INTERNATIONAL DESIGN CONFERENCE - DESIGN 2014

Dubrovnik - Croatia, May 19 - 22, 2014.

KEY PROCESS VARIABLE DRIVEN

MANUFACTURING PROCESS SELECTION

M. S. C. Tuckwood, A. P. Conway and R. I. Whitfield

Keywords: process selection, decision making, design support tools

1. Introduction
Significant effort is being made within the aerospace sector to reduce the unit cost of components and
increase the material utilisation within formed and forged engine components. It is estimated that
around 70% of manufacturing costs are determined during conceptual design stages [Smith et al.
2010], in some cases, manufacturers may have buy-to-fly ratios of less than 20%, prompting questions
as to whether the most efficient process is being selected in the design stage.
There is a lack of objective manufacturing process pre-selection techniques, and subsequently, a need
for a quick, broad and comparative study using key variables to evaluate potential manufacturing
processes, with the intent of down-selecting and prioritising these processes. In the majority of cases,
engineers trust their experience in selecting the process for the manufacture of a specific component.
However this may result in bias and preference of processes and materials being key drivers for their
selection. As a result, optimal manufacturing processes; some of which may be new and emergent
technologies, can be overlooked, simply because of the gap in knowledge in terms of process
capability and applicability. Over the last 20 years in particular, research in this area has resulted in the
development of a range of methods, which attempt to improve the procedure of manufacturing process
selection.
Shercliff and Lovatt [1998] and Ashby et al. [2003] have produced notable works on process selection.
Both agree that there are three different types of selection strategies, loosely based on the stages of the
design process, i.e. conceptual, embodiment and detail design stages [Shercliff and Lovatt 1998].
However, it is suggested that a selection task is one of these three types, whereas Ashby et al. imply
that a combination of approaches are required in order to obtain the most accurate and feasible
outcome. Each selection task is different, so by singling out one approach, the risk of over-constrained
selection and missed opportunity is significantly higher. Process selection should be carried out in the
early stages of the design process when considering aspects of design for manufacture, since this
where the majority of costs are determined [Smith et al. 2010].
A flexible, adaptable decision-based methodology has been developed following the work structure
illustrated in Figure 1. In contrast to previous research, the methodology presented within this paper
does not work under the assumption that each selection task should be approached in the same way.
Instead, design requirements are translated in a logical manner, are case dependant, subsequently
providing the criteria for selection. Appearing simplistic, the methodology is underpinned by hidden
complexities that filter all possible manufacturing processes into a set of viable contenders applicable
to the specific component, allowing selection of the most efficient process.

2. Investigation
Ashby et al. [1992] developed the Cambridge Engineering Selector, the only commercially available
selection software. The focus is primarily on materials, although there are some elements of process

DESIGN SUPPORT TOOLS 1209

selection, but the principles are based purely on well well--rounded
rounded decision making for selection. The
system translates design requi
requirements,
rements, filters through potential solutions, and produces a manageable
set of potentially viable answers that are provided with supporting information. Ashby is also widely
recognised for developing graphical methods, including the Ashby Plot and the Ashby Ashb y Chart [Ashby
1992].
Mathematical methods have been proposed by many, including Raviwonsge et al. [2000] and Kohara
et al. [2009], who attempt to quantify the decision making process. These strategies incorporate
methods such as Self
Self-Organising
Organising Maps and tthe he Analytical Hierarchy Process. Despite their intended
use for selection tasks with a level of uncertainty, Hayes and Akhavi [2008] argue that mathematical
methods are not as widely used as expected, and do not necessarily improve problem solving
performa
performance.
nce. A more effective strategy would be to better reflect the human decision making needs,
while avoiding the need for a lengthy process.
Some methodologies studied rely on specific attributes being pre-defined
pre defined prior to the selection
selectio task
being carried out [Cogun 1993], namely material and production quantity, or have a set order for
requirements to be considered. This can be seen as a somewhat contradictory factor considering the
desire to create a flexible system.

