Topology Optimization of a Jet Engine

Bracket using Solidworks

6101MECH Engineering Project

Final Report

Name: Ahmed Hashhash

Student Number: 841334
Supervisor: Dr. Glynn Rothwell
Programme: Mechanical Engineering BEng(Hons)
Date: 22/3/2021

School of Engineering
Faculty of Engineering and Technology
Byrom St, Liverpool, L3 3AF, UK
6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Acknowledgements
First and foremost, I would like to extend my deepest gratitude to my supervisor Professor
Glynn Rothwell who guided me through theoretical and technical aspects of the subject in the course
of this study. His invaluable advice provided means for tackling complications encountered in the
course of this study. Without his persistent help and dedicated involvement throughout each step
taken towards completing this paper, it would not have been completed.
Additionally, I would like to pay my special regards to my family who supported and encouraged me
throughout this study as well as my previous years of academic study. Through consistent
communication and valuable advice, they have been a source of motivation and optimism. This
accomplishment would not have been realized without them.
Lastly, I wish to show my gratitude to my friends who supported me and encouraged me to complete
this paper and overcome my shortcomings.

Ahmed Hashhash 1 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Abstract
Topology Optimization has been fundamental for the design of lightweight mechanical
components since its introduction to Aerospace industry. Its potential as well as rigorous
development resulted in approaches which were computationally sustainable, this led to its
integration into various Computer Aided Design software. Simultaneous to Topology optimization’s
development, Additive Manufacturing technologies have been greatly improving which resulted in
reliable Additive Manufacturing methods. Through the combination of these two technologies,
optimized parts could be fabricated significantly easier than with traditional manufacturing
techniques.
This paper presents a brief comparison between Topology Optimization and other types of
structural optimization. Additionally, it briefly compares between Additive Manufacturing and other
manufacturing methods. Moreover, a summary of SIMP method which Solidworks utilizes to solve
Topology Optimization problems was presented in this paper, in addition to other theoretical aspects
of the problem. The main aim of the study is to demonstrate Topology Optimization of a Jet Engine
Bracket using Solidworks Simulation. This was achieved through following an organised workflow and
documenting each step in the optimization process. The produced topology was then exported in a
graphics body configuration and was used to rebuild the bracket. The re-built bracket had 50% less
mass than the original bracket and had adequate mechanical performance during FEA simulation for
all given loading conditions. Additionally, this paper presents a discussion of the obtained results
which aimed to identify sources of errors in the study and assess their effect on the performance of
the optimized bracket. Lastly, the paper presents a thorough comparison between feasible
manufacturing methods with regard to their effects on the mechanical properties of the final
bracket.

Ahmed Hashhash 2 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Table of Contents
Acknowledgements .................................................................................................................................. 1
Abstract .................................................................................................................................................... 2
List of Figures ............................................................................................................................................ 5
Nomenclature ........................................................................................................................................... 6
Abbreviations ...................................................................................................................... 6
Notation .............................................................................................................................. 6
1. Introduction .......................................................................................................................................... 7
1.1 Topology Optimization ............................................................................................................. 7
1.2 Additive Manufacturing ............................................................................................................ 9
1.3 GE Engine Bracket Challenge .................................................................................................. 10
1.4 Aims and Objectives. .............................................................................................................. 12
2. Background Theory ............................................................................................................................. 13
2.1 Topology Optimization in Solidworks ..................................................................................... 13
2.1.1 SIMP based Topology optimization. ......................................................................... 13
2.1.2 Multi Load Optimization........................................................................................... 13
2.2 Additive Manufacturing of TO parts ....................................................................................... 14
2.3 Yield criteria............................................................................................................................ 15
2.4 Anticipated issues ................................................................................................................... 16
3. Methodology and Workflow ............................................................................................................... 17
3.1 Specification ........................................................................................................................... 17
3.2 Setting up the model. ............................................................................................................. 18
3.3 Initial static study ................................................................................................................... 18
3.4 Pre-processing ........................................................................................................................ 20
3.4.1 Design Domain ......................................................................................................... 20
3.4.2 Boundary Conditions ................................................................................................ 20
3.4.3 Shape control ........................................................................................................... 21
3.5 Optimization ........................................................................................................................... 21
3.5.1 Optimization goals ................................................................................................... 21
3.5.2 Mesh parameters ..................................................................................................... 21
3.5.3 Convergence graphs................................................................................................. 22
3.6 Filtering .................................................................................................................................. 24
3.7 Interpretation ......................................................................................................................... 25
3.8 Re-design ................................................................................................................................ 26
4. Results ................................................................................................................................................ 29

Ahmed Hashhash 3 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

4.1 Configurations of the optimized geometry ............................................................................ 29

4.1.1 Surface body ............................................................................................................ 29
4.1.2 Solid body ................................................................................................................. 29
4.1.3 Graphics body .......................................................................................................... 31
4.2 Simulation results ................................................................................................................... 33
4.3 Results comparison ................................................................................................................ 35
5. Discussion ........................................................................................................................................... 36
5.1 Review of the results .............................................................................................................. 36
5.2 Manufacturability ................................................................................................................... 37
6. Conclusions ......................................................................................................................................... 39
7. Future Work........................................................................................................................................ 41
References .............................................................................................................................................. 42
Appendix A: Technical Drawings ............................................................................................................. 45

Ahmed Hashhash 4 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

List of Figures
Figure 1: Topology Optimization by distribution of isotropic material (Bendsøe and Sigmund, 2003). 7
Figure 2: Simple structural component before and after TO (Garcia-Granada, Catafal-Pedragosa and
Lemu, 2019)............................................................................................................................................. 8
Figure 3: Manufacturing classifications including schematic representation of core advantages and
technical challenges (Leary, 2020). ......................................................................................................... 9
Figure 4: Original geometry of the Engine Bracket (GE jet engine bracket challenge, 2013). .............. 10
Figure 5: Winning optimized bracket designs (Carter, 2014)................................................................ 11
Figure 6: An upright fabricated using EBM (Walton and Moztarzadeh, 2017). .................................... 15
Figure 7: Stages of Topology Optimization (Walton and Moztarzadeh, 2017). .................................... 17
Figure 8: Loading conditions specified by GE (GE jet engine bracket challenge, 2013)........................ 18
Figure 9: Static FEA study of the original Engine Bracket. .................................................................... 19
Figure 10: Convergence data of two optimizations. ............................................................................. 23
Figure 11: Optimized Engine Bracket in filtering phase. ....................................................................... 24
Figure 12: Optimized Engine bracket during Interpretation. ................................................................ 25
Figure 13: Steps to remove material manually. .................................................................................... 27
Figure 14: Optimized graphics and solid body brackets. ...................................................................... 28
Figure 15: Optimized bracket exported in solid body configuration. ................................................... 30
Figure 16: Re-designed optimized bracket with inconsistent geometry. ............................................. 32
Figure 17: Post-optimization FEA static study results. .......................................................................... 34
Figure 18: Performance comparison of a Ti-4V-6Al part based on method of fabrication (Lewandowski
and Seifi, 2016)...................................................................................................................................... 38
Figure 19: Technical Drawing of the achieved bracket. ........................................................................ 45

