Department of Mechanical Engineering

Post-processing on the aerodynamics of the underbody

of a Formula 1 single-seater in accordance with the new
FIA technical regulation from 2022

Supervisor: Student:
Prof. Ing. Jan Oscar Pralits Alberto Toccafondi

Co-supervisor:
Prof. Ing. Joel Enrique Rivas
Ing. Marco Giachi

July 2022
 Abstract

The purpose of this investigation is to learn how to associate aerodynamics

events outside of a Formula 1 car, conveniently designed in compliance with
the new project rules of the 2022 Formula 1 World Championship To this end,
this thesis starts from numerical and models’ results belonged to older thesis
by other students at University of Genoa and these results will be analysed
here throughout the application of some software, like Ansys EnSight and Par-
aView, in order to study how the Post-Processing works and how to improve
the performance.The numerical results will be focused on the floor and the main
phenomena will be shown and described through pictures and the descriptions
of the physical laws that participate. The aim of this thesis is to learn how to
correlate the aerodynamic phenomena present around a Formula 1 car, appro-
priately designed with the new regulations in force since 2022, through the use
of the main Post-Processing tools such as Ansys EnSight and ParaView soft-
ware. The project of this thesis starts from the numerical results and models
obtained from previous thesis projects by students of the University of Genoa.
The numerical results will be focused particularly on the car’s floor, and the
main phenomena will be described and properly argued through the use of im-
ages and the description of the physical laws involved.
Contents

1 List of symbols 5
1.1 List of the symbols . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Introduction 7
2.1 Description of the thesis subject . . . . . . . . . . . . . . . . . . 7
2.2 Starting points of the project . . . . . . . . . . . . . . . . . . . . 7
2.3 Hardware and software tools used . . . . . . . . . . . . . . . . . . 8
2.4 Thesis objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3 Formula 1 2022 Technical Regulations 9

3.1 General Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 What Changes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

4 Definitions 13
4.1 Aerodynamic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.2 Lift and Lift Coefficient (Cz ) . . . . . . . . . . . . . . . . . . . . 14
4.3 Drag and Drag Coefficient (Cx ) . . . . . . . . . . . . . . . . . . . 15
4.3.1 Friction drag . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.3.2 Form drag . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3.3 Induced drag . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3.4 Wave drag . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3.5 Aerodynamic efficiency . . . . . . . . . . . . . . . . . . . 19
4.4 Steady flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.5 Significant Applications . . . . . . . . . . . . . . . . . . . . . . . 20
4.5.1 Venturi Channel . . . . . . . . . . . . . . . . . . . . . . . 20
4.5.2 Bernoulli principle . . . . . . . . . . . . . . . . . . . . . . 22
4.5.3 Ground effect . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.6 Navier-Stokes equations . . . . . . . . . . . . . . . . . . . . . . . 24
4.7 Vorticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.8 Computational Fluid Dynamic (CFD) . . . . . . . . . . . . . . . 26
4.8.1 Polygonal mesh . . . . . . . . . . . . . . . . . . . . . . . . 27
4.9 Post-Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.9.1 Ansys CFX . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.9.2 OpenFoam . . . . . . . . . . . . . . . . . . . . . . . . . . 28
4.9.3 ParaView . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

1
5 Tools 31
5.1 Streamlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.2 Oil-Film . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.3 Pressure maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.3.1 Pressure coefficient (Cp ) . . . . . . . . . . . . . . . . . . . 33
5.3.2 Total pressure coefficient (Cp tot ) . . . . . . . . . . . . . . 34
5.4 Q-Criterion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

6 Cases examined 35
6.1 Isolated Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6.2 Intermediate model . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.3 Complete model . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

7 Results and observations 39

7.1 Stagnation point . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
7.2 Separation and reattachment point . . . . . . . . . . . . . . . . . 40
7.3 Slipstream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
7.4 Velocity and pressure variation . . . . . . . . . . . . . . . . . . . 43
7.5 Vortices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

8 Conclusions 55

2
List of Figures

3.1 Rear wing of the Aston Martin AMR22-B . . . . . . . . . . . . . 10

3.2 Front wing of the Ferrari F1-75 . . . . . . . . . . . . . . . . . . . 10
3.3 Underbody of the Ferrari F1-75 . . . . . . . . . . . . . . . . . . . 11
3.4 Volumes under the new regulations . . . . . . . . . . . . . . . . . 12

4.1 Forces acting on an airplane . . . . . . . . . . . . . . . . . . . . . 14

4.2 Forces acting on a car . . . . . . . . . . . . . . . . . . . . . . . . 15
4.3 Ferrari 312 F1, 1968 . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.4 Boundary layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.5 Variation of Cz and Cx as a function of angle of attack (AoA) . . 18
4.6 Variation of angle of attack (AoA) . . . . . . . . . . . . . . . . . 18
4.7 Wind tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.8 Venturi channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.9 Velocity difference between the intrados and extrados of a wing . 23
4.10 Lotus-78 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.11 Venturi channel and underbody . . . . . . . . . . . . . . . . . . . 24
4.12 Polygonal mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.13 Example of studying an ogive using ParaView software, front view 29
4.14 Example of studying an ogive using ParaView software, back view 30

5.1 Streamlines on the front part of the F14-T, top view . . . . . . . 32

5.2 Application of oil film on the RB-19 . . . . . . . . . . . . . . . . 33
5.3 Wall-Shear-Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

6.1 Bottom view of the isolated case . . . . . . . . . . . . . . . . . . 35

6.2 Entrance of the bottom of the isolated case . . . . . . . . . . . . 36
6.3 Bottom view of the intermediate case . . . . . . . . . . . . . . . . 36
6.4 Front view of the intermediate case . . . . . . . . . . . . . . . . . 37
6.5 Bottom view of the complete case . . . . . . . . . . . . . . . . . . 37
6.6 Front view of the complete case . . . . . . . . . . . . . . . . . . . 38

7.1 Cp and Oil-Film in the isolated case . . . . . . . . . . . . . . . . 39

7.2 Cp and Oil-Film in the complete case . . . . . . . . . . . . . . . . 40
7.3 Plane at x = cost at the entrance of the bottom in the isolated
case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3
7.4 Plane at x = cost at the entrance of the bottom in the interme-
diate case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
7.5 Plane at x = cost at the entrance of the bottom in the complete
case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
7.6 Pressure map with Cp tot as a parameter in the isolated case . . . 43
7.7 Pressure map with Cp tot as a parameter in the intermediate case 44
7.8 Pressure map with Cp tot as a parameter in the complete case . . 45
7.9 Pressure map with Cp as a parameter in the isolated case . . . . 46
7.10 Pressure map with Cp as a parameter in the intermediate case . 47
7.11 Pressure map with Cp as a parameter in the complete case . . . 47
7.12 Cp on the wall and Vadim on the plane in the isolated case . . . . 48
7.13 Cp on the wall and Vadim on the plane in the intermediate case . 48
7.14 Cp on the wall and Vadim on the plane in the complete case . . . 49
7.15 Arrangement of the plane from which the streamlines originate . 49
7.16 Stremlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
7.17 Tangential component on the cutting plane of the velocity vector
and Vadim on the cutting plane . . . . . . . . . . . . . . . . . . . 50
7.18 Boundary layer in the isolated case . . . . . . . . . . . . . . . . . 51
7.19 Boundary layer in the complete case . . . . . . . . . . . . . . . . 51
7.20 Boundary layer at the rear of the bottom in the isolated case . . 52
7.21 Boundary layer at the rear of the bottom in the complete case . 52
7.22 Y -component of velocity in the complete case . . . . . . . . . . . 53
7.23 Cp and Oil-Film on the bottom of the complete case . . . . . . . 53
7.24 Q-Criterion with Cp as a parameter in the complete case . . . . . 54
7.25 Q-Criterion with the velocity as a parameter in the complete case 54