Figure 1. Project methodology

2.1 Prioritisation mmethods

Creating an adaptable methodology requires integration of methods which can be used to aid the
translation of design requirements into selection criteria. It can
can be argued that no matter which order
questions are asked, the same answer
answer will result. Decision makers must have a need to use the
methodology
methodology,, in order for it to be successful. To be used it must appear viable to the decision making
team – with justifiable results. This can be addressed by considering selection criteria in a logical
order, reflective of the end requirements
requirements of the design previously set.
A variety of different method
methodss for prioritisation exist, eight of which were analysed, compared and
considered for use as part of the investigation stage. The method does not need to be exhaustive, nor
does it have to provide a complex output, all that is required is that it is simple, and quick to carry out.
Reasoning behind designing a methodology like this is partially to cut down on the overall time and/or
ensure focus in tthe
he areas that need it.

2.2 CARVER method

CARVER is a method originally developed by the US Special Forces, used in planning for both target
selection and identifying potentially high risk targets. Although devised for a different application, the
principle
principless can be adapted and used to generate a matrix, highlighting which elements are most
immediately necessary and what should come later. CARVER is an acronym, as described in Table 1.

1210 DESIGN SUPPORT TOOLS

 Table 1. CARVER method
Criticality Criticality with respect to the main objective?

Accessibility Do we have the resources/capability? Immediate start or are there prerequisites?

Return Greatness of expected return?

Vulnerability Vulnerability of requirement? Risks?

Effect Impact upon fulfillment of the aim? Wider effect?

Recognition Is the requirement clear? How easy is it to recognise the required steps?

2.3 Conclusions from investigation

It has become evident that existing research has identified a requirement for a simplistic, yet functional
strategy that will provide comprehensible results, while steering the decision maker away from
preference and bias choices – a quality that has become contradicted in some cases. It seems that many
of the discussed proposals, although recognised, have resulted in complex and time-consuming
processes, expensive systems and services, or requiring significant analysis effort; not exactly a
driving factor in encouraging use by decision makers. Derived from the investigation stage is a re-
emphasis of the need for a quicker, broader methodology for manufacturing process selection.

3. Proposed solution
Following the initial research phase, it had become clear which elements of previous methods were
suitable, such as the combination of approaches proposed by Ashby [2003], and those which were less
successful, such as the mathematical approach proposed by Raviwonsge et al. [2000]. This helped to
define what exactly, in the context of this work, the characteristics of the methodology being proposed
should be. The primary requirement is to create something which would use process attributes to carry
out broad comparisons and support the evaluation of multiple manufacturing processes.
A notable issue with some of the current process selection methodologies is that there is a lack of
flexibility. The proposed methodology should have the ability to be easily tailored and adapted
dependent on the task in hand. Each task subject to manufacturing process selection will be different,
and different outputs will be required, e.g. in some cases, perhaps the surface detail is the driving
factor, whereas others may be focussed on geometric precision. Having set criteria for selection would
be presuming that each task working through the methodology has the same level of requirements,
often not the case, so by allowing for a degree of flexibility, it can be much more focussed and
specific.

3.1 Methodology
The proposed methodology, shown in Figure 2, tackles the issue of manufacturing process selection in
a different way. Essentially, the core function is to be the means of converting a set of inputs - the
requirements, into a set of outputs - the selection of a process [Ashby et al. 2003]. More specifically, it
is a transition from the set of all manufacturing processes, to a manageable sub-set which can then be
considered. A sub-set as an outcome is more realistic than one mathematically optimised solution - a
sub- set of processes will facilitate the decision maker(s) in making an informed choice rather than
making the decision for them. This may take into account factors such as accessibility to equipment.
A key requirement is to address the lack of flexibility, common in existing process selection methods.
The proposed methodology is flexible as a result of two main attributes:
1. The overall structure,
2. The way in which requirements are considered.
Starting with the overall structure: the methodology primarily flows horizontally, although is also
segmented vertically. Each of the vertical sections, termed Levels divides the process into more
manageable portions, and can be added, removed and tailored as required. These levels question
different task-dependent requirements.