Ahmed Hashhash 5 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Nomenclature
Abbreviations
AM Additive Manufacturing
CAD Computer Aided Design
CNC Computer Numerical Control
DMD Direct Metal Deposition
DMLS Direct Metal Laser Sintering
EBM Electron Beam Melting
FDM Fused Deposition Modelling
FEA Finite Element Analysis
FOS Factor of Safety
GE General Electric
HIP Hot Isostatic Pressing
HT Heat Treatment
IGES Initial Graphics Exchange Specification
LENS Laser Engineered Net Shaping
RPT Rapid Prototyping Technologies
SIMP Solid Isotropic Material Penalization
SLM Selective Laser Melting
STEP/STP Standard Exchange in Product model data
STL Standard Tessellation Language
TO Topology Optimization
UTS Ultimate Tensile Stress
YS Yield Stress
3D Three Dimensional

Notation
Symbol Description SI Unit
𝜎𝑛 Principal Stress (where n relates to the direction) 𝑃𝑎
𝜎𝑣−𝑀 Von-Mises Stress 𝑃𝑎
𝜎𝑦 Yield Stress 𝑃𝑎
𝜌 Density 𝑘𝑔⁄𝑚3
𝜌𝑚𝑖𝑛 Minimum Allowable Relative density -

Ahmed Hashhash 6 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

1. Introduction
1.1 Topology Optimization
Topology optimization (TO) is a non-parametric optimization tool used in the initial phases of
the design process which gained a great interest from the engineering design community since its
introduction (Sigala, 2020). It is a structural optimization approach usually applied in the preliminary
design phase. Figure 1 illustrates three types of structural optimization and their effects on a similar
structure (Bendsøe and Sigmund, 2003).

Figure 1: Topology Optimization by distribution of isotropic material (Bendsøe and Sigmund, 2003).
Figure 1(a) demonstrates a simple truss structure of equally sized members on the left side
which represents the structure before undergoing optimization. While on the right side, some
members increased in thickness which represents the geometry after optimization. This category of
optimization is referred to as sizing optimization. Sizing optimization is a detailed design optimization
where cross-sectional areas, dimensions and thicknesses of model features are set as design
variables. While figure 1(b) illustrates a structural component with six hollow circles on the left side
of the figure, the right side demonstrates how these holes were optimized. This process is known as
shape optimization, it is a detailed design tool which aims to enhance the performance of a
component through optimizing a geometry that experiences a high stress concentration when a load
is applied (Gao, Zhu and Zhang, 2017). By changing the geometry of targeted features as shown on
the right side of figure 1(b), stress concentration factors were decreased. Lastly, figure 1(c) shows a
solid structural component on the left side of the figure, and the right side illustrates the component
after undergoing TO. TO alters the topology of a mechanical component through material re-
distribution based on set conditions in a specified design domain. Consequently, it results in a
structure that is lighter yet stiff enough to withstand the loadings it is rated for.

Ahmed Hashhash 7 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Due to its potential in various industries where weight reduction resulted in saving energy,
materials, and money. It quickly became popular among the aerospace, aeronautics, and automotive
industries. Due to the rapid development of commercial computational abilities and the rigorous
development of different TO approaches. This technology has been introduced to various Computer
Aided Design (CAD) software. Consequently, it became available to other industries at lower
computational costs. Subsequently, it undergone considerable development which yielded in
multiple algorithms which have been utilized to optimize microstructures as well as macrostructures
of various components. One software that has this feature is Solidworks which was used exclusively
throughout this study. Figure 2 shows an example of an optimized structural component using the TO
tool available in Solidworks Simulation package.

Figure 2: Simple structural component before and after TO (Garcia-Granada, Catafal-Pedragosa and Lemu,
2019).

Ahmed Hashhash 8 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 2(a) shows the original geometry which represents the maximum design domain in a
Topology study. Subsequently, the loads, fixtures and material plot are shown in figure 2(b), and
outline which elements of the component are safe to remove which are blue coloured. While figure
2(c) shows the Optimized geometry after it has been smoothed out using a fine mesh. This example
displays some landmark phases of the process on a simple three dimensional (3D) structural
component, which resulted in removing 50% of the original mass (Garcia-Granada, Catafal-Pedragosa
and Lemu, 2019).

1.2 Additive Manufacturing

Additive Manufacturing (AM) could be defined as the use of fusion, binding, or solidifying
materials to make layers of the material cohere (Abdulhameed, Al-Ahmari, Ameen and Mian, 2019).
Contrary to G and M codes used in most Computer Numerical Control (CNC) machines, AM machines
interpret the model and find the most efficient patterns based on Standard Tessellation Language
(STL) or 3D data of a given CAD model.

Figure 3: Manufacturing classifications including schematic representation of core advantages and technical
challenges (Leary, 2020).

Ahmed Hashhash 9 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 3 outlines the main differences between AM and other manufacturing processes. The
advantages of AM over formative and subtractive manufacturing could be summarized in the
following points (Leary, 2020):
• Achieving complex geometry: the common nature of AM processes which is supplementing
material to make layers that adhere to one another allows for far more complex geometry to
be achieved compared to other fabrication techniques.
• Minimizing waste: AM processes only feed enough material to acquire the geometry
specified in the design envelope which leads to less cost per unit. In some cases, support
material would be required to achieve design features such as long overhangs. This is
considered the only source of waste in most AM applications.
• Distributed production: obtaining a product could be as easy as downloading a part over the
internet and fabricating it on site.

1.3 GE Engine Bracket Challenge

In 2013 General Electric (GE) in association with Grabcad.com released a challenge to
optimize a Jet engine bracket. The original part is shown in figure 4, was machined from Ti-6Al-4V bar
stock and weighs about 2.09 kg. The part was an engine bracket used during engine maintenance and
was not used in flights (Carter, 2014).

Figure 4: Original geometry of the Engine Bracket (GE jet engine bracket challenge, 2013).
About 320 members of the Grabcad community competed in the challenge which resulted in lots of
different and interesting designs (Morgan et al., 2016). The best 8 models which passed the
simulation phase were manufactured and tested.

Ahmed Hashhash 10 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 5: Winning optimized bracket designs (Carter, 2014).

Figure 5 shows the top 8 entries with their rankings, designer name and country. The winner
of the first place was able to reduce 80% of the original weight without compromising the load-
bearing capacity. In addition, some designers chose to optimize the part considering the possibility of

Ahmed Hashhash 11 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

traditional manufacturing methods. However, all optimized brackets were manufactured using Direct
Metal Laser Sintering (DMLS) (Morgan et al., 2016).

1.4 Aims and Objectives.

This paper aims to investigate the capabilities of the Topology Optimization tool in Solidworks
Simulation. Additionally, it aims to identify and address the limitations concerning rebuilding the
optimized geometry. This would result in a stable CAD model without errors which would allow for
further analysis such as FEA. Consequently, values of mechanical properties would be measures
which would provide insight into the optimized part’s performance. The last aim was to revise a
range of possible manufacturing methods to which would help control the mechanical properties of
the fabricated optimized part. These aims could be achieved through attempting the following
objectives:
• Revision of literature and outlining concepts which are relevant to this paper such as Multi-
load Optimization.
• Optimization of the bracket given in GE Engine Bracket Challenge using Solidworks and
providing a detailed guide on the workflow of the TO process.
• Classification of available methods of re-building the optimized geometry and deciding
methods that are viable to use in the rebuild of bracket optimized in this paper.
• Analysis of the mechanical properties of the achieved bracket under static loading conditions
using Solidworks FEA and compare it with the properties of the original bracket.
• Discussion of the results obtained at each stage of the procedure to address inaccuracies
observed in the study and assess limitations relevant to each stage.
• Selection of feasible fabrication methods based on material data obtained from tensile tests
of various manufacturing methods.

Ahmed Hashhash 12 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

2. Background Theory
This section of the study presents relevant research and theory that were used to complete
this study. It serves various purposes as it briefly outlines the history of the problem. Additionally, it
presents examples of AM technologies successfully fabricating topologically optimized parts.
Furthermore, it briefly outlines the relevant yield criterion which aids with interpreting results
obtained in the study. Lastly, it outlines issues regarding the application of TO.

