To Be Done:

EE290 & ME272

Senior Design Projects

EE Team 1/ ME Team 15

2000-2001 Formula SAE Race Car

Suspension/Steering &

Engine Control System Design

Team Members:

Andy Bilmanis, ME

Robert Hotaling, EE

Jason Mangual, CMPE

Michael McGee, EE

Nnamdi Okam, EE

Greg Pineau, ME

Justin Pribanic, ME









Advisors:

Professor John Ayers: jayers@engr.uconn.edu

Professor Jim Cowart: jscowart@engr.uconn.edu





SAE Collegiate Design Series

http://www.sae.org/students/formula.htm

To Be Done:

1. Label figures

2. Possibly leave out definition section, define each term as it is being used.

3. Analyze the validity of using SuspensionGen and Pro/E modeling software

a. Compare and include in the appendix a sketch of a simple suspension system

(not parallel) to include in the appendix.

Abstract:

This report describes two combined design projects that are to be implemented

through inter-disciplinary teamwork as components of the 2000-2001 Formula SAE

vehicle being built by the Formula SAE Chapter at the University of Connecticut. The

design project is motivated by the Formula SAE student competition, and is funded by

sponsorship support and the Electrical and Computer Engineering Department. The first

project to be discussed is designed through a mechanical engineering senior design

project while the second project is the interdisciplinary senior design project between one

Mechanical, one Electrical and Computer Engineering, and three Electrical Engineering

Students.

The Formula SAE race car suspension design incorporates features that are

designed to make the car’s turning predictable as well as adjustable. The adjustability of

the car extends to meet the needs of different sized drivers, driver preference, and varying

track conditions. The goal of the suspension design is to maintain the maximum

accelerations in the lateral, positive, and negative directions that can be achieved while

competing during the Formula SAE challenge. However, existing components are re-

used such as the rear five link suspension arms and drive train that constrain several

design aspects. Similarly, the steering is designed to minimize bump steer while reusing

existing components such as the rack, uprights, rim spacing, and spindles.

Similarly, the engine control system for the Formula SAE race car will offer a

high performance engine control system. It is specifically designed for a Honda F3

600cc motorcycle engine. The system will provide electronic control of the fuel injection

system, and the primary ignition system. The system will use feedback from the engine

combined with pre-programmed information to deliver metered fuel, and spark firing to

the cylinders at the optimum time for peak engine performance. This system will be

competitive in both performance and price with the current complete fuel injection

systems available from various manufacturers. The design will also offer sequential fuel

injection without the use of a camshaft sensor. This will allow better engine control

without adding mechanical complexity to the design.

Table of Contents





Introduction

Introduction:





The objective of the Formula SAE competition is to design, build, and enter a

prototype car that will perform well in the areas of acceleration, braking, and handling.

The car will be designed and built assuming a manufacturing firm could produce the

design at a rate of four cars per day in a limited production run costing less than $30,000

for a prototype vehicle. This year’s goal is to produce a competitive car that out performs

previous designs in the areas of engine performance and suspension dynamics, through

interdisciplinary team work between Mechanical, Electrical, and Computer Engineers.

Each Team’s design will be compared and judged with other competing designs to

determine the best overall car.

The purpose of a suspension system is to maintain the maximum accelerations





Y

Z

Rear

X

X

Front









during lateral acceleration, forward, and rearward accelerations while responding to body

roll and small vertical wheel movements. The forces generated by cornering,

accelerating, and braking are directed to the vehicle’s center of mass. The center of mass

is a located by three planes represented by the SAE standard. The origin of the y

component is centered on the vehicle’s line of symmetry facing the passenger side in the

positive direction, the positive x direction is to the rear of the vehicle and positive z

direction is referenced from the bottom of the vehicle up.

This coordinate system will be used throughout the car’s design, however it is

local to front and rear suspension systems with the front suspension and steering

combined. There is also a vocabulary that suspension designers use which can seem

foreign, so included with the nomenclature is a list of suspension design vocabulary.

One way to build a successful first year car is to study winning designs from

previous competitions. Ideas were first generated during a visit to the May 2000

competition in Pontiac Michigan, and continued research through internet sources

brought the two suspension designers up to speed on what features the current winning

designs include. Winning designs such as Texas A&M were considered in brainstorming

sessions to develop an improved suspension design for the 2001 FSAE car. Input from

sources such as faculty, car clubs, and classmates have been a good source of ideas.

The Engine Control System to be implemented by the EE/CMPE and ME teams is

an electronic fuel injection and electronic ignition system. An electronic fuel injection

system delivers fuel to an engine’s intake manifold, and an ignition system delivers spark

to each cylinder at the correct time. Both of these systems must be controlled so that they

will change with different temperature and load conditions. This is done through the use

of feedback from various sensors to a microcontroller, and pre-programmed code in the

microcontroller.

There are many different fuel injection systems in use in today’s gasoline engines.

Virtually all automobiles now use electronic fuel injection systems to deliver fuel to the

cylinders. Three common types of electronic fuel injection (EFI) systems used today are

throttle body injection (TBI), multi-port fuel injection (MFI) and sequential fuel injection

(SFI) systems. In throttle body injection systems, all of the fuel for the engine is

delivered to the same point at the intake manifold inlet through one large fuel injector,

somewhat like a hybrid between fuel injection and a carburetor.

There are several different ignition systems in modern automobiles. Early

systems used mechanical points to trigger the ignition coils, and a mechanical distributor

was used to apply the secondary voltage to the correct cylinder at the correct time. Since

the advent of electronic ignition, the triggering of ignition coils has been accomplished

through the use of power transistors. Some automobiles still use a mechanical

distributor to get the spark to the cylinders, but most systems presently use a

distributorless ignition system (DIS). Two common implementations of this system are,

a single ignition coil used for two companion cylinders, and coil-on plug systems that

have a single coil for each individual cylinder.

Although this senior design project at the University of Connecticut only

concentrates on a few aspects of the vehicle it is necessary to refer to the overall

competition. The references are included in order to explain the needs of a competitive

car and the importance of the project to the overall car performance. The objective of the

Formula SAE competition, restated from page eight of the rule book:

The Formula SAE® competition is for SAE student members to conceive, design,

fabricate, and compete with small formula-style racing cars. The restrictions on

the car frame and engine are limited so that the knowledge, creativity, and

imagination of the students are challenged. The cars are built with a team effort

over a period of about one year and are taken to the annual competition for

judging and comparison with approximately 100 other vehicles from colleges and

universities throughout the world. The end result is a great experience for young

engineers in a meaningful engineering project as well as the opportunity of

working in a dedicated team effort.





The car will be judged both statically and dynamically. Each car must be

inspected before being permitted to enter any aspect of the competition to ensure safety.

The static events include Presentation, Engineering Design, and Cost & Manufacturing

Analysis. Dynamic events are judged in areas of Acceleration, Skid-pad, Autocross, Fuel

Economy, and Endurance events. The combined total of both static and dynamic events

is 1000 points.





The design of the suspension and engine controls is described in this document

along with the cost analysis and the various constraints involved with the design. The

discussion section begins with design alternatives and selection considerations. Next, the

system overview and the details of the final design scheme are discussed. Finally, the

overall limitations of the design project are discussed. These limitations include budget,

time line, and competition rules. The rest of the report will examine each of these

components and explain as thoroughly as possible. The focus of this report is design, and

therefore it does not include manufacturing costs.









Discussion:



**********************************************

ME leftover: Racecar suspension dynamics is a complex subject itself, and while

this report thoroughly examines several important aspects, it in no way touches upon the

many complexities of the dynamics achieved during autocross racing*********

The autocross course is constructed of several components that are typically

found in formula racing; straight-aways, slaloms, constant radius turns, hairpin turns,

chicanes, multiple turns, and varying radius turns. The course is on an asphalt parking lot

surface generally termed “asphalt lake” in Michigan. The length of one lap of the course

is roughly ½ mile where several laps will be taken. The straight-aways will be no longer

than 200 feet with hairpins at either end or no longer than 150 feet with wide turns on the

end. The minimum track width will be 11.5 feet, constant turns 75 to 148 feet diameter,

hairpins no less than 23 feet outside diameter, and slalom cones will have a spacing of 25

to 40 feet. These dimensions directly correspond to maximum accelerations developed

while competing. From previous races it is known that these dimensions suggest average

speeds between 25 to 30 mph. Maximum speeds are also suggested to be as high as 65

mph. Next, the forces generated are calculated using physics and are used to design the

suspension geometry and components and steering mechanism.

