Paper Number 03

Design of Formula SAE Brake Systems

Eric Hamilton
2008 FSAE Team Member
Eric Klang
Faculty Advisor
Wolfpack Motorsports
College of Engineering
North Carolina State University
Raleigh, NC 27695-7910

Copyright 2009 Wolfpack Motorsports
Introduction

The Wolfpack Motorsports team designs, builds, and
races a car in the Formula SAE Collegiate Design
Series. Formula SAE [FSAE] is a competition organized
by the Society of Automotive Engineers that draws over
100 international schools to the main competition in
Detroit, Michigan. Wolfpack Motorsports has a history of
strong finishes in the competition and continues to work
towards winning.

The objective of this project was to expand the students
knowledge of mechanical engineering by developing
software tools to help design a FSAE brakes system.
These tools were be used to help design the brakes
system on the 2008 Wolfpack Motorsports FSAE car.
The software was used to determine the pedal ratio,
master cylinders, calipers, and the number of rotors and
their geometry. The guiding factors of the design
process were maximum braking power, minimal weight,
and manageable temperatures while maintaining
reliability. These factors were analyzed in the student
developed software program. The brake rotors for the
2008 car were designed in Solidworks and analyzed with
Cosmo Express based on the results of this analysis.
This report aims to explain the details of the software
program as well as each step of the design process for
the brakes system. It also serves as a reference to
future members of Wolfpack Motorsports for brake
system design.

Design Objectives

The primary design consideration of a FSAE brakes
system is the number of rotors and subsequently their
placement inboard or outboard. The Wolfpack
Motorsports car has primarily been a three rotor car with
one foray into a 4 rotor (Front outboard, Rear inboard)
design in 2006. There has been much debate with
judges and amongst the team as to which setup is ideal
and as fine tuning the handling and team members
understandings of vehicle dynamics have improved the
4 rotor design has come back in 2008 with outboard
rotors front and rear. This allows for more tunable and
controllable dynamics - especially braking in corners. It
also helps with packaging issues to run the two smaller
rear rotors outboard.

Also associated with the rotors is their mounting. As of
2007 Wolfpack Motorsports has started using a floating
rotor with a fixed, opposed-piston caliper setup. This
allows the rotor to center itself in between the pads to
ensure efficient braking as well as a quick return to non-
braking once the brake pedal is released. This also
narrows down the caliper selection, which is generally
determined before sizing any other components.

Another important factor in a brakes system is the
balance of the car under braking. A car that is effectively
balanced will be able achieve the maximum braking the
system components are capable of as maintain stable
handling under braking. A bias bar allows for moderate
adjustment of the balance which is used primarily for fine
tuning the handling of the car. But to be utilized correctly
the cars brake system must be designed around the
geometry and mass of the vehicle. Thus the sizes and
number of rotors, diameters of master cylinders, and
caliper pistons must be calculated with regard to the CG
height of the car, wheelbase, and mass of the car and
driver.

Design Process

Component sizing and selection:
The Matlab program was written to streamline the
component sizing process. It allows a basic
configuration to be entered and will iterate to help
determine what size components to use. As more
variables are determined, the braking of the car can be
fine tuned further. This section will discuss the
functionality of the program in relation to how it is used
to size components.

The program consists of three windows: Car Properties,
Kinematics, and Heat Transfer. They are accessed one
at a time with buttons at the bottom of the program
window. Also at the bottom of the program window are
buttons to save and load configurations. These buttons
will either save or load all variables contained in the
program. The Car Properties window contains all the
variables related to the car that do not change due to
changes in the brakes setup but that said setup depends
on. The Kinematics window contains the bulk of the
program and is used to size components and determine
balance and braking forces. The Heat Transfer window
is used for checking the rotor designs to make sure they
meet fade and thermal heat failure requirements. The
background color of a box is white if it requires a user
input and grey if the program calculates the value. The
following flowchart explains how to use the program.
Details of each section follow the flowchart.


Figure 1: Flowchart for Brakes Program

Car Properties:
When the program starts it opens to the Car Properties
window. All of these values are necessary to perform the
brakes calculations and should be determined in
advance. The first box will show what, if any, file is
loaded. This program allows all the variables to be
saved and loaded to streamline comparison between
different vehicles. Weight distribution will automatically
be calculated as values are entered. The coefficients of
friction for the tires are estimated values. 0.65 has been
used for the kinetic value and 1.35 for static. The mass
moment of inertia can be determined with Solidworks,
but 3.85 lb ft s^2 should be a fairly accurate assumption.


