Page 1
2
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
Centrifugal Compressor Optimization
Final Report
Xavier Medina
Gianfranco Pisani
Fernando Lopez
Advisor: Professor George Dulikravich
November 17, 2014
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.
Page 2
ii
Ethics Statement and Signatures
The work submitted in this B.S thesis is solely prepared by a team consisting of Xavier
Medina, Fernando Lopez, and Gianfranco Pisani, 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.
Page 3
iii
Table of Contents Ethics Statement and Signatures ..................................................................................................... ii
List of Figures ................................................................................................................................ vi
List of Tables ............................................................................................................................... viii
Abstract ........................................................................................................................................... 2
1. Introduction ............................................................................................................................. 3
1.1 Problem Statement ........................................................................................................... 3
1.2 Motivation ........................................................................................................................ 3
1.3 Literature Survey .............................................................................................................. 4
1.4 Survey of Related Standards ............................................................................................ 4
1.5 Discussion ........................................................................................................................ 5
2. Project Formulation ................................................................................................................. 6
Overview .......................................................................................................................... 6
Project Objectives ............................................................................................................ 6
Design Specifications ....................................................................................................... 7
Addressing Global Design ............................................................................................... 8
Constraints and Other Considerations .............................................................................. 9
Discussion ...................................................................................................................... 10
3. Design Alternatives ............................................................................................................... 11
3.1 Overview of Conceptual Designs Developed ................................................................ 11
3.2 Optimized Design Alternatives ...................................................................................... 11
3.2.1 ModeFRONTIER® Algorithms ............................................................................. 13
3.2.2 Graphs and data validation...................................................................................... 15
3.3 Design Alternative.......................................................................................................... 20
3.4 Integration of Global Design Elements .......................................................................... 20
3.5 Feasibility Assessment ................................................................................................... 21
3.6 Proposed Design ............................................................................................................. 21
3.7 Discussion ...................................................................................................................... 22
4. Project Management .............................................................................................................. 23
4.1 Overview ........................................................................................................................ 23
4.2 Breakdown of Work into Specific Tasks ....................................................................... 23
4.3 Gantt Chart for the Organization of Work and Timeline ............................................... 24
4.4 Breakdown of Responsibilities Among Team Members ............................................... 24
Page 4
iv
4.5 Commercialization of the Final Product ........................................................................ 25
4.6 Discussion ...................................................................................................................... 25
5. Engineering Design and Analysis .......................................................................................... 26
5.1 Overview ........................................................................................................................ 26
5.2 Kinematic Analysis and Animation ............................................................................... 26
5.3 Dynamic/Vibration Analysis of the System ................................................................... 27
5.4 Structural Design ............................................................................................................ 28
5.5 Force Analysis ................................................................................................................ 30
5.6 Stress Analysis ............................................................................................................... 31
5.7 Material Selection .......................................................................................................... 32
5.8 Design Based on Static and Fatigue Failure Design Theories ....................................... 33
5.9 Deflection Analysis ........................................................................................................ 34
5.10 Component Design/Selection ......................................................................................... 35
5.11 Component Design Simulation ...................................................................................... 36
5.12 Finite Element Analysis ................................................................................................. 37
5.13 Design Overview ............................................................................................................ 38
5.14 Cost Analysis.................................................................................................................. 38
5.15 Discussion ...................................................................................................................... 39
6. Prototype Construction .......................................................................................................... 40
6.1 Overview ........................................................................................................................ 40
6.2 Description of Prototype ................................................................................................ 40
6.3 Prototype Design ............................................................................................................ 41
6.4 Parts List ......................................................................................................................... 42
6.5 Construction ................................................................................................................... 42
6.6 Prototype Cost Analysis ................................................................................................. 44
6.7 Discussion ...................................................................................................................... 44
7. Testing and Evaluation .......................................................................................................... 46
7.1 Overview ........................................................................................................................ 46
7.2 Design of Experiments - Description of Experiments ................................................... 47
7.2.1 Experiment 1 – Average Miles per gallon .............................................................. 47
7.2.2 Experiment 2 – Dynamometer Testing ................................................................... 48
7.2.3 Experiment 3 – Compressor Test Bench ................................................................ 49
7.3 Test Results and Data ..................................................................................................... 50
7.4 Evaluation of Experimental Results ............................................................................... 51
Page 5
v
7.5 Improvement of the Design ............................................................................................ 51
7.6 Discussion ...................................................................................................................... 52
8. Design Considerations ........................................................................................................... 53
8.1 Assembly and Disassembly ............................................................................................ 53
8.2 Maintenance of the System ............................................................................................ 53
8.2.1 Regular Maintenance .................................................................................................. 53
8.3 Environmental Impact .................................................................................................... 54
8.4 Risk Assessment ............................................................................................................. 54
9. Design Experience ................................................................................................................. 56
9.1 Overview ........................................................................................................................ 56
9.2 Standards Used in the Project......................................................................................... 56
9.3 Contemporary Issues ...................................................................................................... 56
9.4 Impact of Design in a Global and Societal Context ....................................................... 57
9.5 Professional and Ethical Responsibility ......................................................................... 57
9.6 Life-Long Learning Experience ..................................................................................... 57
9.7 Discussion ...................................................................................................................... 58
10. Conclusion ......................................................................................................................... 59
10.1 Conclusion and Discussion ............................................................................................ 59
10.2 Evaluation of Integrated Global Design Aspects ........................................................... 59
10.3 Evaluation of Intangible Experiences ............................................................................ 59
10.4 Commercialization Prospects of the Product ................................................................. 60
10.5 Future Work ................................................................................................................... 60
References ..................................................................................................................................... 62
Appendices .................................................................................................................................... 64
A. Detailed Engineering Drawings of All Parts, Subsystems and Assemblies ................... 64
B. Copies of Used Commercial Machine Element Catalogs .............................................. 65
C. Detailed Raw Design Calculations and Analysis ........................................................... 66
D. Project Photo Album ...................................................................................................... 69
Page 6
vi
List of Figures
Figure 1. Breakaway commencing at the aft edge of the shock wave [1] ___________________________________ 5
Figure 2. Shrouded Impeller. ____________________________________________________________________ 1
Figure 3. Unshrouded impeller. These impellers also tend to have complex 3-D shaped blades. ________________ 1
Figure 4. Two-dimensional impeller blades. ________________________________________________________ 1
Figure 5. Zones on compressor map of turbocharger that are most important from design/research and
development standpoints [4]. ____________________________________________________________________ 2
Figure 6. Illustration of the inducer and exducer of compressor wheel [9]. ________________________________ 3
Figure 7. Impeller removed from a Mitsubishi TD04 turbocharger. ______________________________________ 7
Figure 8. TD04-13T compressor housing. __________________________________________________________ 7
Figure 9. 3D sketch guideline for impeller blade. To the right are the eight parameters which will be altered. ____ 12
Figure 10. Explanation of how evolutionary algorithms work just as selective breeding. _____________________ 14
Figure 11. Above shows eight parameters and desired goal objectives. __________________________________ 15
Figure 12. 3D Graph of angle and initial angle with respect to the average velocity. _______________________ 16
Figure 13. Plot of Average Velocity vs Initial Angle _________________________________________________ 17
Figure 14. Plot of Avergae Velocity and Angle _____________________________________________________ 17
Figure 15. Design Generation of Angle ___________________________________________________________ 18
Figure 16. Design Population of Initial Angle ______________________________________________________ 19
Figure 17. Final Solidworks CFD Study __________________________________________________________ 19
Figure 18. Gantt chart. _______________________________________________________________________ 24
Figure 19. Frequency simulation performed to assess vibration. _______________________________________ 28
Figure 20. Where the rounded edges needed to be made. _____________________________________________ 29
Figure 21. Life of the blade after going through 𝟏𝟎𝟕cycles. ___________________________________________ 29
Figure 22. Damage that is present after 𝟏𝟎𝟕cycles. _________________________________________________ 30
Figure 23. Direction of rotation and the forces that are felt. ___________________________________________ 31
Figure 24. Image showing areas where stress is experienced. __________________________________________ 32
Figure 25. Where failure would occur by dropping below a certain Factor Of Safety. _______________________ 34
Figure 26. Deflection could occur in the red areas. __________________________________________________ 35
Page 7
vii
Figure 27. 3D printed prototype done by Ryan Lucia. ________________________________________________ 41
Figure 28. Polyjet 3D printed acrylic of final design from WB Engineering. ______________________________ 43
Figure 29. Visual example of a turbocharger system undergoing fatigue and performance tests. ______________ 46
Figure 30. Air compressor that will blow air through turbine side of turbocharger. _________________________ 50
Figure 31. Computer that is directly connected with air compressor to get all readings. _____________________ 50
Figure 32. Optimized wheel avg velovity vs. Stock wheel avg velocity. ___________________________________ 51
Figure 33. Wastegate Actuator. _________________________________________________________________ 54
Figure 34. Turbocharger clamp holding the turbine housing with the cartridge. ___________________________ 54
Figure 35. Coolant lines with 17mm banjo bolts and 4 copper washers. __________________________________ 55
Figure 36. Oil return line connected to the CHRA. __________________________________________________ 55
Figure 37. Destroyed turbine blade from Subaru turbo. [15] __________________________________________ 54
Figure 38. Technical drawing for concept design. ___________________________________________________ 64
Figure 39. Mitsubishi TD04-13T Turbo Rebuild. ____________________________________________________ 65
Figure 40. Adaptation of Bernoulli’s Equation. _____________________________________________________ 66
Figure 41. Compressor wheel from used turbo. _____________________________________________________ 69
Figure 42. Top view of 3D printed acrylic. ________________________________________________________ 70
Figure 43. Stock compressor wheel alongside 3D printed ABS plastic final design. _________________________ 70
Figure 44. Rendering of final design. _____________________________________________________________ 71
Figure 45. Used compressor wheel, 3D printed ABS plastic, 3D printed polyjet acrylic. _____________________ 71
Figure 46. Clearance between final design and housing.______________________________________________ 72
Figure 47. Measurement with L ruler. ____________________________________________________________ 72
Figure 48. Isometric view of final design in 3D printed acrylic. ________________________________________ 73
Figure 49. First clearance test with rough 3D printed ABS plastic. _____________________________________ 73
Figure 50. First clearance test. _________________________________________________________________ 74
Page 8
viii
List of Tables
Table 1. Average required fuel economy (mpg) under final CAFÉ standards. ............................................................. 8
Table 2. Table of responsibilities. ................................................................................................................................ 23
Table 3. Tabulated Cost Analysis ................................................................................................................................ 39
Table 4. Southern Gear final estimated quote price. ................................................................................................... 44
Table 5. Incodema3D final estimated quote price. ...................................................................................................... 44
Table 6. 1click3Dprint final estimate quote price. ...................................................................................................... 44
Table 7. Full tabulated data with simulation results of 90 random generated designs. .............................................. 66
Page 9
2
Abstract
In an effort to reduce fuel consumption and remain competitive in an automotive industry that
continues to increase regulations for environmental protection there has been an increase the use
of turbochargers. Due to this increase in demand, there is now a greater need for a centrifugal
compressor that provides a wide and stable operating range that delivers a high-pressure output at
a competitive price. The analysis of the phenomena that occurs within a centrifugal compressor is
extremely challenging and presents three-dimensional unsteady flow analysis with the addition of
thermal flow and expansion. Based on experimental and numerical simulation analysis, this team
will develop a blade design for a centrifugal compressor using computational optimization
methods. The ultimate goal of this project is to create a blade design that results in a wider optimum
operating range whilst increasing the pressure ratio without decreasing the mass flow rate that the
engine requires for stable combustion.
