Professional Documents
Culture Documents
Duan
SAFETY SHEET
The Illinois Junior Academy of Science
Directions: The student is asked to read these introductions carefully and fill out the bottom of this sheet. The
science teacher and/or advisor must sign in the indicated space. By signing this sheet, the sponsor
assumes all responsibilities related to this project.
Safety and the Student: Experimentation or design may involve an element of risk or injury to the student, test
subjects and to others. Recognition of such hazards and provision for adequate control measures are joint
responsibilities of the student and the sponsor. Some of the more common risks encountered in research are
those of electrical shock, infection from pathogenic organisms, uncontrolled reactions of incompatible
chemicals, eye injury from materials or procedures, and fire in apparatus or work area. Countering these
hazards and others with suitable safety practices is an integral part of good scientific research. In the chart below,
list the principal hazards associated with your project, if any, and what specific precautions
you have used as safeguards. Be sure to read the entire section in the Policy and Procedure Manual of the Illinois Jun
Academy of Science entitled "Safety Guidelines for Experimentation" before completing this form.
Possible hazards
Bus moving during measurement.
Specific safety practices related to materials requiring endorsement sheets should be detailed on the specific
endorsement sheet and not included on this safety sheet.
Please check off any other possible endorsements needed. Include these documents in your paper and on your board.
_____Humans as Test Subjects for any projects involving humans including survey administration;
_____Microorganisms-for any projects involving bacteria, viruses, yeasts, fungi or protazoa;
_____Non-Human Vertebrates -for any projects involving fish, amphibians, reptiles, birds or mammals;
_____Tissue Culture-for any projects involving growing eukaryotic tissues or cell cultures;
_____Letter from institution where research was done or IJAS SRC, if an exception to the IJAS rules has been grant
ed...
SIGNED
Student Exhibitor(s)
SIGNED
Sponsor *
*As a sponsor, I assume all responsibilities related to this project.
Duan
Duan
Table of Contents:
Abstract
Safety Sheet
Title Page 3
Table of Contents
Acknowledgements
Background Research 7
Hypotheses
Design Plan
18
Flowchart Outline18
Explanation of Process 19
1. Brainstorming and Criteria
19
2. Obtaining Dimensions 19
3. Making 3-d Models
19
4. Simulation Set Up
22
5. Redesign 23
6. Analysis
Results
24
26
Images
26
Data
30
Cost Analysis
34
Data Analysis
35
Conclusion 37
References 39
Duan
Acknowledgements
I would like to thank my SIRS teacher Mrs. Christine Camel for teaching the class that allowed
us to pursue our own projects and for helping me whenever I was unsure and helping me set up
the computer Id use to work on this project. Id also like to give big thanks to Dr. Thomas
Wallner of Argonne National Laboratory who clarified many concepts and answered questions I
had as well as granting me access to certain articles that helped a lot during the background
research stage. Id like to thank Mr. Benjamin Brzezinski for letting me work on a computer
during his class and guiding me along as I learned how to create the bus 3-d models as well as
loaning me materials to help learn Autodesk Inventor. Im also thankful for Mr. Al-Bazi, a
technician at Niles North, who helped me get the CFD software running on the computer.
Duan
Question: Can an simple drag reduction add-on device for a school bus be designed that is
effective as well as economically viable?
Purpose: To design, optimize, and evaluate the practicality of 3 different drag reduction
devices using CFD and then optimize their geometries to achieve the highest reduction in Cd per
monetary unit.
Duan
Background Research
When the first steam-powered vehicles were created, aerodynamics was not an
important concern; the vehicles travelled at low speeds and mechanical inefficiency far
outweighed the role air drag played in fuel consumption. Today, the limited amount of fossil
fuels, their relatively high prices, and the threat global warming poses has placed a focus on the
improving the fuel economy of newer vehicles. Current school buses are not very fuel efficient; a
diesel type C -- standard frontal engine -- school bus only gets around 7 mpg on average
(Laughlin, 2004). At highway speeds, drag plays the dominating role in fuel consumption for a
large vehicle. At lower speeds, drag is still a sizable contributor to fuel consumption. Although
great improvements in reducing air drag have been made for cars, commercial vehicles like
trucks and buses still have a box-like shape that creates much more drag than a passenger
vehicle. The situation does not have to stay this way; many external add-ons been tested in the
past few years through wind tunnels and CFD, computational fluid dynamics, and many of these
devices have shown a significant effect on reducing drag. With the aid of computers, engineers
can design and optimize much faster than before. CFD allows designs to be tested faster and
cheaper. Simple add on devices have shown to reduce drag on a tractor-trailer truck upwards of
20% (Schoon & Pan, 2007). An drag reduction add-on device or devices could be used on a
school bus which also has an aerodynamically inefficient shape. With over 450,000 school
buses consuming over 550 million gallons of fuel annually (Laughlin, 2004), a small
improvement to the fuel economy of each would be significant.
