Progress in Additive Manufacturing (2018) 3:145–163

https://doi.org/10.1007/s40964-017-0034-y

FULL RESEARCH ARTICLE

Using finite element analysis to influence the infill design of fused

deposition modelled parts
James A. Gopsill1 · Jonathan Shindler2 · Ben J. Hicks2

Received: 5 May 2017 / Accepted: 21 November 2017 / Published online: 28 November 2017
© The Author(s) 2017. This article is an open access publication

Abstract
Additive manufacturing (AM) has and continues to experience considerable market and technological growth with many
forecasting a tripling in market value over the next decade. One of the primary drivers for this growth is the increased freedom
afforded to the design of both the external form and internal structure of fabricated parts. This freedom presents greater oppor-
tunities in optimising a parts mechanical properties, (such as strength, stiffness and mass), which in turn leads to enhanced
performance whilst potentially reducing material use and hence, environmental impact. Realising this potential will further
increase the viability of AM for a greater range of engineering contexts. Correspondingly, the contribution of this paper lies
in the creation and validation of a method for the topological optimisation of the infill structure of fused deposition modelled
(FDM) components. The proposed method uses results attained from finite element analysis (FEA) to influence the design
of the internal structure (i.e. infill) by locally varying the composition of the infill based upon the associated stress values.
This paper presents and discusses the proposed method, and demonstrates the generalisability of the method through its
ability to handle complex geometries and loading conditions, and manufacturing process constraints. In addition, the paper
validates the method through testing of FDM beams comprised of FEA influenced and standard honeycomb infill designs
undergoing four different loading scenarios. The validation reveals that a three and a half times increase in strength can be
achieved where the stress profiles are well defined within the structure. In addition, the FEA-influenced beams exhibited
more consistent failure mode profiles, which maybe desirable for designing parts with specific failure mode characteristics.

Keywords Topology optimisation · Internal geometry · Infill · Internal structure · Fused deposition modelling (FDM) ·
Finite element analysis (FEA)

1 Introduction • highly customised and individualised consumer products;

and,
Additive manufacturing (AM) continues to experience con- • rapid prototyping products in engineering design consul-
siderable market and technological growth with forecasts tancies.
indicating a tripling in market value over the next decade
[1]. In this market, fused deposition modelling (FDM) is one One of the main affordances of AM is the capability to pro-
of the most commonly applied techniques [2] and has been duce highly complex external and internal (infill) geom-
used in a variety of contexts including: etries that enable engineers to optimise the deposition of
material for a given application. This provides benefits in
• supporting and facilitating engineering degree courses; terms of reducing component cost through a decrease in both
• creating novel structures for bio-medical research; material use and manufacturing time. It is also a common
goal for manufacturers to optimise the performance of their
* James A. Gopsill parts through their strength-to-weight ratio. The prevalence
[email protected] of which is exemplified by the extensive literature on the
subject of topology optimisation (see, for example, [3–5]).
1
Department of Mechanical Engineering, University of Bath, Although many techniques and algorithms exist for
Bath, UK
defining the theoretical optimum topology of a compo-
2
Department of Mechanical Engineering, University nent, the ability to translate the optimised geometry into
of Bristol, Bristol, UK