DESIGN SUPPORT TOOLS 1211

 selection groundwork or Level 0 as it has become known.
Underpinning the main Levels is the pre-selection
Level 0 is illustrated with a much clearer divide than the others - considerable focus was in these
initial stages as it is the foundation for which the remainder of the selection process is built.
built Any
design activity, including any potential alterations or improvements required before the selection task
is carried out here. Clear definition of this stage helps to eliminate the risk of it being omitted; it can
often be tempting for decision makers to jump to a conclusion based on their their expertise and
preferences.

Figure 2.
2 Down Select Methodology
Methodology

3.1.1 Level 0
The methodology then focusses on how to translate the design requirements of a specific task into
selection criteria
criteria. This could be approached
approached in multiple ways; however, clarifying the second of the
flexibility attributes, is the decision to incorporate a prioritisation method. Logically, the method will
prioritise the part requirements in order of their importance respective of the final output. outp ut. These will
then feed the input factors vertically into each Level. The most crucial aspects are also likely to be the
most discriminating in terms of process suitability, according to [Shercliff
Shercliff and Lovatt 2001],
2001 , these are
the questions that should be asked first. This will also ensure than there is a clear ‘funnelling down’
throughout the progression of the process. Fewer processes will then remain to be assessed on the
more complex
mplex requirements of the part.
The CARVER method, se selected
lected for its simplist
simplistic
ic nature,
nature, has the ability to be used in a number of
ways; it could be completed by one, a small group or a team of people. How it is approached could
also vary, perhaps a team will carry out the method individually before collating and discussing their
answers,
swers, or maybe it is part of a teamwork focussed exercise. There is scope for stakeholders to work
with this in a way that suits their needs, as well as fitting in with the context of the task in hand.
The values given to each of the requirements with re respect
spect to the CARVER attributes are collated and
summed within an Excel spreadsheet templat template, resulting in a prioritised scoring. This his provides the
basis for Levels 11-44 of the methodology. The key requirements identified will be ordered as selection
criteriaa within.
A collaborative meeting is held within the prioritisation stage involving relevant stakeholders, in
conjunction with the initial generation of the part or product specification
specification.. It is referred to in this work
as a Stakeholder Symposium. There are potentially other benefits that this could promote, particularly
in terms of customer focus and satisfaction. Customers would have a clearer understanding of the
correlation between their expectations and what is actually realistic. It would help encourage encourag e them to
consider what features actually need to be included, and those that perhaps do not matter as much.
Essentially, it can help to ensure that all stakeholders are clear and in agreement to the expectations
and requirements prior to the selection task and subsubsequent
sequent work being carried out.
This activity is crucial. Conventionally, stakeholders will give a specification; and manufacturing
engineers will select a process. The main issue is often the lack of an in in-depth
depth consideration and
rationalization of why a particular process was selected.
selected Alternatives
rnatives are commonly not considered,
contradictory of the fact that design in general is open
open-ended;
ended; design engineers need the willingness to
consider all possibilities [Ashby
Ashby 1992].
1992 . This method ensures that multiple processes are analysed.

1212 DESIGN SUPPORT TOOLS

There is a lack of discussion and involvement of the stakeholders within the decision making task.
Ultimately, they are the judge as to whether a manufacturing task is successful or not.

3.1.2 Level 1
Level 1 can be considered as being the pre-selection stage - this is the level where the processes that
are clearly not feasible can be eliminated. Input in this stage is the highest scoring requirements (two
are shown in Figure 2) previously identified in the prioritisation method. These factors will be the
most discriminating in selection, so the majority of processes will be disregarded, leaving a smaller set
of candidate processes to follow through to the next stages.

