the entire CAD model of the Full Wing
Main Purpose
O
nce the CADCFD Integrator program has been developed was possible to
start a widespread airfoil investigation. Using the Airfoil Editor tool, many
of the profiles, originally found in literature, were changed in order to
withstand our requirements. In this chapter will be reported a short description of each
profile, and eventually how it has been obtained. On the second part of the chapter will
be presented a summary table with the comparison between the results of every airfoil.
All the test have been ran using the cruise condition, that was proven to be the worst for
the airfoil. Indeed in cruise condition separation might occur, as was reported with the
first profile analyzed. In the third part of the chapter a widespread investigation on the
two most interesting profiles will be presented. This investigation will analyze the
cruise, maneuver and takeoff conditions. In the last part of the chapter one profile will
be selected as it fits better for our purposes, finally a comparison between the original
and new geometry of the Flying Wing, using linearized VLM method is presented.
Description
In next pages a brief description will be presented for every airfoil analyzed. If possible
a reference or method to obtain the airfoil will be reported. When an airfoil was ana-
lyzed as it was found in literature, the source of the airfoil will be presented.
NACA 633017
Obviously the first airfoil to be analyzed was the original one. Even if from the pre-
vious study was not reported how this airfoil was obtained, using the Ladson [19]
method a nice fit has been found (refer to Central Airfoil Chapter to verify the compari-
son). The Fig. 5.1 present the plot of the profile. This airfoil is symmetric so its mean
line does not present any camber, thus is not reflex. The profile is not supercritical.
Airfoil n63017.txt Plot, Meanline, Thickness
0.1
0.05
y/c [−]
−0.05
−0.1
0.2
0.15
y/c [−]
0.1
0.05
TsAGI B Series
This profile was the second choice presented in previous study. Even if this airfoil
present a reflex camber, is not supercritical and presents a pronounced thickness in the
leading edge zone that almost immediately collapse to a slim trailing edge. Since the
equation to plot the airfoil was given, only the geometry plot is presented in Fig. 5.2.
Airfoil TSagi.txt Plot, Meanline, Thickness
0.15
0.1
0.05
y/c [−]
−0.05
−0.1
0.2
0.15
y/c [−]
0.1
0.05
0.05
0.3
y/c [−]
−0.05
0.2
−0.1
0.1
0.2
y [m]
0
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
−0.3
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x [m] x/c [−]
SC 165 R 200
This profile was obtained using a symmetrical supercritical airfoil [33]. The original
plot has been thickened and a severe reflex camber was added, resulting plot and key
passages to reach it are presented in Fig 5.4. This profile is supercritical (note the
blunted leading edge and leaves his original symmetry to acquire a reflex camber.
Airfoil SC12.TXT Airfoil SC165r200.txt Plot, Meanline, Thickness
0.4
0.15
0.1
0.3
0.05
y/c [−]
0
0.2
−0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1 y/c [−]
0.1
−0.2
0.05
Original Airfoil
−0.3 Thickness 0.165
Rotation 0.23635 °
Camber 200 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] x/c [−]
SC 165 m 150
This profile was obtained using a symmetrical supercritical airfoil [34]. The original
plot has been thickened and a considerable reflex camber was added, resulting plot and
key passages to reach it are presented in Fig 5.5. This profile is supercritical (note the
blunted leading edge) and leaves his original symmetry to acquire a reflex camber.
Airfoil sc20414.dat Airfoil scm165r150.txt Plot, Meanline, Thickness
0.4
0.15
0.1
0.3
0.05
y/c [−]
0
0.2
−0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
Original Airfoil
−0.3 Thickness 0.165
Rotation 0.22252 °
Camber 150 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/c [−]
X [m]
SC 165 r 100
This profile was obtained using a symmetrical supercritical airfoil [33]. The original
plot has been thickened and a considerable reflex camber was added, resulting plot and
key passages to reach it are presented in Fig 5.6. This profile is supercritical (note the
blunted leading edge and leaves his original symmetry to acquire a reflex camber.
Airfoil SC165r100.txt Plot, Meanline, Thickness
Airfoil SC12.txt
0.15
0.1
0.3 0.05
y/c [−]
0
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
Original Airfoil
−0.3 Thickness 0.165
Rotation 0.23635 °
Camber 100 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] x/c [−]
0.1
0.3
0.05
y/c [−]
0
0.2
−0.05
−0.1
0.1
Y [m]
0 0.2
−0.1 0.15
y/c [−]
−0.2 0.1
NASA R 75
This profile was reported in [33]. The original plot has been thickened in order to fit
our requirements. A tiny reflex camber was added. Resulting plot and key passages to
reach it are presented in Fig 5.8. This profile is supercritical (note the blunted leading
edge) and presents a supercritical camber. This happen because the amount of reflex
camber added was not enough to change the original geometry.
Airfoil nasasc165r75.txt Plot, Meanline, Thickness
Airfoil nasasc2−0714.dat
0.4 0.15
0.1
0.3 0.05
y/c [−]
0
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
Original Airfoil
−0.3 Thickness 0.165
Rotation −0.70532 °
Camber 75 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/c [−]
X [m]
N00165 R 50
This profile was reported in literature (University of Illinois). The original plot has
been thickened in order to fit our requirements. Resulting plot and key passages to reach
it are presented in Fig 5.9. This profile was symmetrical and presents a weird geometry,
even though it has been tested. After that a little amount of reflex camber was added, in
order to compare the results.
Airfoil n0011scr50165.txt Plot, Meanline, Thickness
Airfoil n0011sc.dat
0.1
0.3 0.05
y/c [−]
0
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
−0.3 Original Airfoil
Thickness 0.165
Camber 50 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] x/c [−]
SC 165 sym
This profile was obtained using a symmetrical supercritical airfoil [34]. The original
plot has been thickened. Resulting plot and key passages to reach it are presented in Fig
5.10. This profile is supercritical as might be noticed from the blunted leading edge and
symmetric.
Airfoil SC12.txt Airfoil SC165sym.txt Plot, Meanline, Thickness
0.1
0.3 0.05
y/c [−]
0
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
−0.3 Original Airfoil
Thickness 0.165
Rotation 0.23635 °
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
x/c [−]
SC 165 R 50
This profile was obtained using a symmetrical supercritical airfoil in reference [34].
The original plot has been thickened. A little amount of reflex camber was also added.
Resulting plot and key passages to reach it are presented in Fig 5.11. This profile is su-
percritical, note the blunted leading edge.
Airfoil SC165r50.txt Plot, Meanline, Thickness
Airfoil SC12.txt
0.1
0.3 0.05
y/c [−]
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
Original Airfoil
−0.3 Thickness 0.165
Rotation 0.23635 °
Camber 50 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] x/c [−]
RAE SC 155
This profile was obtained using a supercritical airfoil [34]. The original plot has been
thickened. A little amount of reflex camber was also added. Resulting plot and key pas-
sages to reach it are presented in Fig 5.12. This profile is supercritical (note the blunted
leading edge).
Airfoil nplx.dat Plot, Meanline, Thickness
Airfoil nplx.dat
0.1
0.3 0.05
y/c [−]
0
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0
0.15
−0.1
0.1
y/c [−]
−0.2
0.05
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] x/c [−]
RAE SC 165 R 40
This profile was obtained using a supercritical airfoil [34]. The original plot has been
thickened. A little amount of reflex camber was also added. Resulting plot and key pas-
sages to reach it are presented in Fig 5.13. This profile is supercritical, note the blunted
leading edge.
Airfoil nplx16r40.txt Plot, Meanline, Thickness
Airfoil nplx.dat
0.1
0.3 0.05
y/c [−]
0.2 −0.05
−0.1
0.1
0.2
Y [m]
0.15
−0.1
y/c [−]
0.1
−0.2
0.05
Original Airfoil
−0.3 Thickness 0.165
Rotation 0.89507 °
Camber 40 % 0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
X [m] x/c [−]
Comparison
The result of the analysis of each profile is now presented in Table 5.1.
The nine columns of the chart store crucial informations: the airfoil plot, the airfoil
name, if it is symmetrical (tick means yes and cross means no), if it has a reflex camber
(this include every profile in which a percent of camber was added), the thickness, if the
profile separate (in this case the cross means that separate), lift coefficient, drag coeffi-
cient and momentum coefficient.
Low Lift BWB 1° 16,5% -0.056 [-] 0.058 [-] -0.063 [-]
Best Candidates
Among the airfoils few appear interesting for our purposes, those airfoils are the
RAE SC 165 and the NASA R 75. Those airfoils present a considerable lift coefficient
and a drag coefficient that is within standards. The most important aspect is now to de-
termine which has the best (lowest) momentum coefficient. Both airfoils do not present
a pronounced reflex camber, this means that a momentum coefficient close to 0 can not
be reach. In order to deeply investigate each airfoils’ characteristic a huge analysis ses-
sion was carried out.Using the CAD CFD Integrator was possible to investigate the two
profiles in the same boundary conditions with a range of ± 3° degrees, both in cruise,
maneuver and takeoff conditions. This analysis was made of some 30 sub–analysis. A
simplified program within Matlab™ was built in order to present the results (the feature
of the program is that it can actually append over a present figure, a new analysis).
Cruise
In the table below are reported the values for the boundary conditions of the cruise
case.
CRUISE
Boundary Condition Value
Re 13134466 [–]
! "
kg
Density 0.330
m3
! "
kg
Dynamic Viscosity 1.4325 · 10 −5
m·s
Note that in Fluent a lower Mach number was used, in order to take in count the
sweep angle of the wing. Even if we are close to the midline, an effect on the speed that
the airfoil “sees”, might still be noticeable. In literature [25] is reported this empiric
law:
!
MΛ=α = cos(α) · M
(5.1)
!
MΛ=30 = cos(30) · 0.8 = 0.744
In the next figures are reported the results for lift coefficient, drag coefficient and
momentum coefficient for the cruise case.
Angle of Attack α [deg]
Angle of Attack α [deg]
−3.4 −2.3 −1.1 0.0 1.1 2.3 3.4
0.8 −3.4 −2.3 −1.1 0.0 1.1 2.3 3.4
0.07
0.7
CL − α for the two profiles 0.065 Cd − α for the two profiles
0.6 0.06
0.055
0.5
Lift Coefficient C [ − ]
Drag Coefficient Cd [ − ]
0.05
L
0.4
0.045
0.3
0.04
0.2
0.035
0.1
0.03
0
CL Nasa Cruise 0.025 Cd Nasa Cruise
CL Rae Cruise Cd Rae Cruise
−0.1 0.02
−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad] Angle of Attack α [rad]
0.15
0.14
Moment Coefficient Cm [ − ]
0.13
0.12
0.11
0.1
Cm − α for the two profiles Cm Nasa Cruise
C Rae Cruise
m
0.09
−0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad]
It appears that even if lift coefficient and drag coefficient do not vary much between
the two airfoils, there is an huge difference when comparing the moment coefficient.
The RAE has a momentum coefficient that is almost everywhere a 33% less than the
one of NASA. This feature is really important because can mean an easier plane to
equilibrate. The analysis has been made using a routine that rotates the airfoils, so that
what on the results appears as a -1 degree is usually closer to 0. This mean that each re-
sult should be better down scaled of a degree.
Since this analysis was the most important of all the three, it has been conducted with
steps of 1 degree. In the other cases: maneuver and takeoff, a step of 3 degrees was
used.
Maneuver
In table 5.3 boundary conditions for the maneuver case are presented. The maneuver
was meant to investigate the behavior of the airfoil at cruise altitude, with a lower
speed. Eventually investigation could provide an higher angle of attack, but to keep re-
sults comparable the same range was chosen.
MANEUVER
Boundary Condition Value
Mach 0.43 [–]
!m"
Inlet Speed 126
s
Outlet Pressure 21446.69 [Pa]
Re 63451960 [–]
! "
kg
Density 0.330
m3
! "
kg
Dynamic Viscosity 1.4325 · 10 −5
m·s
The same considerations made for the cruise case are still valid here in the maneuver
case. Since the point used are less (only three), unfortunately the parabola of the drag
coefficient is not so perceptible. What appears clear is the same slope of the momentum
coefficient of the two airfoils. In the next figures the plots for the maneuver case will be
presented.
0.021
0.5
0.02
Drag Coefficient Cd [ − ]
Lift Coefficient C [ − ]
0.4
0.019
L
0.3 0.018
0.017
0.2
0.016
0.1
0.015
0
0.014 Cd Nasa MAN
CL Nasa MAN
CL Rae MAN Cd Rae MAN
−0.1 0.013
−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad] Angle of Attack α [rad]
0.105
0.1
Moment Coefficient Cm [ − ]
0.095
0.09
0.085
0.08
Cm − α for the two profiles
0.075
−0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad]
As in the previous analysis, even if the lift coefficient is slightly lower than the one
of NASA, the RAE airfoil has a better drag coefficient (lower) and a considerably lower
moment coefficient. Those characteristics leads the RAE airfoil to be our best candidate.
In order to clear all doubts over this comparison, the take off condition was studied.
Takeoff
In table 5..4 the boundary conditions for the takeoff case are presented. This condi-
tions are calculated using the very hot day standard as reported in [Ref. 36, 37]. An alti-
tude of 1665 m was used, this altitude is the one of Denver International Airport. This
city is famous for its summer hot season. Both the altitude and the temperature (and
humidity) worst takeoff capabilities for the turbofan engines. Takeoff speed was calcu-
lated using the equation reported below:
! !
WMTOW 205277·9.81 "m#
VTakeoff = 1.3 · S
= 1.3 · 894.375
= 52.96 (5.2)
ρ0 · CLMax 1.0434 · 1.3 s
The equation reported is the equation relative to the lift arranged to evidence the
speed. A corrective term of 1.3 was added as safety factor.
TAKEOFF
Boundary Condition Value
Mach 0.14 [–]
!m"
Inlet Speed 52.98
s
Outlet Pressure 82791 [Pa]
Re 117520218 [–]
! "
kg
Density 1.0434
m3
! "
kg
Dynamic Viscosity 1.7581 · 10−5
m·s
In Fig. 5.16 the comparison between the two airfoils in the takeoff condition is pre-
sented.
This analysis also proves that the RAE airfoil has better characteristics in terms of
moment coefficient and drag coefficient, the only field where NASA airfoil looks better
is the lift coefficient, but the gain that the profile gives in therms of lift are insufficient if
compared to the loss in the other two coefficients.
Angle of Attack α [deg]
Angle of Attack α [deg]
−3.4 −2.3 −1.1 0.0 1.1 2.3 3.4
−3.4 −2.3 −1.1 0.0 1.1 2.3 3.4 0.024
0.6
0.022
0.4
0.021
Drag Coefficient Cd [ − ]
Lift Coefficient C [ − ]
L
0.3 0.02
0.019
0.2
0.018
0.1
0.017
0
0.016
CL Nasa TO Cd Nasa TO
CL Rae TO Cd Rae TO
−0.1 0.015
−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad] Angle of Attack α [rad]
0.078
0.076
Moment Coefficient Cm [ − ]
0.074
0.072
0.07
0.068
Cm − α for the two profiles
0.066
−0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad]
Summary
In order to give to the reader a complete look of the analysis, three summary plot
have been created, those plot are presented in Fig. 5.17. Each condition have a different
weight on the line: solid line for the cruise condition, dotted line for the maneuver con-
dition and dashed line for the takeoff condition. Each airfoil can be distinguished from
its color: RAE airfoil is plotted in red, while NASA airfoil is plotted in blue.
0.7
CL − α for the two profiles Cd − α for the two profiles
0.06
0.6
0.5 0.05
Drag Coefficient Cd [ − ]
Lift Coefficient CL [ − ]
0.4
0.04
0.3
0.2 0.03
CL Nasa TO Cd Nasa TO
0.1
CL Rae TO Cd Rae TO
C Nasa MAN 0.02 Cd Nasa MAN
L
CL Rae MAN Cd Rae MAN
0
CL Nasa Cruise Cd Nasa Cruise
C Rae Cruise Cd Rae Cruise
L
−0.1 0.01
−0.06 −0.04 −0.02 0 0.02 0.04 0.06 −0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad] Angle of Attack α [rad]
0.15
0.14
0.13
Moment Coefficient Cm [ − ]
0.12
0.11
0.1
0.09
Cm Nasa TO
0.08 C Rae TO
m
Cm Nasa MAN
C Rae MAN
0.07 Cm − α for the two profiles m
Cm Nasa Cruise
Cm Rae Cruise
0.06
−0.06 −0.04 −0.02 0 0.02 0.04 0.06
Angle of Attack α [rad]
RAE SC 165 R 40
Finally a new airfoil for the central part of the wing has been chosen. This profile
grants better characteristics in terms of lift coefficient, drag coefficient and separation
compared with the original NACA 633017. It has an higher value of momentum coeffi-
cient but it lies between controllable range. More than fifty analysis were performed in
order to find a substitute candidate. Having defined the airfoil shape in the central sec-
tion is now possible to study the boundary layer near the aft part so that an additional
study might be performed on the shape and fluid dynamics of the Engine S–Duct.
In Fig 5.18 the RAE SC 165 R40 airfoil is presented.
0.1
0.05
y/c [−]
−0.05
−0.1
0.2
0.15
y/c [−]
0.1
0.05
0.5
Original Airfoil
New Airfoil
0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
Angle of Attack α [rad]
Figure 5.19 – Comparison in the Maneuver case between the original and new profiles.
From this comparison appears that the new profile (that is used only in the first sec-
tion of the wing) slightly improves the overall characteristics of the airplane. This gain
has a mayor drawback that is due to the growth of the momentum coefficient.
Main Purpose
O
nce the new airfoil has been chosen, is now possible to start the computa-
tional fluid dynamics 2D study on the S–Duct geometry. The first step was
to perform an analysis of the boundary layer of the simple airfoil, through
this analysis an estimation on the thickness of the boundary layer can be performed.
This chapter will explain the analysis’ process used to perform a standardized study
while optimizing the geometry of the intake and of the boundary layer intake. Some
new tools were added to CADCFD Integrator, that allow the user to easily design the
geometry of the S–Duct via CATIA, and then export the geometry to Gambit, the soft-
ware developed adds the mesh around the airfoil (only in the zones that have not been
modified), so that the mesh design process speeds up.
Once a nice geometry has been found, an extended analysis was performed. Results
are presented throughout the chapter.
Boundary Layer
In order to describe with a better detail the fluid around the airfoil, a study has been
conducted on the RAE profile in cruise conditions.
In particular the study should tell what fluid conditions are around the 65% of the
profile, in this zone should start the engine intake. Knowing the boundary layer thick-
ness in this zone is crucial to define the geometry of the intakes. In the table below are
summarized the boundary conditions used for the cruise case.
Re 13134466 [–]
! "
kg
Density TABLE 2.- Nominal test conditions. 0.330
m3
M TT0, °F PT0, psi Re/ft Re,D2 W2C/Ai, lbm/s·ft!2 " A0AH
6.9 x 10 1.4325
kg 0.98 – 1.58
10 −5
6 6
0.25 Viscosity 34 x 106
-279Dynamic 65 20.3
· – 33.4
0.40 -279 62 51 x 10 10.4 x 106 20.3 – 33.8 m · s 0.64 – 1.06
0.60 -279 62 68 x 106 13.9 x 106 20.3 – 35.4 0.47 – 0.82
0.80 -279 52 68 x 106 13.9 x 106 20.3 – 35.8 0.40 – 0.72
0.83 -200 Table
30 6.1–
25 xCruise
106 boundary
5.1 x 106 conditions.
20.3 – 37.1 0.39 – 0.72
0.83 -200 36 30 x 106 6.1 x 106 20.3 – 37.1 0.39 – 0.72
0.83 -200 41 35 x 106 7.1 x 106 20.3 – 37.1 0.39 – 0.72
0.83 -200 50 42 x 106 8.6 x 106 20.3 – 37.1 0.39 – 0.72
0.83 -279 30 42 x 106 8.6 x 106 20.3 – 36.5 0.39 – 0.72
The first attempt geometry is obtained using the study made in [13], as seen in Fig.
0.83 -279 40 57 x 106 11.6 x 106 20.3 – 36.5 0.39 – 0.72
0.83 -279 50 68 x 106 13.9 x 106 20.3 – 36.5 0.39 – 0.72
6.1. In the image the four different candidates designed by [13] are visualized, there are
two semi–circular geometries and two semi–elliptical. The best resulting candidate was
used as test geometry (it was the semi–circular one).