Page 10
3
1. Introduction
1.1 Problem Statement
Due to the constant increase in fuel cost and increased regulations seen in the automotive
industry, a trend has been noted in the downsizing of engines, resulting in a increase in the demand
of turbochargers. Turbochargers create the benefit of equally powerful engines with a reduction in
fuel consumption.
The purpose of this project is to create a blade design that results in a wider optimum
operating range by optimizing the compressor wheel’s blade shapes. The specific compressor that
will be optimized will be a Mitsubishi TD04-13L. The design constraints will include the housing
of the compressor in order to focus solely on the optimization of the blade design. The goal of this
project is to increase the maximum efficiency of the system by at least 7% while also improving
the efficiency in the important regions of the compressor’s operating range, such as the areas near
the surge region and the choke region.
1.2 Motivation
The motivation for studying a centrifugal compressor is mainly due to the increased in
price in fossil fuels and a growth of interest in the area of efficiency in the car industry. The uses
of centrifugal compressors are currently downplayed heavily in the car industry due to the lack of
power that they produce in the racing industry. Approaching the optimization of the centrifugal
compressor through this senior design project will show that there is an interest and that there are
practical applications for this type of technology. The team has selected this project for the
opportunity to apply advanced concepts and techniques that have been presented during the pursuit
of a Bachelor’s degree such as thermal and fluid mechanics.
Page 11
4
1.3 Literature Survey
Centrifugal compressors are a result of complex phenomena at play that manages to
achieve a pressure rise while adding kinetic energy to a continuous flow of fluid through either a
rotor or an impeller. This kinetic energy is then converted to an increase in potential energy by
means of a diffuser that slows down the flow of fluid creating static pressure at the outlet.
The inlet to a centrifugal compressor is typically a cylindrical pipe that may have features
such as vanes in order to help rotate the flow, as well as pressure and temperature measuring
devices. These vanes add what is known as, prewhirl, to the fluid making it possible to reduce the
Mach number at the inlet. The relative Mach number at the inlet will be given by equation 1 below:
𝑀1 =
𝑉1
√𝛾𝑅𝑇1
(Eq. 1)
where,
𝑇1 is the static temperature at the inlet.
𝑉1 is the Velocity at inlet
𝛾 is the specific weight of the fluid
𝑅 is the gas constant (R=8.314 J/mol x K)
The Mach number is important at the inlet because shockwaves can occur on the curved
part of the impeller from the air that is breaking away as shown below in Figure 1.
Page 12
5
Figure 1. Breakaway commencing at the aft edge of the shock wave [1]
As the relative velocity reaches the speed of sound, this separation causes huge pressure
losses and drastically affects the efficiency of the compressor. A disadvantage that arises when
using prewhirl is that the work capacity is reduced since a whirl component has already been added
to the fluid before entering the impeller.
Once the fluid has entered the impeller, it enters the only rotating aerodynamic part in a
centrifugal compressor. It is the most critical component, and provides all of the kinetic energy
that will be added to the gas as well as being responsible for up to 70% of the static pressure rise
in the system. A well-designed impeller will result in a well-designed compressor because they
have been calculated to achieve efficiencies in excess of 96%. There are two categories for
impellers: shrouded (Figure 2) or unshrouded (Figure 3), and their blades can be either two-
dimensional (Figure 4) or three-dimensional (Figure 3).
Page 13
1
Figure 2. Shrouded Impeller.
Figure 3. Unshrouded impeller. These impellers also
tend to have complex 3-D shaped blades.
Figure 4. Two-dimensional impeller blades.
Each different design has its own advantages and disadvantages. Unshrouded (open)
impellers can operate at much higher rotational speeds, equating to higher pressure ratios.
Shrouded (closed) impellers deliver pressure ratios of 3:1 or less, while open impellers can reach
pressure ratios upwards of 10:1 [3]. While that may be an advantage, open impellers also tend to
be less efficient due to losses associated with tip leakage flow, which does not occur in closed
impellers. Tip leakage flow is the flow of fluid the leaks over the rotating blades and flows in-
between the blades and the housing causing losses in pressure. Different factors are taken into
consideration when selecting the type of impeller to be used in a compressor, one of them being
the impeller flow coefficient, 𝜑 [1].
𝜑 =
𝑄
𝑁 × 𝐷23
(Eq. 2)
Page 14
2
Where:
Q is the impeller flow capacity
N is the operating speed
𝐷2 is the exit diameter
The impeller flow coefficient relies on the flow range that the compressor will be subjected
to. Impellers that have low flow coefficients are described by long, narrow passages, while
impellers with high flow coefficients have wider passages in order to adjust for the higher flow
rates [2]. For this study, the mass flow rate of a typical engine for a typical American driver will
be used as the base line for the compressor’s size. The composition of gas to be compressed as
well as the ranges of pressures and temperatures of the gas as it enters the compressor must also
be specified for a certain application.
Figure 5. Zones on compressor map of turbocharger that are most important from design/research and development
standpoints [4].
From a design point, there are certain zones within the operating range of the compressor
that are more important than others shown below in Figure 5. A turbocharger operates within the
green zone when a vehicle is in idle and while accelerating from a standing start, making this
Pre
ssure
Ratio
Eff
icie
ncy
Volume Flow (m3/s)
Page 15
3
region very important during city driving. This region is also critical in terms of eliminating turbo
lag, which dictates the peak torque point of the engine and the amount of emissions produced. The
red zone in Figure 5 has to do with supplying the engine with high volumes of air that is needed
in driving at higher speeds. In this region, the stress subjected to the impeller is at its highest. The
deciding factor related to the performance in the red zone is the compressor’s ability to meet the
additional demands of high-altitude driving since it is this variable that causes the impeller to reach
its highest speed [5].
Figure 6. Illustration of the inducer and exducer of compressor wheel [9].
The configuration of these blades play a major role in the efficiency and the pressure rise
that is seen when the fluid is exiting the impeller. Once the fluid (air in this case) has passed the
inducer, it goes through the exducer, which converts all of the kinetic energy of the fluid into
pressure by slowing down its velocity as shown in Figure 6. At the very tip of the exducer the
velocity of the fluid is at its fastest which leads to the Mach number to be in excess of unity. These
high Mach numbers can cause shockwaves, which propagate downstream causing cyclic loading
of the impeller [2]. This loading can lead to early fatigue failure if the frequency is of the same
order as one of the natural frequencies of the impeller vanes. Once the fluid has passed the diffuser
it is received by a collector, which also can take several forms. If it is a large chamber, it is called
a Plenum, but when it is collected into an object that looks similar to a snail shell it is termed as a
Page 16
4
volute or scroll. The collector’s purpose is to gather the flow that has been discharged from the
diffuser and deliver it towards a downstream pipe that could also contain valves and
instrumentation that gauge the performance of the compressor [4].
1.4 Survey of Related Standards
Standards are created with regulation and requirements whether it be for safety, feasibility,
machinability, or fitting. These regulations are put in place to keep everything organized and make
designing follow a certain rubric. Here are a few of the standards that standard turbocharger
systems have to follow:
1. American National Standards Association (ANSI)
The American National Standards Institute is a private non-profit organization that
oversees the development of voluntary consensus standards for products, services,
processes, systems, and personnel in the United States. The standards used for the
manufacturing of this prototype follow the guidelines from this association.
2. American Society for Testing and Materials (ASTM)
ASTM International, known until 2001 as the American Society for Testing and
Materials, is an international standards organization that develops and publishes voluntary
consensus technical standards
3. Society for Automotive Engineer (SAE)
SAE International, initially established as the Society of Automotive Engineers, is
a U.S.-based, globally active professional association and standards
organization for engineering professionals in various industries. Principal emphasis is
placed on transport industries such as automotive, aerospace, and commercial vehicles.
4. American Bearing Manufacturing Association (ABMA)
Page 17
5
It’s an industry trade group that deals with all aspects of bearing technology. This
project does not directly relate to this association, but the surrounding components, mainly
the center housing and rotating assembly of a turbocharger make use of their standards.
5. Unified Thread Standards (UTS)
Defines a standard thread form and series—along with allowances, tolerances,
and designations—for screw threads commonly used in the United States and Canada. It
is the main standard for bolts, nuts, and a wide variety of other threaded fasteners used in
these countries.
1.5 Discussion
The purpose of this project is to bridge the divide between the better performing high quality
compressors, and the more economically viable compressors. Using the same high quality
manufacturing processes while maintaining performance and durability will be the top priorities
for this project. In the process of creating this centrifugal compressor, the cost of manufacturing
will become the main challenge in creating this product in order to make it appealing for
commercial use. Another priority for this project is an effective design that improves the
performance and efficiency of the compressor over a larger range. Achieving these goals will result
in more fuel-efficient cars producing the same power with less fuel.
Page 18
6
2. Project Formulation
Overview
Current turbocharger and supercharger companies use designs that tend to be simple and
cheap to produce. The performance of their product is largely overlooked when it comes to
commercial vehicles due to the greater importance of decreasing cost. These lower price options
use simply designed wheels, poor quality manufacturing processes and low cost materials. The
designs that are used in modern day compressors in automotive turbochargers have a range of
seven to five impeller blades depending on the make and model. The combination of material and
manufacturing processes as well as outdated designs reduce the overall performance of
commercial turbochargers. On the other hand, high performance companies use high quality
materials, with high tolerances in the manufacturing processes, with improved wheel designs and
strength.
Project Objectives
The purpose of this study is to bridge the divide between the better performing high quality
compressors, and the more economically viable compressors created by cast molding. Using the
same high quality manufacturing processes while maintaining performance and durability is vital
to the success of this study. In the process of creating this centrifugal compressor, the cost of
manufacturing will become the main challenge in creating this product in order to make it
appealing for commercial use.
For the impeller design in this study, a desirable outcome would be a high efficiency value
throughout the operational speed range that would not abruptly become inefficient along the upper
limits of the operational speeds. With this in mind, the specific goal for this project becomes
creating an impeller blade that will increase the pressure difference at a set operational speed. This
Page 19
7
goal can be achieved by the process of multi-objective optimization, which will use CFD data from
90 randomly generated designs and the genetic algorithms used in modeFRONTIER® in order to
maximize the pressure ratio and efficiency, while reducing the moment of inertia. If this goal can
be achieved, then the increase in fuel efficiency in a vehicle with the same parameters could be
improved with just the modification of the compressor in the turbocharger system.
Design Specifications
In order to create an effective centrifugal compressor, certain physical considerations must
be made such as manufacturing material and processes, tolerances, and overall blade design. But
before any of these considerations can be taken, the type of system that will be supported by this
compressor must be determined. It became apparent that a turbocharger that was already available
to test would become the logical choice for a case study in the optimization of a centrifugal
compressor. This led to the selection of a Mitsubishi TD04 turbocharger, which was currently
installed inside a team member’s Subaru Impreza WRX. The dimensions for this impeller are
detailed further along this report in the design prototype section.
Figure 7. Impeller removed from a Mitsubishi TD04
turbocharger.