4
Chapter 1

List of symbols

1.1 List of the symbols

Here a list of symbols used in the manuscript

Symbol Meaning IS Unit

F IA International Automobile Federation −
CAD Computer Aided Design −
CF D Computational Fluid Dynamics −
Cz Lift Coefficient −
L Lift N
ρ Density of the fluid kg m−3
ρ1 Density of the fluid at the inlet kg m−3
ρ2 Density of the fluid at the outlet kg m−3
ρa Density of the fluid at the section a kg m−3
ρb Density of the fluid at the section b kg m−3
V∞ Velocity at which the body penetrates the fluid m/sec
V1 Velocity of the fluid at the inlet m/sec
V2 Velocity of the fluid at the outlet m/sec
Va Velocity of the fluid through section a m/sec
Vb Velocity of the fluid through section b m/sec
A Frontal area of the body m2
Re Reynolds number −
Ma Mach number −
Cx Drag coefficient −
D Drag N

Table 1.1: List of symbols

5
Symbol Meaning IS Unit
u Component of the fluid velocity along the î axis m/sec
u0 Initial velocity of the fluid m/sec
u(y) Velocity calculated in point y m/sec
AoA Angle of attack deg
E Aerodynamic efficiency −
S Cross section area m2
S1 Cross section area at the inlet m2
S2 Cross section area at the outlet m2
Sa Cross section area at the section a m2
Sb Cross section area at the section b m2
ṁ Flow rate kg/sec
ṁA Flow rate through section A kg/sec
ṁB Flow rate through section B kg/sec
p0 Total pressure Pa
ps Static pressure Pa
ps∞ Static pressure at an infinitely distant asymptotic point Pa
pd Dynamic pressure Pa
g Acceleration due to the gravity m/sec2
h Average potential elevation of the section m
µ Dynamic viscosity Pa sec
θ Pitch angle rad
t time variable sec
ω Velocity rotor m/sec
Γ Circulation of a velocity field along a closed path m2 /sec
Ω Instantaneous rigid rotation rate m/sec
⃗t Direction of the tangent vector to the trajectory m
DN S Direct Numerical Simulation −
RAN S Reynolds Averaged Navier Stokes equations −
LES Large Eddy Simulation −
τw Wall shear stress Pa
y Position along the ĵ axis m
Cp Pressure coefficient −
Cptot Total pressure coefficient −
Q second invariant of the velocity gradient m/sec

Table 1.2: List of symbols

6
Chapter 2

Introduction

Aerodynamics undoubtedly represents one of the fundamental aspects of a high-

performance vehicle. The aim of this thesis is to observe some of the most
important aerodynamic phenomena through the use of images and combinations
of data provided by simulations of three different cases, and then highlight
their differences, thus analyzing the main phenomena that characterize their
performance. Throughout the discussion, studies based on steady flow and
incompressible fluid will be conducted.

2.1 Description of the thesis subject

Throughout this work, we will aim to describe and delve into various aerody-
namic aspects occurring in the underbody of a Formula 1 car. In the initial
section, there will be a brief overview of the new technical regulations for the
2022 Formula 1 World Championship, followed by a focus on the key definitions
and physical laws involved when discussing the results obtained in the latter
part. The section concerning definitions will strive to be as comprehensive as
possible to enhance understanding of the results, hence it will be quite exten-
sive. Subsequently, the tools used for visualizing the results will be introduced
and described, followed by an analysis of these results and concluding remarks.

2.2 Starting points of the project

As previously mentioned in the thesis subject, we begin with models and studies
conducted in the past by students of the University of Genoa; thus, this work
can be seen as a continuation of those previous efforts, albeit delving into an
entirely new topic, Post-Processing. Specifically, we will revisit the three models
that will be addressed: the case involving only the underbody (isolated model),
the case involving the underbody, front wing, and front wheel (intermediate
model), and the case involving the underbody, front wing, front wheel, and
fences (complete model).

7
2.3 Hardware and software tools used
The results were obtained using tools provided by the university. Specifically,
a computer with a Linux operating system and 128 GB of RAM was used for
long and complex calculations. The Ansys license was used for simulations
performed on EnSight, and the ParaView program was used to become familiar
with a simple model (a blunt body).

2.4 Thesis objective

The objective of this project is to learn to identify the most significant phenom-
ena that can significantly influence the aerodynamics of a car (both positively
and negatively) through the use of various tools such as pressure maps, stream-
lines, Oil-Film, velocity vectors, and various dimensionless coefficients. In par-
ticular, the aim is to learn to correlate their behavior with the aerodynamic phe-
nomenon being studied. For example, turbulence (vortex), a stagnation point,
a separation/reattachment point, but above all, how the pressure distribution
varies with velocity.

8
Chapter 3

Formula 1 2022 Technical

Regulations

3.1 General Aspects

In 2022, the Formula 1 World Championship undergoes a technical upheaval
from various perspectives. To mention a few aspects relevant to the subject
of this thesis, one recalls the reintroduction of ground effect (ch. 4.5.3), which
was heavily criticized in previous editions due to its danger; cost reduction; the
inclination to construct aesthetically pleasing cars; and the maximal limitation
of numerous aerodynamic appendices that generate a series of vortices (ch. 4.7)
around them (and behind them), which disturb the aerodynamics of the trailing
car. The aim is to enhance the spectacle by facilitating overtaking, precisely
by reducing the effect of ”dirty slipstream” and thus facilitating close pursuit.
One of the most significant features of the new cars will be the ability of the
underbody to generate downforce (ch. 4.2). Even before seeing the cars on
the track, it can be said that they will have higher straight-line speeds (due
to reduced aerodynamic resistance (ch. 4.3) generated by the front and rear
wings).

3.2 What Changes

The rear wing will be higher and wider compared to previous years, and, as
mentioned earlier, its aerodynamic load will be significantly reduced in favor
of lower drag and decreased turbulent flow around it (fig. 3.1). The reduced
aerodynamic load will also be due to the elimination of the side winglets that
in the past served to delimit the high-pressure area from the low-pressure area.
The front wing is arguably one of the most important parts of a single-seater
because it is the first part of the car subjected to airflow, and thus its task is
to generate downforce and guide the airflow in the best possible way for the

9
 Figure 3.1: Rear wing of the Aston Martin AMR22-B

remaining part of the car. It will need to be directly connected to the nose,
without utilizing the supports used so far (introduced in the early ’90s), which
led to having a ’raised nose’ car, revolutionizing the airflow passing over the
underbody (fig. 3.2).

Figure 3.2: Front wing of the Ferrari F1-75

The new shape of the underbody has been defined based on wind tunnel testing
results with Computational Fluid Dynamics (CFD) simulation programs (ch.
4.8). With the new technical regulations, the underbody has essentially become
a single large Venturi tunnel (ch. 4.5.1): by doing so, the vertical load (down-
force) of the car is much less influenced by the negative effects generated by the
preceding car. Fig. 3.3 illustrates Ferrari’s interpretation of the regulations.
The International Automobile Federation (FIA) has released a model to the ten
teams participating in the Championship to design the car. It consists of a file
containing a 3D CAD drawing of a set of volumes within which the car must be
’sculpted’ (fig. 3.4).
Within these volumes, designers will be able to shape their prototypes with al-
most complete freedom, or rather, almost. In fact, all the fins, sharp edges, tips,

10
 Figure 3.3: Underbody of the Ferrari F1-75

and appendages, widely used in previous years, will no longer be allowed from
2022. This is because, according to the FIA, the cars appear more aesthetically
pleasing, but the main reason remains that all these aerodynamic details create
vortices. In this regard, the FIA has designed flow deflectors, which all teams
must use (identical for everyone), placed on the upper part of the front wheels.
These also aim to ”comb” the flows disturbed by the rotation of the tires. For
the same purpose, we find other flow deflectors in the area of the rear tires as
well. These are also designed by the FIA and are identical for all teams.