The dynamics of the vehicle can be summed up by two objectives. The first

objective is to maintain maximum lateral, positive, and negative accelerations in a

specified range of operating conditions. The second is to produce a reliable engine

management system that out performs other methods of aspiration, spark, and fuel

delivery. While the two objectives respond to the competition each are designated as

senior design projects.





Suspension and Geometry Factors

One of the most difficult parts of designing a suspension system is compromising.

There is no optimum suspension for all conditions, therefore for every improvement there

is a sacrifice. The key is to decide what is most important for your particular application.

In our case we must account for a race on a smooth track that contains many tight turns

but can be subject to a variety of weather conditions. To locate the front uprights we

chose to use a four-link SLA (short-long arm) setup, with the rack and pinion system

while the rear employs a five-link system that can handle the torque transmitted by the

drive train. There are other types of suspension systems such as McPherson struts, solid

axle, and trailing arm, but easily eliminated by our budget constraints. The rack and

pinion steering was also chosen for the same reason.

To fulfill our goal of maintaining the largest accelerations possible we examined

the many components of a suspension design; the definitions of these terms are

conveniently defined during the discussion. The first heavily debated design component





Car Body









L,rc1





L,rc2

1 2



1 > 2

Ground



to be examined is the roll axis, specifically the line that connects the front and rear roll

centers









are the roll axis, the line around which the car body rotates when lateral forces are

applied. Then one may as what defines the roll center: the roll center is the effective

center that the body will appear to rotate about. For roll centers with small radii one

could image a box suspended by two short strings at two corners suspended as something

of a pendulum. The longer the string the larger the pendulum and the more minute the

angular displacements the box will achieve. As shown below, the body roll 1 is greater

that 2 with a longer roll center located below ground level.

Altair Engineering, a sponsor of the team, has provided kinematic software for the

next two semesters. The software is considered kinematic because suspension geometry

changes through wheel displacement. Much of the geometry mentioned was calculated

using this software and then verified by a string calculator, Figure xx in Appendix.

However the software does not analyze the dynamic response associated with a mass-

spring-damper system and is used purely as an efficient tool to generated geometrical

changes with wheel displacement in all specific geometrical areas considered in the

design.

In order to understand the importance of the roll axis of a vehicle the dynamics

must first be examined. Essentially the roll of the vehicle is a result of the moment

formed by the distance of the CG from the roll axis. The three factors that affect the roll

axis of the vehicle are the front roll center, the rear roll center, and the CG (center of

gravity). When a car goes into a turn a lateral force is exerted at the center of gravity. If

the roll center is below the CG, a moment is formed and the car appears to lean during

the turn.

One of the difficulties involved with roll center selection is that it is that it

changes with wheel displacement. One approach is to choose the region, which you want

to keep the roll center within for all displacements. The three basic regions for the roll

center location are the same height as the CG, below the ground, and between the ground

and the CG. It is usually not desirable to have your roll center change between regions

since this makes the handling less predictable. Another method is to design a suspension

in which the roll center stays stay constant, such as using parallel links, but this involves

sacrificing other aspects.

At first analysis it would appear the best arrangement would be to have the CG

and both roll centers on the same plane. This would create no moment and would result

in no body roll in cornering; but there are a few problems involved. The first is that

jacking may develop on high velocity turns. Jacking is when the tires skip around the

turn, this is not desirable since contact with the ground is not maintained and the ability

to fulfill the maximum lateral acceleration has failed. The second problem involves the

tuneablity of the system. A problem with no roll is that the shocks are not being

compressed very much and thus the absorption of corning forces through the suspension

has been limited. A car with some degree of body roll can be adjusted via spring rates of

the shocks and anti-roll bar, but a car with no roll cannot be. Due to the number of

factors effecting handling it is desirable to be able to adjust the system as needed.

The resultant moment must be investigated if the roll center is located below the

CG. If the roll center is too far below the generated moment will be very large causing

the inner tires to lift off the ground. This is highly undesirable since the result will be

loss of both power and control. Our car will have a CG approximately 11 inches above

the ground. With this in mind, we decided the moment for any roll center below the

ground would be too great. Having decided on a roll center between the ground and the

CG we had to determine the optimum range for it.

Our main controlling factor of the roll axis was the front suspension geometry.

The front is critical since that is where the steering is taking place. Any geometry is not

sufficient unless it incorporates positive steering and handling aspects. We can only

change the rear to a limited extent since the rear wheels must be powered by existing

drive train. One suspension limitation was that we decided to reuse the old spindles.

This choice was made for two basic reasons. The first is that we found the setup was

very much what we wanted and the second is just simply to meet budget.

Wheel Displacement









3

2

1

(in)









0

-1 0 2 4 6 8

-2

-3

Roll Center Height(in)





Figure klj – Graph of Roll Center Height vs. Wheel Displacement







Camber is a very important feature in controlling the handling of a

vehicle during turning. Camber is tilting of the tires about the horizontal axis

perpendicular to the direction of rotation. Camber is adjusted in two

methods, the first would be through static camber the second is dynamic

camber. Static camber is gained through small adjustments to the control arm

lengths. Dynamic camber is determined by the suspensions geometry

through the path of a vertically deflected tire. This tilting action can be used

as a valuable handling characteristic. It also can be a very negative

characteristic if it is not properly studied and implemented.









(-) (+)

.









Figure jlj – Demonstration of Negative and Positive Camber







A small amount of negative camber is desirable in a turn, as

determined through tire manufacturers’ data. Specific tire data can be

difficult to achieve but a variety of tire types were available as represented in

figures xx of Appendix…(show plots from Race car vehicle dynamics)

But, the geometry must be such that the camber is zero for straight driving. If

camber exists during positive and negative accelerations the tire patch area is

reduced and maximum possible traction as well as uneven tire wear results.

The desirable aspect of camber is that it can be used to increase tire patch

area when the vehicle experiences body roll. To achieve this effect camber

must be positive when wheel displacement is negative (wheel droop) and

negative when wheel displacement is positive (jounce).

Below is a graph for our suspension system. It achieves just more

than a degree of camber per inch of displacement. Comparing this slope to

the tire data in figure xx of Appendix xx the results support that an increase

in lateral force can be achieved.

Wheel Displacement









3

2

1

(in)









0

-4 -2 -1 0 2 4

-2

-3

Camber (degrees)





Figure ljjlk – Graph of Camber vs. Wheel Displacement

A lot of complications are added to the suspension geometry when the

steering control arm is added. One of the biggest effects it can cause is toe-

in/out, commonly known as bump steer. Toe-in is when the front of the tires

angle in towards each other and toe-out is when they angle away from each

other. It is undesirable to have the tires independently steer the vehicle when

the vehicle hits bump. This characteristic complicates tuning the vehicle by

adding responses that are unpredictable to the driver. Both kinds of toe are a

result of the position of the steering linkages. Since we are using an existing

steering rack its position has several constraints. To steer the vehicle the

control arms must be a distance from the axis about which the tire turns

(specifically the king pin inclination) to induce a moment, thus turning the

car.

Briefly the axis just mentioned is shown here and ads as another

consideration. As the figure below demonstrates, the kingpin inclination





KPI









Scrub radius (+)









affects scrub radius, which is pre-determined by re-using the vertical

uprights.





Bump steer is a very difficult characteristic to accommodate. In a

majority of geometries tested the amount of bump steer cannot be zero for all

wheel displacements. With this in mind we designed our suspension system

with a minimal amount of Toe-In/Out. In previous car designs there has been

as much as 5.5 degrees of toe out over a 3-inch wheel displacement. Thus,

our results were a considerable improvement, that is a maximum amount of

less than a 0.4-degree angle over 4-inch wheel displacement. This amount

should not affect to the predictability of the vehicle.

Wheel Displacement

3

2

1

(in)







0

-0.4 -0.3 -0.2 -0.1 -1 0 0.1 0.2

-2

-3

Toe-In/Out (degrees)





Figure klj – Toe-In/Out vs. Wheel Displacement





Another aspect that must be considered is the Caster Angle. Looking from the

side Caster Angle is the angle between the steering axis and the vertical plane.

