Investigating Perceived Meanings and Scopes of Design for Additive Manufacturing

Abstract

The concept of Design for Additive Manufacturing (DfAM) is gaining popularity along with AM, despite its scopes are not well established. In particular, in the last few years, DfAM methods have been intuitively subdivided into opportunistic and restrictive. This distinction is gaining traction despite a lack of formalization. In this context, the paper investigates experts' understanding of DfAM. In particular, the authors have targeted educators, as the perception of DfAM scopes in the future will likely depend on teachers' view. A bespoke survey has been launched, which has been answer by 100 worldwide-distributed respondents. The gathered data has undergone several analyses, markedly answers to open questions asking for individual definitions of DfAM, and evaluations of the pertinence of meanings and acceptations from the literature. The results show that the main DfAM aspects focused on by first standardization attempts have been targeted, especially products, processes, opportunities and constraints. Beyond opportunistic and restrictive nuances, DfAM different understandings are characterized by different extents of cognitive endeavor, convergence vs. divergence in the design process, theoretical vs. hands on approaches.

Full text

Cite this article: Berni, A., Borgianni, Y., Obi, M., Pradel, P., Bibb, R. (2021) ‘Investigating Perceived Meanings and
Scopes of Design for Additive Manufacturing’, in Proceedings of the International Conference on Engineering Design
(ICED21), Gothenburg, Sweden, 16-20 August 2021. DOI:10.1017/pds.2021.455
ICED21 1937
INTERNATIONAL CONFERENCE ON ENGINEERING DESIGN, ICED21
16-20 AUGUST 2021, GOTHENBURG, SWEDEN
ICED21 1
INVESTIGATING PERCEIVED MEANINGS AND SCOPES OF
DESIGN FOR ADDITIVE MANUFACTURING
Berni, Aurora (1);
Borgianni, Yuri (1);
Obi, Martins (2);
Pradel, Patrick (2);
Bibb, Richard (2)
1: Free University of Bolzano-Bozen;
2: Loughborough University
ABSTRACT
The concept of Design for Additive Manufacturing (DfAM) is gaining popularity along with AM,
despite its scopes are not well established. In particular, in the last few years, DfAM methods have
been intuitively subdivided into opportunistic and restrictive. This distinction is gaining traction
despite a lack of formalization. In this context, the paper investigates experts' understanding of DfAM.
In particular, the authors have targeted educators, as the perception of DfAM scopes in the future will
likely depend on teachers' view. A bespoke survey has been launched, which has been answer by 100
worldwide-distributed respondents. The gathered data has undergone several analyses, markedly
answers to open questions asking for individual definitions of DfAM, and evaluations of the
pertinence of meanings and acceptations from the literature. The results show that the main DfAM
aspects focused on by first standardization attempts have been targeted, especially products, processes,
opportunities and constraints. Beyond opportunistic and restrictive nuances, DfAM different
understandings are characterized by different extents of cognitive endeavor, convergence vs.
divergence in the design process, theoretical vs. hands on approaches.
Keywords: Design for Additive Manufacturing (DfAM), Early design phases, Design methods,
opportunistic DfAM, restrictive DfAM
Contact:
Borgianni, Yuri
Free University of Bolzano-Bozen
Faculty of Science and Technology
Italy
yuri.borgianni@unibz.it

1938 ICED21
1 INTRODUCTION
The enhanced capabilities of new digital manufacturing techniques such as Additive Manufacturing
(AM) and 3D Printing are capable of drastically changing the mindset of designers and engineers
toward the design process. This change of mindset is beginning to gain traction but remains far from
completed and designers’ ability to fully exploit the potential of AM is questioned (Thompson et al.,
2016). In this context, the need arises for new design methodologies, Design for Additive
Manufacturing (DfAM), that can support designers in exploiting the capabilities of these technologies.
DfAM guides designers to consider the opportunities provided by AM such as shape complexity,
material complexity, no-need for tooling and customization to mention a few (Kumke et al. 2018).
This new way of re-thinking the whole design process raised interest both in academia and industry,
leading to the development of methods and frameworks over the past decade (Thompson et al 2016).
However, the objectives of these methods are diverse, and some of them target design constraints,
rather than new opportunities, mirroring the scope of traditional Design for Manufacturing and
Assembly (Boothroyd, 1994).
In consideration of this plurality of objectives, Laverne et al. (2015) reviewed and classified, for the
first time, the main DfAM methods into two main categories: opportunistic and restrictive DfAM. As
more widely explained in the following section, this differentiation is increasingly diffused in the AM
literature, although it has not undergone a validation process yet. This classification might be useful in
formalizing DfAM knowledge, which is still an open issue (Kim et al., 2019), as well as in its ongoing
standardization (Mani et al., 2017).
