SAE TECHNICAL

PAPER SERIES 983035

Comparison of V10 and V12 F1 Engines

Alberto A. Boretti
Fiat Research Center, Engines Dept.

Giuseppe Cantore
University of Modena

Reprinted From: 1998 Motorsports Engineering Conference Proceedings

Volume 2: Engines and Drivetrains
(P-340/2)

Motorsports Engineering
Conference and Exposition
Dearborn, Michigan
November 16-19, 1998

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760
The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the
paper may be made for personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc.
Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sec-
tions 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as
copying for general distribution, for advertising or promotional purposes, for creating new collective works,
or for resale.

SAE routinely stocks printed papers for a period of three years following date of publication. Direct your
orders to SAE Customer Sales and Satisfaction Department.

Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.

To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
other works, contact the SAE Publications Group.

All SAE papers, standards, and selected

books are abstracted and indexed in the
Global Mobility Database

No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
permission of the publisher.

ISSN 0148-7191
Copyright 1998 Society of Automotive Engineers, Inc.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in
SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group.

Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 983035

Comparison of V10 and V12 F1 Engines

Alberto A. Boretti
Fiat Research Center, Engines Dept.

Giuseppe Cantore
University of Modena

Copyright © 1998 Society of Automotive Engineers, Inc.

ABSTRACT journal, where connecting rods from opposing cylinders

are paired. Each camshaft has a number of bearings
The paper compares 3.0 liter F1 engines having different equal to the number of main bearings. Usually camshafts
architectures and developed in compliance with the 1998 directly operate inverted bucket-type valve tappets with
FIA Technical Regulations. Similarity rules and non shim adjustment. Pneumatic systems are used for valve
dimensional parameters from previous projects define motion control. Piston is aluminum alloy, short skirt and
key geometric and operating parameters for V10 and V12 carry two small, light rings. Some sort of piston cooling,
engines having equal degree of sophistication. The paper as spray on the underside of the piston crown, is used.
presents computed classical engine outputs versus Cylinders have a pent roof, four valves per cylinder, cen-
engine speed, including brake, indicated and friction val- tral spark plug, head surface, and a flat crown piston sur-
ues. The V12 solution shows clear advantages in terms face with valve clearance notches. Tumbling vortices
of pure engine performances. characterize the in-cylinder flow field, as tumble is an
integral part of the four valves operation. However, dump
INTRODUCTION ports, where tumble is avoided, are usually preferred to
tumble ports, where tumble is enhanced.
FIA Technical Regulations for F1 engines [1] set very few The engine management system controls injection, igni-
constraints to designers. These rules prescribe four- tion, a fly-by-wire throttle and active variable length trum-
stroke engines, with reciprocating pistons only, and an pet (with closed loop control and engine and gearbox
engine capacity not exceeding 3.0 liters. The maximum control units interfaced). Lambda sensors constantly feed
number of cylinders is twelve. The normal section of each the engine control unit information on the state of com-
cylinder must be circular. Engines may have no more bustion as revealed by exhaust gas composition. Individ-
than five valves per cylinder. Supercharging is forbidden. ually running solenoid injectors and one coil per plug
Any device decreasing the temperature of the intake air ignitions accurately control fueling and spark. Engines
or of the charge of the engine is prohibited. Internal or are designed to act as a stressed member of the chassis
external spraying of water or any substance other than (block stiffness is important for car performances,
fuel for the normal purpose of combustion is prohibited. because it provides a stiffer connection between chassis
Variable geometric length exhaust systems are forbidden. and gear box).
The basic structure of crankshaft and camshafts must be
made from steel or cast iron. Pistons, cylinder heads and Over the last 30 years, the power of 3.0 liter F1 engines
cylinder blocks may not be composite structures that use have virtually doubled. This has been the result of higher
carbon or aramid fiber reinforcing materials. engine speed and lower frictional losses. Ford Cosworth
F1 engines' data [19] clearly show this tendency. The 3.0
Engines developed in compliance with these technical liter normally aspirated V8 engine DFV of 1967 was deliv-
regulations follow some common, general guidelines. ering maximum power with a brake mean effective pres-
They have an aluminum alloy structure. Block, heads, sure mepb of about 14.0 bar at 8,500 Rpm. The DFY of
cam covers and lower crankcase are cast in that material. 1983, the final version of the DFV as a 3.0 liter engine
The block carries separate Nikasil coated aluminum alloy before the turbo revolution of the early eighties, was
liners. The head carries bronze valve seats and bronze developed to give maximum power with about 14.0 bar
valve guides. Camshafts and crankshaft are machined mepb at 11,200 Rpm. Following the introduction of 3.5
from solid steel. Connecting rod and valve are titanium. liter normally aspirated engines in the late eighties, a
The connecting rod is I rather than H section. Crankshaft highly developed version of the DFV, the DFR of 1989,
journals are running in plain, shell-type, main bearings. was delivering about 14.1 bar mepb at 10,750 Rpm. The
There is a main bearing journal each side of each big end HB of 1993 was delivering about 13.6 bar mepb at 13,200