Figure 8. TD04-13T compressor housing.
Figures 7 and 8 above is the compressor housing and impeller that were removed from a
Subaru Impreza WRX. This particular impeller had already experienced 170,000 miles of driving
Page 20
8
before being removed from the engine, showcasing the excellent quality of construction that
Mitsubishi is known for and what this case study strives to accomplish in the prototype design.
This turbocharger system is an ideal candidate for optimization due to its small size, and
its capability of being tested in real-life working conditions alongside an exhaust turbine. Its small
size also means that this particular compressor could be used with many other downsized
turbocharged engines that are becoming more popular in the automotive market [10] in order to
increase fuel efficiency.
Addressing Global Design
Many countries around the world are implementing higher fuel efficiency standards for all
automotive companies within the next 10 to 20 years. In the US, the Corporate Average Fuel
Economy (CAFE) created in 1970, to increase the average fuel efficiency of all automotive
companies to 34.1 mpg for combined cars and trucks by the year 2016 described by Table 1 below
[11].
Table 1. Average required fuel economy (mpg) under final CAFÉ standards.
In the future it seems that the electric car alongside hydrogen fuel cell technology will
become the most prominent power source for the general consumer’s vehicle, but the dependency
in fossil fuels will remain. The addition of turbochargers to downsized engines can increase the
vehicles fuel efficiency without any compromises in performance while the electric car matures
into an economically feasible alternative.
“During the past decade, turbochargers have moved from a niche technology in the high-
performance market segment to an integral part of manufacturers’ mainstream emission control
Page 21
9
and fuel economy strategies,” Gott said. “Because they are compatible with virtually all engine
technologies, they represent a global growth opportunity, making substantial gains even in the
limited-growth markets of the U.S., Japan and Europe [14].” With that stated, the design created
for this study will address the global market by means of Multilanguage technical drawings and
service manuals (i.e., English, Spanish, German, French, and Portuguese) if one desired to install
this impeller in their turbocharger system.
Constraints and Other Considerations
The process of producing a turbocharger takes into account of many different aspects of
the engine in order to create an efficient system that could perform effectively for the lifetime of
the vehicle. In this case study, only the aspect of the compressor was taken into account, which is
an integral part of a turbocharged engine system. It affects the amount of spool time, pressure ratio,
inlet air temperature, efficiency, and capacity of the engine, which affects many parts of the system
such as the intercooler, the turbine, and the piston seals among other things within the engine. Any
changes that will be done to a turbocharging system affect other aspects of a vehicle and should
be taken into consideration. With an increase pressure ratio as this prototype aims to achieve, a
more effective intercooler is necessary to cool the air in order to not increase the temperature
experienced within the combustion chamber [12].
Cost is another consideration that needs to be taken into account for the scope of this study.
The process of developing a blade design for an impeller could perhaps be the largest expenditure
due to computing time, man-hours, and a verification of simulations. The method of manufacture
also affects the cost of the product and depends on the intricacies of the design, manufacturing
process, and tolerances. Several considerations were taken for this study, but all of the processes
called for a prototype production cost in excess of $1000 dollars. For an initial sizing in order to
Page 22
10
test for clearances, rapid prototype methods such as plastic ABS plastic 3-D printing will be used.
It is paramount that for the final prototype a material that is suitable for the amount of stresses that
will be applied when revolving at extremely high speeds won’t deform and cause a failure to the
system.
Discussion
A turbocharger is a rather simple device that recycles energy from the engine and produces
more power with the same engine displacement. In order to improve an existing system, there are
several components within the system that could be enhanced such as the compressor wheel. The
preparation for this project was created based on many books and journal articles of companies
that continuously improve the many components that are present in a turbocharged engine system.
The difference in this case study versus the others seen in many journal articles is the
implementation of a random-geometry generator paired with powerful multi-objective optimizing
algorithms in order to create an efficient impeller at a certain rpm. This method of design
formulation allows for the creation of one prototype that could be bench tested in the real world in
order to verify the FEA and CFD analysis that were performed indeed increased the performance
of the impeller without the need of a trial-and-error process.
Page 23
11
3. Design Alternatives
3.1 Overview of Conceptual Designs Developed
The conceptual designs developed for this study are based on the current geometry of a
TD04-13T compressor housing. In order to use a multi-objective optimization algorithms, it is
necessary to produce a large amount of data that reflect the fluid dynamics of the system. In order
to produce this data many alternate designs are required to be simulated. Using a random-geometry
generator code gracefully provided by Sohail Reddy of the MAIDROC laboratory, 90 alternate
designs were created and altered with eight parameters.
3.2 Optimized Design Alternatives
The optimization of any component is based on controlling and altering particular
parameters that are characteristic of a given design. These specific parameter that have been chosen
as vital design points, done either at random or at a specific trend, produce a design component
that will outperform the original component. These parameters are related to mathematical
phenomena, which make this process more intricate. However, this process also produces the best
possible design in the shortest amount of time, which equates to an overall lower cost. A trial and
error process can be used in order to achieve the same outcome, but that process may take a longer
time and will not guarantee a desired result. As mentioned before, it requires a complex analysis
and development of the data, a process that will be explained shortly.
The geometry of the blades on the compressor wheel determines the air velocity that exits
from the compressor housing. This geometry has different parameters and points; all of these
parameters are linked between them. For the optimization process, parameters are modified in a
certain ways in order to achieve the desired result. The idea of doing this is to obtain different
results from different geometries.
Page 24
12
Figure 9. 3D sketch guideline for impeller blade. To the right are the eight parameters which will be altered.
The ultimate goal for changing these parameters is to achieve the highest average velocity from
the compressor, which will result in the highest efficiency and performance. Each parameter
changed produces a different geometry that will allow for the study of computational fluid
dynamics (CFD) results. The average velocity will be used instead of the maximum because it will
indicate roughly how much pressure can be created from the compressor. In order to optimize the
results each parameter must be modified at least 10 times, subsequent of 90 designs.
These modifications are at a random, following no particular pattern, just within a specific
range that was determined by the geometry of the blade. Once these random numbers are generated
and the parameters are modified on the CAD model a CFD simulation is performed in order to
calculate the average velocity and the maximum velocity. The full tabulated data of all simulation
results of the randomly generated designs can be found in Appendix C.
Page 25
13
Once the studies for each design have been performed, the data is put into the optimization
software, modeFRONTIER®, a program that performs multi-objective and multi-disciplinary
optimization. The software modeFRONTIER® was provided by the FIU MAIDROC laboratory
and the resident graduate research assistant Rajesh Jha. This program streamlines the
multidisciplinary approach in the design process by simultaneously considering multiple
parameters to determine the optimum solution. The main advantage to using this optimization
process is the ability to choose the optimization strategy, depending on the design space boundaries
along with the accuracy of a reliable, improved design. Another advantage to this process is being
able to create a design from the ground up on the computer, with improved performance without
ever needing to perform real world tests until the fabrication of the prototype. This cuts cost on the
design process as well as time, depending on the computing processing power that is available.
Once modeFRONTIER® has generated a batch of geometries, they will then be verified through
more CFD and FEM analysis. The geometry that produces the most appealing results, in this case
the highest average velocity, is selected for prototyping and testing.
3.2.1 ModeFRONTIER® Algorithms
In order to obtain the design population, evolutionary algorithms were used together with
a response surface, both of which were needed for the design generation. An evolutionary
algorithm (EA) basically uses mechanisms inspired by biological evolution, such as reproduction,
mutation, recombination, and selection. Similar to selective breeding in plants and animals, the
solution to these optimization problems, play the role of individuals in a population, and the fitness
function determines the quality (performance) of the solution. The advantage to using this
algorithm is that it does not make any assumptions with regards to how close or far the different
individuals and their properties affect the performance.
Page 26
14
ModeFRONTIER® has a number of evolutionary algorithms to choose from. The
algorithm that was chosen for the project’s objective was NSGA-II, which creates optimal
designs by grabbing multiple objectives and optimizing the goal desired. It optimizes designs by
using mathematical algorithms to solve multi-objective problems. The centrifugal compressor
required the use of multi-objective optimization due to all the different parameters that were
being considered.
The figure below shows how the evolutionary algorithm works with the analogy of the
breeding in horses. In the first generation we have 4 different horses, or parameters, these 4
parameters will be altered to generate new values, but each horse will be combined with another
one to generate a child, or design. This new design will use both parents information, and it can
also mutate to generate random creations.
Figure 10. Explanation of how evolutionary algorithms work just as selective breeding.
Response surface algorithms define how the different dependent and independent variables
affect each other in terms of optimizing the final goal. The two main RSA that were used were
the Gaussian Processes and the Radial Basic Functions. The main reason for the use of two
Page 27
15
different algorithm was to compare the result, making sure that the output was within a 1% or less
difference. The method used for the final design was radial basic function.
After all the variables were used on the optimization, the final output generated over 8,000
designs, which show increased average velocity. After filtering the plethora of designs, the
maximum desired values were found and stood out with these 3 designs shown below.
Figure 11. Above shows eight parameters and desired goal objectives.
3.2.2 Graphs and data validation
The first graphs shown below (Figure 12) is an illustration of a 3D graph of how two of the
most influencing parameters, the twist angle of the blades and initial (lower) angle of the loft, are
improving and maximixing the air velocity. One can notice that the maximun value of the graph
for average velocity is 124 m/s, which doesn’t match the final output of 134 m/s, this is because
the 2 parameters named before alone improved the design in the most radical ways while the other
6 remaining paramenters were resposable for the extra 10 m/s.
Page 28
16
Figure 12. 3D Graph of angle and initial angle with respect to the average velocity.
The graph below (Figure 13) shows how the initial angle of the loft as it varies, changes
the average output velocity. As the angle between the horizontal plane and the initial loft increase,
there is a section where it exhibits a trigonometric function, but the further the angle is increased,
the more linear the relationship becomes. The spot indicated on the graph is the angle that was
nominated for the prototype, resulting in an initial angle of 67.22 degrees.
Page 29
17
Figure 13. Plot of Average Velocity vs Initial Angle
The graph below shows how as the angle between blades varies, changes the average output
velocity.
Figure 14. Plot of Avergae Velocity and Angle
Page 30
18
This graph shows the design generation as the angle between the blades varies, at the
bottom of it can be seen that the design generation, creates many design on this region since the
maximized value was between 75 and 74 degrees
Figure 15. Design Generation of Angle
A similar situation happens with the initial angle of the loft.
Page 31
19
Figure 16. Design Population of Initial Angle
Part of the validation process was to verify that the output from the optimization will match
the output from the CFD study on Solidworks, lastly the output design parameters were modified
on the compressor wheel to allow for the simulation to be ran. The optimization average air
velocity was 134 m/s and the output from the CFD study was 130.28 m/s, which meant that the
results were within the limits of tolerance, which was a 2.85% error, between the optimization and
CFD study.
Figure 17. Final Solidworks CFD Study
Page 32
20
When optimizing there is a process that needs to be followed on how many different
designs need to be created. A set number of parameters are chosen and then according to the
number of parameters that are going to be modified a number of minimum designs can be made.
ModeFRONTIER® created over 8,000 designs that could be chosen by using mathematical
algorithms. A data table, which can be found in Appendix C, has to be provided and then using
the inputs and outputs, certain output goals can be chosen and those goals will be optimized.