The distinction between opportunistic and restrictive methods represents a starting point; however, as
it has been currently proposed, this distinction is affected by the following shortcomings:
 Whilst the distinction is intuitively sound, the belonging of specific methods and approaches to
one of the two categories has not been universally accepted or subject to rigorous evaluation. It
remains an arbitrary decision whether a technique should be considered opportunistic or
restrictive.
 The ready acceptance of these two classifications has neglected the exploration of other
potentially useful classifications.
 Any definitive classification should be evaluated, recognized as a guiding principle, accepted and
internalised by all stakeholders in the field. This would support efficient communication between
different actors within the discipline.
Starting from these premises, this paper explores the interpretation of the DfAM concept, which is
relatively novel and consequently still fluid. In particular, the paper aims to ascertain the degree of
agreement on the distinction between opportunistic and restrictive methods, to understand whether
some experts perceive the two classes as antithetic, and to investigate the presence of other latent and
yet unelicited classification criteria. The study addresses these issues by gathering feedback from a
representative sample of educators active in the teaching of AM and DfAM. Here, the role of
educators is considered critical as the future understanding of DfAM will depend on their viewpoint.
At the same time, consensus, shared definitions and classifications would benefit educators in
achieving international consistency and compatibility of DfAM teaching and learning.
Therefore, this paper aims to investigate this distinction between opportunistic and restrictive methods,
look for consensus among experts and explore whether there are other potentially more useful
classification criteria.
Section 2 analyses the current uptake of the distinction between opportunistic and restrictive methods.
Section 3 describes the approach to determining different interpretations of DfAM, with the results
presented in Section 4 and discussed in Section 5. Conclusions are drawn in Section 6 with emphasis
on the pursuance of the research objectives and the major implications of this study.
2 BACKGROUND
Laverne et al. (2015) characterised opportunistic DfAM methods as those that lead to the creative
exploration of new shapes and concepts. The authors considered methods based on optimization
techniques (e.g. parametric and topological), elementary shapes (e.g. lattice structures), bionic
structures and features produced by specific AM technologies. Such methods provide designers with
an opportunity to extend the possibilities of design thinking, aiming to overcome the fixation on

ICED21 1939
constraints imposed by conventional manufacturing techniques (Booth et al., 2017). On the other
hand, restrictive DfAM methods address the limitations and constraints of AM technologies such as
limited materials and properties, performances and characteristics of AM machines, product
manufacturability and quality. Laverne and colleagues also investigated the distribution of the
implementation of these methods suggesting that opportunistic and restrictive methods are equally
distributed. They also provided case studies where both categories of DfAM were involved. The
distinction between opportunistic and restrictive methods became a landmark for successive studies,
which embraced and shared this classification. Consistently, the literature shows that AM is
characterised by great opportunities and considerable constraints while highlighting the relevance of
this process for industrial applications (Thompson et al., 2016).
In general, many scholars agree that opportunistic DfAM is more relevant to the early, more
explorative divergent stages of the design process (Kumke et al., 2018; Laverne et al., 2015; Blösch-
Paidosh and Shea, 2019; Design Council, 2005) since these methods aim to stimulate creativity and
creative ideas (Watschke et al., 2017; Prabhu et al., 2020a; Barclift et al., 2017). On the other hand,
restrictive DfAM methods are proven to be more effective as guidelines (Watschke et al., 2017) to be
followed in the later, convergent stages of the design process (Kumke et al., 2018; Prabhu et al.,
2018a; Reichwein et al., 2020) where the concept is fixed and the design progresses towards
manufacture.
Beyond scholars’ stances on the role of opportunistic and restrictive DfAM approaches, the two
categories have been purposefully used in specific applications concerning design processes and
training activities. While the former has focused more on opportunistic methods as a means to improve
ideation, the latter has involved predominantly restrictive methods and markedly the use of guidelines
to ensure successful printing.
In particular, the integration of specific methods in design and product development processes can be
found in the contributions that follow.
 Zhu et al. (2017) focused on restrictive DfAM only, since they were interested in design
optimization to ensure geometric consistency of the designed product.
 Laverne et al. (2017) prepared a questionnaire where they asked designers which of the two
DfAM methods they usually use. The scope of the questionnaire was to understand the
relationship between AM and innovation.
 Blösch-Paidosh & Shea (2019) encouraged the integration of opportunistic DfAM in the early
design stages.