2.1 Topology Optimization in Solidworks

2.1.1 SIMP based Topology optimization.
Topology optimization is the most common type of structural optimization. It is used in the
preliminary design phase to predict the optimal material distribution within a given initial design
domain and considers functional specifications and manufacturing constraints. It was firstly deemed
as a binary discrete problem, which made it computationally expensive to optimize large-scale
models given the explicit use of gradient free algorithms (Xia, 2016). After the introduction of the
Homogenization method in (Bendsøe and Kikuchi, 1988), it was realized that no general solution
could be obtained to the problem unless material distribution was taken into account. However, the
introduction of Solid Isotropic Material Penalty for intermediate densities (SIMP) in (Bendsøe, 1989)
which utilized density and other material properties allowed the optimization to be relatively less
difficult to compute.
Solidworks TO uses SIMP method which utilizes a continuous relative density distribution
function for discretized elements (2020 SOLIDWORKS Help - SIMP Method for Topology
Optimization, 2020). This aids with overcoming the binary nature often associated with the results
and enables the software to assign intermediate densities for each element, the value of those
densities varies between ρmin and 1 where ρmin represents the lowest allowable relative density value
of a given element. Where elements with a relative density of 1 are considered critical elements and
cannot be removed. While elements with a relative density below ρmin are void elements and are
removed from the design domain. Finally, elements with a relative density between ρmin and 1 are
categorized as porous elements. The algorithm used in this method which generally follows the
power law introduces the penalty factor which minimizes the contribution of porous elements to the
total stiffness of the model. (Rozvany, 2001) presents a thorough review of SIMP method which
discusses the evolution, performance, and formulation.

2.1.2 Multi Load Optimization.

Solidworks TO provides a multi load case tool which allows separate loading conditions to be
considered during the optimization. This tool utilizes Min/Max Formulation which attempts to

Ahmed Hashhash 13 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

maximize the stiffness of the model in each load case independently then summate the result of all
the load cases to obtain the total stiffness. In this case, the optimized geometry achieves minimum
compliance to each load case that acts independently (2020 SOLIDWORKS Help - Min Max
Formulation for Topology, 2020).

2.2 Additive Manufacturing of TO parts

The issue with Topology optimization lies in the irregular and complex geometry produced by
the process. In the past, it has been extremely difficult to achieve the complex geometric features
resulted by TO using traditional machining and manufacturing techniques. Which in turn made it
difficult to obtain experimental results of Topology optimized parts. However, with the improvement
several additive manufacturing techniques such as Fused Deposition Modelling (FDM) and Rapid
Prototyping Technologies (RPT), it has been significantly easier to achieve those complex geometries.
(Walton and Moztarzadeh, 2017) presents an example on the optimization of a formula
student car upright. Where the authors achieved a 36% weight reduction of the original upright and
successfully fabricated the optimized upright using Electron Beam Melting (EBM). It is also notable to
mention that a pair of these uprights were fabricated in a single run using the ARCAM A2X EBM
machine. Figure 6 illustrates the manufactured upright. It is obvious that the upright is shown in the
same orientation of build since it is fully resting on the support material. The amount of support
material this part required seemed to be excessive yet justifiable due to various overhangs and
detailed features in the optimized geometry. As a result, this upright has been machined to separate
the support material and end up with the required part. Therefore, it is important to recognize the
amount of support material needed during additive manufacturing processes and aim to minimize it
as it can save time, material, and money. These findings demonstrated a successful application of AM
techniques to achieve topologically optimized parts.

Ahmed Hashhash 14 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 6: An upright fabricated using EBM (Walton and Moztarzadeh, 2017).

2.3 Yield criteria

The conclusion of generalized mathematical formulae that estimate the initial plastic yielding
conditions of a given ductile material element is difficult when compound stresses are accounted for
(Dieter, 1961). Several methods have been proposed to ameliorate this problem however the most
famous are Maximum-Shear-Stress Theory and von-Mises or Distortion-energy, Theory. While both
theories aim to present a mathematical model that combines the principal stresses in a way that
simplifies stress combinations, von-Mises criterion is commonly accepted as the better expression of
strain distortion energy. Solidworks Simulation default Stress plot is set to von-Mises stress although
other types of stress could be plotted such as Principal stress, Directional stress, and Normal stress.
The formula of von-Mises Stress is given by:

(𝜎1 − 𝜎2 )2 + (𝜎2 − 𝜎3 )2 + (𝜎1 − 𝜎3 )2

𝜎𝑣−𝑀 =√ (1)
2

This simplifies design calculations and aids with interpreting results obtained from FEA software as
the failure of a component would be observed if von-Mises stress exceeds the yield strength of the
material. As a result, the Factor of Safety (FOS) is given by the following equation:

Ahmed Hashhash 15 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

𝜎𝑦
𝐹𝑂𝑆 =
𝜎𝑣−𝑀 (2)

Where a FOS of 2 theoretically means that the design can withstand stresses that are twice as much
it was originally designed for. On the other hand, a FOS of less than 1 indicates plastic yielding in the
design.

2.4 Anticipated issues

During research, it was observed that bodies generated by Solidworks topology optimization tool
could only be exported in an STL format. While this format is adequate for most AM applications, it
makes it difficult to edit and manipulate features on the model in a CAD specific software (Gebhardt
and Hötter, 2016). The severeness of this problem increases as the complexity of the generated
Topology increases. In industry, third party additions to Solidworks such as 3DXpert are used to
manipulate STL based files (Geomagic - 3DXpert for Solidworks, 2013). While this is a tool that allows
for better control of STL files, models rendered using software like this often result in files with large
amounts of data. This in turn makes it difficult to perform further simulation operations since
simulating the model would entail the use of this large amount of data in the simulation. Since this
study required the use of FEA to verify results of the optimized topology, this problem had to be
eliminated. Consequently, an indirect approach was taken to achieve the optimized solid body
bracket while only using the tools that are native to Solidworks.

Ahmed Hashhash 16 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

3. Methodology and Workflow

This section of the study outlines the steps performed to topologically optimize the GE
Engine Bracket shown in figure 7. Despite having similar simulation tools with other types of analyses
in Solidworks, TO has features that are not available in those analyses such as shape control. As a
result, the workflow of the process is slightly different from FEA static analysis.

Figure 7: Stages of Topology Optimization (Walton and Moztarzadeh, 2017).

In this paper, the sequence of operation presented in (Walton and Moztarzadeh, 2017) was followed
as it was an efficient and simple to follow. Figure 7 illustrates key steps that are characteristic of TO.

3.1 Specification
GE released a set of specifications to follow during the optimization process. The part chosen
to be optimized was the Engine Bracket shown in Figure 4. The specifications could be summarized in
the following points:
• The use of any CAD software was permitted as long as the submitted part is in STEP or IGES
format.
• The geometry generated by the optimization process must fit the original design envelope.
• The part was made of Ti-6Al-4V and was assumed to possess a yield strength of about 903.2
MPa.
• Interface 1: 0.75-inch diameter pin. The pin was to be considered infinitely stiff for analysis
purposes.

Ahmed Hashhash 17 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

• Interfaces 2–5: 0.375-24 AS3239-26 machine bolt. Nut face 0.405 in. max ID and 0.558 in.
min OD. The bolts were to be considered infinitely stiff.
• the loading conditions were given as shown in figure 8:

Figure 8: Loading conditions specified by GE (GE jet engine bracket challenge, 2013).
Figure 8 shows the bracket connected to a flat plate at interfaces 2-5 which was accounted for in the
optimization process.