The combination of Caster Angle and Kingpin inclination greatly affect the

handling of the car. They are very important since they influence both the

steering forces during lateral acceleration and the self-centering effect of the

steering. As with Toe it is desirable to minimize both the Caster Angle and the

Kingpin Angle for all wheel displacements. The combination of the two has a

large effect on the rate of camber change during wheel displacement. The final

design uses pneumatic trail to provide steering center effect at higher speeds.

Pneumatic trail differs from the mechanical trail defined by the KPI and caster

angle by specific tire characteristics. Pneumatic trail is a result of the tire patch

area shape. The patch area forms roughly a triangle thus providing a wedge effect

with the ground and provides a horizontal centering force. Both types of trails act

as weather vanes to the steering wheel but have varying effects at over a given

speed range. Mechanical trail is dominant at low speeds while pneumatic trail at

high speeds. The advantages of the low speed mechanical trail are not as high as

the geometry advantages and therefore are not used. Skid warning is also

maintained by minimizing mechanical trail and since the effect of pneumatic trail

is non-linear with vehicle speed the driver will be able to sense when there is a

significant decrease in pneumatic trail. This provides an important source of

driver feedback at higher speeds and the vehicle will exhibit under steer and feels

“loose”.









Wheel Displacement

3

2

1

(in)



0

-0.04 -0.02 -1 0 0.02 0.04 0.06

-2

-3

Caster Angle (degrees)





Figure kjl – Graph of Caster Angle vs. Wheel Displacement

Wheel Displacement









3

2

1

(in)









0

-4 -2 -1 0 2 4

-2

-3

Kingpin Angle (degrees)





Figure lks – Graph of Kingpin Angle vs. Wheel Displacement







Steering is another important factor of vehicle control on driver response.

A major focus of our system was to incorporate Ackermann’s principal. The basic

idea of Ackermann’s is that the control arm must connect to the spindle along an

imaginary line that connects the spindle and the center of the rear axis. By doing

this, the fact that the outer tire has a larger turning radius than the inner is

supported. Our control arm will connect in front of the spindle and therefore must

be set into the wheel. For this reason, as mentioned above, the bump steer cannot

be reduced considerably with the existing wheel offset.









Control arm connection





spindle

Center of rear axle

Angle for Ackermann’s

Figure ljj – Diagram of Ackermann’s Principle

Suspension Set-Up Wheel Disp. Camber Roll Center





Wheel Disp. Negative

Body

None

Wheel Disp. Positive





Wheel Disp. Camber

Body

Always Negative

Wheel Disp. Camber





Wheel Disp. Camber

Body

Always Positive

Wheel Disp. Camber





Wheel Disp. Camber

Body

Always Positive

Wheel Disp. Camber





Wheel Disp. Camber Positive Majority of

Body

the Displacements

Wheel Disp. Camber





Wheel Disp. Camber

Body

Always Negative

Wheel Disp. Camber





Wheel Disp. Camber Negative Majority of

Body

the Displacements

Wheel Disp. Camber



Figure lkj – Suspension Set-Up Types

Force Calculations

Load transfer Calculations, steady state:

1. Static weight:

The load transfer calculations using the following parameters to model the forces

generated. The static forces are calculated using a driver weight of 175 lbs, vehicle

weight of 520 lbs estimated from previous vehicles, front-to-rear weight distribution

50/50, and left-to-right weight distribution 50/50. The static roll centers are

geometrically determined using a string computer then checked using SuspensionGen

software. The height of the center of gravity, cg is estimated from average of moment of

inertias of major components.

Height of cg wheel base, wb

11 65

front track rear track

50 51.6





weight driver, wd front static roll center

175 4

weight vehicle, wv rear static roll center

520 6

total vehicle weight, tvw = wd + wv

695

front bias Positive Acceleration

0.5 1.5

rear bias Negative Acceleration

0.5 1.5

left bias Lateral Acceleration

0.5 1.5

right bias

0.5

Figure jlj – Acceleration Calculations

2. Lateral Load transfer due to lateral acceleration:

The lateral forces generated act through the center of gravity and are summed as a

torque or moment to determine the vertical force on a tire. The moments are summed

about the roll center axis where roll centers are calculated using tire deflections of xx

positive left and xx negative right modeling a left hand turn with body roll. The two

moments are summed to find Ftire then split front to rear by multiplying the bias ratio.

The lateral acceleration is a high value of 1.5 times the force of gravity, 1.5 g’s. This

number is used for two reasons, the first is that during the competition the car is tilted 45

to inspect fluid leaks under maximum lateral acceleration and also that most cars are not

capable of holding more than 1.2g’s.

F2



z L4

x Fz,lr L7

Lrc,r



y L5

L6

F1

L2



Fz,lf Lrc,f





Figure jlj – Graph of Relevant Forces







Front Rear

L1=Lcg-Lrc,f L3=Lcg-Lrc,r L5

6 4 11

L6=wb*front

L2=front track/2 L4=rear track/2 bias

25 25.8 32.5

F1=al*(Fz,sfl+Fz,s L7=wb*rear

fr) F2=al*(Fz,srl+Fz,srr) bias

521.25 521.25 32.5

Ft=(F1*L1+F2*L3)/(L2+L4)

102.6082677

Figure lks – Vertical Tire Force Calculation







Fz, Lateral Acceleration Loads, steady

state

Front, 1 tire

51.30413

Rear, 1 tire

51.30413

Figure lsd – Lateral Acceleration Loads









3. Longitudinal weight transfer due to negative acceleration:

The rear-to-front weight transfer due to braking is a sum of moments about the y-

y axis defined by the intersection of the x coordinate of the cg projected on the ground.





F1 L5

L7 L6









Tire Fz, Rear Tire Fz, Front

Figure ljs – Tire Force Schematic









Fz, Longitudinal weight transfer, negative acceleration, steady

state

Front Front Left Front Right

264.6346 132.3173 132.3173

Rear Rear Left Rear Right

-264.635 -132.317 -132.317

Figure sds – Longitudinal Weight Transfer

4. Maximum Loads achieved:

The maximum vertical loads that could be reached correspond to the combined

forces of negative and lateral accelerations with the static weight of the vehicle.





Fz, Maximum achievable loads; lateral + negative accelerations

static + lateral + negative

Front Left Front Right

357.3714 357.3714

Rear Left Rear Right

92.73683 92.73683

Figure ses – Maximum Achievable loads

5. Maximum Tractive Forces:

The traction generated by a tire, Fx,y is a function of vertical force exerted on a

tire and the coefficient of friction,  between the tire and ground. The coefficient of

friction is a function of many variables, including velocity, temperature, tire wear, wet,

dry, and/or sandy surface conditions. The coefficient is estimate to be high, 1.5 so that

the x and y components of the forces developed are high since these forces will be used

when determining component materials and dimensions. It is not a factor of safety but

only an added margin of safety.









z

x





y



Figure sle – Schematic of tire with axis







Horizontal tire force, Fx,y

Fy: Cornering=u*(static + lateral acceleration

loads)

Front Left Front Right

337.5812 337.5812

Rear Left Rear Right

337.5812 337.5812

Fx,y: Cornering & braking=u*(static + lateral + negative acceleration

loads)

Front Left Front Right

536.0572 536.05716

Rear Left Rear Right

139.1052 139.10524

Fx: Braking=u*(static + negative acceleration

loads)

Front Left Front Right

459.101 459.10096

Rear Left Rear Right

62.14904 62.149038

Figure sld – Horizontal Tire Force







.

Engine Group:

The engines that are used in a Formula SAE race car are 4-stroke piston engines

with no more than 610cc of displacement. The majority of the teams use mid-size

motorcycle engines because of their exact match to the SAE rules. However other

engines can be used as long as they meet these displacement and type requirements. The

engines are allowed to be supercharged or turbo-charged. The only other major

restriction is the fuel types; SAE allows the use of 94 and 100 octane Sunoco gasoline or

M85 (methanol) fuel. These are the major constraints that are placed on the team when

designing the drivetrain.

The engine group’s responsibilities are to decide on a suitable powerplant and the

systems that will control it. These controls will be used to reliably run the engine, while

also increase its performance. Some of these controls will be new designs and others will

be an improvement to existing equipment. The major controls that are needed to run the

engine are a fuel injection system and the integration of the existing ignition system to

work with the fuel injection system. These controls will allow the team the ability to tune

the engine for various conditions.