 Meisel et al. (2017) took into consideration the whole design process and presented a case study
where an existing product had been redesigned with AM techniques. The study showed how
using specific DfAM guidelines can significantly reduce manufacturing time and costs, while
obtaining new and novel design geometries. In this process, opportunistic methods were mainly
used to widen the exploration of the design space, while restrictive ones were leveraged to
consider fabrication issues.
 Reichwein et al. (2020) focused on the functionality of the product and used an opportunistic
DfAM approach to a larger extent. Their contribution took into account the detailed design
phases too.
Other contributions consider DfAM in the training of practitioners and the education of students
through workshops and other practical activities, as in the following.
 Prabhu et al. (2020a) underlined the need to further investigate opportunistic and restrictive
DfAM as training methods for industry professionals. The scholars used a workshop-based study
to introduce opportunistic and restrictive DfAM to industrial practitioners.
 Conversely, Watschke et al. (2017) proposed an academic workshop where opportunistic and
restrictive approaches were combined to support the generation of creative solutions concepts in
design education.
 Barclift et al. (2017) evaluated students’ creativity when performing a design task using
opportunistic methods only.
 Prabhu et al.(2018a) investigated the effect of opportunistic and restrictive DfAM methods in
education, focusing on engineering students’ creative process.

1940 ICED21
 Prabhu et al. (2018b) studied the impact of timing in teaching opportunistic and restrictive
methods on the students’ design processes. The scholars compared two DfAM educational
interventions conducted at different points in the academic semester.
 Prabhu et al. (2020b) proved that the order of teaching opportunistic and restrictive DfAM on
engineering students can affect the way they approach the design process. This highlights the
need for educators to account for the order of presenting content to students.
The above examples show that the distinction between opportunistic and restrictive DfAM is no longer
limited to theoretical constructs but is also becoming operationalized. It follows that the classification
of DfAM methods is perceived as a necessity by scholars and that more proven categorizations can
increase the robustness of future research on the topic. To the best of author’s knowledge and based on
the analysis of the literature, there are no further classifications of DfAM methods and approaches.
3 METHOD
The introduction revealed the need for increasing formalization of DfAM which would align with
ongoing standardization initiatives and inform and support the development of education and training
in DfAM.
To investigate educators’ current understanding of DfAM and the current diffusion of this topic in
tertiary education, the authors implemented an online survey. While a full description of the survey,
which is available in an article currently under review, goes beyond the scope of the present paper, a
clone is accessible at https://lboro.onlinesurveys.ac.uk/dfam-education-programmes-copy-2 (data will
not be saved). Briefly, the survey included questions to pursue the objectives that follow.
 Evaluate the uptake of DfAM in education and specifically to the broader AM field.
 Characterise units of study in which DfAM is taught in terms of the number of students,
University level (e.g. Bachelor, Master), training activities (e.g. lectures, hands-on projects),
assessment methods, taught DfAM contents, and the relevance of DfAM-related subjects.
 Characterize educators in terms of their experience, University grade, attitude to research DfAM
and AM.
 Characterize institutions and programmes that include DfAM education in terms of discipline
(e.g. engineering, design), country, etc.
 Elucidating educators’ understanding of DfAM, taught DfAM-related topics out of a tentative list
extrapolated from Thompson et al. (2016), familiarity with DfAM, and agreement on possible
definitions of DfAM, which had been previously collected in the literature.
The survey was completed in spring 2020 by 100 internationally distributed educators that teach AM
and claimed to be familiar with DfAM. Out of this sample, 71 educators include DfAM in their taught
content. The issues in the last bullet point above were investigated to address the problems identified
in this paper. In particular, we considered the answers to the two survey questions below.
1. “What does ‘Design for Additive Manufacturing’ mean to you?”, which represents an open
question and, therefore, the respondents were free to answer in an unconstrained way — the
analysis of the answers to this question is given in Section 4.1.
2. How pertinent are the definitions or topics below to your understanding of ‘Design for Additive
Manufacturing’?”, which was followed by 17 randomly ordered alternative DfAM meanings
extracted from the literature and largely inferable from (Thompson et al., 2016). The respondents
answered through a five-point Likert scale ranging from “not at all pertinent” to “extremely
pertinent”. The analysis of the answers to this question is given in Section 4.2., which includes
the 17 options.
The latter followed the former to avoid bias and maximize respondents’ freedom of answering the first
question.