3.2 Setting up the model.

The STP model given in the challenge was not aligned with the global co-ordinate system so,
it had to be re-oriented. Re-orienting the model helps during the re-building phase. In addition, if
gravity condition were to be included in later analysis, it would affect the model relative to the global
co-ordinate system. Despite the availability of various methods to re-orient the model, a simple
sketch was drawn with the appropriate relations to the features the global co-ordinate system in the
software should be aligned with. Subsequently, a co-ordinate system was inserted using this sketch
as a reference to its axes. Then, the part was exported as an STP file with the new co-ordinate system
set as the global co-ordinate system. Lastly, the exported file was saved and re-opened showing the
part in the correct orientation.

3.3 Initial static study

Prior to the Topology Optimization, a FEA static study was conducted to obtain some insight
into the bracket’s performance under the loading conditions given in the challenge. This was done
using the load case manager tool which allows the analysis of multiple loads independently or
collectively. The mesh generated for the model in this study was chosen to be a very fine mesh since

Ahmed Hashhash 18 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

it helped with obtaining more converged results. Figure 9 demonstrates the results of the four
loading conditions with an exaggerated deformation scale.

Figure 9: Static FEA study of the original Engine Bracket.

Ahmed Hashhash 19 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 9(a) shows the von-Mises stress plot of the bracket when subjected to the first loading
conditions illustrated in figure 8, the maximum von-Mises stress experienced by the model was 413
MPa. While figure 9(b) shows the bracket under the second loading conditions resulted in a
maximum von-Mises stress of 279 MPa. Figure 9(c) illustrates the third loading conditions which
resulted in the largest von-Mises stress plotted 514 MPa, this indicates that these results are sensible
since the 9500 lb (42.258 kN) force included in this loading scenario was the largest. Figure 9(d)
shows the last loading condition of the challenge which resulted in a maximum von-Mises stress of
291 MPa. Furthermore, a factor of safety plot aided with obtaining the values of the minimum factor
of safety calculated in each loading scenario which were used as references to some constraints in
the optimization.

3.4 Pre-processing
3.4.1 Design Domain
The design domain represents the maximum geometry the software can optimize. In this
case, it was the engine bracket. In addition, general optimization options were reviewed at the
properties window of the study. The software was set to keeping material at areas of forces and
fixtures. However, it was left free to optimize areas of global contact. Despite this serving as the best
setup for the optimization over various trials, it meant that all interfacing bodies had to be replaced
with appropriate boundary conditions later in the study.

3.4.2 Boundary Conditions

This part of the process is of great importance since it affects the output directly and
significantly. The first boundary condition to be set was defining the material of the engine bracket,
which in this study was Ti-6Al-4V and could be found under Titanium alloys in Solidworks materials
library. Secondly, the horizontal plate beneath the bracket was defined as a Rigid virtual wall
coincident to the base of the bracket. Subsequently, the bolts were defined as fixed geometry that
prevents movement in all degrees of freedom at the locations of the bolt holes. This simplifies the
study as no pre-loading information was given for the bolts, which affect the outcome of the study
significantly. In addition, fixed geometry was chosen because it ensured that no material would be
removed from the bolt holes.
In order to replicate the forces, the circular faces of the pin holes were split into 4 separate faces on
each hole. This was done by using a 3D sketch to draw lines where the circles needed to be split and
using the split line feature to split the circular faces within the pin holes. Additionally, forces were
grouped appropriately into four loading scenarios using the multi load case tool.

Ahmed Hashhash 20 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

3.4.3 Shape control

In addition to the general optimization options outlined in (3.4.1), Solidworks offers features
for direct shape control which are preserved regions, De-mold direction, thickness control and
symmetry condition. However, in this study only the preserved regions tool has been used since the
part is not symmetrical and could be manufactured using additive manufacturing. The member
thickness could be reviewed during the re-design phase to ensure the minimum member thickness
meets the required metric in the challenge. The preserved regions chosen were the bolt holes and
the surrounding circular faces and the pin holes where the forces were applied.

3.5 Optimization
3.5.1 Optimization goals
Optimization goals are the most important factor during the optimization since they define
the result. Solidworks offers three types of TO goals which are best weight to stiffness ratio, minimize
maximum displacement, frequency optimization and minimize mass with displacement constraint. In
this study only best weight to stiffness goal was applied. This goal utilizes an algorithm that yields the
shape of a component with the largest stiffness considering the given amount of mass that will be
removed from the initial maximum design space. The algorithm achieves this result through
minimizing global compliance in the model, compliance is simply the reciprocal of stiffness and
represents the sum of strain energies in all elements (2020 SOLIDWORKS Help - Topology Study,
2020). A weight reduction target of 80% was chosen based on results of the winner of the GE
challenge.
Additionally, a global FOS constraint with a value of 2 was applied. This value was estimated
as an average to the minimum FOS values obtained from the initial static study since independent
FOS cannot be applied to individual load cases in Solidworks TO. This constraint utilizes an algorithm
that seeks to maintain the specified FOS if the optimized under all loading scenarios.

3.5.2 Mesh parameters

Solidworks offers various features and options for meshing which are influential to the
optimization process as well as the convergence of the specified goals. However, mesh refinement
varies directly with the computational power required and the total time of the optimization. In
addition, the mesh used in Solidworks TO is programmed to have smaller mesh elements than
normal FEA studies. In this study, it was decided to begin with a default mesh. Additionally, the
automatic transition feature was enabled which allows the software to refine the mesh generated on
features such as ribs, fillets, and holes.

Ahmed Hashhash 21 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

3.5.3 Convergence graphs

During the optimization process, the iterative solver displays the number of iterations
calculated as well as the elapsed time. In addition, convergence graphs of each optimization goal and
constraint could be monitored during the optimization. This provides an insight into the results and
quality of optimization by reviewing targeted values and their convergence over several iterations. In
addition, monitoring convergence graphs can indicate an error early in the study which is important
considering the long-time of the process. The solver could then be stopped to rectify the error and
re-run the optimization.

Ahmed Hashhash 22 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 10: Convergence data of two optimizations.

Figure 10(a) shows the convergence data of the optimization when a coarse mesh was used,
green and red colours indicate whether results of a specific goal have converged throughout the
optimization or failed to converge. Since there was no target value for the stiffness goal, the results
from this goal have converged. However, the mass target could not be met and resulted in a 76%

Ahmed Hashhash 23 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

mass reduction on the best iteration. Additionally, FOS constraint for each Loading condition were
presented in order below the mass target. While FOS constraints for the first two loading scenarios
converged at the specified value of 2, they have not converged for the last two loading scenarios.
Figure 10(b) illustrates the convergence data when a default mesh was used. While stiffness result
was improved slightly than when coarse mesh was used, all FOS constraints have converged. This
indicates that the optimized geometry now has a FOS of 2 in all Loading scenarios. Despite not
reaching the targeted mass, 73% mass reduction was achieved using a default mesh which was
acceptable.

3.6 Filtering
After the optimization problem was solved, a plot was generated in the design domain as
shown in figure 11.

Figure 11: Optimized Engine Bracket in filtering phase.

