High performance Turbo Exhaust Manifold, Elmer Gutierrez, Michael

Transcription

EML 4905 Senior Design Project
A B.S. THESIS
PREPARED IN PARTIAL FULFILLMENT OF THE
REQUIREMENT FOR THE DEGREE OF
BACHELOR OF SCIENCE
IN
MECHANICAL ENGINEERING
HIGH PERFORMANCE TURBO
EXHAUST MANIFOLD
Final Report
Elmer Gutierrez
Michael Martinez
Rogelio Muro
Advisor: Professor Andres Tremante
November 16, 2011
This B.S. thesis is written in partial fulfillment of the requirements in EML 4905.
The contents represent the opinion of the authors and not the Department of
Mechanical and Materials Engineering.
P a g e | ii
Ethics Statement and Signatures
The work submitted in this B.S. thesis is solely prepared by a team consisting of ELMER
GUTIERREZ, MICHAEL MARTINEZ, and ROGELIO MURO and it is original. Excerpts from
others’ work have been clearly identified, their work acknowledged within the text and listed in
the list of references. All of the engineering drawings, computer programs, formulations, design
work, prototype development and testing reported in this document are also original and
prepared by the same team of students.
Elmer Gutierrez
Michael Martinez
Dr. Andres Tremante
Faculty Advisor
Rogelio Muro
P a g e | iii
Table of Contents
Abstract ........................................................................................................................................... 1
Chapter 1 Introduction .................................................................................................................... 2
1.1 Problem Statement .................................................................................................................... 2
1.2 Motivation ................................................................................................................................. 3
1.3 Literature Survey ...................................................................................................................... 6
1.4 Discussion ............................................................................................................................... 13
Chapter 2
Project Formulation ................................................................................................... 14
2.1 Overview ................................................................................................................................. 14
2.2 Project Objectives ................................................................................................................... 15
2.3 Constraints and Other Considerations .................................................................................... 15
Chapter 3
Design Alternatives ................................................................................................... 16
3.1 Overview of Conceptual Designs Developed ......................................................................... 16
3.2 Design Alternative 1 ............................................................................................................... 17
3.6 Feasibility Assessment ............................................................................................................ 23
3.7 Proposed Design ..................................................................................................................... 23
Chapter 4 Project Management ..................................................................................................... 32
4.1 Overview ................................................................................................................................. 32
4.2 Specific Tasks Comprising the Project ................................................................................... 32
4.3 Organization of Work and Timeline ....................................................................................... 33
4.4 Breakdown of Responsibilities among Team Members ......................................................... 34
4.5 Patent/Copyright Application ................................................................................................. 35
4.6 Commercialization of the Final Product ................................................................................. 35
Chapter 5 Engineering Design and Analysis ................................................................................ 36
5.1 Design Based on Static and Fatigue Failure Design Theories: ............................................... 36
5.2 Material selection .................................................................................................................... 36
5.3 Deflection Analysis ................................................................................................................. 49
P a g e | iv
Chapter 6 Prototype Construction ................................................................................................ 54
6.1 Description of Prototype ......................................................................................................... 54
6.2 Prototype Design ..................................................................................................................... 54
6.3 Parts List ................................................................................................................................. 55
6.4 Construction ............................................................................................................................ 55
6.5 Prototype Cost Analysis.......................................................................................................... 68
6.6 Discussion ............................................................................................................................... 70
Chapter 7 Testing and Evaluating ................................................................................................. 72
7.1 Overview ................................................................................................................................. 72
7.2 Design of Experiments - Description of Experiments ............................................................ 72
7.3 Test Results and Data............................................................................................................ 111
7.4 Improvement of design ......................................................................................................... 129
7.5 Discussion ............................................................................................................................. 130
Chapter 8. Design Considerations............................................................................................... 131
8.1 Sustainability Report ............................................................................................................. 131
Chapter 9. Conclusion ................................................................................................................. 135
9.1 Conclusion and Discussion ................................................................................................... 135
9.2 Commercialization of the Manifold ...................................................................................... 137
9.3 Future Work .......................................................................................................................... 140
Appendix A, Engineering drawing sheets .................................................................................. 143
Appendix B, Receipts/Invoices ................................................................................................... 145
Appendix C, Pictures .................................................................................................................. 151
Page |v
Table of Figures
Figure 1 Mazda Miata ..................................................................................................................... 3
Figure 2 After Market Mazda Miata 1.8L Turbo Manifold (Reference 3, OBX Racing) .............. 7
Figure 3 1.8L Turbo Manifold (Reference 4, OBX Racing) .......................................................... 8
Figure 4 Thermal Efficiency vs. Back Pressure(Smith 1) ............................................................ 10
Figure 5 Exhaust Manifold Miata 1.8L......................................................................................... 17
Figure 8 Ram horn manifold configuration .................................................................................. 20
Figure 10 Exhaust Manifold Miata 1.8L....................................................................................... 22
Figure 11 Flow Analysis distribution of the proposed design ...................................................... 24
Figure 12 CFD Analysis of Alternative Design ............................................................................ 25
Figure 13 Pressure Analysis in Solid Works ................................................................................ 26
Figure 14 Velocity Distribution Analysis ..................................................................................... 26
Figure 15 Exhaust Manifold velocity flow distribution ................................................................ 29
Figure 16 Exhaust Manifold velocity flow distribution without collector ................................... 30
Figure 17 . Exhaust Manifold velocity flow distribution .............................................................. 31
Figure 18 Time Frame Senior Design Project .............................................................................. 34
Figure 19 Stress Simulation Analysis using AISI 304................................................................. 40
Figure 20 Displacement Simulation Analysis Using AISI 304 stainless steel ............................ 41
Figure 21 Strain Simulation of 304 stainless steel ....................................................................... 41
Figure 22 Factor of safety of 304 stainless steel .......................................................................... 42
Figure 23 Analysis of maximum load supported using 304 stainless steel ................................. 43
Figure 24 Stress Simulation Analysis using 316 stainless steel ................................................. 43
Figure 25 displacement of 316l stainless steel ............................................................................. 44
Figure 26 Strain Distribution of 316l stainless steel .................................................................... 44
Figure 27 Factor of safety 316l stainless steel ............................................................................. 45
Figure 28 Analysis of maximum load of 316l stainless steel ...................................................... 46
Figure 29 Stress Simulation Analysis using Gray Cast Iron ........................................................ 46
Figure 30 Displacement of Stress Simulation Analysis using Gray Cast Iron ............................ 47
Figure 31 Strain Gray Cast Iron ................................................................................................... 48
Figure 32 Factor of safety Gray Cast Iron ................................................................................... 48
Figure 33 Analysis of Maximum Load Supported using Gray Cast Iron .................................... 49
Figure 34 Deflection Analysis ..................................................................................................... 49
Figure 35 Crack prone Manifold.................................................................................................. 50
Figure 36 Crack........................................................................................................................... 51
Figure 37 Deflection Analysis ..................................................................................................... 53
Figure 38 Knee mill working on first hole, cylinder 2 ................................................................ 56
Figure 39 Knee mill working on first hole, cylinder 2 ................................................................ 56
P a g e | vi
Figure 40 Manifold flange in the process of being cut by using MasterCam software ............... 57
Figure 41 Manifold flange completed successfully by using MasterCam software .................... 58
Figure 42 Center section cuts ....................................................................................................... 59
Figure 43 Center section cuts assembled ..................................................................................... 59
Figure 44 Cut of the 90 degrees elbow in process ....................................................................... 60
Figure 45 Spot weld process ........................................................................................................ 61
Figure 46 Spot Welding MIG ....................................................................................................... 61
Figure 47 Partial assembly view .................................................................................................. 62
Figure 48 Partial assembly with tubes connecting cylinders spot welded ................................... 63
Figure 49 Partial assembly with collector base leveled ............................................................... 64
Figure 50 Team #1 working on middle part of the turbocharger flange ...................................... 65
Figure 51 Partial assembly the turbocharger flange .................................................................... 65
Figure 52 Manifold assembly right after welding process was completed .................................. 67
Figure 53 Finish and polishing process of manifold assembly .................................................... 68
Figure 54 Assembly with the welding. ........................................................................................ 71
Figure 55 Part Number of the Flow Machine ............................................................................... 73
Figure 56 Outlet of the Duct ......................................................................................................... 73
Figure 57 Control Board ............................................................................................................... 74
Figure 58 Installing Nozzle ........................................................................................................... 74
Figure 59 First Nozzle .................................................................................................................. 75
Figure 60 Third Nozzle Installed .................................................................................................. 75
Figure 61 Box install already in nozzle ........................................................................................ 76
Figure 62 Box preparation for manifold mounting. ...................................................................... 76
Figure 63 Box preparation for manifold mounting. ...................................................................... 77
Figure 64 Manifold already mounted on box ready for flow test. ................................................ 77
Figure 65 Installation of the pitot tube into based ........................................................................ 78
Figure 66 Setting up the engineering sheet ................................................................................... 78
Figure 67 Points on the x and y directions .................................................................................... 79
Figure 68 Testing Measuring ........................................................................................................ 80
Figure 69 Data recording during testing using Excel. .................................................................. 80
Figure 70 Preparing for individual testing .................................................................................... 81
Figure 71 Testing flow in individual pipes ................................................................................... 82
Figure 72 Stopwatch used to measure the time step ..................................................................... 82
Figure 73 Baker Aviation School Logo ........................................................................................ 83
Figure 74 Material used to create gasket for pressurization ......................................................... 83
Figure 75 Manifold with bottom gasket........................................................................................ 84
Figure 76 Testing Table ................................................................................................................ 84
Figure 77 Manifold fixed in place for testing. .............................................................................. 85
Figure 78 Manifold with spot welds ............................................................................................. 86
Figure 79 The sections wear the leak is located is been marked with the tool shown.................. 87
P a g e | vii
Figure 80 Example of leakage during test .................................................................................... 87
Figure 81 Preparations for third test ............................................................................................. 88
Figure 82 Manifold with Air Pressure .......................................................................................... 88
Figure 83 Air Pressure Gage ......................................................................................................... 89
Figure 84 Water with soap ready to detect any leaks ................................................................... 89
Figure 85 Manifold Polished after the testing was done............................................................... 90
Figure 86 Manifold painted .......................................................................................................... 90
Figure 87 The three components of the Leak Test ....................................................................... 91
Figure 88 Rear view of the three chemicals .................................................................................. 91
Figure 89 Manifold Clean ............................................................................................................. 92
Figure 90 First Application ........................................................................................................... 93
Figure 91 Applying Penetrating Agent ......................................................................................... 93
Figure 92 Manifold Entirely Sprayed ........................................................................................... 94
Figure 93 Excess Removed ........................................................................................................... 94
Figure 94 Developer just applied .................................................................................................. 95
Figure 95 Developer Two Minutes after ....................................................................................... 95
Figure 96 Developer Three Minutes After.................................................................................... 96
Figure 97 Developer SKD-2 fully developed. .............................................................................. 96
Figure 98 Stock Manifold ............................................................................................................. 98
Figure 99 Prototype Installed ........................................................................................................ 98
Figure 100 Comparison of body temperature and starting test temperature as seen by the camera.
....................................................................................................................................................... 99
Figure 101 Using laser thermometer as soon as the engine was started at the first pipe to the left.
....................................................................................................................................................... 99
Figure 102 Corresponding image to support figure 101. Is notable the temperature path through
the manifold. ............................................................................................................................... 100
Figure 103 Reading the laser thermometer at the second pipe and one of the middle pair. ....... 100
Figure 104 Thermal Imaging showing average temperature through the whole picture ............ 101
Figure 105 Reading to the right interior pipe and third of the four pipes in the prototype......... 102
Figure 106 Thermal Imaging showing the heat shield and temperature pattern ......................... 102
Figure 107 Reading at the fourth pipe with the laser thermometer. ........................................... 103
Figure 108 Corresponding image to support figure 107 ............................................................. 103
Figure 109 Reading at the collector center with laser................................................................. 104
Figure 110 Highest reading with thermometer ........................................................................... 104
Figure 111 Corresponding image to support figure 110 ............................................................. 105
Figure 112 2000RPM .................................................................................................................. 105
Figure 113 Physical aspect of the manifold at 2000 revolutions per minutes. ........................... 106
Figure 114 Manifold at 2000RPM ~368F .................................................................................. 106
Figure 115 3000RPM .................................................................................................................. 107
Figure 116 Physical aspect of the manifold at 3000 revolutions per minutes. ........................... 107
P a g e | viii
Figure 117 Cooling Off ............................................................................................................... 108
Figure 118 Temperature Pattern ................................................................................................. 108
Figure 119 Testing the Stock Manifold ...................................................................................... 109
Figure 120 Header warms up a lot faster .................................................................................... 109
Figure 121 Stock Manifold Heat Signature ................................................................................ 110
Figure 122 Shows a few seconds after getting to 3000RPM, the manifold reaches 655F,finally
reaching 706F after just a few seconds. ...................................................................................... 110
Figure 123 Data obtained from flow test at center point (PT 0) ................................................. 112
Figure 124 Data obtained from flow test at point one left (PT 1L) ........................................... 112
Figure 125 Data obtained from flow test at point two left (PT 2L) ............................................ 113
Figure 126 Data obtained from flow test at point three left (PT 3L) .......................................... 113
Figure 127 Data obtained from flow test at point one right (PT 1R) .......................................... 114
Figure 128 Data Obtained from Flow Test at point two right (PT 2R) ...................................... 114
Figure 129 Data obtained from flow test at point three right (PT 3R) ....................................... 115
Figure 130 Data obtained from flow test at center point (PT 0) ................................................. 116
Figure 131 Data obtained from flow test at point one up (PT 1U) ............................................. 116
Figure 132 Data obtained from flow test at point two up (PT 2U) ............................................. 117
Figure 133 Data obtained from flow test at point one down (PT 1D) ........................................ 117
Figure 134 Data obtained from flow test at point two down (PT 2D) ........................................ 118
Figure 135 Data obtained from flow test at each individual header ........................................... 119
Figure 136 Uncertainty Analysis ................................................................................................ 122
Figure 137 Pressure 1.................................................................................................................. 123
Figure 138 Several Points at Pressure 1 ...................................................................................... 123
Figure 139 Points at Pressure 2 ................................................................................................... 124
Figure 140 Pressure 2.................................................................................................................. 124
Figure 141 Pressure 3.................................................................................................................. 125
Figure 143 Pressure 4.................................................................................................................. 126
Figure 145 Pressure 5.................................................................................................................. 127
Figure 147 Location of Production and Distribution .................................................................. 131
Figure 148 Definition of Use of Design ..................................................................................... 131
Figure 149 Life Cycle of a Product............................................................................................. 132
Figure 150 Pie Charts and Environmental Impact ...................................................................... 133
Figure 151 MiataTurbo.net ......................................................................................................... 138
Figure 152 Vendor Forum .......................................................................................................... 139
Figure 153 Usage Info ................................................................................................................ 140
Figure 154 Front Flange.............................................................................................................. 143
Figure 155 Back flange ............................................................................................................... 144
P a g e | ix
Figure 156 Invoice of purchase made at P&M Supply Inc. ....................................................... 145
Figure 157 Invoice of purchase made at P&M Supply Inc. ....................................................... 146
Figure 158 Invoice ...................................................................................................................... 147
Figure 159 Invoice ...................................................................................................................... 148
Figure 160 Invoice ...................................................................................................................... 149
Figure 161 Invoice ...................................................................................................................... 150
Figure 162 Compressor Used...................................................................................................... 151
List of Tables
Table 1 AISI Type 316L Stainless Steel Properties...................................................................... 11
Table 2 Gray Cast Iron Properties ................................................................................................ 12
Table 3 Different angular configurations for inner and outer pipes. ............................................ 27
Table 4 Average velocity and Maximum velocity for multiple angular configurations............... 28
Table 5 Timeline Data for Senior Design Project Completion ..................................................... 33
Table 6 AISI 316L Stainless Steel ................................................................................................ 38
Table 7 AISI 304L Stainless Steel ................................................................................................ 38
Table 8 Gray Cast Iron .................................................................................................................. 39
Table 9 Detailed list of the parts, cost, and supplier information ................................................. 69
Table 10 Values of Student-t distribution for finite sample sizes ............................................... 120
Table 11 Real life test results for inletTP equal to 124.4 Pa at point 0 ...................................... 122
Table 12 Real life test results for inlet TP equal to 186.6 Pa at point 0 ..................................... 124
Table 13 Real life test results for inlet TP equal to 248.8 Pa at point 0 ..................................... 125
Table 14 Real life test results for inlet TP equal to 311 Pa at point 0 ........................................ 126
Table 15 Real life test results for TP equal to 373.2 Pa at point 0.............................................. 127
Table 16 Error percentage analysis ............................................................................................. 129
Page |1
Abstract
Technology has always been used to achieve better performance in any mechanical system.
In the world of automotive engineering, high performance ideas and designs are essential
especially in the advent of a new generation of vehicles coming in the future. Turbocharged
engines have proved in the past to provide better fuel economy and more power compare to the
naturally aspirated counterparts. Our senior design project is to design a performance exhaust
manifold for the Mazda Miata from 1994-2005; this vehicle has never being equipped with a
turbocharger from factory in its standard form. The goal is to make a new high performance
turbo header designed to increase the power and efficiency out of the 1.8 liter engine at an
affordable price. The three design alternatives are based on small restriction which creates less
friction and a cooler working manifold for better performance and longer life. This exhaust
manifold is equipped with a T3 turbo charger base plate. This product will help to increase
horsepower for racing purposes and can be used alternatively to manage fuel economy for
average performance. The design will face some constrains due to the space that this vehicle
provide from factory for the simple reason that the chassis is not designed for a T3 turbo charger
around the exhaust area. Providing a higher flow and a colder system from the exhaust valve
cavity to the turbo impeller it is certain a higher efficiency system and less destructive in a matter
of parts conservation.
Page |2
Chapter 1 Introduction
The 1994-2005 Mazda Miata is a car model equipped with a 1.8 liter naturally aspirated
engine. Bell Engineering or BEGI is a company that’s builds turbocharger systems as well as
pistons, turbo, intake coolers, piping systems, wastegate and some other high performance parts.
The cost in the market for a similar exhaust header created by BEGI is around 800 US dollars for
the stainless steel and 530 dollars for the cast iron counterpart; this exhaust header may include a
wastegate port. Our project is based on a design and the study of a revolutionary turbo manifold
that will surpass the only company producing this exhausts in price, quality and performance, the
studies that will be accomplished to achieve these goals are based on CFD analysis, simulations
studies of structure, thermal analysis of the fluid and the metal involved. Also, manufacturing of
the prototype is involved in the progress of this project; there will be construction and physical
testing of the design.
1.1 Problem Statement
Race and track Mazda Miata need turbo chargers for competition purposes, where the
conditions are extreme and cry for a solid design and structure of a turbo manifold. The reason
why there is a need for a redesigned turbo charger manifold is because the existing ones are
cracking up, breaking, overheating, and melting, some with just a few hours of use ; but, most of
them fail to get to the consumers hands due to the higher cost of the quality built manifolds like
the models from BEGI. The other concern besides the durability or the resistance of this
manifold is the flow achieved of these often rough designs. For example, Racing Beat Inc. does
not used CFD analysis or any type of computer simulation to create efficient exhausts manifold;
Page |3
this company does trial and error until they hit a specific target. There is nothing wrong with this
approach, but the process is too expensive since it creates a lot of waste material and the final
result is not exactly accurate as the best possible alternative. Imagine if every manifold designed
in the industry had to made as a prototype first and then tested, just to gain an idea on actual
performance, the waste of material, money and energy that it would take.
1.2 Motivation
For 20 years in a row, the Mazda Miata has been the world bestselling sports car. The Miata
is not only a very agile car, but also a great looking sports car that takes handling to a whole
different dimension and makes other cars, more expensive cars, look ridiculous around the track
and around the slalom. It was an obvious choice as we have had experience working first hand
with the car. The team admits the Miata was a very easy choice to make since the car has a few
advantages and special design considerations that are unique to it.
Figure 1 Mazda Miata
Page |4
A very unique set of mechanical characteristics of the Miata are:
1. Rear Wheel Drive configuration: ideal for sports car and track cars, needed to obtain
balance and proper handling in the track.
2. Inline 4 Engine: A four cylinder engine that has the proper factory design, a closed deck
block and a long stroke characteristic of high torque engine. A generous redline of 7500
rpm and an oil pump capable of withstanding continued abuse.
3. The Miata engine is derived from the BPT engine used in previous generation Mazda
cars, which was already a turbocharged engine, so the BP was in theory and design ready
for a turbocharger.
4. The engine bay has enough space to accommodate not only the turbo but also a different
range of manifolds, making either top mounted; side mounted or lower mounted
configurations possible.
5. A very strong powertrain especially in the higher trims of the Miata. The more expensive
LS Miata came with a very strong RX7 derived 6 speed transmission that although fitted
with very closed ratio gears, important for high torque but not so good for turbocharger
spool, works perfectly in conjunction with a turbocharged Miata.
The question was, where had Mazda cut corners in the Miata?
Most owners will agree that the Miata is a car that not only deserves but needs 250 hp, it
comes natural to the car. The chassis feels right and the power delivery of the turbocharged car
mimics what the natural aspirated engine power band provides. Since power was the factor
where the Miata fell short, it was natural for the team to concentrate in ways of making the car
more powerful with the slightest amount of money. Turbocharging came as a definitive answer
Page |5
since it provides the biggest power increase while remaining safe for the stock engine, fuel
efficiency is not sacrificed, and the car exhibits behavior that could be confused with the stock
car. The only difference is that now torque comes a lot stronger and faster.
After a small market research and coming from the experience of the team, it was logical to
start our quest for power from the manifold. The current manifold design suffers from a few fatal
design problems and constraints.