Model A – Wi/2Hi = 0.95, a/b = 2.0 Model C – Wi/2Hi = 1.42, a/b = 2.0
Model B – Wi/2Hi = 0.95, a/b = 3.0 Model D – Wi/2Hi = 1.42, a/b = 3.0
9
American Institute of Aeronautics and Astronautics
It is fundamental is to understand where is the best zone to start the engines intake
geometry, considering integration with the primary structure, cabin layout and fluid dy-
namics of the simple airfoil. In order to unveil the fluid dynamics of the airfoil, result
are presented.
Fig. 6.2 shows the static pressure contours of the simulation.
3.08e+04
2.98e+04
2.89e+04
2.79e+04
2.69e+04
2.60e+04
2.50e+04
2.40e+04
2.30e+04
2.21e+04
2.11e+04
2.01e+04
1.92e+04
1.82e+04
1.72e+04
1.63e+04
1.53e+04
1.43e+04
1.33e+04
1.24e+04
1.14e+04
Another key parameter to look at are the temperature profiles of the simulation, if
some abrupt change is found, probably a shock is present, and thus the drag of the pro-
file increases.
2.38e+02
2.36e+02
2.33e+02
2.30e+02
2.28e+02
2.25e+02
2.22e+02
2.20e+02
2.17e+02
2.14e+02
2.12e+02
2.09e+02
2.06e+02
2.04e+02
2.01e+02
1.98e+02
1.96e+02
1.93e+02
1.90e+02
1.88e+02
1.85e+02
Power requirements
Engines Gross dimensions
Boundary Layer Area
Engine’s Intake Geometry, based on [Ref. 13]
As referred in previous chapter the FW-300 has a thrust to weight ratio of 0.25, this
joined with the MTOW gave an initial estimation of the thrust requirements of the air-
plane.
Thrust
= 0.25
Weight (6.1)
Thrust = 0.25 · 205200 · 9.81 ≈ 510 [kN]
In this calculation there are already some safety factors so that this is the total thrust
needed from the airplane.
Knowing that, and looking to common turbofans, a research was made in order to
choose the best one to fit the thrust requirements, with a particular care on the dimen-
sions.
Among the many different turbofan that actually power transport aircraft of similar
size, there are few of interest. The main difference between them is the chance to be op-
timized for a bleed or bleed–less powered ECS.
Aviation history has used Bleed since decades. This technique actually takes a per-
centage of the hot high pressure air from different (depending on the requirement and
rpm of the engines) compressor stages. This hot high pressure air is then used as a
power supply and as fluid for the ECS or environmental control system and pneumatic
system. Since this air is supplied directly from the compressor, it is a power loss, in
terms of possible thrust not used. In particular turbines must spin also for this hot air
that does not go through the full engine cycle. A new technology has been (and is still)
developed during this last years. The growing reliability and controllability of electrical
power devices has open a new field also in aviation. The interest on a more (or all) elec-
tric airplane has taken some of the most important military aeronautical enterprises to
build the first all electric fighter. After that the Boeing enterprise announced the Boeing
787 Dreamliner.
This airplane is the first more electric transport airplane, that almost eliminates the
pneumatic system, substituting it with an ECS that uses is own air probes. In this way
an almost no bleed engine (optimized) was designed, the Genx (General Electric Next
Generation).
This General electric turbofan was used as a mean of comparison for our purposes.
Both the thrust and dimensions where interesting, but above all was interesting this new
bleed–less technology, that could be adopted and as we will see optimized for our de-
sign purposes. In the next table a brief summary of the general dimensions and perform-
ances of the turbofan is presented.
The Genx is a twin counteracting spool turbofan whose pressure ratio is of 19:2 and
the 10 stages compressor pressure ratio is 23:1. Its composite–titanium leading edge fan
blades present lighter and more reliable performances than previous one.
GEnx-2B-67
Dimension Value
Fan stages 1
Boundary Layer calculation where made using the equation reported below.
! b
ṁ = ρ (x) · V (x) dx
a (6.2)
Since the boundary layer velocity profiles grows fast until reaching almost the mean
velocity of the stream, the limit of the boundary layer was selected to be at some 80% of
the velocity of the stream.
The Log law represent the turbulent boundary layer between the inner and outer lay-
ers bounds. Where u+ represent the dimensionless speed, K and B are two constant of
the flow.
1
u+ = · ln y + + B
K (6.3)
K = 0.41 B = 5.2
18
16
14
12
10
u+
0
0 50 100 150 200 250 300
y+
Using the data provided by General electric was possible to perform the calculations
of the Matlab™ program.
Engine S−Duct Intake Geometry
4
Area Boundary Layer=0.752 m2
Engine Intake
3.5 Area Intake=4.8106 m2 BLI Lip
Radius (without BLI) =1.75 m
3 Reference Engine Radius =1.225 m
2.5
2
y [m]
1.5
0.5
−0.5
−1
−3 −2 −1 0 1 2 3
x [m]
This first attempt of the frontal geometry developed with Matlab™ needs to be inte-
grated with the model in order to develop a computational fluid dynamics simulation.
From
To
.stp CADCFD Integrator .txt
The tool developed allows the user to draw the geometry with ease using a particular
CATIA file, then through CAD CFD Integrator (Fig. 6.8) only the most peculiar points
are saved, in Matlab™ a plot is presented with the geometry (Fig. 6.9). The program ask
the user if he wants any kind of conversion between units of measurement, to allow a
better accessibility.
Using CATIA as design tool is possible to adjust the shape only by moving some key-
points, as they move the geometry automatically refresh, allowing the user to know bet-
ter how the changes influence the whole geometry. It is also very easy to verify the
length and in general the dimensions of the profile. In theory it would be possible to let
Gambit read directly a “.CATPart” (CATIA model) file, but since the “.CATPart” is
really a complex file, it is better a prudent approach, leaving Matlab™ do the conver-
sion and leaving Gambit with its own geometrical spline routines. This approach avoid
any tolerance and proximity error that might occur when importing a geometry with dif-
ferent math laws.
'
&
$
-)*+,
!$
!&
!%
! " #! #" $!
()*+,
This resulting plot in Matlab™ is given by a conversion of the .stp file to a raw coor-
dinate text file. The two different file formats are presented below, in Fig. 6.10. The
program scans all the file in order to identify the keywords relating to the coordinate
point object in the .stp file. It then save the coordinates to a more friendly text file.
Figure 6.10 – Different file formats, before and after the conversion.
Using the Gambit journal file approach (explained in previous chapters), it was then
possible to create a mesh file almost complete, with only the intake geometry left blank
from the mesh. This process is completely automated and grant an huge save in time
when trying to find the better geometry. After using CAD CFD Integrator the first step
the user has to do is manage to finish the mesh. In Fig. 6.11 is presented the state of the
mesh when the user has to perform the first operation.
++
+
+
+
++
+
Gy + ++ + + ++ ++ ++ ++ ++ + ++ ++++ + ++
++++++++++ +
++++++++ +
+
++++
+
+ ++
+Gz
++ Gx+ + + +
++++++ ++
++++++++++++ + +
+
+++++ ++ + + ++ ++ ++
++ ++ ++ ++++ ++
Z X
As reported in previous picture, the user can speed up the optimization process
through this software. Once a nice mesh has been obtained is then possible to use the
Fluent tool in order to speed up the setup of the processor.
In the Fig. 6.12 are presented the different geometries that have been studied, in a
solid black line is drawn the final geometry used.
It can be observed that few of the points have not been moved throughout the opti-
mization process. Those points are the one referring to the turbofan inlet disk, known
data, turbofan case length, boundary layer intake height, obtained by the previous, sim-
ple, airfoil study.
Mesh
This particular geometry needed attention on some key–points. The mesh, close to
the profile, due to boundary layer is really coarse. But what happen when the mesh ap-
proach the intake and the boundary layer lip? Using the standard grid would have re-
sulted in a sparse mesh near the boundary of the intake and S-Duct. That was unaccept-
able since the target of this study is to analyze this region. On the other hand, letting the
number of nodes grow wildly is never a brilliant idea, since basically we are only man-
aging to create a more complex problem to solve.
It was then necessary to perform some assumptions: the design criterion of the mesh
was that to ensure the mean stream lines to be sought by the mesh. In this way the cal-
culation performed by the processor ensure less errors between cell and cell, helping to
find a convergence solution. In order to ensure this criterion a non conformal grid was
used. This particular grid is well known in literature and even if it must be used in zones
without any big change in flow quantities, it ensure a better defined domain and much
detail. The non conformal grid is based on the hang nodes, they are used in some grid
adaption tools of Fluent, and are referred in the manual of both Gambit and Fluent. The
most interesting feature of this element is that it hangs on one side, while on the other
start a new set of grid mesh. An example of the concept is reported in Fig. 6.13.
Using this technique the mesh was detailed enough around the walls, so that also the
boundary layer lip and the S-Duct were defined to be studied.
In the Fig. 6.14 is drawn the final mesh used for the study of the S–Duct. The reader
can notice the more defined points that start from the 50% of the chord to the boundary
of the upper engine intake lip. The geometry here presented is taken starting from the
profile at its max chord, 21.3 m. Then adding the engine and the intake added some 3
meters, so that the final dimension are far bigger than the one should expect at 5 m of
wing span. This choice was made in order to perform a simplified study, since the mesh
of the two blended airfoil was difficult to replicate and than adding the engine in a 3D
wing, with dihedral, swept and tapered, pose some serious integration problems, with a
3D case to be model, and smooth shapes.
+ +
+
+ +
+ +++
++
+
+ +
+
+
+ + + + + + + + + + + + + + + +++
+ + + + + + +
++ +
++++++++ +
Gy +++++++++ ++
++
+++++ ++
+Gz
+ Gx
+
+ + + +
++++ +
++++ ++
++++++ +
+++++++++ + +
++ + + ++
+ + + +
+ + + + + + + + + + ++ + + +
Z X
Boundary Conditions
Once ended with the mesh it is now important to set up the all new sets of boundary
conditions that inherently this geometry requires.
p2 , u2 , M2 , ρ2 , A2
p1 , u1 , M1 , ρ1 , A1
A boundary condition at the turbofan intake and at the boundary layer lip intake
should be defined. Since few of the information at the intake are known, like pressure,
area and mach at turbofan intake, and all the inlet values have been defined, by previ-
ous, simple airfoil analysis, a compressible problem can be solved.
The procedure here described is found on the [38] Zucker “Fundamental of Gas Dy-
namics”. In order to study the problem some assumptions have been made:
∂
=0
Steady one-dimensional flow ∂t
Adiabatic q̇ = 0
No shaft work Li = 0
Perfect gas p · ν = RT
Neglect potential g · ∆z ≈ 0
Starting from the perfect gas equation of state, just defined, we analyze continuity:
ṁ = ρ · V · A ≡ const (6.1)
ρ1 · V1 · A1 = ρ2 · V2 · A2 (6.2)
A2 ρ1 · V 1
=
A1 ρ2 · V 2 (6.3)
We substitute for the densities using the equation of state and for velocities
from the definition of Mach number:
! "! "
A2 p1 RT2 M1 · a1 p1 T2 M1 a1
= =
A1 RT1 p2 M2 · a2 p2 T1 M2 a2 (6.4)
!
a= γ·R·T (6.5)
!
A2 p1 · M 1 T2
=
A1 p2 · M 2 T1 (6.6)
Now we find a means to express the pressure and temperature ratios in terms of
M1 , M2 , γ and ∆s .
Since for a perfect gas enthalpy is a function of temperature only, we can write:
Using the relation between static and stagnation temperatures for a perfect gas:
! "
γ−1
Tt = T 1+ · M2
2 (6.10)
! " ! "
γ−1 γ−1
T1 1+ · M12 = T2 1+ · M22
2 2 (6.11)
or:
γ−1
T2 1+ · M12
= 2
γ−1
T1 1+ 2 · M22 (6.12)
Since no subscript has been used on the specific heats ratio γ , a constant specific
heats ratio value has been assumed. This assumption is close to reality when tempera-
tures gradients are negligible. Through the polytropic process a relation between static
and stagnation pressure can be written.
! " γ−1
γ
γ−1
pt = p 1+ · M2
2 (6.13)
Using the previous equations and rearranging them it is possible to define completely
the S–Duct, in order to find a result a Matlab™ program was written, so that the calcu-
lation could be repeated with different boundary conditions.
In the [38] the procedure added the chance to adjust the solution using a ∆s to
model a gas loss due to friction or other losses. Since this is a simplified simulation, the
entropy gradient has been neglected.
In the table 6.3 calculated values for the S–Duct are presented.
Inlet Intake
p1 = 12076.58 [P a] p2 = 26831.5 [P a]
γ = 1.4 [–]
Results
Once the mesh has been created, and boundary conditions calculated, is now possible
to perform the computational fluid dynamics simulation. The results are presented in the
next figures. In the first picture, Fig. 6.16, the static pressure is presented. The two ver-
tical dark lines are the zones where a non conformal grid was used. The figure does not
present abrupt gradient of pressure inside the duct, thus a pretty undistorted flow is pre-
sent. A little separation on the lower part can be noticed. Those results are in line with
what was expected, the simplified nature of the study does not allow a detailed 3D
simulation of the S–Duct. In this way some mayor distortion effects can not be noticed,
by the way in the 2D simulation the area the CFD code sees grows from 1.75 to 2.5 m,
this worst the separation of the flow.
3.08e+04
2.96e+04
2.85e+04
2.74e+04
2.62e+04
2.51e+04
2.39e+04
2.28e+04
2.16e+04
2.05e+04
1.94e+04
1.82e+04
1.71e+04
1.59e+04
1.48e+04
1.36e+04
1.25e+04
1.14e+04
1.02e+04
9.06e+03
7.92e+03
In the next figures static temperature, and Mach are presented. A weak shock is evi-
dent in the lower part of the airfoil, while in the upper part after the case an evident
separation can be seen. The separation on the case could be avoided through a better
design of the nacelle.
2.39e+02
2.35e+02
2.31e+02
2.28e+02
2.24e+02
2.21e+02
2.17e+02
2.14e+02
2.10e+02
2.07e+02
2.03e+02
2.00e+02
1.96e+02
1.93e+02
1.89e+02
1.86e+02
1.82e+02
1.79e+02
1.75e+02
1.72e+02
1.68e+02
1.47e+00
1.40e+00
1.33e+00
1.25e+00
1.18e+00
1.11e+00
1.04e+00
9.62e-01
8.89e-01
8.17e-01
7.44e-01
6.71e-01
5.98e-01
5.25e-01
4.52e-01
3.79e-01
3.06e-01
2.34e-01
1.61e-01
8.79e-02
1.50e-02
Figure 6.17 – Static Temperature and Mach Number contours of the S–Duct.
Main Purpose
S
ince the geometry of the S–Duct has been studied, now the results let us know
the quantities both in the boundary layer lip and in the proper S–Duct. The air
that passes through the boundary layer lip is considerable for one engine, and
obviously duplicates since this is a twin engine configuration airplane. Some studies
were conducted in order to understand if a no bleed ECS, environmental control system,
should be used for our purposes. Some calculations were made and the compressor and
heat exchanger were preliminary designed. Then a revised 3D CATIA model was cre-
ated in order to verify the compatibility between the old and the new configuration.
Results
In the table below, there is the summary of the results of both boundary layer lip and
S–Duct, those values are taken in the inlet of both the conducts. In order to better com-
prehend the geometry of the model, a draw is presented in Fig. 7.1 which describe the
different zones with the proper name.
Zone Label Area Average Speed Average Density Mass Flow Value
Boundary Layer Inlet 0.752 [m2] 145.42 [m/s] 0.35744 [kg/m3] 39.09 [kg/s]
Engine Intake 4.810 [m2] 237.76 [m/s] 0.31686 [kg/m3] 362.36 [kg/s]
Engine Inlet 4.9 [m2] 193.13 [m/s] 0.31404 [kg/m3] 362.41 [kg/s]
Let us focus on the boundary layer inlet: those values are referred to a single engine
so that the mass that flows, in cruise condition, inside this lip is considerable. This huge
amount of air can not be used to breath the engines because of its turbulent low speed
nature (almost 100 m/s) of average difference. If such air could reach the engine disk, a
considerable cyclic load on each blade of the turbofan would be present. This kind of
load lead to an high cycle fatigue issue and worst the engine maintainability and reli-
ability. Finally this kind of turbulent flow could induce significative engine efficiency
losses [11].
It appears that this semi–sinked configuration inherently leads to manage an huge
mass of air, the integrative nature of airplane designs give us the chance to perform
some study on what system could benefit from such mass of air and eventually elimi-
nate the one in excess.
The results reported on previous table are obtained through the computational fluid
dynamic simulation performed on the full geometry of the S–Duct.
Bleed system
The typical aircraft ECS system configuration is the bleed system. High pressure air
is obtained through a bleed on the compressor of the turbofan engine. Based on the en-
gine RPM or revolutions per minute, in order to match air temperature and pressure re-
quirements, bleed is taken either from low pressure, intermediate pressure or high pres-
sure compressor stages.
This air bleed produce a power loss since the turbines must spin to compress that
amount of air, even if optimized the efficiency loss are always present in the engines.
In the Fig. 7.2 the typical airliner ECS system is reported. In the picture various sys-
tems are presented in different colors: green, sky–blue, yellow and red the different
bleed stations. On the right of the picture there is the built–in test module that checks
and monitors the values of the bleed so that the different valves can be shut down. With
this system architecture the amount of wasted energy can reach a 30% [39], most of it
thrown away by the precooler. Due to the high safety reason the high temperature and
pressure air can not be used around the airplane, so it is immediately cooled down.
F Fan
starter valve
(reference)
fan air mod valve
F T
check valve
I
to cowl
H high pressure precooler anti ice
shut off valve
temperature
to aircraft
sensor to wing
anti ice
The Bleed System automatically provides air at the proper temperature and
pressure required to meet needs of all pneumatic services on the aircraft
Part of the air is then used for anti ice on the blade and on the case of the engine, or
eventually for the anti ice system in general (some example of electrical anti ice system,
and bleed system are nowadays present). Temperature and pressure in the cabin must lie
between a specified range, that is almost constant throughout the seasons. The effi-
ciency of this system is particularly low during taxing or on ground operations in gen-
eral, while the engines are spinning at a low RPM and thus the bleed must be obtained
by high pressure compressor stages. Even if presenting many limitations, this system
has been used almost always on commercial airplanes. The reason of this choice can be
found on the difficulty to find an adequate power source that could provide both higher
efficiency and higher or at least comparable reliability.
During the last years a growing interest on the more electric airplane or all electric
airplane was reported. This new interest on a different architecture was probably
strengthened by the new generation of electrical motors that provided more power while
reducing many of the weight / reliability problem of the older one.
Bleed–less system
The most noticeable example of this architecture is the Boeing 787 Dreamliner. This
airplane is also known as more electric airplane. Even if this is the first example in the
commercial transport aviation, a milestone airplane was the F-35 Fighter. This military
airplane was the first aircraft with all–electric characteristics. The various system of the
Boeing 787 were designed to be electrical, where possible. The hydraulic system is
powered by an electrical motor so that the pipes and hydraulic system in general has
been considerably reduced.
The pneumatic system was also reduced so that the only real component that is still
pneumatical (and thus require a little bleed) is the engine’s blade and case anti ice sys-
tem.
The architecture of the Boeing 787 Dreamliner is presented in Fig. 7.3.
787 NO-BLEED SYSTEMS ARCHITECTURE
Figure 1
RAT
P G
RAT
P G
P M SG
P Engine
HX P M SG
M SG
Engine
ECS
HX P
XFR
CTAI
M M P M
SG Battery
ECS XFR CTAI
Battery
M
P M HX Heat Exchanger
M
M SG
XFR Trans / Rectifier
NGS NGS APU SG
SG APU Starter Generator
P SG SG
M P M
CTAI Ram Air Turbine
ECS XFR RAT
M SG
HX P Hydraulic Pump
M P M P XFR
Engine
ECS M SG CTAI P M Motor
M
HX SG CTAI Cowl Thermal Anti–Ice
P Engine
Electrical
P M SG
Hydraulic
HX Heat Exchanger
ELECTRICAL
POWER IS MORE
Figure 7.3 – Boeing 787 Dreamliner more electric airplane.