1
Rpm while the Zetec-R of 1994 was delivering about 13.5 Torque and power depend on engine size and speed. A
bar mepb at 13,800 Rpm. After the displacement cut of more useful relative engine parameter related to torque
3.0 liter in 1995, Cosworth developed V8 ED customer and power is the brake mean effective pressure:
engines by using the HB as a basis, and V8 Zetec-R,
then V10 Zetec-R factory engines. The V8 ED5 engine of  2 ⋅ Pb   4 ⋅ π ⋅ Tb 
1997 is believed to have a peak power at about 14,500 mepb =   = 
Rpm with around 13.7 bar mepb. The V10 Zetec-R of  N ⋅ Vd   Vd 
1997 is understood to have a peak power at about 16,000
Rpm with around 13.2 bar mepb. The majority of contem- where Vd is the displaced volume. The brake mean effec-
porary V10 rival engines are recognized to have peak tive pressure may be compared with the indicated mean
power at about 16,250÷16,750 Rpm with around effective pressure, the work delivered to the piston over
13.6÷13.8 bar mepb. the cycle divided by the cylinder volume displaced per
cycle:
Fundamentally all the F1 engines of this season have in
 p ⋅ dV 
∫
common the V10 architecture (Mecachrome, Ferrari,
Mercedes, Honda, Peugeot, Yamaha, Ford). Despite a
mepi =  
further increase in the number of cylinders should pro-  Vd 
duce better engine performances, the V12 architecture is  
not utilized since 1995. As the authors have some legiti-
mate doubts about the actuality of this choice, they where p is the in-cylinder pressure. Their difference is the
decided to perform a study of V10 and V12 F1 engines friction mean effective pressure, mepf=mepi-mepb. The
having equal degree of sophistication. The aim of the indicated mean effective pressure is proportional to ther-
study is to estimate benefits and shortcomings of these mal and charging efficiencies,
solutions on a logical basis.
 p ⋅ dV 
Similarity rules and non dimensional parameters from
ηt = 
∫  η =  m f + ma 
previous projects define basic geometric and operating  mf ⋅ H  c
 ρ ⋅ Vd 
parameters of these engines. Classical engine outputs  
versus engine speed, including brake, indicated and fric-
tion values are then computed. Indicated values are where mf and ma are the mass of fuel and air trapped in a
obtained by using a well-established analytical method. cycle, H the fuel lower heating value and ρ a reference
Cycle simulations are in use by more than 25 years and density. Therefore:
many efficient and reliable commercial codes performing
these simulations are available on the market. These  ρ⋅H 
mepi = ηt ⋅ ηc ⋅  
codes have undergone extensive validation by both  1 + AF 
developers and customers and they generally provide
very satisfactory results in terms of breathing parame- where AF=ma/mf is the air-to-fuel ratio.
ters. Computations are performed here by using the
WAVE code that is developed by Ricardo North America, As engines run faster, friction losses increase as the
Inc. of Burr Ridge, IL, [21¸26]. Friction values are speed or the square of speed. Therefore an increase of
obtained by using a certainly less reliable empirical mepf running higher speeds is inevitable. Engines run-
method. Such technique represents typical rapid meth- ning faster with adequate breathing properties and com-
ods that are often used to obtain a feel for a complex sub- parably smaller frictional losses require a bigger bore/
ject based on a limited amount of experimental data. stroke ratio (and lightweight parts). The bore/stroke ratio
Consequently, results in term of friction and brake values thus increases with engine speed for keeping constant
have to be considered very carefully. the mean piston speed then reducing mepf. This ratio
also increases for keeping constant gas velocities within
FRICTION AND INDICATED MEAN EFECTIVE larger ports to maintain adequate h c at higher speed.
However, the form of the combustion chamber tends to
PRESSURE
become compromised as long as the bore/stroke ratio
increases. With bigger bores the combustion volume at
Engine performances are given in terms of torque and
top dead center becomes thinner and valve clearance
power. Torque is a measure of an engine's ability to do
notches tend to become an ever greater proportion of it.
work, while power is the rate at which the work is done. If
Consequently, at very high speed it is difficult to attain a
Pb is the power, N the engine speed and Tb the torque,
satisfactory chamber form while at the same time achiev-
then Pb=2·π·N·Tb. The goal of the designer is, in theory,
ing a high compression ratio, and ht tends to reduce. As a
to push both peak power and torque curve up, while, in
result, running higher speeds, a small reduction of mepi
practice, is to find the best compromise between torque
is inevitable.
curve and peak power. It is important not to exchange too
much area under the power curve in the search for top
end power.