Powerplant Selection:

With the limitations on the engine displacement, there are no automobile engines

that would meet the design criteria. However, mid-size motorcycles have four-stroke

engines with a displacement of 600cc. The engines in this class are generally used for

high performance and racing applications, which makes this type of engine a suitable

option for the Formula SAE car. With the cars weight just above that of the motorcycle

that the engine is coming from, the power and torque of the engine are also suitable for

this type of application.

Before the engine selection process was to begin, some constraints had to be set.

We wanted to select an engine that has had proven reliability and performance, while

maintaining cost within our budget. Another recommendation that was made early on

was that we should have two motors for this project, one for dynamometer (dyno) testing

and one for the FSAE competition. The reason for two motors is due to the hard stress

and extensive wear associated with dyno testing and tuning. This would also help to

reduce the chance of engine failure in the competition.

There are a number of motorcycle manufactures that produce the mid-size 600cc

high performance motorcycles. These companies are Honda, Yamaha, Suzuki, and

Kawasaki. Through researching these manufactures we found that all of them have been

manufacturing these types of motorcycles for some time and have had good results with

them in motorsports competitions. But through contact with Momentum Racing in

Fairfield, CT, we learned more about these manufactures. The engines from Yamaha and

Kawasaki have shown evidence of failure under hard loading in a short period of time.

These engines also have a high cost for repairs and parts. The Suzuki manufacture has

had great success in recent months with their new 600GSXR engine in races. However

due to its recent development the cost associated with this was still high, and would not

allow for two engines under the budget. The engines produced by Honda have had very

good success and are well respected by racers. The 600cc engines have been in

development for almost 15 years now, with four different generations of engines. The

most recent generation of engines is known as the F4, which is only two years old. This

engine produces the most horsepower out of all of the Honda motors and is known for is

reliability. However it too is still pricey due to is recent development and availability.

For us to afford two engines for the project we looked at the Honda F3. This motor is

also known for its power and reliability. The F3 and F4 engines are similar but have

different dimensions, the power is also decreased with the F3, but the torque is higher

than that in the F4. The F4 engine is rated at 92-95 Hp and the F3 is rated at 88-90 Hp,

these power ratings are from the manufacturer. The increase of torque over the F4 is due

to the shorter piston stroke and this increase was an important consideration due to that it

would help the cars performance due to the driving conditions. Another note was that

some of the components for the F3 and F2 engine were the same. This was a

consideration due to the abundance of F2 engine components from previous teams.

The decision to use the Honda CBR 600 F3 engine was the most practical option to meet

the driving conditions and budget constraints. The F3 engine is rated at 90 Hp, which is a

significant increase over previous year models and only a 5-7 Hp decrease in power from

the F4 model. The high torque rating from the F3 also would work better for the car

performance due to the layout of the competition. The power output is also comparable to

other manufactures engines. The extra parts from previous teams also were a deciding

factor in the choice of engines.





Additional Engine Decisions:

Under the rules from SAE we have the options as far as induction and fuel. For

the induction we can use natural aspiration or forced aspiration. A naturally aspirated

induction for the engine would be a tube or scoop to pull air to the carburetor or fuel

injection system. A forced aspiration set-up would utilize a supercharger or turbo-

charger to intake air to the carburetor or fuel injection system. The benefit if properly

tuned to that particular engine is between a 30 and 40% power increase. From looking at

the previous teams at UConn and other teams, there has been the difficulty of finding a

turbo small enough to match the engines displacement. When forced induction is used

the compression ratio needs to be lowered to handle the extra air mass being delivered.

This would require internal engine modifications to lower the compression, such as

different pistons. Another difficulty has been the tuning of the turbo to produce

maximum power. It has been noted by other teams that their vehicles have also not been

able to achieve decent fuel economy when forced induction is used. From looking also at

the power output numbers from the dynamometer from last year’s competition, more

horsepower was created with natural aspiration. Therefore we have decided that we

would use a naturally aspirated system, and avoid the problems of forced induction.

The other decision was as to what fuel type and octane to use. The SAE offers

three choices on fuel; they are Sunoco 94 and 100 octane gasoline and M85 (methanol).

Looking as the pros and cons of gasoline and methanol we were able to select our fuel

type. Methanol is rated at about 118 octane, and has a high energy associated with it than

gasoline. However methanol is corrosive, and can do serious damage to engine

components, which leads to power losses if the components are not replaced. With this

side effect, it was clear that methanol would not be used for this reason. This was also

based on a tight budget already. Now the choice was to what octane level to use. The

engine we will be using has a compression ratio of 12:1. This number is relatively high

for a gasoline engine. Therefore to prevent pre-detonation “knocking”, the higher octane

would be used. Generally the higher the compression ratio the higher the octane level

needed to prevent knocking. The 94-octane fuel would be used for forced induction

systems, where the compression ratio is lower for the use of the turbo or supercharger.





Fuel Injection:

The major component that is needed for the engine is a fuel injection system. The

reason for a fuel injection system is that the stock carburetors for motorcycles are not

designed to take lateral forces. If the stock carburetors were used in our car, as the car

made hard turns the lateral forces in the turn would cause the fuel to spill from the

carburetors and the engine would be starved. The reason that this carburetion system is

sufficient in a motorcycle is because when the bike turns it will lean into the turn and the

force from the turn would be transmitted in the vertical plane of the bike. Therefore the

lateral acceleration felt in a four-wheeled car, is a vertical acceleration on a bike with

respect to the position of the engine. The rules for SAE also require that all fluids such as

fuel be contained, therefore the spilling of fuel is illegal.

There are a number of fuel injection manufactures that produce programmable

electronic fuel injection (EFI) systems for this type of application. Some of the

companies that manufacture these types of EFI systems are HALTECH, Edlebrock, and

Accel. These systems tend to cost around $1500. The systems from these companies are

complete, with all the wiring harnesses, sensors, and EFI computer. The computers for

these systems just need to be programmed for engine constraints such as displacement

and redline (maximum engine speed) limits. These EFI systems do have their limitations

as far as tuning to maximize performance.

Therefore another consideration was to design our own EFI system, which would

be similar to the aftermarket systems. This approach would allow us to design the EFI

computer to the exact constraints of the F3 engine. This would also allow us to

customize the system to our needs, with features that we want. One of these was the

ability to easily set the computer to either performance or economy mode. In the

performance mode we would be able to produce the most power and this would be used

in events when maximum power is needed.

A number of different designs were considered before the final control system

scheme was chosen. Initially the fuel injection was going to be accomplished using

multi-port injection, firing all the injectors at once, and the spark would be delivered to

pairs of companion cylinders together in a waste spark system. This was based on the

assumption that the engine to be used was a Honda F4 engine having an ignition system

similar to that of the F3 engine. This would be a system that has two ignition coils and

four cylinders. In this type of system the method of firing the coils is done using a series

configuration of the ignition coils that fires two spark plugs simultaneously in a scheme

known as waste spark. This type of system would also not have a camshaft position

sensor, so the multi-port fuel injection system would be the necessary fuel delivery

scheme.

When the wiring diagrams for the F4 engine became available it was discovered

that the ignition system was a coil on plug system. This system has an individual coil for

each cylinder, making the waste spark system somewhat redundant and inefficient since

it is no longer necessary to fire two ignition coils at once. The F4 system has no

camshaft position sensor, so the engine position with respect to which cylinder is on a

power stroke is not readily available. This system does not lead to an easy design for

sequential fuel and ignition control.

A scheme has been devised by the EE/CMPE team to obtain the engine’s power

stroke position from the voltage difference found at two companion ignition coils primary

windings during the inductive discharge. The way that this is done is to start the engine

in a waste spark and multi-port injection mode. Since at start up the only information

available when the engine is the overall engine position from the crankshaft position

sensor. This is only half of the information needed for sequential fuel injection. A four-

stroke engine’s cylinders travel up and down twice to make a complete cycle. These

means that the engine is on either number one, or number four cylinder compression

stroke when the crankshaft is at 0. It is therefore necessary to start the engine in a waste

spark mode. Number one and number four spark plugs are fired simultaneously just

before 0, and number two and number three spark plugs simultaneously just before 90.

Once the engine is running the primary voltage drop across two coils will be measured

simultaneously. This will be done at the 1-4 or the 2-3 cylinder pairs. This has been

successfully simulated using Pspice software and a comparator circuit. A voltage pulse

4ms in duration and 100 to 140 volts in amplitude was used based on the following

assumption.





The electrical engineering portion of the Formula SAE racecar project involves

designing an Engine Control System that consists of two subsections, ignition control

system, and a fuel injection control system.