3.3 Design Alternative
One of the design alternates is choosing one family of blades. When choosing only one
family of blades more blades will need to be added to get similar results to the chosen design.
Most high-performance turbochargers are benefitting from using two families of blades. The GTX
series turbochargers from Garrett are one of the industry professionals that currently have one
family of blades and 11 blades in total to produce high boost pressure. The problem with creating
a one family, many blade configuration is that machining becomes complicated with the really
small tight tolerances that are needed. CNC machining of this type of configuration is very
expensive because custom sized tool-bits are needed to make the small gap between blades.
3.4 Integration of Global Design Elements
The method of optimization that was used to create the proposed design, could also be
used for many other fields as diverse as art, biology, marketing, genetics, operations research,
social sciences, physics, politics and chemistry. This is mainly due to the unique property
evolutionary algorithms possess in not making any assumptions of the capabilities certain
Page 33
21
parameters can produce in the right combination. The production of more efficient and better
performing products would benefit from using the algorithms ModeFRONIER® provides the
users with, and could possibly change the manner in which many designers create their product
prototypes. The only downside to using this method is the amount of data needed for accurate
results, which would depend on the number of parameters that could need to be altered.
3.5 Feasibility Assessment
Turbochargers have a big global market that is always increasing in demand and thus
supply must increase as well. Many automotive manufacturers around the world are being pushed
to increase fuel economy in all cars. Most car manufacturers are using turbocharger technology
to provide this increase in fuel economy. This technology is in high demand and keeps increasing,
therefore the value to be attained outweighs the cost of research and development. Following the
supply and demand curve of this technology can dictate maximum feasibility of this project.
3.6 Proposed Design
The centrifugal compressor that will be created for this project will compose of two major
components: the impeller and the housing. The CAD design of the compressor blade impeller was
created by using the modeling software, SolidWorks and is one complete solid piece. This design
was finalized once the process of blade optimization was completed. The physical design consists
of an aluminum wheel with seven main blades and a set of smaller vanes in-between these blades.
In order to create an impeller with the highest coefficient of efficiency, the spacing between the
impeller and the housing must be minimized, requiring extremely high tolerances in the
manufacturing process in order to reduce the tip leakage flow. Which leads to the other major
component of this centrifugal compressor system, the housing.
Page 34
22
The effect that the housing has on the performance of a centrifugal compressor is
not as apparent as the impeller that is found inside. Noting the difficulty of creating
compressor housing out of a mold it was decided to use a housing component that was
already fabricated. A Mitsubishi TD04 housing component was chosen for this product due
to its small size relative to the majority of high performance turbochargers in the market
with the idea that a small turbocharger will be used in downsized engines with the intention
of improving fuel efficiency. Figure 7 shows the compressor housing that will be used in
conjunction with the compressor wheel. Another reason why this specific housing was
selected is for the opportunity to test the compressor with the current turbocharger installed
on a 2004 Subaru Impreza WRX. This vehicle is an example of engine downsizing in the
market, which has already been noted as a trend in the automotive industry. The process of
testing the compressor will be elaborated in the Prototype Analysis section of this report.
3.7 Discussion
Optimization needs to be elaborate and detailed and requires a lot of understanding of the
mathematical algorithms that are used in the modeFRONTIER® software. It requires previous
knowledge of how the program works and how it generate the designs. The feasibility of the
project scope is very important because it deals directly with if the project can and will be done or
not.
Page 35
23
4. Project Management
4.1 Overview
The timeline depicted below is a projected overview of how the project is going to be
managed. Tasks will be split accordingly to each strength of each team member. Being able to
work with the strengths of each team member maximizes the efficiency of the team collaboration.
The estimated completion date for this project is November 20, 2014. This date is ideal for all of
the team members in that it does not overlap onto other responsibilities or holidays.
4.2 Breakdown of Work into Specific Tasks
As shown below in Table 2, the responsibilities for this project were shared amongst the
team members evenly. All of the team members were actively involved in conducting research,
CAD modeling, performing CFD analysis, testing, and writing. The colored boxes represent the
responsibility that a team member was associated with throughout the duration of the project. The
blank boxes show where the team member had no involvement during the project.
Table 2. Table of responsibilities.
Responsibilities Xavier Medina
Fernando Lopez
Gianfranco Pisani
Research
CAD Modeling
CFD Analysis
Optimization
Manufacturing
Testing
Writing
Page 36
24
4.3 Gantt Chart for the Organization of Work and Timeline
Figure 18. Gantt chart.
4.4 Breakdown of Responsibilities Among Team Members
As team leader, Xavier Medina was involved in all of the processes of this project, but it
does not mean that his workload was greater than that of the other team members. He had a major
role in the CFD analysis due to circumstances such as processing power and flexibility. He also
had a role in the manufacturing process of the prototype along with the testing. Gianfranco had a
major role in the CAD modeling aspect of this project and created the base computer model needed
in order to create the random designs and run the software simulations. In addition to creating the
CAD modeling, Gianfranco along with Xavier handled the optimization process of the project
using modeFRONTIER®. Fernando’s major role in this project was the editing of the report along
with the testing and manufacturing of the prototype. His involvement in the testing and
manufacturing is because he is the owner of the vehicle that is being used in this project and he
will be around the prototype the most.
1/11 3/2 4/21 6/10 7/30 9/18 11/7
Identify Team & Project
Develop Detailed Plan
Conduct Research
CAD Modeling
Software Simulation (FEM Analysis)
Software Simulation (CFD Analysis)
ModeFRONTIER Optimization
Initial Test
Preliminary Manufacturing
CNC Machining
Final Testing
Page 37
25
4.5 Commercialization of the Final Product
Marketing campaign for this product could be geared toward a number of companies.
Design could be shown to Mitsubishi, since stock turbocharger that was used for constraints on
project is from the Mitsubishi line. Another alternative would be DieselSite Wicked Wheels
Company, which specialize in turbo-diesel engine technology. They deal with creating custom
blade configurations that outperform the blade profile set in standard diesel-turbochargers. All
custom designs are made and manufactured to specifically follow all dimensions of stock housings,
so modifications can be done by only changing the compressor wheel.
4.6 Discussion
Tasks were divided up according to the strength of each team member to maximize the
efficiency and flow of the team’s efforts. In the end everyone had a role in every single aspect of
the project. Proper team chemistry allowed for the best configuration in doing certain tasks
simultaneously alongside each other. Dividing this big project into manageable chunks makes
the task of managing the sub-phases of this study easier to handle.
Page 38
26
5. Engineering Design and Analysis
5.1 Overview
This section of the study involves the actual design and analysis of the prototype that will
be manufactured. Some of these analyses include structural, kinematics/computational fluid
dynamics (CFD), vibrational dynamics, stress, and deflection. Using these analysis, a final
prototype can be created in order to validate the claims produced by the simulations. All of these
simulations were created by using Dassault Systemes’ SolidWorks.
5.2 Kinematic Analysis and Animation
The primary analysis that are performed for this project is kinematic in nature.
Computational fluid dynamics (CFD) were used to simulate the flow of air through all of the
designs that were created. An important aspect of this computation are the boundary conditions
that dictate the environment surrounding the compressor wheel. The boundary conditions
consisted of atmospheric conditions for both the inlet and the outlet, and an environmental
temperature of 27 degrees Celsius (298 K). The surface was assumed to be perfectly smooth and
adiabatic. These conditions were used in order to simplify the simulations and shorten the
computational time that will require the majority of time.
The rotational speed of the impeller dictates the performance of the compressor, therefore
the optimization parameter that needs to be isolated is the angular velocity. Since the objective of
this project is to increase the efficiency throughout the operational speed of the compressor a range
needed to be established. It was estimated that the maximum velocity of smaller turbochargers to
be roughly 150,000 rpm, therefore the design speed of the prototype to be optimized was selected
to be 140,000 rpm. This would ensure that the maximum compressor efficiency would not be
Page 39
27
reached towards the higher end of the engine speed, and result in a higher efficiency throughout
the different volumetric flows of the engine.
5.3 Dynamic/Vibration Analysis of the System
The dynamic analysis consists of a centrifugal test set to rotate the compressor wheel to
the limit of its rpm, which is roughly 150,000 rpm or 15,700 rad/s. It must also be noted that this
limit could be reached by the compressor, but only be reached at the very limits of the entire engine
system.
The compressor wheel when spun at its limit could experience vibration from the point of
contact of the shaft, and propagate through the blades causing minor deflections (fluttering). This
vibration analysis could be done by performing a frequency simulation on the compressor wheel
component. The frequency achieved is 8311.1 Hertz, and at this frequency is when the deflection
can be seen on the second family of blades as shown below.
Page 40
28
Figure 19. Frequency simulation performed to assess vibration.
5.4 Structural Design
The compressor wheel blades are the biggest point of concern when it comes to structural
design. Since the compressor wheel is being spun in excess of 100,000 rpms the edge of where
the blades meet the body need to have rounded fillets and edges. These rounded edges allow for
dissipation of the centrifugal force that is generated by the high velocity rotations.
The following picture will show where these modifications were made. The stress analysis
will go in more depth on the reasons for why these rounded edges are needed.
Page 41
29
Figure 20. Where the rounded edges needed to be made.
The life cycle of the compressor wheel has to be very high since it is operating under
extreme conditions. A fatigue life test simulation was created to analyze how long it would take
under operating conditions for the compressor wheel to finally fail. The results are illustrated in
the figure below.
Figure 21. Life of the blade after going through 𝟏𝟎𝟕cycles.
Fillet
Fillet
Page 42
30
Figure 22. Damage that is present after 𝟏𝟎𝟕cycles.
5.5 Force Analysis
The main force acting on the compressor wheel is centrifugal force, which is generated by
the high speed of rotation. This force is created because of its moment of inertia being in the center
of where the bore is for the shaft. When rotating, the blades are being subjected to a reaction force
from the air molecules when increasing the velocity of the air. This reaction force turns into
centrifugal force as shown below.
Page 43
31
Figure 23. Direction of rotation and the forces that are felt.
5.6 Stress Analysis
Stress analysis needs to be done to assess the compactibility of the design and how it deals
with the forces that it is subjected to. It would indicate as to where certain stresses are being seen
and the areas that they affect. The main problem that was found was that at the point where the
blades and the main core meet, the curve had a lot of stress due to the area that was involded. This
area needed to be changed and 5 mm fillets were done at the points in order to increase the surface
area and decrease the stress. This allowed for a more sustainable stress dissipation. Finally, the
highest stress was 92 MPa which is less than the yield strength.
was considered to determine the different points on which modifications were needed as
stated before. On this section the reason and sizing of the fillets will be covered.
Centrifugal
Force
Page 44
32
Figure 24. Image showing areas where stress is experienced.
5.7 Material Selection
When making material selection several possibilities were contemplated. Since the
material that had to be used didn’t only have to be strong enough to withstand the continuous
punishment of operating conditions, but also had to be light to decrease the inertia and allow for
an easier spool. The material had to be machinable and rust resistant in order to be able to be
manufactured.
Initially the best material would be magnesium WE43, which has the lowest density and
very high strength, followed by aluminum and finally titanium in respect to density. With respect
to strength, titanium takes the top position, followed by aluminum and magnesium, respectively.
Page 45
33
For pricing purposes the cheapest option was aluminum, since magnesium and titanium alloys are
particularly expensive.