Mass produced manifolds are too cheaply made and this results in cracks and yellowing
of the manifold. Cracks in the exhaust can lead to a few disastrous problems, most
importantly, to the turbo not spooling at the correct time and, therefore, the engine runs
rich in gasoline. This in turn affects performance, efficiency, and the overall running of
the car. Materials used on these manifolds are usually made out of stainless steel T304.
T304 is a great choice for a manifold. However, the issue with most Chinese and a nonlegit manifold is the use of extremely thin stainless steel. Being thin reduces cost but
makes cracking under cyclic loading inevitable.

Brand manifolds are too expensive because the companies that make them are reputable
companies that invest a lot of money in materials and manufacturing, but these manifolds
are very expensive and out of the reach of many enthusiasts.

Hobby or homemade manifolds are usually very well made material wise, but the makers
lack the understanding of the fluid mechanics of a well-designed manifold. Therefore,
they are not as efficient as they could be. They tend to be expensive in so far as the parts
being used.
Page |6
Therefore, our objective is the creation of a manifold that can be cheaply made, yet is strong
enough to withstand any abuse in daily driving or in the track and that would provide horsepower
gains compared to others as well.
1.3 Literature Survey
An exhaust manifold is a device that has the function of transmitting the exhaust gases from
multiple cylinders through tubes into a collector. Exhaust manifolds come in different shapes and
sizes, and in most cases are made of either stainless steel or cast iron. Several companies
specialize in designing and manufacturing exhaust manifolds for the Miata in today’s market.
Companies such as: Flying Miata and BEGI bring to the market multiple configurations and
styles of manifolds every year.
History
The design of exhaust manifolds has been developed for many years. As engines have been
improved so have exhaust manifolds. Early exhaust manifolds were much less efficient than
today’s exhaust manifolds. Designs that did not contribute to a smooth flow were very typical in
early times. The back pressure built up was much greater, this increased the work done by the
particular piston at the exhaust stroke. Large amounts of residual gases remain in the
compression chamber and, as a consequence, the temperature increases. Sometimes when
working hard the manifold glows red-hot. They even used asbestos, a highly heat-resistant
fibrous silicate mineral, to protect paintwork. Nowadays, exhaust manifold designs have been
transformed completely. In order to improve earlier configurations, designers have come up with
different designs that decrease the flow resistance by using a much improved pipe layout and
Page |7
increasing the average exhaust velocity of the gases, which improves the power output. The
collector has also been modified throughout the years through the use of different geometries for
the purpose of obtaining a better flow efficiency.
Manifold Configurations
The designs of exhaust manifolds for the Miata 1.8 L vary from very simple designs to
designs with higher level of complexity. Several manifolds are designed in such a way that the
exhaust head pipes run individually and converge at the collector. There are other designs in
which the pipes intersect, and only two and sometimes three pipes converge at the collector. The
head pipes can have multiple lengths, diameters, and shapes making each design unique since
these factors have a direct effect in the volumetric flow efficiency of the manifold and power
output.
Figure 2 After Market Mazda Miata 1.8L Turbo Manifold (Reference 3, OBX Racing)
Page |8
Figure 2 shows a Mazda Miata 1.8L Turbo Manifold manufactured by OBX Racing. The
design uses tubular pipes and combines straight and inclined pipes. Some specifications of this
product design include Stainless Steel T304, a wall thickness of 1.6 mm and a T3 turbo flange.
Another example of designs made by OBX Racing is the one shown in figure 3. In this design all
tubular pipes have several inclination angles before they converge at the collector, and the length
of the pipes also changes in comparison to the length of pipes in figure 2. The manifold in figure
3 is made of stainless steel and includes a T25/T26 turbo flange.
Figure 3 1.8L Turbo Manifold (Reference 4, OBX Racing)
Page |9
Important Parameters
There are several parameters we should have in mind when designing an exhaust manifold.
The geometry of the manifold which includes parameters such as length, diameter, and shape
essential for a successful design. Multiple shapes involve the use of oval pipes or circular pipes
in the manifold design. The fact that circular pipes have a larger cross-sectional area than oval
pipes and provides a better flow distribution is a main advantage. The pipes can be designed with
an inclination angle or as a combination of straight and angular pipes. High inclination angles of
pipes in a manifold will increase the flow resistance which is an undesired effect. The diameters
of the pipes should also be taken into account since pipes with higher diameters are more suited
for engines at higher speeds because more exhaust gases are expelled at higher speeds. On the
other hand, pipes with lower diameters are better used for lower engine speeds. The length of the
pipes is another important factor when designing a manifold. This is required to get the most out
the space we have available to design the manifold and select the appropriate pipe length. Pipes
that are too long allow the exhaust gases to expand, thus reducing the exhaust speed. This
phenomenon is not desired because once the exhaust gases speed is reduced, more residues of
gases remain in the cylinder which decrease the power and efficiency of the engine. That is the
reason why selecting a good length for the pipes will help reduce to a minimum the residue of
gases that remain in the cylinder.
Temperature is another important parameter that needs attention when designing a manifold
due to the high exhaust temperatures and the heat transfer. Sometimes some kind of insulation is
provided to the manifold in order to reduce the heat transfer via radiation to the engine bay. In
most of the cases a ceramic coating is sprayed to the manifold to provide thermal insulation. The
thicker the ceramic mixture the better the insulation will be. Another type of insulation used is
P a g e | 10
wrapping for the exhaust manifold. This method is in fact cheaper than the ceramic coating but
offers less insulation and also a much lower cosmetic value.
Special interest must be given to the back pressure parameter. Back pressure is an undesired
effect because as the back pressure increases so does the amount of residuals left in the head. The
increase in weight of residuals will decrease the volume of the fresh charge increasing the
temperature at the beginning of compression as well. Several experiments show how the increase
in back pressure influences the thermal efficiency for different induction manifold pressures.
Figure 4 shows five different values for induction manifold pressure. We can observe how the
thermal efficiency changes when there is an increase in back pressure from 6 lb/in2 to 20 lb/in2.
Figure 4 Thermal Efficiency vs. Back Pressure(Smith 1)
P a g e | 11
As shown in Figure 4, for any case of induction manifold pressure there is a decrease in the
thermal efficiency when the back pressure increases. It is also important to note that for the
highest induction manifold pressure we obtain the lowest thermal efficiency reduction, and for
the lowest value shown of induction manifold pressure we get the highest reduction in thermal
efficiency. In general the lowest induction manifold pressure value and the highest back pressure
value will reduce considerably the net output.
Materials
The materials used to manufacture manifolds are mainly stainless steel or cast iron. When
these two materials are analyzed and compared, several differences can be observed in aspects
such as: elastic modulus, density, thermal expansion coefficient, thermal conductivity, specific
heat, and others. Table 1 shows the main properties of AISI Type 316L Stainless Steel.
Table 1 AISI Type 316L Stainless Steel Properties
Property
Value
Units
Elastic Modulus
2e+011
N/m^2
Poisson's ratio
Shear Modulus
0.265
8.2e+010
N/A
N/m^2
Density
8027
kg/m^3
Tensile Strength
Compressive Strength in X
485000000
N/m^2
N/m^2
Yield Strength
170000000
N/m^2
Thermal Expansion Coefficient
1.65e-005
/K
Thermal Conductivity
14.6
W/(m·K)
Specific Heat
450
J/(kg·K)
Material Damping Ratio
N/A
P a g e | 12
Table 2 shows the main properties of gray cast iron and the main differences between
these two materials. One of the properties in which where is a huge difference is the thermal
conductivity. Stainless Steel has a thermal conductivity of 14.6 W/ (m·K) while the gray cast
iron has a value of 45 W/ (m·K). This difference means that stainless steel material would take
longer time to warm up and in this case make the engine bay cooler, which makes the engine
work in a more desirable environment and make more power. Also stainless steel has a higher
density and elastic modulus value than gray Cast Iron while cast iron has a higher specific heat.
All these properties need to be taken into account when selecting the material for a manifold,
specially the elastic modulus and thermal conductivity.
Table 2 Gray Cast Iron Properties
Property
Value
Units
Elastic Modulus
6.61781e+010
N/m^2
Poisson Ratio
0.27
N/A
Shear Modulus
5e+010
N/m^2
Density
7.2e03
kg/m^3
Tensile Strength
1.51e+08
N/m^2
Compressive Strength
5.72e+08
N/m^2
Yield Strength
N/m^2
Thermal
Expansion
Coefficient
1.2e-005
/K
45
W/(m·K)
Specific Heat
510
J/(kg·K)
N/A
P a g e | 13
1.4 Discussion
As mentioned before, the Mazda Miata is an incredible vehicle. The car is not only sporty
and fun, but also a little devil at the track. Since the car is so popular along enthusiast we thought
that it will make a clear market sense to invest R&D time and money in the design of parts for
the Miata since the clientele and the fan base is already there for the car. The team’s idea is quite
simple, to design, prototype and construct a manifold under the most rigorous of engineering
design methods, proving to be a superior solution to other manifolds in the market as well as
making it a better value for the everyday enthusiast.
P a g e | 14
Chapter 2
Project Formulation
Chapter 2 will reveal the origins of the components to be modified as well as explain the
objectives, design specifications, and constraints of this project. Additionally, main and partial
objectives, including global learning components, will be identified and lead to a discussion of
possible design resolutions.
2.1 Overview
When the team sat down and started discussing options and ideas, it became clear that in
order to reduce cost it would be a great idea to make the manifold out of cast iron. Cast iron is
not only very strong but also has great thermal characteristics that make it the material of choice
for OEM like Ford, BMW etc. The team main concern at the moment is that casting practices are
very expensive and require large orders to manufacture within budget. Stainless steel, on the
other hand, was readily available and, although the cost of admission for stainless was a lot
higher than cast iron, at least for the sake of a prototype, stainless made more sense. We ran
calculations and simulations for both materials since the end product could be made both ways
and a choice can be given to the consumer as to what material is desired. Stainless steel polishing
and refinishing is also discussed in order to improve the appearance of a manifold as opposed to
cast iron which tends to have rougher look both in color and texture.
P a g e | 15
2.2 Project Objectives
The main objectives of this design project are as follows:
1. Create a Mazda Miata Turbocharger manifold with incredible flow characteristics.
2. Easy to manufacture which could fits the budget of an amateur enthusiast. Team is trying
to reach a 200 dollar maximum prototype price for the manifold for the cast iron unit and
about 500 dollar for the stainless steel variety.
3. Create a choice of materials either stainless steel or cast iron and let the end user choose
which is desired.
4. The manifold must be easy to install and allow for normal installation in the Miata’s
engine bay without modification of any of the internal components.
5. Simulate as many variables present in the manifold in order to accurately describe the
dynamics of the manifold.
2.3 Constraints and Other Considerations
Constraints that the team was facing with the proposed design are based on the effect of the
extremes or higher abuses in the metal selected, rain, drastic changes in the air temperature can
damage the properties of the material in a second, and bend the metal or make it more vulnerable
to future ruptures at the welds. Another constrain, the shape of the curvatures will tremendously
affect the flow, as will temperature which is essential for the functionality and performance of
an engine whether made of stainless steel or cast iron.
P a g e | 16
Chapter 3
Design Alternatives
This section is to demonstrate the different design alternatives. Out of these design
alternatives the team will select the higher efficiency model. The selected model will be
optimized to its maximums potential. Note: all alternative designs will be tested and optimized
equally to assess their specific performance and compare them through CFD analysis. The
proposed design selection will be tested with real flow in a flow bench.
Our team is working on developing different design alternatives. The team conducted
research in order to find information about exhaust manifolds for Miata 1.8L available in the
market. This information helped us know what has been made to date and how we can come up
with something new and efficient.
3.1 Overview of Conceptual Designs Developed
Every design has its own geometry in terms of angles, shapes and length. All three design
alternatives have been made with the purpose of being creative, unique, in order to have three
good options to eventually select the optimal design. Having this in mind we developed each
design alternative in such a way that inclination angles and intersections of the pipes are
completely different one from another, as well as the collector design. For all three designs we
used the same values and shapes for parameters such as, internal diameter of the pipes and size
of the leads. Having such parameters with the same values will allow us to compare the results
obtained after a flow simulation analysis on each specific design. The flow simulation study
analyzes parameters such as, average velocity, maximum velocity, total pressure and average
pressure of the exhaust gases. As part of the results we are also able to visualize the areas in the
manifold where these parameters are maximum and minimum. Such information will help us
P a g e | 17
later on to optimize the selected design alternative. The total length available is five inches,
which means that each design can have a length up to five inches or less, but not more than five.
3.2 Design Alternative 1
Figure 5 Exhaust Manifold Miata 1.8L
Design alternative 1 is an assembly composed of twenty pieces that all together cover five
inches in length. One of the main specifications of this design are the fact that the two middle
pipes have an inclination of twelve degrees with the horizontal while the two outer pipes have
eighty six degrees of inclination. These angles were selected intentionally for the purpose of
increasing the average speed of the exhaust gases to obtain a smooth flow distribution. The
twelve degrees angle inclination of the middle pipes provides a better flow efficiency than other
designs
that use straight middle pipes because of the fact that these inclined pipes will
eventually meet at a point when they are about to enter the collector reducing the flow resistance
P a g e | 18
to a minimum and enhancing a smooth flow distribution. On the other hand middle pipes with a
straight configuration reduce the average speed of the exhaust gases considerably when entering
the collector because of the separation between these two pipes. The outer pipes eighty six
degrees inclination also offers a reduction in four degrees compared to many designs that use
ninety degrees for the outer pipes providing an increase in flow velocity in that area. The outer
pipes intersect the middle pipes right before they enter the collector. This was also strategically
selected with the purpose of improving the flow efficiency. If the outer pipes meet the middle
pipes at a shorter distance, then the angle of the outer pipes would have to be greater and hence
the flow resistance will increase, which is not desired. The collector was also designed in such a
way that no obstruction to the exhaust gas flow will occur minimizing the output loss. Figure 5
shows exhaust manifold design alternative 1composed of the manifold flange, system of pipes,
collector and turbo T3 flange. Some specifications of the design alternative are an inner diameter
of 2.0 inches, outer diameter of 2.38 inches with a thickness of 0.15 inches.
P a g e | 19
Design alternative number two is based on a symmetrical type model. This model is straight.
What the team calls straight is that the flange where the turbo will be mounted is perpendicular
to the horizontal axis. This design has pipes with 2.077 inch inside diameter and thickness 0.154
inches (OD: 2.375) as well as most of the conceptual designs. What makes this turbo exhaust
header a conceptual selection in this project is the way that the pipe angles are distributed
throughout the model. The outer pipes have smooth radial angle of 3.57 inches with an arc
diameter of 4.34 inches. The two pipes that are located at the middle have a radius of 9.24 inches
and an arc of 4.06 inches. Open arcs create less constrictions and coincidently more flow and
lower temperatures, as well as friction is diminished.
P a g e | 20
The last design the team attempted was a top mounted ram horn. The ram horn is very
popular among enthusiast with high HP builds. The main concern the team had with trying a
design with this configuration was the increased cost of making such a complicated geometry.
Although the ram horn could also be molded and made out of cast iron, the team thought that it
would defeat the purpose of our design. The team tested the design of the ram horn as far as flow
and thermal and decided to mimic its flowing characteristics. If the team could design a modified
log manifold and make it flow as close as possible as the ram horn, the team knew we had hit the
sweet spot of cost and performance. The team knew that a manifold with a ram horn decided was
difficult to surpass since the average velocity in those manifolds are extremely high compared to
anything we tried. In figure 7 can be appreciated the ram horn manifold configuration, while in
figure 8 can be observed a flow simulation analysis.
Figure 8 Ram horn manifold configuration
P a g e | 21
A design with independent runners works best since there is no turbulence between one flow
and the other. Even the transient analysis which takes in consideration firing order and timing
proved that the lack of turbulence in the center created all the difference in flow dynamics.
Design Alternative 4 was more of a very simple test. We wanted to evaluate the amount of
turbulence present in a design where all the runners fused together and then had a pressure driven
collector. The result was loss of flow compared to any other design with individual runners. This
model is actually quite popular among cast iron manifolds for simplicity but even in the
community they are known for poor numbers.
P a g e | 22
This design has no round pipes, is the only design alternative without the expansion of a
round pipe, and has a center tank where all the pipes meet to exit through the turbo flange. This
geometry is great with cast iron and cast techniques in general; the material selection for most of
the motor vehicles in the world, one of the main reasons is cost of the material and how it
behaves through the time used. This design is also one of the most comfortable in terms of
installation of the part yet the length of the manifold might make turbocharger selection a lot
more difficult.
P a g e | 23
3.6 Feasibility Assessment
Any of the designs worked on by the team are feasible, any of them could be made into a real
model and later tested and successfully used. We chose design number 1 simply because it had
the best characteristics possible as compared to anything else the team could have built. The
design chose has some key benefits:

Simple design with small contractions and bents renders great flow numbers.

Simple to cast and weld since the amount of turns and cuts is minimal.

Geometry does not restrict materials in any way.

Tolerances are not as tight as with any other design except at the center where good
closure is key to prevent having to make a greater bead of weld.
3.7 Proposed Design
The team had to design a manifold with many constraints and ideas in mind. Like most
aspects in life, there is a tradeoff and pros and cons in any decision the team makes. Just as when
buying a diamond, the customer must choose between clarity, color, size and price, designing a
manifold is no different. There are always limitations and compromises made throughout the
process. The most important element for the team was to design a cost effective, good flowing
manifold. The team knew that superior flow characteristics could be achieved by tweaking the
design and making a manifold based on current CFD analysis tools and techniques. The
industries do not use them so we had a distinct advantage over them. The idea was to make the
best out of the least. In figure 11 we can see the flow analysis distribution of the proposed
design.
P a g e | 24
Figure 11 Flow Analysis distribution of the proposed design
The team started with design alternative 3 as a reference design. Design Alternative 3 has a
main manufacturing constraint that comes from the fact that the tubing is so bent that it requires
numerous cuts and more welding than the other designs. Also the collector in the RAM design is
difficult to precisely since it compresses gases into the turbine and failure to do so smoothly
would create flow issues. The team set out to try to emulate the ram horn flowing characteristics
in a well-designed log.
One of the first phenomenon we learned about is the center of low pressure of nonindependent manifolds.
P a g e | 25
Figure 12 CFD Analysis of Alternative Design
Right at the intersection of the two middle tubes there forms a low velocity area, created by a
high pressure zone that forms from the intersection of the low speed currents moving along the
wall of the steel, when these two walls of low speed flow meet they create a tough zone where
flow is broken into and rotated like a blender would move air. This phenomenon can be observed
in figure 12.
P a g e | 26
Figure 13 Pressure Analysis in Solid Works
Figure 14 Velocity Distribution Analysis
P a g e | 27
The higher pressure creates so much interference in the area before the collector that the team
thought about using a separator plate to see if the interference could be broken but soon found
out that the divider acted like yet another wall and created an indentation in the pressure zones
towards the exit, further reducing the flow mechanics of the manifold and reducing average
speeds and exit speeds. Therefore we turned into the final design which minimized the center
area and developed by far the best flow patterns.
3.8 Discussion
After several flow simulation studies that we conducted to the proposed design by using
multiple manifold lengths such as, 5.0 inches, 5.9 inches, 6.0inches and 6.1inches, different
angle combinations and collector lengths, we can say that the maximum average speed was
obtained for a length equal to 6.1inches for all configurations. Unfortunately, the actual physical
space available to place the exhaust manifold is 5.0 inches. We optimized our proposed design
with the objective to manufacture it with the geometry that provides the maximum average speed
possible to obtain the highest possible volumetric flow efficiency and power output. In order to
accomplish this task and having in mind the flow simulation results from previous studies, we
analyzed the proposed design for a maximum length possible of 5.0 inches. The different angular
combinations for the inner and outer pipes analyzed can be observed in table 3.
Table 3 Different angular configurations for inner and outer pipes.
Angular Configurations
Inner Pipes
Outer Pipes
Angular Configuration 1
10°
11°
73°
82°
12°
13°
86°
90°
P a g e | 28
These angular configurations also affect the straight inner and outer pipes length as well
as the collector length while keeping the same original inner and outer diameters for all
configurations analyzed. Small angular configurations such as: 10° and 73° for inner and outer
pipes will result in longer straight inner and outer pipes but a shorter collector length. The
highest angular configuration studied which is 13° and 90° for inner and outer pipes will affect
the geometry by obtaining shorter inner and outer straight pipes and the highest collector length
possible. Very important is the fact that all configurations analyzed have a total length of 5.0
inches. The proposed design that we can see in figure 5 is design in such a way that the outer
pipes meet the inner pipes right before entering the collector with the purpose of increasing the
average velocity of the exhaust gases and improve flow efficiency. This fact is the reason why
there is such an angular difference when comparing one configuration with another. We also use
the same values of input pressure for all the cases; this allows us to obtain values of velocity
based only in the geometry of the assembly. The results obtained for average velocity and
maximum velocity are shown in table 4.
Table 4 Average velocity and Maximum velocity for multiple angular configurations
Angular Configurations
Inner Pipes
Outer Pipes
Average Velocity Maximum
(m/s)
velocity(m/s)
10°
73°
89.098
330.521
11°
82°
94.943
328.628
12°
86°
98.112
324.352
13°
90°
97.036
324.642
P a g e | 29
As we can see in table 4, angular configuration 3 which is the one with 12° inner pipes and
86° outer pipes, was the design that obtained the highest average velocity value. We can notice
from table 4 that the average velocity increases as we increase the different pipes angular
combination, however an angular configuration higher than configuration 4 will experience a
decline in the average velocity due to the fact that as the outer pipes angles increases to a value
greater than 86°, so does the area of slow flow distribution inside the exhaust manifold. The
maximum velocity parameter is an important parameter as well, but not as important as the
average velocity because maximum velocity represents just a point where maximum velocity is
obtained; while average velocity represent the average velocity of the entire design. Figure 15
shows the flow distribution of the entire assembly representing angular configuration 3 with 12
degrees inner pipes and 86 degrees outer pipes.
Figure 15 Exhaust Manifold velocity flow distribution
P a g e | 30
Figure 15 shows the areas where the average velocity is the lowest is in the outer pipes which
because of the 86° angle inclination will affect negatively the overall maximum velocity. On the
other hand we can see that the inner pipes experience a much higher velocity that the outer pipes
due to its little angular inclination. As we can see in figure 15 the highest velocity is obtained
inside the collector which due to the designed geometry of the pipes provides an almost perfect
flow distribution. Just very small areas in the corners of the collector result in low average
velocities, while the rest is flowing at the maximum speed possible. The flow analysis was also
conducted to the optimized design but without collector in order to find out how much is the lost
in velocity. It turns out that the optimized designed without the collector has an average velocity
of 111.189 m/s which means that the lost in velocity is around 11.76 %. This results probe that
the collector shape and design is very efficient providing the minimum lost possible. There were
others designs alternative tested in which the lost in average velocity was as high as 40%. In
figure 16 can be appreciate the exhaust manifold flow distribution without collector.
Figure 16 Exhaust Manifold velocity flow distribution without collector
P a g e | 31
The team also conducted an additional study to an exhaust manifold very frequently used
now a day by owners of the Miata 1.8L (1994-2005). This design was not a design alternative
but allowed us to compare our proposed and optimized design to designs that already exist in the
market. The design tested has straight inner pipes and outer pipes with 84° of inclination which
is 2° less than our proposed design. The design flow distribution can be appreciated in figure 17.
Figure 17 . Exhaust Manifold velocity flow distribution
As we can see from figure 17 the velocity in the outer pipes is reduced considerably in
comparison to the proposed design because of the straight pipes and the geometry of the
collector. Also the speed in the inner pipes as well as in the collector in which due to the
separation between the two inner pipes provokes a considerably speed reduction in the collector.
The maximum average velocity obtained in this study was a value equal to 84.4 m/s which is
much smaller than the average velocity obtained in the proposed design equal to 98.112 m/s.
These results demonstrate that that the proposed design is not only going to be efficient and very
unique but is proven that already possess much higher volumetric flow efficiency than models
and design that exist in the market nowadays.
P a g e | 32
Chapter 4 Project Management
This chapter estimates duration of the project, how it is organized, and overview of the design, and
breakdown of project responsibilities by the team members in specifics tasks. Patent of the conceptual
design and commercialization will be considered.
4.1 Overview
Time lines and tables are presented to obtain a better visualization of the time and progress
made by the team and what to expect for future development. Also to indicate the breakdown of
responsibilities among team members. These tables and timelines also indicate and predict team
effort dedication and cooperation.
4.2 Specific Tasks Comprising the Project
At the beginning of the project each team member was assigned with specifics tasks and
main goals to manage the process of the project. These tasks were assigned with the consent and
approval of all team members. The only counter time or change that this personal assign task
could have is that team members have to work in conjunct to fulfill it. The weight of the tasks
assigned are evenly distributed or compensated with extra work in other areas. Many times some
team members have to wait for the others to finish in order to start working on a dependent task,
This time delay is minimized if the member who is waiting starts helping the one seeking this
result. Team work and cooperation is always present. The main purpose is to make a product that
is more efficient, less noisy, and affordable.
P a g e | 33
4.3 Organization of Work and Timeline
Table 5 shows the time line in a tabulated format for the senior design project completion.
Major tasks to be completed during the entire project are outlined as well as their respective
starting and ending dates for each task. The duration in days and the end date scheduled for the
tasks to be completed are also listed.
Table 5 Timeline Data for Senior Design Project Completion
Tasks
Start Date
End Date
6/1/2011
Duration
(Days)
20
Project Formulation
Literature Survey
6/10/2011
30
7/10/2011
Design
AlternativesProposed Design
Material Selection
7/15/2011
15
7/30/2011
7/25/2011
10
8/5/2011
Finite Element Analysis
7/20/2011
40
8/30/2011
Cost Analysis
9/5/2011
15
9/20/2011
Prototype Construction
10/1/2011
20
10/21/2011
Prototype Testing
10/25/2011
15
11/10/2011
Report
6/1/2011
165
11/16/2011
6/21/2011
Figure 18 shows the time frame for the completion of the Senior Design Project. The Gantt
chart illustrates the major tasks that have been completed and those to be completed for each
individual task. The bars in red represent the major tasks that have been completed so far, while
the bars in green represent the remaining tasks that will be developed later on.
P a g e | 34
Figure 18 Time Frame Senior Design Project
4.4 Breakdown of Responsibilities among Team Members
In order for a project to be completed successfully all participants need to be actively
involved throughout the process. Team number one analyzed the different tasks that needed to be
done and distributed them among all team members. Some tasks are distributed individually to
team members while others require the participation of the entire team. Project formulation and
material selection was assigned to Elmer Gutierrez, Rogelio Muro is in charge of cost analysis as
well as Thermal and Static studies, while Michael Martinez conducts the literature survey and
vibration study. There are several major tasks that because of time constraints and the
importance of the design require participation of all team members. The report, for example, is a
major task that must be equally and simultaneously addressed by all project participants. Design
P a g e | 35
alternatives will be conducted by all three members where each member will come up with a
design alternative.
Flow simulation analysis to each design alternative will also have the
participation of the entire team. Another task that requires the presence of all team members is
the prototype construction as well as testing and evaluation of the final design.
4.5 Patent/Copyright Application
After an extensive research in the US Patent Office database online, no significant patent design was
found related to any exhaust manifold for Miata 1.8 L. Our team intends to apply for a copyright and
patent since we believe our proposed design concept is unique and we have not found a similar
design among all the manufacturers we have researched.
4.6 Commercialization of the Final Product
The material of the proposed design have not been specified yet due to simulations
analysis as explained above one of the material selections was eliminated due to the cost and
complications of the casting process (cast iron). Stainless steel 304 is the one that is in the lead in
terms of materials for the proposed design; but, based on the structure and shapes of the exhaust
the team was able to draw some conclusions. This is one of the most desirable shapes in the
market due to the position and the space available to install the part. Due to the velocity accurate
in the propose design this manifold will sell itself
The team is planning to sell the product first at local auto parts dealers and local high
performance shops. Depending on the reaction to these sales, global commercialization of the
final product through the internet will be considered.
P a g e | 36
Chapter 5 Engineering Design and Analysis
5.1 Design Based on Static and Fatigue Failure Design Theories:
Miatas are equipped with aluminum blocks. These blocks and cylinder heads are very sturdy
and strong, yet when a load is present in the head, in the sense of a torque being applied in the
exhaust studs via the turbocharger, downpipe and exhaust hanging from the head, special
considerations must be taken in order to prevent exhaust leaks and other even more catastrophic
failures like the studs becoming so elongated that they actually damage the cylinder head or the
actual stud breaks and lets the exhaust assembly fall. The Miata is prone to these problems and,
therefore, some companies have made Income Cylinder Head Studs. However, these are so
expensive that they are out of the reach of most enthusiasts. These four studs sell for 100 dollars
including the hardware.
At this point in the design process, the team hit a crossroad. The manifold could be made
longer and could be maintained in the engine bay as the factory designed header or it could be
made shorter, lose some flow since there is a highest level of overall contraction in the manifold,
and yet be safe.
5.2 Material selection
This is the selection of materials for the design. Two out of the three are commonly used in
the market and aftermarket exhaust manifolds. These materials are being studied through
simulations to select the most convenient one for our selection of manifold design. Due to
extensive research and studies of the casting process for the gray cast iron, the team noticed that
the process is too expensive and cannot be created at any lab of Florida international University.
P a g e | 37
That is the main reason why it was eliminated the gray cast iron of our propose design. The
simulations are provided for the three materials to compare and to make future references.
Analyzing the material properties of all three materials, AISI 316L, AISI 304, and gray cast,
iron, we noticed a huge difference between the gray cast iron and the stainless steel materials.
The difference that we first noticed was the thermal conductivity which is 45 W/ (m·K) for the
gray cast iron, 14.6 W/ (m·K) for AISI 316L and 14.6 W/ (m·K) for AISI 304. These values
guarantee much faster heat dissipation for the gray cast iron material followed by AISI 304 and
finally AISI 316 with the lowest thermal conductivity. A thermal study of the proposed design is
also going to be conducted analyzing all three materials in order to obtain accurate results and
compare how each material responds to the high temperatures existing in the manifold. The
tensile strength is another property with a huge difference between stainless steel materials and
gray cast iron. AISI 304 is the material with the highest tensile strength value which is 5.17e+08
N/m^2 followed by AISI 316L with 4.85e+08 N/m^2 and gray cast iron with a very low tensile
strength equal to 1.51e+08 N/m^2. The higher tensile strength of these two Stainless Steel
materials compared to gray cast iron will allow these materials to withstand much higher tensile
loads than those the gray cast iron would be able to support. Also a static analysis will be
conducted for these three materials in order to confirm this statement and to obtain real values of
stress, displacement, strain and factor of safety. In general, gray cast iron has better thermal
properties and will transfer heat at a faster rate than these two stainless steel materials. On the
other hand the stainless steel materials analyzed have higher tensile strength properties than gray
cast iron. Higher tensile strength will give a lower stress concentration for the Stainless Steel
materials when this kind of force is applied. Specifically comparing AISI 304 and AISI 316L we
P a g e | 38
can say that AISI 304 has better thermal and strength properties and is less dense which will
make this material the preferred one between the two stainless steel materials studied.
AISI Type 316L Stainless Steel
Table 6 AISI 316L Stainless Steel
Property
Elastic Modulus
Poisson's ratio
Shear Modulus
Density
Tensile Strength
Yield Strength
Value
2e+011
0.265
8.2e+010
8027
4.85e+08
1.7e+08
1.65e-005
Units
N/m^2
N/A
N/m^2
kg/m^3
N/m^2
N/m^2
/K
Specific Heat
14.6
450
W/(m·K)
J/(kg·K)
N/A
AISI Type 304 Stainless Steel
Table 7 AISI 304L Stainless Steel
Property
Value
Units
Elastic Modulus
Poisson's ratio
Shear Modulus
1.9e+011
0.29
7.5e+010
N/m^2
N/A
N/m^2
Density
Tensile Strength
Yield Strength
8.0e+03
5.17017e+08
2.06807e+08
1.8e-005
kg/m^3
N/m^2
N/m^2
/K
Specific Heat
16
500
W/(m·K)
J/(kg·K)
N/A
P a g e | 39
Gray Cast Iron
Table 8 Gray Cast Iron
Property
Value
Units
Elastic Modulus
6.61781e+010
N/m^2
Poisson’s Ratio
0.27
N/A
Shear Modulus
Density
Tensile Strength
5e+010
7.2e+03
151658e+008
N/m^2
kg/m^3
N/m^2
Compressive Strength
Yield Strength
Thermal
Expansion
Coefficient
Specific Heat
5.72165+e008
N/m^2
N/m^2
/K
1.2e-005
45
510
W/(m·K)
J/(kg·K)
N/A
Static Analysis
AISI 304 Stainless Steel
One of the studies that the team conducted for the proposed design was a static analysis. This
study was necessary since the exhaust manifold needs to support a load once the T3 turbo is
attached to it. In this particular study we fixed the nine holes of manifold flange and applied a
load of 150 N in the top face of the T3 flange. The mesh used for this study was 4.5 mm which is
the finest possible, using AISI 304 stainless steel, AISI 316L stainless steel and gray cast iron.
Parameters such as stress, displacement, strain and factor of safety were obtained from this
simulation.
P a g e | 40
Figure 19 Stress Simulation Analysis using AISI 304
As we can see from Figure 19, the maximum stress value obtained is 7.213602e+006
N/mm2 and is located in the collector area. This maximum value is much smaller than the yield
strength for the AISI 304 which is 2.06807e+008 N/m which means that the design is totally
safe. High stress concentration areas are also located near the holes. Figure 20 shows the results
obtained for the displacement analysis.
P a g e | 41
Figure 20 Displacement Simulation Analysis Using AISI 304 stainless steel
We can notice in figure 20 that the maximum displacement value is located in the center of
the edge where the load is applied. The maximum displacement has a value equal to 0.006897
mm which is insignificant for the load applied. Figure 21 shows the graph results corresponding
to the strain analysis conducted.
Figure 21 Strain Simulation of 304 stainless steel
P a g e | 42
As shown in figure 21 the relative displacement between particles in the assembly is
extremely small. The maximum ratio value obtained was 0.00001934 which is insignificant.
This study allows us to determine how a given displacement differs locally from a rigid body
displacement. In figure 22 is represented the last result from the static analysis conducted.
Figure 22 Factor of safety of 304 stainless steel
This analysis represents the factor of safety of the assembly including locally minimum and
maximum values. The minimum factor of safety value is 28.67 which guarantee total safety for
the proposed design. We also determined the maximum load the assembly will be able to
support, and it turns out that this value is 4300.5 N. Applying such load will result in a minimum
factor of safety of 1 as we can see in figure 23. This maximum load supported will never be
reached in real life application of the assembly, which ensures a statically safe design using AISI
304 Stainless Steel.
P a g e | 43
Figure 23 Analysis of maximum load supported using 304 stainless steel
AISI 316L Stainless Steel
We did the same study for AISI 316L in which we analyzed the corresponding results. The
maximum stress value using AISI 316L is 7,252,747 N/mm2 which is higher than the maximum
stress obtained by using AISI 304. The complete stress distribution of the assembly can be
appreciated in figure 24.
Figure 24 Stress Simulation Analysis using 316 stainless steel
P a g e | 44
The displacement values obtained using AISI 316L are once again very small with a
maximum value of 0.006528 mm as we can see in figure 25. This value is relatively smaller than
the maximum displacement for AISI 304.
Figure 25 displacement of 316l stainless steel
The strain parameters are very similar to the study conducted using AISI 304. In this case for
AISI 316L the maximum strain value is equal to 0.00001813 and is shown in figure 26.
Figure 26 Strain Distribution of 316l stainless steel
P a g e | 45
The corresponding factor of safety analyzed for this material is shown in figure 27. We
obtained a minimum factor of safety equal to 23.44 which puts the system in the safety side by a
huge stretch. Even though this results shows a save design, the 23.44 factor of safety is smaller
than the 28.67 factor of safety obtained using AISI 304 due to the fact that higher stresses are
present when using AISI 316L and the mention load is applied.
Figure 27 Factor of safety 316l stainless steel
The maximum load supported by the assembly using AISI 316L was also determined. It turns
out that the assembly will be able to withstand a total maximum load equal to 3516 N as we can
see in figure 28.
P a g e | 46
Figure 28 Analysis of maximum load of 316l stainless steel
Gray Cast Iron
Gray cast iron was the other material we analyzed as part of our static analysis study and
materials selection task. The same conditions were applied for this study obtaining the results for
stress distribution, displacement, strain and factor of safety. Figure 29 shows the Von Misses
stress distribution for the proposed design. The maximum value is equal to 7,249,642 N/mm2 and
is located in the collector area.
Figure 29 Stress Simulation Analysis using Gray Cast Iron
P a g e | 47
Figure 30 shows the scale with the displacement values located throughout the assembly. The
maximum value of displacement is 0.1974 mm which is the maximum value obtained among all
studies conducted so far.
Figure 30 Displacement of Stress Simulation Analysis using Gray Cast Iron
Figure 31 represents the graph for the strain analysis conducted on gray cast iron. The
maximum value of 0.00005496 obtained is located in the collector area. As the strain ratio is
directly related to the deformation value this maximum value of strain is also the highest value
obtain among all studies.
P a g e | 48
Figure 31 Strain Gray Cast Iron
The final result that we analyzed was the factor of safety of the assembly for the applied load.
The minimum factor of safety value is equal to 29.04 which is very similar to the factor of safety
obtained for the AISI 304 stainless steel and higher than the minimum factor of safety for AISI
316L stainless steel. The graph with full detail of the factor of safety distribution is shown in
figure 32.
Figure 32 Factor of safety Gray Cast Iron
P a g e | 49
In figure 33 we can observe a graph with the maximum load that the assembly using gray
cast iron is able to support. The maximum load allowed is 4356 N which is the maximum load
value obtained.
Figure 33 Analysis of Maximum Load Supported using Gray Cast Iron
5.3 Deflection Analysis
Figure 34 Deflection Analysis
P a g e | 50
The manifold needed to achieve enough flexibility so as to not crack under the cyclic heat
load, yet be strong enough to support the weight efficiently plus resist the vibrations and uneven
loading provided by the hanging exhaust. The team had already seen and analyzed a few of the
failed manifolds around the web and began working on solutions to the biggest and most
common problems found in the inexpensive manifolds. The first common problem found with
manifold came from the metal cracking under stress, cyclic loading and temperature fluctuations
of a turbocharged car. The EBay manifolds, as they are called, lack not only the right
composition of materials, but also the wrong tolerances and thicknesses. The EBay manifolds are
made for the most part of schedule 5 steel, which is not thick enough to withstand the abuse to
which these cars are subject.
Figure 35 Crack prone Manifold
P a g e | 51
The typical EBay log manifold, offers a few team tested drawbacks:

Two middle plenums with large separation create a dead spot in the middle of the
manifold that creates losses in the system.

If the outer tubes are too arched, they created turbulence in the center section that
slows that the overall performance of the manifold.

Material is not to specs as far as ASME.

Geometry and thickness subpar with minimum requirements

Failure rates very high in the course of 30 days
When all these characteristics are combined, the end user is left with a manifold that although
really inexpensive, is very prone to failure. Failure of the manifold does not only render the car
inoperational but also, could damage other components as well.
Figure 36 Crack
P a g e | 52
A Miata Chinese made manifold made out of schedule 10 stainless steel T304 failing at the
support of the flange. The torque made by the weight of the turbocharger combined with the
expanding/contracting stress generated by the engine’s exhaust gases
5.4 Thermal Analysis:
One of the manifold biggest design constraints is the fact that the manifold heats and cools
down every time the car goes from point A to point B. The heat flux through the manifold is also
non- linear, meaning that manifold gets a given amount of heat when the motor is at 1000rpms
but a different heat load when the motor is at 7000rpm since the amount of air being moved
through the manifold at the time is a lot more. To complicate matters even more, is very difficult
to measure the temperature that the manifold experiences because of a few specific factors:

EGT or Exhaust Gas Temperatures fluctuate with the tune of the turbocharged car and
therefore it’s almost impossible to set an specific temperature at which the manifold will
operate; besides the fact that with different power levels, temperatures escalate and range
differently.

The cooling effect of the ambient air to the manifold is very difficult to measure and
reproduce since the ambient temperature varies as well as the specific configuration of
the engine bay.

Different fuel mixes alter EGTs.