The design approach for the Engine Control System has been largely done using

assumptions, since the engine has not been available for testing here on campus. The

data that is required for the final design and implementation is not readily available.

There is some basic information available through service manuals and similar

documents. However, the manufacturers of engines do not make their engineering

information available to the public. For this reason some of the details of the design will

not be known exactly until the engine is on premises. Nevertheless, there is enough

information available to complete the design, aside from the specific code for the

microcontroller, and some scaling of resistor and capacitor values.





The ignition system is the device, responsible for supplying sufficient voltage at

the correct time to the spark plugs for combustion to occur. As the name implies, the

spark plugs create a spark, which comes in contact with fuel and subsequently ignites an

explosion in the cylinders of the engine. The explosive force on the pistons forces engine

motion to occur.









The fuel injection system is the device, which supplies fuel to the cylinders where

the spark occurs. The fuel injectors are mechanical devices, which rely on DC voltage

pulses to activate them. The duration of time that this voltage is applied to the injectors

determines how long the injectors are open. In addition, the frequency at which the

injector signal is applied determines the volume of fuel supplied per second.

The most common types of fuel injection systems currently used are multi-port

fuel injection and sequential fuel injection. These are both systems where the fuel is

delivered to each individual cylinder by using separate fuel injectors for all the cylinders.

These injectors are located in the intake manifold runners in close proximity to each

individual cylinder. In multi-port fuel injection the injectors are actuated in groups (also

known as bank firing). In sequential fuel injection the injectors are actuated individually

just before with each cylinder’s combustion stroke. For sequential fuel injection the

engine stroke position must be known. This is usually accomplished through the use of a

camshaft position sensor. If multi-port fuel injection is used, there is no need for a

camshaft position sensor to give an indication of the engine’s stroke position. Sequential

Fuel injection offers the lowest exhaust emissions and the most efficient use of fuel, so it

is usually preferred over multi-port fuel injection by automobile manufacturers

The type of ignition control system designed is a sequential ignition system. This

is a method, which activates the spark on cylinder individually pairs. A voltage pulse is

sent to the ignition system that activates the spark to each cylinder. The spark is faster as

the frequency increases of the pulse increases. The pulse width of the spark is set at a

constant because the amount of spark needed for combustion does not change with

different speeds.

The inputs to the ignition control system as a whole are engine speed, crankshaft

position, and engine load. Engine speed is used because as speed increases the frequency

of the sparks will increase. Engine speed is directly proportional to the frequency of the

sparks. Crankshaft position is required as an input to the ignition system because the

position of the pistons determines when the spark plugs should fire. Engine load is used

because as the engine load is increased there is a need to adjust the timing. All of theses

inputs are sent to a microcontroller. Based on the inputs, the microcontroller is coded

appropriately to send the required pulse to the driver circuits. The driver circuit is placed

to act as a switch, which provides 12 volts to the primary ignition coil based on the

position of the pistons. Then the 12-volt supply voltage is applied at the primary coil, and

then transformed (step-up) voltage resides at the secondary coil. The interruption of the

voltage supply at the primary coil causes an inductive discharge at the secondary coil.

This amount of voltage at the secondary coil fires the spark plug. The high voltage at the

secondary coil induces a large transient voltage at the primary coil during the secondary

discharge. In order to protect the microcontroller, and to source enough current to the

ignition coil, an insulated gate bipolar transistor (IGBT) is placed as the driver circuit for

the ignition coil.

The inputs to the fuel injection control system are the engine speed, engine load,

throttle position and ambient air temperature. The fuel injector is a mechanical system

and is activated by a voltage pulse. The amount of fuel injected and the frequency of the

injections are determined by the pulse duration (or pulse width) and the frequency of the

voltage signal respectively. The pulse duration changes with respect to the load on the

engine i.e with a bigger load, the amount of fuel supplied increases and therefore the

pulse width is increased. As engine speed changes, the frequency of pulse changes with

it. Throttle position measures the position of the throttle plate. The throttle plate is

located near the intake, which opens and closes to let more air in as the accelerator is

depressed. When more is let in engine load increases also. If the throttle is wide open

the fuel injection system will increase the pulse width to supply more fuel. The pulse

width fluctuates with throttle position. Ambient air temperature affects the air to fuel

ratio. When the temperature is lower the density of the air changes and more fuel is

required to achieve the stoichiometric ratio.

The crankshaft sensor determines engine speed and position. On the end of the

crankshaft, there are nine teeth and then a large gap. As seen in figure 3. The sensor

mounted on the engine case will send a signal when a tooth has gone by. The frequency

of these signals will tell the speed of the engine. When the large gap is encountered, the

exact position of the crankshaft is known. Knowing the position of the crankshaft will

tell you exactly where all the pistons are. Knowing where each piston is, is not enough to

find the exact position of the engine. Since the pistons are moving in pairs the crankshaft

sensor will not tell us which of the cylinders is on its power stroke and which is on its

exhaust stroke. To find this a different approach was taken to this problem. When the

car is first started up the position of the engine for that is need for proper sequential

injection and spark is not known. To solve this problem the ignition system is first

started in a waste spark mode. In the pair of cylinders, one of the cylinders is in the

power stroke and the other is in the exhaust stroke. When the pair of spark plugs fire, the

spark to the cylinder in the power stroke is only spark that will induce combustion. Since

the other cylinder is in its exhaust stroke, the spark will fire into exhaust and no

combustion will occur. This spark is the “waste” spark. At this time the injection is

started in a bank-firing mode. Bank firing is firing all the injector at once. When all of

the injectors are fired at once the car will run very rough because of fuel that is injected

into an open valve or called open valve injection (OVI). OVI is very fuel inefficient.

When the waste spark mode is used, one of the spark plugs is firing into exhaust while

the other is igniting compressed air/fuel. When the spark plugs are fired there is a

transient voltage present at both ignition coils. The size of the voltage is directly

proportional to the pressure in each cylinder. When the pressure is greater, the voltage is

greater. When a cylinder is in its power stroke the piston comes nearly to the top of the

cylinder compressing the fuel mixture thus creating a large amount of pressure on the

magnitude of 8-40 atmospheres (ATMs) from idle to high engine speeds. When a

cylinder is in its exhaust stroke the pressure in the cylinder is only 1-3 ATMs. There is a

vast pressure difference between the two cylinders so there is a large difference in the

transient voltage. If both of the voltages are put into a comparator the output of it will

tell us which cylinder has the greater voltage hence telling us which cylinder is in the

power stroke. Once the correct engine position is known the injection and ignition

system can be synced up and fired sequentially









The Manifold Absolute Pressure (MAP) sensor measures the amount of pressure

or vacuum in the manifold. This sensor tells the engine load. With more pressure, the

engine load increases. With greater load on the engine, more fuel is required. This is an

input to the microcontroller and will be a factor when setting the voltage pulse width for

the fuel injectors.

The Throttle Position Sensor (TPS) tells what the position of the throttle plate is.

When the TPS measures a wider opening this will lead to greater engine load. When the

TPS is suddenly forced into WOT the accelerator has been depressed rapidly causing a

need for more fuel. The TPS senses the position by using a potentiometer that turns

when the position of the plate changes. Since this is a potentiometer there will be a

change in voltage in different positions. This voltage is then sent back to the

microcontroller.

The ambient air temperature is measured using a thermistor. A thermistor is a

variable resistor that varies with the temperature of the air. If a voltage is applied to this

thermistor then each voltage will have a corresponding temperature.





EFI Requirements:

The EFI system uses particular constraints to properly control the engine.

Therefore before the computer code can be finished the system needs to know how much

fuel is needed, air/fuel ratio, and injecting timing. Other system configurations need to

also be addressed to proper design the computer to the engine and competition

requirements.

The fuel required by the engine can be determined based on the engine's

displacement. An engine is essentially a pump and the displaced mass the pump creates

is what the engine needs to intake. Therefore the first variable that needs to be

determined is the mass of the air that the engine displaces. This can be done using the

Ideal Gas Law.

p V = m R T v

This law will be used to model one cylinder to determine its requirements; this is due to

that each cylinder will have its own injector.