13
Vol.:(0123456789)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

146 Progress in Additive Manufacturing (2018) 3:145–163

Fig. 1  Infill strategies from MakerWare (a–e) and Slic3r (f–j) software

a manufacturable component continues to be a challenge. though this will not lead to a truly optimised structure, it
This is due to the constraints imposed by the manufactur- is argued that significant improvements can be attained in
ing techniques being applied. In the case of selective laser terms of the strength-to-weight ratio of the component. In
sintering (SLS), research has already begun to overcome addition, the components’ external form may be crucial to
these challenges. In comparison, the development of this its function, such as brand representation [10]. By provid-
underlying capability for FDM remains relatively immature ing an optimisation process that can meet these constraints
[6–8]. This is in part due to the differing manufacturing con- will lead to further democratisation of the design process by
straints between the processes preventing the same topology enabling human-machine co-creation, where the engineer
optimisation techniques being applied. focuses on the external geometry and the machine focuses
In contrast to SLS, where the process involves the layer- on the internal geometry.
ing of granular material and sintering at specific locations, Consequently, this paper describes a process that uses
FDM involves the layer-by-layer deposition of material and the results from FEA to influence the automatic genera-
this introduces the challenge of defining the deposition path tion of a components infill. The paper first discusses related
as the layers are highly dependent upon the previous to pro- work concerning the design and manufacture of AM parts,
vide support for the material being extruded. Rafts, supports which has primarily focused on the accuracy and mechani-
and bridging are all techniques that have been introduced to cal properties of parts, and design of infill patterns. The
enable the manufacture of geometrically complex parts using paper then continues by discussing the proposed process
FDM. It is argued that the very nature of extruding material of using FEA to influence the design of the infill where the
is a key-contributing factor to why two-dimensional repeat- focus has been on manufacturability, generalisability and
ing patterns are commonly applied for the infill of FDM automation. To evaluate the proposed method, a compari-
components (Fig. 1). This is alongside the reduction in com- son of the stress–strain profiles of parts with hexagonal and
putational time to produce the part as the pattern is consist- FEA-influenced infill has been conducted. The paper then
ent across layers and thus, only the intersections between the concludes by expressing the key findings of this method as
pattern and model surfaces have to be calculated. In addi- well as discussing potential areas of future work.
tion, the infill percentage can be easily controlled by scaling
the pattern.
Given the extensive literature on topology optimisation 2 Related work
and the ability to manipulate the internal structure of FDM
parts, it is contended that there is potential to influence the To date, the majority of research concerning the optimisa-
infill design of a part using finite element analysis (FEA) tion of fused deposition modelled (FDM) parts has focused
results whilst also considering the manufacturing constraints on either the accuracy of the manufactured part compared
of FDM and not effecting the external form of the compo- to the digitally designed model, or strength of the part
nent. Factoring the manufacturing constraints within design given changes to the parameters within common slicing
tools has been highlighted as a key requirement in empower- approaches (for example, infill percentage, infill orientation
ing the development of innovative products [9]. And, even and layer thickness).

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 147

Fig. 2  Tensile strength of FDM ABS compared to injection moulded

(from: [11])

The accuracy of the part to its digital counterpart has been

analysed by Sood et al. [12] who applied Taguchi’s design of
experiments on five common parameters1 of the FDM pro-
cess and determined that the shrinkage phenomena is most
influential. This has been confirmed by Lanzotti et al. [13]
Fig. 3  Topology-optimised parts using AM technologies
in their study of the RepRap open-source 3D printer, which
revealed that layer thickness and the consequent shrinkage
was the leading factor in attaining geometrically accurate of studies have examined tensile and flexural strength of
parts. It was also revealed that to optimise accuracy in all FDM parts with all results corroborating the fact that layer
Cartesian co-ordinates generally necessitates a compromise thickness is the most influential factor [23–27]. However,
as optimising in a single direction negatively impacts on decreasing the layer height to increase strength has a nega-
the others. Further, analysis of part accuracy by Pennington tive impact on the surface roughness. Thus, a compromise
et al. [14] highlighted that part position on the bed space, must again be sought. In addition, reductions in layer height
part size and temperature of the work area also contribute can increase factors such as distortion, layer cracking and
significantly to accuracy. To improve the accuracy of the de-lamination leading to premature part failure [23]. While
printed part, Gregorian et al. [15] found that the optimum Jones et al. [24] highlight that Acrylonitrile Butadiene
shrinkage compensation factor to be applied to the manufac- Styrene (ABS) has been shown to lose 20% of its strength
turing process was 1.007 for their particular machine. Whilst when extruded using FDM whilst polylactic acid’s (PLA’s)
work by Zhang and Chou [16] have looked to model the strength remains unaffected.
shrinkage phenomena through FEA to automatically gener- Initial research into topology-optimised FDM structures
ate the compensation factors for FDM machines. In addition, attempted to emulate SLS techniques with Leary et al. [18]
Pandey et al. [17] show that improvements to accuracy and discovering that considerable use of rafts and supports were
quality can be made through adaptive layer height control required to produce the component (Fig. 3a). And even with
in the generation of the tool-head path. This highlights the these additions, there was a high likelihood of the part failing
potential to optimise part geometry through adjustment of during manufacture. In addition, those that were manufac-
the numerical control rather than the machine itself. tured successfully produced a considerable amount of waste
In terms of investigating the mechanical properties of material and required significant post-processing time in
parts produced by FDM, it has been shown that the deposi- terms of removing the support material. In contrast, Rezaie
tion process itself has profound anisotropic effects on the et al. [19] and Lei et al. [28] have developed techniques that
part, which leads it to behave much more like a laminate [20, locally varies the thickness of the linear mesh within a FDM
21]. Ahn et al. [11] show that a decrease of 80% in tensile part (Fig. 3b). Their results show that the approach can lead
strength can occur if the material is deposited in an inap- to potential gains of up to 25% in the strength-to-weight ratio
propriate orientation (Fig. 2). This has since led to the devel- of the component.
opment of algorithms to automatically determine the opti- This growing interest in the optimisation of internal
mum orientation for a given part [22]. In addition, a number geometry based on the predicted loading scenario has also
been investigated by Gopsill and Hicks [6], who demonstrate
that a 79% increase in the stiffness of a part can be attained
1
layer thickness, air gap, raster angle, orientation and raster width through the alignment of infill to the direction of stress