Water and elements can create sudden change in temperature with catastrophic effects on
any part with high temperatures including distortion.
With all these elements and limitations in mind, the team decided to perform thermal analysis
and thermal simulations that would produce an insight of the real phenomenon with
P a g e | 53
understanding of the limitations and the consequences of these limitations. The team first sailed
out to survey what was the average value of EGTs in a turbocharged Miata and the consensus in
the community is the temperature should be anywhere between 1200-1300F which seemed
accurate compared to other data from similar vehicles. This rendered a value of 922K to be used
in simulations all around. The team then combined this value with the Flow Works simulation to
see as if, given that the manifold was at ambient temperature and then heated gases came into the
manifold and heated it to the average operating temperature. The results were as follows:
Figure 37 Deflection Analysis
The manifold showed extremely high temperatures at the center of the two middle pipes,
therefore creating a great sense of urgency to reinforce both the weld and the assure tight
tolerances between the two pipes that touch. The manifold heat analysis also showed a very
uniform warming of the central tubes which would mean a very good contraction/expansion ratio
making the manifold less prone to cracks after time and after the 304 becomes brittle over time.
Flange and turbo flange show temperatures as expected, a lot lower than any of the gas moving
pipes, although by the effect of conduction by itself they get to 700F.
P a g e | 54
Chapter 6 Prototype Construction
6.1 Description of Prototype
When the team sat out to build the prototype, the team decided to build the prototype as
realistic as possible compared to what the commercial part would actually be. The team decided
to use the same quality steel and the same dimensions. Also the team tried almost to exhaustion,
to keep the prototype as exact as possible to the Solid Works model in order to get the closest
numbers out of the testing that would follow. This meant that not only the prototype would be as
accurate as possible, but also that some engineering insight was to be gained from constructing
the real manifold. The team new this task was going to be a challenge due to the complexity and
precisions that required some of the cuts of the assembly, but the team was decided to accept the
challenge and put in practice our creative and team work spirit in order to get the job done
successfully.
6.2 Prototype Design
The design of the prototype was done through a series of steps. First we sat in front of the
SolidWorks assembly, then isolated each part until we had an exact model in stainless steel and
followed we spot welded all pieces to keep them in place. We also grinded out any part that
needed to be checked back into tolerance. Then the manifold was welded together and the
finishing process started. Various options existed for the finishing aspect of the construction
process, the team sanded down the manifold until a polished, yet rough surface appeared, then
proceeded to mirror polish the manifold.
P a g e | 55
6.3 Parts List
Raw Materials:

One plate ½” thick, 14” by 5” Stainless Steel 304

Two plates ¼” thick, 14” by 5” Stainless Steel 304

Two 90° Elbows SCH40 Stainless Steel 304

Two wide round pipes 12” long SCH40 by 2” ID Stainless Steel 304
6.4 Construction
1. One ½” thick slab was drilled in order to start the process of the flange. The bolt pattern
of the Miata’s head was drilled into it. The team at this very young construction stage
realized how limited was the student machine shop, the team struggle to find a sharp drill
bit to penetrate the steel and there was no cooling system for the machine, so we had to
cool it by manually spraying water into the drill bit with a plant spray.
2. Once the bolt pattern was done, the team proceeded to make the holes for the exhaust
chambers. The first hole, cylinder 2, was done by the team in one of the mills available at
the student machine shop. After a while, the team realized it was going to take mayor
time to penetrate the material. The bits were not sharp and the machine lacked coolant
system. The team had to literally pour water without discrimination in order to keep the
cutter at working temperature for most of the time. The process of doing the first hole,
cylinder 2, can be observed in figure 38.
P a g e | 56
Figure 38 Knee mill working on first hole, cylinder 2
In figure 39 we can observe two of the team members, Elmer Gutierrez and Michael
Martinez working on the first hole, cylinder 2. The most difficult part of this step was to
complete the round areas with the necessary precision being the fact it was done manually and
with a lack of coolant which delayed the whole process much more than what the team had
initially planned. After a while the hole was successfully completed with the required
dimensions.
Figure 39 Knee mill working on first hole, cylinder 2
P a g e | 57
3. After the first pocket was properly finished, the manifold was starting to look more and
more like the flange we had set out to do. By the beginning of the second hole and having
tried every single drill bit in the student machine shop, we realized we had neither
machinery nor tooling necessary to keep going on. The second hole proved impossible.
4. The team at this time realized it was wasting precious time. Before we had set out to
build the flange by ourselves, the team even had the CNC code to have it perfectly cut by
machine, as can be observed in figure 40. Constraints of money and availability proved in
our way and therefore the CNC idea was unfortunately out of reach.
Figure 40 Manifold flange in the process of being cut by using MasterCam software
P a g e | 58
Figure 41 Manifold flange completed successfully by using MasterCam software
The team had not only the G code for the part but also a simulation made of the part
completely finished as shown in figure 41. At this stage we had to finish the flange to start the
cutting of the other parts, the flange was supposed to be the easy part. The team had to outsource
the pocket drilling to a machine shop close by, where the remaining holes where completed.
5. Once the flange was actually completed the team started assembling the parts together.
The team started working from the ground up beginning with the elbows and the center
section. The parts required filing and at times the team felt like assembling a difficult
puzzle. Angles were very specific in the design and the team understood a difference of 5
degrees had a big impact in the flow as seen in the previous flow analysis, therefore
attention was provided to the right cut. In figure 42 we can see the cuts for the center
section parts which include two twelve degrees joins as well as two pipes with 2.76
P a g e | 59
inches long by 2.0 inches ID. The twelve degrees joins will be attached to the two center
holes in the flange and the pipes will go attached to each join and intersect at a certain
point as we can appreciate in figure 43.
Figure 42 Center section cuts
Figure 43 Center section cuts assembled
P a g e | 60
The outer pipes required an angular inclination of 86 degrees. The team marked the 86
degrees position in the 90 degrees elbow and proceed to do the cutting using a grinder machine
which is property of the team and using a very thin cutting disc in order to make the angular cut
precise. Part of this process can be observed in figure 44 in which one of the 90 degrees elbows
is being cut down to 86 degrees. Once the parts had the basic shape and the elbows were cut to
the precise angle or 86 degrees to be more specific, it was time for a spot weld to hold the
prototype together.
Figure 44 Cut of the 90 degrees elbow in process
Having only basic experience with welding, the team practiced first with a few pieces of
scrap metal until the basic technique was understood. Then the team proceeded to fix the
assembly using very basic TIG welding with small spots used in the lower side of the manifold
P a g e | 61
in order to create a more cosmetically appealing part. Figure 45 shows part of the spot weld
process performed at FIU student machine shop.
Figure 45 Spot weld process
Figure 46 Spot Welding MIG
P a g e | 62
6. At this point the manifold was starting to get in shape, the assembly was looking solid
and the design was being followed as specified by our drawing. It was a great sight and a
great feeling of success that we could actually transfer design from virtual to the physical
world. In figure 47 we can see a partial assembly view with spot weld already performed
fixing several parts of the assembly together.
Figure 47 Partial assembly view
7. One of the hardest parts to build was the tube connecting the two cylinders together, this
small but geometrically challenging parts had to be filed and cut several times in order to
have the perfect shape we needed. The cutting of these special parts required a lot of time
consuming and detailed spirit in order to be completed. Once on, the manifold was spot
welded again and it was a close race to build the collector. In figure 48 is shown the
manifold assembly more completed with the tubes connecting the two cylinders already
spot welded.
P a g e | 63
Figure 48 Partial assembly with tubes connecting cylinders spot welded
The parts required cutting and work in order to make then fit together, not only was this
complicated matter where overcutting could render the part useless, but also a very time
consuming task. Working with AISI 304 is not an easy task, that is why most of the cut at this
point where done with the grinder, as well as the mating of specific surfaces, since the weld can
only fill so much of a gap, the process is requires attention to detail and patience.
8. Definitely the toughest part of the whole assembly was the collector, it felt like building a
card tower, with the same easiness to collapse at any given point since it has to be
measured first, then spot welded. As we can see in figure 49 some of the components of
the base of the manifold were measured, cut and leveled.
P a g e | 64
Figure 49 Partial assembly with collector base leveled
9. Once the collector was cut and tested to be right and accurate the team sat out to machine
the last component, the turbocharger flange. The flange was again pulled from the digital
environment, SolidWorks, printed in real scale and then laid over the ¼” sheet of steel
and draw. The team was able to produce the four flange holes for the screws as well as
the middle part. The middle part was what took more time to finish since precision was
required for this specific part of the assembly. As we can see in figure 50
is the team
working on the middle part of the turbocharger flange at FIU student machine shop.
P a g e | 65
Figure 50 Team #1 working on middle part of the turbocharger flange
The final collector assembly with the rectified flange and the other components that were
previously cut was tested for symmetry and tolerance. Figure 51 shows the partial assembly of
the collector which was measured and corrected several times until desired dimensions were
obtained.
Figure 51 Partial assembly the turbocharger flange
P a g e | 66
10. After all the parts were cut, filed and ready to be welded the team proceeded to perform
the welding process. The initial intentions of the team was to do the welding using TIG
technology. This technology offers several advantages over other welding technologies,
generating high quality, clean welds. After looking for several estimates to weld our
assembly by using TIG technology all prices ranged between 250 and 300 dollars. Since
the prices to weld the manifold using TIG technology was out of the team budget we
decided to use a simpler welding method in order to balance cost and quality. The parts
were taken in front of an expert welder which through years of experience brought our
manifold to life using the simplest of the welding methods, stick welding. Of course, on
the hands of an expert and by supplying him with the best of welding rods, in this case a
package of 24 3/32 inches stainless steel stick electrodes that the team purchased at a
local welding supplier, P & M Welding Supply Inc. The welding came out excellent with
perfect penetration and cosmetically speaking a lot of potential. In figure 52 can be
observed the welding procedure already done with the manifold assembly ready to be
submitted to the finish process.
P a g e | 67
Figure 52 Manifold assembly right after welding process was completed
11. Some sanding, milling and slag removing were made after the bare manifold was out.
Almost like a recent born creature from the far horizons of the universe, it was a sight
that resembled a mixed feeling of strength, beauty and monstrosity.
12. After even the most basic of polishing of the metal, the true colors of the beast were
starting to appear. In order to obtain a better finished we polished the assembly a little bit
more by using a Rex Cut disc specially design for polishing stainless steel and aluminum
materials. After this process was completed our manifolds looks in a better shape in terms
of finishing which consume a lot of time been the fact there were areas in which the Rex
Cut disc could not reach and was necessary to use a hand file to perform the polishing. A
detailed view of the finish and polishing process can be observed in figure 53.
P a g e | 68
Figure 53 Finish and polishing process of manifold assembly
6.5 Prototype Cost Analysis
The prototype cost analysis done by the team covers every single tool, material and extra
accessories purchased by the team once the construction process started. The team did not have
any setback to obtain the necessary materials and cutting tools on time and turned to some of the
most local construction and materials suppliers which are listed in table 9. The total cost
including tax turn out to be 246.89 dollars. This amount is a little bit more expensive than what
the team initially expected, but do to the fact that essential tools such as: Weld Tig Torch, ,
Hand File, Disc cutter, grinding and polishing accessories where not provided to us, these
unexpected purchases had to be made.
P a g e | 69
Table 9 Detailed list of the parts, cost, and supplier information
Parts
Cost
Supplier
2" SCH 40 Pipe T304 20"
28.83
Surplus
and
1/2"x5"x14" T304 Flat Bar
51.33
Simmons
Stainless
Simmons
Stainless
Surplus
and
2" SCH 40 90° Elbow(2)
28.30
Simmons
Stainless
Surplus
and
1/4"x4x14" T304 SS Bar(2)
42.04
Simmons
Stainless
Surplus
and
AK2 Best Weld Tig Torch
25.00
US Welding
Supply
3/32 Excalibur 308 Weld Rod
15.45
P and M Welding Supply
4-1 80A Rex Cut
12.49
Pipeliner
2.99
8" Hand File
18.34
Disc Cutter Accessories
5.97
Home Depot
Total Cost Without Tax
230.74
Total Cost Including Tax
246.89
and
Safety
P a g e | 70
6.6 Discussion
All the design alternatives were evaluated to its maximum potential; out of these design
alternatives a proposed design is chosen. The proposed design it is not the simplest nor the most
complicated of the alternatives, but still it is not an easy prototype to build due to the resources
that we have at the FIU student machine shop. Complicated welds and assemblies did not stop
the team, tight tolerances and the precision necessary to achieve the model. The process began
by buying the material selected by its properties and finite element analysis explored in previous
sections. The construction of the prototype involved machining, welding, vertical band saw,
horizontal band saw, grinder, polishing material, hand filing and precise measurements to
accomplish the goals that the proposed design requires to acquire the desire shape. All the shapes
were built in pairs to match perfect geometry. The detail job required time and dedication, when
these two are accomplish the results are extremely accurate. The manifold is composed of two
flanges, one for the turbo inlet or exhaust outlet and the one for the cylinder head outlet or
exhaust inlet, also is equipped with two 86 degrees elbows and two 12 degrees inclined pipes for
better performance. The two straight pipes are connected by a triangular shape also shown in the
Figure 54that in part fix the straight pipes and give a critical separation to the lower collector
wall. The collector is composed of eight walls also illustrated in Figure 54.
P a g e | 71
Triple cut coupling
Triangular
unification
86 degree elbows
Head side flange
12 degree
incline
Straight Pipes
Figure 54 Assembly with the welding.
P a g e | 72
Chapter 7 Testing and Evaluating
7.1 Overview
The first test was a flow test designed to compare the results simulated at
SolidWorks, in other words the theoretical part of the experiment. This test was developed at the
transport phenomena lab here at the university. There were two types of flow test accomplished,
the first test is a full flow test, this would involve the four pipes of the manifold at five different
total pressures with a time delay between them of ten second for stability purposes and 12 points
tested at the turbo flange cavity. The second flow test accomplish was the single pipe test where
all the pipes of the manifold where tested but this time with only one pipe at a time. Same
procedure for the two flow test and same points of interest. The third test was a welding test,
facility provided by Baker Aviation School, the manifold was pressurize and the welding defects
were market with red ink for localization and fixing purposes. The team then proceed to test the
manifold in its ultimate test, a working Miata.
7.2 Design of Experiments - Description of Experiments
The experiment was designed to test static pressure total pressure and velocity at many
points at the exit flange. The set up consists in a flow bench, a magnetic base with movement in
the three axis and, a pitot tube with a scale on it.
P a g e | 73
Figure 55 Part Number of the Flow Machine
The team conducts several tests on the prototype. The exhaust manifold was tested at full
flow through all the pipes and at each pipe individually. Also it was measured at different
pressures at several points in the turbo flange to establish the fluid behavior at the turbo housing.
As follows there is a walk through the experiment set up:
Figure 56 Outlet of the Duct
P a g e | 74
Figure 57 Control Board
Step 1
Figure 58 Installing Nozzle
The diameter at the end of the flow pipe is 8 inches. The team had to adapt three contractions
to be able to place the box.
P a g e | 75
Figure 59 First Nozzle
First contraction in place and ready to mount the second one.
Figure 60 Third Nozzle Installed
P a g e | 76
The three contractions already installed and with dot tape to seal any leak at the joins are shown
in figure 61.
Step 2
Figure 61 Box install already in nozzle
Step 3
Figure 62 Box preparation for manifold mounting.
P a g e | 77
Figure 63 shows Team 1 losing the box with dot tape for leaks purposes and making
arrangements to mount the exhaust manifold for four pipes flow test.
Figure 63 Box preparation for manifold mounting.