Electrical
Hydraulic
EFFICIENT THAN
Pneumatics
ENGINE-GENERATED
Fuel
PNEUMATIC POWER.
Ram Air
07
WWW.BOEING.COM/COMMERCIAL/AEROMAGAZINE
In the Fig. 7.3 different colors are used to represent different systems: sky–blue for
electrical, brown for hydraulic, purple for pneumatic yellow for fuel and black represent
ram air. Those color are also used to represent other components with the exception of
green that is used for RAT (or ram air turbine) and the heat exchangers.
The two engines provide the electrical power to four starter generators, then this
power reach the hydraulic pumps, the two ECS air conditioning packs, electrical anti ice
system on the wings and the NGS or Nitrogen Generator System. The NGS is a system
that pumps nitrogen inside the fuel tanks, providing a safer inert environment for the
tanks so that they should not ignite or explode in case of crash.
In this particular architecture a no bleed engine is designed, this engine’s efficiency
can reach an higher value, not only due to the better thermodynamic cycle (more stable
since the air that enters the engine equals the air that flows away from it), but also be-
cause there is no more the variable request in term of air from the ECS, this means an-
other time a more stable cycle so that the cruise condition can be optimized with an
higher detail. Let us have a detailed look on the ECS system, Fig. 7.4.
Legend:
Non Return Valves: NRV
Bypass Valve
Fwd Aft
Cargo Cargo
Pressure Regulating Valve
Mix
Outflow valves
Mainfold
Extraction fans ACP II
Turbine Water Extractor Heat Exchanger
Filters
Motor
Ozone Remover
Compressor
There are six main compartments to be filled with pressurized air: flight deck, for-
ward cabin area central cabin area, aft cabin area, forward cargo bay, aft cargo bay. Each
of those compartments can be reach through one of the two air conditioning packs.
The different colors used in the previous image are meant to provide a graphical help
on the temperatures and pressure of the air. In this particular architecture the air required
for pressurization and breath issues is provided through some static probes. The light
and rarefied air is compressed through the compressor and then cooled through the mix
main fold.
The power required from the air cycle is provided by an electrical motor. At the end
of the cycle some filter clear the air that will be reused for recirculation purposes. The
system is designed to have some leakage so that the air inside the cabin is changed
every few minutes. This architecture provides a new configuration and even if a bit
more complex when compared with the classical bleed architecture, reaches better effi-
ciency values.
Another concern of this technology is that it requires higher maintenance, but since
this is the first example of new technology, this field can be highly improved.
As explained previously the only part left on site of the pneumatic system is the anti
ice engine blades and case system.
The configuration of the ECS probes and position of the engines suggest this to be
the the best solution for the anti ice problem.
Cabin requirements
Typical cruise altitude is of 10 to 11 kilometers from the sea level, at this height the
environment is non survivable. The two mayor normative regarding commercial trans-
port aircrafts, the FAR-25 and CS-25 (United States and European normative), define
the ranges of temperatures pressure and density that must be respected in the cabin. A
detailed prescription is also posed in terms of fresh air that must be available for each
passenger [36] [37].
Designing an ECS involves some mayor calculations, and many terms should be
considered:
Only few of them can be calculated, like the thermal power due to heat of the pas-
sengers. An rudimental approximation can give some 100 W for passenger if he is calm,
this value can be raised to 150 – 200 W if he becomes nervous. When designing a 300
seat aircraft the total thermal power produced can be calculated to reach 45 – 60 kW.
Since the design is still in a preliminary phase, instead of the complete design and esti-
mation of the ECS system, a simplified compressor and heat exchanger estimation will
conducted.
!"#$%%&!'()*+#!,%#$*-*%#)!./01234!
In order to define the previous values, some references were used. Looking at the
normative the value of +,=0>2!
;<,! 02+**8! temperature should
3*2+0?0*2=! 5=,+!
(
be within
02! /*1+! 31/35/1?0*2=! -5=?! the American
7,! 3*2=0=?,2?! @0?<! ?<,! Society of Heating,
8,3*--,2+1?0*2=! 02! ABCD EACFGAH! C?12+18+! ::$(&&%I! ;<,8-1/! H2J08*2-,2?1/!
)*2+0?0*2=! K*8! F5-12! L335.1239'! ! ! )*-K*8?! +,.,2+=! *2! 7*?<! ?<,! +89$75/7!
Refrigerating and Air-Conditioning Engineers or ASHRAE comfort envelope. In the
?,-.,81?58,!12+!?<,!1-*52?!*K!-*0=?58,!02!?<,!108'!!;<,!3*-K*8?!,2J,/*.,!3*2=0=?=!*K!
1//! 3*-7021?0*2=! *K! +89$75/7! ?,-.,81?58,! 12+! -*0=?58,! K*8! @<03<! -*=?! .,8=*2=! 18,!
Fig. 7.5 the ASHRAE standard graphic is presented.
3*-K*8?17/,'! ! ;<,! ,2J,/*.,! 0=! =.,30K0,+! =,.181?,/9! K*8! =5--,8! 12+! @02?,8! 12+! 0=!
=<*@2!02!60>58,!"#(A'!!!
!!"#$%&'!()*+!,-'!+./0+1!23453&6!178'93:'!
!
°F
70
15
RH
%
70
0%
10
65
HUMIDITY RATIO (G/KG OR LB/1000 LB)
%
50
%
60
DEW POINT TEMPERATURE
60
10
55
EFFECT
%
R 30
MME
R
50
NTE SU
IVE TEM
WI
45
PERA
40
5
TURE ET
35
30
* (°F)
25
20
15
5 68°F 79°F
60 70 80 90°F
OPERATIVE TEMPERATURE !
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
(!A-,80312!B1?0*21/!C?12+18+=!D2=?0?5?,! Figure 7.5 – ASHRAE Comfort envelope.
"##$%! &#'(!)*+,!)*-./0123,!41251/!$!6,785189!(&&:!
If looking to the chart it can bee noticed that the ranges during winter and summer
intersects in a zone so that we will consider this range to be the desired range. In par-
ticular a minimum value of 20° C for the basic temperature requirements in the aircraft
could be chosen. However, nowadays temperature is a criterion of comfort for airliners
and equipment manufacturers. This temperature tends to reach 22°C in most aircraft and
combined with an increased level of humidity, it contributes to the well being of the oc-
cupants. Cabin pressure can vary within different planes and in the same airplane ac-
cording to its altitude. As a matter of example the Cabin pressure versus outside pres-
sure in a Boeing 767 is presented in the chart of Fig. 7.6. Since the designed flying wing
2.5
Cabin altitude in 1000 ! [m]
00 [m] ≈ 34450 [f t] 2
0.5
!0.5
!2 0 2 4 6 8 10 12 14
Airplane altitude in 1000 ! [m]
The last value that needs to be defined is now the fresh air required for the cabin. Ac-
cording to the EASA CS-25 requirements, the minimum fresh airflow to be delivered to
the cabin and therefore to be conditioned is 0.35 kg/min/pax.
“Each passenger and crew compartment must be ventilated and each crew
compartment must have enough fresh air (but not less than 10 cubic ft per minute
per crew member) to enable crew members to perform their duties without undue
discomfort or fatigue”
This quantity grows to a 1.75 kg/s if considering the total amount of air required
from the crew and passengers. As explained previously this value is considerably lower
than the 60 kg/s that the two boundary layer lips ingest. Even if managing to use some
of this air to cool down the heat exchanger, and also with anti ice purposes, the available
mass is still considerable. A diverter with some static blinds should be the proper con-
figuration in terms of total efficiency values.
In the [39] some more information were found. There is a break even point between
the two different architectures: bleed and no–bleed. This break even point is presented
in terms of PAX, under a 250 the bleed system is still the best solution, since the penal-
ties due to compressor and heat exchanger requirements lower the general efficiency of
the no–bleed architecture. When crossing the range of 250 passenger things changes,
and no–bleed architecture becomes advantageous.
In Fig. 7.7 the overall system fuel weight penalties are presented. There are four dif-
ferent architecture proposed in the [39], bleed and no–bleed (generators), have been de-
scribed, while Auxiliary Power Unit or APU and Fuel Cells are other two options. The
APU architecture dedicate a gas turbine only to the ECS system, so that it should be
used through out the full flight envelope (usual airplanes uses APU only on ground or in
some emergency procedures at low altitude in order to re ignite the turbofans). The
fourth architecture studied on [39] is the Fuel Cells, this promising configuration uses
the hydrogen powered fuel cell system in order to supply the power requirements of the
ECS, the key feature of this technology is that it is a “green” technology, that lowers the
emissions and do not require any fossil combustible.
1.20
1.00
% Design Fuel Mass
0.80
Bleed
Generators
0.60
Fuel cells
APU
0.40
0.20
0.00
1
150 2002 4003 > 5004
PAX
Figure
Figure 27 : ECS7.7 – Overall
Overall system
system fuel fuel weight
weight penalties penalties
- % Design Fuel Mass[39].
Another interesting1. graphic provided by [39] describes the overall systems weight
ELECTRICAL ECS
penalties expressed in Newton. Each system architecture has its own sets of components
that as it can be As
easily assumed do not weight the same. As in previous case, the no–
we can see on Figure 26 and Figure 27, fuel cell systems generate the highest
overall system fuel weight penalties of electrical ECS.
bleed configuration appears to be recommendable for a 300 seats aircraft, because,
This is mainly because the mass of the system remains important compared to
while providing a low fuel mass required, it grants a weight save. With the growing in-
other electrical power systems and induces in addition to the fuel required by the fuel
terest on a less
cell(no) emission
an increase aircraft
in fuel mass even
penalties. But a weight
most saveis generated
of this mass of few by grams can imply tons of
additional
devices adopted upstream and downstream the fuel cell. The fuel cell itself is an
CO2 reduction in an year of airplane life. In the Fig. 7.8 the ECS overall system weight
efficient mechanism with an attractive power density.
All the previous results are gathered in the chart below. This enables to compare
the impact of the different power systems, bleed and electrical, on the performance of
the aircraft, and define the most suitable solution for each category of aircraft. Figure 26
and Figure 27 represent the overall system fuel weight penalties respectively Newton
and % of the Aircraft Design Fuel Mass. Giovanni Medici ~ Engine Integration on a Flying Wing
180,000.00
160,000.00
140,000.00
Individual Research Project ECS for the All Electric Aircraft
120,000.00
ψ % "
passenger are presented.
The details of the calculations are reported in appendix A. The results are
summarised in
Figure 16.
550
Approx. Electrical Power Consumption ( kW )
500
450
400
350
300
250
200
150
100
50
0
< 10 PAX < 100 PAX 150 PAX 200 PAX 400 PAX 550 PAX
Aircraft Size ( PAX )
Now that all the desired features have been defined a summary of the problem data
can be presented, Table 7.2.
CABIN VENTILATION
Condition Value
The values of the desired pressure and density have been derived using the interna-
tional standard atmosphere data. The procedure was the same as the one described in
previous chapters.
0.8
0.7
!ratio Tratio pratio
0.6
Pressure
0.5 Density
Temperature
0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30
Altitude [km]
Matlab™ program
Once all the initial data have been collected, was possible to write a simple Matlab™
program that calculated the performance of the compressor, using a simplified model.
The compressor is the component that raise the pressure (and thus the temperature) of
the air swallowed by the boundary layer lip. Pressure must be almost quadruplicated
and density almost triplicated. The expected values of power requirements should lie
between 200 and 350 kW. This huge amount of power is referred to the complete sys-
tem, but it should be noticed that the most power demanding sub–system is the com-
pressor. For this reason an estimation of the power requirements, even if simplified, is
recommendable. In the Table 7.3 a summary of the conditions at the boundary layer lip
is provided.
Condition Value
Since for a perfect gas enthalpy is a function of temperature only, we can write:
h1 − h2i T1 − T2i
ηis, comp = =
h1 − h2 T1 − T2 (7.4)
and:
! " γ−1
p2 γ
T2i = T1 ·
p1 (7.5)
In order to calculate the previous temperatures a little iterative procedure was used
since variable heat coefficient ratio were adopted. Those value do not vary much with
slight adjustment of temperature, but when some relevant temperature gradients are pre-
sent, the calculation can take advantage of the variable heat coefficients ratio.
Once the final temperature was calculated it was possible to determine the work
made and thus the power required to perform the thermodynamic change.
Li = cp (T2 − T1 ) (7.6)
&'$*( &"*"
7.3!38,01379,::;7;,<134=>?@>56
!3!37,01389,::;8;,<13/01;934!6
&'$*) &")"
&'$*% &"("
&'$*! &"'"
&'$* &"#"
&'$(( &"%"
&'$() &"$"
!"" !#" $"" $#" %"" !"" !#" $"" $#" %""
+,-.,/012/,3456 +,-.,/012/,3456
Where Li is the work performed during the compression. Once the work has been
calculated, the power can be estimated.
ṁ · Li
P =
ηc (7.7)
The results of the previous calculation are presented in the table below.
CABIN VENTILATION
Condition Value
The first important consideration is on the temperature. The 334 K value is obviously
too high to be comfortable for the passenger but in case of failure of some part of the
equipment (the heat exchanger), does provide an uncomfortable but survivable envi-
ronment. There is thus the need to dimension an heat exchanger.
A simplified model of heat exchange has been adopted with the main focus to calcu-
late the thermal power to be taken away from the flow, and the mass of air required to
perform this task.
Another Matlab™ program was created to perform those calculations. A thermic
power balance can be used but two variables are still present.
In order to determine the values required the main focus was on the outlet tempera-
ture of the heat exchanger, in the calculation a lower great thermal gradient for this
quantity was preferred. Resulting calculations are presented
The total mass of air required from the ECS and heat exchanger does not reach the
5% of the total mass of air that the boundary layer lip ingests. It appears logical then to
design a diverter rather then a simple throat. On the diverter some blind should be used
as static probes, the difference between static and dynamic probes is the inherent effi-
ciency benefit of the static ones.
In order to verify the calculations that have been made, a comparison between the
trend presented in the Fig. 7.9 and the results found is made.
& −b ε −b #
(
W 1 net = m 1 ⋅ T 11 ⋅ $ P 11 − 1 + ⋅ (ψ − 1) ⋅ ψ − P 11 !)
% ψ "
Giovanni
The details of the calculations are reported MediciA.~The
in appendix Engine Integration
results are on a Flying Wing
summarised in
Figure 16.
550
450
400
350
300
250
200
150
100
50
0
< 10 PAX < 100 PAX 150 PAX 200 PAX 400 PAX 550 PAX
Aircraft Size ( PAX )
The results found are in line with what expected. Even if a little more of air should
Fabienne COUAILLAC MSc. AVD 2006-2007 37
be compressed, for anti ice purposes, the value appears to be reasonable.
Compatibility
In this section will be placed some technical drawings of the design. Note that at the
end of this work in the Appendix II are summarized all drawings and placed in the
proper scale. The multidisciplinary nature of aircraft design does not allow any abstrac-
tion. The designer always has to think in a multidisciplinary way in particular in the
case of study. Particular care should be posed to verify some issues. Some of them have
been calculated previously, like the first two entries while other must be verified:
ECS architecture.
Air Conditioning power requirements (compressor and heat exchanger).
Compatibility with primary structure, in particular ribs and spars.
Compatibility with cabin.
Cabin internal volume.
Compatibility between the S–Duct and the main gear.
Main gear clearance.
As previously described, the primary structure has been designed to have two princi-
pal spars, respectively at 11% and 67% of the chord, and different principal ribs in the
central section, then at 3.9, 7.8 and 12.8 m of wingspan. The main problem in the 3D
model design of the S–Duct is then to place it within the geometrical limits.
In order to give the reader the dimensions of the flying wing FW-300, in the Fig.
7.14 a technical design is presented. In the figure are drawn both the three views and the
3D scaled view. The principal dimensions are reported and the S–Duct frontal area can
be seen.
The geometry has been designed in order to respect the geometrical constrains pro-
vided by the primary structure. In the Fig. 7.15 the primary structure in 3D is presented.
In the next figure is presented the primary structure in a top view, it appears that both
the main gear and the spar/ribs could be interested on intersection or compatibility prob-
lems.
The first vane has not been used for propulsions purposes because in this area a Type
A emergency exit is designed. Since the aft spar is at some 67% of the chord passengers
should run through the remaining 30% to reach the end of the airplane.
A removable structure (like on some tail cones) has been designed to open like a
beak. In the Fig. 7.16 the top view of the FW-300 is presented.
In order to comprehend the three dimensional geometry of the model a section in the
axis of symmetry of the intake is made.
Looking at the section it appears clear that neither the aft spar nor the second rib in-
tersect the geometry of the S–Duct. In the previous picture also the dimensions are re-
ported. Those dimensions respect the size of the Genx engine (Boeing 787 Dreamliner
is equipped with such no–bleed engines). The total length of the airfoil and engine is
bigger than the one of the airfoil at its max chord (in the central section). The choice to
put the engine in such backward position is made both for balance and in order to pro-
vide a long enough S–Duct.
Another concern that emerged during the analysis was that since the central airfoil
has been changed, a new check on the cabin volumes and heights should be performed.
The most important zone is the passenger bay. This zone runs up to 7.8 m on each side
of the wing, providing a considerable 15.6 m of total cabin span. As emerged in previ-
ous chapters this value is higher than the usual architecture airplanes, and some study
were performed in order to check passenger comfort during turbulence flights and ma-
neuvers. The main deck should have an height of more than 1.83 m so that the passen-
ger do not feel to be stuck. The Fig. 7.18 provide the dimension of two section the cen-
tral one and one at some 11 m of span (far away from the 7.8 limit), and testify that the
airplane can handle standard containers in the cargo bay.
A closed up view of the 3D geometry can be used to confirm that there is not incom-
patibility between the S–Duct and the main gear. In order to check this out a cutaway
image has been created and is presented in the Fig. 7.19.
In the picture appears clear that the main gear does not collide with the geometry re-
vealing that the position chosen for it has sense. In particular please note that the engine
blades plane do not cross the main gear in any of its part, so that in case of explosive
failure of the turbofan, the main gear should not be affected by any kind of damage. In
the next figure a check has been made on the gear clearance that is provided by this po-
sition. Remembering that the normative [36] [37] define as a minimum angle of roll on
the gear clearance the value of 8°.
In the next figure another detail of the 3D model is presented, this time all the ge-
ometry is visible with the exception of the wing skin. This particular back view allows
the reader to notice the structure (an hypothesis) that keeps the engine in place. This
structure has been chosen to be close to one rib so that a both longitudinal and lateral
loads could be sustained either by the parallel part of the structure or by the oblique one.
A detailed 3D CATIA model has been drawn with the main purpose of verifying the
integration between the designed engine architecture and the previous studied compo-
nents. A complete compatibility between the two design has been found.
Even the change of the airfoil does not provide any remarkable change in the cabin
and cargo bay volumes.
Matlab™
M
any programs have been developed during this study, each program has
been developed using Matlab™, this well known mathematical suite
offer a sound development kit with many integrated sub functions ready
to use. In this appendix those programs are presented. Since many programs require
subroutines, in this list they will be presented as close as possible to the logic written in
the code, some times it is not possible to seek the algorithm of the program, but each
program written during this work will be reported in the appendix so that it could be
easy enough to find it. In the first section will be presented the programs developed to
build the CAD CFD Integrator suite, within this section the reader can easily find the
subsection that explain which module uses the described program. The other section
deal with the other programs that have been written as help in this analysis. Each pro-
gram is reported with a description, this description is the one written in the file itself
(as help for the user or software developer). Each line of the file that starts with the per-
cent “%” symbol, is a line of comment, that does not affect the code.
Airfoil Editor
VLM: Vortex Lattice Method
Gambit: CFD Preprocessor (create the mesh)
Fluent: CFD Processor
CATIA: CAD Software
In the figure below there is a representation of the sub calls between the routines of
the program.
addTopXAxis.m
processairfoilVLM.m
GambitRev.m processairfoil.m
ro_ratio.m
temperature_ratio.m
MathtoIGES2.m
CADCFD.m
This program stores the graphical information for the CADCFD Integrator suite, it
also gives some help to the user when dealing with a new command.
% ‐‐‐ Executes on button press in Catia.
function Catia_Callback(hObject, eventdata, handles)
% hObject handle to Catia (see GCBO) % ‐‐‐ Executes on button press in VLM.