2
INDICATED MEAN EFFECTIVE PRESSURE lence controlled combustion is limited by the turbulent
MODEL mixing at the flame front. It depends on turbulent velocity
and length scales. The turbulent velocity scale is propor-
We consider engines of similar design, having the same tional to the mean piston speed. The length scale is pro-
displacement, but a different number of cylinders. portional to the bore. The time duration of this
Engines have equal bore/stroke, connecting rod length/ combustion phase is therefore about proportional to the
stroke, intake and exhaust valve diameter/bore ratios, inverse of the engine speed, while its crank angle dura-
and equal maximum power mean piston speed. Engines tion is about constant. The rate of kinetically controlled
have a similar four valves, single spark plug, pent roof combustion is limited by the averaged rate of chemical
combustion chamber design. reactions. It does not depend on turbulent velocity and
length scales. The time duration of this combustion
If B is the bore, S the stroke, R the connecting rod length, phase is therefore about constant, while its crank angle
nc the number of cylinders, NM the maximum power duration increases with engine speed. Finally, the time
engine speed, Div the intake valve diameter and Dev the required for flame travel between spark plug and bore
exhaust valve diameter, we have nc·B2·S=C1, B/S=C2, R/ with equal flame speed is about proportional to the bore
S=C3, NM·S=C4, Div/B=C5, Dev/B=C6 were C1, C2, C3, size. As a consequence of these hypotheses, it is
C4, C5 and C6 are design constants. assumed that engines have about the same distribution
Head and piston geometry has all its dimensions scaled of cumulative mass fraction burned versus crank angle,
proportionally to the bore. Intake and exhaust ports have and a Wiebe function with identical shape and duration
all their relevant dimensions scaled proportionally to the parameters is used to represent combustion. The heat
bore. Since the bore/stroke ratio does not change, transfer coefficient is made proportional to mean and tur-
engines also have the same compression ratio e, another bulent gas velocities, and therefore to mean piston
design constant. speed, through the Woschni correlation.

From the geometrical similarity, engines have the same If all the cam-follower parameters are scaled with the
distribution of steady discharge coefficients (ratio of bore, engines have equal contact stresses at maximum
effective to bore area) and tumble coefficients (ratio of power speed. Cam lift, curvature radius, geometric veloc-
tumble to axial momentum) versus the non dimensional ity, acceleration and jerk, and cam-follower contact width
valve lifts (valve lifts to bore ratio). The effective area is is proportional to the bore. The force at the cam-follower
the ratio of the volume flow rate to a characteristic veloc- linkage is the sum of inertia and spring force, the latter
ity related to the pressure drop across the valve. being determined by the inertia force at engine over
speed. The engine over speed is proportional to maxi-
Engines have similar valve lift profiles and timings, i.e., mum power speed. The force at the cam-follower linkage
equal distributions of non dimensional valve lift versus is therefore F≡mf·a·NM2 [10], mf being the follower equiv-
crank angle, and therefore equal valve timings, θIVO, θIVC, alent mass and a the geometric acceleration. The Hertz
θEVO and θEVC, and equal intake and exhaust duration contact stress is σ2≡F/(ϕ·λ) [10], where ϕ is the width of
δθi, δθe. Since engines have equal distributions of non the contact zone and λ is the radius of curvature of the
dimensional valve lift versus crank angle and discharge cam. Considering mf≡B3, a≡B, NM≡B-2, then F≡B2, and
coefficient versus non dimensional valve lift, they have since ϕ≡Β and λ≡Β, the Hertz stress does not change.
the same distribution of effective area per unit displaced
volume versus crank angle. The scaling rule for intake primary pipe diameters is
obtained by considering equal gas velocities at maximum
If Dj is the equivalent diameter of the j-th section of the power speed [11], and a relation similar to the one previ-
intake or the exhaust port, located at a distance L j from ously provided for port flow velocities. Equivalent diame-
valve seat, then Dj½B, Lj½B. Consequently, at maximum ters of relevant sections along the intake pipe (including
power speed, engines have about the same velocities of seat, throat, head inlet, throttle, taper inlet and bell
gases within ports [11]: mouth) may be obtained by using C 7,j and ug,j values
from previous projects.
 B2 ⋅ S ⋅ N M 
u g , j = C7 , j ⋅   The scaling rule for intake primary pipe lengths is
 δθ ⋅ D j 
2
obtained by considering similar pressure traces at maxi-
mum power speed. The best tuned length of the intake
where C7,j are section dependent flow coefficients, about pipe is found when the pressure upstream of the intake
constants between engine projects, and δθ is the valve valve exhibits three peaks [11,12]. A strong peak, due to
opening duration. As a result, engines have similar pres- wave reflections during valve opening, is located shortly
sure losses and dynamic pressure effects. before θIVC, while one of the two following peaks, due to
wave reflections during valve closure, is located about
The similarity of in-cylinder geometry and intake flow pro- θIVO. This “three peaks”, or “second pulse”, configuration
duces about the same mean and turbulent in-cylinder requires [11]:
flow fields at maximum power speed. The rate of turbu-