3.1.3 Levels 2-4

The levels which follow can be added or removed as required for the specific task. Levels 2-4 will
foresee a more gradual funnelling down process, taking into consideration the more specific
requirements, and resulting in a refinement of candidate processes each time. These requirements may
be categorised, e.g. Economical Requirements, with an underlying set of more detailed needs which
can be considered during progression of the level.
These levels work vertically as well as contributing to the overall horizontal flow of the methodology,
resulting in focussed attention given at each stage. It is also possible that the decision makers could
result in a confined enough shortlist after any of these levels, although four is recommended.

3.2 The significance of levels

The levels are more than just a means of dividing the process into more manageable stages.
Generic decision making tools for process selection have an input factor, and an output factor - but
lack in detail where the transition is concerned. It is clear that there are issues, which have been
attempted to be solved, although many have failed to direct decision makers away from the tool that
they know ‘works best’: their own preference. This is precisely what this methodology is attempting to
address.
The level breakdown prompts consideration for the steps in-between achieving the A – B objective in
selection. Decision makers have an element of control over the filters used to down-select. Breaking
down the process encourages evaluation at each stage, perhaps a problem may arise, things may have
changed - it is much simpler to go back a level, particularly as each stage is recorded, than starting
from the beginning again.
Furthermore, it is in the nature of engineers and designers to question things: why are things the way
they are? Where did those solutions come from? Particularly as the majority of design tasks are open-
ended, with no set answer – designers seek rationale. A logical systematic approach is needed, one that
is robust in the decisions made – this is the logic behind the consideration of requirements on different
levels. The process down-select problem has only been left un-solved because decision makers do not
use past solutions. Incorporating the level structure allows for them to see, but more importantly
understand and justify where each of the answers (in this case process short-lists) are coming from. As
a result, the methodology will appear reliable, and a desirable contribution to decision making.
“Creating effective decision aids is not simply a matter of finding a method that computes the most
correct answer or the interface that best presents the data, but also of finding the most effective way to
integrate tools with the human problem solving needs.” [Hayes and Akhavi 2008, p. 153]

3.3 Selection Support

From a user perspecitive, a methodology of this nature should appear simple, although requires
supporting hidden complexities. The first of these has already been discussed in the prioritisation
method, however crucial to the decision making framework is the Selection Support.
The key requirements of the selection support is that it is able to capture the capabilities of a broad
range of manufacturing processes while enabling decision makers to filter these, using the criteria
defined during the prioritization method. Although not the primary focus of the project, consideration
of selection support was critical, particularly in allowing the methodology to be demonstrated.

DESIGN SUPPORT TOOLS 1213

 Figure 3. SharpCloud
For experimentation purposes, SharpCloud was the selection support tool used; software allowing for
the creation of vis
visual
ual experiences to communicate a story [SharpCloud 2013]. The stories help to
convey the overall view (Figure 3), with the integration of timelines, relationships and virtual worlds
being some of the features that help these stories stand out and provoke vvisual
isual appeal. It provided a
broad comparison of a wide range of manufacturing processes based on different process attributes,
such as finishing costs and cycle times, although other tools can be us
used
ed to undertake the same task.

3.4 Database Support

upport
Giachetti
tti [1998] identifies the significance of incorporating both decision and database support within
the process selection methodology. Databases of technical and process data underpin the selection
support. Again, another hidden complexity, databases should be updated frequently, ensuring a
reliable source of information. Some previous selection methods have revolved around either selection
or database support, but not collaborating the two. As illustrated in Figure 2, the database support is
shown to work in two directions. Not only is the process data essential for process selection to take
place, it should also be subject to feedback as a method of selection – adding to the database and
analysing where necessary upon completion of manufacturing tasks. Over time this will capture past- past
case data, potentially useful and applicable to future selection tasks, and even for compiling case
studies. The supporting database can be as simple or as sophisticated as needed. In this case, it was
generated within an Excel spreadsheet, data can easily be imported and exported from an Excel file,
and can be easily integrated into different selection support tools.
tools

4. Exemplar Components
omponents
Demonstration of the proposed methodology is critical in justifying its logic. Described below
below is an
example use case based on an aerospace application. There are, however, a potentially large range of
applications of the methodology due to its adaptability depending on the task; the approach discussed
can also be used for more generic parts.
parts