The decision was to use aluminum 7075, which offered the best properties out of the three
proposed series. A wide variety of materials could be used for the project, but from all of them
the most suitable was the ASTM 7075 (aluminum alloy). Aluminum offered properties that are
similar to all the previous materials, but aluminum could make the design cheaper and easily
machined. Titanium is a lighter choice and has more strength, along with magnesium, but the
price is not suitable for the project’s purpose. Aluminum offered economic cost efficiency without
sacrificing quality which is an important part of the design.
5.8 Design Based on Static and Fatigue Failure Design Theories
For the project’s application, static analysis was not taken into account because of its nature
of being a purely dynamic component. The fatigue failure design theory specifies on the cyclic
loading of the component under operating conditions. The load in question in this project would
be centrifugal forces acting on the compressor’s blades. One of the biggest problems would be
having sharp edges on the inside of the blades. Fillets are needed to make machinability easier
and also to address the fatigue failure that could happen. With fillets the blades have less of a
chance of experiencing critical stresses at the base of the blade and thus decreasing the chance of
fatigue failure.
Fatigue failure could be seen on the following figure. The compressor wheel has no
specific point as to where it would fail first. The figure shows where the component would fail
which is the whole component would fail, if failure were to occur. The lowest factor of safety
obtained was 1.3 at the limit of the rpm. This further shows that as the rpms drops the factor of
safety increases.
Page 46
34
Figure 25. Where failure would occur by dropping below a certain Factor Of Safety.
5.9 Deflection Analysis
The displacement or deflection provides a disadvantage. The rpms of the design show how
centrifugal force could predominantly demand careful considerations. Deflection can be seen on
the design at the red points. As was explained previously the vibration that occurs from the point
of contact where the shaft and compressor wheel touches and propagates through to the blades of
the wheel. This propagation of vibration produces a small fluttering effect on the small blades that
can be seen as the point of deflection. Although there are deflections they are within the
performance limits and do not affect the performance of the compressor wheel. Even at the highest
Page 47
35
rpm setting, the deflection is very minimal showing a deflection of 0.003187 mm to almost no
deflection.
Figure 26. Deflection could occur in the red areas.
5.10 Component Design/Selection
The basic shape of the compressor wheel was based on standard designs that were
commercially available. The procedure of the used shape of blades came from trying to first
resemble existing design, and then, improve the existing ones by using every tool that was
available and could develop the final design, meaning, flow simulation and optimization. Also
considerations on 2 or 1 family of blades were taken.
Page 48
36
5.11 Component Design Simulation
In order to determine the performance of the design, a flow simulation was created in
SolidWorks with a rotating region. This rotating region was the blade design, but in order to able
to test the wheel a compressor housing was needed as well. The compressor housing used for the
simulation was modeled using similar techniques that were used for the wheel. These two parts
had to work together in order to generate velocity values that would then be used to calculate the
pressure difference created by the system. After weeks of design considerations and modeling
changes, an assembly with proper flow conditions was achieved. After the complete assembly of
the 3D model, including wheel and housing, the project was ready for the next step, simulation
testing and initial conditions.
For the initial conditions it was important to have the correct ones, since the compressor
system works in a particular way, and these conditions were needed to generate accurate results.
Initially there are three main conditions or parameters, inlet, outlet and rpm. These three
components will determine the performance. To be able to understand the reason of these three
conditions, a review of how the compressor wheel generates the difference in pressure is needed.
Compressor wheels are inside compressor housings and these two parts work together to
create pressure difference. This is achieved by a divergent circular cross section that gradually
increases its diameter as its turns around the compressor wheel. The compressor wheel accelerates
the incoming air by drastically increasing its velocity, therefore, increasing its kinetic energy.
Following Bernoulli equations:
𝑔ℎ1 +
𝑝1
𝜌+
𝑣12
2= 𝑔ℎ1 +
𝑝2
𝜌+
𝑣22
2
(Eq. 3)
Where:
g is gravity
Page 49
37
h is height of fluid
p is pressure
𝜌 is density of the fluid
v is velocity of the fluid
In the compressor side system some assumptions can be made:
ℎ1 = ℎ2 (Eq. 4)
𝜌 = 1𝑘𝑔
𝑚3⁄ (Eq. 5)
𝑝1 = 101 𝐾𝑃𝑎 (Eq. 6)
Following these assumptions a simplified version of Bernoulli’s equation can be used:
𝑝1 − 𝑝2 = (
𝑣22 − 𝑣1
2
2)𝜌
(Eq. 7)
It is possible to determine the pressure difference by utilizing Eq. 7 above. As the circular
cross section of the housing increases its diameter, the velocity of the flow decreases, increasing
the pressure. So the objective was to generate the highest pressure difference at the limit of the
compressor wheel which was 12500 rad/s, or 140000 rpm, which is a common limit used in today’s
turbocharger due to lubrication and bearing technology. So far one of the conditions was defined,
now the inlet and outlet conditions will be defined as well. The conditions for the inlet and outlet
are atmospheric pressure the reason behind this is because the pressure difference is being tested
by determining how high the air velocity is in the housing. Since a higher velocity difference will
automatically generate a higher pressure difference, it results in a better design.
5.12 Finite Element Analysis
SolidWorks is a program that doubles up as a Finite Element Analysis solver. When
designating mesh sizes in all simulations, they were generally set to be coarse. Mesh sizes required
Page 50
38
to be coarse because of limited processing power on the computers that ran the simulations.
Attempting to run a finer mesh size resulted in a catastrophic error in SolidWorks and would shut
down the program. The mesh that also needed to be created due to the complex geometry of the
blade is a 3D dynamic mesh that is very difficult to simulate in SolidWorks. ANSYS is another
software simulation program that could be used to create and run Computational Fluid Dynamics,
but due to limited processing power and time, this simulation could not be generated.
5.13 Design Overview
The design objectives were always kept in mind across all the design aspect. Simulations
could have been performed better by using more advanced simulation programs. Boundary
conditions could have been researched to create a better, ideal testing simulation of CFD.
5.14 Cost Analysis
In the cost analysis, every expense will be reported and added to a final amount, three
categories which would best sum up and take into account the amount of work and cost necessary
are as follows: materials, design, manufacturing, and testing. The following sections go in-depth
for each category but the following table sums up the total cost associated with this project
involving everything from man-hours to materials.
Page 51
39
Table 3. Tabulated Cost Analysis
ID Name Res Duration Cost
ID (hours)
1.1 Conduct Research EIC Computer Lab 85 $21.25
1.2 CAD Modeling
1.2.1 SolidWorks Modeling EIC Computer Lab 140 $35.00
1.2.2 Visual Basic Coding Random Geometry (Sohail) 1 $-
1.2.3 Measuring Physical Model Impeller and Housing 2 $120.00
1.3 Software Simulation
1.3.1 FEM Analysis Analysis 90 $22.50
1.3.2 CFD Analysis Analysis 90 $22.50
1.4 Optimization
1.4.1 ModeFrontier Optimization Personal Computer 1 $0.25
1.5 Manufacturing
1.5.1 Raw Materials ASTM 7075 (Aluminum alloy) 0 $60.00
1.5.2 Plastic 3-D Printing Concept Manufacturing 4 $30.00
1.5.3 CNC Machining 5-axis CNC Machining 12 $280.00
1.6 Testing
1.6.1 Initial Testing Dyno runs/MPG testing 8 $320.00
1.6.2 Test of Prototype Dyno runs/MPG testing 8 $120.00
Total Hours Total Cost
441 $1,031.50
5.15 Discussion
Certain issues came up when designing the component, which was expected. The issues had
to be resolved to be able to make a viable component. Being able to identify these problems and
resolving them confirms the understanding of the analysis and simulations. Also understanding
the limitation of processing power better helps understand the problems that arose when generating
the CFD analyses.
Page 52
40
6. Prototype Construction
6.1 Overview
Manufacturing a prototype is a very important aspect of every design project. Making sure
that the correct type of manufacturing process is chosen is fundamental in producing a prototype.
With turbochargers every process is expensive because it is such an intensive and intricate piece
of machinery that needs to be manufactured.
6.2 Description of Prototype
The compressor wheel was initially manufactured through the process of 3-D printing in
ABS plastic. Ryan Lucia, a current undergraduate student at FIU, and his MakerBot® 3-D printer
created the plastic model at a cost of $20 per model which included the material that was used to
create it. This first prototype was used to be able to check if the design is close to the physical
housing of the Mitsubishi TD04-13T turbocharger.
Page 53
41
Figure 27. 3D printed prototype done by Ryan Lucia.
The final design prototype that would be ideal is to be made of aluminum 7075 through the
process of CNC machining using a 5-axis machine. This type of detail that is needed is crucial
when dealing with an object such as a turbocharger compressor wheel. The piece not only needs
to be machined with a lot of detail, but the tolerances are extremely tight. Rotating components
also have an extra thing that needs to be worried about which is correct rotational balancing. Any
slight imbalance in the compressor wheel when rotating in excess of 100,000 rpms could result in
catastrophic failure.
6.3 Prototype Design
A random design is chosen from the thousands of designs that were tabulated by the
modeFRONTIER program. From this random design it is simulated one more time to make sure
Page 54
42
that the goals that were supposed to be achieved through the optimizing algorithm are achieved.
The prototype that was chosen is one with curves that make machining easier and would not be as
steep to make sure manufacturing is easier.
6.4 Parts List
Rolled barstock of aluminum 7075 was needed to be provided to the machining company
Southern Gear. The material would be used to create the small compressor wheel geometry. A
small piece would be mounted on the CNC 5-axis machine and then machined out of the billet of
aluminum. The CNC 5-axis machine also would need two specific modules from GibbsCam to
be able to program the pathway for the tools to create the turbo machinery geometries.
6.5 Construction
The construction was going to be done with current IAB board member, Mr. Allan Arch,
CEO of Southern Gear. Southern Gear had an onsite 5-axis CNC machine that could perform the
machining of our prototype design. Problems arose in this area, with Southern Gear not having
the specific module that is needed to program the CNC machine.
They do not create compressor wheels, so naturally the two modules that would be needed
to program the machine to create the pathway would be missing. Without two modules that are
needed to program the CNC machine so that it can make a pathway for the tool, the machining in
aluminum could not be done. The level one 5-axis Multiblade module comes out to about
$3,375.00 with a yearly maintenance of $375.00. In addition to the level one module, the level
two module also needed to be added one to create the more complex geometry that was needed for
the prototype design. The level two 5-axis Multiblade module is priced at approximately
$13,275.00 with a yearly maintenance of $1,475.00. A rough estimate quote was still drafted by
Southern Gear’s manufacturing engineers to be able to reflect as much as possible on the actual
Page 55
43
cost of this project. The quote for machining the prototype out of aluminum was about $1,800.00.
It was assessed that for the project objective of only creating one compressor wheel, the price was
out of reach.
Other options were explored such as aluminum 3D printing done by Incodema3D. The
quote for the prototype came out to about $1,337.65, which was also too much for a self-sustained
project. After exploring further options with manufacturing professor, Richard Zicarelli, it was
decided to pursue a high resolution 3D printing option. WB Engineering is a company located in-
state that utilizes a polyjet process of 3D printing the compressor wheel in acrylic. A quote was
drafted up by WB Engineering to be priced at $149.04.