3
 ( )
of primary pipe exit and secondary pipe inlet, respectively
 720 ⋅ a ⋅ 1 − M 2 De,p and De,s, are preferably related by the simpler rela-
≅
p p
Li, p
  tion De,s2= De,p2·(nc/2)·( δθe/720).
 12 ⋅ 6 ⋅ N M 
The scaling rule for lengths of exhaust primary pipes, col-
lector and secondary pipe is obtained by considering
ap being the averaged speed of sound and Mp the aver-
similar pressure traces at maximum power speed. The
aged Mach number within the pipe during valve opening.
best tuned lengths are found when the pressure down-
The global length is split into different components. The stream of the exhaust valve exhibits a wide deep during
taper length should be about proportional to the differ- valve overlapping [11,12]. If θw is the angle between
ence between inlet and exit diameters. The length of the intake top dead center and the end of the deep, Le,p is
collector, between throttle and head inlet, should be the length of the primary pipe, Le,c is the length of the
taken proportional to the difference between inlet and exit junction, and Le,s is the length of the secondary pipe, a
equivalent diameters. Since both lengths and diameters “second pulse” configuration requires [11]:
are scaled proportionally to the bore, all these ratios are
conserved.
Le , p ≅
p (
a ⋅ 1− M 2
p )  ⋅  θ EVO + 180 + θw 

The use of fully variable telescopic trumpets produces a
 4   6⋅ N M 
length of the intake primary pipe variable between a min-  
imum and a maximum. The previous value of Li,p may be
considered the minimum primary pipe length, while the
maximum pipe length is obtained by adding to Li,p the
 2⋅ L
 e, p
≅
(
 2 ⋅ L + L
e ,s e ,c ) 
length of the cylindrical portion of the trumpet δLi,p. At
maximum power speed, the primary pipe length is Li,p.  (
 a p ⋅ 1 − M p2 )   a s ⋅ (1 − M s2 )
 


As engine speed reduces, the primary pipe length
increases about inversely proportional to engine speed, ap being the averaged speed of sound and Mp the aver-
thus preserving an optimum “three peaks” configuration. aged Mach number within the primary pipe during valve
When the maximum primary pipe length L i,p+δLi,p is opening, as the averaged speed of sound and Ms the
reached at a first critical speed, the primary pipe length is averaged Mach number within the secondary pipe during
then reduced to the minimum primary pipe length L i,p, the cycle. Since ap≈as, up≈us, then Le,s should be taken
thus producing about an optimum “four peaks” configura- equal to Le,p-Le,c. The length of the junction should be
tion. The primary pipe length then increases again as the taken about proportional to the difference between inlet
speed further reduces, until the maximum primary pipe and exit equivalent diameters.
length is reached at a second critical speed. The primary As a consequence of the previous hypothesis, engines
pipe length is then reduced again to the minimum pri- have Dj≡B, Lj≡B along the intake pipe, from the valve seat
mary pipe length, thus achieving about an optimum “five to the bell mouth section, and along the exhaust primary
peaks” configuration. pipe, from the valve seat section to the inlet of the
The scaling rule for exhaust primary pipe diameters is exhaust junction. Conversely, diameters and lengths of
obtained by considering equal gas velocities at maximum the exhaust junction and secondary pipe are also
power speed [11], and a relation similar to the one previ- affected by the number of cylinders coupled through the
ously provided for port flow velocities. Equivalent diame- junction.
ters of relevant sections along the exhaust primary pipe
(including seat, throat, head exit and cylindrical primary FRICTION MEAN EFFECTIVE PRESSURE
pipe) may be derived by using C7,j and ug,j values from MODEL
previous projects.
Friction losses result from relative motion between solid
At their end, the primary pipes from one bank of cylinders
surfaces in the engine, i.e., motion between piston and
empty into the exhaust junction. The scaling rule for
cylinder wall or crankshaft journal and main bearing. Rel-
exhaust secondary pipe diameters may also be obtained
ative motion does not require the two solids to be in con-
by considering equal gas velocities at maximum power
tact with each other, as it is generally the case that there
speed [11], and a relation similar to the one previously
is a film of lubricant between surfaces. The crankshaft
provided for port flow velocities now introducing the
friction includes friction from the main bearings and the
dependence on the number of cylinders coupled through
front and rear main bearing oil seals. The reciprocating
the junction:
component friction includes friction from the connecting
rod bearings, piston skirts and piston rings. The valve
n   B ⋅S⋅ NM 
2
train friction includes friction from the camshaft bearings,
u g , j = C7 , j ⋅  c  ⋅  
 2   δθe ⋅ D 2j  cam-follower interfaces and valve actuation mechanisms.
Also relevant are pumping losses due to turbulent dissi-
pation in hydrodynamic journal bearings, windage losses
However, due to the relatively large variation between dif-
and finally miscellaneous and accessories' losses. Pis-
ferent projects of C7,j far from engine cylinders, diameters