4 RESULTS
4.1 Open-ended definitions of DfAM
The open-ended questions were analysed using Nvivo (version 12). Keywords indicating individual
concepts were identified and coded. To gain additional insight, an affinity diagram was developed to
categorize the codes that were cited twice or more. Affinity diagrams are useful for organizing ideas

ICED21 1941
into categories based on their underlying semantic similarity and assist the identification of patterns
and categories that exist in qualitative datasets (Pyzdek, 2003).
To develop the affinity diagram shown in Table 1, one of the authors worked independently to code
the entries and cluster them into categories. At that point, the preliminary categorisation was shared
with other two authors, who independently reviewed the pertinence of the coding and clustering.
Thereafter, the three authors met and discussed the categories to reach a final consensus., Each author
has at least 5 years’ research experience in the field of DfAM evidenced by peer reviewed papers.,
The content analysis provided 241 different codes. Codes were subsequently compared to determine
which belonged together, thereby forming a Category. Category names are short overarching
keywords. The codes were grouped into 7 broader categories (Erlingsson and Brysiewicz, 2017).
Table 1 reports the categories with at least five entries (the number of entries in brackets), while the
codes with one entry only are grouped for the sake of brevity.
Table 1 Definitions of DfAM (n = 241 themes, n = 100 respondents).
Category
Codes
Frequency
Outcomes (63)
Parts
18
Products
12
Performance
10
Component
4
Desired function
3
New functionalities
2
Lightweight
2
Other subcategories with an entry
12
Design approaches (48)
Design rules
10
Design methods
9
Design tools
4
Design guidelines
4
Design thinking
3
Design process
3
Other subcategories with an entry
15
Opportunistic DfAM (43)
Capabilities
10
Advantages
8
Benefits
6
Design freedom
5
Potentials
3
Use AM freedoms
2
Investigating design possibilities
2
Understand and capitalise opportunities
2
Other subcategories with an entry
5
Restrictive DfAM (42)
Design for Manufacturability
13
Limitations
10
Constraints
7
Restrictions
4
Factors to consider
2
Other subcategories with an entry
6
Individual AM processes (7)
Subcategories with an entry
7
Optimisation (6)
Optimal use of the manufacturing process
2
Other subcategories with an entry
4
Uncategorized (32)
32
Four main categories ‘Outcomes’, ‘Design approaches’, ‘Opportunistic DfAM’ and ‘Restrictive
DfAM’ were identified from the codes while the categories ‘Individual AM processes’ and
‘Optimisation’ were also identified, but by a limited number of codes. The majority of the elicited
definitions included the outcomes of DfAM. These consisted of generic and widely mentioned terms
such as ‘parts’ and ‘products’ (illustrative definitions given by respondents are reported in italics in the
followings).

1942 ICED21
‘Any tools or methods that are useful for synthesizing and/or refining parts that can be fabricated with
AM successfully and/or make use of the freedoms enabled by AM’,
More specific product-related outcomes such as ‘lightweight’, ‘quality’, ‘price’, ‘complexity’, were
also frequent.
‘Designing a part with lightweight, optimal shape (complex) and manufacturable using AM’.
The ‘Design approaches’ to achieve these outcomes were also widely reported in the gathered
definitions. These included a wide variety of approaches and synonyms such as ‘Design rules’,
‘Design methods’ and ‘Design tools.’
‘Any tools or methods that are useful for synthesizing and/or refining parts that can be fabricated with
AM successfully and/or make use of the freedoms enabled by AM.’
The interpretation of DfAM as Restrictive or Opportunistic was also well recognized among our
participants, although these classes cannot be considered as fully representative of DfAM-related
principles. As evident from Table 1, the distribution of definitions compliant with the two classes is
extremely balanced. The most cited Opportunistic DfAM meanings were generic synonyms of
opportunities such as ‘Capabilities’, ‘Advantages’, ‘Benefits’, etc.
‘Exploiting the potential of AM in the design step and optimize geometries and shape.’
In terms of Restrictive, the most cited terms were ‘Design for manufacturability’ and ‘Limitations’,
‘Constraints’ and ‘Restrictions’.
‘The process of designing parts or products to take into account the constraints and abilities of an
additive manufacturing process.’
Besides the mention of specific AM technologies, which the authors judge misleading to characterize
DfAM, and optimization-oriented definitions, there was a long tail of codes that did not clearly match
the categories and other codes, or they were mentioned too rarely to justify a new category. These
themes included for instance ‘Adapted design to end-user’, ‘Life cycle requirements’ or ‘New
materials’.
‘Basically, it can be design methods or tools which consider the functional and mechanical
performance and key product life-cycle requirements, including manufacturability, complexity,
reliability, time and cost can be optimized leveraging on the unique capabilities of additive
manufacturing.’