Before the mass of air can be calculated there are some variables that need to be

determined. These are the pressure in the cylinder on the intake stroke, the volume of

one cylinder, and the temperature of the air within the cylinder. When an engine is at idle

the pressure in the manifold is at about 1/3 atm, and when the engine is at wide-open

throttle (WOT) the pressure increases to 1 atm. So when the engine is running at these

conditions the pressure in the cylinder is at the respective pressure based on the throttle

positions. For the cylinder pressure in this calculation we are assuming that the throttle is

wide open, therefore the pressure is 1 atm. The reason for assuming WOT is that

maximum power will be produced around this position of the throttle.

As far as the cylinder volume, this is done by summing the volume the piston

displaces and the cylinder clearance volume together. The volume the cylinder displaces

can be determined by dividing the total engine displacement by the number of cylinders.

In this case the engine has a total displacement of 599cm3 with 4 cylinders, which make

the displaced volume of one cylinder 149.75cm3. The clearance volume is the volume in

the cylinder above the piston when it is at top dead center (TDC) and the volume of the

cylinder head. To determine this the following equation is used:

CR = (Vdisp + Vclearance) / Vclearance

This can be used, using the compression ratio (CR) and the known displacement of the

piston. For our engine a 12.0:1 compression ratio is listed by manufacture making the

clearance volume equal to 13.623cm3. Using the determined volume the total cylinder

volume can be calculated, this is equal to 163.373cm3.

The temperature determination is a variable that is tough to control in this type of

situation; the weather conditions usually dictate the air temperature. When the engine is

operating at WOT, the temperature of the air in the cylinder can be assumed to be equal

to that of the ambient air. This assumption is due to the fact the air velocity is so great

that it does not sit in the manifold to be heated. Also additional heat shields and cooling

ducts can be added to insure this assumption. The EFI computer will have an air

temperature sensor to input the temperature for the current conditions. For the mass of air

determination here, the temperature will be equal to 70 oF, room temperature. This

assumption is based on the temperature when the injectors will be experimentally tested

for verification.

The final variable needed for the Ideal Gas Law is the volumetric efficiency, v.

The volumetric efficiency is the measure of the efficient volume of air that is actually

being inducted to the cylinders. This measurement is essentially a percentage of air

volume inducted. As engine speed and throttle position increase, this value for v will

increase as well. For the injector testing this variable will be equal to one, normalizing

the equation. This assumption is also based on the throttle position set at WOT. The

change in engine speed versus volumetric efficiency can be viewed in the following plot.

This graphical representation will also be used for the microcontroller programming.

Figure ##: Engine speed vs. volumetric efficiency

Now that the assumptions for the Ideal Gas Law have been made the mass of air

can be calculated. The units for this will be metric for ease of use; therefore the constant

R will be equal to 8.314 ______. The mass of air was determined to be equal to

0.006769 kg.

To determine the mass of the fuel, the stoichiometric ratio must be used. The

stoichiometric ratio is also known at the chemically correct ratio, and this is the ratio of

air to fuel. For a gasoline the operation limits are roughly 11:1 (rich) mixture to a 20:1

(lean) mixture. The stoichiometric ratio for optimum operation and lowest emissions is

14.5:1. At this ratio it is represented by saying  = 1. This is what the initial calculations

will be based on. However for fuel economy purposes a ratio 20% lean can be used for

fuel mass calculations, this would be about a 17:1 mix. For maximum performance the

fuel mixture would be richened about 10% to achieve max power, this would be a ratio of

13:1.

Using the stoichiometric ratio it was determined that the mass of fuel needed at

 = 1, to be equal to 4.6686 E -4 kg. This mass can now be converted to a volume using

the specific weight for gasoline, which is equal to 45.9 lb/ft3. This will need to be

converted for metric units. The volume of fuel needed for one cylinder at stoichiometric

conditions is 0.63497cm3. The change in the volume of fuel compared to stoichiometric

ratio can be seen in the graphical representation below.

Figure ##: Fuel volumes compared to stoichiometric ratios

This is the volume that each injector will discharge every cycle at the appropriate time.

The equation posted on the graph, is for the represented curve. This will be used also in

the future programming to fine tune the air/fuel ratio for different engine requirements

and driving conditions. The injection timing will be discussed later.





Pulse Width Determination:

Now that the volumes of fuel have been determined at different stoichiometric

ratios, the pulse widths for the injectors need to be determined so that those volumes can

be delivered. Fuel injectors are voltage dependent and this voltage input is used to open

the pintel and allow fuel to flow by. The duration of the voltage signal is a variable,

which controls the volume of fuel discharged. The fuel pressure is another variable

consideration. To determine the pulse width for the injectors, a test rig was designed to

represent the fuel delivery system. This can be seen in the layout below.









Figure ##: Fuel Injector Test Rig

Using this test rig, a particular pulse width can be produced from the signal

generator. This signal from the generator will be a square wave. This signal will be

transmitted to fuel injector driver chips. These chips are specially designed to control

and operate fuel injectors. This will then send the conditioned signal to the injector. As

for the fuel delivery to the injector, this will be done just as it would in an actual car set-

up. The fuel will be delivered from the tank via a high-pressure fuel pump (95-psi max.)

to the injector. The pressure at the injector will be regulated, to control the pressure

input. The fuel pressure regulator is critical to maintain a constant pressure at the

injector. This will also allow us to vary the pressure during the test. Most fuel injection

systems operate a fuel pressure of 40 psi. However it is recommended that the fuel

pressure should not exceed 60 psi, this recommendation was made by the fuel injector

manufactures due to internal component design constraints.

As mentioned earlier the duration of the pulse width and the fuel pressure inputted

to the injector have an effect on the volume of fuel discharged. In an automotive

application, the fuel pressure will vary with respect to the manifold pressure. This is

because a fuel injector works on the principle of differential pressure across the injector.

And this differential pressure is the systems operating pressure. In an earlier section it

was noted that as throttle position increases the manifold pressure begins to increase,

therefore to maintain a constant differential pressure the absolute fuel pressure must

increase. The fuel regulator will control the absolute fuel pressure from a vacuum tube

from the intake manifold. The diagram below shows the layout of the fuel rail, fuel

regulator and manifold.









Figure ##: Injector differential pressure layout

The differential pressure is governed by Bernoulli’s equation. This can be applied

to determine an absolute fuel pressure based on the desired fuel volume.

P + V2 = constant

 2

The determination of the proper differential fuel pressure is critical, if this is not

maintained and the pulse width is constant it could mean that the engine would be starved

for fuel. This starvation will cause the engine to operate at lean conditions causing power

losses and possible engine damage if the ratio is very lean.





Valve Train Timing:

The timing of the valves is an important variable that is needed for fuel injection

systems. This timing of the valves is the point at which the intake and exhaust valves for

each cylinder open and close. This measurement is based on the position of the crank

shaft, and is measured in degrees. Using the manufactures specifications from the

owners manual a linear valve timing diagram can be produced. This is a graphical

representation using linear bars to indicate when each of the individual valves are

opening and closing with respect to one another. A valve timing plot was developed for

the Honda F3 engine and can be viewed in Appendix ??. This tool will allow the EFI

programs a measure of when to fire the injectors in either the bank firing or sequential

firing mode.





Engine Intake Design:

Since a new fuel injection system will be utilized an intake manifold will be used

to deliver the fuel and air to each of the cylinders. The manifold will consist of a throttle

plate, restrictor, plenum, runners, and fuel rail. The main focus as far as design will be

the restrictor, plenum, and runners, since this is where the majority of tuning will be

done. Through the use of equations these components can be tuned to maximize engine

performance. A typical 4-cylinder intake manifold would look like something like the

figure below.

Figure ##: Intake manifold for a 4-cylinder engine

The restrictor is a requirement through the SAE rulebook. This is used to restrict

the engines from making their ultimate power. Restrictors are used in other racing sports

to reduce power and in turn lower speeds. This is more of a safety item, to limit the

vehicle from extreme high speeds. The rules state that a 20mm restriction will be placed

on the intake after the throttle plate. The restrictor does not have to be the normal

restrictor plate design, which is a thin plate (about ¼” thick) with the required bore. This

is a restriction anywhere in the intake after the throttle plate and before the intake valves.

Therefore a venturi can be utilized to restrict the intake system, also with the used of a

venturi a considerable amount of the air mass lost when using a conventional restrictor

can be regained. It should be noted that due to the restrictor the engine would only make

good power up to about 9000 RPM. This will also be tested later on the dyno for

verification. The venturi will be designed to minimize the air mass loss. This is done

with the following equations:

A2/A* = (A2/At) (At/A*)

where A2 = exit area

At = venturi throat area

A* = inlet area

m = (P A V) / (R T)

where m = mass flow rate

A = cross sectional area





The next portion of the intake that needs to be studied is the plenum and intake

runners. These two work together to delever the air to the engine. Inside the intake

manifold the air flowing through can experience pulsations or waves. This occurs due to

the air flowing in and hitting a closed valve then traveling back up the intake runner.