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

148 Progress in Additive Manufacturing (2018) 3:145–163

Fig. 4  Infill designs generated from the predicted planar stress profiles of parts (from: [6])

(Fig. 4). In addition, the topology-optimised parts showed manufacturing constraints of FDM into account, does not
more consistent buckling and failure mode behaviour. This alter the external geometry, is applicable to as many com-
phenomena gives rise to the potential for designing parts ponents as possible, and requires minimal user intervention.
that provide visual indicators of structures exceeding their The strategy comprises of five stages:
designed loading and/or sacrificial parts that prevent com-
ponents from being damaged if an engineering system oper- 1. Model construction
ates outside its normal range. Although the potential benefits 2. Determining mesh size and perimeter identification
have been demonstrated, the process is still limited to planar 3. Linear infill design
stress profiles and has yet to be evaluated for more complex 4. FEA-influenced infill design
loading scenarios. 5. G-code generation
In summary, the related work reveals that the majority
of current research has focused on optimising the manu- The first stage consists of the construction of the stereo-
facturing process with regards to enhancing the accuracy, lithography (STL) file detailing the geometry of the part
tolerance, print time and material use. In addition, there alongside the exported results from a Finite Element model
is a growing trend towards studies that provide processes in the form of a comma-separated variable (csv) file (Fig. 6a,
that optimise the topology of a component. Supporting the b). It is important to note that the role of the FEA is to guide
design of components through FDM remains an emerging the design of the infill in stage four and does not reflect the
field with many potential areas that could be explored. In actual stress within the final structure. Once constructed,
particular, it is contended that the results from modelling the process can continue to stage two where the optimum
techniques, such as FEA, have the potential to positively internal mesh sizing is derived. This is followed by a loop
influence the internal geometry of a component. Conse- that generates the relevant G-Code for each deposition layer.
quently, the contribution of this paper is the presentation The start of each deposition layer involves the identification
of a method that uses the results from FEA to influence the of the geometry perimeter (Fig. 6c) and the generation of the
generation of the internal geometry of FDM parts. related number of shells that will be produced. Shells being
the numbers of times the FDM machine will deposit material
to form the perimeter of the part where an offset is applied to
3 Influencing the infill design using FEA ensure no overlapping of material occurs (Fig. 6e).
Once completed, the process moves to the third stage
The process for influencing the infill design for generic FDM which generates the linear infill design. The objective of the
parts is presented in Fig. 5 alongside an illustration of the linear infill is to provide a consistent structure that the FEA-
outputs from the various stages in Fig. 6. The key objec- influenced infill design (stage four) can attach and bridge
tive of the process is to use the results from FEA to influ- across (Fig. 6g). This is an important feature as it enables
ence the infill design whilst ensuring the strategy takes the the FEA-influenced infill designs to alter as the layer height