4.2 Evaluated pertinence of DfAM meanings
The pertinence evaluations attributed by the 100 respondents to the 17 DfAM meanings (see Table 2)
were transformed into 17 corresponding ordinal variables. Consistent with the Likert scale used, the
values attributed to these variables ranged from 1 to 5. The statistical analyses that follow were
performed with the software Stata SE 13.
The variables were firstly used to verify the redundancy of the meanings by submitting them to a
Spearman correlation. As a common rule of thumb, values over 0.8 could be considered as an (almost)
perfect agreement (Landis and Koch, 1977), which, in the specific case, could be interpreted as a
repetition of a definition because of a systematic repetition of pertinence evaluations. The correlation
values ranged from -0.03 to 0.75, which led us to conclude that none of the variables’ pairs represent
fully overlapping DfAM-related concepts.
Therefore, to merge these meanings and capture independent dimensions of DfAM definitions, the
same variables were processed with a Principal Component Analysis (PCA). The PCA provides a
number of independent dimensions, namely the components, which are featured by their capability of
describing the variability of the sample, which is expressed by the eigenvalue and, accordingly, the
proportion of explained variability. The extracted components are a linear combination of input
variables (here the stated agreement on DfAM meanings) and their calculated weights. In the
presented case, each respondent could therefore feature a number of independent variables and their
associated values. However, the key point here is to extract and characterize the fundamental
independent components and interpret their significance to DfAM.
As a common rule of thumb, those PCs reporting an eigenvalue greater than 1 were processed further
to infer the relevant independent dimensions, leading us to focus on five PCs accounting for 67.7% of
the data variability. The weights of the original components (first column) on the five selected PCs are
reported in Table 2.

ICED21 1943
Table 2. Extraction of Principal Components out of the evaluation of pertinence of DfAM
definitions (n=100 respondents)
PCA weights
Meanings
PC1
PC2
PC3
PC4
PC5
1. Design methods aimed at better exploiting the potentials
of Additive Manufacturing
0.11
0.07
0.72
-0.05
-0.11
2. Exploring design methods in which the functional
performance and lifecycle considerations of products are
optimized to Additive Manufacturing capabilities
0.20
0.25
0.33
-0.34
0.02
3. Design freedoms enabled by Additive Manufacturing
0.26
0.03
0.16
-0.21
-0.04
4. Understanding the constraints in the design of parts to be
produced by Additive Manufacturing
0.21
-0.33
0.07
-0.03
0.39
5. Design rules, guidelines and suggestions for parts
conform with Additive Manufacturing production
technologies in light of their current affordances and
constraints
0.19
-0.30
0.21
-0.04
0.31
6. Supporting the choice between traditional and Additive
Manufacturing technologies
0.26
-0.33
0.04
0.10
-0.12
7. Supporting the choice between the most suitable AM
process for the production of parts or products
0.25
-0.39
-0.12
0.21
-0.16
8. Part consolidation, reducing the number of components
0.31
-0.07
-0.10
-0.16
0.15
9. Designing parts with lattice or cellular structures
0.27
-0.05
-0.12
-0.34
-0.04
10. Utilising Topology Optimization or targeting the
creation of lightweight structures
0.30
0.00
-0.30
-0.28
-0.07
11. Design of parts with multiple or composite materials
0.24
0.21
-0.28
-0.13
-0.33
12. Design of products with enhanced functionalities that
are enabled by Additive Manufacturing
0.20
0.48
0.02
0.08
-0.08
13. The redesign of a part to make it compatible with
Additive Manufacturing production
0.25
0.20
-0.23
-0.08
0.38
14. The development and optimal use of CAD systems for
the design of parts that match Additive Manufacturing
capabilities
0.24
-0.01
0.04
0.16
-0.45
15. The overcoming of cognitive barriers due to the
knowledge of the limitations of traditional technologies
0.27
0.27
0.12
0.33
0.27
16. The change of thinking style previously imposed by
past experience and conventional fabrication techniques,
which leads to an extension of the design space
0.23
0.22
-0.10
0.57
0.20
17. The research aimed at pushing the boundaries of
Additive Manufacturing technologies and its repercussions
on design capabilities
0.26
-0.16
0.12
0.27
-0.31
The subsequent interpretation of the PCs based on the weight of the original variables led to the
determination of the following DfAM-related concepts (see the attributed names in italics), which are
ordered according to diminishing proportion of variability attributed to each PC (in brackets). Where
applicable, a tentative definition of DfAM classes featured by high (a.) and low (b.) values of the PCs
is given.