11
Figure 3.4: Volumes under the new regulations

12
Chapter 4

Definitions

4.1 Aerodynamic
Aerodynamics is a branch of fluid dynamics that focuses on studying the dy-
namics of gases, usually air, and their interaction with solid bodies. Solving
aerodynamic problems is by no means trivial, and despite the existence of many
important governing equations, it is still very difficult to accurately describe
the behavior of a fluid. However, nowadays, there are several software pro-
grams capable of studying these aerodynamic flows through simulations, thanks
to which we can develop new shapes that allow us to exploit all the potential
of the fluid to our advantage. There are several parameters that determine the
main characteristics of a fluid and define the potential that can be obtained from
it: velocity, pressure, temperature, and density. These parameters vary greatly
over time and space. The fields of application of aerodynamics are manifold: it
is common to classify aerodynamic problems based on the environment in which
the flow is defined. The first, which will be of interest in this thesis project,
is the so-called external aerodynamics, which studies flows around solid bodies.
Typical applications include determining drag and lift, for example, of an air-
foil (for aeronautical or automotive purposes). Also important is the study of
aerodynamics around a structure in civil engineering or for vehicle design, not
only terrestrial.The second concerns internal aerodynamics, which, as one might
imagine, focuses on studying flows within solid bodies. A classic example of this
application is simulating flows inside an internal combustion engine. Another
crucial aspect to consider is the flow velocity. This, in fact, results in completely
different effects depending on whether the flow in which the body is immersed
travels at a speed greater than the speed of sound (supersonic regime) or if all
speeds within the flow field are lower than the speed of sound (subsonic regime).
The aerodynamic forces (figg. 4.1, 4.2) acting on a solid body in relative motion
to any fluid (such as an aircraft or an automobile) are primarily drag, which
opposes thrust, and lift, which balances with the weight of the body.

13
 Figure 4.1: Forces acting on an airplane

4.2 Lift and Lift Coefficient (Cz )

Lift is the aerodynamic force acting perpendicular to the direction of the flow. It
plays an extremely important role in terms of performance both in the case of a
vehicle where negative lift (directed downward) is desired to maximize (referred
to as downforce) and in the case of an aircraft, where this force allows it to lift
off the ground. In the automotive field, to create lift, the wing was introduced,
a mechanical system capable of generating a significant difference in pressure
between its upper and lower surfaces. This mechanism was introduced for the
first time in the Formula 1 World Championship in the second half of 1968 by
Ferrari (fig 4.3).
The particular shape of this object allows for a flow velocity over the wing’s
upper surface (extrados) that is faster than the flow velocity over the lower sur-
face (intrados), creating a difference in pressure and thus the desired downforce.
At this point, the lift coefficient (Cz ) is defined: it is a dimensionless number
that allows for the calculation of the lift of the solid body in question and is
commonly used, for example, for various airfoil shapes.

L
Cz = . (4.1)
1
2
ρV∞ 2 A
Where ρ [kg m−3 ] represents the density of the fluid the body is moving through,
V∞ [m s−1 ] the velocity at which the body is penetrating it, and A [m2 ] the
frontal area of the body’s surface in contact with the fluid. This value varies
depending on the geometry of the object, the angle of attack, and other dimen-
sionless parameters such as the Reynolds number (Re) and the Mach number
(Ma).

14
 Figure 4.2: Forces acting on a car

4.3 Drag and Drag Coefficient (Cx )

Drag is the aerodynamic force acting parallel to the direction of the flow. This
type of force is caused by four main factors: friction drag, form drag, induced
drag, and wave drag.

4.3.1 Friction drag

This type of resistance is due to the viscosity of the fluid and therefore to
the force that arises between it and the external surface of the body. It often
constitutes the largest portion of the overall drag. In particular, examining the
boundary layer (fig. 4.4), once the fluid encounters the object’s surface, due
to viscosity, and therefore friction, a significant difference in velocity is created
between the flow molecules in contact with the surface (which are stationary)
and those distant from it. Colliding with each other, these molecules influence
the motion of all adjacent particles, and as I move away from the surface, the
velocity magnitude of the particles gradually returns to be more congruent with
that of the flow. The longer the surface develops, the greater the friction at
play, thus increasing the thickness of the boundary layer more and more.
The boundary layer can be of three types: laminar, turbulent, or in transitional
phase. The condition of laminar flow is a characteristic directly dependent on
the velocity and the distance from the leading edge and occurs at low Reynolds
numbers. In the absence of other disturbance factors, increasing the velocity or
the distance traveled by the fluid over the body will result in the transition from
laminar to turbulent flow. Since the friction drag generated between the fluid
and the body is much greater in a turbulent layer due to the loss of momentum

15
 Figure 4.3: Ferrari 312 F1, 1968

Figure 4.4: Boundary layer

of the air molecules caused by mixing and consequent heating, it is clear that
maintaining laminar flow conditions on transport aircraft can lead to signifi-
cant advantages in terms of autonomy and/or fuel savings. For longer distances
traveled by the fluid over the body, it may happen that the pressure gradi-
ent increases to the point of causing flow separation from the body, generating
vortices, recirculation currents, and, in aerodynamic profiles at high angles of
attack, a considerable increase in form drag. To delay the transition and conse-
quently the point of separation, different boundary layer control systems have
been studied, including the removal of the less energetic layer or its energization.
Mentioned techniques include boundary layer suction and blowing.

• Boundary Layer Suction: The technique of boundary layer suction involves

the extraction, along the direction of flow advancement, of the fluid closest
to the body, in order to remove the less energetic layer from the flow and
shift the separation point forward. This is achieved continuously using

16
 microporous walls or discretely, with slots arranged perpendicular to the
fluid motion, connected to a pump that generates the necessary depression
to enable the extraction of the slower particles from the boundary layer.
• Boundary Layer Blowing: An alternative way to delay the separation of
the boundary layer is to inject a high-speed air stream tangentially to the
profile surface into the flow. This increase in momentum again accelerates
the slower particles of the boundary layer that had been slowed down by
wall viscous effects, thus allowing to increase the angle of attack at which
the profile stalls.

4.3.2 Form drag

Form drag, as the name suggests, is due to the shape of the body, the separation
of the boundary layer, and in particular to the force (dissipative) generated due
to the pressure difference. In this case, we can distinguish two types of bodies:
• Bodies where the separation of the boundary layer occurs in very limited
areas and the wake generated is minimal.
• Bodies with blunt shapes where the boundary layer does not remain at-
tached to the body and the wake generated is very large.
Based on what has been said so far, it is clear that the drag will be greater in
blunt bodies.

4.3.3 Induced drag

This type of drag is due to the presence of lift. As explained earlier, on the
upper surface of an airfoil, the pressure is lower compared to the lower surface.
In accordance with the Navier-Stokes equations (4.11), under such conditions,
the flow will tend spontaneously to move to a region of lower pressure (thus on
the upper surface).

4.3.4 Wave drag

Wave drag occurs when traveling at speeds close to the speed of sound (approxi-
mately 343 m/s). These motion conditions lead to the generation of shock waves
that increase aerodynamic drag. In reality, this type of phenomenon begins to
manifest for speeds even slightly lower than the speed of sound, for example for
values of Ma > 0.85 (thus still in a subsonic regime). The subject of this thesis
involves a study conducted at lower speeds, and for this reason, this type of
drag will not be present. Similar to lift, a drag coefficient has also been defined
in this case. It is also a dimensionless number that quantifies the aerodynamic
drag of a solid immersed in a fluid:

D
Cx = . (4.2)
1
2
ρV∞ 2 A

17
Where ρ [kg m−3 ] represents the density of the fluid the body is moving through,
V∞ [m s−1 ] the velocity at which the body is penetrating it, and A [m2 ] the
frontal area of the body’s surface in contact with the fluid. This value varies
depending on the geometry of the object, the angle of attack, and other dimen-
sionless parameters such as the Reynolds number (Re) and the Mach number
(Ma). Both the lift coefficient and the drag coefficient vary their values depend-
ing on the angle of attack, as can be observed in figg. 4.5 and 4.6 below.

Figure 4.5: Variation of Cz and Cx as a function of angle of attack (AoA)

Figure 4.6: Variation of angle of attack (AoA)

Thanks to advances in computational fluid dynamics and the ever-increasing

performance of electronic computers, it is now possible to make a very precise
estimate of various aerodynamic coefficients (including Cx and Cz ). However,
given the considerable cost of simulations, the mathematical complexity of aero-

18
dynamic problems, and the incomplete understanding of all the physical phe-
nomena governing them, the coefficients are still determined empirically through
the use of wind tunnels and scale models (either reduced or natural scale). In
fig. 4.7, you can see the wind tunnel used by Renault with the 2020 prototype
ready for simulation.