Figure 28. Polyjet 3D printed acrylic of final design from WB Engineering.
Page 56
44
6.6 Prototype Cost Analysis
Tables 4, 5, and 6 below show the estimated quote prices that were provided by trhe
separate companies that were researched as alternative options for manufacturing. As seen
Southern Gear would’ve been the optimal choice since it would actually have been manufactured
and machined. The budget being a self-funded project meant that it was not a viable option.
Table 4. Southern Gear final estimated quote price.
Company Description Cost
Southern
Gear
GibbsCam 5-Axis MultiBlade Level 1 Module $3,375.00
GibbsCam 5-Axis MultiBlade Level 2 Module $13,275.00
Machining Cost $1,800.00
Total $18,450.00
Table 5. Incodema3D final estimated quote price.
Company Description Cost
Incodema3D Additive Manufactured Part (DMLS) Aluminum AlSi10Mg $1,337.65
Total $1,337.65
Table 6. 1click3Dprint final estimate quote price.
Company Description Cost
1click3Dprint Multi-Jet-Printing Acrylic-Like $149.04
Total $149.04
6.7 Discussion
The manufacturing process of this project was unfortunately not able to be done to the most
desirable outcomes. Due to budget constraints, the aluminum compressor wheel could not be
manufactured for proper real-world testing.
Page 58
46
7. Testing and Evaluation
7.1 Overview
Testing on turbocharger compressor wheels are very intensive. Turbochargers have to be
able to function continuously at the limits of its operating conditions for many cycles. The figure
below shows a test bench of a turbocharger system, where the extreme temperatures and rotational
forces at work cause the turbine housing to glow red after a certain amount of time and exhaust
pressure loading.
Figure 29. Visual example of a turbocharger system undergoing fatigue and performance tests.
Only during extreme conditions and for short periods of time will the turbocharger system
experience this amount of heat generation. Turbo manufactures must test these systems in this
manner in order to verify that there won’t be any failure, typically running these test continuously
for upwards of 16 hours. For the scope of this project and with the resources available, these types
of tests are economically feasible but there are alternatives to this extreme testing. The following
section will go into depth of the possible experiments that could verify the design of the prototype
compressor created through multi-objective optimization.
Page 59
47
7.2 Design of Experiments - Description of Experiments
Performance data will be collected from the stock turbo within Fernando Lopez’s 2004
Subaru WRX. There are two different ways of getting performance data out of the stock turbo that
is currently in the car, by calculating miles per gallon (mpg) and by performing dynamometer
testing. A third way in which a comparison test could be created is in creation of a test bench for
the compressor that would create a controlled environment for the system. This third method does
not take into account several parameters such as a variation in temperature, environmental
pressure, vibrations, and losses due to turbine inefficiency.
7.2.1 Experiment 1 – Average Miles per gallon
The first experiment will be approximating the gas mileage of the car before and after the
modification of the compressor wheel. A path of approximately 100 miles will be mapped out
using Florida’s highway system where the car will be ran under constant conditions as shown
below in Figure . Weather data will need to be collected on the days that gas mileage runs will be
done in an attempt to recreate every gas mileage run, such as humidity, temperature, time of day,
and barometric pressure. The car will be taken to the same gas station and to the same gas pump
in order to minimize the number of variables in these approximations of mpg. After machining the
final prototype and installing it into the car, the gas mileage tests will be done again under the same
conditions. This type of experimentation will directly relate the effect of a compressor impeller to
the mpg or the fuel efficiency of a car. It has been hypothesized that a compressor wheel that has
been optimized to perform more efficiently within a certain rotational velocity could increase the
overall fuel efficiency of a vehicle by 5%. The results from this experiment validate the prototype’s
performance from a practical approach that is simple for the masses to understand.
Page 60
48
7.2.2 Experiment 2 – Dynamometer Testing
The second experiment that will be performed will involve placing the car on a chassis
dynamometer. Certain mechanic shops have these performance-reading machines that collect real-
time instantaneous data from the car such as horsepower, power generated, air-to-fuel ratio ratings,
carbon emission output, torque curve readings, boost generated, etc. A dynamometer in a machine
shops performs all of these readings within a controlled environment. The car is situated on drive
rollers with sensors which measure the torque being delivered, which could be used along other
sensors for a summary of the vehicle's performance. The dynamometer system being used should
have emissions sampling, measurements of engine speed and load control, data acquisition, and
safety monitoring enveloped in a complete test cell system.
There are three types of tests that can performed using a chassis dynamometer:
1. The engine is held at a steady state, meaning that the engine is held at a specific rpm, or at
several different rpms for a given time.
2. The engine is tested under a load and allowed to rev up in rpm, between a certain ranges.
This is commonly known as a sweep test.
3. This final test type involves changing the engine power and speed throughout the test
cycle. This type of test could follow standards used by the EPA in order to determine the
different dynamics of a car or engine.
These preliminary tests will be done to collect as much performance data from the stock
turbo as well as the entire vehicle. Once the prototype has been created and properly installed, the
same bench tests will be performed using the same equipment and parameters. The tests that will
be specified in the performance mechanic shop will have to do with the boost pressure buildup
(spool time), power and torque increase (or decrease), and fuel efficiency. These dyno pulls will
also be used to test and affirm that the final desired product can withstand the revolutions that a
Page 61
49
turbo is subject to experience. The dynamometer tests will push the turbo to produce a full range
of performance data alongside the performance of the engine.
New and old data will be charted and plotted side-by-side to show comparisons between
stock turbo and final product turbo data. This data will be used as a validation for the SolidWorks
simulation results that have been charted. The results from this experiment are more reliable than
the other forms of experimentation due to the nature of dynamometer tests and demonstrate a real-
world performance validation of the compressor in relation to the entire system.
7.2.3 Experiment 3 – Compressor Test Bench
This experiment being performed will validate the performance of the compressor to the
simulations performed within SolidWorks. Turbochargers rely on a source that provides high
temperature and high pressure gases. Normally these gases will come from the exhaust stroke of
an internal combustion engine. Since an internal combustion engine is a complex system that
emits harmful carbon monoxide gases testing the turbocharger outside of the car becomes
expensive and difficult to do even without including an engine as the source of exhaust gas.
Since the scope of this project was only the optimization of the compressor wheel, a
feasible test bench could be created that did not involve the hazardous conditions turbochargers
were designed to endure. This also led to the certain conditions that were used in SolidWorks.
Two parameters will need to be measured: the rotational speed and the output air velocity. The
boundary conditions used in the SolidWorks simulation need to be recreated within this test
bench include environmental pressure (101 kPa) and temperature (25 C).
Page 62
50
Figure 30. Air compressor that will blow air through
turbine side of turbocharger.
Figure 31. Computer that is directly connected with air
compressor to get all readings.
7.3 Test Results and Data
As mentioned in the section above, a test bench would need to be created that can rotate
the turbine wheel in excess of 100,000 rpms by pushing pure compressed air through the turbine
exhaust gas pipe. This compressed air will spin the turbine which is directly connected to the
compressor wheel blade and thus the compressor wheel will rotate in unison with the turbine
wheel. As mentioned above in the prototype section, the optimal material could not be used to
manufacture the prototype, thus debilitating the testing process.
It was stated that the prototype that was created is a polyjet 3D printed acrylic prototype.
Although this material is stronger than the conventional ABS and PLA plastic 3D printed pieces,
it still cannot withstand operational conditions for testing. Rotating this piece in excess of 100,000
rpms can potentially force the component to explode from the centrifugal force that it is being
exerted to.
Page 63
51
7.4 Evaluation of Experimental Results
Since no testing can be done there are no real-world physical experiment results to evaluate.
All processes of designing the experiments are detailed so that for future work testing can be done
when funding is available for further research and development.
7.5 Improvement of the Design
After getting test results from running the designed experiments, discrepancies can be
noted in the simulations and changes can be made in the development process.
From a simulation standpoint, it can be seen by the figure below, that the optimized wheel
does produce better overall average velocity across the spectrum of operating conditions.
Following from the bottom spectrum of the rpms, to the top, the optimized concept design can
produce a higher overall average velocity. With a higher average velocity it produces a more
stable increased pressure ratio.
Figure 32. Optimized wheel avg velovity vs. Stock wheel avg velocity.
0
20
40
60
80
100
120
140
0 2000 4000 6000 8000 10000 12000 14000
Ave
rage
Air
Vel
oci
ty
RPM in Rad/s
Wheel Comparison
Optimized Wheel Stock Wheel
Page 64
52
7.6 Discussion
The testing of the compressor wheel component tends to be very arduous and demanding. The
compressor wheel has to be able to meet the cyclic loading of the operating conditions throughout
cycles. As stated in the structural design section, the compressor wheel could be put under
operating conditions and pushed to the limits for many cycles and almost no damage would be
noticed. Unfortunately, due to the extreme cost in manufacturing an impeller blade that could
withstand the immense rotational forces a functioning prototype could not be created. A bench test
was formulated that could theoretically test the compressor, by means of rotating the turbine blades
with the help of compressed air. The fuel efficiency test was also not able to be performed, but it
is hypothesized that there would be an increase of roughly 3-5%. Over the lifetime of the
compressor, there are major savings to this minor increase, aiding in reducing CO2 emissions.
Page 65
53
8. Design Considerations
8.1 Assembly and Disassembly
The disassembly and assembly of a turbocharger unit is specified within this report only refers to
the Mitsubishi TD04-13T installed on a 2004 Subaru Impreza WRX. If this modification is to be
made for another unit, then there is a different set of instructions tailored specifically for said unit.
It has been assumed that the turbocharger has already been removed from the vehicle. Everything
shown in this section has been adapted from ©Stage Infinity, under the Mitsubishi TD04-13T
Turbo Rebuild section of their website. It must be noted that not balancing the shaft with the new
compressor could be extremely dangerous to the condition of the turbocharger and is not
recommended. Below is a list of tools that will be needed in order to disassemble the turbocharger
unit:
10,12 mm wrench
8mm socket
hammer
flathead screwdriver
large-bore snap-ring pliers (3” bore minimum, a high quality one is recommended such as
a McMaster-Carr PN# 5415A64)
Penetrating lubricant like WD40, Liquid Wrench, etc.
Touch-up paint, White Out, or similar painting tool with small brush (substitutions possible
– read full page to understand what it is needed for)
A reasonably sized area should be used for this project, with some rags and paper towels handy as
well as a small container to store the parts that will be removed. The most delicate part in this
Page 66
54
turbocharger is the compressor wheel and the turbine wheel. And the most troublesome part of the
disassembly and assembly of this unit is the snap ring for the compressor. This is where safety is
important: always wear safety goggles when working with snap rings. Under high compression if
they slip uncontrollably, snap rings could be shot across the room at a very high speed.
Disassembly:
1. Remove all hose clamps and hoses. There should be hoses going to the wastegate actuator
and compressor housing. Set these parts in a container, as they will be reused.
2. Remove the wastegate actuator (2 12mm bolts and 1 retaining “E” clip) as shown in Figure
15.
Figure 33. Wastegate Actuator.
Figure 34. Turbocharger clamp holding the turbine
housing with the cartridge.
3. Remove the clamp around the center housing and rotating assembly (CHRA - 1 10mm bolt)
shown in Figure 16. Remove the bolt and nut completely from the clamp and put these in
your parts container, as they will be reused.