1. General (38.0%): the weights of the original variables are substantially evenly distributed; as
such, all the meanings contribute to a general understanding of the whole scope of DfAM
regardless of restrictive or opportunistic nuances.
2. Explorative (9.9%): the variables with the highest positive weights feature the overcoming of
barriers enabled by AM and divergent design processes. Conversely, variables associated with
the most significant negative weights deal with constraints and decision-making processes. This
PC has expectedly high values for those respondents with a predominant opportunistic
interpretation of DfAM.

1944 ICED21
a. Explorative DfAM methods are those intended to guide designers in divergent design
activities keeping in mind the opportunities offered by AM.
b. Non-explorative DfAM includes criteria and tools for making decisions on design
alternatives with a particular focus on designs’ feasibility by means of AM devices.
3. Conceptual (7.1%): the variables with highest positive weights feature theoretical constructs
concerning the importance of exploiting AM potential, as opposed to markedly negative weights
attributed to tasks that involve detailed design phases (topology optimization, choice of
materials).
a. Conceptual DfAM includes techniques that support the consideration of AM potential
in the early design phases.
b. Non-conceptual DfAM tools are those meant to finalize the design of products to be
manufactured with AM.
4. Cognitive (6.6%): the variables with highest positive weights feature DfAM meanings concerning
the change of designers’ mindset and the overcoming of their psychological inertia, as opposed
to markedly negative weights attributed to practical computer-supported design tasks (topology
optimization, lattice structures).
a. Cognitive DfAM involves the research aimed to change designer’s mindset towards
the understanding of benefits enabled by the existence and the progress of AM.
b. Non-cognitive DfAM includes methods and tools that support geometric
modifications of designed parts, whose fabrication is enabled by AM capabilities.
5. Constraint (6.1%): the variables with highest positive weights feature constraints and guidelines
for product design and redesign, mostly oriented to feasibility and compliance with AM
technologies. Conversely, variables associated with the most significant negative weights deal
with new opportunities offered by AM. This PC has expectedly high values for those
respondents with a predominant restrictive interpretation of DfAM.
5 DISCUSSION, LIMITATIONS, AND NEED FOR FUTURE WORK
The analysis of the definitions attributed by the survey’s respondents reveals that product and process
aspects constitute the foci of DfAM. This finding can be interpreted as the relevance of both
procedural/methodological aspects and the outcomes thereof when DfAM is approached. The clear
emergence of assigned definitions or clauses ascribable to opportunistic and restrictive DfAM from
both the affinity diagram and the outcomes of the PCA shows that the scope of both classes is well
internalised by educators in DfAM, irrespective of their awareness and their knowledge of the
terminology introduced by Laverne et al. (2015). It follows that, overall, the answers suggest that for
our sample of educators, DfAM is a process which implies design methods, tools and knowledge to
achieve a specific outcome by exploiting AM characteristics and mitigating its limitations. This is in
accordance with the published definitions of DfAM (e.g. BS EN ISO/ASTM 52910:2019), showing
that there is both a shared agreement of what DfAM means among our sample and between our
sample and the published definitions.
The fundamental alignment between what is generally claimed as being part of DfAM and the
perceived pertinence of meanings emerges as an outcome of the PCA too. Here, the main PC was a
balanced combination of an extensive list of meanings, including opportunistic and restrictive
interpretations, as well as other facets. Taken in isolation, this main PC is capable of explaining 40%
of data variability. This implies that our sample of respondents is attributed varying values on this PC.
This apparently strong result could indicate that the respondents had a different degree of confidence
and familiarity towards definitions included in the DfAM literature. However, some other terms
distinguishing educators’ understanding of DfAM emerged. While opportunistic and restrictive
nuances arose, other fundamental interpretations of DfAM are worth highlighting. Other
interpretations of the study outcomes include:
 DfAM methods might be attributed to the capability of both expanding the design space enabled
by AM and supporting decision-making, especially when it comes to selecting manufacturing
technologies.
 DfAM methods might be featured by their varying level of abstraction.
 DfAM ranges from actions to increase engineers’ and designers’ awareness of the potential
benefits of AM to practice-oriented tools targeting redesign and operating on 3D models.