4
ton, rod and crankshaft assemblies are generally consid- having equal parameters. By assuming equal ring pres-
ered to be responsible for the major frictional losses. sure and viscosity coefficient, and ring height propor-
tional to the bore, and introducing an empirical constant
Relating interface friction mean effective pressure to
C9, we have:
design and operating parameters requires the determina-
tion of characteristic friction coefficient ζ, velocity u and
 S
load F. The assumption made about the type of lubrica- mep f ,r = C9 ⋅  ⋅ N
tion determines the relationship between the friction coef-  B
ficient and a non dimensional duty parameter µ·u/F,
where µ is the viscosity of the oil. The friction coefficient For a piston skirt having height hp, normal force, friction
versus the duty parameter varies in the boundary, mixed coefficient and velocity at each piston skirt-liner interface
and lubrication regimes. In boundary lubrication (signifi- are given as follows:
cant asperity, little or no lubricant film), the friction coeffi-
cient is about constant. In mixed lubrication (some   S 
asperity contact, some lubricant film), the friction coeffi- F ≡ ( p g ⋅ B 2 + C10 ⋅ ρ ⋅ B 3 ⋅ N 2 ⋅ S ) ⋅  tan 
cient varies approximately with the inverse of the duty   2 ⋅ R 
parameter. In hydrodynamic lubrication (no asperity con-
µ ⋅ u ⋅ hp
tact, full film), the friction coefficient varies about propor- ζ≡ u ≡ N ⋅S
tionally to the duty parameter. The friction coefficient ζ F
multiplied by the normal force F determines the friction
force. The friction force multiplied by the friction velocity u where pg is the gas pressure, r the material density of the
determines the friction power loss Pf. The friction power reciprocating parts, C10 the relative weight of pressure
loss divided by engine speed and displacement deter- and inertia forces. By assuming equal material density,
mines the friction mean effective pressure mepf. This gas pressure and viscosity coefficient, and skirt height
term is finally calibrated on the available experimental proportional to the bore, and introducing empirical con-
data by introducing an empirical constant C as follows: stants C10 and C11, we have:

 ζ ⋅ F ⋅u   S 
mep f , p = C11 ⋅ p g + C10 ⋅ B ⋅ N 2 ⋅ S ⋅ tan
mep f = C ⋅   
 nc ⋅ B 2 ⋅ S ⋅ N   2 ⋅ R
S
For a bearing having diameter D b and length Lb, friction ⋅ ⋅ N
force and velocity at each bearing-journal interface are
B
the following:
For a valve with diameter Dv, friction force and velocity at
each cam-follower interface are given as follows:
 µ ⋅ Db ⋅ Lb ⋅ u 
ζ ⋅F ≡   u ≡ Db ⋅ N
 δb 
2
N  hM
ζ ⋅ F ≡ ρ ⋅ Dv ⋅ h M ⋅  M 
3
u ≡ N ⋅ δθ ⋅ (rb + )
 δθ  2
where µ is the viscosity coefficient and δb is the operating
clearance. There are nc/2+1 main bearings, nc connect- where δθ is the valve opening angle, rb the base radius,
ing rod bearings, and 4·(nc/2+1) camshaft bearings, each hM the maximum valve lift, ρ the material density. There
bearing of a type having equal parameters. By assuming are 2·nc intake and 2·nc exhaust valve, having equal
bearing length and diameter proportional to the bore, and parameters. By assuming base radius, valve diameter
equal operating clearance and viscosity coefficient, and and maximum lift proportional to the bore and equal
introducing an empirical constant C8, we have: material density, and introducing an empirical constant
C12, we have:
 B2   n + 2
mep f ,b = C8 ⋅   ⋅  c ⋅N
 S   2 ⋅ nc   N M 2 ⋅ B3 
mep f ,v = C12 ⋅  
 S ⋅ δθ 
For a piston ring having height hr, normal force, friction
coefficient and velocity at each ring-liner interface are
given as follows: The previous relations determine “basic” friction losses.
Additional sources of engine friction are supposed to give
an identical distribution of mepf,a versus engine speed.
µ ⋅u⋅ B
F ≡ pr ⋅ B ⋅ hr ζ≡ u ≡ N ⋅S The sum of all these sources of losses produces a poly-
F nomial approximation of total mepf versus engine speed
having coefficients explicitly introducing basic engine
where µ is the viscosity coefficient and the ring pressure parameters. These relations are useful to indicate a trend
pr is proportional to both gases' pressure and ring elastic by interpolating a set of experimental data.
pressure. There are 2·nc piston rings, each ring of a type