4. Remove the 17mm banjo bolts and coolant line from the CHRA as shown in Figure 17. Some
penetrating fluids may be needed in order to remove these bolts. With the lines removed,
thread the bolts back in to keep debris out or plug the holes with something else and put the
Page 67
55
bolts and line in the parts container. Make sure to keep the 4 copper washers, unless they are
being replaced.
Figure 35. Coolant lines with 17mm banjo bolts and 4
copper washers.
Figure 36. Oil return line connected to the CHRA.
5. Remove the 2 10mm bolts holding the oil return line to the CHRA as shown in figure 18. The
flange may have trouble coming off, but once it does take note on the condition of the gasket
material that will be stuck to the surfaces. A new gasket might be needed upon assembly. Set
the oil return line in the parts container, and thread the bolts back into the CHRA, and plug
the oil hole in order to not dirty the inside.
6. Use the large bore snap pliers to remove the snap ring holding the compressor from the
CHRA.
7. Before removing the compressor wheel, the nut must be marked with paint in relation the
CHRA. When being assembled the markings must be lined up.
8. Using an 8mm socket for the compressor wheel nut and a 12mm socket that fit well enough
to hold the turbine nut, remove the compressor wheel nut and put it in your parts
container. NOTE: the TD04L-13G’s compressor wheel nut is reverse threaded, so go
clockwise.
9. It is recommended that the new compressor wheel be rebalanced alongside the shaft and the
turbine in order to ensure that there will not be any unwanted vibrations. This can be done at
Page 68
53
a local shop for a service fee. The nut should be tightened to the torque specs supplied in the
service manual, roughly 6-8 ft-lbs and lined up with the paint along the CHRA
For reassembly, the following steps may be used.
1. Slide the new compressor wheel back onto the shaft and line up the marks that were marked
on the nut and the CHRA.
1. Clean the oil return line’s gasket surface well and remove all old gasket material. Clean the
gasket surface on the CHRA in the same manner.
2. Reassemble the oil return line with the new gasket. Torque the bolts to 3-4 ft-lbs.
3. Reassemble the coolant line with the copper washers (one between the pipe and the CHRA,
another between the bolt head and the pipe). Torque the bolts to 22 ft-lbs.
4. Reattach the wastegate actuator/controller and vacuum hoses. Give the bolts a “good snug”
but don’t overtighten them, as they are threaded into the semi-soft aluminum compressor
housing.
8.2 Maintenance of the System
8.2.1 Regular Maintenance
In general, a turbocharger needs little maintenance for the entirety of its lifetime. The
typical lifespan for a turbocharger depends on many factors, such as time spent creating boost,
and any modifications to the engine system that do not favor the loading conditions the turbo was
designed for. In the case of the TD04, the lime time could be rather long, in some cases lasting
more than 180,000+ miles without much maintenance. Since this particular study is only focused
on the compressor, there isn’t any maintenance that could be directly related to this component.
Typically, the maintenance of a turbocharger would result in a complete overhaul of the bearings
and seals, as well as the cleaning of both the turbine and compressor blades.
Page 69
54
Figure 37. Destroyed turbine blade from Subaru turbo. [15]
The image above depicts the damage that could be cause from the failure of the bearings
in a turbocharger due to the lack of maintenance. In the specific case, a banjo bolt which filters
the oil before reaching the bearings was clogged, starving the bearings of oil. This simple
replacement could have prevented this sort of catastrophic damaged that would render a
turbocharger useless.
8.3 Environmental Impact
The impact that is caused by the creation of this prototype is an overall neutral impact.
The amount of fuel that could be saved in the lifetime of the compressor could be equated to the
similar environmental cost of creating this prototype out of a billet of aluminum. If this
compressor were to be created out of 3D printed methods, then the impact could be positive due
to the reduced amount of metal that is used to create each part, as well as a reduced energy cost.
8.4 Risk Assessment
The risk associated with this part is rather large due to the extreme conditions that will be
experienced during the lifetime of the component. The worst case loading is used in order to
evaluate the risk associated during extreme loading of the blades. A harmonic response analysis is
vital in order to evaluate the stresses associated during resonance values. Using all of the CFD
Page 70
55
results and using the Von Mises criteria it was calculated that the resulting impeller had a factor
of safety of 1.68. The harmonic response analysis resulted in a deflection of 0.003187 mm. Which
is less than the deflection experienced from the load the blades bear from the fluid.
Page 71
56
9. Design Experience
9.1 Overview
Designing this compressor wheel is a complex job that required a lot of SolidWorks
experience to be able to make the concept and have it automated so that changing parameters were
easy enough. Creating the housing with ideal dimensions to properly make the simulations that
were needed.
9.2 Standards Used in the Project
There is no association that provide set standards for turbochargers, but there are certain
testing standards that are regulated by the Society of Automotive Engineers (SAE). There is a
manuscript that is available for the price of $70.00 to have a detailed description of all the testing
standards and regulations that need to be followed.
ASTM standards were also used when deciding on the type of material that could be
machined for the concept design. The proper regulations need to be followed when testing the
safety of this material.
9.3 Contemporary Issues
The contemporary issues that are being addressed in the scope of this project deal with
fossil fuel dependency and improvement of fuel economy in cars. The downsizing of engines
require turbochargers to allow for more power buildup, while still providing great fuel economy.
The dependency of fossil fuels is something that needs to be addressed and be helped along
the way. Many car manufacturers are going through research stages of electric and hydrogen fuel
alternatives for cars. Until that research can be validated and developed fossil fuels are still being
used, thus creating the demand for better fuel economic cars. Car manufacturers decide to go with
Page 72
57
turbocharged engine systems to be able to have smaller engines in cars. They want to provide
maximum power without sacrificing fuel economy as much.
9.4 Impact of Design in a Global and Societal Context
The demand for turbocharger technology is increasing globally and in conjunction with the
demand for increased fuel economy. Since this demand is growing globally, the impact that the
project has is relevant in today’s global market. The impact that this project has on everyday
society is that people normally don’t want to sacrifice performance for fuel efficiency. Car
manufacturers are attempting to bridge that gap to provide drivability enjoyment and still be
environmentally-conscious.
9.5 Professional and Ethical Responsibility
As an engineer there are many responsibilities that have to be upheld. It is understood that
as a group of people who design things that are being utilized by people every day, worldwide, all
things have to be safe for consumer use. With this project it affects how society commutes and
thus needs to be impeccable when it comes to performance. The project objective would be to
make sure that the design and use of this product will be safe for people’s cars and perform as
advertised.
9.6 Life-Long Learning Experience
The full research and development of this project has shown that theories and information
is always changing. Existing theories can always be applied in new and different ways and
information is always being adapted into new technologies. Life-long learning is paramount to
being a well-rounded individual with understanding of all modern-day engineering phenomena.
Page 73
58
9.7 Discussion
This project is suitable even for professional engineers, since as stated, optimization is a
very complex and arduous process. It is challenging for undergraduate and post-graduate
engineers that are looking to further their understanding of a common mechanical component that
is being used more and more in today’s market. It delves deeply into advanced fluid dynamics
phenomena that remains a mystery even amongst today’s engineers. It allows an engineer to
challenge the understanding of fluid dynamics, manufacturing abilities, and manufacturability
assessment.
Page 74
59
10. Conclusion
10.1 Conclusion and Discussion
The motivation is the biggest driving force for the passion of optimizing the compressor.
Being able to optimize this technology to make it optimal for everyday use is something that can
produce a big demand. Fuel economy is a rising problem that more and more people are becoming
aware of. Society is becoming more aware of being environmentally friendly and this technology
could provide a different option that allows expanding into a field to create more efficient engines.
Improving compressors is a big step on the fuel consumption field, since it will make an already
efficient system even more efficient, this technology will for sure impact today’s markets, and
making a change is a big possibility and with the proper study and development.
10.2 Evaluation of Integrated Global Design Aspects
As been noted in all global design aspects, turbocharger technology is a huge increasing
global market. There will be an increasing demand for it in many years to come. This demand
creates a supply for engineers that understand the concepts behind how turbochargers work.
Turbocharger technology will always be around as long as there is a dependency for fossil fuels.
10.3 Evaluation of Intangible Experiences
When taking on a new project there is always an uneasy feeling of where to start or how.
In engineering there is an abundance of information about physical phenomena, and yet among all
that information there is still a vast expanse of things yet to be discovered. In the midst of this
whole project, the team has been subjected to tremendous pressure, obstacles, and satisfaction.
Being that the team consists of a group of engineers that live, eat, and breathe automotive sports
performance, this project exudes feelings of familiarity. Turbocharger technology is used a lot in
the racing performance world, and being able to research ways to provide the excitement that
Page 75
60
comes with modifying a car and enhancing its performance, while still achieving a clean conscious
among environmental standards is extremely gratifying. There is a deep sense of pride and joy
when recognizing that a lot of what has been learned in the length of the team’s collegiate career,
can be used to fuel a passion that burns from within. When something that is enjoyed as a hobby
can actually be crafted into a career, it is in most times, unheard of. There is a quote that resonates
all these feelings in one simple sentence, “If you do what you love, you’ll never work a day in
your life”.
10.4 Commercialization Prospects of the Product
The feasibility assessment of the project explains how there is a huge demand from this
market and many chances to grow. Following the feasibility assessment and pursuing turbocharger
companies, the commercialization future of this project is very bright. DieselSite Wicked Wheels
is a company that would be the first of many that is looked at for commercializing this project’s
idea and design. There are many turbocharger companies that this idea can be pitched to such as
Garrett, Honeywell, Cummins, Caterpillar, Mitsubishi, Borg Warner, Bosch, Turbonetics, and
many more.
10.5 Future Work
In retrospect, there are a lot of areas where things can definitely be done more intensively,
such as the simulations. It is common knowledge that when dealing with Computational Fluid
Dynamics (CFD) there are better simulation software programs than SolidWorks. ANSYS is a
software that can run intensive CFD analysis and was not used in this project’s scope. There are
also many companies that can evaluate the overall design of the compressor wheel design, such as
CFturbo Software and Engineering. When looking for machining possibilities, DieselSite Wicked
Page 76
61
Wheels could be a possibility that can deliver a quality product. More importantly, the biggest
obstacle that was faced was funding and it is something that could be improved.
Page 77
62
References
1. Whitfield, A. & Baines, N. C. (1990). “Design of Radial Turbomachinery.”, Longman
Scientific and Technical.
2. Aungier, Ronald H., (2000). “Centrifugal Compressors: Strategy for Aerodynamic Design
and Analysis.”, American Society of Mechanical Engineers.
3. Goria, Rama S.R (2003). “Turbomachinery: Design and Theory.”, Decker Mechanical
Engineering
4. Robinson, Chris & Hutchinson, Brad. (2013). “Turbo compressor optimization.”, ANSYS
Inc.
5. Zangeneh, M., Nikpor, B., Watanabe, H. (2010). “Development of a high performance
centrifugal compressor using a 3D inverse design technique.”, Cummin Turbo
Technologies Ltd.
6. Ibaraki, S.,Tomita, I., Ebisu, M., Shiraishi, T. (2012). “Development of a Wide-range
Centrifugal Compressor for Automotive Turbochargers.”, Mitsubishi Heavy Industries
Technical Review Vol.49 No.1.