Figure 64 shows box ready for exhaust mounting and elevation arrangements for fixture of box.
Step 4
Figure 64 Manifold already mounted on box ready for flow test.
P a g e | 78
Arrangement of weight and elevation of manifold for zero movement and perfect elevation is
shown in figure 65.
Step 5
Figure 65 Installation of the pitot tube into based
Figure 66 shows preparation to position the pitot tube in the three axis arm.
Step 6
Figure 66 Setting up the engineering sheet
.
P a g e | 79
Set up marks into the graphing paper to fix points of interest and position of base. Once the
arm is in the position required for the experiment, it is fixed and it will be move in terms of its
base through the graphing paper for the X-axis and the pitot tube with the measurement printed
on it body for the Y-axis.
Step 7
Figure 67 Points on the x and y directions
Figure 67 shows position of one of the points of interests reflected at the pitot tube.
Also Y-axis calibration at point 3 right. As explain before many of test points where taken to
acquire a reliable data and all these points where tested 3 times to reinforce credibility and
accuracy of the measurements taking place at the exit flow point.
P a g e | 80
Step 8
Figure 68 Testing Measuring
Top view of the pitot tube position in one of the points selected for testing purposes is shown in
figure 68. Ready for data acquisition.
Step 9
Figure 69 Data recording during testing using Excel.
P a g e | 81
The testing operation is conduct by three members, one of the members is operating the flow
bench and fan speed, the other positioning the pitot tube at the points of interest and transmitting
the data to the third member that is recording all the data acquired from the experiment.
Step 10
Figure 70 Preparing for individual testing
This step it is where the isolation of the pipes begin. The purpose of the isolation is to
get a reading corresponding to a single pipe at a time with the same parameters taken in
consideration before for full pipe test.
P a g e | 82
Figure 71 Testing flow in individual pipes
Step 11
Figure 72 Stopwatch used to measure the time step
The stopwatch is use to make an exact elapse of time between testing and transient period of
the fan motor to record an exact data to match and compare theoretical with experimental
results.
P a g e | 83
Welding Test I
Figure 73 Baker Aviation School Logo
Welding test at Baker Aviation School was based on a test provided by the institution to
make sure that the welding was correct and with no leaks.
Step 1
Figure 74 Material used to create gasket for pressurization
P a g e | 84
Two gaskets where created for this experiment to be accomplish, one made for the turbo flange
and the other for the cylinder head side.
Figure 75 Manifold with bottom gasket
Step 2
Figure 76 Testing Table
P a g e | 85
Once the manifold gasket for pressurization are made the next step is to find a straight
section in the metal table to fix the manifold.
Step 3
Special Flange
C-clamp
Fitting
Bench
Gasket
s
Exhaust
Manifold
Figure 77 Manifold fixed in place for testing.
The manifold is placed with the two gaskets in a straight section of the table and with the
special flange in top of it as you can see in the Figure 77. At this experiment nine C- clamps
where used at the beginning and ten at the end for bottom leakage elimination. At the fitting in
the top of the manifold was applied 120 psi / 827.37 kPa of pressure. Soap and water was used to
detect any leak that the manifold could have. Some porous sections were detected and marked
instantaneously with a permanent marker for welding revision purposes. As soon as the manifold
P a g e | 86
passes this test it will be ready for the ultrasound and magnetic welding test, taking place at
Baker Aviation School.
Figure 78 Manifold with spot welds
After the localization of the holes and porous welding around the manifold, again some
touches wear made to it as shown in figure 79 to seal any linkages and prepare the prototype for
the ultra sound test. Also another pressure test was done to make sure it was ready. Two more
cracks were found during these test as shown in figure 80.
P a g e | 87
Figure 79 The sections wear the leak is located is been marked with the tool shown.
Figure 80 Example of leakage during test
P a g e | 88
Figure 81 Preparations for third test
After this test there was no leaks in any sections of the manifold.
Figure 82 Manifold with Air Pressure
P a g e | 89
Figure 83 Air Pressure Gage
This is a gauge that includes a regulator valve use to control the pressure inside the
manifold.
Figure 84 Water with soap ready to detect any leaks
P a g e | 90
Figure 85 Manifold Polished after the testing was done
Figure 86 Manifold painted
P a g e | 91
Welding Test II
Figure 87 The three components of the Leak Test
Figure 88 Rear view of the three chemicals
P a g e | 92
To get to the second round welding test it was necessary to pass the first welding test
(Pressurization of the manifold). The testing requires penetrate (SKC-S), cleaner (SKL-SP1), and
developer(SKD-S2). These three components are essential for the test. This test consist in a
series of stapes and the chronological order goes as follows:
Step 1
Figure 89 Manifold Clean
The manifold need to be spray and wiped with a rag, to make sure there is no grease
residue or soap from previous testing.
P a g e | 93
Step 2
Figure 90 First Application
Figure 91 Applying Penetrating Agent
P a g e | 94
Figure 92 Manifold Entirely Sprayed
The manifold is divided into sections to start covering the welding sections with the
penetrating spray. Each section needs to be sprayed and let it sit for 15 minutes to let the
penetrate agent get into the cracks and dry up.
Step 3
Figure 93 Excess Removed
P a g e | 95
Once the penetrating agent is dried it is needed to remove the excess to get the section
ready for the developer. The excess will be remove with a rag and the cleaner without spraying
the manifold with the cleaner because if sprayed the penetrant could be remove from the cracks.
The excess need to be gently remove from the piece of work.
Step 4
The manifold is ready to be sprayed with developer and let it sit for 5 minutes.
Figure 94 Developer just applied
Figure 95 Developer Two Minutes after
P a g e | 96
Figure 96 Developer Three Minutes After
Figure 97 Developer SKD-2 fully developed.
P a g e | 97
After five minutes the developer will show the cracks if any, the developer will take
the penetrant out of the crack and show all the porous sections and defects of the welding. In
Figure 98 one is able to note how the developer is able to extract the penetrant out of the holes,
in this case there is no cracks due to the rigorous pressure testing, but is clear that the test is
effective and consistent. The red dots at the collector are light what it means is that the penetrant
is been extract from imperfections of the metal finish," it is nothing to worry about" (Eugene
King, professor at Baker Aviation), told the team.
Thermal Behavior Test
To produce an accurate thermal analysis it was necessary to install the prototype in the
actual engine. This analysis would then reflect the actual temperature gradients in the manifold,
location of the highest and lowest temperature and behavior of the part in terms of deformation
or even possible cracking.
All the data for each infrared picture will be available in a CD for further research
or just information to understand at a higher level the experiment and the numbers reflected
under the figures . The classifications of colors in the infrared pictures are representing the
temperature at a part of the manifold at an specific time. Colors going from blue, green, yellow,
orange and red the order is from higher to lower.
P a g e | 98
The procedure for the thermal analysis is as follows:
Step 1
Figure 98 Stock Manifold
Step 2
Figure 99 Prototype Installed
P a g e | 99
Figure 100 Comparison of body temperature and starting test temperature as seen by the camera.
Figure 100 shows the temperature similarity of the manifold and a hand of a human body to
demonstrate the temperature at the starting point with the engine off to give a perspective of
room temperature . Also used for the calibration of the infrared camera used to perform the
pictures containing the data points of the manifold at different position.
Step 3
Figure 101 Using laser thermometer as soon as the engine was started at the first pipe to the left.
P a g e | 100
The laser was used to determine the individual temperatures of the four pipes and the
collector. The reading as shown at Figure 101 is 206 degrees Fahrenheit engine at idle 800
revolutions per minute.
Figure 102 Corresponding image to support figure 101. Is notable the temperature path through
the manifold.
Figure 103 Reading the laser thermometer at the second pipe and one of the middle pair.
P a g e | 101
Temperature reading at idle with 800 revolutions per minutes is 197 Fahrenheit.
Figure 104 Thermal Imaging showing average temperature through the whole picture
Figure 105 Corresponding image to support figure 103. The behavior is compatible with the
image and the data point is very accurate as explained above the pipes at the center work are
cooler that the pipes at the side. A difference from 6-to-20 Fahrenheit is observed, due to the
different calibration of the equipment.
P a g e | 102
Figure 105 Reading to the right interior pipe and third of the four pipes in the prototype
Reading at idle 194.3 Fahrenheit.
Figure 106 Thermal Imaging showing the heat shield and temperature pattern
P a g e | 103
Figure 107 Reading at the fourth pipe with the laser thermometer.
Figure 108 Corresponding image to support figure 107
P a g e | 104
Figure 109 Reading at the collector center with laser.
In this step we note that the pipe number 1 and pipe number 4 the readings are very close
and higher temperatures as the pipe 2 and 3 due to friction of the fluid with the interior walls of
the prototype and length of the pipes also the shape is essential. All this theoretical data is
supported by the team previous studies and simulations.
Step 4
Figure 110 Highest reading with thermometer
P a g e | 105
Figure 110 shows a team member using the laser thermometer to measure the highest
temperature section of prototype after the transient period of idle at 800 revolutions per minuets.
The reading is 340 Fahrenheit, and is already visible the color changing at the highest
temperature sections in the prototype.
Figure 111 Corresponding image to support figure 110
Step 5
Figure 112 2000RPM
P a g e | 106
Figure 113 Physical aspect of the manifold at 2000 revolutions per minutes.
Figure 114 Manifold at 2000RPM ~368F
Figure 114 Corresponding image to support figure 113. The maximum temperature acquired at
these step is 368 Fahrenheit.
P a g e | 107
Figure 115 3000RPM
Figure 116 Physical aspect of the manifold at 3000 revolutions per minutes.
P a g e | 108
Figure 117 Cooling Off
Figure 117 Corresponding image to support figure 116. At 3000 revolutions per minutes the
team is able to see the temperature that the manifold has acquired . The maximum temperature of
this image is 447 Fahrenheit.
Figure 118 Temperature Pattern
P a g e | 109
Step 6
Figure 119 Testing the Stock Manifold
Figure 120 Header warms up a lot faster
Figure 120 shows Factory heaters temperature just seconds after start at the inner pipe.
P a g e | 110
Figure 121 Stock Manifold Heat Signature
Figure 121 supports the reading of figure 120 and it specify how fast this factory manifold is
dissipating heat to the surroundings and the magnitude of it. Within minutes this manifold is
dissipating a heat of 500 Fahrenheit. The maximum temperature that the team obtain during
testing of the prototype is less than the minimum time-temperature related obtain at the factory
one.
Figure 122 Shows a few seconds after getting to 3000RPM, the manifold reaches 655F,finally
reaching 706F after just a few seconds.
P a g e | 111
7.3 Test Results and Data
After an intensive session of flow analysis, our team recorded several times the pressure
difference in several points at the T3 turbo flange. In our first experiment we measured the
pressure difference in the output T3 turbo flange after inputting air through the four manifold
headers at the same time. This input of air was set to five different total pressures: 0.5, 0.75, 1.0,
1.25,and 1.5 inches of H2O by adjusting the current of the control motor that drives a centrifugal
fan. The instrument used to measure the output pressure difference was the Pitot tube which was
connected to a manometer instrument from which data was taken to later do the calculation of
the respective velocities. The air velocities were calculated by simplifying the Bernoulli equation
until we obtained the equation seen below:
√
(
)
The value of
is the difference between the total or stagnation pressure and the static
pressure that the manometer is reading in inches of H2O. This value read in inches of H2O is
converted to Pa in order to obtain the velocity of the air in m/s. The value of
represents the air
density value which is equal to 1.2041 kg/m3 at 20 °C. The air velocity values were calculated by
using Microsoft Excel software. The several figures that follow represent the data obtained from
the horizontal flow test at seven different points: a center point, three points to the left and three
points to the right. In figure 123 we can see the values obtained at point 0, which is the center
point.
P a g e | 112
Inlet TP (in of H2O)
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
Veloc at Inlet (m/s)
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
ΔP at PT 0 (in of H2O)
0.31
0.31
0.31
0.45
0.47
0.47
0.61
0.63
0.62
0.74
0.77
0.78
0.91
0.95
0.93
ΔP at PT 0 (Pa)
77.13
77.13
77.13
ΔP at PT 0 (Pa)
111.96
116.94
116.94
ΔP at PT 0 (Pa)
151.77
156.74
154.26
ΔP at PT 0 (Pa)
184.11
191.58
194.06
ΔP at PT 0 (Pa)
226.41
236.36
231.38
Veloc at PT 0 (m/s)
11.32
11.32
11.32
Veloc at PT 0 (m/s)
13.64
13.94
13.94
Veloc at PT 0 (m/s)
15.88
16.14
16.01
Veloc at PT 0 (m/s)
17.49
17.84
17.95
Veloc at PT 0 (m/s)
19.39
19.81
19.60
Figure 123 Data obtained from flow test at center point (PT 0)
Figure 124 shows the results obtained at point one left (PT 1L). As we can notice the results
obtained at point one left are very similar to those values obtained in the center point.
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
ΔP at PT 1L (in of H2O)
0.31
0.31
0.30
0.45
0.46
0.46
0.65
0.63
0.62
0.76
0.78
0.77
0.91
0.92
0.92
ΔP at PT 1L (Pa)
77.13
77.13
74.64
ΔP at PT 1L (Pa)
111.96
114.45
114.45
ΔP at PT 1L (Pa)
161.72
156.74
154.26
ΔP at PT 1L (Pa)
189.09
194.06
191.58
ΔP at PT 1L (Pa)
226.41
228.90
228.90
Figure 124 Data obtained from flow test at point one left (PT 1L)
Veloc at PT 1L (m/s)
11.32
11.32
11.13
13.64
13.79
13.79
16.39
16.14
16.01
17.72
17.95
17.84
19.39
19.50
19.50
P a g e | 113
In figure 125 we can see the data for point two left (PT 2L), which is a little bit more further
than point one left to the center point.
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.30
0.30
0.29
0.46
0.44
0.44
0.62
0.62
0.62
0.75
0.75
0.73
0.89
0.92
0.88
ΔP at PT 2L (Pa)
74.64
74.64
72.15
ΔP at PT 2L (Pa)
114.45
109.47
109.47
ΔP at PT 2L (Pa)
154.26
154.26
154.26
ΔP at PT 2L (Pa)
186.60
186.60
181.62
ΔP at PT 2L (Pa)
221.43
228.90
218.94
11.13
11.13
10.95
13.79
13.48
13.48
16.01
16.01
16.01
17.61
17.61
17.37
19.18
19.50
19.07
Figure 125 Data obtained from flow test at point two left (PT 2L)
The results for the furthest point from the center point is observed in figure 126. This point
was called point three left (PT 3L).
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.13
0.17
0.14
0.18
0.24
0.22
0.22
0.30
0.26
0.30
0.36
0.33
0.34
0.42
0.40
ΔP at PT 3L (Pa)
32.34
42.30
34.83
ΔP at PT 3L (Pa)
44.78
59.71
54.74
ΔP at PT 3L (Pa)
54.74
74.64
64.69
ΔP at PT 3L (Pa)
74.64
89.57