This resonants can be used to help increase performance if the runners and plenum are

designed correctly. The runners need to be a certain length so that as the pulse waves

travel away from they engine, they bounce back at the proper moment the intake valve

opens again. This is called a tuned induction or “ramming” and can result in considerable

power improvement of 10–20%. The length of the induction pipe will influence the

engine speed as which maximum benefit is obtained from the pulsating flow. This can be

seen in the following plot of the pipe length versus engine speed. (Stone pg. 310)









Figure ##: Pipe length benefits at particular engine speeds

This plot is only applicable to engines, which have single carburetors or fuel injectors per

cylinder. This is primarily due to the difficulty to design an intake manifold with a single

carburetor or fuel injector, because of the difficulty of optimizing the volumetric

efficiency and the mixture distribution. From the plot it is shown that a shorter pipe is

used when engine speeds increase. In our case the intake will be designed to take

advantage of this phenomenon at the optimum power band, which is

The intake can also be acoustically modeled to study the propagation of these

pulsating waves. A tuned induction system can be considered an organ pipe or

Helmholtz resonator. This is modeled by the Helmholtz equation below.



fH = C / 2 [ A / ( l*V ) ] ½

where C = speed of sound

A = pipe area

L = pipe length

V = resonator volume

The pipe variables are representatives of the intake runners, and the resonator is the

plenum. The other model for the resonates frequency is of the organ pipe, and is

represented by the following equation

fp = C / 4L

L = l + 0.3d

where L = effective length

l = pipe length

d = pipe diameter

Either of these can be used for the determination of the resonate frequencies. This tuning

technique will benefit the volumetric efficiency, and in turn power. However the

resonants frequency will only be good for particular engine speeds. The intake can

experience more than one resonating frequency. Therefore a focus needs to be made to

ensure that these resonating frequencies are present at particular engine speeds, mainly at

speed the engine will be run at for the competition.

Other design considerations are the intake runners from the plenum. From the

plenum air needs to be funneled into each of the intake runners. There is a tendency in

intake design to just connect the two. But at this junction a boundary layer is present.

Therefore to help minimize the boundary layer experienced at the mouth of the runners,

the runners are tapered or flared. This shape is sometimes called a velocity stack. This

runner shape at the mouth minimizes the boundary layer by helping the air to flow more

smoothly. With the reduction of the boundary layer the runner is less prone to choking

the flow. These runner shapes at the mouth can be seen in the following figure.









Figure ##: Runner Shapes at the mouth

As far as the rest of the intake systems components they do not need as much

attentions to maximize the performance. The throttle plate needs to have a matching bore

to the front end of the venturi to minimize airflow losses. The throttle plate also needs to

have a linear opening motion. If this motion is not achieved then, when the engine is run

the driver will not be able to adequately control the engine speed and output. A throttle

with non-linear motion will result in maximum engine speeds with out have fully opening

the throttle plate. This is undesirable for driving conditions.

The fuel rail is the last point of concern for the intake system. The fuel rail will

provide fuel to each of the injectors through a common pipe. The only concern for the

fuel rail is adequate fuel delivery for the injectors. Therefore the volume of the rail needs

to be looked at to ensure under hard accelerations that they rail holds enough fuel to

provide the injectors adequately. However this should not be to big of a concern with a

fuel regulator with a quick time response to meet the need for a higher fuel pressure in

the rail. It was also suggested that mild steel or stainless steel be used for the

construction of the fuel rail, instead of aluminum. This is because after a hot shutdown

steel gathers far less heat, which could lead to fuel boiling in the rail and causing vapor

lock. (sdsefi.com/tech)





Engine Control System Design:





This section discusses the parameters of the control system, which consists of the

Fuel Injection system and the Ignition system. A large part of our design requires the use

of a microcontroller to control the operation of these systems. All of the necessary inputs

to the microcontroller are further described in this section.

The table below shows each control variable and the sensor from which each

receives its signal.





CONTROL VARIABLE SENSOR

Fuel supply frequency Crank Sensor

Spark frequency Crank Sensor

Fuel supply amount (volume) MAP Sensor/Crank sensor/Air

temperature sensor

Spark timing (advance/retard) Crank Sensor/MAP Sensor

Fuel quality(rich/lean) Driver controls









MAP Sensor:

The MAP sensor detects the pressure in the intake manifold. The pressure is

important because amount of fuel necessary is dependent on the pressure in the intake

manifold. The output of the MAP sensor is a voltage level, which is dependent on the

load on the engine. The heavier the load the larger the voltage. Typically, the voltage

level ranges from 1V to 5V.

Based on the voltage present at the MAP sensor, the corresponding pressure can

be deduced by means of a pressure-voltage graph exclusive to the particular MAP sensor

being used. With the pressure and the RPM of the engine known, the volumetric

efficiency factor (v) can be found from the v vs RPM graphs or from look up tables

developed by the mechanical engineering team. The mass of air in the cylinders can then

be found using the equation below.

 v Pm V

ma 

RT

Where Pm is the manifold pressure, V is the volume, R is an ideal gas law constant and T

is the temperature of the engine.





With ma known, the mass of fuel is then calculated using the equation below.

mf = ma/14.5

Knowing the mass of fuel, the volume can be determined.









The MAP sensor signal is proportional to the fuel that is necessary at any given engine

speed. The MAP sensor signal is based on the following:



Low pressure – Intake vacuum is high – At idle – 1.25V (typical)

High pressure – Intake vacuum is low – At wide open throttle – Just under 5V (~4.8V)

Light load – Engine vacuum is high

Heavy load – Engine vacuum is low

Light load – low pressure

Heavy load – high pressure

Light load – low pressure – Intake vacuum high – lean fuel mixture and advance spark

timing.

Heavy load – high pressure – Intake vacuum low – rich fuel mixture and retard spark

timing.

Low voltage – High vacuum – Low pressure



High voltage – Low vacuum – High pressure



This depicts how intake vacuum, pressure and load affect the MAP sensor output.

The spark timing is varied based on the engine speed and the pressure in the intake

manifold. The intake manifold pressure and MAP sensor voltage have a linear

relationship. Using the pressure-voltage graph, the pressure in the intake manifold can be

determined since the voltage is known. This pressure along with the RPM of the engine

is used to determine a parameter known as the spark-angle (θspk). θspk is determined from

the θspk vs RPM graph or from look-up tables developed by the ME team. Spark-angle is

simply the number of degrees of rotation left in the crankshaft before the piston gets to

the top of the cylinder





Crankshaft Sensor (Variable Reluctance Sensor)

The crankshaft sensor detects the speed and position of the engine. The Variable

Reluctance Sensor (VRS) reacts to variations in flux density created by a rotating

multiple toothed wheel with a missing tooth region. The sensor contains a coil of wire, a

magnet, and a pole piece. The changing flux field induces an Alternating Current (A/C)

voltage in the VRS. One A/C cycle is generated for each tooth on the wheel. The signal

is typically 150 milivolts at minimum cranking speed, 30 RPM, and 300 volts at high

engine speeds. When the missing tooth region passes by the sensor the zero crossing of

the A/C signal changes. The signal will then clipped and used as an input to the

microcontroller. This signal returns both engine speed and position.

Two additional sensors are the Ambient Air Temperature (AAT) and the Throttle

Position sensors. The AAT sensor is a thermistor, a device that changes resistance with

air temperature. This device is used by the microcontroller to adjust fuel volume. The

TPS is a potentiometer, a device that changes resistance when it’s center terminal

changes position. The microcontroller uses this input to supply more fuel for hard

accelerations.





Microcontroller Applications:

The microcontroller supplies a voltage pulse to the ignition driver and the injector

driver circuits at the correct time.





Fuel Injection:

Obtaining the volume of fuel requires mathematical calculations. Subroutines

will be developed for the microcontroller to perform these calculations. After volumetric

efficiency is found, this value will be stored as an 8-bit binary number in a register

internal to the microcontroller. All other constants and variables would also be stored as

8-bit binary numbers. This is done because the microcontroller operates by retrieving

data from its internal registers. The math subroutines developed will be implemented to

perform 8-bit multiplications and 8-bit divisions. The calculated value for the mass of

fuel will act as a variable for determining the pulse width of the signal to the fuel

injectors.