Figure 2: Car Properties Figure Window

Kinematics:
The kinematics window contains all the calculations
relating to the dynamics of the vehicle under braking.
This is where most of the work will take place. It
calculates the deceleration of the car in Gs based on
the force of the pedal, the configuration of the bias bar,
the master cylinder sizes, caliper sizes and
configuration, pad type, rotor geometry, and the
interaction of the tires wit the road. For the program to
calculate the braking values and help the user determine
the optimal setup values must be filled out in every white
box.
Start:
Either start with a blank setup or load a setup. Loading an old setup should be easier for
setting up a new car as many of the variables will not change from car to car
Car Properties:
Input all properties of the car that do not depend on the brakes setup


Figure 3: Kinematics Program Window

The first box contains information on the pedal and
pedal ratio. While this will not affect the balance of the
car The main considerations here are making sure the
force required of the driver for lockup is not too high (120
is generally accepted as a maximum) and that the pedal,
Heat Transfer:
Input necessary data for heat transfer calculations and calculate. Check to make sure
variables are within their acceptable limits. Braking power can be compared to previous
car and rotor setups to ensure rotors and pads will not be overheated.
3D Plot: Find potential master cylinder and rotor sizes
This plot will help to determine what master cylinder selection will provide strong braking
at a reasonable pedal force and offset. The information found here is fine tuned in the 2d
plot.
Final Checks:
Input maximum pedal force without lockup and Calculate Current Settings
Ensure all details look acceptable like pedal force, offset, and line pressure.
2D Plot: Tune the balance
Compare best master cylinder and offset setups from the 3d plots. Can have several
windows open at once. Look for optimal braking Gs at a reasonable pedal force and
offset. Also look for acceptable lockup characteristics.
Kinematics:
Input necessary data for kinematics calculations. Force on pedal and offset do not matter
yet as they will be calculated with the 3d plot function next.
bias bar, and master cylinders are able to fit in the car
and be effectively actuated by the driver.
The bias bar is the way the balance of the car is
adjusted during testing. It allows pedal force to be
biased towards one master cylinder or the other by
turning the threaded rod. Turning it towards a master
cylinder will put more force on that cylinder during
braking. The aim of this program is to find a setup that
has a bias bar offset near zero to allow for plenty of
tuning in either direction during testing.

Master Cylinder
Pedal Force
L1 L2
Master Cylinder




R
Outer
R
Effective






R
Inner


The master cylinders determine the line pressure
delivered to the calipers and therefore the potential
clamping force of the calipers. Check the line pressures
to ensure that they meet the safety requirements of the
lines and fittings; an approximate limit for steel fittings
and lines is 2000 psi. Generally the master cylinder type
will be decided in advance and it is up to the brake
engineer to decide what size to use. There is generally
more opportunity to tune the balance of the car here
than with caliper or rotor sizes because there is a wider
range of sizes available in a package that will fit most
pedal box configurations. Always check master cylinder
sizes with suppliers. The most common sizes that are
useful in this application are 0.625, 0.7, and 0.75. It is
also worthwhile to check the shop to see what is
available.

The caliper bores are equally important as the master
cylinder bores in determining the clamping force of the
calipers. However the calipers are usually decided on
based on functionality, packaging, and weight. Therefore
the options for caliper bores are very limited and it is
generally better pick one and tune for it with master
cylinders. If the overall braking is still unacceptable then
a different caliper bore may be necessary, but it is
important to know the availability ahead of time.
Clamping force is double the force exerted by one side
of a caliper. A floating calipers piston force will be
balanced by the opposite side and therefore will be
double the force one side of the caliper exerts. An
opposed piston caliper will exert equal force through
each piston to balance the forces and therefore the
clamping force on the rotor will be equal to the sum of
the forces exerted by pistons.

The value for the coefficient of friction for the brake pads
is an estimate based on pads that have been used in the
past. This is a very dynamic value especially with
respect to temperature and the 0.47 is used as an
estimated value at operating temperatures around 600
to 700 degrees F. This value will not affect the balance
of the car as long as the same values are used across
different theoretical scenarios. It will, however, affect the
maximum braking of the vehicle.
The rotors require the most design time for the formula
car, but with respect to this program and sizing of
components in general the most important piece of
information is the effective radius of the rotor. This is a
point at which the pad is assumed to be acting and for
our purposes we assume it to be the midpoint of the
inner and outer radii of the rotor.








Figure 4: Bias Bar Free Body Diagram





Figure 5: Radii Diagram


The brunt of the design work for the rotors is in the
specifics of the mounting and surface geometry and will
be discussed in detail later. The type of rotor/caliper
setup must also be determined. The setup had been a
floating caliper with a fixed rotor up front and a fixed or
floating caliper with a fixed rotor in the rear. In 2008 all of
the rotors will be floating with fixed calipers. This offers
many advantages. The primary advantage is in braking
efficiency. It is much easier for the rotor to float and
react to the pads rather than the opposite. Also several
of the floating calipers had very poor floating
characteristics and could cause drag issues. The floating
rotor setup also helps to alleviate extra stresses on the
rotor by allowing it free movement to react to both the
pads and thermal expansion.