7. Zhou, Yayun. (2010). “Optimization with Ruled Surface.”, Logos Verlag.
8. Wilcox, Ed. (2000). “Practical Methods For Field Performance Testing Centrifugal
Compressors.”, Conoco, Inc.
9. Garret by Honeywell. (2012). "Wheel Trim." Retrieved September 20, 2014, from:
http://www.turbobygarrett.com/turbobygarrett/wheel_trim
10. U.S. Department of Energy. "Energy Efficiency & Renewable Energy." 2012 Annual
Merit Review, Vehicle Technologies Program. (2012).
Page 78
63
11. National Highway Traffic Safety Administration (NHTSA). "CAFE-Fuel Economy."
Retrieved November 20, 2014 from: http://www.nhtsa.gov/fuel-economy
12. Bell, Corky (1997). "Maximum Boost: Designing, Testing, and Installing Turbocharger
Systems", Bentley Publishers.
13. ©Stage Infinity. "Mitsubishi TD04-13T Turbo Rebuild." Retrieved November 23, 2014
from: http://www.stage-infinity.com/2007/09/28/mitsubishi-td04-13t-turbo-rebuild/
14. Honeywell Turbo Technologies (2014). "2014 Global Turbo Forecast." Retrieved
November 20, 2014 from: http://turbo.honeywell.com/whats-new-in-turbo/2014-turbo-
forecast/
15. 1A Blog (2012). "Is your turbo Subaru on borrowed time?." Retrieved November 25,
2014 from: http://blog.1aauto.com/2012/05/30/got-subaru-turbo-read/
Page 79
64
Appendices
A. Detailed Engineering Drawings of All Parts, Subsystems and Assemblies
Figure 38. Technical drawing for concept design.
Page 80
65
B. Copies of Used Commercial Machine Element Catalogs
Figure 39. Mitsubishi TD04-13T Turbo Rebuild.
Page 81
66
C. Detailed Raw Design Calculations and Analysis
Figure 40. Adaptation of Bernoulli’s Equation.
Table 7. Full tabulated data with simulation results of 90 random generated designs.
BSR TSR Angle BSPy BSPz TSPy TSPz InitialAngle Average V Max V
1 1.3 3.1 88 10.875 -13.1 16.48 -18.4 56 116.111 701.408
2 1.3 2.2 94 10.625 -13.3 16.09 -17.7 26 105.905 442.422
3 2.8 3.7 79 10 -14.3 16.74 -19.45 56
Page 82
67
4 3.4 1.3 73 11 -13.7 15.83 -20.15 44 122.956 761.344
5 1.9 2.8 88 10.375 -12.7 16.48 -18.4 74 111.054 458.486
6 2.2 1.9 85 10.5 -13.5 16.61 -17.35 62 117.615 556.984
7 3.7 3.4 100 9.875 -12.5 15.96 -19.1 32 103.272 456.158
8 3.1 2.5 91 10.875 -13.1 16.87 -19.8 80 117.91 455.994
9 1.6 4 76 10.25 -14.1 16.22 -18.05 50 120.802 411.994
10 1.6 1.3 88 9.875 -14.3 16.35 -17.35 80 118.522 457.038
11 3.1 2.8 73 10.5 -13.3 17 -19.1 50 120.079 608.834
12 4 1.9 82 10.125 -12.9 15.96 -20.15 68
13 2.5 3.4 97 10.75 -13.9 16.61 -18.4 38 101.283 735.641
14 2.2 2.2 79 10 -12.7 16.48 -17 44
15 3.7 3.7 94 10.625 -13.7 15.83 -18.75 74 97.965 628.581
16 2.8 1.6 100 10.25 -14.1 16.87 -19.8 32 108.284 525.369
17 1.3 3.1 85 10.875 -13.1 16.22 -18.05 62 119.712 441.177
18 1.3 1.6 73 10.875 -12.7 15.83 -19.45 74 128.399 452.815
19 2.8 3.1 88 10.25 -13.7 16.48 -17.7 44 109.733 720.78
20 3.7 2.5 97 10.5 -14.3 16.22 -18.4 56 110.661 487.945
21 2.2 4 82 9.875 -13.3 16.87 -20.15 26 110.657 463.037
22 2.5 2.2 91 11 -13.9 17 -18.75 50 111.101 492.726
23 4 3.7 76 10.375 -12.9 16.35 -17 80 119.812 413.706
24 3.4 1.3 85 10.75 -12.5 16.74 -18.05 38 120.122 833.287
25 1.9 2.8 100 10.125 -13.5 16.09 -19.8 68 105.807 459.986
26 1.6 2.5 79 10.375 -13.9 15.96 -19.1 32
27 3.1 4 94 11 -12.9 16.61 -17.35 62 105.841 521.627
28 4 1.9 88 10 -13.3 16.35 -18.4 44 105.841 521.627
29 2.5 3.4 73 10.625 -14.3 17 -20.15 74 118.834 500.628
30 1.9 1.6 97 10.125 -13.1 16.74 -18.75 56 109.64 520.03
31 3.4 3.1 82 10.75 -14.1 16.09 -17 26 114.942 432.097
32 2.8 2.2 76 9.875 -13.5 16.48 -17.7 68
33 1.3 3.7 91 10.5 -12.5 15.83 -19.45 38 114.295 475.831
34 1.3 1.3 82 10 -14.1 16.87 -18.4 62 121.638 511.517
35 2.8 2.8 97 10.625 -13.1 16.22 -20.15 32 108.102 508.753
36 3.4 2.2 88 10.25 -12.7 16.61 -19.45 74 121.634 668.243
37 1.9 3.7 73 10.875 -13.7 15.96 -17.7 44 121.043 689.082
38 2.5 2.5 100 9.875 -12.9 15.83 -18.05 26 99.603 454.196
39 4 4 85 10.5 -13.9 16.48 -19.8 56 116.071 590.602
40 3.1 1.6 79 10.125 -14.3 16.09 -19.1 38 120.805 778.342
41 1.6 3.1 94 10.75 -13.3 16.74 -17.35 68 107.114 459.192
42 1.9 1.9 73 10.5 -12.9 16.87 -18.4 44 123.855 792.368
43 3.4 3.4 88 9.875 -13.9 16.22 -20.15 74 117.226 673.151
44 4 1.3 94 10.875 -13.5 16.48 -19.1 26 117.717 449.969
Page 83
68
45 2.5 2.8 79 10.25 -12.5 15.83 -17.35 56 116.63 463.709
46 2.2 1.6 91 10.75 -13.7 16.09 -17.7 80 119.856 586.095
47 3.7 3.1 76 10.125 -12.7 16.74 -19.45 50 119.006 632.36
48 3.1 2.2 82 11 -13.3 16.35 -18.75 62 118.388 459.318
49 1.6 3.7 97 10.375 -14.3 17 -17 32 96.856 520.948
50 1.3 2.5 88 10.125 -13.5 16.74 -20.15 50 119.094 433.789
51 2.8 4 73 10.75 -12.5 16.09 -18.4 80 120.295 454.629
52 3.7 1.6 82 9.875 -13.1 17 -17.35 38 120.052 807.531
53 2.2 3.1 97 10.5 -14.1 16.35 -19.1 68 106.4 479.721
54 2.5 1.3 76 10.375 -13.3 16.22 -19.8 74 125.066 503.446
55 4 2.8 91 11 -14.3 16.87 -18.05 44 114.898 514.692
56 3.4 1.9 100 10 -13.9 15.96 -17 62 109.791 528.285
57 1.9 3.4 85 10.625 -12.9 16.61 -18.75 32 120.986 488.721
58 1.6 1.9 97 11 -12.5 16.48 -19.8 56 123.958 481.339
59 3.1 3.4 82 10.375 -13.5 15.83 -18.05 26 124.317 504.685
60 3.7 2.5 76 10.75 -13.9 16.74 -17.35 68 129.993 496.261
61 2.2 4 91 10.125 -12.9 16.09 -19.1 38 121.09 510.223
62 1.9 2.2 85 10.875 -14.1 16.35 -19.45 32 107.536 431.058
63 3.4 3.7 100 10.25 -13.1 17 -17.7 62 98.43 493.191
64 2.8 1.3 94 10.625 -12.7 15.96 -17 50 119.38 476.699
65 1.3 2.8 79 10 -13.7 16.61 -18.75 80 126.109 479.253
66 1.3 1.9 94 10.375 -13.1 16.61 -18.05 68 105.54 429.295
67 2.8 3.4 79 11 -14.1 15.96 -19.8 38 122.135 440.775
68 3.4 2.5 85 10.125 -13.7 16.87 -18.75 80 119.732 471.428
69 1.9 4 100 10.75 -12.7 16.22 -17 50 105.888 530.54
70 2.2 2.2 76 10.25 -13.9 16.35 -18.4 32 124.881 475.854
71 3.7 3.7 91 10.875 -12.9 17 -20.15 62 111.132 446.68
72 3.1 1.3 97 10 -13.3 16.09 -19.45 44 117.43 516.313
73 1.6 2.8 82 10.625 -14.3 16.74 -17.7 74 128.095 470.584
74 1.9 2.2 100 10.875 -13.9 16.61 -17.7 44 107.772 460.685
75 3.4 3.7 85 10.25 -12.9 15.96 -19.45 74 119.21 677.719
76 4 1.6 76 10.5 -12.5 17 -18.75 26 122.234 452.31
77 2.5 3.1 91 9.875 -13.5 16.35 -17 56 107.992 451.221
78 2.2 1.3 79 11 -12.7 16.22 -18.4 80 127.881 484.661
79 3.7 2.8 94 10.375 -13.7 16.87 -20.15 50 109.894 662.817
80 2.8 1.9 88 10.625 -14.3 15.83 -19.1 62 117.992 446.076
81 1.3 3.4 73 10 -13.3 16.48 -17.35 32 126.001 453.217
82 1.6 1.9 82 9.875 -12.5 16.74 -19.8 50 120.193 826.157
83 3.1 3.4 97 10.5 -13.5 16.09 -18.05 80 108.432 621.51
84 3.7 1.3 91 10.25 -14.1 16.48 -17 32 109.192 509.27
85 2.2 2.8 76 10.875 -13.1 15.83 -18.75 62 114.489 462.523
Page 84
69
86 2.5 1.6 94 10.125 -14.3 15.96 -20.15 68 111.776 497.388
87 4 3.1 79 10.75 -13.3 16.61 -18.4 38 117.588 444.523
88 3.4 2.5 73 10.375 -12.7 16.22 -17.35 56 118.789 459.106
89 1.9 4 88 11 -13.7 16.87 -19.1 26 109.487 462.956
90 1.6 1.6 76 10.625 -13.5 17 -19.45 56
D. Project Photo Album
Figure 41. Compressor wheel from used turbo.
Page 85
70
Figure 42. Top view of 3D printed acrylic.
Figure 43. Stock compressor wheel alongside 3D printed ABS plastic final design.
Page 86
71
Figure 44. Rendering of final design.
Figure 45. Used compressor wheel, 3D printed ABS plastic, 3D printed polyjet acrylic.
Page 87
72
Figure 46. Clearance between final design and housing.
Figure 47. Measurement with L ruler.
Page 88
73
Figure 48. Isometric view of final design in 3D printed acrylic.
Figure 49. First clearance test with rough 3D printed ABS plastic.
Page 89
74
Figure 50. First clearance test.