4.1
.1 Engine Shaft
Shafts are central components to the aero-engine,
aero engine, running through the majority of the length, and are
required to connect and mount other parts. As a result, it is critical that they have high strength and
structural integrity. Continuous effort
effort is being made to manufacture them thinner and resultantly more
lightweight to help reduce fuel and other costs, but also stronger. These characteristics could be
considered contradictory, so the key to achieving the objective lies in the design process,
process, namely in the
selection of the best manufacturing process. This is a prime example of cost reduction in aerospace
components, the main driver for this project.
An example part specification of a typical engine shaft has been drawn up for demonstrative purposes:

1214 DESIGN SUPPORT TOOLS

  Must be manufactured from high strength material, Titanium/Nickel based super alloys
common but in this case, high strength steel will be used for its integrity and toughness.
 Geometric requirements are that the part is tubular, of varying cross sections with a flanged
end.
 Material utilisation should be excellent, main aim is to reduce material waste.
 The part requires a smooth surface detail to be fit for purpose.
 Finishing costs should be minimal, secondary processing should be avoided.
 Good process flexibility to allow for similar components of varied sizes to be manufactured
using the same technique.
 Overall quality should be excellent, close tolerances and high mechanical strength necessary.

Level 0
Although the requirements of a component with the caliber of an aero-engine shaft tend to be greatly
detailed and specific, it is used for demonstrative purposes, not to solve a problem, hence the generic
nature of the specification. As the methodology describes, the specification is broken down and
prioritised using the CARVER method, Table 2, to produce an order for the requirements to be
considered in selection.

Table 2. CARVER Analysis

C A R V E R Total:
Material 5 4 3 4 2 3 21
Geometric Requirements 5 5 4 5 2 4 25
Material Waste 5 5 5 3 5 3 26
Surface Detail 5 3 4 2 4 4 22
Finishing Costs 4 3 5 2 4 5 23
Flexibility 2 4 2 3 3 4 18
Quality 4 4 3 5 5 4 25

The results from the CARVER matrix are then fed into the methodology, resulting in a model tailored
to the specific manufacturing task.

Level 1
Material waste/utilisation and overall part quality were the factors used in Level 1, the stage where the
majority of processes which are clearly not viable will be eliminated. After filtering within the
SharpCloud system taking the processes with the lowest material waste and highest quality, the
candidate process list, consisting of 19 processes was generated – derived from the original 45.

Level 2
The next highest scoring requirement is considered within Level 2. In this particular case, it can be
seen that there are in fact two attributes with the second highest score. Decision was made for the
geometric requirements to solely dominate Level 2 due to the fact that it is essentially a set of
requirements. This approach to filtration worked well due to the tag feature within the system; the 19
candidate processes were shortlisted to 10.

Level 3
Working with the 10 processes of the first shortlist, Level 3 considered the processes with the lowest
finishing costs, and higher surface detail values. This resulted in a further process being eliminated to
form the second shortlist.

DESIGN SUPPORT TOOLS 1215

Level 4
The remaining processes were filtered depending on their changeover flexibility and applicable
materials. Two manufacturing processes, considered “viable contendercontenders”
s” resulted. This is a prime
example of where an informed choice is required. In this case, one of the two processes, flow forming,
is superior in terms of surface detail, so is the selected process.

Figure 4. Down Select Results;

esults; Engine S
Shaft

Summary
Manufacture of an engine shaft is an ideal example of an aerospace component where continuous
effort into part cost reduction and improved buy buy-to-fly
fly ratio is being made. Selection of better
manufacturing processes can significantly improve the part prospe prospects.
cts. In this example, a generic
engine shaft has been used, which after following the methodological steps resulted in flow forming
being the best process for the task, one which is subject to research and development work for
potential use in critical mai
mainline
nline engine shafts.
Recording of the results after eacheach level of selection, as seen in Figure 44,, is critical when using
SharpCloud as the selection support. The system allows for advanced filtering of the processes,
although filters are limited in terms of
of how many can be applied at once, so some processes eliminated
in Level 1 may reappear in Level 3. Keeping track of the shortlist allows decision makers to ignore
these processes and continue with the selection task. If more detailed input attributes were to be used,
use of the system may be unfeasible.