The number of rotors is an important design point as
well. It has been thought that is better to run three rotors
because of the weight savings of only one (albeit larger)
rotor, caliper, and mounting. The use of new calipers
has made this weight difference almost negligible this
year. However as the understanding of vehicle dynamics
of the team has increased the need for a 4 rotor setup
has increased as well. It was also thought that tuning
through the differential could create acceptable braking
in corners with a 3 rotor setup. However the car
generates a moment about its vertical axis as it brakes
into a corner and the tuning potential of the previous
differentials has not been able to overcome this.

The third column of data in the kinematics window is
used to calculate theoretical and actual braking forces
and deceleration. First values are calculated based on
the brakes setup and the force input to determine the
weight transfer, forces acting on the tires, and finally the
braking force. This information must be passed through
a script to determine if the tires are actually capable of
handling this level of braking. If the forces are too high
the coefficient of friction will change from static to
dynamic which corresponds to a tire sliding in the real
world. Therefore an iterative process compares the
theoretical braking deceleration to the actual
deceleration the tires are capable of to determine at
what point tires lockup and what the actual deceleration
of the car is.

With all this information entered it is possible to start
tuning the system. At his point a 3-dimensional plot is
utilized. This plot compares pedal force to bias bar offset
to braking deceleration of the car. This plot is used to
determine which master cylinder sizes are most
effective. The goal is to find a combination that put the
peak deceleration at a 0 offset on the bias bar and
somewhere around 80 to 120 pounds of pedal force. To
activate this plot, press the Plot Pedal Force vs. Offset
vs. Braking Gs button in the bottom left. It will iterate the
braking values across bias bar offsets from -10 to 10
and pedal forces from 0 to 200 pounds. A button at the
top of the figure window allows the user to rotate the
figure. This makes it easier to see what is going on and
determine where the peak braking is in relation to other
variables. With increasing pedal force beyond the peak,
front or rear wheels have locked up. This situation is to
be avoided in the real world and the data is not useful
theoretically either. Therefore this area of the graph is
not of much importance. However it is important to make
sure that the maximum braking possible occurs before
the front or rear tires lock up. In extreme cases is it
possible to generate a graph where maximum braking
occurs after a set of tires has theoretically locked.


The next plot utilized is Pedal Force vs. Braking Gs. This
is essentially a slice of the previous graph. It makes it
much easier to analyze exactly what is going on with a
given master cylinder selection. This is also where the
brake balance is dialed in. The other advantage of this
graph is that it allows for easy comparison of setups.
Data on the master cylinder sizes, offset, and rotor sizes
is printed in the bottom corner of the graph. Several of
these plots can be opened at once. It allows maximum
braking, force required for maximum braking, and lockup
behavior to be compared very easily. Once the ideal
setup is found here the component sizing is finished.
There is one more calculation function to calculate the
current settings. This is useful to determine specific
details about a brakes setup. For instance it can be used
to determine exactly what force the maximum braking is
at and then check line pressures and other forces to
make sure they are reasonable.

Figure 7: Brake Force versus Deceleration Gs Plot


Rotor Design

The design of the brake rotors for the Formula car is
based on a history of rotors that have worked and one
that failed. By analyzing previous systems it is possible
to tell approximately what mass of rotor is necessary to
ensure safe, repeatable braking. It also helps to
determine how much volume of rotor is necessary to
keep from overheating the pads.

Since 2003 the following people have worked on rotor
design: Luke Lambert, Eric Hamilton, Matt Anderson,
and J osh Weeks. The rotors were generally all designed
from laser cut 1020 steel with a thickness of 0.185.
There were slots cut into the rotors to help reduce
weight while providing a larger effective radius. The slots
also served to reduce thermally induced stresses in the
rotor. The same reasoning was used for the scallops
Figure 6: Pedal Force vs. Bias Bar Offset vs. Braking G's
around the outside. At one point rotors had been
designed with holes rather than slots but the fabrication
costs were several times more per rotor due to
complicated laser cutter setup. These rotors have
always been ordered from Martin Custom Products and
it is advisable to get the order in early because of a
sometimes long turnaround time.
Figure 8: Brake Rotor with Load Applied to Pad