Figure 11 illustrates a wireframe sketch; this sketch is displayed as a reference to the maximum
design domain to indicate areas where material could be removed. The slider on the left-hand side
allows navigation between iterations calculated during the optimization which are ordered by
weight. Selecting different iterations results in changing the colours of the plot in the design domain.
The Calculated Element Mass outlines the mass of the optimized geometry based on a chosen
iteration. In addition, figure 11 shows the percentage of original mass which indicates that 73% of
the bracket’s mass was removed during the optimization. The material mass plot demonstrates
yellow areas which are critical and cannot be removed. While blue coloured areas can be removed

Ahmed Hashhash 24 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

but would decrease mass and stiffness of the part. The legend shown on the right side of the bracket
refers relative mass densities of different elements with yellow being 1 and blue being 0.3 (2020
SOLIDWORKS Help - Material Mass PropertyManager, 2020). This stage is important as it allows for
reduction of elements to be included in the final model and in turn, total mass. Additionally, it is
necessary to monitor for any elements that are separated from the brackets main body. As these
elements could form entities which would not contribute towards the stiffness and would increase
the total mass. This is referred to as checkerboarding effect and indicates instability in a given
iteration (Bendsøe and Sigmund, 2003).

3.7 Interpretation
After filtering the results and choosing a suitable iteration, the calculate smoothed mesh tool
was used to create a smoother surface mesh as shown in figure 12.

Figure 12: Optimized Engine bracket during Interpretation.

Figure 12 shows the optimized geometry after an extremely smooth mesh was applied which helped
interpret how features throughout the bracket were optimized. These optimized features seemed
irregular throughout the model with some features resembling organic structures. However, it was
expected that the optimized geometry might not be as aesthetic as the original. Overall, the
optimized geometry seemed to have the same general idea of the top 8 winners in the challenge
(refer to figure 5).

Ahmed Hashhash 25 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

3.8 Re-design
Following filtering and interpretation, the optimized geometry could be exported as a
graphics body, solid body, or surface body. For additive manufacturing applications, exporting the
optimized geometry as a solid body would be adequate for most prototyping applications. However,
the feature recognize tool in Solidworks would not be able to identify features which are irregular
and would result in a mesh based solid body where each mesh element is a face. This made it
difficult to edit unwanted features as well as preparing the optimized geometry for further analysis.
On the other hand, surface bodies only export the surface geometry of the smoothed mesh data
obtained in the interpretation phase.
Despite the availability of various re-design methods, the general method of superimposing a
graphics body on the original bracket was chosen. This method was chosen based on several factors
including availability of various tools to edit the original geometry, ease of application and low
computational requirements. However, the major limitation to this method is that only a rough
representation of the optimized geometry. Figure 13(a) illustrates the optimized graphics body
superimposed on the original bracket.

Ahmed Hashhash 26 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 13: Steps to remove material manually.

After overlaying both geometries, several contours were sketched at areas where both
bodies did not overlap as illustrated in figure 13(b). Subsequently, these contours were used to

Ahmed Hashhash 27 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

remove material from the original body using the extruded cut tool which resulted in the bracket
shown in figure 13(c). This process was repeated using planes aligned at different angles with respect
to the bracket, which allowed for easy removal of a considerable amount of material. However, this
method has several limitations. Firstly, these extruded cuts could form edges with length that is less
than the allowable tolerance of the software resulting in a geometric inconsistency called short
edges. This can lead to undesirable features such as broken features and edges which can hinder the
meshing process in later analysis. Consequently, only few extruded cuts were made while constantly
monitoring for short edges using the check tool in the evaluation tab. After 6 extruded cuts using
different planes, the bracket’s topology roughly resembled the optimized topology. Subsequently,
fillets of various radii were used to smooth the edges that were resulted by the extruded cuts.

Figure 14: Optimized graphics and solid body brackets.

Figure 14(a) demonstrates the optimized graphics body which was used to rebuild the model.
While figure 14(b) illustrates the final shape of the solid body bracket after it was carved to resemble
the optimized topology. It is apparent that the volume of model 14(a) is much smaller than model
14(b) despite sharing relatively similar topologies. This is due to the limitations associated with this
re-design method which in turn, resulted in less material being removed than expected. The final re-
designed bracket was evaluated to have a mass of 1029 grams which is about 50% of the original
mass (for dimensions refer to appendix). Despite not achieving a final re-designed bracket with the
original mass reduction value intended in the optimization, the percentage of mass removed was
acceptable for further analysis.

Ahmed Hashhash 28 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

4. Results
This section demonstrates the configurations available to export directly from the Topology study
and analysis each configuration in terms of its general applications as well as its relevance to this
paper. Additionally, it demonstrates the results obtained from the FEA static studies performed on
the original bracket as well as the re-built solid body bracket. It compares relevant performance
parameters obtained from said studies.

4.1 Configurations of the optimized geometry

After the interpretation phase, the optimized bracket could be exported as either a graphics
body, solid body or a surface body (2020 SOLIDWORKS Help - Export Smoothed Mesh, 2021). Based
on various factors, other configurations could have been chosen for the re-build phase. This section
discusses these factors and presents a justification for the chosen configuration.
4.1.1 Surface body
Surface bodies only represent the surface data of the smooth mesh and their format is STL
which is used in some additive manufacturing applications. However, it could not be used in this
study as no further FE analysis could be performed on a surface body. therefore, it was excluded
from this study.
4.1.2 Solid body
While this option allows smoothed mesh data to be exported as a solid body, this
configuration presented various challenges while trying to manipulate its features. Solidworks tries
to automatically recognize features which renders broken CAD geometry and identifies standard
engineering features throughout solid geometry. This was not suitable for application on the
optimized solid body due to the large angular deviation between mesh elements. Despite the
availability of interactive feature recognition which provided more manual control of the process, it
failed to work properly due to the same reason. Figure 15 shows the optimized bracket in solid body
configuration and wireframe display.

Ahmed Hashhash 29 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 15: Optimized bracket exported in solid body configuration.

Figure 15 shows the optimized geometry in its solid body configuration, the part in the figure
has undergone automatic feature recognition which managed to merge some facets into
recognizable faces. This mostly occurred at the areas where no material has been removed.
Additionally, automatic feature recognition recognized the bolt holes as extruded cuts as highlighted
in figure 15. this indicates the severity of angular deviation present in this configuration which caused
the holes to be recognized as a sketch closer to an octagon than a circle. Since most mesh data failed
to render into recognizable features during recognition, each smoothed mesh element was now
considered face on the solid body. This resulted in hundreds of broken and short edges all over the
model which are geometric discontinuities. These discontinuities foiled the meshing process of the
part in further FE analysis and had to be repaired manually. Repairing these errors by hand was too
complicated and mundane to perform. Lastly, this configuration file was significantly large which
made its performance slower hence, this configuration was not used to re-build the bracket.
While it was not necessarily of much use in this study, the optimized solid body configuration
provided insight on handling mesh based solid bodies in Solidworks. There are various tools available
in the mesh modelling tab which are useful during editing and manipulating features in this

Ahmed Hashhash 30 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

configuration. However, these tools were not effective for the optimized bracket and seemed more
suitable for less complex models. While this configuration could be used to fabricate the model using
a given additive manufacturing technique, it would be poorly represented due to previously
discussed inaccuracies and could result in a low-quality part only fit for prototyping.

4.1.3 Graphics body

Despite not providing direct means to edit the optimized geometry or perform FE analysis,
graphics bodies provide a good reference for sketching new geometry or editing an existing
geometry. The optimized part was too complicated to sketch from the beginning. Therefore, it was
decided that the best course of action was super imposing the optimized graphics body over the
original bracket and reducing the transparency of the original geometry. This allowed for easy
material removal using extruded cuts as shown in figure 13 and provided more control over the
targeted final optimized solid body. hence, it was the method chosen for re-building the model.
The main limitation to this method was the restricted amount of material that could be
removed from the original part. The first reason to this limitation was the complex geometry of the
optimization which had a big overhanging space at the back of the bracket and other complex
features. Obtaining a geometry on the solid body that identically resembles the geometry of the
optimized part using a graphics body and material cutting was a difficult and time-consuming task. It
was done using more extruded cuts at different angles and directions. Despite yielding the closest
results to the optimized geometry in both shape and weight. It caused more problems similar to the
problems discussed in (4.1.2). Figure 16 illustrates the original solid body after 70 extruded cuts and
123 fillets.