The output signal of the crankshaft sensor has a shape that clearly depicts the

completion of one revolution. This signal will be converted to a digital signal using A/D

converters and stored in an 8-bit register. The microcontroller will be programmed to

send out a signal to the fuel injectors every time the crankshaft position sensor indicates

the missing tooth region has passed. This controls the fuel supply frequency









Ignition System:

Using the information from the crankshaft position sensor the microcontroller will

supply the frequency of the spark and adjust the timing as needed. The spark angle

indicates how many degrees before top center for spark activation. The speed of the

engine is of importance because the equation below would be used to determine the time

it would take to reach top dead center.

 spk 60 sec s

t= *

360 rpm





The microcontroller will be programmed to advance spark by this calculated time.





The outputs from the microcontroller are interfaced to the engine using driver

circuits. These circuits are designed to source enough current to the ignition coils and

injector solenoids. The microcontroller will also use the outputs of a comparator to

detect which cylinder is in it’s in compression stroke. The information is used to inject

fuel and spark sequentially.





Engine Interface Circuits:





Injector Driver Circuit:





The driver circuit for the fuel injection system was chosen based on integrated

circuits (IC’s) that are specifically designed for this application. The ON Semiconductor

CS452 injector solenoid driver was selected due to the fact that they meet the needs of the

system we are implementing, and they were offered as free samples from the

manufacturer.

The CS452’s are low side drivers, meaning they switch the ground side on and

off, that are controlled via a 5V transistor transistor logic (TTL) signal. They provide a

path to ground for a 2.4A current at the low side of the injector solenoid, which is

connected to a 12V DC source at its high side. They also possess a specific current drive

characteristic designed for driving fuel injector solenoids. These drivers initially provide

a low impedance saturated power switch (essentially a short circuit) until the drive

current reaches 2.4A, they then drop to about .6A until the injector driver is switched off.

This supplies the high initial current that is necessary to open the injector, and then drops

to a lower current level that maintains the injector open until the driver shuts off. This

prevents the injector solenoid from becoming too hot. A 40V zener diode configured in

parallel with the injector and the driver will protect the driver from transients above the

40V limit of the CS4352 driver.

Fuel Injection Driver Circuit (Including Microcrontroller)





Ignition Coil Driver:

Driving the ignition coils presents a distinct design problem due to the high

current and voltage levels present at the ignition coils. The ignition coils are actually

transformers that step 12V up to as much as 50kV. A pair of inductors that are in close

proximity form a transformer. The transformer steps up, or steps down the primary

voltage based on the turns ratio of the inductors, given by

n = Ns/Np

where Ns and Np are the number of turns in the primary and secondary windings.

The voltages and currents in a transformer are proportional to the turns ratio

n = Vs/Vp = Ip/ Is

When the current changes in one inductor it induces a voltage in the other inductor

known as mutual inductance. The induced voltage is given by

V = L(di/dt)

Where L is the inductance in Henrys

This phenomenon is known as mutual inductance, and it is what is used to fire the spark

plugs in an internal combustion engine. When the voltage supply is interrupted at the

primary side it causes the magnetic field developed in the primary side to collapse (V =

L(di/dt)) as the high secondary voltage arcs across the spark gap. The resultant

secondary current in turn induces a voltage in the primary windings.

The current drawn by the ignition coils is in the 10A to 15A range as the primary

side charges, and the induced voltage at the primary side of the ignition coil caused by

the secondary discharge is in the 100V to 200V range. These are both orders of

magnitude too high for the microcontroller to handle. This requires a driver circuit

between the microcontroller and the ignition coils.

The device that best handles these conditions is an insulated gate bipolar transistor

(IGBT). This is a combination of a bipolar transistor and a field effect transistor (FET).

This device can source large amounts of current, and it can handle the high voltage

transients found at the primary side of the ignition coil during secondary discharge. This

specialized transistor has been used extensively as an automotive ignition coil driver.

Recently Intersil introduced an IGBT, the HGTP14N36G3VL, specifically designed for

coil on plug applications. This a 14A 350V ignition IGBT .

These IGBT’s were selected for their low cost, and their availability as free

samples from the manufacturer. These drivers should be than sufficient for the

application at hand. The actual current draw and transient voltage at the primary ignition

coil will have to be measured when the engine arrives. At this time the design could

change slightly. However, the assumptions that have been made should hold.

One important control variable is the charge time for the primary ignition coil.

The time to fully charge the ignition coil needs to be known in order to deliver the spark

at the right time. Since the interruption of the primary voltage triggers the secondary

discharge, the coil should be fully charged when the spark is needed. This is

accomplished by working backwards from the necessary time for ignition spark. The

microcontroller will turn on the ignition driver circuit at a time just before the spark is

actually needed. The time must be at least sufficient to fully charge the ignition coil for

the control system to maximize engine performance. The primary ignition time to

saturation (full charge) is estimated by

Vpt = IsatL

where Vp is the primary voltage, Isat is primary saturation current, and L is inductance.

Rearranging gives

t = IsatL/Vp.

Using typical values of saturation current and inductance from Intersil data sheets, and

the primary voltage, which is known to be 12V, the saturation time is found to be 3.75ms.

The actual time to saturation will be determined when the ignition coils are tested. The

ignition driver circuit will source the necessary current for the ignition coils and insulate

the microcontroller from the high voltage transients found at the primary ignition coils.

Cylinder Compression Detector:





In order to perform sequential injection and ignition with out a camshaft sensor, it

is necessary to compare the transient voltage spikes at the ignition coil for cylinder

compression detection. Using the voltage divider rule,

 R2 

 R  R  * Vcyl

VR2   

 2 1

we dropped the corresponding voltages at the primary ignition coils from 130V to 13V

and 100V to 10V, respectively, for companion cylinders. Using 0.1uF capacitors in

parallel created a low-pass filter that spreads the voltage spikes for easy comparison

through an op-amp. The cutoff frequency had to be determined to find out which portion

of the signal to use. The equation is as follows.

1

fc 

2RthC

The circuit now supplies the microcontroller with an indication of which companion

cylinder is in its compression stroke.

An example of the logic of the Op-Amp output is:





If V cylinder 4 > V cylinder 1, then the comparator output is positive.

If V cylinder 4 2.5°,

Toe-in=-.36->.08°

No caster

KPI=3

Rear: Track=51”

Roll center height=4.8-8”

Camber=-2->2°

No caster

Predicted Center of gravity height=11”

Proposed weight distribution 50/50

True Ackerman steering angle

Ground static ground clearance of 2.5”

Spring rate=350lbf/in

Damping Coefficients of selected Fox Vanilla RC damper: Adjustable compression

between 0-8.75lbf*s/in, Adjustable rebound between 0-87.5lbf*s/in.

Market Research

Although there are several commercial fuel injection systems available, none fit our

needs as a complete system. They are either too expensive or do not provide all of the

controls the engine needs. A search of prices was done on www.rancefi.com and

www.sdsefi.com. The most notable models and manufacturers along with the cost and

explanation of the problems with each are summarized in Table 1 below.

Table 1: Various Commercially Available Fuel Injection/Ignition Systems:

Product Manufacturer Cost Description/Comments

74022A - Accel Accel $880.00 The 74022A is a general conversion kit

Closed Loop made to replace a carburetor. Therefore, it

Auto/RV does not include an ignition or timing

Computer and system control.

Harness Kit

EM-3 F Simple Digital $1134 - Provides fuel injection and crank triggered,

Systems (SDS) $1300 distributor-less ignition control on 4 and 6

cylinder engines.

E6K Haltech $1495 Provides fuel injection and ignition timing.

Includes a one bar manifold air pressure

(MAP) sensor, a 3 wire O2 sensor, a throttle

position sensor (TPS), air and temperature

sensors, software and cables.

According to price, the best low-end solution is the Accel 74022A closed loop injection

controller for 4, 6, 8, and 12 cylinder engines that are 600cc or higher. This system allows

for re-programmability of fuel injection and costs $880, but does not include ignition

control. A more advanced commercially available solution is the Haltech E6K. This

device is programmable and provides multiple feedbacks and more engine control, but at

a much higher price. There are additional solutions at different costs between these two

extremes.