The Cosmos Express package in Solidworks has been
used to analyze the rotors. At the time the program was
used it was rather limited in its analysis ability. Cosmos
did not allow for torque loading and could only do hard
restraints. Therefore a few methods were developed to
analyze the rotors that appear to yield acceptable
results. The methods involved mimicking a torque load
through a series of loads or restraints around the rotor.
The first change to the Solidworks model is a
split line to differentiate the area the pad would contact
the rotor from the rest of the rotor. This provides a point
to apply loads or restraints for the FEA. Next planes are
drawn perpendicular to the mounting points as well as
distributed across the face of the rotor and in the center
of the pad outline. The first method is to restrain all the
mounting points of the rotor. Next a load is applied to the
rotor on the pad parallel to the rotor. This should be
done to both sides of the rotor. Values for the force can
be determined from the Force of Friction (front or rear)
box in the kinematics window of the program. Generally
this number has been rounded up to the nearest 100
pounds. The pedal force should be set to the maximum
possible pedal force before tire lockup to determine this
value. The loading should be applied as one load across
the two sides of the rotor. The analysis for this method
can now be run. The second method is to switch the
forces and restraints. The forces on the mounting points
must be recalculated to determine the value of the force
at the new radius to maintain the same torque. The
forces should also be arranged to create a torque on the
rotor.


The next parts of the analysis will provide options to see
wherever the Factor of Safety of the rotor is below a
specified value. All the rotors have been designed to
have at the absolute minimum of a 1.0 Factor of Safety,
but it is definitely better to try to get a little higher. Most
models have had a FOS somewhere around 1.3 with
anywhere from 100 200 pounds added to the applied
load as well. The most important stresses to look for are
at mounting points and around the ends of the slots. The
only rotor failure in the past 4 years was on a rear rotor
on which the distribution of slots coincided incorrectly
with the scallops on the rotor; thus leading to a crack
and then complete failure of the rotor. When adequate
space, as determined through FEA analysis, has been
designed into the rotor there have been no failures.
Some experience is necessary here to help
determine what stresses the rotor will actually see and
what is a result of the way it is modeled. Sometimes
there are very high stress concentrations due to the
models defined lines for the forces and restraints that do
not exist in the real world. Cosmos also continues to
improve so it is very possible that this method of
modeling will be outdated and replaced by much more
accurate FEA models.

Figure 9: Brake Rotor Stress Distribution and Deflection
Other

The car is plumbed with 3/16 stainless steel hardline as
much as possible to reduce weight. Flex lines (-3) are
used to connect hardlines to brake calipers to allow for
movement of the wheel. AN-3 fittings are used for all
connections. While aluminum fittings have been used in
the past, it is much safer to use steel fittings. The brake
line pressure has increased some the past few years
and a DNF due to a failed fitting is not worth the weight
savings of the aluminum fittings. Also because the lines
are stainless steel and the available flaring tools are
simple single flares are used for all the hardlines. If
these are done carefully and matched with steel fittings
they will perform without leaking. A pressure switch is
plumbed into a rear line to operate the brake light. Also
note that for the past 4 carbon cars the chassis has
been designed to split where the rear meets the front.
This means that the brake and clutch lines must be able
to split here as well so fittings for this are required. One
method that has worked well for determining plumbing
requirements is to draw a complete chart of the brakes
system to determine exactly what is needed and then
order a few spares. There should be a
compartmentalized box in the shop with extra fittings,
but its a good idea to make sure there will always be a
spare.

There has only been one significant brakes real world
brakes test so far. There were IR heat sensors mounted
to the rear rotor and a front rotor. It turned out that the
operating range of the sensors was below the operating
temperature of the rotors. However the data showed that
the temperature would exceed the range of the sensors
but tend to oscillate back and forth in and out of range at
steady state. Therefore it was determined that the
brakes operate not much above the maximum range of
the sensor, or around 650 to 700 degrees Fahrenheit. It
also showed that the pads and rotors would be at least
300 400 degrees Fahrenheit within a couple braking
zones. This information is very useful when picking out
brake pads. It showed definitely what temperature range
was necessary, but also that it gets close to that range
very quickly. Future tests that could be beneficial are
stopping distance tests to give a comparison point for
different cars and a way to check the accuracy of the
program. It would also be a good idea to do some type
of repeated braking test to determine fade
characteristics.

Conclusion

Brakes are an essential part of a racecar, but in general
large performance gains are not made here. It is
important to design a system that is well balanced, offers
tuning potential during testing, and is reliable. It would
be exciting to see further research look into the heat
capacity of rotors more and balance analytical work with
real world testing.

References

Stone, Richard. And J effrey K. Bell. Automotive
Engineering Fundamentals. Warrendale, Pa: Society of
Automotive Engineers, Inc. 2004.

Limpert, Rudolf. Brake Design and Safety. Warrendale,
Pa: Society of Automotive Engineers, Inc. 1999.

Walker, J r., J ames. The Physics of Braking Systems
StopTech LLC. 2005.

http://www.stoptech.com/tech_info/The%20Physics%20
of%20Braking%20Systems.pdf
Accessed March 31, 2008.*

*The original information used from this site was
accessed in 2004 and is no longer available. The link
provided is an updated and much more in-depth version.