Fluids are substances that do not have a fixed shape and deform when a force is applied
to them, like water and air. Depending on factors such as viscosity, velocity, and density, a
moving fluid will flow either laminarly or turbulently around an object in it. Laminar flow is when a
fluid flows smoothly in parallel layers around the object while turbulent flow is when a fluid
behaves chaotically and unpredictably. A dimensionless quantity called the Reynolds Number
predicts the way a fluid will flow at a certain velocity. It is defined as
Duan
R=
pVl
For low Reynolds numbers, flow is laminar. As the number increases, fluid flow becomes
turbulent and flow becomes fully turbulent at high Reynolds numbers like the ones seen in
vehicle and airplane aerodynamics (Talay, 1975). The Reynolds number at which flow
transitions from laminar to turbulent depends on the shape of the object the fluid is flowing
around. The value used for characteristic length for a vehicle is commonly the length of the
vehicle. Since density, length, and viscosity are constant at a given air temperature, Reynolds
number directly varies with velocity.
A moving object in a fluid experiences a force that opposes its motion called drag. The
drag force is approximated by the drag equation (Elert, n.d.):
FD= CDAv2
FD is the drag force in N
is the density of the fluid, in kg/m3
CD is the coefficient of drag, unitless
A is the cross sectional area, m2
The coefficient of drag is a number that sums up all the different drag forces on an object at a
given Reynolds number and gives a measurement of how easily it moves through the air .
Generally, for turbulent flows with high Reynolds numbers, the shape of an object determines its
coefficient of drag. Bluff bodies, which are non-streamlined objects like a box or sphere, have
high drag coefficients while streamlined shapes like an airplane wing have low ones. A flat plate
perpendicular to the airflow has a CD value around 1.28 while an airfoil, the shape of an airplane
Duan
wing, has a low CD value of around ~0.05 (Shape Effects on Drag, n.d.). The behavior of air
flow around a sphere and its drag coefficient is highly dependent on Reynolds number, causing
the CD to range from 0.07 to 0.5. Since the air flow around a vehicle is turbulent and the
Reynolds number is large for almost all speeds, drag coefficient does not vary much at different
Reynolds numbers. One can find the drag coefficient of an object if they know the drag force,
density of the fluid, cross-sectional area of the object, and velocity of the object by rearranging
the equation for CD to get
CD=2FD/(Av2)
The drag force can be determined through experimentation or simulation. The cross sectional
area and velocity can be measured, and the density of air is known. Seen in the CD equation,
drag force varies with the square of speed and therefore the power needed to overcome drag
increases with the cube of velocity. Thus, drag becomes the main force a vehicle has to
overcome at highway speeds(The tyre: Rolling resistance and fuel savings, 2003).
Fluid motion is described by the Navier-Stokes equations, a set of partial differential
equations. The problem with these equations is that solving them directly for real world
scenarios like the flow around a vehicle is almost impossible due to the chaotic behavior of
fluids. Instead, numerous equations that approximate the solution have been developed and
enable computer simulations of fluid motion to produce relatively accurate results with
computational fluid dynamics(CFD). CFD uses computers to approximate a solution using
simplified versions of the Navier-Stokes equation, RANS, Reynolds-Averaged Navier Stokes
equations, which can be iterated along time (Reynolds-Averaged Navier Stokes Equations,
n.d.). A turbulence model is used with RANS simulation, the standard being the k-epsilon model.
Another turbulence model commonly used in external flow analysis is the SST k-omega model,
which can produce more accurate results and is recommend for external incompressible flows
in Autodesk Simulation CFD software(Autodesk Sustainability Workshop, n.d.).
Duan
CFD software is commonly used to analyze the aerodynamic properties of a design.