Ahmed Hashhash 31 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 16: Re-designed optimized bracket with inconsistent geometry.

The edited part shown in figure 16 had a mass of 0.627 kg which is significantly close to the
result of the optimization. However, due to the intersection of sketches and features problems such
as short edges, broken faces and invalid faces occurred in the solid body being edited. These
inconsistencies prevented the body being meshed in further analysis. Hence, it was not used for the
final static analysis despite being the best representation of the optimized geometry re-built.
Re-building the model using this method needed to be performed in a much simpler way to
prevent added cuts and features from generating unwanted features in the model. Therefore,
constant monitoring of these problems was required after applying each cut or feature to the original
bracket using the check tool available in the evaluation tool. Additionally, only 6 extruded cuts were
used which were enough to remove 50% of the original bracket’s mass as well as obtain a rough
representation of the optimized topology as shown in figure 14(b). This resulted in a model that had
no inconsistencies or broken features and was used in the post-optimization FEA static study. This
however was a trade off with having almost twice as much mass as the bracket shown in figure 16.
This method resulted in a part that was acceptable for the purposes of this study and had adequate
performance under all loading conditions.

Ahmed Hashhash 32 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

4.2 Simulation results

After achieving the final solid body bracket shown in figure 14(b), a FEA static analysis similar
to the initial static study was performed on the edited bracket to verify its performance. All boundary
conditions used in the initial static study were replicated in this simulation as none of the interfacing
faces was affected. Additionally, mesh parameters used in both FEA studies were extremely fine
meshes to ensure all parameters are constant in both studies. Figure 17 demonstrates the results of
the static analysis performed on the optimized solid body bracket where each load case is shown
independently.

Ahmed Hashhash 33 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Figure 17: Post-optimization FEA static study results.

Figure 17(a) corresponds to the first load case of 8000lb (35.586 kN) and illustrates the von-Mises
stress plot generated throughout the model. The maximum von-Mises stress experienced by the

Ahmed Hashhash 34 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

edited bracket when subjected to this load was evaluated to be 570 MPa which was the highest of all
load cases. von-Mises stress results of the second loading condition are illustrated in figure 17(b) and
the maximum value outlined was 343 MPa. Furthermore, the third load case results are shown in
figure 17(c) and the maximum von-Mises stress in the Model was plotted at 511 MPa. Lastly, the
static torque loading condition results are illustrated in figure 17(d), this case had the lowest plotted
von-Mises stress of all cases at 287 MPa.

4.3 Results comparison

For ease of comparison between the results obtained from both static studies, relevant data
were compiled as shown in Table 1.
Table 1: Relevant results of FEA static studies.
Original bracket FEA static study Optimized bracket FEA static study
Max. von- Minimum Maximum Max. von- Minimum Maximum
Mises stress Factor of Deformation Mises stress Factor of Deformation
(MPa) Safety (mm) (MPa) Safety (mm)
Load 413 2 0.165 570 1.45 0.335
case 1
Load 279 2.96 0.129 343 2.4 0.239
case 2
Load 514 1.6 0.142 511 1.6 0.228
case 3
Load 291 2.84 0.0652 287 2.88 0.0747
case 4

Table 1 outlines the maximum von-Mises stress experienced during each load case before and after
the bracket was optimized. Additionally, minimum FOS values were also added as they provided
insight into the final performance of the bracket. The optimized bracket’s performance under the
first loading condition was negatively affected as it experienced more von-Mises stress compared to
the original bracket. Its minimum FOS dropped from 2 to 1.45 which is significant but not critical. As a
result, it was the lowest FOS calculated in all studies. Furthermore, the optimized bracket’s
performance was hindered compared to the original when subjected to the second loading
condition. It experienced a relatively higher von-Mises stress with 64 MPa more than the original
bracket under the same load. This led to a drop in the minimum FOS from 2.96 to 2.4 between the
original and optimized geometry. On the contrary, the performance of the optimized bracket was
slightly improved under the third and fourth load cases. Additionally, the maximum deformation

Ahmed Hashhash 35 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

results of the optimized bracket were almost double those of the original bracket for the first three
load cases. This could be linked with the tensile nature of the loading conditions as well as the
percentage of weight reduction achieved during optimization. While this could be controlled by
adding a displacement goal in the optimization, it did not seem critical to the performance of the
bracket.
The performance of the optimized bracket in the first two load cases was not improved
compared to the original geometry and the factor of safety was decreased by about 20-26% for each
case. However, the optimized bracket did exceed the yield stress of the material (827 MPa) in any
load case and managed to maintain a minimum FOS of 1.45 in the worst case. Additionally, the
performance of the optimized bracket was slightly better in the 3rd and 4th load cases as the
maximum von-Mises experienced by the bracket was reduced by about 4MPa compared to the
original bracket. These results are acceptable since the mass of the bracket was reduced from 2.09 kg
to 1.03 kg. Hindrance of the bracket’s performance following the optimization could be associated
with inaccuracies during the re-building phase and the location of removed material. Additionally,
the increase of maximum deformation after the optimization was important to monitor as it could
result in a structure that deforms more than it should. However, the maximum deformation
calculated for the optimized bracket was 0.335 mm in the worst case. Therefore, these deformation
results were considered negligible and would not affect the performance significantly. All in all, the
static performance results of the optimized bracket design obtained from Solidworks FEA showed
adequate factors of safety as well as negligible deformation readings. As a result, the optimized
bracket was acceptable, and no further optimization was performed.

5. Discussion
This part of the study was dedicated to critical review to the results obtained in terms of
performance and manufacturability. It aims to identify sources of inaccuracy encountered during the
study and assess their effect on the results.

5.1 Review of the results

One limitation to TO in Solidworks is the absence of an additive manufacturing constraint.
This could be due to the isotropic nature of the process, this is not the case in additively
manufactured parts which are considered anisotropic (Hitzler et al., 2017). As a result, some
unwanted features such as overhangs were characteristic of the optimized bracket. An additive
manufacturing constraint should generally try to avoid these features which could result in a less
complex topology. However, the re-build method allowed for greater control over which features the
author saw fit to incorporate in the final design. As a result, the intensity of overhangs in the

Ahmed Hashhash 36 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

optimized model was reduced and a more stable geometry was achieved. This shows the versatility
of the method used in re-building the model and the control it provides to the designer.
Since TO tool in Solidworks utilizes SIMP method, this could result in very inaccurate
experimental results if the part was to be manufactured using AM. This would also be due to the
anisotropic nature of AM parts and the severity of these inaccuracies would depend on the type and
level of anisotropy in the finished part. Anisotropy could be briefly defined as the difference in
structural behavior when the same load is applied at different directions. For example, (Tempelman,
Shercliff and van Eyben, 2014) state that parts fabricated using powder bed fusion tend to have
different performance in terms of strength and elongation depending on the orientation of the load.
Additionally, the material was assumed to be isotropic during FEA studies. Therefore, any efforts in
testing the bracket would have to account for the anisotropy of the finished bracket.
The result of the Topology study performed in this paper seemed valid since the achieved
topology shared many features of the winning brackets in the GE Engine bracket competition shown
in figure 5. It has been noticed that all wining models had a large portion of material removed from
the back of the bracket which was referred to as ‘open mouth design’ (Morgan et al., 2016).
Additionally, the bracket achieved in this study did not yield when subjected to the service loads
specified by GE. Which aided with performance validation of the optimized topology. Despite the
increase of maximum deformation in the optimized bracket, this increase would not be critical to the
functionality of the bracket under static loading conditions. However, this is under the assumption
that the material is completely isotropic. Therefore, these results could be further verified through
the fabrication of the bracket and testing it under similar operational conditions.