% eventdata reserved ‐ to be defined in a future version of function VLM_Callback(hObject, eventdata, handles)
MATLAB run VLM\VLMSetup.m
% handles structure with handles and user data (see % hObject handle to VLM (see GCBO)
GUIDATA) % eventdata reserved ‐ to be defined in a future version of
run CATIA\MathtoIGES2.m MATLAB
%pos_size = get(handles.figure1,'Position'); % handles structure with handles and user data (see
%delete(handles.figure1) GUIDATA)
% ‐‐‐ Executes on button press in Quit.
function Quit_Callback(hObject, eventdata, handles) % ‐‐‐ Executes on button press in q4.
% hObject handle to Quit (see GCBO) function q4_Callback(hObject, eventdata, handles)
% eventdata reserved ‐ to be defined in a future version of helpdlg('This program will modify the Geometry of VLM
MATLAB tornado to match the new airfoil profile. You will be asked to
% handles structure with handles and user data (see locate the 2D vertex data of the new airfoil, and a file that
GUIDATA) can be read by Tornado will be saved in its directory. Once
performed all the calculation (you must select a state
pos_size = get(handles.figure1,'Position'); condition), you can use the same suite to plot the results of
delete(handles.figure1) an Alpha Sweep. During the program you may also plot all
the graphics of Tornado.');
% ‐‐‐ Executes on button press in q1. % hObject handle to q4 (see GCBO)
function q1_Callback(hObject, eventdata, handles) % eventdata reserved ‐ to be defined in a future version of
% hObject handle to q1 (see GCBO) MATLAB
% eventdata reserved ‐ to be defined in a future version of % handles structure with handles and user data (see
MATLAB GUIDATA)
% handles structure with handles and user data (see
GUIDATA)
helpdlg('In this program you will asked to insert the % ‐‐‐ Executes on button press in AirfoilEdit.
destination folder of the Journal file that will be used in function AirfoilEdit_Callback(hObject, eventdata, handles)
Gambit. In the same folder will be placed a copy of the % hObject handle to AirfoilEdit (see GCBO)
vertex.txt data of the profile you will choose. Note that % eventdata reserved ‐ to be defined in a future version of
profiles data must start/end with [0 0] or [1 0]'); MATLAB
% handles structure with handles and user data (see
% ‐‐‐ Executes on button press in q3. GUIDATA)
function q3_Callback(hObject, eventdata, handles) %cd airfoiledit\
% hObject handle to q3 (see GCBO) run AirfoilEditor
% eventdata reserved ‐ to be defined in a future version of
MATLAB
AirfoilEditor.m
This program allows the user to perform some operations on an airfoil, it reads .txt or
.dat text files and writes .txt, .dat, .pdf files.
% Airfoil Editor % Edit the above text to modify the response to help
% This program has been developed by Giovanni Medici, allows AirfoilEditor
the user to load an
% airfoil file either in the 2D 3D two columns of y up y down % Last Modified by GUIDE v2.5 15‐May‐2009 18:26:53
format.
% It then allows the user to perform some operation over it % Begin initialization code ‐ DO NOT EDIT
such as: gui_Singleton = 1;
% ‐ thickness gui_State = struct('gui_Name', mfilename, ...
% ‐ camber 'gui_Singleton', gui_Singleton, ...
% ‐ rotation 'gui_OpeningFcn', @AirfoilEditor_OpeningFcn, ...
% ‐ compare two airfoils 'gui_OutputFcn', @AirfoilEditor_OutputFcn, ...
% ‐ print the resulting one 'gui_LayoutFcn', [] , ...
% ‐ print the history of changes performed 'gui_Callback', []);
% ‐ clear all if nargin && ischar(varargin{1})
% ‐ save gui_State.gui_Callback = str2func(varargin{1});
end
function varargout = AirfoilEditor(varargin)
% AIRFOILEDITOR M‐file for AirfoilEditor.fig if nargout
% AIRFOILEDITOR, by itself, creates a new AIRFOILEDITOR [varargout{1:nargout}] = gui_mainfcn(gui_State, varargin{:});
or raises the existing else
% singleton*. gui_mainfcn(gui_State, varargin{:});
% end
% H = AIRFOILEDITOR returns the handle to a new % End initialization code ‐ DO NOT EDIT
AIRFOILEDITOR or the handle to
% the existing singleton*. % ‐‐‐ Executes just before AirfoilEditor is made visible.
% function AirfoilEditor_OpeningFcn(hObject, eventdata,
% handles, varargin)
AIRFOILEDITOR('CALLBACK',hObject,eventData,handles,...) % This function has no output args, see OutputFcn.
calls the local % hObject handle to figure
% function named CALLBACK in AIRFOILEDITOR.M with the % eventdata reserved ‐ to be defined in a future version of
given input arguments. MATLAB
% % handles structure with handles and user data (see
% AIRFOILEDITOR('Property','Value',...) creates a new GUIDATA)
AIRFOILEDITOR or raises the % varargin command line arguments to AirfoilEditor (see
% existing singleton*. Starting from the left, property value VARARGIN)
pairs are
% applied to the GUI before AirfoilEditor_OpeningFunction % Choose default command line output for AirfoilEditor
gets called. An handles.output = hObject;
% unrecognized property name or invalid value makes
property application % Update handles structure
% stop. All inputs are passed to AirfoilEditor_OpeningFcn guidata(hObject, handles);
via varargin. if strcmp(get(hObject,'Visible'),'off')
% plot(rand(5));
% *See GUI Options on GUIDE's Tools menu. Choose "GUI end
allows only one
% instance to run (singleton)". initialize_gui(hObject, handles, false);
%
% See also: GUIDE, GUIDATA, GUIHANDLES % This sets up the initial plot ‐ only do when we are invisible
% so window can get raised using AirfoilEditor.
% Copyright 2002‐2003 The MathWorks, Inc.
ylabel('Y [m]'); selection = questdlg(['Close ' get(handles.figure1,'Name') '?'],...
['Close ' get(handles.figure1,'Name') '...'],...
set(handles.airfoil,'String',num2str(XY)); 'Yes','No','Yes');
if strcmp(selection,'No')
fig = 80; return;
figure(fig) end
cla;
set(fig,'visible', 'off'); delete(handles.figure1)
hold on
grid on
axis equal % ‐‐‐ Executes on selection change in popupmenu1.
plot(XY(:,1),XY(:,2),'Marker','o','MarkerSize',6,'DisplayName',' function popupmenu1_Callback(hObject, eventdata, handles)
Original Airfoil'); % hObject handle to popupmenu1 (see GCBO)
xlabel('X [m]'); % eventdata reserved ‐ to be defined in a future version of
ylabel('Y [m]'); MATLAB
title(['Airfoil ',GeomName]); % handles structure with handles and user data (see
legend('Location','SouthEast'); GUIDATA)
hold off
handles.o.XY=XY; % Hints: contents = get(hObject,'String') returns popupmenu1
handles.o.GeomName=GeomName; contents as cell array
handles.o.GeomPath=GeomPath; % contents{get(hObject,'Value')} returns selected item
handles.o.GeomFullPath=GeomFullPath; from popupmenu1
guidata(handles.figure1, handles);
end % ‐‐‐ Executes during object creation, after setting all
properties.
function popupmenu1_CreateFcn(hObject, eventdata,
% file = uigetfile('*.fig'); handles)
% if ~isequal(file, 0) % hObject handle to popupmenu1 (see GCBO)
% open(file); % eventdata reserved ‐ to be defined in a future version of
% end MATLAB
% handles empty ‐ handles not created until after all
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ CreateFcns called
function PrintMenuItem_Callback(hObject, eventdata,
handles) % Hint: popupmenu controls usually have a white background
% hObject handle to PrintMenuItem (see GCBO) on Windows.
% eventdata reserved ‐ to be defined in a future version of % See ISPC and COMPUTER.
MATLAB if ispc
% handles structure with handles and user data (see set(hObject,'BackgroundColor','white');
GUIDATA) else
% axes(handles.axes1);
% figure(2) set(hObject,'BackgroundColor',get(0,'defaultUicontrolBackgro
% hold on undColor'));
% grid on end
% axis equal
% p2rint2=copyobj(handles.figure1.axes1,gca) set(hObject, 'String', {'plot(rand(5))', 'plot(sin(1:0.01:25))',
% hold off 'bar(1:.5:10)', 'plot(membrane)', 'surf(peaks)'});
if max(size(handles.o.GeomName))==1
warndlg('There is nothing to print');
else
fig=80
printdlg(fig) function airfoil_Callback(hObject, eventdata, handles)
cd(handles.i.pathtome); % hObject handle to airfoil (see GCBO)
end % eventdata reserved ‐ to be defined in a future version of
MATLAB
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ % handles structure with handles and user data (see
function CloseMenuItem_Callback(hObject, eventdata, GUIDATA)
handles)
% hObject handle to CloseMenuItem (see GCBO) % Hints: get(hObject,'String') returns contents of airfoil as text
% eventdata reserved ‐ to be defined in a future version of % str2double(get(hObject,'String')) returns contents of
MATLAB airfoil as a double
% handles structure with handles and user data (see
GUIDATA)
% ‐‐‐ Executes during object creation, after setting all % eventdata reserved ‐ to be defined in a future version of
properties. MATLAB
function airfoil_CreateFcn(hObject, eventdata, handles) % handles structure with handles and user data (see
% hObject handle to airfoil (see GCBO) GUIDATA)
% eventdata reserved ‐ to be defined in a future version of
MATLAB
% handles empty ‐ handles not created until after all % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
CreateFcns called function RotateSMenu_Callback(hObject, eventdata, handles)
% hObject handle to RotateSMenu (see GCBO)
% Hint: edit controls usually have a white background on % eventdata reserved ‐ to be defined in a future version of
Windows. MATLAB
% See ISPC and COMPUTER. % handles structure with handles and user data (see
if ispc && isequal(get(hObject,'BackgroundColor'), GUIDATA)
get(0,'defaultUicontrolBackgroundColor')) if max(size(handles.o.GeomName))==1
set(hObject,'BackgroundColor','white'); warndlg('There is no profile loaded');
end else
XY=handles.o.XY;
if isequal(XY(1,1),0) && isequal(XY(1,2),0)==0
% ‐‐‐ Executes on button press in Clear. som=XY(1,2);
function Clear_Callback(hObject, eventdata, handles) XY(:,2)=XY(:,2)‐som;
% hObject handle to Clear (see GCBO) end
% eventdata reserved ‐ to be defined in a future version of %XY(:,2)=XY(:,2)+0.01015;
MATLAB [maxX I]=max(XY(:,1));
% handles structure with handles and user data (see alpha=(asin(XY(I,2)));
GUIDATA) alphadeg=rad2deg(alpha);
ClearSMeny_Callback(hObject, eventdata, handles) str=['h=helpdlg(''The profile will be rotated of
',num2str(alphadeg),' degrees.'',''Rotate'')'];
eval(str);
function airfoilpath_Callback(hObject, eventdata, handles) uiwait(h);
% hObject handle to airfoilpath (see GCBO) %alpha=deg2rad(10);
% eventdata reserved ‐ to be defined in a future version of eulerrot=[cos(alpha),‐sin(alpha);sin(alpha),cos(alpha)];
MATLAB XY=XY*eulerrot;
% handles structure with handles and user data (see [maxXX II]=max(XY(:,1));
GUIDATA) maxXXX=sqrt(maxXX^2+XY(II,2)^2);
XY=XY./maxXXX;
% Hints: get(hObject,'String') returns contents of airfoilpath as %figure(1)
text axes(handles.axes1);
% str2double(get(hObject,'String')) returns contents of hold on
airfoilpath as a double axis equal
airfoilpath=get(hObject,'String') grid on
handles.airfoilpath = airfoilpath; plot(XY(:,1),XY(:,2),'‐^','MarkerSize',4,'Color','red');
guidata(hObject,handles) hold off
% ‐‐‐ Executes during object creation, after setting all handles.o.XY=XY;
properties. set(handles.airfoil,'String',num2str(XY));
function airfoilpath_CreateFcn(hObject, eventdata, handles) fig = 80;
% hObject handle to airfoilpath (see GCBO) figure(fig)
% eventdata reserved ‐ to be defined in a future version of set(fig,'visible', 'off');
MATLAB hold on
% handles empty ‐ handles not created until after all grid on
CreateFcns called axis equal
if abs(alphadeg)<0.0001
% Hint: edit controls usually have a white background on betadeg=0;
Windows. else
% See ISPC and COMPUTER. betadeg=alphadeg;
if ispc && isequal(get(hObject,'BackgroundColor'), end
get(0,'defaultUicontrolBackgroundColor')) NamePlot=['Rotation ',num2str(betadeg),' ∞'];
set(hObject,'BackgroundColor','white'); plot(XY(:,1),XY(:,2),'‐^','MarkerSize',4,'Color','red','DisplayNam
end e',NamePlot);
legend('Location','SouthEast');
hold off
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
function ToolsMenu_Callback(hObject, eventdata, handles)
% hObject handle to ToolsMenu (see GCBO)
if max(size(handles.o.GeomName))==1
end warndlg('There is no profile loaded');
guidata(handles.figure1, handles); else
XY=handles.o.XY;
XYZ=XY;
defanswer=num2str(max(XYZ(:,2)‐min(XYZ(:,2))));
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ p r o m p t = [ ' M a x T h i c k n e s s i s
function CamberSMenu_Callback(hObject, eventdata, ',num2str(max(XYZ(:,2)‐min(XYZ(:,2)))),' change it if you
handles) want.'];
% hObject handle to CamberSMenu (see GCBO) defanswerc=num2cell(max(XYZ(:,2)‐min(XYZ(:,2))));
% eventdata reserved ‐ to be defined in a future version of answer=inputdlg(prompt,'Thickness',1,cellstr(defanswer));
MATLAB if isempty(answer)==0
% handles structure with handles and user data (see NamePlot=['Thickness ',char(answer)];
GUIDATA) answer=str2num(char(answer));
if max(size(handles.o.GeomName))==1 if isequal(max(XYZ(:,2)‐min(XYZ(:,2))),answer)==0
warndlg('There is no profile loaded'); XYZ(:,2)=XYZ(:,2)./max(XYZ(:,2)‐min(XYZ(:,2))).*answer;
else %XYZ(max(size(XYZ)),1)=XYZ(max(size(XYZ)),1).*0.998;
XY=handles.o.XY; %XYZ(max(size(XYZ)),2)=0;
load('airfoiledit\Liebeck_chamber.mat'); %XYZ(max(size(XYZ))/2,1)=XYZ(max(size(XYZ))/2,1).*0.998;
x=XY(:,1); %XYZ(max(size(XYZ))/2,2)=0;
answer = inputdlg('Choose the amount of reflex camber
[%]:','Camber',1,cellstr('100')); %XYZn=[XYZ(1:max(size(XYZ))/2,:);1,0;XYZ(max(size(XYZ))/2
NamePlot=['Camber ',char(answer),' % ']; +1:max(size(XYZ)),:);1,0];
answer=str2num(char(answer)); %XYZ=XYZn;
y=camber(x).*answer.*0.01; XY=XYZ;
xy=[x,y]; axes(handles.axes1);
YY=XY(:,2)+y; hold on
p=diff(y)'; axis equal
px=diff(x)'; grid on
per=0; plot(XY(:,1),XY(:,2),'‐+','MarkerSize',4,'Color','green');
XY=[XY(:,1),YY]; hold off
axes(handles.axes1); handles.o.XY=XY;
hold on
axis equal set(handles.airfoil,'String',num2str(XY));
grid on
plot(XY(:,1),XY(:,2),'‐d','MarkerSize',4,'Color',[1 0 1]); fig = 80;
hold off figure(fig)
handles.o.XY=XY; set(fig,'visible', 'off');
hold on
fig = 80; grid on
figure(fig) axis equal
set(fig,'visible', 'off'); plot(XY(:,1),XY(:,2),'‐+','MarkerSize',4,'Color','green','DisplayNa
hold on me',NamePlot);
grid on legend('Location','SouthEast');
axis equal hold off
p l o t ( X Y ( : , 1 ) , X Y ( : , 2 ) , ' ‐ d ' , ' M a r ke r S i z e ' , 4 , ' C o l o r ' , [ 1 0
1],'DisplayName',NamePlot);
legend('Location','SouthEast'); end
hold off end
end
guidata(handles.figure1, handles);
set(handles.airfoil,'String',num2str(XY)); % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
end function EditMenu_Callback(hObject, eventdata, handles)
guidata(handles.figure1, handles); % hObject handle to EditMenu (see GCBO)
% eventdata reserved ‐ to be defined in a future version of
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ MATLAB
function ThickSubMenu_Callback(hObject, eventdata, % handles structure with handles and user data (see
handles) GUIDATA)
% hObject handle to ThickSubMenu (see GCBO)
% eventdata reserved ‐ to be defined in a future version of
MATLAB % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
% handles structure with handles and user data (see function ClearSMeny_Callback(hObject, eventdata, handles)
GUIDATA) % hObject handle to ClearSMeny (see GCBO)
% ‐‐‐ Executes on button press in pushbutton8. elseif strcmp(button0,'Cancel')==1
function SaveBtn_Callback(hObject, eventdata, handles) return;
% hObject handle to pushbutton8 (see GCBO) end
% eventdata reserved ‐ to be defined in a future version of
MATLAB
% handles structure with handles and user data (see
GUIDATA) % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
Save_Callback(hObject, eventdata, handles) function Print_PDF_Menu_Callback(hObject, eventdata,
handles)
% hObject handle to Print_PDF_Menu (see GCBO)
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ % eventdata reserved ‐ to be defined in a future version of
function CompareMenu_Callback(hObject, eventdata, handles) MATLAB
% hObject handle to CompareMenu (see GCBO) % handles structure with handles and user data (see
% eventdata reserved ‐ to be defined in a future version of GUIDATA)
MATLAB
% handles structure with handles and user data (see if max(size(handles.o.GeomName))==1
GUIDATA) warndlg('There is nothing to print as PDF');
[GeomName,GeomPath] = uigetfile('*.txt;*.DAT','Load the else
first airfoil Vertex file :','Airfoil'); fig=80;
GeomFullPath=[GeomPath,GeomName]; GeomName=handles.o.GeomName;
if isequal(GeomFullPath,[0 0]) if strfind(GeomName,'.txt')
h=warndlg('Program Aborted without anything changed.'); NameGeom= strrep(GeomName,'.txt','');
return; elseif strfind(GeomName,'.dat')
end NameGeom= strrep(GeomName,'.dat','');
XY=load(GeomFullPath); else
[GeomName1,GeomPath1] = uigetfile('*.txt;*.DAT','Load the NameGeom=GeomName;
second airfoil Vertex file :','Airfoil'); end
VLMSetup.m
This program is the core program of the VLM suite, it contains some parts of the
VLM Tornado™ code, but are clearly referenced.