Wahba et al. (2012), Lav (2013), and Raveendran et al. (2009) in their simulations used the
commercial CFD solvers ANSYS Fluent and ANSYS CFX. The drag coefficient can be
determined from a CFD simulation by using the rearranged drag equation mentioned earlier.
Before a 3-D model can be analyzed, it must be meshed. Meshing is the process of discretizing
a model into numerous triangles or tetrahedrons so the RANS equations can be solved for each
one. Generally, a finer mesh gives more accurate solutions but requires more computing power;
a compromise must be made. A fully-detailed model can take an impractical and unnecessary
amount of computational time(Autodesk Sustainability Workshop, n.d.).
In a CFD simulation for external flow, the air can be treated as incompressible if speeds
are low, under 150 m/s(Talay, 1975; Raveendran, 2009). This assumption greatly simplifies the
equations the solver has to process. After the model is meshed, a virtual wind tunnel situation is
constructed by placing the object inside a box with an air inlet on one side and an outlet on the
other and setting up additional conditions. More complicated criteria can be set; Raveendran et
al. (2009) considered the movement of the road relative to the bus and the rotating tires in his
CFD analysis of a new bus design.
The two main components to aerodynamic drag on a vehicle are skin friction and
pressure drag, also known as form drag. Skin friction is caused by the air rubbing against the
surface of the vehicle; the molecules near the surface of the vehicle resist change in motion due
because air has viscosity, creating a force opposing the vehicles motion (Edgar, 2008). It is not
significant for heavy vehicles like trucks and busses, contributing <10% to the total drag force
(Patten, McAuliffe, Mayda, & Tanguay, 2012). Pressure drag accounts for the rest of the drag
force and is created because a moving vehicle creates an area of high pressure at the front an
area of low pressure in the back. This pressure difference pushes the vehicle back. Air flows
turbulently, recirculates, and forms vortices in the area of low pressure formed at the back of the
vehicle known as the wake.The wake is formed because attached air, air that smoothly flows
10
Duan
around a vehicle, separates near the end of the vehicle. The wake significantly contributes to
form drag because the low pressure pulls the vehicle backward (Edgar, 2008).
During vehicle design, the main consideration is function; a truck has a box-like shape in
order to maximize cargo space. The external appearance also dictates the general shape of the
vehicle because consumers will not buy a car that looks strange. Early attempts at reducing the
drag on cars failed because bad roads and low speeds were far larger problems and customers
disliked the new shapes (Hucho & Sovran 1993). Hucho and Sovran state that drag slowly was
reduced as car manufacturers changed the shape of their cars in response to changes in
consumer preferences and the oil crises in the 1970s. They point out that the two oil crises of
the 1970s generated great pressure for improving fuel economy drastically, and a rapid
decrease is seen in CD after the 1970s. There is a large discrepancy between the CD of a
modern car, about 0.26, and that of a bus, 0.6-0.8 (Drag Coefficient, n.d.). A large part of that
difference is due to the two performing different functions, but the high CD of a bus and sharp
edges that tend to increase CD implies that there is a large area for improvement. Global
warming has become a much larger issue, especially with the heavy reliance on buses for
transportation in developing countries, and fuel prices have risen since the 1990s. There is a
higher incentive now to improve the aerodynamics of heavy vehicles.
A redesign of an intercity Indian coach bus that significantly improved aerodynamic
performance was made by Raveendran, Sridhara, Rakesh, & Shankapal (2009). They used a
computer simulation to determine the drag force and CD of the chosen bus and their redesigned
one. The wants of bus users were surveyed and considered in creating the final design. Their
new design had a CD of 0.296 compared to 0.538 for the current one as well as a lower frontal
area, leading to a 60.39% decrease in drag force at 100kph. There were problems the proposed
design created such as less luggage space, less space for fuel tanks, and less seats available.
Nonetheless, the new design still resembled a bus and provided adequate space for the
passengers.
11
Duan
Because a redesign of a vehicle that does not compromise any part of its function, e.g.
the reduced cargo-storing capability of the new bus studied by Raveendran et. al, is difficult and
possibly difficult to adopt, external devices instead can be used to reduce drag. Although the
reduction in the drag coefficient may not be as large as that from a redesign, they can still result
in improvements to CD on the order of 20% or more; Schoon & Pan (2007) tested a model truck
with drag reduction devices in a wind tunnel and achieved a 23% reduction in CD.