5. Recommendations
Process selection is a fundamental issue in design for manufacture of aerospace components. The
research carried out within this project reinforced the evident complexity of the issue. As a result, the
main recommendation is that the work carried out to date is continued, building upon the findings
presented in this report. This project focussed on the development of the decision-based,
decision based, selection
methodology proposed, but what is needed in future is focus on the support tools embedded within, for
instance, the selection and database support. There is clear opportunity, and with such applicability
and need within industry, the potential exists to further develop a valuable process-selection
process tool.

5.1 Specification Translation

ranslation
Further research is required within the pre-selection
pre selection stage, more specifically when extracting design
requirements. Specifications take a variety of formats and level of detail, resulting in potential
inconsistencies where the down
down-select
select method is concerned. The specification is the root of the overall
process, and is the document which can determine much of the cost and final outcome of the part.
Translation of the specification will involve stakeholders and decision makers, clearly defining the
requirements and expectations, as well as simultaneously prioritising these to facilitate process
selection. Wh
What
at must be considered, however, are potential conflicting criteria and the risk of two or
more attributes being rated the same in terms of importance. Incorporation of a secondary tool or

1216 DESIGN SUPPORT TOOLS

tailoring of the current method to address the possibility of this occurring could be a further
development. Further work should begin with emphasis on these underlying aspects.

5.2 Process Chains

Net-shape manufacture is becoming increasingly sought-after, however in the majority of cases, parts
require a sequence of manufacturing steps before being considered finished. Perhaps in some instances
it is not simply the selection of the most efficient manufacturing process: what if the combination of
steps is the key to improved selection? Research into this possibility would prove challenging,
requiring greater detail of work, and possibly an alternative selection support tool. It may be
unrealistic to capture every potential combination, but perhaps selecting the core process first,
followed by selection of the best pre- and post-processing techniques could be effective. Feasibility of
this approach would need to be explored.

5.3 Material Selection

Process and material selection are undeniably linked. Some existing methods researched look at the
potential co-selection of both materials and processes [Giachetti 1998], [Smith et al. 2010], [Albiñana
and Vila 2012], where others work on the basis that the material is a predefined attribute [Cogun
1994], [Swift and Booker 2003]. Examination into the viability of incorporating a material selection
tool could be useful. Some cases may require a specific material type or grade to be used, but others
may not be so certain, with only the qualities of the required material being defined. Research into the
correlation of process and material selection could also contribute to the overall cost-reduction aim.
What cannot be ruled out is that perhaps the most efficient material is not being selection for the same
reasons as the processes – preference and bias. In addition, this could potentially increase the
flexibility of the method, ensuring that it can be used in as many scenarios as possible. Component
cost reduction is a problem that is by no means one-dimensional. There are a number of factors that
contribute; addressing as many as feasible within a methodology would be advantageous.

6. Conclusions
The process down-select project undertaken at the Advanced Forming Research Centre allowed for
considerable progression to be made in understanding the issue of manufacturing process selection,
specifically within aerospace engine components, despite only being the initial stage within a larger
project. Research into significant works of literature in the area combined with identification of the
key requirements and awareness of the underlying issues resulted in the generation of a conceptual
selection methodology.
The proposed methodology is one that tackles the process selection task in a unique way.
Traditionally, the selection methodology is a transfer function between the input of part requirements,
and the output of selected process(es). This is also true of the proposed solution; however the
differentiating factor is the breakdown of the selection task, as well as the decision maker’s ability to
tailor it for each individual selection task. By incorporating different levels, the decision makers are
encouraged to focus more effort into the pre-selection phase, where they involve input from
stakeholders to define and prioritise the requirements for selection. More importantly, it allows for the
decision makers to understand and justify where each of the answers (in this case process short-lists)
are coming from. As a result, the methodology offers traceability and supports rationalization by the
decision-maker. By developing a methodology which supports the human decision making process in
this way, it reflects the engineer’s nature to question things, supporting its potential for successful
adoption. The levels also allow for evaluation at each stage, meaning that if circumstances change or a
problem occurs, it can be addressed as it arises.
Simplicity is desirable in selection methodologies, although it is evident that it should appear simple,
and requires underlying hidden complexities. The nature of the project required a broad comparison
approach, and SharpCloud software was used to achieve this. It provides the selection support, and
allows user to visualize the task for a clearer understanding. Decision makers also need to keep track
of the process shortlists due to the filtering limitations within SharpCloud.