5
The proposed technique represents typical rapid meth- 2.5l V6 ITC 3.0l V10 F1 3.0l V12 F1
ods that are often used to obtain a feel for a complex sub- De,s [mm] 56.0 (2x3) 75.0 (2x5) 77.5 (2x6)
63.5 (3x4)
ject starting from a limited amount of experimental data. 55.0 (4x3)
The technique ignores the description of the fundamental Le,s [mm] 546 315 300
phenomena and prefers to use a few basic engine
parameters and empirical constants to simply interpolate
the available experimental data. where Di,t is the intake throttle diameter. The length of the
cylindrical, telescopic portion of the intake trumpet δLp,i
The proposed empirical relations focus on sources of fric- has been arbitrarily taken equal to 0.05 m.
tional losses relevant to racing engines running high
Rpms. These relations indicate that increasing the num- The maximum power engine speed of F1 engines is
ber of cylinders decreases total friction. This is in accor- expected to be higher than the proposed values. The
dance to many racing engine data, as those provided in 12,000 Rpm maximum power engine speed of the V6
[21] just to name a few in the public domain. It is worth of ITC engine, corresponding to a mean piston speed of 23
note that in series engines running low Rpms, where the m/s, is set by the prescribed FIA-Rpm limiting device, and
trend is opposite and increasing the number of cylinders it is not a technological limit of the engine.
also increases total friction, the previous sources of Following the similarity rules previously outlined, the V12
losses have a different formulation and there are other engine has two six-into-one exhausts (2x6 model), while
relevant sources of losses to be considered. the V6 and V10 engines respectively have two three-into-
one and two five-into-one exhausts (2x3 and 2x5 mod-
BASIC ENGINE MODEL PARAMETERS els). The six-into-one exhaust does not produce better
performances than the classical six-into-two-into-one
The basic engine model parameters are obtained by exhaust, but the latter does not comply with the proposed
using the similarity rules previously outlined and data of similarity rules. Two other solutions have been shown for
the 2.5 liters ITC V6 engine developed by Fiat Auto Corse the V12 engine, four three-into-one exhausts (4x3
for the 1996 season [10,11]. model), and three four-into-one exhausts (3x4 model).
We have C1=(4/π)·3·10+3 cm3. We assume equal C4, i.e., The four three-into-one exhausts may be used in both a
mean piston speed, for a 3.0 liters F1 V10 engine running V12 engine or in a W12 engine made up of four banks of
at about 16,000 Rpm [19] and the 2.5 liters ITC V6 three cylinders, each two banks being combined with a
engine running at 12,000 Rpm. Furthermore, we also narrow angle as a special bank of six cylinders. The three
assume for these engines equal C5 and C6 values, i.e., four-into-one exhausts are not appropriate for a V12
mean intake and exhaust velocity, and finally equal ratio engine, but they may be used in a W12 engine made up
R3/(B2·S). From these assumptions the key model of three banks of four cylinders.
parameters of 3.0l V10 and V12 F1 engines are obtained The proposed values are only a reasonable hypothesis
as follows: for comparison of different solutions. Undoubtedly, some
changes are needed in order to derive fully working mod-
2.5l V6 ITC 3.0l V10 F1 3.0l V12 F1 els from these values. However, aim of the paper is not to
B [mm] 96.0 94.0 89.0 provide detailed information on actual F1 engines. The
S [mm] 57.5 43.2 40.2 aim of the paper is exclusively to evaluate relative bene-
R [mm] 135.5 121.5 115.2 fits and shortcomings of different architectures, specifi-
ε 12.95 12.95 12.95 cally solutions having a different number of cylinders, on
Niv 2 2 2 a logical basis.
Div [mm] 34.0 33.5 31.5
Nev 2 2 2
COMPUTATIONAL RESULTS
Dev [mm] 28.0 27.5 26.0
NM 12,000 16,000 17,250
δθI 315° 315° 315° The computational results obtained for the V10 and V12
θIVO 60° 60° 60° engines previously defined are given here. Computations
θIVC 75° 75° 75° have been performed for the steady, Wide Open Throttle
Hi,M [mm] 13.0 12.7 12.1 (WOT) operation of the engines in the range of speed
δθe 315° 315° 315° 10,000÷18,500 Rpm. The telescopic intake trumpets are
θEVO 80° 80° 80° operated as follows (linear variation of length):
θEVC 55° 55° 55°
He,M [mm] 12.0 11.8 11.1
Lp,i [mm] 237 181 174 3.0l V10 F1 3.0l V12 F1
δLp,i [mm] 0 50 50 δLi,p [mm] N [Rpm] N [Rpm]
Di,t [mm] 49.0 49.0 46.0 0→50 16,500→14,000 17,500→15,000
Lp,e [mm] 676 487 461 0→50 13,500→11,000 14,500→12,000
De,p [mm] 51.0 51.0 48.0 0→50 10,500→ 8,000 11,500→ 9,000
Le,c [mm] 130 172 176