Passive add-ons that have been considered in the past for trucks that could be applied
to a bus include: frontal and rear fairings/spoilers, boat-tail, side skirts, boat-tail like devices,
vortex generators, and lateral rear vanes. These devices are shown in figure 1. Vortex
generators are triangular shaped vanes that create turbulence through vortices that delay flow
separation and reduce the size of the wake (Lav, 2013). A frontal fairing,a piece of material that
can be attached to the front of a vehicle to smooth out airflow, could reduce the pressure and
some flow separation at the front of a bus that usually occurs due to the sharp perpendicular
edge. A rounding of the front corners of a model bus tested in a wind tunnel produced a 38%
reduction in CD (Lajos 2002); a frontal fairing operates on the same principle of smoothing out
the airflow to reduce the high pressure in the front. A rear spoiler is a device on the top end of a
vehicle changes the airflow in the wake to decrease drag and also increase downforce if desired
to give the vehicle more stability(Kim, 2003). Side skirts are usually sheets of metal or plastic
installed on the side of a tractor-trailer to prevent airflow into the underbody region. They
resulted in a 3-7% reduction in fuel consumption for trucks (Patten et al., 2012). Rear guide
vanes were proposed and tested through simulation by Wahba, Al-Marzooqi, Shaath, Shahin &
El-Dhmashawy(2012) and produced a reducing in CD of up to 18%. A boat tail is a device
installed on the rear of the vehicle intended to streamline the flow and decrease the size of the
wake(Patten, 2012). Many boat-tail like devices exist like rear panels that produce most of the
benefits of a full boat tail while being more practical.
12
Duan
For a tractor-trailer truck, the three regions that create the most drag, each contributing
around to the total CD, are the wake region in the back, the front, and the gap between the
tractor and the trailer(Wood & Bauer, 2003). The wheels do contribute about 0.1 to the CD, but
reducing the drag on them is more difficult than reducing the drag on the other areas. The 3
13
Duan
devices tested by Wood and Bauer(2003) targeted the gap region, the wake region, and
underbody. While the gap device is not applicable to a school bus, the other two concepts are.
Wood & Bauers device for reducing drag at the base was a vortex generator using 5 parallel
metal panels angled at around 45 degrees at the base of the trailer. The underbody device
located near the base changed the behavior of the airflow around it to stabilize the wake airflow.
The three devices used in conjunction produced a drag reduction of 30% and a fuel savings of
10%(Wood & Bauer, 2003). Their vortex generator design is shown in figure 2.
14
Duan
Duan
slanted side, width, and the height and length. Kim found the best design for the spoiler by
testing 24 different combinations of these parameters, blocking them and varying one at a time.
The optimal design was calculated to produce a 16.4% reduction in CD. Kim then had that
design built and tested it on a real bus, leading to a 12% reduction in CD and a increase in the
amount of downforce on the bus, which improves its maneuverability and stabilizes it, especially
at high speeds.
Through physical testing and CFD simulation, various drag reduction devices have been
successful in reducing the fuel consumption for tractor-trailer trucks. Not much research has
been done on the aerodynamics of school buses, which have similar shapes and high CDs like
tractor-trailer trucks. CFD has shown to be a viable way of testing the aerodynamic performance
of different designs quickly and by applying these various drag reduction to buses, a similar
improvement in CD could be made. The efficacy and practicality of such a device for a school
bus has not been investigated. A drag reduction device for a school bus would have to comply
with laws and be safe, durable, and inexpensive enough to be practical along with countless
other criteria. It must not obstruct the drivers view or the emergency exits and must be securely
attached to the bus. During the design process, many of these constraints must be taken into
consideration when evaluating possible designs.
Nowadays, with fossil fuels being a nonrenewable resource and the contribution they
have to global warming, small steps should be taken to reduce the collective carbon footprint
like reducing the fuel consumption of a bus with an external device. Technology exists to quickly
model a vehicle and simulate its aerodynamic properties; using it to design an add-on to a bus
would be one step towards stopping global warming. The box shape a school bus does not
have to unnecessarily add to its fuel consumption as an external add on could reduce its drag
without modifying its basic design and could be used on current school buses. Drag-reduction
devices are already being used on tractor-trailers and other heavy vehicles and are
economically viable for them; they could also be economically viable for school buses.
16
Duan
Hypotheses
1: All of the three proposed designs, rear spoiler, back plates, and frontal fairing will reduce the
total drag force and the Cd and the total Fd of the school bus.