Figure 4.7: Wind tunnel

4.3.5 Aerodynamic efficiency

The aerodynamic efficiency, usually indicated by the letter ”E”, is defined as a
combination of the two coefficients in a single and fundamental relationship:

Lif t Cz
E = ef f iciency = = . (4.3)
Drag Cx
In essence, aerodynamic efficiency compares the ability to generate downforce
with the aerodynamic resistance of the body under consideration. In practical
terms, for example, an inefficient wing will generate less downforce, resulting
in higher levels of resistance. Conversely, if the downforce-generating surface
manages to push the vehicle to the ground while achieving low levels of resis-
tance, then it is considered to be a more efficient component. In modern times,
in a car, efforts are made to generate downforce from parts that are less suscep-
tible to resistance, such as the car’s underbody (ground effect), which will be
discussed later.

4.4 Steady flow

A fluid dynamics problem is said to be in steady flow when the velocity of the
fluid, although it may vary from point to point, remains constant over time at

19
each point. For a steady flow, the principle of conservation of mass is verified,
and from it, the equation of continuity is derived, which is given by:

ρ1 V1 S1 = ρ2 V2 S2 . (4.4)
Where the subscript ”1” and ”2” respectively denote the inlet and outlet condi-
tions of any perfectly sealed flow tube with the exterior; ρ represents the density,
V the velocity and S the cross-sectional area.

4.5 Significant Applications

Over the years, the Formula 1 environment has been, and still is, a testing
ground for the development of new technologies. Below, we will delve partic-
ularly into the ground effect and the law governing this technique of creating
downforce.

4.5.1 Venturi Channel

The Venturi effect is nothing but a hydrodynamic phenomenon discovered and
studied by the physicist Giovanni Battista Venturi. The observed physical
phenomenon is the decrease in the static pressure of the fluid as its velocity
increases. Before continuing with the discussion, however, it is necessary to
introduce some mathematical and physical relationships that help to better un-
derstand the concept:

• Flow rate: the flow rate ṁ [kg s−1 ] is a physical quantity that measures
the amount of fluid passing through a section in a certain unit of time.

ṁ = ρV S. (4.5)
Where ρ [kg m−3 ] indicates the density of the fluid in question, V [m s−1 ]
indicates its velocity, and S [m2 ] indicates its cross-sectional area. The re-
lationship is as simple as it is fundamental and forms the basis for a better
understanding of the Venturi effect. From this crucial relationship, we can
already understand that to increase the fluid flow rate, it is necessary to
increase the density of the fluid itself, the velocity, or the cross-sectional
area.
• The total pressure p0 [Pa] of a fluid is composed of two components: the
static pressure ps [Pa] and the dynamic pressure pd [Pa]:

static pressure = ps = p;
1
dynamic pressure = pd = ρV 2 ; (4.6)
2
1
total pressure = p0 = p + ρV 2 .
2

20
 The static pressure (ps ) is the perceptible pressure within the fluid, whether
it is in motion or at rest. It depends on altitude and atmospheric condi-
tions. Without disturbances or induced perturbations, its value coincides
with atmospheric pressure (101325 Pa at sea level). The dynamic pres-
sure (pd ), on the other hand, is the pressure component associated with
the kinetic energy of the fluid when it is in motion. This increases with the
fluid’s flow velocity, as shown by the mathematical formula just presented,
and also increases with an increase in its density. The total pressure (p0 )
is the sum of static pressure and dynamic pressure.

The Venturi channel is nothing more than a duct with a variable cross-section
in which an air current is generated inside (see fig. 4.8).

Figure 4.8: Venturi channel

The fluid passing through the tube in question has its own total pressure value.
From fig. 4.8, it can be observed that the airflow enters from the left through
section ”A” and traverses the entire tube until reaching the exit on the opposite
side of the duct. The gas entering has assumed a total pressure that corresponds
to the sum of the static and dynamic values. This remains constant along the
entire duct due to the lack of new additions or subtractions of energy. As the
gas flows, it encounters a constriction, section ”B,” which acts on the flow by
modifying its pressure and density properties. Indeed, the same quantity of flow
entering section ”A” must also pass through ”B” (as it is a closed and sealed
duct). However, the dimensions of the two sections in question are different
(section ”A” is larger than section ”B”). To pass all molecules through the nar-
rowest point of the channel, they must accelerate. Only in this way is it possible
to maintain the flow equilibrium and thus respect the principle of conservation
of mass:
↑ ↓
ρa Va Sa = ρb Vb Sb . (4.7)

21
At this point, let’s talk again about the variation of fluid pressure during its
flow inside the variable cross-sectional area channel. The first aspect to discuss
concerns the total pressure. As mentioned several times, this remains constant
throughout the Venturi channel. However, the flow velocity varies along the
different sections of the tube. The more the passage narrows, the faster the
flow must pass through. This draws attention to the velocity factor within the
equation:

1
ps + ρV 2 = p0 = constant. (4.8)
2
The velocity increases or decreases depending on the position of the molecules
along the channel. But if the value of the flow velocity increases or decreases
during the passage of the molecules, and the value of the total pressure must
remain constant, what decreases to maintain equilibrium within the formula?
The static pressure:
↑
↓ 1
(ps ) + ( ρV 2 ) = constant. (4.9)
2
Thanks to this balance, the variation in static pressure is obtained: part of it
converts into kinetic energy when the molecules are forced to accelerate, having
to flow through a constriction. This variation will be as evident as the tube
shows differences between the various passage sections. This system of reducing
static pressure, at the expense of dynamic pressure, is widely used and plays a
primary role when applied to the bottom of a racing car. Underneath the car,
in fact, an environment with lower static pressure is created, and therefore the
much-desired downforce is obtained.

4.5.2 Bernoulli principle

The Bernoulli’s principle in fluid dynamics represents a simplified model of a
non-viscous, incompressible flow under steady-state motion.

V2
ps + ρ + ρgh = constant. (4.10)
2
Where ρ is the fluid density, V is the flow velocity, ps is the static pressure
along a streamline, h is the average potential elevation of the section and g is
the acceleration due to gravity. The Bernoulli’s equation describes a phenom-
ena where in an ideal fluid on which no work is applied, for every increase in
drift velocity, there is simultaneously a decrease in pressure or a change in the
potential energy of the fluid (not necessarily gravitational). This relationship is
also able to quantify (within certain limits) the lift generated by the difference
in velocity between, for example, the upper and lower surfaces of a wing. In
fig. 4.9, it is evident how the flow is ”forced” to accelerate on the upper side
to cover a longer path (upper surface) in the same time it takes to cover the

22
shorter path (lower surface) on the lower side. This difference in velocity results
in a difference in pressure.

Figure 4.9: Velocity difference between the intrados and extrados of a wing

4.5.3 Ground effect

The ground effect, thanks to the introduction of the new technical regulations in
2022, has once again become one of the main focuses of study for aerodynamics
engineers, as it is capable of generating a significant amount of downforce while
at the same time only producing a small amount of aerodynamic resistance,
thus allowing for very high aerodynamic efficiency (ch. 4.3.5). In automotive
engineering, ground effect refers to the set of aerodynamic forces exerted by the
layer of air flowing underneath the vehicle’s floor. In vehicles designed with these
principles in mind, the floor and the body of the vehicle are engineered so that
the vehicle as a whole behaves like an inverted wing. In this perspective, the
floor plays the role of the wing’s upper surface. The first time ground effect was
introduced on a car was in the Chaparral Series2 in the CanAm Championship.
The first ground-effect single-seater to compete in Formula 1 was the Lotus − 78
introduced in 1977 (fig. 4.10).
During that era, to maximize the airflow directed to the bottom of the car, sev-
eral techniques were adopted (later prohibited by regulations), such as the use
of side skirts or ”miniskirts” to seal the bottom as much as possible. Referring
to the shapes adopted in racing cars from the mid-1980s to the present day, the
rear diffuser optimizes the airflow underneath, merging it with the airflow that
interacts with the upper part of the vehicle body and the profile of the rear
wing. The purpose of this device is to promote the outflow of air from under
the car to allow the entry of the maximum possible amount of fresh air from
the front, thus generating the greatest possible ground effect. This phenomenon
is a direct consequence of the principles of Bernoulli and the Venturi tube pre-
viously discussed. By designing the bottom of a car like a Venturi tube and
bringing it as close as possible to the ground, it is possible to ensure that the air

23
 Figure 4.10: Lotus-78

pressure passing under the car is lower than that passing over it, thus creating a
depression between the bottom and the asphalt. This depression presses the car
to the ground, thereby creating downforce. In fig. 4.11, in the lower detail, the
bottom shaped like an inverted wing is observed, along with how the pressure
distributes around it.