DESIGN SUPPORT TOOLS 1217

7. Future work
Although developed to a high level at this stage, the methodology presented within this paper has clear
potential to be further refined and utilised as the basis of a more detailed process selection method.
What this project has enabled is the justification of the feasibility of developing a process down
selection method for use within the aerospace sector. There is scope for the development of a purpose-
built toolset, incorporating both database and selection support, subsequently eliminating the reliance
on existing systems. In turn, this will allow the methodology to accommodate selection tasks of a
higher complexity. Relevant examples will be required to effectively test the process selection toolset
in its entirety, particularly in parts which best represent the process selection challenge.

References
Albiñana, J. C, Vila, C., "A Framework for Concurrent Material and Process Selection during Conceptual
Product Design Stages", Journal of Materials and Design, Vol. 41, 2012, pp. 433-446.
Ashby, M. F, "Materials Selection in Mechanical Design", Butterworth Heinemann, ISBN: 0-7506-6168-2, UK,
1992, 1999, 2005.
Ashby, M. F., Bréchet, Y. J. M., Cebon, D., Salvo, L., "Selection Strategies for Materials and Processes",
Materials and Design, Vol. 25, 2004, pp. 51-67.
Cogun, C., "Computer-Aided Preliminary Selection of Non-Traditional Machining Processes", International
Journal of Mechanical Tools Manufacture, Vol. 34 No. 3, 1994, pp. 315-326.
Giachetti, R. E., "A Decision Support System for Material and Manufacturing Process Selection", Journal of
Intelligent Manufacturing, Vol. 9, 1998, pp. 265-276.
Hayes, C. C, Akhavi, F., "Creating Effective Decision Aids for Complex Tasks", Journal of Usability Studies,
Vol. 3, Issue 4, 2008, pp. 152-172.
Kohara, K., Aoki, K., Isomae, M., "Forecasting, Diagnosis and Decision Making with Neural Networks and Self-
Organizing Maps", Automation Control - Theory and Practice, A. D. Rodi (Ed.), ISBN: 978-953-307-039-1,
2009.
Raviwonsge, R., Allada, V., Sandidge Jr., T., "Plastic Manufacturing Process Selection Methodology Using Self-
Organising Map (SOM)/Fuzzy Analysis", International Journal of Advanced Manufacturing Technology, Vol.
16, 2000, pp. 155-161.
SharpCloud, [online], Accessed October 9 2013, Avaliable at <www.SharpCloud.com>, 2013.
Shercliff, H. R., Lovatt, A. M., "Selection of Manufacturing Processes in Design and the Role of Process
Modelling", Journal of Progress in Materials Science, Vol. 46, Issues 3-4, 2001, pp. 429-459.
Smith, C., Wright, P., Séquin, C., "The Manufacturing Advisory Service: Web-based process and material
selection", International Journal of Computer Integrated Manufacturing, Vol. 16:6, 2010, pp. 373-381.
Swift, K. G, Booker, J. D., "Process Selection: From Design to Manufacture", Butterworth Heinemann, UK,
2003.

May Tuckwood
University of Strathclyde, Design Manufacture and Engineering Management Department
131 Rottenrow, G4 0NG Glasgow, United Kingdom
Email: may.tuckwood.2013@uni.strath.ac.uk

1218 DESIGN SUPPORT TOOLS