2: At least one of the designs will be able to reduce Cd by 10% or more.
3: At least one of these devices, after optimization, will be economically viable to implement if
manufacturing costs could be brought down low enough.
Rationale: Previous research has managed to produce simple devices that reduce drag a
significant amount and many like side skirts, rear end plates, and vortex generators have been
adopted by commercial trucking companies.
17
Duan
Design Plan
18
Duan
1. Brainstorming
19
Duan
After looking back at the research and brainstorming what would be simple to implement but
also effective, the three preliminary designs chosen were a frontal fairing, a rear spoiler, and
back plates.
3. Construction of 3D Models
Autodesk Inventor was used to create the 3D models of the bus and the three designs. The 3D
models were simplified versions of the bus and omitted details while retaining the main
geometric shape. The size and position of the emergency exit was marked so neither the spoiler
nor the rear plate could obstruct it. All the units used were metric; the dimensions displayed are
in centimeters.
20
Duan
21
Duan
Rear Spoiler:
22
Duan
23
Duan
The back plates are metal plates that are angled inward. They reduce flow separation at the
back of the bus, decreasing the size of the wake. The two parameters they had were yaw angle
and length(cm), shown above. The original had a length of 1.5m and a yaw angle of 30 degrees.
To test the models in Simulation CFD, a box of air had to be made encapsulating the bus to act
as the virtual wind tunnel. This virtual box of air, the computational domain, was made to be 75
m long, 12 m wide, and 15 m high. This size was sufficiently large so that the boundaries of the
box did not interfere with the airflow much.
24
Duan
was created in the region surrounding the bus and the wake. The mesh is shown below.
5. Redesign
After testing the first iteration of the designs, the frontal fairing and back plates both reduced Cd
and were optimized. The rear spoiler increased Fd so no optimization or cost analysis was done
25
Duan
on it. For the frontal fairing, the alternate designs tested were a flat design and designs with
varying curvature.
For the back plates, the first round of optimization tested 3 different versions of the design:
standard back plates, a curved design, and the standard back plates with an extra plate on top.
The standard design was the best after considering its Fd reduction and its simplicity. The
second round of optimization changed the lengths of the back plate. 0.3 m was found to provide
the most reduction in Fd per unit volume.
The last round of optimization was changing the yaw angle while keeping the length 0.3 m. 30
degrees, the standard, was the most effective.
6. Analysis
A cost analysis was done on the frontal fairing and back plate design. The volumes of the
devices were calculated in Autodesk Inventor. Potential materials for the frontal fairing design
were chosen to be 0.5 cm thick polycarbonate with 30% glass fiber and 0.5cm thick
polypropylene with 30% glass fiber as they are both strong transparent plastics. Potential
materials for the back plates were identified as steel and aluminum.
Data on their costs were gathered from quandl.com, a commodities website, and a website
estimating the costs of custom parts(Cost Estimator, 2015). The data used to calculate average
diesel cost over last 5 years was from the U.S. Department of Energy(U.S. Gasoline and Diesel
Retail Prices,2015).
The measure of efficiency used was %DeltaCD/Volume, as this quantity shows how effective the
device is compared to its materials cost, which is proportional to volume. With the volumes of
the devices known, the materials cost was able to be calculated for each device.
26
Duan
At 60 mph, a 2% reduction of FD for a heavy truck will lead to a 1% reduction in fuel savings. At
40 mph and 20 mph, the necessary FD reduction for a 1% reduction in fuel is 3% and 6%(Wood
& Bauer, 2003). Assuming that the bus does both suburban routes and long distance routes, we
assume that each 1% reduction in FD will result in a 0.25% reduction in fuel consumption.
It must also be considered that the devices add weight to the school bus. It is assumed that the
school bus will weigh around 8500 kg with 24 students(1 to a seat). For every 1% increase in
weight, we assume it takes 0.5% more fuel to move the bus at the same velocity. From this
information, the equation used to calculate percent fuel saved is
1
1 w
Fd (
) is the weight of the device(kg)
4
2 8500
The negative is present in front of the %FD because a reduction in FD is a negative change.
The weight of the device is factored into the analysis. The payback period for the materials cost
was calculated and it ultimately determines whether or not the device is economically feasible. If
the materials take more than a few years to pay off for themselves, then the extra costs of
manufacturing, installation, transportation, etc. will make the device impractical.