This program does call some subroutines that are GeometryVLM.m and
addTopXAxis.m (a third party Matlab™ file that allows the user to use multiple X axis
in graphs, for this reason it is not present in this appendix), in the figure below a de-
scription of the subroutines is reported.
addTopXAxis.m
processairfoilVLM.m
uiwait(h)
% UPDATE FULLWING? return;
button = questdlg('Do you want to update the FullWing file end
w i t h t h e n e w C e n t r a l P r o fi l e ? ' , ' U p d a t e V L M disp('Update of VLM Central Profile Complete');
Tornado','No','Yes','Yes'); h=helpdlg('All files Updated Successfully');
if strcmp(button, 'Yes')==1 uiwait(h)
cd(TornadoPath)
cd aircraft\ % +‐‐‐‐‐ PLOT THE RESULTS ‐‐‐‐‐‐‐+
[GeomName,GeomPath] = uigetfile('*.mat','Load the
FWTOT.mat VLM Tornado file:','FWTOT'); elseif strcmp(button0, 'Plot Results')==1
GeomFullPath=[GeomPath,GeomName]; cd(TornadoPath)
if isequal(GeomFullPath,[0 0]) cd output\
h=warndlg('Program Aborted, an airfoil might be changed, [ResuName,ResuPath] = uigetfile('*.mat','Load the Alpha
but no geometry was updated. Will now quit'); Sweep Results of VLM Tornado .mat file:','‐Cx_alpha.mat');
return; ResuFullPath=[ResuPath,ResuName];
end load(ResuFullPath);
load(GeomFullPath); cd(pathtome)
cd(pathtome) fig=89;
geo.foil(1,1,1)=cellstr(GeonName); try
geo.foil(1,1,2)=cellstr(GeonName); get(fig);
save(GeomFullPath,'geo','tor_version') figure(fig);
cd(TornadoPath) hold on
StateFullPath='state\TakeOff.mat'; cool=rand(1,1);
load(StateFullPath) cool=[1,cool,0];
[lattice,ref]=fLattice_setup2(geo,state,0); str=['text(0.02,0.76,''Airspeed: ',num2str(state.AS),'
geometryplot(lattice,geo,ref); [m/s]'',''FontSize'',10,''Color'',cool'',''Units'',''Normalized'')'];
cd(pathtome) eval(str);
% RUN TORNADO SIMULATION?
button4 = questdlg('Do you want to run the Tornado Alpha plot(results.alpha_sweep,squeeze(results.matrix(1,1,:)),'Line
Sweep Simulation?','Run VLM Tornado','No','Yes','Yes'); Width',2,'Color',cool,'Displayname',ResuName);
if strcmp(button4, 'Yes')==1 legend('Location','SouthEast')
results=[]; hold off
load(GeomFullPath) catch
cd(TornadoPath) figure(fig)
cd state\ hold on
axisH = axes
[StateName,StatePath] = uigetfile('*.mat','Load the state plot(results.alpha_sweep,squeeze(results.matrix(1,1,:)),'Line
.mat VLM Tornado file:','TakeOff'); Width',2,'Displayname',ResuName);
StateFullPath=[StatePath,StateName]; addTopXAxis(axisH, 'expression', '180/pi*(argu)','xLabStr',
load(GeomFullPath); 'Angle of Attack \alpha [deg]')
load(StateFullPath) xlabel('Angle of Attack \alpha [rad]');
cd(TornadoPath) ylabel('Lift Coefficient C_L [ ‐ ]');
[lattice,ref]=fLattice_setup2(geo,state,0); t e x t ( 0 . 02 , 0 . 9 , ' C _ L ‐ \ a l p h a f o r t h e f u l l
cd output\ wing','FontSize',14,'Units','Normalized')
[ResuName,ResuPath] = uiputfile('*.*','Save the result str=['text(0.02,0.84,''Airspeed: ',num2str(state.AS),'
analysis as:','Clalphatest'); [m/s]'',''FontSize'',10,''Color'',''blue'',''Units'',''Normalized'')'];
ResuFullPath=[ResuPath,ResuName]; eval(str);
cd(TornadoPath) grid(axisH,'on')
solverloop7(results,2,ResuName,lattice,state,geo,ref); legend('Location','SouthEast')
cd(pathtome) hold off
elseif strcmp(button4,'No')==1 end
h=helpdlg('Geometry Files and Profiles Updated return;
Successfully'); elseif strcmp(button0, 'Cancel')==1
uiwait(h) return;
return; end
end cd(pathtome)
cd ..
elseif strcmp(button, 'No')==1
h=helpdlg('Profiles Updated Successfully, no Geometry
saved');
GeometryVLM.m
This program perform some calculations given an airfoil to design and print the ge-
ometry of the FW-300 Flying Wing.
%Design the geometry of the FW‐300 Airplane GAMMA=2.5; %Diedrhal angle in degrees
function GeometryVLM Diedro=wingspan*tan(deg2rad(GAMMA));
%close all backroot=24.7125;
%clear all TIP=[up64;down64];
clc
TIP(:,3)=wingspan;
TIP(:,1)=TIP(:,1)+backroot;
filestring='airfoilVLM\NACA64up.txt';%input('Load Upperfile TIP(:,2)=TIP(:,2)+Diedro;
(Enter file name): ','s'); %Wing Tip Left
up64=load(filestring); TIPR=[TIP(:,1),TIP(:,2),‐TIP(:,3)];
filestring='airfoilVLM\NACA64low.txt';%filestring=input('Loa
d Lowerfile (Enter file name): ','s'); %Derive Right Root
down64=load(filestring); zeri=zeros(2*max(length(up64)),1);
filestring='airfoilVLM\NACA0012.txt';%filestring=input('Load profile=N0012.*devchord;
Lowerfile (Enter file name): ','s'); VDERIV=[profile(:,1),zeri,profile(:,2)];
N0012=load(filestring); VDERIV(:,1)=VDERIV(:,1)+backroot;
VDERIV(:,2)=VDERIV(:,2)+0.3+Diedro;
[ProfiName,ProfiPath] =processairfoilGEO(0); VDERIV(:,3)=VDERIV(:,3)+wingspan+0.3;
ProfiFullPath=[ProfiPath,ProfiName];
%filestring=input('Load Lowerfile (Enter file name): ','s'); %Derive Left Root
airfoil63=load(ProfiFullPath); LDERIV=VDERIV;
II=max(length(down64)); LDERIV(:,3)=‐LDERIV(:,3);
for i=II:‐1:1
down64a(‐i+II+1,:)=down64(i,:); %Derive Right Tip
end backrootd=8;
wingspand=11.45;
VDERIV2=VDERIV;
VDERIV2(:,1)=VDERIV2(:,1)+backrootd;
%maxchord=input('Give the chord lenght [m]: '); VDERIV2(:,2)=VDERIV2(:,2)+wingspand‐0.6;
maxchord=21.3;
%Derive Left Tip
airfoil64=[up64;down64]; LDERIV2=LDERIV;
area1=trapz(abs(airfoil63(:,1)),abs(airfoil63(:,2)))*maxchord^ LDERIV2(:,1)=LDERIV2(:,1)+backrootd;
2; LDERIV2(:,2)=LDERIV2(:,2)+wingspand‐0.6;
area2=trapz(abs(airfoil64(:,1)),abs(airfoil64(:,2)))*maxchord^
2; %Horizontal Stabilizer Right Root
% up63=up63.*maxchord; backrooth=9.8;
% down63=down63a.*maxchord; HSTAB=N0012.*devchord;
airfoil63=airfoil63; HSTAB(:,3)=wingspan;
lowchord=2.55; HSTAB(:,1)=HSTAB(:,1)+backroot+backrootd;
devchord=3.3; HSTAB(:,2)=HSTAB(:,2)+wingspand+Diedro;
minichord=2.4; HSTAB(:,3)=HSTAB(:,3);
%up64b=up64.*minichord; %Horizontal Stabilizer Left Root
%down64b=down64a.*minichord; LHSTAB=HSTAB;
up64=up64.*lowchord; LHSTAB(:,3)=‐LHSTAB(:,3);
down64=down64a.*lowchord;
%Horizontal Stabilizer Right Tip
%Wing Tip Right backrootl=7;
wingspan=37.5; HBSTABT=N0012.*minichord;
HBSTABT(:,3)=wingspan‐backrooth; NAME(oldlenght+1:oldlenght+max(size(TIP)),1)=cellstr('Witl'
HBSTABT(:,1)=HBSTABT(:,1)+backroot+backrootd+backroot );
l; oldlenght=max(size(NAME));
HBSTABT(:,2)=HBSTABT(:,2)+wingspand+Diedro; NAME(oldlenght+1:oldlenght+max(size(TIPR)),1)=cellstr('Wit
r');
%Horizontal Stabilizer Right Tip oldlenght=max(size(NAME));
LHBSTABT=HBSTABT; NAME(oldlenght+1:oldlenght+max(size(VDERIV)),1)=cellstr('
LHBSTABT(:,3)=‐LHBSTABT(:,3); Derl');
oldlenght=max(size(NAME));
%Wing Root NAME(oldlenght+1:oldlenght+max(size(LDERIV)),1)=cellstr('
ROOT=airfoil63.*maxchord; Derr');
ROOT(:,3)=0; oldlenght=max(size(NAME));
areoalae=trapz(ROOT(:,1),ROOT(:,2)); NAME(oldlenght+1:oldlenght+max(size(VDERIV2)),1)=cellstr
('Detl');
oldlenght=max(size(NAME));
%GRAPHICS NAME(oldlenght+1:oldlenght+max(size(LDERIV2)),1)=cellstr(
b u t t o n 7 = q u e s t d l g ( ' D o y o u w a n t t o p l o t t h e 'Detr');
Geometry?','Plot','Cancel','Plot Geometry','Plot Geometry'); oldlenght=max(size(NAME));
if strcmp(button7, 'Plot Geometry')==1 NAME(oldlenght+1:oldlenght+max(size(HSTAB)),1)=cellstr('
figure(31) HSrr');
hold on oldlenght=max(size(NAME));
axis equal NAME(oldlenght+1:oldlenght+max(size(LHSTAB)),1)=cellstr('
grid on HSrl');
oldlenght=max(size(NAME));
title('\bf\fontsize{12}Vertex Data Matlab'); NAME(oldlenght+1:oldlenght+max(size(HBSTABT)),1)=cellst
plot3(ROOT(:,1),ROOT(:,2),ROOT(:,3)); r('HStr');
oldlenght=max(size(NAME));
plot3(TIP(:,1),TIP(:,2),TIP(:,3)); NAME(oldlenght+1:oldlenght+max(size(LHBSTABT)),1)=cells
plot3(TIPR(:,1),TIPR(:,2),TIPR(:,3)); tr('HStl');
plot3(VDERIV(:,1),VDERIV(:,2),VDERIV(:,3));
plot3(LDERIV(:,1),LDERIV(:,2),LDERIV(:,3)); VORTEX=VERTEX;
plot3(VDERIV2(:,1),VDERIV2(:,2),VDERIV2(:,3)); VORTEX(:,2)=VERTEX(:,3);
plot3(LDERIV2(:,1),LDERIV2(:,2),LDERIV2(:,3)); VORTEX(:,3)=VERTEX(:,2);
plot3(HSTAB(:,1),HSTAB(:,2),HSTAB(:,3)); VERTEX=VORTEX;
plot3(LHSTAB(:,1),LHSTAB(:,2),LHSTAB(:,3)); cd ..
plot3(HBSTABT(:,1),HBSTABT(:,2),HBSTABT(:,3)); cd CATIA\geometry
plot3(LHBSTABT(:,1),LHBSTABT(:,2),LHBSTABT(:,3)); [GeomName,GeomPath] = uiputfile('*.mat','Save the new
hold off Geometry as:','FullWing');
end GeomFullPath=[GeomPath,GeomName];
if isequal(GeomFullPath,[0 0])
% Set all data in mm to export through ElementFromExcel in h=warndlg('Program Aborted without anything
CATIA. changed.Will now quit');
return;
ROOT=ROOT.*1000; end
TIP=TIP.*1000; cd ..
TIPR=TIPR.*1000; cd ..
VDERIV=VDERIV.*1000; cd VLM
LDERIV=LDERIV.*1000; save(GeomFullPath,'VERTEX','NAME');
VDERIV2=VDERIV2.*1000; %Calcolo Baricentro superfici
LDERIV2=LDERIV2.*1000;
HSTAB=HSTAB.*1000; %Semiala
LHSTAB=LHSTAB.*1000; S e m i w i n g = [ 0 0 0 ; 3 7 . 5 2 4 . 7 1 2 5 D i e d r o ; 3 7 . 5
HBSTABT=HBSTABT.*1000; 27.2625 Diedro;0 21.3 0];
LHBSTABT=LHBSTABT.*1000; A=Semiwing(3,2)‐Semiwing(2,2);
B=Semiwing(4,2);
VERTEX=[ROOT;TIP;TIPR;VDERIV;LDERIV;VDERIV2;LDERIV C=Semiwing(4,2);
2;HSTAB;LHSTAB;HBSTABT;LHBSTABT]; H=Semiwing(2,1);
%NAME(1:max(size(VERTEX)),:)='a'; xgw=(2*A*C+A^2+C*B+A*B+B^2)/3/(A+B);
%NAME=cellstr(NAME); ygw=H*(2*A+B)/3/(A+B);
zgw=0
NAME(1:max(size(ROOT)),1)=cellstr('Root'); Sala=(A+B)*H/2;
oldlenght=max(size(NAME));
%Derive
backdev=8; Z_cg=((Sala*zgw)+(Sderive*zgd)+(SHstab*zgh))/(Sala+Sderi
Derive=[37.5 24.7125 Diedro; ve+SHstab);
37.5 28.0125 Diedro; disp('‐‐‐‐‐‐‐‐‐‐‐‐‐');
37.5 28.0125+backdev 11.45+Diedro; disp(' Center of Gravity of the entire airplane:');
37.5 24.7125+backdev 11.45+Diedro]; strng=['disp(''XG: ',num2str(X_cg),' [m]'');'];
Sderive=(Derive(2,2)‐Derive(1,2))*Derive(3,3); eval(strng);
xgd=(Derive(2,2)‐Derive(1,2))/2+4+Derive(1,2); strng=['disp(''YG: ',num2str(Y_cg),' [m]'');'];
ygd=37.5; eval(strng);
zgd=11.45/2; strng=['disp(''ZG: ',num2str(Z_cg),' [m]'');'];
eval(strng);
%Hstab disp('‐‐‐‐‐‐‐‐‐‐‐‐‐');
backhs=7; Lwing=Semiwing;
hspan=9.8 Lwing(:,1)=‐Lwing(:,1);
HSTAB=[37.5 24.7125+backdev 11.45+Diedro; Perimetro=[Semiwing(1,:);Semiwing(2,:);Semiwing(3,:);Semi
37.5 28.0125+backdev 11.45+Diedro; wing(4,:);Lwing(3,:);Lwing(2,:);Lwing(1,:)];
37.5‐hspan 24.7125+backdev+backhs+minichord if strcmp(button7, 'Plot Geometry')==1
11.45+Diedro; figure(31)
37.5‐hspan 24.7125+backdev+backhs 11.45+Diedro]; hold on
%SHstab1=(3.3+44.36‐41.45)*9.5/2; plot3(Perimetro(:,2),Perimetro(:,3),Perimetro(:,1));
SHstab=(HSTAB(2,2)‐HSTAB(1,2)+HSTAB(3,2)‐HSTAB(4,2))* Derive1=Derive;
(HSTAB(1,1)‐HSTAB(3,1))/2; Derive1(5,:)=[37.5 Derive(1,2) Diedro];
AH=HSTAB(3,2)‐HSTAB(4,2); Ldev=Derive1;
BH=HSTAB(2,2)‐HSTAB(1,2); Ldev(:,1)=‐Ldev(:,1);
CH=HSTAB(4,2)‐HSTAB(1,2); plot3(Derive1(:,2),Derive1(:,3),Derive1(:,1));
HH=(HSTAB(1,1)‐HSTAB(3,1)); plot3(Ldev(:,2),Ldev(:,3),Ldev(:,1));
xgh=(2*AH*CH+AH^2+CH*BH+AH*BH+BH^2)/3/(AH+BH)+ Hst1=HSTAB;
HSTAB(1,2); Hst1(5,:)=[37.5 HSTAB(1,2) 11.45+Diedro];
ygh=HH*(2*AH+BH)/3/(AH+BH)+HSTAB(3,1); LHst=Hst1;
zgh=11.45; LHst(:,1)=‐LHst(:,1);
plot3(Hst1(:,2),Hst1(:,3),Hst1(:,1));
%CalcoloCG (lungo X) plot3(LHst(:,2),LHst(:,3),LHst(:,1));
X_cg=((Sala*xgw)+(Sderive*xgd)+(SHstab*xgh))/(Sala+Sderi plot3(X_cg,Z_cg,Y_cg,'*r','Markersize',12)
ve+SHstab); hold off
Y_cg=0; end
processairfoilGEO.m
This program perform some calculations given an airfoil to match the requirement in
the Gambit routine.
function [GeonName,GeonPath] =processairfoilGEO(plott) % Verifies it contains the same points for upper and lower
% This function manage an airfoil to match the requirements side of airfoil
of the Gambit if mod(sizeXY,2)==0
% routine (158 vertex 3D with a particular X vector); pari=1;
% If plott==1 plots in two window the result before and after elseif mod(sizeXY,2)==1
the halfprofile=inputdlg('The file chosen doesn''t contain even
% conversion (should be minded the first time an airfoil is number of vertex please write the index of the first lower
converted) surface point.','Warning',1,halfprofile‐0.5);
return;
%clear all end
%clc
%plott=1; % Streches the profile to fit the desired chord
maxchord=21.3;
%maxchord=input('Put the max chord length [m]: '); % Creates a Spline with the fit command, in order to export
uniform 158
load('X'); % vertex files with the same X coordinate
processairfoilVLM.m
This program perform some calculations given an airfoil to match the requirement in
the Gambit routine.
function [GeonName,GeonPath] =processairfoilVLM(plott)
% This function manage an airfoil to match the requirements GeonPath=[TornadoPath,'aircraft\airfoil\'];
of the Gambit GeonFullPath=[GeonPath,GeonName];
% routine (158 vertex 3D with a particular X vector);
% If plott==1 plots in two window the result before and after XYZ=XY;
the h = waitbar(0,'Please wait...');
% conversion (should be minded the first time an airfoil is
converted)
fid=fopen(GeonFullPath,'w');
%clear all fprintf(fid,'%% %s\n',GeonName);
%clc fprintf(fid,'%d %d\n',max(size(X))/2,max(size(X))/2);
%plott=0; fprintf(fid,'%% UPPER SURFACE\n');
maxchord=21.3;
%maxchord=input('Put the max chord length [m]: '); for l=1:1:80
fprintf(fid,'%7.6f %7.6f\n',XY(l,1),XY(l,2));
load('X'); waitbar(l/max(size(X)))
load('Tornadodir') end
fprintf(fid,'%% LOWER SURFACE\n');
[GeomName,GeomPath] = uigetfile('*.txt;*.DAT','Load the for l=1:1:80
Airfoil Vertex file for VLM:','Airfoil'); fprintf(fid,'%7.6f %7.6f\n',XY(l,1),XY(l,2));
GeomFullPath=[GeomPath,GeomName]; waitbar((80+l)/max(size(X)))
if isequal(GeomFullPath,[0 0]) end
GeonName=0; fclose(fid);
GeonPath=0; close(h)
%h=warndlg('Program Aborted without anything
changed.Will now quit');
return;
end
XY=load(GeomFullPath,'‐ascii'); %msgbox('The profile was unchanged.','£D Vertex','warn');
if plott==1 str=['h=helpdlg(''The file ',GeonName,' has been saved in the
figure:1 directory ',GeonFullPath,''',''File Saved'')'];
hold on eval(str)
axis equal uiwait(h)
grid on else
title('Original Airfoil Spline and Points'); sizeXY=max(size(XY));
plot(XY(:,1),XY(:,2)); halfprofile=sizeXY/2;
hold off
end % Verifies it contains the same points for upper and lower
sizeXY=max(size(XY)); side of airfoil
msizeXY=min(size(XY)); if mod(sizeXY,2)==0
pari=1;
%This check if the profile needs to be changed, if not, returns elseif mod(sizeXY,2)==1
the correct halfprofile=inputdlg('The file chosen doesn''t contain even
%file number of vertex please write the index of the first lower
surface point.','Warning',1,halfprofile‐0.5);
if msizeXY==2 && isequal(XY(:,1),X(:,1))==1 return;
%GeonName = inputdlg('Save the modified airfoil file end
as:','Save as:',1,'Airfoil.DAT');
GeonName=GeomName; % Streches the profile to fit the desired chord
if max(XY(:,1))==1.0
X=X./maxchord; h = waitbar(0,'Please wait...');
XYup=XY(1:halfprofile,:);
XYdown=XY(halfprofile+1:sizeXY,:);
sup=fit(XYup(:,1),XYup(:,2),'spline'); fid=fopen(GeonFullPath,'w');
sdown=fit(XYdown(:,1),XYdown(:,2),'spline'); fprintf(fid,'%% %s\n',GeonName);
XYupper=[X(1:80).*maxchord, sup(X(1:80).*maxchord)]; fprintf(fid,'%d %d\n',max(size(X))/2,max(size(X))/2);
X Y l o w e r = [ X ( 8 1 : 1 6 0 ) . * m a x c h o r d , fprintf(fid,'%% UPPER SURFACE\n');
sdown(X(81:160).*maxchord)];
XYnew=[X(1:80).*maxchord, sup(X(1:80).*maxchord) ; ... for l=1:1:80
X(81:160).*maxchord,sdown(X(81:160).*maxchord)]; fprintf(fid,'%7.6f %7.6f\n',XYupper(l,1),XYupper(l,2));
if plott==1 waitbar(l/max(size(X)))
figure:2 end
hold on fprintf(fid,'%% LOWER SURFACE\n');
axis equal for l=1:1:80
grid on fprintf(fid,'%7.6f %7.6f\n',XYlower(l,1),XYlower(l,2));
title('Modified Airfoil Spline and Points'); waitbar((80+l)/max(size(X)))
plot(X(1:80).*maxchord,sup(X(1:80).*maxchord),'‐o','Marker end
Size',4) fclose(fid);
plot(X(81:160).*maxchord,sdown(X(81:160).*maxchord),'‐o',' close(h)
MarkerSize',4) str=['h=helpdlg(''The file ',GeonName,' has been saved in the
hold off directory ',GeonFullPath,''',''File Saved'')'];
eval(str)
end uiwait(h)
% % Modifies the profile so that there are not double vertex end
defname=cellstr('Airfoil.DAT');
GambitRev.m
This program perform some calculations given an airfoil to match the requirement in
the Gambit routine.