5.2 Manufacturability
In order to continue with field testing, the optimized bracket would need to be fabricated
through either subtractive or additive manufacturing. Since the material used is Ti-4V-6Al, it would
be harder to be machined compared to other alloys due to material properties such as low thermal
conductivity (Pradhan et al., 2019). Additionally, it was very difficult to obtain technical drawings of
the achieved bracket due to the irregular geometry of the edited features. However, the geometry of
the optimized bracket achieved in this study could be vaguely fabricated using 5-axis CNC Milling
machines which would save time compared to a machine that can only move in 3 axes.
On another note, the winning optimized brackets in the GE bracket competition were
manufactured using DMLS. This is important as further comparison between the bracket achieved in
this study and other brackets would require the use of the same fabrication method as each method
results in its own sets of properties for a given material. Furthermore, DMLS or any appropriate AM
process would result in a significantly better overall representation of the optimized geometry which

Ahmed Hashhash 37 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

is the most important aspect of this study. However, this method is usually more costly than
traditionally machined parts. In addition to the post fabrication machining an additively
manufacturing part receives to remove the support material, parts usually have to be sanded due to
the rough surface finishes. This increases cost, time and tooling required to achieve a final product.
(Facchini et al., 2010) states that when pre-alloyed Ti-4V-6Al powders were used to fabricate a part
using selective laser melting (SLM), the resulted part was stronger than its hot worked counter-part
due to the martensitic microstructure. This however resulted in a significant decrease in ductility.
This could be concerning as von-Mises theory could yield inaccurate results while assessing the
fabricated part due to its nature of only evaluating ductile materials. This could be controlled either
by changing the yield criteria to fit the brittle nature of the material or subject the fabricated part to
post-fabrication heat treatment which could result some of the strength achieved to be traded off for
improved ductility.

Figure 18: Performance comparison of a Ti-4V-6Al part based on method of fabrication (Lewandowski and
Seifi, 2016).
(Fraizer, cited in Lewandowski and Seifi, 2016) presented a summary of the performance of
Titanium alloy test specimens fabricated using various methods, some of which are traditional while
other are AM techniques. Figure 18 illustrates the mechanical properties of the identical specimens
represented in Ultimate tensile stress (UTS), YS and percentage elongation. The most significant
observation is the considerable range between the parameters as the highest YS is about 1100 MPa
achieved using DMLS. While the lowest YS were affiliated with the three traditional manufacturing
techniques and were just below 800 MPa. This accents the fact that additively manufactured parts

Ahmed Hashhash 38 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

generally have higher yield strengths. Furthermore, the percentage elongation in this case relates to
the ductility of each test specimen. While this does not apply to the specimen made using EBM,
specimens made using AM techniques had an improvement in the elongation of the specimen which
could be seen as an improvement to the overall ductility. However, this is shown to have affected the
yield strengths of those specimens. This data is important as it helps with choosing a manufacturing
technique that fits the application. However, for the purposes of this study, the properties of the part
resulted by DMLS and Heat Treatment seems most relevant for two reasons. Firstly, DMLS was the
same manufacturing method used by GE to fabricate the winning brackets. Secondly, the yield
strength of the resulted part (about 850 MPa) is close to the yield strength defined in Solidworks
(827 MPa).

6. Conclusions
Through application of Solidworks Topology Optimization, the author has become
increasingly aware of the strengths as well as the limitations associated with this type of
optimization. The part of choice for this study was a great choice as it had various other counterparts
produced in the original competition, therefore it was easy to virtually compare it to other designs.
Despite successful optimization of the part, it could not undergo FEA using the configurations
presented directly in the Topology study (refer to 4.2). Therefore, it had to be re-built using the
original geometry and a graphics body which was traced to generate sketches. While this method
had implications on the accuracy of the re-designed bracket, it allowed for the use of the general
sketch and features tools available in Solidworks. This greatly aided with the control of the final
geometry, as geometric inconsistencies were constantly monitored and eliminated to achieve a CAD
model suited for FEA as well as manufacturing.
The bracket achieved had 50% less mass than the original bracket and was subjected to an
FEA static study to compare its performance with the original. Despite observing decline in the
optimized bracket’s performance compared to the original, it did not exceed the yield stress specified
for the material under any of the loading conditions. Nevertheless, concerns about the performance
of the optimized bracket were raised because of the isotropic assumption applied in the optimization
as well as the FEA studies presented in this paper. These concerns could be summarized in the
following points.
1. The isotropic nature of Topology Optimization in Solidworks: this reason is important
because the bracket was set to be manufactured using an additive manufacturing technique
which has been proven to result in parts with some considerable degree of anisotropy. This
could result in a considerable difference between the performance of the manufactured
bracket and its FEA tested model. While the anisotropic behaviour associated with AM parts

Ahmed Hashhash 39 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

could be manipulated through different build orientations and post-build treatment, it would
be considerably expensive to test for several build orientations.
2. Absence of an AM constraint: while this was not critical to the progress of this study, it could
have had a considerable effect on the results in terms of eliminating features which often
cause problems during AM processes. This would have also led to a more sustainable part
since this constraint should aim to minimize the support material required.
3. Despite observing an increase in deformation in the achieved bracket, it was not critical as it
did not exceed 0.34 mm in any given case. However, this parameter could be important
when assessing the ductility and the anisotropic behaviour of a manufactured part.
4. Through maintaining simplicity in the features applied in the re-design phase, the resulted
bracket could be fabricated through subtractive manufacturing techniques. However, it
would require expensive machining and tools since it would be machined from Ti-4V-6Al bar
stock.
5. While additively manufactured parts often have higher Yield Stress values, this could be
unfavourable as it is directly related to lower ductility. This could cause fractures specially in
a part with a high level of anisotropy which is characteristic of AM parts. Since these features
could be somewhat controlled through re-orientation of the build and post-built heat
treatments, a part with favourable mechanical properties could be achieved.

Ahmed Hashhash 40 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

7. Future Work
The extent of this project could be furthered to facilitate various endeavours in upcoming
work. Despite difficulty of application, clear parametric sketches could be used in the re-build phase
in future work. This would be significantly difficult to do since the optimized graphics body could only
be traced approximately. Additionally, the irregular nature of geometric features in parts resulted by
TO makes it inherently difficult to achieve using sketches including regular geometric shapes such as
circles, squares, or rectangles. Therefore, an appropriate approximation would have to be made for
dimensions of those sketches. This method would allow for further, more detailed types of
optimizations to be applied such as sizing optimization which could aid with finding optimal
dimensions of previously mentioned sketches. Combining those two optimization methods is a
standard practice in various industries and it would help achieve a better geometry in terms of
control over design parameters and manufacturing requirements.
Furthermore, the optimization could be improved through a number of methods. Firstly, the
increase of deformation in the optimized bracket could be controlled through including a
displacement constraint for each loading case in further optimization. This could result in a different
topology which limits deformation in the optimized model. Additionally, the computational power
could be improved as the power available was only sufficient to generate a default sized mesh.
Refining the mesh could have yielded better optimization results as discussed in (3.5.3).
Despite the significant 50% weight reduction achieved as well as the acceptable post
optimization results, field testing of a fabricated bracket is essential for performance validation. This
could outline performance problems with the achieved design that would entail further optimization.
Additionally, it could provide insight into differences in performance between a bracket
manufactured using additive manufacturing or machining processes due to the difference in
microstructures and mechanical properties.