27
Duan
Results
Images
(The back plates in the image are the 0.3m long, 30 degrees yaw angle design. The spoiler shown is the
preliminary design. The frontal fairing is the curved 175 arc radius design. These were the optimal
configurations)
Fairing:
Spoiler:
28
Duan
Back Plates:
Velocity Profiles:
Control:
29
Duan
Fairing:
Spoiler:
Back Plates:
30
Duan
31
Duan
32
Duan
0.167
Area of body:
5.784
Area spoiler:
1.267
Frontal Area(Body+2wheels)
6.118
Background:
School Buses
480000
1310
4860.10$
Weight(loaded)
8500 kg
Materials:
Density(kg/m^3)
Cost/kg($)
5.34
2.87
Aluminum
2712
1.87
Steel
7850
0.6
Control Bus:
Control Bus
Fd(N)
Cd
%Delta Cd
Fy
Cl
4059
1.223
0.000
-339.573
-0.102
Cl
Volume
Device(cm^3)
1.118 -8.572
-448.227
-0.135
60885.645
-140.791
3501
1.055 -13.737
245.000
0.074
65928.575
-208.356
Arc radius
150
3496
1.054 -13.860
169.634
0.051
64392.940
-215.238
Arc radius
3473
1.047 -14.427
-41.000
-0.012
63581.838
-226.897
Fd(N)
Cd
Straight
3711
Arc radius
125
%Delta
Cd
%DeltaCd/Volum
e(m^3)
33
Duan
175
Arc radius
250
3516
1.060 -13.374
119.300
0.036
62472.902
-214.083
Cd
%Delta
Cd
Fy
50142.870
-229.316
Cl
Volume
Device(cm^3) %DeltaCd/Volume
1.083
-11.499
339.643 -0.102
1.097
-10.352
64.755
0.020
73307.250
-141.212
1.046
-14.480
277.785 0.084
61096.190
-236.999
Cl
Volume
Device(cm^3 %DeltaCd/Volu
)
me
Fd(N)
Cd
%Delta
Cd
1.5m
3592
1.083
-11.499
-339.643
-0.102
50142.870
-229.316
1.2m
3776
1.138
-6.971
-238.790
-0.072
40114.300
-173.767
0.9m
3801
1.146
-6.352
-295.390
-0.089
30085.722
-211.133
0.6m
3796
1.144
-6.478
-143.180
-0.043
20057.148
-322.965
0.3m
3873
1.167
-4.583
-371.870
-0.112
10028.574
-456.992
34
Duan
Cl
Volume
Device(cm^3 %DeltaCd/Volu
)
me
Fd(N)
Cd
%Delta
Cd
10 degrees
4059
1.223
0.002
-325.940
-0.098
10028.574
0.197
20 degrees
3900
1.175
-3.918
-222.540
-0.067
10028.574
-390.654
30 degrees
3873
1.167
-4.583
-371.870
-0.112
10028.574
-456.992
40 degrees
3921
1.182
-3.400
-285.270
-0.086
10028.574
-339.058
35
Duan
Rear Spoiler:
Fd(N) Cd
%Delta Cd Fy
28.050
Cl
Volume
Device(cm^3)
%DeltaCd/Volu
me
NA
Comparison of Designs:
36
Duan
Weight(kg)(Design 1)
92.082
27.197
Material Cost(1) $
491.717
50.859
Weight(kg)(Design 2)
74.696
39.362
Material Cost(2) $
214.377
23.617
Annual Fuel
Consumption(Gallons)
1310
1310
4860.1
4860.1
-2.923
-0.986
-3.026
-0.914
Fuel Savings(material 1)
142.075
47.909
Fuel Savings(material 2)
147.045
44.431
1.062
37
Duan
Cost)(Years)
Payback Period(Materials 2
Cost)(Years)
1.458
0.532
Practical? Yes/No
Yes
Yes
Data Analysis
The control school bus model had a higher CD than theoretically expected; 1.223 compared to a
theoretical 0.6-0.8. This could be due to two reasons: the model made in Autodesk Simulation
CFD was fairly coarse to reduce the mesh complexity and its sharp edges caused more flow
separation and higher surface pressures than the smoother ones found on a real bus. Secondly,
CFD comes with inherent error that was made worse with the relatively low mesh
resolutions(~100-150k elements) due to limited computing power. Despite the error, it is still
viable to make comparisons between the control and the three designs.