Figure 4.11: Venturi channel and underbody

4.6 Navier-Stokes equations

In fluid dynamics, the Navier-Stokes equations are a system of three balance
equations (four if the continuity equation is included) of continuum mechanics
that describe a nonlinear viscous fluid:

24
 ∂ 2 vx ∂ 2 vx ∂ 2 vx

∂vx ∂vx ∂vx ∂vx ∂p
ρ( + vx + vy + vz )=− + µ( + + ) + bx ;


2 2




 ∂t ∂x ∂y ∂z ∂x ∂x ∂y ∂z
∂vy ∂vy ∂vy ∂vy ∂p ∂ 2 vy ∂ 2 vy ∂ 2 vy



 ρ(

 + vx + vy + vz )=− + µ( + + ) + by ;
∂t ∂x ∂y ∂z ∂y ∂x2 ∂y 2 ∂z
2 2 2
(4.11)
 ∂vz ∂vz ∂vz ∂vz ∂p ∂ vz ∂ vz ∂ vz
ρ( + vx + vy + vz )=− + µ( + + ) + bz ;


2 2




 ∂t ∂x ∂y ∂z ∂z ∂x ∂y ∂z
∂vx ∂vy ∂vz


+ + = 0.



∂x ∂y ∂z
Where ρ is the fluid density, v is the fluid velocity, p is the pressure and µ
the dynamic viscosity. The mathematical treatment of this behavior can only
occur under three specific conditions: it is assumed that the fluid is continuous
(although in reality, matter should be considered discontinuous as it is composed
of discrete particles), homogeneous, non-reactive, and without electric charges
within it. In them are introduced, as constitutive laws of the material, the Stokes
Law (in the kinematic balance) and the Fourier Law (in the energy balance).
The equations are named after Claude-Louis Navier and George Stokes.

4.7 Vorticity
In fluid dynamics, vorticity is a measure of the local rotation of a fluid element.
It describes the tendency of fluid elements to rotate about their local axis.
Vorticity is a highly complex fluid dynamic phenomenon to describe, yet it is
also extremely important. When utilized correctly, it can lead to significant
performance improvements for a vehicle. In fluid dynamics, vorticity is defined
as the following vector quantity:

⃗ × V⃗
ω=∇ (4.12)
Where ω represents the velocity rotor. The surface density of circulation:

∂ 2Γ
ω= (4.13)
∂r2
In which Γ precisely represents circulation, namely the circulation of a velocity
field along a closed path.
I
Γ= V⃗ · ⃗t dS (4.14)
∂S
Where ⃗t indicates the direction of the tangent vector to the trajectory.
Vorticity is also related to the rotational velocity of a fluid element, and in
particular, it is demonstrated by studying the deformation that is equal to
twice its instantaneous rigid rotation rate (Ω):

ω = 2Ω (4.15)

25
Vortices are nothing more than flows that move along a helical trajectory. This
characteristic can be exploited to keep the flow attached to the surface around
which it flows and thus, for example, to prevent, where possible, the phenomenon
of separation of the boundary layer (ch. 4.3.1).

4.8 Computational Fluid Dynamic (CFD)

Computational Fluid Dynamics (CFD) is a branch of fluid mechanics that,
through numerical analysis and data structures, analyzes and solves problems
involving fluid flows. Computers are used to perform the calculations necessary
to simulate the free flow of the fluid and its interaction (liquid or gaseous)
with surfaces defined by boundary conditions. Nowadays, there are high-speed
computers capable of solving even very complex problems. There are numerous
ongoing research efforts focused on developing software to improve the accuracy
and speed of calculation for complex simulation scenarios such as transonic or
turbulent flows. To validate such software, data from wind tunnel simulations
(but also larger-scale tests, such as flight tests) are typically collected and then
compared to those provided by the software under development. Of course,
under the same model and environmental parameters. The most well-known
and widely used CFD software include: ANSYS CFX, Fluent, Kiva, NUMECA,
Phoenics, and other open-source software like OpenFOAM and Code Saturne.
The primary use of CFD is to solve 4.11 and the equations related to them.
However, solving these analytical expressions is feasible only in simple cases with
laminar flows (low Reynolds numbers) and simple geometries (spheres, cylinders,
etc.), whereas in real-world scenarios, where turbulent flows are very common
(high Reynolds numbers), numerical approaches are necessarily required. There
are several methods to solve the equations number 6, and since these are usually
computationally expensive procedures, three main models have been developed
to address the problem:

• Direct Numerical Simulation (DNS): This approach provides the most ac-
curate calculations by solving the equations without turbulence models.
Time and space are discretized with grids of user-defined sizes, and calcu-
lations are then performed on them. However, the cost of this process is
extremely high and requires the use of supercomputers. For this reason,
it is rarely used in industrial applications, except for studying the most
fundamental aspects of turbulence.

• Reynolds-Averaged Navier-Stokes equations (RANS): This method is based

on the assumption that turbulent motion can be decomposed into a mean
motion and its fluctuation over time (the fluctuations are modeled, and
there are many types differing in their accuracy). This method requires
significantly reduced computation times because the scales of the mean
motion are much larger than those of the turbulent motion. All results
obtained in this document stem from analyses using the RANS method.

26
 • Large Eddy Simulation (LES): In this approach, the numerical compu-
tation focuses on the behavior of the largest turbulent scales while also
appropriately modeling the smaller scales. This is achieved using specially
designed numerical filters. The results provided by LES are more accu-
rate than those of the RANS method and furthermore, it has a much lower
computational cost compared to DNS.
Depending on the cases studied, it is often necessary to introduce the resolution
of additional equations in addition to those already mentioned, such as the
Navier-Stokes equations and the turbulence model equations. This occurs, for
example, in cases where multi-component flows, reactive flows, two-phase flows,
non-Newtonian flows, and so on are considered.

4.8.1 Polygonal mesh

The term ”mesh” comes from English and translates to ”maglia” or ”rete”
in Italian; in fact, the polygonal mesh is a grid of polygons that defines the
contours, vertices, and faces of a three-dimensional object in space. In recent
years, its use has been increasingly widespread in many areas, especially in
computer graphics and geometry processing. A polygonal mesh (fig 4.12) is the
partitioning of a continuous surface into polygonal cells (usually triangles or
quadrilaterals).

Figure 4.12: Polygonal mesh

4.9 Post-Processing
Post-processing is currently one of the most effective methods used to develop
complex aerodynamic devices. When utilized effectively, and reliable methods
are used in conjunction with CFD calculations with a sufficient number of cells,
good mesh quality, and a reliable turbulence model, valuable data and informa-
tion regarding the flow behavior around the object under study can be obtained.

27
Through post-processing, simulation results and the aerodynamic performance
of an object can be examined quickly and efficiently.

4.9.1 Ansys CFX

Ansys CFX is a commercial software by ANSYS for computational fluid dy-
namics (CFD), used to simulate fluid behavior in a wide variety of applications,
compatible with Unix, Microsoft Windows, and Linux operating systems. Ex-
amples of simulations include: the flow of seawater under a ship’s hull, the
behavior of a turbo-machine, aerodynamics of an aircraft, HVAC systems, ag-
itated vessels, and cyclones. Fluent and EnSight are two commercial software
products by Ansys for CFD widely used in many industries and academic fields,
based on the finite volume method. They are associated with connected soft-
ware Gambit and TGrid for grid generation. Currently, they are among the
world’s leading software in the CFD market.

4.9.2 OpenFoam
OpenFoam (Open Field Operation And Manipulation) is primarily a C + +
toolbox for customizing and extending software solutions for simulation. It is a
solver based on the theory of continuum mechanics that includes computational
fluid dynamics. It comes with an extensive library of solvers, continuously
expanded, applicable to a wide range of problems. OpenFoam is one of the
earliest scientific software written in C + +. There are three main versions of
OpenFoam, all free and released under the GPL license:

• The official version of OpenFOAM released by ESI-OpenCFD Ltd.