Ahmed Hashhash 41 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

References
2013. Geomagic - 3DXpert for Solidworks. UK: 3D Systems. Available at:
<https://uk.3dsystems.com/software>
Abdulhameed, O., Al-Ahmari, A., Ameen, W. and Mian, S., 2019. Additive manufacturing:
Challenges, trends, and applications. Advances in Mechanical Engineering, [online] 11(2), pp.2-3.
Available at: <https://journals.sagepub.com/doi/full/10.1177/1687814018822880> [Accessed 11
March 2021].
Bendsøe, M., 1989. Optimal shape design as a material distribution problem. Structural
Optimization, 1(4), pp.193-202.
Bendsøe, M. and Kikuchi, N., 1988. Generating optimal topologies in structural design using a
homogenization method. Computer Methods in Applied Mechanics and Engineering, [online] 71(2),
pp.197-224. Available at: <https://www.sciencedirect.com/science/article/pii/0045782588900862>.
Bendsøe, M. and Sigmund, O., 2003. Topology Optimization: Theory, Methods, and Applications. 2nd
ed. Berlin: Springer, pp.1-3.
Carter, W. et al., 2014. The GE Aircraft Engine Bracket Challenge: An Experiment in Crowdsourcing
for Mechanical Design Concepts.
Dieter, G., 1961. Mechanical metallurgy. New York: McGRAW-HILL BOOK COMPANY, pp.58-60.
Facchini, L., Magalini, E., Robotti, P., Molinari, A., Höges, S. and Wissenbach, K., 2010. Ductility of a
Ti‐6Al‐4V alloy produced by selective laser melting of prealloyed powders. Rapid Prototyping Journal,
[online] 16(6), pp.450-459. Available at:
<https://www.emerald.com/insight/content/doi/10.1108/13552541011083371/full/html> [Accessed
17 March 2021].
Gao, T., Zhu, J. and Zhang, W., 2017. Topology Optimization in Engineering Structure Design. Elsevier
Science, pp.12-14.
Garcia-Granada, A., Catafal-Pedragosa, J. and Lemu, H., 2019. Topology optimization through
stiffness/weight ratio analysis for a three-point bending test of additive manufactured parts. IOP
Conference Series: Materials Science and Engineering, 700, p.012012.
Gebhardt, A. and Hötter, J., 2016. Additive Manufacturing. Munich: Hanser, pp.21-91.
Help.solidworks.com. 2020. 2020 SOLIDWORKS Help - Export Smoothed Mesh. [online] Available at:
<http://help.solidworks.com/2020/english/SolidWorks/cworks/hidd_save_shape_mesh.htm?id=d97
360d30df64188a23a3bc0271c1676#Pg0> [Accessed 11 March 2021].
Help.solidworks.com. 2020. 2020 SOLIDWORKS Help - Material Mass PropertyManager. [online]
Available at:

Ahmed Hashhash 42 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

<http://help.solidworks.com/2020/english/SolidWorks/cworks/c_material_mass.htm?verRedirect=1
> [Accessed 11 March 2021].
Help.solidworks.com. 2020. 2020 SOLIDWORKS Help - Min Max Formulation for Topology. [online]
Available at:
<http://help.solidworks.com/2020/english/SolidWorks/cworks/c_min_max_formulation.htm>
[Accessed 11 March 2021].
Help.solidworks.com. 2020. 2020 SOLIDWORKS Help - SIMP Method for Topology Optimization.
[online] Available at:
<http://help.solidworks.com/2020/english/SolidWorks/cworks/c_simp_method_topology.htm?id=a3
2c484b48294e25b66440d773aa7b98#Pg0> [Accessed 11 March 2021].
Help.solidworks.com. 2020. 2020 SOLIDWORKS Help - Topology Study. [online] Available at:
<https://help.solidworks.com/2020/English/SolidWorks/cworks/c_generative_design_study.htm>
[Accessed 11 March 2021].
Hitzler, L., Janousch, C., Schanz, J., Merkel, M., Heine, B., Mack, F., Hall, W. and Öchsner, A., 2017.
Direction and location dependency of selective laser melted AlSi10Mg specimens. Journal of
Materials Processing Technology, [online] 243, pp.48-61. Available at:
<https://www.sciencedirect.com/science/article/pii/S0924013616304150?via=ihub> [Accessed 17
March 2021].
Leary, M., 2020. Design for Additive Manufacturing. 1st ed. Elsevier, pp.1-6., ISBN 9780128167212,
https://doi.org/10.1016/B978-0-12-816721-2.00001-4.
(https://www.sciencedirect.com/science/article/pii/B9780128167212000014)
Lewandowski, J. and Seifi, M., 2016. Metal Additive Manufacturing: A Review of Mechanical
Properties. Annual Review of Materials Research, 46(1), pp.151-186.
Morgan, Dawn & Levatti, Héctor & Sienz, J. & Gil, Antonio & Bould, D. (2016). (2016) GE Jet Engine
Bracket Challenge: A Case Study in Sustainable Design. The Journal of Innovation Impact. 7. 95-107.
Pradhan, S., Singh, S., Prakash, C., Królczyk, G., Pramanik, A. and Pruncu, C., 2019. Investigation of
machining characteristics of hard-to-machine Ti-6Al-4V-ELI alloy for biomedical applications. Journal
of Materials Research and Technology, [online] 8(5), pp.4849-4862. Available at:
<https://www.sciencedirect.com/science/article/pii/S2238785419306490> [Accessed 11 March
2021].
Rozvany, G., 2001. Aims, scope, methods, history and unified terminology of computer-aided
topology optimization in structural mechanics. Structural and Multidisciplinary Optimization, 21(2),
pp.90-108.
Sigala, C., 2020. Topology Optimized Design Approach for Additive Manufacturing on an Aircraft
Engine Bracket. Master of Science in Mechanical Engineering. California State University, Long Beach.

Ahmed Hashhash 43 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Tempelman, E., Shercliff, H. and van Eyben, B., 2014. Additive Manufacturing. Manufacturing and
Design, [online] pp.187-200. Available at:
<https://www.sciencedirect.com/science/article/pii/B9780080999227000111> [Accessed 17 March
2021].
Walton, D. and Moztarzadeh, H., 2017. Design and Development of an Additive Manufactured
Component by Topology Optimisation. Procedia CIRP, 60, pp.205-210.
www.grabcad.com. 2013. GE jet engine bracket challenge. [online] Available at:
<https://grabcad.com/challenges/ge-jet-engine-bracket-challenge> [Accessed 5 March 2021].
Xia, L., 2016. Multiscale structural topology optimization. 1st ed. London: ISTE Press Ltd and Elsevier
Ltd, p.12.

Ahmed Hashhash 44 841334

6101MECH Topology Optimization of a Jet Engine Bracket using Solidworks

Appendix A: Technical Drawings

Figure 19: Technical Drawing of the achieved bracket.

The technical drawing in figure 19 demonstrates the main dimensions of the achieved bracket and
does not include the dimensions of the features edited during the rebuild phase due to their
complexity. It is only intended for illustrative purposes which aid with the perception of the
optimized bracket geometry. All dimensions given are in millimeters.

Ahmed Hashhash 45 841334