ICED21 1945
This study is inherently affected by some limitations. On the one hand, the respondents constitute a
large sample of convenience, but the generalizability of the results can be questioned, as the sample is
not sufficiently representative of the total engineering educator population. Moreover, the authors have
already underlined the reasons for targeting educators in a first instance, but their views could be
partially misaligned with the viewpoint of other stakeholders, such as researchers, industrialists,
officers of standards bodies, manufacturers of AM machines. While the respondents to our survey
possessed a range of confidence in DfAM, their confidence level was not been considered in this first
analysis. The authors’ future work will nevertheless investigate the views of a broader range of
stakeholders. It will be of particular interest to investigate if the distinction between opportunistic and
restrictive DfAM is known or intuitively understood also in the AM community at large and markedly
outside of the design field. As a further limitation, the processes followed to create categories of
definitions and interpret PCs involve subjectivity. Although we have followed best practice and
worked in an unbiased way, the analysis performed by different research groups could lead to different
outcomes. Acknowledging this, we offer to share our data with researchers willing to repeat the
analyses and/or to contribute to the formalization of the DfAM concept and its classifications.
6 CONCLUSIONS
Based on the above discussion of the results, we propose the following final conclusions in response to
the research objectives.
 The overall interpretation of DfAM amongst a large international sample of expert educators is in
line with the literature and especially with the first attempts to establish standards, in that
outcomes, processes, opportunities and constraints are well recognized.
 Neither opportunistic nor restrictive interpretations of AM dominated, despite some respondents
expressing a preference for the former or latter. The two classes could be isolated as independent
clusters in the PCA, a general knowledge of all DfAM facets appears prevalent among the
respondents.
 Other classification criteria, beyond opportunistic and restrictive DfAM methods were found.
Although these new classes are not necessarily overlooked within DfAM, our results reveal these
additional criteria and show that the opportunistic/restrictive distinction is insufficient to describe
the whole breadth of DfAM.
The last bullet point indicates the need for the for standardization the scientific formalization of the
field of DfAM without neglecting the breadth of DfAM. In this respect, while many researchers focus
on practical aspects and markedly on products, whose design and fabrication is enabled by
opportunistic DfAM and constrained by restrictive DfAM, the conceptual and intellectual endeavour
behind the development of DfAM (see some well-recognizable PCs) cannot be overlooked.
Consequently, more studies are needed to validate the existence of DfAM categories and classification
criteria. The expected final outcome will be the formalization of methods and concepts belonging to
DfAM classes, articulated according to the opportunistic or restrictive distinction, and, if further
evidence is found, the additional criteria that have been elicited in this paper.
REFERENCES
Barclift, M., Simpson, T.W., Alessandra Nusiner, M., Miller, S. (2017). “An Investigation into the Driving
Factors of Creativity in Design for Additive Manufacturing”, in: IDETC-CIE2017. Volume 3: 19th
International Conference on Advanced Vehicle Technologies; 14th International Conference on Design
Education; 10th Frontiers in Biomedical Devices. https://doi.org/10.1115/DETC2017-68395
Blösch-Paidosh, A., Shea, K. (2019). “Design Heuristics for Additive Manufacturing Validated Through a User
Study”, Journal of Mechanical Design Vol. 141 No.4 p. 041101. https://doi.org/10.1115/1.4041051
Booth, J.W., Alperovich, J., Chawla, P., Ma, J., Reid, T.N., Ramani, K. (2017). “The Design for Additive
Manufacturing Worksheet”, Journal of Mechanical Design Vol. 139 No. 10, p. 100904.
https://doi.org/10.1115/1.4037251
Boothroyd, G. (1994) “Product design for manufacture and assembly”, Computer-Aided Design, Vol. 26 No. 7,
pp. 505-520. https://dx.doi.org/10.1016/0010-4485(94)90082-5.
BS EN ISO/ASTM 52910:2019 (2019) BSI Standards Publication Additive manufacturing - Design -
Requirements, guidelines and recommendations’
Design Council (2005) Eleven lessons: managing design in eleven global brands A study of the design process.
London.

1946 ICED21
Erlingsson, C. and Brysiewicz, P. (2017) “A hands-on guide to doing content analysis”, African Journal of
Emergency Medicine, Vol. 7 No. 3, pp. 93–99. https://dx.doi.org/10.1016/j.afjem.2017.08.001.