• The fork of OpenFOAM-Foundation released by CFD Direct Ltd.
• The fork of the FOAM-Extend project released by Wikki Ltd.

In addition to these main versions, there are several other variants maintained
and released for specific applications by other entities. It has been produced by
the British company OpenCFD Ltd since 2004, and in 2007 the name OpenFoam
was registered as a trademark, guaranteeing the use of its OpenFoam trademark
to third parties through licensing. OpenFoam successfully competes with the
functionalities of major commercial CFD software packages. It relies on users’
choice of third-party pre and post-processing utilities and is provided as:

• A plug-in (ParaFoam) for visualizing finite data and mesh in ParaView.

• A wide variety of converters for three-dimensional grids allowing import
from a range of major commercial packages.

• An automatic hexahedral mesher to create polygonal grids for engineering

configurations.

28
4.9.3 ParaView
ParaView is a cross-platform open-source application for interactive scientific vi-
sualization. It has a client-server architecture to facilitate remote visualization
of data sets and generates level-of-detail (LOD) models to maintain interac-
tive frame rates for large data sets. It is built upon the Visualization Toolkit
(VTK) libraries. ParaView is designed for data parallelism on multi-computer
and shared-memory or distributed-memory cluster systems. It can also run as
a standalone application on a single computer. It is well-known and widely
used in many communities for analyzing and visualizing scientific data. Par-
aView can be used to create visualizations and analyze data using qualitative
and quantitative techniques. Data exploration can be performed interactively
in 3D or using ParaView’s batch processing capabilities. This software has been
developed to analyze extremely large data sets using distributed memory com-
puting resources. It can be run on supercomputers for large-scale data analysis
or on laptops for smaller data sets. ParaView is both an application framework
and a turnkey application. The code base is designed so that all its components
can be reused to quickly develop vertical applications. This flexibility allows
developers to rapidly create applications with specific features for a particular
problem domain. It can be installed on single-processor and distributed-memory
parallel systems. It has been successfully tested on Windows, macOS, Linux,
IBM Blue Gene, Cray Xt3, and various Unix workstations, clusters, and su-
percomputers. In figg. 4.13 and 4.14, the first study conducted on ParaView
software is presented as an attempt to approach the world of CFD.

Figure 4.13: Example of studying an ogive using ParaView software, front view

29
Figure 4.14: Example of studying an ogive using ParaView software, back view

30
Chapter 5

Tools

5.1 Streamlines
In fluid mechanics, a streamline is defined as a curve that is tangent at every
point to the velocity vector of the flow. Along with many other tools, these have
been used throughout the study in this thesis to better visualize and describe
certain phenomena that will be presented later. It is of fundamental importance,
as already mentioned in the introduction, to emphasize that we are referring to
a steady-state study (ch. 4.4) and in a subsonic regime. In the case of non-
steady flow, indeed, the velocity vector changes from moment to moment, and
the streamlines (fig. 5.1) would have only an instantaneous meaning. In this
latter case, one would also proceed with the study of other flow visualization
tools such as pathlines and streaklines (in the steady-state case, streamlines,
pathlines, and streaklines are coincident). Regarding the subsonic regime, it is
important to clarify that speeds close to the speed of sound are never reached.
However, locally, velocity magnitudes 2 or 3 times higher than those in the inlet
conditions can be experienced. This is because, for example, a fluid passing over
an airfoil tends to undergo acceleration as it passes over its upper surface.

5.2 Oil-Film
Oil films are a highly utilized tool in the post-processing phase as they effectively
reproduce streamline patterns around a body. To visualize them, a special paint
(which solidifies only after a certain testing period) is applied to a body and then
submerged in a flow. Once solidified, the ”frozen” streamlines can be visualized
directly on the body under examination. In fig. 5.2, some oil films applied to
the RB − 19 during a test run can be observed. One might argue that the fluid
velocity near the wall is non-zero (ch. 4.3.1). Well, for this reason, in the models
to be studied throughout the treatment, we had to move away from the wall by
5 mm.
The oil film technique is also used because it is minimally invasive: it can be

31
 Figure 5.1: Streamlines on the front part of the F14-T, top view

used both in wind tunnels (fig. 4.7) and in the field (which allows obtaining
very significant results) and provides important information on the wall shear
stress on a surface. The wall shear stress is one of the elementary stresses to
which a body can be subjected, measured in Pa, and typically indicated by the
symbol τw . To calculate the magnitude of the tangential stresses within a fluid,
reference is made to two parallel ”sheets” perpendicular to the wall, one fixed
and one mobile, separated by a differential thickness normal to the wall (fig.
5.3). According to Newton’s law, the tangential stress is equal to:

∂u
τw = −µ (5.1)
∂y
In which µ is the dynamic viscosity of the fluid, ∂u is the velocity difference
between the two fluid layers along the i direction, and ∂y is the position differ-
ence between the two fluid layers along the j direction perpendicular to i. To
trace the Oil-Film, ideally, one should consider the calculation of τw at the fixed
plane (j = 0).

5.3 Pressure maps

Pressure maps are highly effective tools for better understanding the effects and
causes of certain flow behaviors around a body through the variation of specific
coefficients, as will be discussed later. To create a pressure map, one needs to
define a surface (usually flat) that intersects the domain containing the model
under study. After positioning the model at the desired location on this plane,
the variable to be visualized is represented. Often, to enhance visualization,

32
 Figure 5.2: Application of oil film on the RB-19

Figure 5.3: Wall-Shear-Stress

several factors are adjusted, such as color/quantity of colors, coloring (whether

continuous or banded), the range of variation of the variable, etc.

5.3.1 Pressure coefficient (Cp )

The pressure coefficient, in fluid dynamics, is a dimensionless quantity that
describes the relative pressure in a flow field. This coefficient is widely used
in both aerodynamics and hydrodynamics. Each point immersed in a flow has
its own unique pressure coefficient (Cp ). In a steady flow field around a body,
assuming the fluid is incompressible (meaning its density remains constant),
particles flow along one of the many streamlines, decelerating from the va value
of the asymptotic streamline to zero at the stagnation point. The ”stagnation
point” is a point in the flow field where the local instantaneous velocity is zero.

33
This point is located on the surface of the object, in the flow field, where it
interrupts the flow. The Bernoulli Equation shows that static pressure is at
its maximum value when the local velocity is zero, indicating that pressure
is highest at stagnation points. This static pressure, in this case, is called
stagnation pressure. At this point, the particles diverge to form an array of
streamlines, centered at the stagnation point, to continue their path. In the
motion of bodies in fluids, what matters more than the level of pressure is the
pressure difference (ps − ps ∞ ), where ps ∞ is the static pressure at an infinitely
distant asymptotic point and ps is the static pressure at the point where the
measurement is made. The pressure coefficient can then be defined as:
ps − ps ∞
Cp = 1 (5.2)
2
ρV∞2
Wherein the previously mentioned difference is dimensionless with respect to
the dynamic pressure at the asymptotic point. It follows that at the stagna-
tion point, the pressure coefficient is equal to 1, which also corresponds to the
maximum value that Cp can assume.

5.3.2 Total pressure coefficient (Cp tot )

The total pressure coefficient in fluid mechanics represents a value capable of
establishing the energy content of the fluid at a specific point. The relationship
that allows determining this value is:

p0 − ps
Cp tot = 1 (5.3)
2
ρV∞2
Where p0 and ps are both referred to upstream conditions, i.e., at a point far
enough from the model to not be affected by its effects. In the denominator,
however, is the dynamic pressure of the point where the measurement is being
taken. So, it is a ratio of two dynamic pressures.

5.4 Q-Criterion
The Q-Criterion is a very useful tool used for visualizing vortical structures.
With Q, we refer to the second invariant of the velocity gradient ∇V⃗ . This pa-
rameter defines a ”vortex” as a region where every point has a value of Q > 0.
In particular, on the studied models, three-dimensional surfaces can be repre-
sented, identifying regions with Q greater than a user-defined value.