function [GeonName,GeonPath] =processairfoilVLM(plott)
% This function manage an airfoil to match the requirements GeonPath=[TornadoPath,'aircraft\airfoil\'];
of the Gambit GeonFullPath=[GeonPath,GeonName];
% routine (158 vertex 3D with a particular X vector);
% If plott==1 plots in two window the result before and after XYZ=XY;
the h = waitbar(0,'Please wait...');
% conversion (should be minded the first time an airfoil is
converted)
fid=fopen(GeonFullPath,'w');
%clear all fprintf(fid,'%% %s\n',GeonName);
%clc fprintf(fid,'%d %d\n',max(size(X))/2,max(size(X))/2);
%plott=0; fprintf(fid,'%% UPPER SURFACE\n');
maxchord=21.3;
%maxchord=input('Put the max chord length [m]: '); for l=1:1:80
fprintf(fid,'%7.6f %7.6f\n',XY(l,1),XY(l,2));
load('X'); waitbar(l/max(size(X)))
load('Tornadodir') end
fprintf(fid,'%% LOWER SURFACE\n');
[GeomName,GeomPath] = uigetfile('*.txt;*.DAT','Load the for l=1:1:80
Airfoil Vertex file for VLM:','Airfoil'); fprintf(fid,'%7.6f %7.6f\n',XY(l,1),XY(l,2));
GeomFullPath=[GeomPath,GeomName]; waitbar((80+l)/max(size(X)))
if isequal(GeomFullPath,[0 0]) end
GeonName=0; fclose(fid);
GeonPath=0; close(h)
%h=warndlg('Program Aborted without anything
changed.Will now quit');
return;
end
XY=load(GeomFullPath,'‐ascii'); %msgbox('The profile was unchanged.','£D Vertex','warn');
if plott==1 str=['h=helpdlg(''The file ',GeonName,' has been saved in the
figure:1 directory ',GeonFullPath,''',''File Saved'')'];
hold on eval(str)
axis equal uiwait(h)
grid on else
title('Original Airfoil Spline and Points'); sizeXY=max(size(XY));
plot(XY(:,1),XY(:,2)); halfprofile=sizeXY/2;
hold off
end % Verifies it contains the same points for upper and lower
sizeXY=max(size(XY)); side of airfoil
msizeXY=min(size(XY)); if mod(sizeXY,2)==0
pari=1;
%This check if the profile needs to be changed, if not, returns elseif mod(sizeXY,2)==1
the correct halfprofile=inputdlg('The file chosen doesn''t contain even
%file number of vertex please write the index of the first lower
surface point.','Warning',1,halfprofile‐0.5);
if msizeXY==2 && isequal(XY(:,1),X(:,1))==1 return;
%GeonName = inputdlg('Save the modified airfoil file end
as:','Save as:',1,'Airfoil.DAT');
GeonName=GeomName; % Streches the profile to fit the desired chord
if max(XY(:,1))==1.0
X=X./maxchord; h = waitbar(0,'Please wait...');
XYup=XY(1:halfprofile,:);
XYdown=XY(halfprofile+1:sizeXY,:);
sup=fit(XYup(:,1),XYup(:,2),'spline'); fid=fopen(GeonFullPath,'w');
sdown=fit(XYdown(:,1),XYdown(:,2),'spline'); fprintf(fid,'%% %s\n',GeonName);
XYupper=[X(1:80).*maxchord, sup(X(1:80).*maxchord)]; fprintf(fid,'%d %d\n',max(size(X))/2,max(size(X))/2);
X Y l o w e r = [ X ( 8 1 : 1 6 0 ) . * m a x c h o r d , fprintf(fid,'%% UPPER SURFACE\n');
sdown(X(81:160).*maxchord)];
XYnew=[X(1:80).*maxchord, sup(X(1:80).*maxchord) ; ... for l=1:1:80
X(81:160).*maxchord,sdown(X(81:160).*maxchord)]; fprintf(fid,'%7.6f %7.6f\n',XYupper(l,1),XYupper(l,2));
if plott==1 waitbar(l/max(size(X)))
figure:2 end
hold on fprintf(fid,'%% LOWER SURFACE\n');
axis equal for l=1:1:80
grid on fprintf(fid,'%7.6f %7.6f\n',XYlower(l,1),XYlower(l,2));
title('Modified Airfoil Spline and Points'); waitbar((80+l)/max(size(X)))
plot(X(1:80).*maxchord,sup(X(1:80).*maxchord),'‐o','Marker end
Size',4) fclose(fid);
plot(X(81:160).*maxchord,sdown(X(81:160).*maxchord),'‐o',' close(h)
MarkerSize',4) str=['h=helpdlg(''The file ',GeonName,' has been saved in the
hold off directory ',GeonFullPath,''',''File Saved'')'];
eval(str)
end uiwait(h)
% % Modifies the profile so that there are not double vertex end
defname=cellstr('Airfoil.DAT');
processairfoil.m
This program perform some calculations given an airfoil to match the requirement in
the Gambit routine.
function [GeonName,GeonPath] =processairfoil(plott)
% This function manage an airfoil to match the requirements load('XX');
of the Gambit
% routine (158 vertex 3D with a particular X vector); h=helpdlg('Step Two: Choose the airfoil profile for the root of
% If plott==1 plots in two window the result before and after the wing');
the uiwait(h)
% conversion (should be minded the first time an airfoil is pathtome=pwd;
converted) cd([pathtome,'\airfoils\'])
[GeomName,GeomPath] = uigetfile('*.txt;*.DAT','Load the
%clear all Airfoil Vertex file:','Airfoil');
%clc GeomFullPath=[GeomPath,GeomName];
%plott=1; cd(pathtome)
maxchord=21.3;
%maxchord=input('Put the max chord length [m]: '); if isequal(GeomFullPath,[0 0])
GeonName=0; XYup=XY(1:halfprofile,:);
GeonPath=0; XYdown=XY(halfprofile+1:sizeXY,:);
% h=warndlg('Program Aborted without anything sup=fit(XYup(:,1),XYup(:,2),'spline');
changed.Will now quit'); sdown=fit(XYdown(:,1),XYdown(:,2),'spline');
return; XYnew=[XX(1:80).*maxchord, sup(XX(1:80).*maxchord) ;
end ...
XY=load(GeomFullPath,'‐ascii'); XX(81:158).*maxchord,sdown(XX(81:158).*maxchord)];
% XYnew is the new profile with 158 points
sizeXY=max(size(XY)); XYold=[XY(1:halfprofile,1),sup(XY(1:halfprofile,1)); ...
msizeXY=min(size(XY));
XY(halfprofile+2:max(size(XY))‐1,1),sdown(XY(halfprofile+2:
%This check if the profile needs to be changed, if not, returns max(size(XY))‐1,1))];
the correct % XYold is the old profile with the corrected points
%file if plott==1
figure:2
if msizeXY==3 %%&& isequal(XY(:,1),XX(:,1))==1 hold on
GeonName=GeomName; axis equal
GeonPath=GeomPath; grid on
GeonFullPath=GeomFullPath; title('Modified Airfoil Spline and Points');
XYZ=XY; plot(XX(1:80).*maxchord,sup(XX(1:80).*maxchord),'‐o','Mar
% save(GeonFullPath,'XYZ','‐ascii'); kerSize',4)
h=msgbox('The profile was left unchanged.','3D plot(XX(81:158).*maxchord,sdown(XX(81:158).*maxchord),'‐
Vertex','warn'); o','MarkerSize',4)
uiwait(h) hold off
else
if plott==1 end
figure:1 % % Modifies the profile so that there are not double vertex
hold on % XYn=[XY(1:halfprofile,:);XY(halfprofile+2:sizeXY‐1,:)];
axis equal cd([pathtome,'\airfoils\'])
grid on [GeonName,GeonPath] = uiputfile('*.txt','Save the modified
title('Original Airfoil Spline and Points'); airfoil file as:','Airfoil');
plot(XY(:,1),XY(:,2)); cd(pathtome);
hold off GeonFullPath=[GeonPath,GeonName];
end if isequal(GeonFullPath,[0 0])
sizeXY=max(size(XY)); % h=warndlg('Program Aborted without anything
halfprofile=sizeXY/2; changed.Will now quit');
GeonName=0;
% Verifies it contains the same points for upper and lower GeonPath=0;
side of airfoil return;
if mod(sizeXY,2)==0 end
pari=1; % Adds a rows of 0 (the z row);
elseif mod(sizeXY,2)==1 if min(size(XYnew))==2
halfprofile=inputdlg('The file chosen doesn''t contain even XYZ=XYnew;
number of vertex please write the index of the first lower XYZold=XYold;
surface point.','Warning',1,halfprofile‐0.5); XYZ(:,3)=0;
return; XYZold(:,3)=0;
end h=msgbox('A 2D Vertex airfoil file was found; added a third
column of 0.','2D Vertex to 3D','warn');
% Streches the profile to fit the desired chord uiwait(h)
if max(XY(:,1))==1.0 end
XY=XY.*maxchord;
end
XX=XX./maxchord;
Fluent.m
This program allows the user to enter the basic data needed for the Fluent processor
routine. In the Fig. I.3 there is a detailed view of the subcalls that starts from the
Fluent.m program.
ro_ratio.m
temperature_ratio.m
function varargout = Fluent(varargin) % instance to run (singleton)".
% FLUENT M‐file for Fluent.fig %
% FLUENT, by itself, creates a new FLUENT or raises the % See also: GUIDE, GUIDATA, GUIHANDLES
existing
% singleton*. % Copyright 2002‐2003 The MathWorks, Inc.
%
% H = FLUENT returns the handle to a new FLUENT or the % Edit the above text to modify the response to help Fluent
handle to
% the existing singleton*. % Last Modified by GUIDE v2.5 08‐Apr‐2009 13:50:46
%
% FLUENT('CALLBACK',hObject,eventData,handles,...) % Begin initialization code ‐ DO NOT EDIT
calls the local gui_Singleton = 1;
% function named CALLBACK in FLUENT.M with the given gui_State = struct('gui_Name', mfilename, ...
input arguments. 'gui_Singleton', gui_Singleton, ...
% 'gui_OpeningFcn', @Fluent_OpeningFcn, ...
% FLUENT('Property','Value',...) creates a new FLUENT or 'gui_OutputFcn', @Fluent_OutputFcn, ...
raises the 'gui_LayoutFcn', [] , ...
% existing singleton*. Starting from the left, property 'gui_Callback', []);
value pairs are if nargin && ischar(varargin{1})
% applied to the GUI before Fluent_OpeningFunction gets gui_State.gui_Callback = str2func(varargin{1});
called. An end
% unrecognized property name or invalid value makes
property application if nargout
% stop. All inputs are passed to Fluent_OpeningFcn via [varargout{1:nargout}] = gui_mainfcn(gui_State,
varargin. varargin{:});
% else
% *See GUI Options on GUIDE's Tools menu. Choose "GUI gui_mainfcn(gui_State, varargin{:});
allows only one end
% End initialization code ‐ DO NOT EDIT errordlg('Altitude Input must be a number','Error');
end
% ‐‐‐ Executes just before Fluent is made visible. % Save the new altitude value
function Fluent_OpeningFcn(hObject, eventdata, handles, handles.metricdata.altitude = altitude;
varargin) guidata(hObject,handles)
% This function has no output args, see OutputFcn.
% hObject handle to figure
% eventdata reserved ‐ to be defined in a future version of % ‐‐‐ Executes during object creation, after setting all
MATLAB properties.
% handles structure with handles and user data (see function altitude_CreateFcn(hObject, eventdata, handles)
GUIDATA) % hObject handle to altitude (see GCBO)
% varargin command line arguments to Fluent (see % eventdata reserved ‐ to be defined in a future version of
VARARGIN) MATLAB
% handles empty ‐ handles not created until after all
% Choose default command line output for Fluent CreateFcns called
handles.output = hObject;
% Hint: edit controls usually have a white background on
% Update handles structure Windows.
guidata(hObject, handles); % See ISPC and COMPUTER.
initialize_gui(hObject, handles, false); if ispc && isequal(get(hObject,'BackgroundColor'),
% UIWAIT makes Fluent wait for user response (see get(0,'defaultUicontrolBackgroundColor'))
UIRESUME) set(hObject,'BackgroundColor','white');
% uiwait(handles.figure1); end
%varargout{1} = s; % Save the new altitude value
handles.metricdata.alpha = alpha;
varargout{1} = handles.output; guidata(hObject,handles)
end function OK_Callback(hObject, eventdata, handles)
% hObject handle to pushbutton1 (see GCBO)
% eventdata reserved ‐ to be defined in a future version of
MATLAB
function mach_Callback(hObject, eventdata, handles) % handles structure with handles and user data (see
% hObject handle to mach (see GCBO) GUIDATA)
% eventdata reserved ‐ to be defined in a future version of
MATLAB try
% handles structure with handles and user data (see M=handles.metricdata.mach;
GUIDATA) alpha=handles.metricdata.alpha;
altitudine=handles.metricdata.altitude;
% Hints: get(hObject,'String') returns contents of mach as catch
text %errordlg('Can''t read data, may be the user has not changed
% str2double(get(hObject,'String')) returns contents of any field. Will now quit.','Error no data');
mach as a double pos_size = get(handles.figure1,'Position');
mach = str2double(get(hObject, 'String')); delete(handles.figure1)
if isnan(mach) return
set(hObject, 'String', 0); end
errordlg('Mach Input must be a number','Error');
end save('boundary.mat','M','altitudine','alpha','‐mat');
% Save the new altitude value pos_size = get(handles.figure1,'Position');
handles.metricdata.mach = mach; delete(handles.figure1)
guidata(hObject,handles)
FluentSet.m
When the user enter the boundary conditions, the program performs some calcula-
tions and then writes the journal files.
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ t e x t ‐ e n t r y Journal{88,:}=strng;
"Materials*Table1*Frame1*Table1*TextEntry1(Name)" s t r n g = [ ' ( c x‐ g u i ‐ d o c x‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t " F o r c e
"air',num2str(altitudine),'")']; M o n i t o r s * F r a m e 1 * F r a m e 3 * F r a m e 1 ( F o r c e
Journal{8,:}=strng; Vector)*RealEntry1(X)" ''( ',num2str(alphatrimsin),'))'];
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t Journal{93,:}=strng;
"pressure‐far‐field‐4‐1*Frame4*Frame3(Momentum)*Table1 s t r n g = [ ' ( c x‐ g u i ‐ d o c x‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t " F o r c e
*Table10*RealEntry2(Mach Number)" ''(',num2str(M),'))']; M o n i t o r s * F r a m e 1 * F r a m e 3 * F r a m e 1 ( F o r c e
Journal{48,:}=strng; Vector)*RealEntry2(Y)" ''( ',num2str(alphatrim),'))'];
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t Journal{94,:}=strng;
"pressure‐far‐field‐4‐1*Frame4*Frame3(Momentum)*Table1 strng=['(cx‐gui‐do cx‐set‐real‐entry‐list "Reference
*Table11*RealEntry2(X‐Component of Flow Direction)" ''( Values*Frame2(Reference Values)*RealEntry1(Area)" ''(
',num2str(alphatrim),'))']; ',num2str(airfoil),'))'];
Journal{50,:}=strng; Journal{115,:}=strng;
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t strng=['(cx‐gui‐do cx‐set‐real‐entry‐list "Reference
"pressure‐far‐field‐4‐1*Frame4*Frame3(Momentum)*Table1 Values*Frame2(Reference Values)*RealEntry5(Length)" ''(
*Table12*RealEntry2(Y‐Component of Flow Direction)" ''( ',num2str(maxchord),'))'];
',num2str(alphatrimsin),'))']; Journal{116,:}=strng;
Journal{51,:}=strng; strng=['(cx‐gui‐do cx‐set‐real‐entry‐list "Reference
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t Values*Frame2(Reference Values)*RealEntry3(Depth)" ''(
"pressure‐far‐field‐4‐1*Frame4*Frame4(Thermal)*Table1*T ',num2str(depthairfoil),'))'];
able1*RealEntry2(Temperature)" ''( ',num2str(temperat),'))']; Journal{117,:}=strng;
Journal{53,:}=strng; s t r n g = [ ' ( c x‐ g u i ‐ d o c x‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t " F o r c e
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t Reports*Frame1*Frame2*Frame1(Vector)*RealEntry1(X)" ''(
"pressure‐outlet‐3‐1*Frame4*Frame3(Momentum)*Table1* ',num2str(alphatrim),'))'];
Table9*RealEntry2(Gauge Pressure)" ''( ',num2str(pres),'))']; Journal{128,:}=strng;
Journal{59,:}=strng; s t r n g = [ ' ( c x‐ g u i ‐ d o c x‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t " F o r c e
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t Reports*Frame1*Frame2*Frame1(Vector)*RealEntry2(Y)" ''(
"pressure‐outlet‐3‐1*Frame4*Frame4(Thermal)*Table1*Tabl ',num2str(alphatrimsin),'))'];
e 1 * Re a l E n t r y 2 ( B a c k fl o w To t a l Te m p e r a t u r e) " ' ' ( Journal{129,:}=strng;
',num2str(Ttot),'))'];
Journal{62,:}=strng; JournalName= strrep(MeshName, '.msh', '');
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t
"pressure‐outlet‐3‐1*Frame4*Frame3(Momentum)*Table1* JournalName_jou=[MeshPath,JournalName,'FLUENT.jou'];
Table5*RealEntry2(X‐Component of Flow Direction)" ''(
',num2str(alphatrim),'))']; fid = fopen(JournalName_jou,'w');
Journal{67,:}=strng; [M,N] = size(Journal);
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t for (i=1:M)
"pressure‐outlet‐3‐1*Frame4*Frame3(Momentum)*Table1* for (j=1:N)
Table6*RealEntry2(Y‐Component of Flow Direction)" ''( fprintf(fid, '%s\t',Journal{i,j});
',num2str(alphatrimsin),'))']; end
Journal{68,:}=strng; waitbar(i/M)
s t r n g = [ ' ( c x ‐ g u i ‐ d o c x ‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t fprintf(fid, '\n');
"pressure‐far‐field‐4‐1*Frame4*Frame3(Momentum)*Table1 end
*Table9*RealEntry2(Gauge Pressure)" ''( ',num2str(pres),'))']; fclose(fid);
Journal{73,:}=strng; close(h)
s t r n g = [ ' ( c x‐ g u i ‐ d o c x‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t " F o r c e hh=helpdlg('File Saved Successfully.');
M o n i t o r s * F r a m e 1 * F r a m e 3 * F r a m e 1 ( F o r c e uiwait(hh)
Vector)*RealEntry1(X)" ''( ',num2str(alphatrim),'))'];
Journal{87,:}=strng;
s t r n g = [ ' ( c x‐ g u i ‐ d o c x‐ s e t ‐ r e a l ‐ e n t r y ‐ l i s t " F o r c e
M o n i t o r s * F r a m e 1 * F r a m e 3 * F r a m e 1 ( F o r c e
Vector)*RealEntry2(Y)" ''( ',num2str(alphatrimsin),'))'];
pressure_ratio.m
This program returns the pressure ratio given altitude, as referred by international
standard atmosphere.
ro_ratio.m
This program returns the density ratio given altitude, as referred by international
standard atmosphere.
temperature_ratio.m
This program returns the temperature ratio given altitude, as referred by international
standard atmosphere.