82.10
ΔP at PT 3L (Pa)
84.59
104.50
99.52
Figure 126 Data obtained from flow test at point three left (PT 3L)
7.33
8.38
7.61
8.62
9.96
9.53
9.53
11.13
10.37
11.13
12.20
11.68
11.85
13.17
12.86
P a g e | 114
Once the center point and the three points to the left of the center were analyzed it was time
to observe the behavior of the three points to the right of the center.
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
ΔP at PT 1R (in of H2O)
0.31
0.31
0.30
0.45
0.48
0.47
0.65
0.66
0.63
0.76
0.80
0.79
0.91
0.96
0.97
ΔP at PT 1R (Pa)
77.13
77.13
74.64
ΔP at PT 1R (Pa)
111.96
119.42
116.94
ΔP at PT 1R (Pa)
161.72
164.21
156.74
ΔP at PT 1R (Pa)
189.09
199.04
196.55
ΔP at PT 1R (Pa)
226.41
238.85
241.34
Veloc at PT 1R (m/s)
11.32
11.32
11.13
13.64
14.08
13.94
16.39
16.52
16.14
17.72
18.18
18.07
19.39
19.92
20.02
Figure 127 Data obtained from flow test at point one right (PT 1R)
In figure 128 we can observe results for point two right (PT 2R)
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.31
0.30
0.31
0.47
0.42
0.46
0.64
0.57
0.64
0.81
0.67
0.78
0.91
0.81
0.95
ΔP at PT 2R (Pa)
77.13
74.64
77.13
ΔP at PT 2R (Pa)
116.94
104.50
114.45
ΔP at PT 2R (Pa)
159.23
141.82
159.23
ΔP at PT 2R (Pa)
201.53
166.70
194.06
ΔPat PT 2R (Pa)
226.41
201.53
236.36
Figure 128 Data Obtained from Flow Test at point two right (PT 2R)
11.32
11.13
11.32
13.94
13.17
13.79
16.26
15.35
16.26
18.30
16.64
17.95
19.39
18.30
19.81
P a g e | 115
The last point analyzed in the horizontal flow test was point three right (PT 3R) and the results
are shown in figure 129
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.12
0.10
0.09
0.16
0.11
0.14
0.23
0.15
0.16
0.25
0.16
0.17
0.29
0.19
0.20
ΔP at PT 3R (Pa)
29.86
24.88
22.39
ΔP at PT 3R (Pa)
39.81
27.37
34.83
ΔP at PT 3R (Pa)
57.22
37.32
39.81
ΔP at PT 3R (Pa)
62.20
39.81
42.30
ΔP at PT 3R (Pa)
72.15
47.27
49.76
7.04
6.43
6.10
8.13
6.74
7.61
9.75
7.87
8.13
10.16
8.13
8.38
10.95
8.86
9.09
Figure 129 Data obtained from flow test at point three right (PT 3R)
The same method we applied horizontally was applied vertically with the exception that
instead of analyze seven points we analyzed five points this time due to the lack of vertical
space. The points analyzed were a center point, two points up and two points down. Again, input
parameters are represented in each figure as well as the results measured and the output
velocities calculated. The first point measured was the center point which we had already
measured previously in the horizontal flow test. We obtained very similar values as those
obtained in the previous study which indicate that the measurements are accurate. The result data
for the center point (PT 0) can be observed in figure 130.
P a g e | 116
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.31
0.31
0.31
0.47
0.47
0.47
0.63
0.63
0.63
0.76
0.76
0.76
0.93
0.93
0.93
ΔP at PT 0 (Pa)
77.13
77.13
77.13
ΔP at PT 0 (Pa)
116.94
116.94
116.94
ΔP at PT 0 (Pa)
156.74
156.74
156.74
ΔP at PT 0 (Pa)
189.09
189.09
189.09
ΔP at PT 0 (Pa)
231.38
231.38
231.38
Veloc at PT 0 (m/s)
11.32
11.32
11.32
Veloc at PT 0 (m/s)
13.94
13.94
13.94
Veloc at PT 0 (m/s)
16.14
16.14
16.14
Veloc at PT 0 (m/s)
17.72
17.72
17.72
Veloc at PT 0 (m/s)
19.60
19.60
19.60
Figure 130 Data obtained from flow test at center point (PT 0)
Now point one up (PT 1U) is measured and the results are shown in figure 131
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
ΔP at PT 1U (in of H2O)
0.33
0.33
0.33
0.53
0.50
0.51
0.69
0.70
0.70
0.83
0.82
0.85
0.99
1.00
1.01
ΔP at PT 1U (Pa)
82.10
82.10
82.10
ΔP at PT 1U (Pa)
131.86
124.40
126.89
ΔP at PT 1U (Pa)
171.67
174.16
174.16
ΔP at PT 1U (Pa)
206.50
204.02
211.48
ΔP at PT 1U (Pa)
246.31
248.80
251.29
Figure 131 Data obtained from flow test at point one up (PT 1U)
Veloc at PT 1U (m/s)
11.68
11.68
11.68
14.80
14.37
14.52
16.89
17.01
17.01
18.52
18.41
18.74
20.23
20.33
20.43
P a g e | 117
In figure 132 we can observe the measurement results for point two up (PT 2U)
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.36
0.34
0.36
0.54
0.56
0.54
0.75
0.75
0.76
0.91
0.91
0.92
1.07
1.07
1.08
ΔP at PT 2U (Pa)
89.57
84.59
89.57
ΔP at PT 2U (Pa)
134.35
139.33
134.35
ΔP at PT 2U (Pa)
186.60
186.60
189.09
ΔP at PT 2U (Pa)
226.41
226.41
228.90
ΔP at PT 2U (Pa)
266.22
266.22
268.70
12.20
11.85
12.20
14.94
15.21
14.94
17.61
17.61
17.72
19.39
19.39
19.50
21.03
21.03
21.13
Figure 132 Data obtained from flow test at point two up (PT 2U)
Now is the time to analyze the two points from the center moving down. The first point which is
point one down (PT 1D) can be seen in figure 133.
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
ΔP at PT 1D (in of H2O)
0.34
0.34
0.34
0.53
0.54
0.53
0.70
0.70
0.70
0.86
0.85
0.86
1.02
1.02
1.02
ΔP at PT 1D
84.59
84.59
84.59
ΔP at PT 1D
131.86
134.35
131.86
ΔP at PT 1D
174.16
174.16
174.16
ΔP at PT 1D
213.97
211.48
213.97
ΔP at PT 1D
253.78
253.78
253.78
Figure 133 Data obtained from flow test at point one down (PT 1D)
(Pa) Veloc at PT 1D
11.85
11.85
11.85
(Pa) Veloc at PT 1D
14.80
14.94
14.80
(Pa) Veloc at PT 1D
17.01
17.01
17.01
(Pa) Veloc at PT 1D
18.85
18.74
18.85
(Pa) Veloc at PT 1D
20.53
20.53
20.53
(m/s)
(m/s)
(m/s)
(m/s)
(m/s)
P a g e | 118
The last point that we analyzed for the vertical test was the point two down (PT 2D) and can be
observed in figure 134:
0.50
0.50
0.50
0.75
0.75
0.75
1.00
1.00
1.00
1.25
1.25
1.25
1.50
1.50
1.50
Inlet TP (Pa)
124.40
124.40
124.40
Inlet TP (Pa)
186.60
186.60
186.60
Inlet TP (Pa)
248.80
248.80
248.80
Inlet TP (Pa)
311.00
311.00
311.00
Inlet TP (Pa)
373.20
373.20
373.20
14.37
14.37
14.37
17.61
17.61
17.61
20.33
20.33
20.33
22.73
22.73
22.73
24.90
24.90
24.90
Amps
7.00
7.00
7.00
Amps
7.25
7.25
7.25
Amps
7.50
7.50
7.50
Amps
7.75
7.75
7.75
Amps
8.25
8.25
8.25
RPM
500.00
500.00
500.00
RPM
650.00
650.00
650.00
RPM
700.00
700.00
700.00
RPM
750.00
750.00
750.00
RPM
800.00
800.00
800.00
0.34
0.35
0.34
0.53
0.54
0.52
0.74
0.72
0.72
0.90
0.91
0.90
1.05
1.03
1.04
ΔP at PT 2D (Pa)
84.59
87.08
84.59
ΔP at PT 2D (Pa)
131.86
134.35
129.38
ΔP at PT 2D (Pa)
184.11
179.14
179.14
ΔP at PT 2D (Pa)
223.92
226.41
223.92
ΔP at PT 2D (Pa)
261.24
256.26
258.75
Veloc at PT 2D (m/s)
11.85
12.03
11.85
14.80
14.94
14.66
17.49
17.25
17.25
19.29
19.39
19.29
20.83
20.63
20.73
Figure 134 Data obtained from flow test at point two down (PT 2D)
After this experiment we can notice from the different values obtained that as we increase the
current of the control motor that drives the centrifugal fan so does the output velocity at the point
measured. For example in figure 134 which describes the results obtained for point two down
(PT 2D), we can notice that for a current equal to 7 Amps, 500 RPM and a Total Pressure (TP)
equal to 0.5 inches of H2O we obtained approximately 12.03 m/s of output velocity, while for a
current of 8.25 Amps, 800 RPM and 1.5 inches of H2O of Total Pressure (TP) we obtained an
output velocity of approximately 20.63 m/s at the same point measured. These values are in
correspondence to what the team expected before conducted the experiment. Also from the
horizontal flow analysis we can observe that the highest values of velocity are obtained at the
center point, point one and two left, as well as point one and two right. There is a drop in output
velocity in points one left and one right for all the cases studied. The vertical flow analysis show
P a g e | 119
different results compared to the horizontal analysis. In this case all points studied such as :
center point, point one and two up, as well as point one and two down, show similar output
velocities.
Another experiment that the team conducted was the individual flow analysis. In this case air
was applied to each individual header and the output pressure difference in the center point was
measured and the velocity was calculated. The parameters used for each individual header were a
current equal to 8.25 Amps, 800 RPM and a Total Pressure (TP) equal to 1.5 inches of H2O. The
results obtained are shown in figure135.
Header 1
Header 2
Header 3
Header 4
Inlet TP (in of H2O) Inlet TP (Pa) Veloc at Inlet (m/s)
1.50
373.20
24.90
1.50
373.20
24.90
1.50
373.20
24.90
1.50
373.20
24.90
Amps
8.25
8.25
8.25
8.25
RPM ΔP at PT 0 (in of H2O) ΔP at PT 0 (Pa) Veloc at PT 0 (m/s)
800.00
0.11
26.12
6.59
800.00
0.15
37.32
7.87
800.00
0.15
37.32
7.87
800.00
0.11
26.12
6.59
Figure 135 Data obtained from flow test at each individual header
As we can see from figure 136 the velocity at each individual header was calculated at the
center point, and as the team expected, header two and three which represent the middle section
obtained the highest output velocity since these sections predicted the lowest loss rate due to the
fact that the angular inclination is smaller. Also the values of pressure difference that the
manometer read for header one and header four were the same, such is the case for headers two
and three, this a good result which indicates that the assembly is completely symmetrical and the
losses will be the same at both sides. Since the difference in pressure between header two and
header one is not that much we can also assume that such a small loss in the output velocity is
due to the angular inclination difference and not because of air going into another pipe.
P a g e | 120
Uncertainty Analysis
Uncertainty Analysis or Error Analysis is the process in which the uncertainties in the
measurement are determined. Generally measurements vary from one measurement to the next
and statics can be used to analyze such variations. The average X, and Standard Deviation Sx,
can be calculated using the following formulas:
∑
;
= √[
∑
(
) ]
The measurement uncertainty can be determined for two types of sample sizes: the large
sample size (N > 60), and the small sample size (N ≤ 60), where N represents the number of
measurements. In our experiment we used N = 3, so the small sample size will be applied. For
the small sample size, the uncertainty increases due to the limited number of samples available
for the statistics. In this case the Student-t distribution is used to determine the uncertainty from
the standard deviation Sx. Table 10 shows the values of Student-t distribution for finite sample
sizes.
Table 10 Values of Student-t distribution for finite sample sizes
V
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
t50
1
0.816
0.765
0.741
0.727
0.718
0.711
0.706
0.703
0.7
0.697
0.695
0.694
0.692
0.691
t90
6.314
2.92
2.353
2.132
2.015
1.943
1.895
1.86
1.833
1.812
1.796
1.782
1.771
1.761
1.753
t95
12.706
4.303
3.182
2.77
2.571
2.447
2.365
2.306
2.262
2.228
2.201
2.179
2.16
2.145
2.131
t99
63.657
9.925
5.841
4.604
4.032
3.707
3.499
3.355
3.25
3.169
3.106
3.55
3.012
2.977
2.947
P a g e | 121
16
17
18
19
20
21
30
40
50
60
∞
0.69
0.689
0.688
0.688
0.687
0.686
0.683
0.681
0.68
0.679
0.674
Once the corresponding value of
1.746
1.74
1.734
1.729
1.725
1.721
1.697
1.684
1.679
1.671
1.645
2.12
2.11
2.101
2093
2.086
2.08
2.042
2.021
2.01
2
1.96
2.921
2.898
2.878
2.861
2.845
2.831
2.75
2.704
2.679
2.66
2.576
is obtained from the Student-t distribution table, by
intersecting the value of N (number of measurements) and P (confidence level desired) the
uncertainty value is obtained by using the following formula:
In our case we are going to determine the measurement uncertainty in the center point which
is the point in which we obtained one of the highest velocities. The confidence interval used was
95%, and from the Student-t distribution table we obtained a value of
equal to 3.182. In
figure 137 we can find the velocities calculated at the center point, average velocity, standard
deviation, uncertainty and range of uncertainty. As we can see in figure 137 the range of
uncertainty is obtained for all cases studied. These values represent the range in which the
velocity will lie 95% of the time. For the first total input pressure which is 0.5 inches of H 2O the
uncertainty is 0 which means that there is no range of uncertainty since all measurements
collected were the same. For a total input pressure equal to 0.75 inches of H2O the velocity will
lie between 14.39 m/s and 13.29 m/s. In the case of the total input pressure equal to 1.0 inches of
H2O the measurements will result in a range of values between 16.42 m/s and 15.6 m/s. For a
total input pressure equal to 1.25 inches of H2O the maximum range of uncertainty is obtained
P a g e | 122
which lies between 18.53 m/s and 16.99 m/s. Finally for a total input pressure equal to 1.5 inches
of H2O we get velocity that lies between 20.27 m/s and 18.93 m/s in 95% of the time.
Veloc at PT 0 (m/s)
11.32
11.32
11.32
Veloc at PT 0 (m/s)
13.64
13.94
13.94
Veloc at PT 0 (m/s)
15.88
16.14
16.01
Veloc at PT 0 (m/s)
17.49
17.84
17.95
Veloc at PT 0 (m/s)
19.39
19.81
19.60
Average Veloc (X)
Standard Deviation (Sx) Uncertainty (Ux,mean)
Range of Uncertainty
11.32
Average Veloc (X)
0
0
Standard Deviation (X) Uncertainty (Ux,mean)
V = 11.32 ± 0
Uncertainty Range
13.84
Average Veloc (X)
0.17
0.55
V = 13.84 ± 0.55
Uncertainty Range
16.01
Average Veloc (X)
0.13
0.41
V = 16.01 ± 0.41
Uncertainty Range
17.76
Average Veloc (X)
0.24
0.77
V = 17.76 ± 0.77
Uncertainty Range
19.60
0.21
0.67
V = 19.60 ± 0.67
Figure 136 Uncertainty Analysis
7.3 Evaluation of Experimental Results
SolidWorks vs. Real Life Test
Table 11 Real life test results for inlet TP equal to 124.4 Pa at point 0
Inlet TP (in of Inlet
H2O)
(Pa)
0.5
124.4
TP Am
ps
7
RP
M
500
ΔP at PT 0 (in of ΔP at PT 0 Veloc at PT 0
H2O)
(Pa)
(m/s)
0.31
77.128
11.31852292
P a g e | 123
Figure 137 Pressure 1
Figure 138 Several Points at Pressure 1
P a g e | 124
H2O)
(Pa)
0.75
186.6
TP Am
ps
7.25
RP
M
650
H2O)
(Pa)
(m/s)
0.45
111.96
13.63688457
Figure 139 Points at Pressure 2
P a g e | 125
H2O)
(Pa)
1
248.8
TP Am
ps
7.5
RP
M
700
H2O)
(Pa)
(m/s)
0.61
151.768
15.8771967
P a g e | 126
Table 14 Real life test results for inlet TP equal to 311 Pa at point 0
H2O)
(Pa)
1.25
311
TP Am
ps
7.75
RP
M
750
H2O)
(Pa)
(m/s)
0.74
184.112
17.48738081
P a g e | 127
Table 15 Real life test results for TP equal to 373.2 Pa at point 0
H2O)
(Pa)
1.5
373.2
TP Am
ps
8.25
RP
M
750
H2O)
(Pa)
(m/s)
0.91
226.408
19.39231262
P a g e | 128
Several tests were performed in the manifold in order to obtain a good picture of how the
manifold was behaving in real life and how did the extensive CFD and FEA simulations that we
had made translated into real world. The team had attempted to build a manifold as close as
possible to the outlined model in SolidWorks, in order to get test results as close as possible but
also to create the best flowing manifold. It is important to notice that during the design process
the team noticed that even a difference of two degrees in the right junction could create flow
differences for the whole manifold, therefore making the prototyping process time consuming
and necessarily precise.
The first and most important test made by the team was the flow test. Since the school
lacked a flow bench, the team used another available machine and converted it to the equivalent
of a flow bench. The team measured pressured at both entrance and exit and determined flow.
By using a pitot tube and SolidWorks; the two results were compared.
The results were amazingly close and the errors were so low the whole team was
surprised. The team had achieved to reproduce the manifold within a 3% error from the model
with very basic tooling and instrumentation. This result meant that the model had been built
almost exactly like the virtual model and that the error the team was getting was mostly due to
measurement error and pitot tube placement. The team did place the pitot tube in specific
location but specially because of the nature of the pitot, and the fact that it measures spatially
pressure but without being specific as far as location, some of the error was generated. Also the
box used to create a semi equal pattern of flow thru all the runners might have had some type of