6
Results of computations are presented in figure 1÷10. Figures 9 and 10 present the indicated and brake specific
Figure 1 shows the charging efficiency, mass of fresh fuel consumption. There is a considerable overlapping of
mixture trapped to reference mass versus engine speed. these curves. Therefore, neither the V10, nor the V12
The efficiency of the gas exchange processes is similar, engine shows clear advantages under this point of view.
due to the similar design. Nevertheless, despite the tun-
ing for higher Rpms, the V12 engine has a wider range of CONCLUSIONS
high charging efficiency. The accuracy of this estimation
is expected to be fairly reliable especially in the range of A simple, preliminary study of V12 and V10 engines has
high engine speed. This kind of computations has been been performed by using similarity rules and a small data
abundantly verified and may produce accurate results base to derive their models for indicated and friction
when the model has been properly set up. The actual mean effective pressure.
model has been set up only for the operation where both
the intake and the exhaust stay tuned. Consequently, the These models have shown some of the clear advantages
accuracy of the computation may rapidly deteriorate for the V12 solution has in terms of pure engine perfor-
the operation at low engine speed where the exhaust mances. Other advantages, as those arising from balanc-
goes out of tuning and significant back flows appear. Dif- ing of inertia forces, have been disregarded.
ferences between V12 models become relevant only Thanks to the lower frictional losses and the better bal-
when the accuracy is expected to deteriorate. ancing of inertia forces, a V12 engine with a slightly
smaller bore/stroke ratio may be optimized to give lower
Figure 2 shows the indicated mean effective pressure
specific fuel consumption and higher power output at
versus engine speed. The efficiency of the heat release
marginally larger engine speeds.
and heat transfer processes is similar, due to the similar
design. However, the V10 engine has a smaller surface- The paper does not cover all these aspects pertinent to
to-volume ratio, and therefore a better thermal efficiency. general car performances. There is certainly more to the
The V10 engine has a higher maximum mepi, while the performance of a F1 car than the pure engine theory cov-
V12 engine has a wider range of high mepi. The accu- ered here. Advantages of the V12 architecture in terms of
racy of this estimation is expected to be less reliable than general car performances are difficult to be assessed,
the one provided for breathing parameters, due to the and they require a more complete approach.
many additional hypotheses involved.
A more precise evaluation of benefits and shortcomings
Figure 3 shows the friction mean effective pressure ver- of the V12 architecture may be obtained by performing a
sus engine speed. The accuracy of this estimation is con- full “pre-study”, including engine and car theory and
siderably less reliable than both the estimations of experiment. This “pre-study” could be carried out with a
charging efficiencies and indicated mean effective pres- relatively small budget, but clearly it requires an effort
sures. A precise estimation of the friction mean effective considerably greater than the one performed for writing
pressure is very difficult to be achieved. The V12 engine this paper.
has a mepf always smaller than the mepf of the V10
engine running at the same speed. Opposite to series REFERENCES
engines running low Rpms, where increasing the number
of cylinders means increased friction, in racing engines 1. Federation Internationale de L’Automobile, “Technical Reg-
running high Rpms, friction reduces by increasing the ulations for Formula One Cars.” Valid through January, 1,
number of cylinders. 1997 to December 31, 1997. (www.fia.com)
2. Boretti, A.A., et al., “Advanced Tools for Intake and Exhaust
Figure 4 shows the brake mean effective pressure versus Port Design in Four Valve Spark Ignition Engines”, paper
engine speed. The V12 engine has a wider range of high presented at the “ATA Seminar High Performances Spark
Ignition Engines for Passenger Cars”, Milan, Italy, Novem-
mepb and higher values of mepb. This is primarily the
ber 1992.
result of a generally better distribution of mepf, and then 3. Boretti, A.A., Cantore, G., and Mattarelli, E., “Numerical
of the wider range of high mepi. Correlation of Combustion Evolution and Port and Com-
bustion Chamber Shape in a High Speed, Four Valve
Figure 5 and 6 show the indicated torque and power, Spark Ignition Engine,” paper presented at the “SAE Motor
while figure 7 and 8 present the brake torque and power Sport Engineering Conference and Exposition”, Dear-
versus engine speed. The V12 engine has lower values borne, Michigan, December 1994. SAE P.942534.
of torque, but a wider range of high values. Furthermore, 4. Boretti, A.A., Borghi, M., and Cantore, G., “Numerical
Study of Volumetric Efficiencies in a High Speed, Four
the V12 engine has a greater power output. Shifting from
Valve, Four Cylinder Spark Ignition Engine,” paper pre-
indicated to brake values, differences in favor of the V12 sented at the “SAE Motor Sport Engineering Conference
engine increase. and Exposition”, Dearborne, Michigan, December 1994.
SAE P.942533.
The V12 engine provides a significantly larger maximum 5. Boretti, A.A., Borghi, M., Marmorini, L., and Mattarelli, E.,
indicated power Pi (+6%) at a slightly increased Rpm “Experimental and Computational Methods for Port Design
(+3.0%), while it provides a still remarkably larger maxi- in High Performance Four Valve Spark Ignition Engines,”
mum brake power Pb (+6.0%) at a notably increased paper presented at the “Convegno Modena Motori - Auto-
mobili e motori high-tech”, Modena, Italy, May 1995.
Rpm (+6%).
6. Boretti, A.A., Borghi, M., Dominici, A., Guerrini, G., and