34
Chapter 6

Cases examined

In this chapter, we aim to present the three models that have been analyzed
throughout the project. As mentioned in the early pages, these are models
designed and studied in previous thesis projects by students of the University
of Genoa. However, here we intend to revisit and delve deeper into some new
and interesting aspects.

6.1 Isolated Model

The first model is undoubtedly the simplest of the three presented in this project
(figg. 6.1 and 6.2). It consists only of the bottom without any type of aerody-
namic disturbance in the front section. Therefore, we will have a ”clean” flow
entering the bottom, which will then generate the desired effects that will be
examined in more detail in ch. 7.

Figure 6.1: Bottom view of the isolated case

35
 Figure 6.2: Entrance of the bottom of the isolated case

6.2 Intermediate model

The intermediate model (figg. 6.3 and 6.4) includes the ”simple” bottom as seen
before, but in addition, two elements (wheel and front wing) are inserted at the
front, before the flow enters the bottom. These elements, as will be observed,
cause various aerodynamic disturbances.

Figure 6.3: Bottom view of the intermediate case

36
 Figure 6.4: Front view of the intermediate case

6.3 Complete model

The third model discussed (figg. 6.5, 6.6) is already more complex as it is
equipped with fences, which are three flow deflectors clearly visible at the en-
trance of the bottom. The purpose of these elements is to regulate the flow in
the car bottom to maximize the downforce effect generated.

Figure 6.5: Bottom view of the complete case

37
Figure 6.6: Front view of the complete case

38
Chapter 7

Results and observations

7.1 Stagnation point

The stagnation point is a very particular point that identifies areas on the
body surface where the instantaneous velocity of the flow field is zero. The
stagnation point causes a true interruption of the flow, and at this point, as
shown by the Bernoulli equation (ch. 4.5.2), the static pressure reaches its
maximum value, called ”stagnation pressure.” For incompressible flows, it can
also be noted that the stagnation pressure corresponds to the total pressure,
as the dynamic pressure is zero. Additionally, at the stagnation points, the
pressure coefficient will reach its maximum value (typically shown in red), and
the Oil-Film will tend to diverge with a radial arrangement centered at the
stagnation point (figg. 7.1 and 7.2).

Figure 7.1: Cp and Oil-Film in the isolated case

39
 Figure 7.2: Cp and Oil-Film in the complete case

One might spontaneously wonder why Oil-Films appear so visible and pro-
nounced if they are nothing more than lines that reproduce the path of flow
velocity around the body. What happens to the boundary layer? Well, in re-
ality, the flow velocity at the body’s surface is zero, precisely because of the
boundary layer. To visualize the Oil-Films, it was necessary to move away from
the walls by 10 mm.

7.2 Separation and reattachment point

The local pressure at a given point in the boundary layer on the surface of
the body is slightly lower than the pressure slightly downstream. This slight
downstream pressure may cause the intrusion of downstream pressure beneath
the boundary layer, leading to its detachment. If this separation occurs in (or
under) a laminar boundary layer, it is called laminar separation. In this sce-
nario, if the boundary layer transitions from laminar to turbulent flow (termed
turbulent boundary layer), it consists of numerous small vortices. These small
vortices in the turbulent boundary layer have higher kinetic energy compared to
the laminar boundary layer’s laminar filaments, making them more resistant to
the intrusion of adverse pressure gradient from downstream and more resistant
to separation (if separation occurs, it will happen further downstream). Sepa-
ration (or detachment) of the boundary layer due to pressure gradient is not an
inertial phenomenon because the boundary layer lacks inertia (it is not moving
fast enough) and cannot withstand the intrusion of adverse pressure gradient.
Sometimes, after separating from the body’s surface, the boundary layer may
reattach at a point known as the ”reattachment point”. Occasionally, the area
between the separation point and the reattachment point is referred to as the
”separation bubble” or ”turbulent bubble”.

40
7.3 Slipstream
One of the most immediate methods to visualize the wake released behind a
body, in this case at the entrance of the vehicle floor, is to display pressure
maps on planes at x = const immediately at the entrance of the vehicle floor
(figg. 7.3, 7.4 and 7.5), using the total pressure coefficient as a parameter (figg.
7.6, 7.7 and 7.8).

Figure 7.3: Plane at x = cost at the entrance of the bottom in the isolated case

Figure 7.4: Plane at x = cost at the entrance of the bottom in the intermediate
case

By drawing pressure maps on these planes, the following results are obtained:
From the various colors observed, different zones can be distinguished: a high-

41
Figure 7.5: Plane at x = cost at the entrance of the bottom in the complete
case

energy density zone characterized by a color tending towards red where the
total pressure coefficient tends to unity, a low-energy density zone characterized
by a color tending towards blue where the total pressure coefficient tends to
negative unity, and a transitional phase tending towards green where the total
pressure coefficient tends to zero. To quantify the incoming wake into the floor,
a parameter called ”blockage” is defined as the ratio between the area where
Cptot < −0.5 (Aslipstream ) and a reference area (Atot ), the latter being the same
for all three models. The value is then expressed in percentages, multiplied by
one hundred.

Blockage Isolated Intermediate Complete

Aslipstream
Atot 100 0% 19.57% 14.71%

The fact that in the ”isolated” case the blockage turns out to be zero is not sur-
prising since the fluid encounters no obstacle before entering the floor. However,
upon analyzing the ”intermediate” model and subsequently the ”complete” one,
it is noticeable that the introduction of fences has a decidedly beneficial effect.
As the flow is in subsonic regime, these fences alter the wake even upstream of
them, reducing it by about five percentage points. At this point, we can fur-
ther analyze the floor to see how the wake caused by the wheel and front wing
affects the pressure distribution beneath it. To do this, we utilize the pressure
coefficient this time (figg. 7.9, 7.10 and 7.11).
From the various colors observed, three distinct zones can be distinguished: a
low-pressure zone characterized by a dark blue color where the pressure coeffi-
cient tends towards negative unity, a high-pressure zone depicted by a red color
where the pressure coefficient tends towards positive unity, and a transitional
phase represented by a green color where the pressure coefficient tends to nullify.

42
 Figure 7.6: Pressure map with Cp tot as a parameter in the isolated case

Similarly, to quantify the distribution of depression on the floor, a parameter

called ”depression” has been defined. It is calculated as the ratio between the
area with Cp < −0.5 (Adep ) and the total area of the floor (Atot ), the latter
being the same for all three cases. The value is then expressed in percentages,
multiplied by one hundred as in the previous case.

Depression Isolated Intermediate Complete

Adep
Atot 100 84.24% 38.70% 66.43%

As one might expect, in the ”isolated” model, the air entering the underbody
is relatively less turbulent, resulting in significantly higher depression values
compared to the ”intermediate” model, where dirty air enters due to the wake
released by the front wing and the wheel. However, when the fences are added,
the flow tends to partially stabilize. Therefore, in the ”complete” model, a
significant portion of the depression lost is recovered.

7.4 Velocity and pressure variation

To better understand how the pressure on the underbody depends strongly on
the flow velocity, a new dimensionless parameter, Vadim , is defined as the ratio
between the magnitude of the velocity at a certain point in the domain and the
magnitude of the velocity under inlet conditions (55 m/s), both squared. As
explained earlier, with the new regulations of the 2022 F1 World Championship
(ch. 3), the ground effect returns. To create the much-desired downforce, the
flow must be accelerated. By doing so, the dynamic pressure will increase at
the expense of the static pressure (figg. 7.12, 7.13 and 7.14).
In all three cases, it is evident that as the flow velocity increases, there is an

43
 Figure 7.7: Pressure map with Cp tot as a parameter in the intermediate case

increase in depression and consequently a negative pressure coefficient, indicated

by the blue color, tends to dominate.