The rear spoiler, even though it decreased the size of the wake the most and did increase its
pressure, added an extra 1.267m^2 to the cross sectional area and increased the overall drag
force. The images reveal that the spoiler has reduced the size of the wake as well as reduce
recirculation in the wake. A cost analysis was not done for it because the design did not reduce
the overall drag force.
38
Duan
All the frontal fairing designs reduced Fd by a considerable amount. The image of frontal
pressure shows the fairing reducing the pressure on the front face of the bus. The worst
performing, as expected, was the flat design. All the other designs were fairly similar in their
%DeltaCd/Volume scores, showing that they were fairly close in efficiency. The most efficient
design with a 175 cm arc radius reduced Fd and Cd by 14.427% and had a materials payback
period of 3.461 years if made with 0.5cm polycarbonate and 1.458 years if made with 0.5cm
polypropylene, both reinforced with 30% glass fiber.
The back plates were all fairly effective in reducing Fd and Cd. The images of back pressure
show that the back plates, although relatively small, had a large effect on the wake region. From
the first optimization, the efficiency, measured through the quantity -%DeltaCd/Volume,
constantly increased as length decreases. 0.3m was the smallest plate size for practical
reasons. In the 2nd round of optimization, the most efficient yaw angle was 30 degrees, with the
efficiency decreasing as angle increased or decreased. The back plate design was also
successful, with a payback period of 1.062 years for the aluminum design and 0.532 years for
the steel.
All of the designs reduced the amount of recirculation in the wake compared to the control. The
frontal fairing also led to less flow separation on the front of the bus. Comparing the velocity
profiles, the wakes are smaller for each of the three proposed designs with the spoiler having
the largest effect on wake size. Although the overall size seems similar, the region of lowest
velocity is smaller when devices were used.
Possible Improvements:
39
Duan
To reduce simulation error and achieve more accurate results, a more detailed model of the bus
could be created that has more features of a real school bus like the side mirrors and the minor
details. The mesh could also be made to have more elements and more refinement could be
added to the region around the bus to get a more detailed understanding of the airflow.
Autodesk Simulation CFD also has the ability to add moving parts; the wheels could be made to
rotate but this comes at the cost of using more computing power. All these improvements would
require a faster computer to run the simulations on.
Conclusion
The purpose of this project was to design, evaluate, and optimize 3 different drag reduction
devices for a school bus. Out of the three hypotheses, two were met; the spoiler did not
decrease FD.
The frontal fairing and back plates both would be economically viable to produce if the
manufacturing, transportation, and other extra costs could be kept low by mass production, but
the frontal fairing design would have many flaws in the real world looking back at the criteria
from earlier. The transparent material would cover the windshield; if the windshield was
damaged it would be inconvenient to fix. Also, the fairing would have to be safety tested and get
approval, a difficult process. The fairings most fatal flaw is when it gets dirty or when the
weather is bad. It would be very unsafe if rain fell on the fairing as there are no wipers on the
fairing. Although the fairing is very effective at reducing FD, it would not be practical.
40
Duan
The back plates on the other hand present a promising opportunity. The fuel savings would pay
off for the material cost of the steel version in only half a year. Due to the relative ease of
manufacturing, installing, and transporting steel plates, these could be very inexpensive. The
design meets all of the criteria mentioned: it does not pose any safety threat or obstruction to
the emergency exit, it is cheap, it is effective, and it adds less than a foot to the buss length.
The 0 .6 m design, although not as efficient, would still pay off its materials cost in less than a
year if made from steel and could be considered as a possible alternative to the 0.3m design.
There are drawbacks to this analysis. Depending on the type of school bus and the routes it
travels, a drag reduction device could never be practical. In urban areas where school busses
stop very frequently and rarely drive at high speeds, stopping and re-accelerating consumes
most of the buss fuel.
The U.S. depends on school buses to transport half the nations school-age children. Even with
a modest 1-2% reduction in fuel consumption, the potential money saved and potential fossil
fuels prevented from being burned would be great if most current school buses were outfitted
with the back plates device. Already being implemented by trucking companies, having back
plates on current school buses would be a small step in the right direction towards slowing down
the pollution of our planet.