leak or even a small turbulence pattern due to the tape that we couldn’t account for.
P a g e | 129
The percent of error is very low as this table shows, the numbers obtained by the team at the
different prove the manifold is working well according to spec:
Table 16 Error percentage analysis
Inlet TP (Pa)
124.4
Inlet TP (Pa)
186.6
Inlet TP (Pa)
248.8
Inlet TP (Pa)
311
Inlet TP (Pa)
373.2
Veloc at PT 0 (m/s)
11.31852292
Veloc at PT 0 (m/s)
13.63688457
Veloc at PT 0 (m/s)
15.8771967
Veloc at PT 0 (m/s)
17.48738081
Veloc at PT 0 (m/s)
19.39231262
FlowWorks
Value
11
Closest
Error %
2.90
13.3
2.53
15.5
2.43
17.5
-0.07
19.4
-0.04
7.4 Improvement of design
The design was improved several times before the construction of the prototype. As
explained above the proposed design was the best of all the design alternatives; furthermore, this
proposed design was optimized to its top potential. Once all the testing was done at the prototype
the team realized that stick welding is to dirty for this job, and for a more efficient design it will
be more clean and precise to use TIG welding. Another improvement that will reduce hours of
work will be the way the team weld the exhaust; this part of the building process need to be
reconsidered. This improvement does not affect the data, but it does affect time and money
invested to get a perfect seal necessary in the part.
P a g e | 130
7.5 Discussion
The testing overall was a complete success; not only did the team achieve to obtain data
that reflects a very good manifold design, but also the team passed and exceeded any metric or
expectation originally set out at the beginning of this project. Most importantly, the manifold was
able to be tested right on a vehicle, which the team was able to pull at the last minute after weeks
waiting for a test. It was a great sight to be able to test the manifold in a working motor that
produced temperatures that would have cracked an otherwise weak manifold. The manifold even
changed color to a yellowish silver, which seemed to be an indication of the manifold achieving
a level of maturity beyond the first stage. In all the testing that we performed the results achieved
were in accord with SolidWorks Thermal and FlowWorks analysis, which also highlighted the
fact that the model with had created was performing with a very small degree of deviation from
the model with had virtual designed.
Our testing methods also proved to be in the level of proficiency that allowed the team to
obtain data within five percent of error from most of the calculations. Most of the testing we had
done proved to be logical and straightforward, which help to keep operator error and
instrumentation error at minimum.
P a g e | 131
Chapter 8. Design Considerations
8.1 Sustainability Report
Figure 147 Location of Production and Distribution
Figure 148 Definition of Use of Design
P a g e | 132
Figure 149 Life Cycle of a Product
P a g e | 133
Figure 150 Pie Charts and Environmental Impact
Air Acidification - Sulfur dioxide, nitrous oxides and other acidic emissions to air cause
an increase in the acidity of rainwater, which in turn acidifies lakes and soil. These acids can
make the land and water toxic for plants and aquatic life. Acid rain can also slowly dissolve
manmade building materials such as concrete. This impact is typically measured in units of
P a g e | 134
either kg sulfur dioxide equivalent (SO2), or moles H+ equivalent. Carbon Footprint - Carbondioxide and other gasses which result from the burning of fossil fuels accumulate in the
atmosphere which in turn increases the earth’s average temperature. Carbon footprint acts as a
proxy for the larger impact factor referred to as Global Warming Potential (GWP). Global
warming is blamed for problems like loss of glaciers, extinction of species, and more extreme
weather, among others. Total Energy Consumed - A measure of the non-renewable energy
sources associated with the part’s lifecycle in units of mega joules (MJ). This impact includes
not only the electricity or fuels used during the product’s lifecycle, but also the upstream energy
required to obtain and process these fuels, and the embodied energy of materials which would be
released if burned. Total Energy Consumed is expressed as the net calorific value of energy
demand from non-renewable resources (e.g. petroleum, natural gas, etc.). Efficiencies in energy
conversion (e.g. power, heat, steam, etc.) are taken into account. Water Eutrophication - When
an overabundance of nutrients are added to a water ecosystem, eutrophication occurs. Nitrogen
and phosphorous from waste water and agricultural fertilizers causes an overabundance of algae
to bloom, which then depletes the water of oxygen and results in the death of both plant and
animal life. This impact is typically measured in either kg phosphate equivalent (PO4) or kg
nitrogen (N) equivalent. Life Cycle Assessment (LCA) - This is a method to quantitatively assess
the environmental impact of a product throughout its entire lifecycle, from the procurement of
the raw materials, through the production, distribution, use, disposal and recycling of that
product.
P a g e | 135
Chapter 9. Conclusion
9.1 Conclusion and Discussion
After an extensive work the team was able to successfully complete the Senior Design
requirements in spite of the obstacles found along the way. The team presented five different
design alternatives that were analyzed in terms of flow distribution using standard parameters in
SolidWorks. The proposed design was selected by comparing each proposed design and
analyzing the advantages and disadvantages of the following factors: flow speed, manufacturing
feasibility, and material selection. In this stage of the project the team also optimized the
proposed design by examining the different angular configurations possible and selecting the
configuration with the highest flow speed. A comparison of flow distribution between the
optimized proposed design and a very frequently used exhaust manifold model was made and the
results show a 16.6% increase in the flow velocity for the proposed design. As mention
previously the team stuck to our project management plan which includes the breakdown of
work into specific tasks, timeline for completion and responsibilities among team members. This
section allows us to complete each individual task on time and manage our time more efficiently.
Several studies were conducted to the exhaust manifold, such as: thermal, static, cyclic, transient
thermal, transient flow, as well as flow and thermal. Out of these studies we conclude that the
design was safe enough to hold the load applied with a factor of safety of 28. We also obtained
high values of flow speed and a very nice flow distribution along the entire assembly. The loss in
the collector was also calculated by running a flow simulation of the proposed design with
collector and without it. It turns out that the loss is only 11% meaning that the collector was
design in such a way that the loss is as minimum as possible. Once we had all the parts listed the
process of constructing the prototype started. In this stage the teams encountered several
P a g e | 136
obstacles especially with the tools in the students machine shop which were in extremely critical
conditions, which delayed the whole manufacturing process requiring extra time of work to do
the drilling and cutting with the precision required. After approximately a week of work in the
machine shop the assembly was ready to be welded. The welding process was completed in one
day with the help of a professional welder and the team used stick welding as the preferred
welding method since it is the cheapest and simplest method. The quality of the welding was
analyzed latter on while doing the testing and evaluation to make sure we a good penetration of
the welding material was obtained. It turns out that the assembly presented several spots with
leaks; this was verified by conducting a pressurization test by using one of the lab resources at
Baker Aviation School in Miami. Once the leaks were corrected, another welding test was made
by using a red penetration spray to verify the quality of the welding. The results show a few
pores but nothing that indicates the presence of any leak. A flow analysis was also made by the
team in the transport phenomenon lab at the Engineering Center in FIU. The results obtained
from the flow experiments compared to the flow simulation in Solid Works show very small
error percentages, the maximum error obtained was 2.9% and the minimum 0.04% for the all the
cases compared. These small values in error percentages indicate the accuracy of the simulation
studies analyzed by the team throughout this project. The last test the team realized was the
thermal testing in which we mounted the exhaust manifold in a Miata 2001 model. This test was
made by using and infrared camera and a laser thermometer. With these two instruments we
measured maximum temperatures at 800, 2000 and 5000 rpm in the exhaust manifold. As a
result the maximum temperature recorded was 380 °F at 3000 rpm. The original manifold
mounted on the Miata was also tested and the maximum temperature recorded was 700 °F at
3000rpm. This huge difference in temperature represents an 84% increase in temperature
P a g e | 137
meaning that the original manifold will heat up the bay of the engine which is a major
disadvantage. As the temperature in the surface of our proposed design is much smaller than the
one in the original manifold, the heat transfer to the bay of the engine is minimum using the
proposed design. In general our proposed design offers a unique design with the maximum flow
speed possible, made out of stainless steel 304 schedule 40 that provide minimum heat transfer to
the engine bay and enough resistance to support the loads. Several test comparing our design to
designs already in the market show the advantages in terms of quality and durability that out
product offers, and how the flow distribution contributes to reduce parameters such as back
pressure increasing the efficiency of our proposed design.
9.2 Commercialization of the Manifold
There used to be a saying before the Great Depression among entrepreneurs that claimed
“Supply will always create its own demand”; a few years after nothing was far from the truth and
the country’s economy collapsed. It wasn’t until a few bright minds claimed the necessity of
desirable demand that companies and countries alike understood that behind any successful
product there must be a demand for it, in other words; just because you build it doesn’t guarantee
it will sell.
In our case though, the team has picked a product that although part of a niche market,
the product has sustainable and consistent demand such that if a mechanism of mass production
were to be implement, a single manifold like this custom log manifold may keep a company
afloat. That is not to say of course that the experience and design techniques learned while
producing this manifold cannot be used to produce manifolds for any car.
P a g e | 138
It seems like a natural endeavor to begin our quest to recognition and therefore sharing a
piece of the market by the forums. Many small companies start this way and once a name has
been made for then, it is very easy to move into private websites and even mainstream
commercial websites like EBAY.
Figure 151 MiataTurbo.net
A forum like the ones shown above is the perfect way to get into the market. Once the
account has been set up and the introductory fees have been paid, the team can move up to the
VENDOR status and the Vendor Classifieds.
P a g e | 139
Figure 152 Vendor Forum
From there on, besides being part of a very active community, the race is on to conquer the
hearts and pockets of those savvy Miata owners. Something to mention and to make note of, it is
the type of clientele available at the forums. Not only is the forums filled with enthusiasts but
also enthusiasts specific to the product the team fabricates. The level of exposure is also
tremendous. Notice the numbers for activity:
P a g e | 140
Figure 153 Usage Info
After a name has been founded in the forum community the horizons opens to places like
EBay where the whole worlds has access to our products.
9.3 Future Work
The future holds great challenges in the form of new designs and new ideas that bring
together higher efficiency and lower cost. Not every car is as special as the Miata and therefore
the possibilities have never been as great. Something is clear, with more time and more capable
machinery, the same exact manifold that has already been designed could be built with even
tighter tolerances and better characteristics. With even more powerful CFD and with a
workstation capable of handling the load, even more exact models could be made. The future
brings progress, and progress is unstoppable.
P a g e | 141
References
1. Ashby, M. F and D. R. H. Jones, Engineering Materials 1, An Introduction to Their
Properties and Applications. 3rd Edition. ButterworthHeinemann, Woburn, UK, 2005.
2. Clayton T. Crowe, Donald F. Elger, John A. Roberson. Engineering Fluid Mechanics. 8 th
Edition. 2005.
3. Faye C. McQuiston, Jerald D. Parker, Jeffrey D. Spitler. . Heating, Ventilating and Air
Conditioning Analysis and Design. 5th Edition. 2000.
4. Faye C. McQuiston, Jerald D. Parker, Jeffrey D. Spitler. Heating, Ventilating and Air
Conditioning Analysis and Design. 6th Edition. 2005.
5. Ferdinand P. Beer, E. Russell Johnston, Jr, John T. DeWolf. Mechanics of Materials. 4 th
Edition. 2006.
6. Frank P. Incropera, David P. Dewitt, Theodore L. Bergman, Adriene S. Lavine. Introduction
to Heat Transfer. 5th Edition. 2007.
7. Michael J. Moran, Howard N. Shapiro. Fundamentals of Engineering Thermodynamics. 6 th
Edition. 2008.
8. Jay Storer. Sport Compact Performance. Sparkford Nr Yeovil Somerset BA22 7JJ England.
2003.
9. Philip H. Smith. The scientific design of exhaust and intake systems. 3rd Edition.
Cambridge, Massachusetts, 1972.
P a g e | 142
10. Robert Maddox, John H Haynes. Automotive Repair Manual. Sparkford Nr Yeovil
Somerset BA22 7JJ England. 2005.
11. Shigley’s. Mechanical Engineering Design. 8th Edition. New York, 2008.
12. William E. Howard, Joseph C. Musto. Introduction to Solid Modeling Using SolidWorks
2008. 4th Edition. New York, NY. 2009.
13. William K. Toboldt, Larry Johnson, Steven W. Olive. Automotive Encyclopedia,
Fundamental Principles, Operation, Construction, Service, Repair. South Holland, Illinois.
14. "Google Images." Google. Web. Aug 23, 2011.
http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?VISuperSize&item=180711613654
15. "Google Images." Google. Web. Feb 18, 2011.
http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?VISuperSize&item=190500476870
16. http://www.mitchell1.com
17. http://www.alldata.com
18. http://roymech.com
19. http://gowelding.com
20. 2001 Mazda Miata. 1.8L Motor with VVT, coupled with a manual transmission.
P a g e | 143
Appendix A, Engineering drawing sheets
Figure 154 Front Flange
P a g e | 144
Figure 155 Back flange
P a g e | 145
Appendix B, Receipts/Invoices
Figure 156 Invoice of purchase made at P&M Supply Inc.
P a g e | 146
Figure 157 Invoice of purchase made at P&M Supply Inc.
.
P a g e | 147
Figure 158 Invoice
P a g e | 148
Figure 159 Invoice
P a g e | 149
Figure 160 Invoice
P a g e | 150
Figure 161 Invoice
P a g e | 151
Appendix C, Pictures
Figure 162 Compressor Used

High performance Turbo Exhaust Manifold, Elmer Gutierrez, Michael

Transcription

Similar documents

MANIFOLD FITMENT MG used a number of different inlet manifold

mortu ary operational e quipment - Worldwide Disaster Response

Kernel methods on Riemannian manifolds

KENNE BELL MAMMOTH TWIN SCREW vs TVS 2.3 KITS* KENNE

Product Guide

MG100395 - BANCO DI COLLAUDO PER MANIFOLD E SISTEMI

george mangoletsi`s gm manifold modifier

Composite Manifold

QHPM - Ferguson Chicago