7.5 Vortices
From fig. 7.11, it can be observed how a long zone of low pressure is created
immediately after the innermost fence. This could imply the presence of a vortex
since vortices typically develop approximately in the direction of the flow and
can indeed cause areas of low pressure. To verify its actual existence, one can
investigate by tracing streamlines (fig. 7.16), starting from where it is believed
the vortex originates. In this case, they were initiated from a flat surface just
after the innermost fence (fig. 7.15).
It is worth noting that the streamlines were also colored according to the flow
velocity at that point, following a scale ranging from 35 m/s (blue) to 70 m/s
(red). Consistent with what was mentioned earlier, there will also be more red
hues in points with higher downforce. At this point, one can already begin to
confirm the actual presence of a vortex. In addition to the previously listed
clues, one can now also observe the classic spiral shape that the streamlines
tend to form. However, to further highlight the presence of turbulence, one can
analyze the velocity vector components, the boundary layer, the Oil-Film, and
the Q-Criterion.
By slicing the underbody with longitudinal planes and plotting the tangential
components of the velocity vector on these planes, it can be seen the result in
fig. 7.17:
In this way, besides the evident presence of turbulent motion, it is also possible to
infer another important piece of information, which is the direction of rotation
of the vortex. The boundary layer in the two cases (isolated and complete)

44
 Figure 7.8: Pressure map with Cp tot as a parameter in the complete case

appears to be significantly different, as shown in figg. 7.18 and 7.19

It can be observed that the boundary layer in the complete case is much less
pronounced and more uniform. The biggest difference is found in the rear part
of the underbody. While in the isolated case it becomes thicker and thicker,
in the complete case, due to the vortex, the thickness increases with a lower
velocity gradient. This happens because the vortex, by rotating, energizes the
flow even near the wall without giving time for the boundary layer to expand.
This is in contrast to what happens in the isolated case due to the absence of
significant turbulence (figg. 7.20 and 7.21).
If on the same vector plane, as seen previously, instead of projecting the tangen-
tial components of the flow velocity modulus, only the Y component is projected,
it will be observed that, at the first special point, the flow tends to split into
two due to the presence of the obstacle (the bottom wall), as represented in fig.
7.22.
In the black circled area, it is highlighted the point where the vortex, while ro-
tating, impacts against the lower wall of the floor and splits into two: one part
continues to rotate counterclockwise and therefore will have a positive Vy (red
color), while the other part splits and proceeds in the opposite direction with
a negative Vy value (blue color). The Oil-Film, based on their arrangement,
can also suggest the presence of a vortex. Indeed, the vortex, while rotating,
determines two specific points. The first one where it impacts against the wall
and the second one where it moves away from it (fig. 7.23).
At the points where the flow separates, there will be a more ”dirty” area with
many Oil-Films, as they tend to converge and accumulate, creating the tradi-
tional darker line (black circle). However, at the opposite point of the vortex,
there will be an exactly opposite situation: here, the fluid reattaches, and, un-
like before, the Oil-Films tend to diverge, highlighting the traditional ”clean”
area where a lower accumulation is observed (red circle).

45
 Figure 7.9: Pressure map with Cp as a parameter in the isolated case

Finally, as the last tool to highlight a vortical zone, the Q-Criterion was em-
ployed:
From figg. 7.24 and 7.25, once again, the presence of the vortex emerges. It
is noteworthy how this tool is capable of better highlighting the entire path
of the vortex being discussed and how it manages to highlight many others,
more or less significant. It is particularly important to adjust the scale properly
to visualize the vortices correctly and avoid cluttering the model with small
turbulences that would make visualization very difficult. Always observing the
figures, it is clear how the vortical zone is characterized by a negative Cp and
relatively low velocities.

46
Figure 7.10: Pressure map with Cp as a parameter in the intermediate case

Figure 7.11: Pressure map with Cp as a parameter in the complete case

47
 Figure 7.12: Cp on the wall and Vadim on the plane in the isolated case

Figure 7.13: Cp on the wall and Vadim on the plane in the intermediate case

48
 Figure 7.14: Cp on the wall and Vadim on the plane in the complete case

Figure 7.15: Arrangement of the plane from which the streamlines originate

49
 Figure 7.16: Stremlines

Figure 7.17: Tangential component on the cutting plane of the velocity vector
and Vadim on the cutting plane

50
Figure 7.18: Boundary layer in the isolated case

Figure 7.19: Boundary layer in the complete case

51
Figure 7.20: Boundary layer at the rear of the bottom in the isolated case

Figure 7.21: Boundary layer at the rear of the bottom in the complete case

52
 Figure 7.22: Y -component of velocity in the complete case

Figure 7.23: Cp and Oil-Film on the bottom of the complete case

53
 Figure 7.24: Q-Criterion with Cp as a parameter in the complete case

Figure 7.25: Q-Criterion with the velocity as a parameter in the complete case

54
Chapter 8

Conclusions

Based on what has been shown throughout this project, it has been possible
to observe and experience some of the aerodynamic phenomena that are often
encountered during post-processing. The three models presented have demon-
strated different characteristics and, therefore, different performances in prac-
tice. In particular, it has been highlighted how much greater the downforce
effect is in the ”isolated” model compared to the ”intermediate” model, not to
mention the blocking effect, which is null in the former for obvious reasons...
At this point, a further comparison was made with the ”complete” model, which
highlighted how the introduction of three ”simple” flow deflectors (fences) can
positively influence the previously mentioned values: both in terms of blocking
and depression, significant improvements have been observed. In the first case,
there was an improvement of almost five percentage points, while in the second
case, it was nearly twenty-eight.
However, the introduction of the fences has also led to the formation of an im-
portant vortex (widely discussed and visualized through various tools) which, if
managed correctly, can lead to further performance improvements.

55
Ringraziamenti

Desidero ringraziare tutti i miei familiari. In particolare miei genitori e mio

fratello per aver reso possibile il raggiungimento di questo traguardo.
Ringrazio Camilla per essere stata al mio fianco e per aver sempre creduto in
me.
Ringrazio Anna, Giacomo, Annalisa, Silvia, Elena, Pietro, e tutte le persone
che ho incontrato lungo questo cammino che si sono rivelati essere, oltre che dei
compagni di corso, dei grandi amici.
Ringrazio il mio relatore e i miei correlatori che mi hanno dato la possibilità di
prendere parte a questo progetto di tesi che tanto desideravo.
Ringrazio infine Christian per avermi aiutato a fare i primi passi in un campo
che conoscevo ancora poco.

56
Bibliography

[1] Turbulence. introduction to theory and applications of turbulent flows.

Springer, 2, J. Westerweel F. Nieuwstadt, B. Boersma.

[2] Capire la Formula 1. Minerva, 2022, Marco Giachi e Mauro Forghieri.

[3] 2022 formula 1 technical regulations,
https://www.fia.com/regulation/category/969, Accessed: 30 March 2020.
[4] Web page, https://www.wolfdynamics.com/.

[5] Two-equation eddy-viscosity turbulence models for engineering applications.

AIAA Journal, 32:1598–1605, 1994, F.R.Menter.
[6] The aerodynamics of an isolated wheel rotating in contact with the ground.
Thesis for the degree of Doctor of Philosophy in the Faculty of Engineering,
1974, John Edward Fackrell.

[7] Optimization Methods:From Theory to Design. Springer,2013,

Marco Cavazzuti.
[8] Ansys fluent gradient-based optimization. 2020, ANSYS.

[9] Adjoint Navier-Stokes Methods for Hydrodynamic Shape Optimization.

TUHH, 2012, Arthur Stuck.
[10] Race Car Aerodynamics Designing for Speed (Engineering and Perfor-
mance), 2003, Joseph Katz.
[11] Capire l’aerodinamica delle auto da corsa con semplici esempi (2020), Al-
berto Aimar.
[12] The Free Encyclopedia of Wikipedia,
[13] Race Tech, Motorsport Engineering, Willem Toem.
[14] Nuovo Regolamento F1 2022, RED, Federico Marino.

[15] F1, il nuovo regolamento 2022: aerodinamica, OAsport, Francesco Paone.

57
[16] Monoposto con effetto suolo, caratteristiche e nuovo regolamento F1 2022,
News Auto, Manuel Cerfeda.
[17] Le grandi innovazioni della Formula 1: l’effetto suolo, FunoWorld, David
Bianucci.

[18] Analysis of oil film interferometry implementation in non-ideal conditions,

2010, Kevin Chen.
[19] Analisi Tecnica F1, Aerodinamica: Turbolenze, vortici, ala anteriore,
F1ingenerale.com, Cosimo Piazzolla.

58