MathtoIGES2.m
This program perform all the calculations calculations for the CATIA routine.
% This Routine ask the user to load a File of Vertex and load(GeomFullPath);
Names (must have cd ..
% the same number of rows), and then ask where to save the %load('VERTEX');
.igs file that
% will be imported in CATIA if max(size(NAME))~=max(size(VERTEX))
errordlg('Both Name and Vertex Data must have the same
clear all number of Rows','Error');
%clc break;
format long end
hh=helpdlg('Step One: Load the Geometry file saved before %function MathtoIGES=
(in point 1), that will be added to the CATIA model'); hh=helpdlg('Step Two: Choose where and how save the .IGS
uiwait(hh) file that will be imported in CAITA');
cd geometry\ uiwait(hh)
[GeomName,GeomPath] = uigetfile('*.mat','Load the [IGesName,IGesPath] = uiputfile('*.igs','Save the Geometry
Geometry file:','FullWing'); file as:','Wing');
GeomFullPath=[GeomPath,GeomName]; IGesFullPath=[IGesPath,IGesName];
if isequal(GeomFullPath,[0 0])
h=warndlg('Program Aborted without anything if isequal(IGesFullPath,[0 0])
changed.Will now quit'); h=warndlg('Program Aborted without anything
return; changed.Will now quit');
end return;
end
h = waitbar(0,'Please wait...'); iii=1;
BaseName='BaseWing.CATPart'; [M,N] = size(VERTEX);
BasePath=[IGesPath,BaseName]; Rootold=NAME(1,1);
copyfile(BaseName,BasePath);
%copyfile('Macro.catvbs',IGesPath);
Macro=textread('Macro.catvbs', '%s','delimiter','\n'); for (i=1:M)
s t r n g = [ ' S e t d o c u m e n t 1 = waitbar(1/2+i/M/4)
documents1.Open("',IGesFullPath,'")']; ii=i*2‐1;
Macro{7,:}=strng; if strcmp(NAME(i,1),Rootold)==0
s t r n g = [ ' S e t p a r t D o c u m e n t 2 = iii=1;
documents1.Open("',[IGesPath,'BaseWing.CATPart'],'")']; end
Macro{981,:}=strng;
fod= fopen([IGesPath,'Macro.catvbs'],'w'); fprintf(fid, ' 116 %7.d 0 0 10000 0 0
[MMM,NNN] = size(Macro); 000000001D %6.d\n',i,ii);
for (i=1:MMM)
for (j=1:NNN) fprintf(fid, ' 116 0 0 1 0 %19s.%03d 0D
fprintf(fod, '%s\t',Macro{i,j}); %6.d\n',char(NAME(i,1)),iii,ii+1);
end iii=iii+1;
waitbar(i/MMM/2) Rootold=NAME(i);
fprintf(fod, '\n'); end
end
fclose(fod); for (i=1:M)
fid = fopen(IGesFullPath,'w'); waitbar(1/2+(M+i)/M/4)
ii=i*2‐1;
f p r i n t f ( fi d , ' S T A R T R E C O R D G O H E R E . fprintf(fid,'116,%010.3f,%010.3f,%010.3f,0,0,0; %28dP
S 1\n'); %6d\n',VERTEX(i,1),VERTEX(i,2),VERTEX(i,3),ii,i);
f p r i n t f ( fi d , ' 1 H , , 1 H ; , 2 0 H C N E X T ‐ I G E S end
PRODUCT,5H1.igs,44HIBM CATIA IGES ‐ CATIA VersioG
1\n'); fprintf(fid,'S 1G 4D %6dP %6d T
fprintf(fid,'n 5 Release 17 ,27HCATIA Version 5 Release 17 1',ii+1,i);
,32,75,6,75,15,5HPart2, G 2\n'); fclose(fid);
fprintf(fid,'1.0,2,2HMM,1000,1.0,15H20090409.150950,0.001, close(h)
10000.0,10HGiovanniPC,8HGIOG 3\n'); hh=helpdlg('File Saved Successfully.');
f p r i n t f ( fi d , ' VA N N I , 1 1 , 0 , 1 5 H 2 0 0 9 0 4 0 9 . 1 5 0 9 5 0 , ; uiwait(hh)
G 4\n');
fprintf(fid,'Matlab2IGSVertex by Giovanni Medici
G 5\n');
Other Programs
Many other programs have been developed in order to automatize and fasten some
operations within the study.
PlotBlackAndWhite.m
This program reads an airfoil and then plots the shape in black and white with a
transparent background, it has been used during the presentations and with table design
purposes
liebeck.m
This program was used in order to build the geometry of the Liebeck airfoil pre-
sented in [34], a fit and smooth operation was then manually added.
clear all %SASB=solve('0.8*(0.8‐1)*(0.8‐b)*(0.8‐a)=‐0.00629565','0.3
clc 75*(0.375‐1)*(0.375‐b)*(0.375‐a)=0.0121507','b');
format long XFull=LFullCamber(:,1);
[GeomName,GeomPath] = uigetfile('*.mat','Load the airfoil YFull=LFullCamber(:,2);
Vertex file :','Airfoil'); WeightsFull=zeros(max(size(XFull)),1);
GeomFullPath=[GeomPath,GeomName]; WeightsFull=WeightsFull+0.3;
if isequal(GeomFullPath,[0 0]) WeightsFull(1,1)=1;
h=helpdlg('Program Aborted without anything WeightsFull(max(size(XFull)),1)=1;
changed.Will now quit'); WeightsFull(55,1)=1;
return; [val II]=max(YFull);
end [vall III]=min(YFull);
load(GeomFullPath); WeightsFull(II,1)=0.8;
%GeomPath_c=Geompath WeightsFull(III,1)=0.8;
load([GeomPath,'Liebeck_chamber.mat']); XFS=smoothweight.x;
%plot(XYLiebeck(:,1),XYLiebeck(:,2)) YFS=smoothweight.y;
figure(1) XFS(1,1)=0;
hold on XFS(max(size(XFS)),1)=1;
plot(LiebeckChamber(:,1),LiebeckChamber(:,2)) YFS(1,1)=0;
[XYy IX]=sort(XYLiebeck,2); YFS(max(size(YFS)),1)=0;
Xy=XYLiebeck(IX); %solve('x*(x‐1)*(x‐aa)(x‐a)=0','')
Xup=XYLiebeck(1:82,1); %diff(func)
Yup=XYLiebeck(1:82,2); %a=2.048+1.92*(‐1‐aa)+1.6*(aa)/(1.6*(1+aa)‐1.92‐aa);
Xdown=XYLiebeck(83:164,1); x=0:0.025:1;
Ydown=XYLiebeck(83:164,2); y=x.*(x‐1).*(x‐aa).*(x‐double(S));
y2=x.*(x‐1).*(x‐aa).*(x‐double(SA));
% S=solve(P)
%plot(Xy,XYy) plot(x,y,'green');
% cd airfoils plot(x,spline(x),'black');
% save('Liebeck.txt','XYLiebeck','‐ascii'); plot(x,y2,'Color',[0.6 0.2 0.2]);
% cd .. plot(LFullCamber(:,1),LFullCamber(:,2),'o','MarkerSize',6)
x=0:0.025:1; hold off
hold on figure(2)
grid on hold on
%axis equal grid on
axis([0 1 ‐0.02 0.02]) axis([0 1 ‐0.02 0.02])
plot(x,fittedmodel1(x),'red');
aa=2/3; plot(LFullCamber(:,1),LFullCamber(:,2),'o','MarkerSize',6)
sym a y3=x.*(x‐1).*(x‐double(SASA.a(1))).*(x‐double(SASA.b(1)));
sym b y4=x.*(x‐1).*(x‐double(SASA.a(2))).*(x‐double(SASA.b(2)));
x=0.8; y5=shape(x);
S=solve('4*0.8^3+3*0.8^2*(‐1‐2/3‐a)+2*0.8*(a+2/3+2/3*a)‐2 plot(x,y3,'Color',[0.6 0.2 0.2]);
/3*a=0','a'); plot(x,y4,'Color',[1 0.1 0.1]);
plot(x,y5,'Color','red','LineWidth',2)
SA=solve('0.8*(0.8‐1)*(0.8‐2/3)*(0.8‐a)=‐0.00629565');
SASA=solve('0.8*(0.8‐1)*(0.8‐b)*(0.8‐a)=‐0.00629565','0.375
*(0.375‐1)*(0.375‐b)*(0.375‐a)=0.0121507');
Tsagi.m
clear all
clc PerGambit=[AIII,zizi];
PerGambit=PerGambit.*21.3;
AX=PerGambit(:,1);
AY=PerGambit(:,2);
AREAX=trapz(AX,abs(AY));
thickness=17; save('TSAGIGambit.DAT','PerGambit','‐ascii')
%i=0;
%LL=1/79; plot(AIII(:,1),AIII(:,2))
load('XX'); save('TSAGI2.DAT','AIII','‐ascii')
x=XX'; II=max(length(x));
x=‐x+1; for i=II:‐1:1
for i=1:1:80 xdown(:,‐i+II+1)=x(:,i);
ydown1(:,‐i+II+1)=ydown(:,i);
yup(i)=thickness/100*[1/4*(‐x(i)+7*x(i)^2‐6*x(i)^3)+x(i)^(0.87 end
)*((1‐x(i)^2)^(0.56))]; XTsagi=[xdown,x];
ydown(i)=thickness/100*[1/4*(‐x(i)+7*x(i)^2‐6*x(i)^3)‐x(i)^(0. XTsagi=XTsagi';
87)*((1‐x(i)^2)^(0.56))]; YTsagi=[yup,ydown1];
end YTsagi=YTsagi';
%x=0:LL:1; XYTsagi=[XTsagi,YTsagi];
%x=x'; save('Tsagi.txt','XYTsagi','‐ascii');
AI=[x;yup]'; %hold on
grid on
AII=[x;ydown]'; axis equal
AI(:,1)=‐AI(:,1)+1; %plot(XTsagi,YTsagi);
AII(:,1)=‐AII(:,1)+1;
TsagiN.m
This program was used to build a renewed profile starting from the Tsagi curve.
ECS_calculations.m
This program perform some calculations over the amount of power needed for the air
conditioning packs.
mc=4.0;%kg/s
mmm=mm*ro;% kg/s entire plane mh=1.75;%kg/s
T1i=225.77;%K
% COMPRESSOR DATA ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ T2ii=T2;%K
etac=0.88; T2u=300;%K
ri_re=0.3;% Ratio between shaft radius and external anulus K = 7 . 9 ; % W / m ^ 2 / K a s r e f e r r e d i n
radius http://www.engineeringtoolbox.com/overall‐heat‐transfer‐co
T0=225.772; efficients‐d_284.html
p0=23086.105; cph=pure_air_cp((T2ii+T2u)/2);%J/kg/K
rho0=0.35614181; cpc=pure_air_cp(T1i);%J/kg/K
u0=137.77251;%m/s P_t=mh*cph*(T2ii‐T2u);%W
T2i=T0*((pres/p0)^((g‐1)/g)); % isoentropic process T1u=P_t/mc/cpc+T1i;%K
temperature DTo=T2ii‐T1i;
T2=(T2i‐T0)/etac+T0; % temperature at the end of the DTs=T2u‐T1u;
process
DTm=(DTo‐DTs)/log(DTo/DTs);
SUP=P_t/K/DTm;
% Referring to Centrifugal Compressor:
T20media=(T2+T0)/2; %‐‐‐‐‐‐‐SAVES THE WORKSPACE‐‐‐‐‐‐‐‐‐‐‐‐
cpef=pure_air_cp(T20media); Names=who;
Namesm=max(size(Names));
Li=cpef*(T2‐T0); %Work fid=fopen('provaaspettando2.txt','w');
Pi=Li*mmm/etac; %Power for i=1:1:Namesm;
pi_comp=pres/p0;% Compression Ratio s t r i n g a = [ ' f p r i n t f ( fi d , ' ' ' , c h a r ( N a m e s { i , 1 } ) , ' = % 1 0 . 6 f
dg=sqrt(mmm*4/rho0/u0/pi/(1‐ri_re^2));%inlet diameter \n'',',char(Names(i,1)),');'];
eval(stringa);
end
% HEAT EXCHANGER ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ fclose(fid);
Athmosphere.m
This program is a revised (non automatic) version of the Fluent routine, it can handle
each type of profile, so that it return correct values for the boundary conditions. Since it
can read a text file, no particular mesh file is required.
vel=(M)*sqrt(temperat*R*gamma); %m/s PuntoR=ro/roref;
%vel=265.3; %m/s PuntoP=pres/pref;
%alpha=1.1895;
alphatrim=cos(deg2rad(alpha)); %risp al coseno
alphatrimsin=sin(deg2rad(alpha)); x=0:0.1:30;
Xvel=vel*alphatrim; pres=pressure_ratio(x,pref)/101300;
Yvel=vel*sin(deg2rad(alpha)); dens=ro_ratio(x,roref)/1.225;
x1=0:0.1:10.9;
Temp1=temperature_ratio(x1,Tref)/Tref;
ptot=pres*(TtotT)^((gamma)/(gamma‐1)); x2=11:0.1:19.9;
XYZ=load('nacagambit2.txt'); Temp2=temperature_ratio(x2,Tref)/Tref;
X=XYZ(:,1,1); x3=20:0.1:30;
Y=XYZ(:,2,1); Temp3=temperature_ratio(x3,Tref)/Tref;
YY=abs(Y);
areatsagi=53.9540358495598; Temp=[Temp1,Temp2,Temp3];
airfoil=trapz(X(1:80),Y(1:80))*2;
disp('‐‐‐‐‐‐‐‐‐‐‐‐‐'); %plot(x,pres,x,dens,x,Temp)
strng=['disp(''ISO Atmosphere at ',num2str(altitudine),' x1=x;
[m]'');']; y1=pres;
eval(strng); y2=dens;
y3=Temp;
strng=['disp(''Temperature: ',num2str(temperat),' [K]'');'];
eval(strng); hold on
strng=['disp(''T_tot: ',num2str(Ttot),' [K]'');']; grid on
eval(strng); %axis equal;
strng=['disp(''Density: ',num2str(ro),' [kg/m^3]'');']; %% Create axes
eval(strng); axis([0 30 0 1]);
strng=['disp(''Pressure: ',num2str(pres),' [Pa]'');']; %ylim(axes1,[0 1]);
eval(strng); title('\bf\fontsize{12}Standard Atmosphere ISO');
strng=['disp(''P_tot: ',num2str(ptot),' [Pa]'');']; text(15.5,0.95,'\rho_r_e_f=1.225 kg/m^3 p_r_e_f=10130 Pa
eval(strng); T_r_e_f=288.2 K','FontSize',8);
strng=['disp(''mu: ',num2str(mu),' [kg/m/s]'');']; xlabel('Altitude [km]');
eval(strng); ylabel({'\rho_r_a_t_i_o T_r_a_t_i_o p_r_a_t_i_o'});
disp('‐‐‐‐‐‐‐‐‐‐‐‐‐');
disp('Reference values:');
strng=['disp(''T_ref:',num2str(Tref),' [K]'');']; %% Create plot
eval(strng); plot(x1,y1,'r','LineWidth',2,'DisplayName','pressure ratio');
strng=['disp(''Ro_ref: ',num2str(roref),' [kg/m^3]'');'];
eval(strng); %% Create plot
strng=['disp(''Pref: ',num2str(pref),' [Pa]'');']; plot(x1,y2,'LineWidth',2,'DisplayName','\rho ratio');
eval(strng);
strng=['disp(''Airfoil Area: ',num2str(airfoil),' [m^2]'');']; %% Create plot
eval(strng); p l o t ( x 1 , y 3 , ' C o l o r ' , [ 0 0 . 8
strng=['disp(''Lenght: ',num2str(maxchord),' [m]'');']; 0],'LineWidth',2,'DisplayName','Temperature ratio');
eval(strng);
disp('‐‐‐‐‐‐‐‐‐‐‐‐‐'); plot(altitudine/1000,PuntoT,'*','Color','black','MarkerSize',10)
strng=['disp(''M: ',num2str(M),' [ ‐ ]'');']; ;
eval(strng); plot(altitudine/1000,PuntoR,'*','Color','black','MarkerSize',10
strng=['disp(''|V|: ',num2str(vel),' [m/s]'');']; );
eval(strng); plot(altitudine/1000,PuntoP,'*','Color','black','MarkerSize',10)
strng=['disp(''Vx: ',num2str(Xvel),' [m/s]'');']; ;
eval(strng); plot([altitudine/1000,altitudine/1000],[PuntoT,0],'‐‐','Color','b
strng=['disp(''Vy: ',num2str(Yvel),' [m/s]'');']; lack');
eval(strng); legend('Pressure','Density','Temperature','Location','East');
strng=['disp(''XComp: ',num2str(alphatrim),' [‐]'');']; %% Create legend
eval(strng); % l e g e n d 1 = l e g e n d ( a x e s 1 , { ' p r e s s u r e r a t i o ' , ' \ r h o
strng=['disp(''YComp: ',num2str(alphatrimsin),' [‐]'');']; ratio','Temperature ratio'},'Position',[0.1486 0.1381 0.2245
eval(strng); 0.1631]);
strng=['disp(''Area TsAGI: ',num2str(areatsagi),' [m^2]'');'];
eval(strng); %Mesh for cruise:
disp('‐‐‐‐‐‐‐‐‐‐‐‐‐'); %Upstream ‐20 c
%Points of interest %downstream 25 c , 30 c
PuntoT=temperat/Tref;
engine_calculations.m
This program perform some calculations by comparing known data on turbofan en-
gines of common airplanes.
% CALCULATIONS ON ENGINE'S DIMENSION: CFM_l=92; % lenght of the engine
CFM=CFM_T*lbf2kN;
%Thrust Needed
T=510; % kN %‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
lbf2kN=0.00444822162; % CFM56‐5C2 A‐340
in2m=0.0254; %‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
Tlbf=T/lbf2kN/2; %Thrust needed lbf single engine
%‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ CFM3_T=34000*2; % lbf of the two engines
% CFM56‐7B B737‐800 CFM3_f=72; % inches of diameter of the fan
%‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ CFM3_l=103; % lenght of the engine
CFM3=CFM3_T*lbf2kN;
CFM_T=27300*2; % lbf of the two engines leng=132*in2m;
CFM_f=61; % inches of diameter of the fan
scambiatore.m
%SCAMBIATORE
DTm=(DTo‐DTs)/log(DTo/DTs);
clear all SUP=P_t/K/DTm;
clc
%x=1:1:floor(SUP);
mc=4.0;%kg/s
mh=1.75;%kg/s %‐‐‐‐‐‐‐SAVES THE WORKSPACE‐‐‐‐‐‐‐‐‐‐‐‐
T1i=225.77;%K Names=who;
T2i=330.79;%K Namesm=max(size(Names));
T2u=300;%K fid=fopen('Scambiatore.txt','w');
K = 7 . 9 ; % W / m ^ 2 / K a s r e f e r r e d i n for i=1:1:Namesm;
http://www.engineeringtoolbox.com/overall‐heat‐transfer‐co s t r i n g a = [ ' f p r i n t f ( fi d , ' ' % 9 s = % 1 5 . 6 f
efficients‐d_284.html \n'',char(Names(i,1)),',char(Names(i,1)),');'];
cph=pure_air_cp((T2i+T2u)/2);%J/kg/K % stringa=['fprintf(fid,''',char(Names{i,1}),'=%10.6f
cpc=pure_air_cp(T1i);%J/kg/K \n'',',char(Names(i,1)),');'];
P_t=mh*cph*(T2i‐T2u);%W eval(stringa);
T1u=P_t/mc/cpc+T1i;%K end
DTo=T2i‐T1i; fclose(fid);
DTs=T2u‐T1u;
pure_air_cp.m
This program perform some calculations on the isoentropic exponent changes with
temperature in pure air fluid.
pressure767.m
This program perform some calculations on the differential pressure between outside
and inside the fuselage of a Boeing 767 as reference [39].