7
 Mattarelli, E., “Influence of Intake Port Design on Perme- LIST OF SYMBOLS
ability and Rate of Combustion in a Four Valve High Perfor-
mance Engine”, paper presented at the “ATA 2nd
International Seminar High Performances Spark Ignition a: acceleration or speed of sound
Engines for Passenger Cars”, Milan, Italy, November 1995. AF: air to fuel ratio
7. Boretti, A.A., Borghi, M., Cantore, G., and Mattarelli, E., B: bore
“Computational methods for the Fluid Dynamic Optimiza- C: constant
tion of High Performance Engines”, paper presented at the D: diameter
“2° Convegno Automobili e motori high-tech”, Modena,
Italy, May 1996. F: force
8. Boretti, A.A., Borghi, M., Cantore, G., and Mattarelli, E., h: height
“Numerical Optimization of a F1 Engine with variable Intake H: lower heating value
and Exhaust Geometry and Valve Events”, paper pre- L: length
sented at the “2nd SAE Motor Sport Engineering Confer- m: mass
ence and Exposition”, Dearborne, Michigan, December
1996. mep: mean effective pressure
9. Boretti, A.A., Cantore, G., Mattarelli, E., and Preziosi, M., M: Mach number
“Experimental and Computational Analysis of a High Per- n: number
formance Motorcycle Engine”, paper presented at the “2 nd N: engine speed
SAE Motor Sport Engineering Conference and Exposition”, p: pressure
Dearborne, Michigan, December 1996.
10. Boretti, A.A., and Villa, F., “Design of the Fiat Auto Corse P: power
ITC 96 racing engine. Part I: valve lift profiles and timings”, r: radius
paper to be presented to the 1997 SAE International Con- R: connecting rod length
gress and Exposition , Detroit, Michigan, February 1997. S: stroke
11. Boretti, A.A., and Secchi, D., “Design of the Fiat Auto Corse T: torque
ITC 96 racing engine. Part II: pipe lengths and diameters”,
paper to be presented to the 1997 SAE International Con- u: velocity
gress and Exposition , Detroit, Michigan, February 1997. V: volume
12. Pignone, G.A., and Vercelli, U.R., Motori ad alta potenza µ: viscosity coefficient
specifica, Giorgio Nada Editore, Vimodrone (Milan, Italy), δ: clearance
1995. δθ:
δθ valve opening angle
13. Rothbart, H.A., Cams, Design, Dynamic and Accuracy,
John Wiley & Sons, Inc., New York, 1956. δL: length variation
14. Bocchi, G., Motori a quattro tempi, Hoepli, Milan, Italy, ε: compression ratio
1987. ϕ: width of contact zone
15. Stone, R., Introduction to Internal Combustion Engines, λ: curvature radius
SAE, Warrendale, PA, 1995. η: efficiency
16. Heywood, J.B., Nitschke, R.G., and Patton, K.J., "Develop-
ment and Evaluation of a Friction Model for Spark-Ignition θ: crank angle
Engines", SAE P. 890836. σ: Hertz stress
17. Yagi, S., et al.,"Total Engine Friction in Four Stroke S.I. ρ: density
Motorcycle Engine", SAE P. 880268. ζ: friction coefficient
18. Yagi, S., et al.,"Estimate of Total Engine Loss and Engine
Output in Four Stroke S.I. Engines", SAE P. 910347. SUBSCRIPTS
19. ”Engine Tech”, Race Tech, Racing Technique & Race Car
Technology, vol.3, no.1, 1997. a: air or additional
20. Henein, N.A., et al., “A New Approach to Determine Lubri- b: brake or bearings
cation Regimes of Piston-Ring Assembles”, Journal of Tri- c: cylinders, collector or charging
bology, Vol.119, October 1997, pp. 808³816 d: displaced
21. Ricardo North America Inc., “WAVE User’s Manual”, Burr
Ridge, IL, USA, 1997. e: exhaust
22. Morel, T., and La Pointe, L.A., "Concurrent Simulation and EVC: exhaust valve closure
Testing Concept in Engine Development", SAE P.940207. EVO: exhaust valve opening
23. Morel, T., Silvestri, J.J., and Costello, M., "Study of Intake f: friction, follower or fuel
System Wave Dynamics and Acoustics by Simulation and g: gas
Experiments", SAE P.940206.
24. Morel, T., Keribar, R., and Blumberg, P.N., A New Approach I: indicated or intake
to Integrating Engine Performance and Component Design IVC: intake valve closure
Analysis Through Simulation", SAE P.880131. IVO: intake valve opening
25. Morel, T., and Lowe, A.S.H., "A New Generation of Tools for j: section index
Accurate Thermo-Mechanical Finite Element Analyses of m: minimum
Engine Components", SAE P. 920681.
26. Jackson, N.S., Pilley, A.D., and Owen, N.J., "Instantaneous M: maximum
Heat Transfer in a Highly Rated DI Truck Engine," SAE p: piston or primary
P.900692. r: ring
s: secondary
t: thermal or throttle
v: valve
w: pressure wave