41
Duan
References
Autodesk Sustainability Workshop. (n.d.). Retrieved November 3, 2014, from
http://sustainabilityworkshop.autodesk.com/
Cost Estimator. (n.d.). Retrieved March 2, 2015, from
<http://www.custompartnet.com/estimate/injection-molding/>
"Drag Coefficient." Drag Coefficient. The Engineering Toolbox. Web. Retrieved 1 Nov. 2014,
from http://www.engineeringtoolbox.com/drag-coefficient-d_627.html.
Edgar, J. (2008, April 29). Analysing Road Car Drag. Retrieved October 31, 2014, from
http://www.autospeed.com/cms/article.html?&A=110351
Elert, G. (n.d.). Aerodynamic Drag. Retrieved October 30, 2014, from
<http://physics.info/drag/>
Find and Use Data. Easily. (n.d.). Retrieved March 2, 2015, from
<https://www.quandl.com/>
Hucho, W., and G. Sovran. "Aerodynamics of Road Vehicles." Annual Review of Fluid
Mechanics, Vol. 25 (1993): 485-537. Print.
Kim, M., "Numerical Study on the Wake Flow and Rear-Spoiler Effect of a Commercial Bus
Body," SAE Technical Paper 2003-01-1253, 2003, doi:10.4271/2003-01-1253.
42
Duan
Lajos, Tamas. "Basics of Vehicle Aerodynamics." Budapest University of Technology and
Economics. 1 Jan. 2002. Web. 3 Nov. 2014. Retrieved from
http://www.ara.bme.hu/oktatas/letolt/Vehicleaerodyn/Vehicleaerodyn.pdf
Laughlin, M. (2004, April 1). Analysis of U.S. School Bus Populations and Alternative Fuel
Potential. Retrieved November 3, 2014, from
http://www.nrel.gov/docs/fy04osti/35765.pdf
Lav, C., "Three Dimensional CFD Analysis on Aerodynamic Drag Reduction of a Bluff Tractor
Trailer Body using Vortex Generators," SAE Technical Paper 2013-01-2458, 2013,
doi:10.4271/2013-01-2458.
Patten, J., McAuliffe, B., Mayda, W., & Tanguay, B. National Research Council Canada, Centre
for Surface Transportation Technology. (2012). Review of aerodynamic drag reduction
devices for heavy trucks and buses. From:
https://www.tc.gc.ca/media/documents/programs/AERODYNAMICS_REPORTMAY_2012.pdf
Raveendran, A., Sridhara, S., Rakesh, D., and Shankapal, S., "Exterior Styling of an Intercity
Transport Bus for Improved Aerodynamic Performance," SAE Technical Paper 2009-280060, 2009, doi:10.4271/2009-28-0060.
Reynolds-Averaged Navier-Stokes Equations. (2009, June 9). Retrieved November 3, 2014,
from http://www.symscape.com/reynolds-averaged-navier-stokes-equations
School Bus Safety Data. (2009, October 1). Retrieved December 10, 2014, from
http://stnonline.com/index.php/school-bus-safety-data
Schoon, R. and Pan, F., "Practical Devices for Heavy Truck Aerodynamic Drag Reduction," SAE
Technical Paper 2007-01-1781, 2007, doi:10.4271/2007-01-1781.
Shape Effects on Drag. n.d. Retrieved October 31, 2014, from http://www.grc.nasa.gov/WWW/k12/airplane/shaped.html
Talay, T. (1975). III. FLUID FLOW. In Introduction to the aerodynamics of flight. Washington:
NASA.
43
Duan
The tyre: Rolling resistance and fuel savings. (2003). Clermont-Ferrand: Societe de technologie
Michelin.
U.S. Gasoline and Diesel Retail Prices. (n.d.). Retrieved March 2, 2015, from
http://www.eia.gov/dnav/pet/pet_pri_gnd_dcus_nus_a.htm
Wahba, E., Al-Marzooqi,, H., Shaath, M., Shahin, M., & El-Dhmashawy, T. (2012). Aerodynamic
Drag Reduction for Ground Vehicles using Lateral Guide Vanes. CFD Letters, 4(2), 6879. Retrieved October 4, 2014, from
http://www.issres.net/journal/index.php/cfdl/article/viewFile/S21801363(12)4268-X/154
Wood, R. and Bauer, S., "Simple and Low-Cost Aerodynamic Drag Reduction Devices for
Tractor-Trailer Trucks," SAE Technical Paper 2003-01-3377, 2003, doi:10.4271/2003-013377.
44