Kim, S., Rosen, D. W., Witherell, P. Ko, H. (2019). “A design for additive manufacturing ontology to support
manufacturability analysis”, Journal of Computing and Information Science in Engineering, Vol. 19 No. 4,
p. 041014. https://doi.org/10.1115/1.4043531
Kumke, M., Watschke, H., Hartogh, P., Bavendiek, A.-K., Vietor, T. (2018). “Methods and tools for identifying
and leveraging additive manufacturing design potentials”, International Journal of Interactive Design and
Manufacturing Vol. 12, pp. 481–493. https://doi.org/10.1007/s12008-017-0399-7
Landis, J. R., Koch, G. G. (1977). “The measurement of observer agreement for categorical data”, Biometrics,
Vol. 33, pp. 159-174. https://doi.org/10.2307/2529310
Laverne, F., Segonds, F., Anwer, N., Le Coq, M. (2015). “Assembly Based Methods to Support Product
Innovation in Design for Additive Manufacturing: An Exploratory Case Study”, Journal of Mechanical
Design Vol. 137 No. 12, p. 121701. https://doi.org/10.1115/1.4031589
Laverne, F., Segonds, F., D’Antonio, G., Le Coq, M. (2017). “Enriching design with X through tailored additive
manufacturing knowledge: a methodological proposal”, International Journal of Interactive Design and
Manufacturing Vol. 11, pp. 279–288. https://doi.org/10.1007/s12008-016-0314-7
Mani, M., Witherell, P., Jee, H. (2017). “Design rules for additive manufacturing: a categorization”. In
International Design Engineering Technical Conferences and Computers and Information in Engineering
Conference. American Society of Mechanical Engineers. https://doi.org/10.1115/DETC2017-68446
Meisel, N.A., Woods, M.R., Simpson, T.W., Dickman, C.J. (2017). “Redesigning a Reaction Control Thruster
for Metal-Based Additive Manufacturing: A Case Study in Design for Additive Manufacturing”, Journal of
Mechanical Design Vol. 139 No. 10, p. 100903. https://doi.org/10.1115/1.4037250
Prabhu, R., Brackena, J., Clinton B, A., Jablokowa, K., Simpson, T.W., Meisel, N.A. (2020a). “Additive
creativity: investigating the use of design for additive manufacturing to encourage creativity in the
engineering design industry”, International Journal of Design Creativity and Innovation Vol. 8 No. 4,
pp. 198–222. https://doi.org/10.1080/21650349.2020.1813633
Prabhu, R., Miller, S.R., Simpson, T.W., Meisel, N.A. (2020b). “Fresh in My Mind! Investigating the Effects of
the Order of Presenting Opportunistic and Restrictive Design for Additive Manufacturing Content on
Creativity”. Presented at the ASME 2020 International Design Engineering Technical Conferences and
Computers and Information in Engineering Conference, American Society of Mechanical Engineers Digital
Collection. https://doi.org/10.1115/DETC2020-22449
Prabhu, R., Miller, S.R., Simpson, T.W., Meisel, N.A. (2018a). “Teaching Design Freedom: Exploring the
Effects of Design for Additive Manufacturing Education on the Cognitive Components of Students’
Creativity”. Presented at the ASME 2018 International Design Engineering Technical Conferences and
Computers and Information in Engineering Conference, American Society of Mechanical Engineers Digital
Collection. https://doi.org/10.1115/DETC2018-85938
Prabhu, R., Miller, S.R., Simpson, T.W., Meisel, N.A. (2018b). “The Earlier the Better? Investigating the
Importance of Timing on Effectiveness of Design for Additive Manufacturing Education”. Presented at the
ASME 2018 International Design Engineering Technical Conferences and Computers and Information in
Engineering Conference, American Society of Mechanical Engineers Digital Collection.
https://doi.org/10.1115/DETC2018-85953
Pyzdek T. (2003). The six sigma handbook, McGraw-Hill, New York
Reichwein, J., Vogel, S., Schork, S., Kirchner, E., 2020. “On the Applicability of Agile Development Methods to
Design for Additive Manufacturing”, Procedia CIRP Vol. 91, pp. 653–658.
https://doi.org/10.1016/j.procir.2020.03.112
Thompson, M.K., Moroni, G., Vaneker, T., Fadel, G., Campbell, R.I., Gibson, I., Bernard, A., Schulz, J., Graf,
P., Ahuja, B., Martina, F. (2016). “Design for Additive Manufacturing: Trends, opportunities,
considerations, and constraints”. CIRP Annals Vol. 65 No. 2, pp. 737–760.
https://doi.org/10.1016/j.cirp.2016.05.004
Watschke, H., Bavendiek, A.-K., Giannakos, A., Vietor, T. (2017). “A methodical approach to support ideation
for additive manufacturing in design education”. In: DS 87-5 Proceedings of the 21st International
Conference on Engineering Design (ICED17), 041–050.
Zhu, Z., Anwer, N., Mathieu, L. (2017). “Deviation Modeling and Shape transformation in Design for Additive
Manufacturing”. Procedia CIRP Vol. 60, pp. 211–216. https://doi.org/10.1016/j.procir.2017.01.023