% This program prints the plot of the differential pressure t i t l e ( ' C a b i n P r e s s u r e v s . O u t s i d e P r e s s u r e i n a
value in a 767 767','FontWeight','Demi');
clc xlabel('Airplane altitude in 1000 \times [ft]','FontSize',10);
close all ylabel('Cabin altitude in 1000 \times [ft]','FontSize',10);
clear all plot(x,y,'LineWidth',2);
hold off
% X Y = c s v r e a d ( ' C : \ T e s i \ p e r Xm=x.*0.3048;
Engauge\GraficoPressioni767.csv',1); %1000 of feet Ym=y.*0.3048;
% X=XY(:,1); fog=800;
% Y=XY(:,2); figure(fog)
load('pressur767.mat') hold on
x=‐2:0.05:45; % this values are in 1000 feet grid on
y=pressure767(x); % this values are in 1000 feet %axis equal
x=x'; t i t l e ( ' C a b i n P r e s s u r e v s . O u t s i d e P r e s s u r e i n a
fig=900; 767','FontWeight','Demi');
figure(fig) xlabel('Airplane altitude in 1000 \times [m]','FontSize',10);
hold on ylabel('Cabin altitude in 1000 \times [m]','FontSize',10);
grid on plot(Xm,Ym,'LineWidth',2);
%axis equal hold off
CL2profilesplot_crs.m
This program perform some calculations for the comparative lift coefficient analysis
between the RAE and the NASA airfoils. Then plots the result in an updatable graphic.
As example only the cruise file is here reported.
figure(fog); axisH = axes;
hold on
plot(xrad',yl1(x'),'‐‐','LineWidth',1,'Color','blue','DisplayName'
plot(xrad,y21,'+‐','LineWidth',2,'Color','red','Displayname','C_ ,'C_L Nasa linearized');
d Rae CRS'); addTopXAxis(axisH, 'expression', '180/pi*(argu)','xLabStr',
legend('Location','SouthEast'); 'Angle of Attack \alpha [deg]');
set(legend,'FontSize',8); xlabel('Angle of Attack \alpha [rad]');
hold off ylabel('Lift Coefficient C_L [ ‐ ]');
end text(0.02,0.9,'C_L ‐ \alpha for the two
% CREATE THE C_m PLOT profiles','FontSize',14,'Units','Normalized');
grid(axisH,'on');
fug=93; hold off
try get(fig1);
get(fug); figure(fig1);
figure(fug); hold on
hold on
plot(xrad,y12,'^‐','LineWidth',2,'Displayname','C_m Nasa plot(xrad',yl2(x'),'‐‐','LineWidth',1,'Color','red','DisplayName','
CRS'); C_L Rae linearized');
legend('Location','SouthEast');
plot(xrad,y22,'+‐','LineWidth',2,'Color','red','Displayname','C_ hold off
m Rae CRS');
legend('Location','SouthEast'); % CREATE THE C_D PLOT Linearized
set(legend,'FontSize',8); % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
hold off
catch fog1=82;
figure(fug); figure(fog1);
hold on hold on
axisH = axes; axisH = axes;
plot(xrad,y12,'^‐','LineWidth',2,'Displayname','C_m Nasa
CRS'); plot(xrad',yd1(x'),'‐‐','LineWidth',1,'Color','blue','DisplayName
addTopXAxis(axisH, 'expression', '180/pi*(argu)','xLabStr', ','C_d Nasa linearized');
'Angle of Attack \alpha [deg]'); addTopXAxis(axisH, 'expression', '180/pi*(argu)','xLabStr',
xlabel('Angle of Attack \alpha [rad]'); 'Angle of Attack \alpha [deg]');
ylabel('Moment Coefficient C_m [ ‐ ]'); xlabel('Angle of Attack \alpha [rad]');
text(0.02,0.1,'C_m ‐ \alpha for the two ylabel('Drag Coefficient C_d [ ‐ ]');
profiles','FontSize',14,'Units','Normalized'); text(0.02,0.9,'C_d ‐ \alpha for the two
grid(axisH,'on'); profiles','FontSize',14,'Units','Normalized');
hold off grid(axisH,'on');
get(fug); hold off
figure(fug); get(fog1);
hold on figure(fog1);
hold on
plot(xrad,y22,'+‐','LineWidth',2,'Color','red','Displayname','C_
m Rae CRS'); plot(xrad',yd2(x'),'‐‐','LineWidth',1,'Color','red','DisplayName'
legend('Location','NorthEast'); ,'C_d Rae linearized');
set(legend,'FontSize',8); legend('Location','SouthEast');
hold off hold off
end
saveas(fig,'CL_alpha.eps','eps'); % CREATE THE C_m PLOT Linearized
saveas(fig,'CL_alpha.pdf','pdf'); % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
saveas(fog,'Cd_alpha.eps','eps'); fug1=83;
saveas(fog,'Cd_alpha.pdf','pdf'); figure(fug1);
saveas(fug,'Cm_alpha.eps','eps'); hold on
saveas(fug,'Cm_alpha.pdf','pdf'); axisH = axes;
button0 = questdlg('Do you want to plot also the linearized
plots?','Plot Linearized','Cancel','Yes','Cancel'); plot(xrad',ym1(x'),'‐‐','LineWidth',1,'Color','blue','DisplayNam
if strcmp(button0, 'Yes')==1 e','C_L Nasa linearized');
% CREATE THE C_L PLOT Linearized addTopXAxis(axisH, 'expression', '180/pi*(argu)','xLabStr',
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ 'Angle of Attack \alpha [deg]');
xlabel('Angle of Attack \alpha [rad]');
fig1=81; ylabel('Moment Coefficient C_m [ ‐ ]');
figure(fig1); text(0.02,0.1,'C_m ‐ \alpha for the two
hold on profiles','FontSize',14,'Units','Normalized');
LogLaw.m
% This Application plots the Log Law
yplus=1:1:300;
clear all uplus=1/K.*log(yplus)+B;
clc yplus=[0;yplus'];
K=0.41; uplus=[0;uplus'];
B=5.2; yb=yplus./ustar*ni;
%LOW BOUNDARY
tauw=9.6365547; %Pa fig=5;
rho=0.33694708;%kg/m^3 figure(fig);
ustar=sqrt(tauw/rho); axisH = axes;
U=83.964241; hold on
uplusb=U/ustar; grid on
mu=1.4523035e‐05; %title('u^+=f(y^+)','FontWeight','Demi','FontSize',10);
ni=mu/rho; plot(yplus,uplus,'DisplayName','u^+=f(y^+)','LineWidth',2);
a d d T o p X A x i s ( a x i s H , ' e x p r e s s i o n ' ,
%HIGH BOUNDARY 'argu/ustar*ni*1000','xLabStr', 'y [mm]');
tauwh=44.470158; %Pa
rhoh=0.41186216;%kg/m^3 ylabel('u^+');
ustarh=sqrt(tauwh/rhoh); xlabel('y^+');
Uh=188.08511; % m/s (control point a little aft
uplush=Uh/ustar; hold off
mass.m
This program perform some calculations on the boundary layer lip dimensions, and
then plot a figure that summarize the results.
clear all % ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
clc center=[0 delta];
NOP=300;
% BOUNDARY LAYER COORDINATES THETA=linspace(0,pi,NOP);
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ RHO=ones(1,NOP)*radius_intake;
X=[16.44138 16.40475]; [Xc,Yc] = pol2cart(THETA,RHO);
Y=[1.172649 1.023583]; Xc=Xc+center(1);
Yc=Yc+center(2);
XX=(X(1)‐X(2));
YY=(Y(1)‐Y(2)); %PLOTS THE FIGURE (in red is presented the boundary layer
delta=sqrt(XX^2+YY^2); ingesting lip)
areabl=0.752;%m^2 %‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐
speed=235;%m/s fig=99;
rho=0.33304;% kg/m^3 at 11500 m figure(fig);
engine_rad=2.45/2; hold on
area=(pi*(engine_rad)^2);% m^2 area of the engine inlet axis([‐3 3 ‐1 4]);
speedbl=speed/2; %m/s this is the speed of the lip of fluid axis equal
(slower), conservative since this is the upper bound
portatacruise=areabl*rho*speedbl; grid on
t i t l e ( ' E n g i n e S ‐ D u c t I n t a k e
Geometry','FontSize',10,'FontWeight','Demi');
% Calculate intake area; testo={['Area Boundary Layer=',num2str(areabl),'
for i=1:0.05:10 m^2'];['Area Intake=',num2str(area_intake),' m^2']; ...
area_intake=pi*i^2/2; ['Radius (without BLI) =',num2str(radius_intake),' m^
if area_intake>=area; '];['Reference Engine Radius =',num2str(engine_rad),' m^ ']};
radius_intake=i; text(0.03,0.87,testo,'FontSize',8,'Units','Normalized','Linewid
break; th',1,'EdgeColor','black','BackGround','White','Margin',5)
end xlabel('x [m]');
end ylabel('y [m]');
plot(Xc,Yc,'LineWidth',2,'DisplayName','Engine Intake');
%Calculate Boundary Layer ingesting Lip plot(Xb,Yb,'LineWidth',2,'Color','red','DisplayName','BLI
%‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ Lip');
legend('Location','NorthEast');
X b = [ ‐ r a d i u s _ i n t a ke r a d i u s _ i n t a ke r a d i u s _ i n t a ke set(legend,'FontSize',8);
‐radius_intake ‐radius_intake]; hold off
Yb=[0 0 delta delta 0];
% SAVES THE SEMI CIRCLE OF THE ENGINE INTAKE (Thanks
to circle.m)
SDuct.m
This program perform some fluid dynamics calculations on the S Duct shape, it uses
the theory reported in [38].
% This program calculates the conditions of the flow that is A2=2.5;% m4.9; %m^2
passing M2=0.6;
% through the S duct, it uses the method described in Zucker M1=1.2151933;
Fundamentals T10=240.64508;%K
% of Gas Dynamics T1=185.88234;%K
% ‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐ p1=12076.584;%Pa
% Hypothesis: u1=331.99658;%m/s
% the flow is: rho1=0.22633789;
% ‐stationary ds=0;
% ‐adiabatic g=1.4;
% ‐no work inside the volume R=287;
% ‐perfect gas cp=1080;
% ‐neglet potential %p2=p1/(((1+(g‐1)/2*M2^2)/(1+(g‐1)/2*M1^2))^(g/(g‐1)))/exp(
% ds/R);
% Through this Hypothesis a simplified form of the Euler T2=T1*(1+(g‐1)/2*M1^2)/(1+(g‐1)/2*M2^2);
Equation's might p2=A1/A2*p1*M1/M2*sqrt(T2/T1);
% be used. rho2=rho1*(T2/T1)^(1/(g‐1))*exp(‐ds/R);
% An additional term of friction might be inserted through T20=T2+M2*(g*R*T2)/cp/2;
the Ds term.
%‐‐‐‐‐‐‐SAVES THE WORKSPACE‐‐‐‐‐‐‐‐‐‐‐‐
Names=who;
clear all Namesm=max(size(Names));
clc fid=fopen('SDuct.txt','w');
format long for i=1:1:Namesm;
s t r i n g a = [ ' f p r i n t f ( fi d , ' ' ' , c h a r ( N a m e s { i , 1 } ) , ' = % 1 0 . 6 f
% A1=input('Insert the Value of the surface inlet A1= '); \n'',',char(Names(i,1)),');'];
% disp('A2=4.9 m^2 ‐‐default value due to a 2.5 m diameter eval(stringa);
fan'); end
fclose(fid);
A1=1.740; %m
saveastext.m
Even if this program does not affect in any way the study performed it is very useful
and might be used in other applications. Those few lines of code, if added to an existing
program create a text file with the name of the variables used in the program paired with
the calculated value. This program is very useful to be used as report on computers that
do not have Matlab™ installed, or to copy data with ease.
Drawings
S
ome technical designs have been made using the CATIA software in order to
better represent the integration issues, and the geometries in general. At the
end of this chapter are presented some graphical images that have the intent to
describe with an higher detail some sections of the 3D model. The drawings are pre-
sented only on the right page in order to offer an easier readability. On the bottom of the
left page a little caption describing the image on the right page was added.
Geometry:
A A
7580
1740,03
3267,73
B B
5444
2500
130
C 22457,73 C
CHK'D
D Engine S-Duct
APPV'D
Giovanni Medici ~ Engine Integration on a Flying Wing
Q.A MATERIAL:
DWG NO.
A4
Alluminium 1
1 2 WEIGHT: --- SCALE:1:100 SHEET 1 OF 1
Giovanni Medici ~ Appendix II Tech Designs
F Fan
starter valve
(reference)
fan air mod valve
F T
check valve
I
to cowl
H high pressure precooler anti ice
shut off valve
temperature
to aircraft
sensor to wing
anti ice
The Bleed System automatically provides air at the proper temperature and
pressure required to meet needs of all pneumatic services on the aircraft
Giovanni Medici ~ Engine Integration on a Flying Wing
RAT
P G
RAT
P G
P M SG
P M
P Engine
HX SG
M SG
HX P Engine
ECS CTAI
M M SG XFR Battery
ECS P M Battery
M XFR CTAI
P M HX Heat Exchanger
M
M SG
APU SG XFR Trans / Rectifier
NGS NGS SG APU
P SG SG Starter Generator
M P M
CTAI
ECS XFR RAT Ram Air Turbine
M SG
HX P Hydraulic Pump
M P M P Engine
XFR
ECS M SG CTAI M Motor
M P
HX SG CTAI Cowl Thermal Anti–Ice
P Engine
Electrical
P M SG
Hydraulic
HX Heat Exchanger
ELECTRICAL Electrical
POWER IS MORE
Hydraulic
EFFICIENT THAN
Pneumatics
ENGINE-GENERATED
Fuel
PNEUMATIC POWER.
Ram Air
Giovanni Medici ~ Engine Integration on a Flying Wing
Legend:
Non Return Valves: NRV
Bypass Valve
Fwd Aft
Cargo Cargo
Pressure Regulating Valve
Mix
Outflow valves
Mainfold
Extraction fans ACP II
Turbine Water Extractor Heat Exchanger
Filters
Motor
Ozone Remover
Compressor
References
T
his chapter collects all the papers and books that where used as reference for
this work. Every list’s item was added chronological, so that there are not
sections for each area (ie. aerodynamics fluid dynamics, systems...). Main
source for the search was NASA Technical References, and Journal of Aircrafts.
[1] Martinez-Val R., Pérez E. et al., “Conceptual design of a medium size flying
wing”, JAERO90, Vol. 221 pag. 57–66, July 2007.
[2] Liebeck R. H., “Design of the Blended wing body subsonic transport”, Journal
of Aircraft, Vol. 41 No. 1 pag. 10–25, January 2004.
[3] Bolsunovsky A. L., Buzoverya N. P. Et al., “Flying Wing ∼ Problems and deci-
sions” , Von Karman Institute Lecture Series 2005-2006, Vol.1 January 2005.
[4] Martinez-Val R. Pérez E. Et al.,“Medium Size Flying Wing” , Von Karman Insti-
tute Lecture Series 2005-2006, Vol.1 January 2005.
[5] Kroo I., “Non planar wing concepts for increased aircraft efficiency” , Von Kar-
man Institute Lecture Series 2005-2006, Vol.1 January 2005.
[6] McMasters J. H., Cummings R. M., “From Farther Faster, Higher to Leaner,
Meaner, Greener”, Journal of Aircraft, Vol. 41 No. 1, January 2004.
[8] Sedillie M., “La propulsion par réaction en combinacion avec l’aspiration de la
couche limite”, Congres Nationale De L’Aviation Froncoise, 1946.
[9] Williams P. R. G., Steward D. J., “An aircraft designer’s review of some aircraft
and engine integration concept”, British Aircraft Corp., 1973.
[10] Owens L. R., Allan B. G., et al., “Boundary–Layer–Ingesting Inlet Flow Con-
trol” Journal of Aircraft American Institute of Aeronautics and Astronautics,
2005.
[11] Plas A. P., Sargeant M. A., Madani V., et al, “Performance of a Boundary Layer
Ingesting (BLI) Propulsion System” AIAA 45th Science Meeting Exhibit, Vol 8–
11, January 2007.
[12] Lian, Zhang X., Wu L., “Flow Control Analysis of S-duct Diffuser Inlet”,
AJCPP, Vol 6. N. 8, 2008.
[14] Österheld C., Heinze W., et al., “Preliminary design of a Blended Wing Body
Configuration using the Design Tool PrADO”, IFL Institut für Flugzeugbay und
Leichtbau TU Braunschweig, 2004.
[15] Kawai R. T., Friedman M., et al., “Blended Wing Body (BWB) Boundary Layer
Ingestion (BLI) inlet configuration and System studies”, NASA/
CR−2006−214534 , November 2006.
[16] Daggett L. D., Kawai R. T., et al., “Blended Wing Body (BWB) Boundary
Laryer Ingestion Inlets with Active Flow Control”, NASA/TM-2003-212670,
December 2003.
[18] Abbott I. H., Von Doenhoff A. E., et al., “Summary of Airfoil data”, National
Advisory Commitee for Aeronautics NACA Report n° 824, Langley Memorial
Aeronautical Laboratory, 1945.
[19] Ladson C., Brooks C., “Development of a computer program to obtain coordi-
nates for NACA 6 and 6ª- series Airfoil” , NASA/TM 4741, Hampton, Virginia,
December 1996.
[20] Bradley K., “A sizing methodology for the conceptual design of Blended Wing
Body (BWB) Transport”, NASA/CR 2004 213016, September 2004.
[21] Martínez-Val R., Cuerno C., et al., “Potential effects of Blended Wing Bodies on
air transportation system”, 9th AIAA Aviation Technology Integration and Op-
eration Conference (ATIO), Hilton Head South Carolina, USA, September 2009.
[22] Scholz D., “A student project of a Blended Wing Body aircraft, from conceptual
design to flight testing”, 8th European Workshop on aircraft design education,
Samara, Russia, June 2007.
[23] Vicroy D., “Blended Wing Body low speed flight dynamics: summary of ground
test and sample result (invited)”, NASA Langley Research Center, Hampton,
January 2009.
[24] Roskam J., “Airplane Design – Part III: Layout design of cockpit, fuselage, wing
and empennage: cutaways and inboard profiles”, Dar Corporation, Larwence
Kansas, 1999.
[25] Roskam J., “Airplane Design– Part VI: Preliminary calculation of aerodynamic
thrust and power characteristics”, Dar Corporation, Larwence Kansas, 1999.
[26] Nickel K., Wohlfahrt M., “Tailless Aircraft in Theory and Practice”, Edward Ar-
nold, London, 1994.
[27] Loftin L. K., Bursnall W. J, “The effects of variations in Reynolds number be-
tween 3.0 105 and 25.0 105 upon the aerodynamic characteristic of a number of
NACA 6–series airfoil sections”, National Advisory Committee for aeronautics,
Report 964, Langley, October 1946.
[28] Ladson C., “Two dimensional airfoil characteristic of four NACA 6–series air-
foils at transonic Mach numbers up to 1.25”, Naca Research Memorandum,
Langley, August 1957.
[29] Melin T., “A Vortex Lattice MATLAB Implementation for Linear Aerodynamic
Wing Applications.”, Royal Institute of Technology Stockholm, December 2000.
[30] Mason W. H., “Transonic Aerodynamics of Airfoils and Wings”, Applied Com-
putational Aerodynamics, August 1997.
[32] Roman D., Allen J.B., “Aerodynamic design challenges of the Blended–W-
ing–Body Subsonic Transport”, AIAA–2000–4335, 2000.
[33] Harris C., “NASA Supercritical Airfoils: A matrix of family related airfoils”,
NASA Technical paper, March 1990.
[34] Liebeck R., “Design of Subsonic Airfoils for High Lift”, Journal of Aircraft Vol.
15, No. 9, pag 547–561, September 1978.
[35] Sinnett M., “787 No Bleed–Systems: Saving Fuel and enhancing operational
efficiencies”, AeroMagazine Boeing, December 2007.
[38] Zucker R., “Fundamental of Gas Dynamics”, John Wiley & Sons, Monterey
California, 2002.
[39] Couaillac F., “Environmental Control Systems for the all–electric aircraft”,
Cranfield University, 2006.
[40] Hunt H., Reid H., “Commercial Airliners Environmental Control System”, Pre-
sented at Aerospace Medical Association Annual Meeting, Anaheim, CA, May
1995.
[41] Payne R., “Stuck on the Drawing Board: Unbuilt British Commercial Aircraft
Since 1945” Tempus, January 2004.