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 149

Fig. 5  Infill designs workflow

increases and thus, provide appropriate support for three- is achieved by extending the linear infill design to form
dimensional loading scenarios (this is demonstrated and a lattice design. Although limiting the topology to a lat-
later discussed in Fig. 13). The linear infill is also used to tice design will prevent the strategy from producing truly
counter the potential of introducing instabilities and unsup- optimal solutions, lattice structures have been selected due
ported features by not placing material within low stress to their:
areas [3].
Stage four continues the development of the deposition • proven capability to transfer and withstand high load-
layer by generating the FEA-influenced infill design. This ing;

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

150 Progress in Additive Manufacturing (2018) 3:145–163

Fig. 6  Illustrative examples of the stages taken to generate an infill design that has been influenced by FEA

• highly connected nature that enables bridging, which is • ability to be fitted to complex external geometries of
important as the optimised internal structure may evolve FDM components; and,
layer-on-layer (c.f. Fig. 13); • generation being computationally inexpensive when
• ability to be reliably manufactured using FDM; compared to more complex geometries such as the opti-

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 151

Fig. 7  Examples geometries for

strategy

Fig. 8  Finite element analysis of

beam and bracket

mal positioning of bezier splines along force pathways particular, the bracket geometry is non-uniform along the
[6]. axis of print, contains multiple-perimeter geometry (for
example, holes) and has been placed under a three-dimen-
The local stress values are then mapped to the lattice ele- sional loading scenario.
ments (Fig. 6f) and a removal of elements with least stress
occurs until the desired infill percentage is met (Fig. 6h). 3.1 Model construction (stage one)
Following this, a post-processing of the infill occurs where
the objective is to identify and remove elements that are The initial step is to create the computer-aided design
unable to support any loading as they terminate without con- (CAD) model of the part to be optimised. From this model,
necting to the rest of the structure. both the STL file for the identification of the part perim-
Upon completion of the post-processing, merging of the eter and the FEA model are generated. In the case of this
FEA influenced and linear infill occurs and involves the paper, AutoDesk Inventor 2015 and AutoDesk Simulation
union of the two designs. The final aspect of Stage Four is Mechanical 2015 were used.
to then define the optimum print path for the resulting infill The process uses the FEA model to perform an evolution-
design to reduce the time taken to manufacture the part. The ary structure optimisation (ESO) of a continuous structure
process then progresses to Stage Five, where the G-Code is within a design domain [29, 30]. The design domain being
written and is either used directly in the printing process or the model geometry with the loads being applied and the
further processed into the relevant binary print file. premise is to identify the load paths through the continuum
The following sections provide further details of each structure. These load paths highlight the areas where it is
stage alongside an example of a beam being optimised for a desirable to maintain a high concentration of material and
three-point bend test, and a bracket in a contrived 3-dimen- the non-load paths are the areas where material could be
sional stress scenario. The STL geometries for both exam- reduced or removed.
ples are shown in Fig. 7. The beam has been chosen as it is Figure 8a shows the von Mises stress calculated by the
used later to evaluate the process in terms of the potential FEA analysis for a three-point bend test with the contours
performance improvement. Whilst the bracket is consid- highlighting lines of constant stress. It can be seen that there
ered to demonstrate the generalisability of the process. In are two main load paths that both emanate from the applied

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

152 Progress in Additive Manufacturing (2018) 3:145–163

Fig. 9  Mesh size and perimeter

identification
Perimeter
Offset

Modulo

Mesh Width
z y

x
Appropriate Non-Appropriate

(a) Determining the appropriate mesh sizing (b) Cutting planes through the beam STL

load and go directly to the points of reactions at the pin- both the x- and y-directions for the range of mesh sizing. The
joints. Less obvious, although visible due to the contours, is x and y lengths are determined by the minimum and maxi-
an area of high stress on the underside of the beam, between mum values for each direction within the STL model along-
the two reaction locations. As the beam is under high com- side an offset to ensure the inner mesh does not interfere
pressive loads, this area is undergoing significant bending with the perimeter when being printed. Once the modulo for
stress that also needs to be taken into account in the optimi- both dimensions are calculated, these are summed and form
sation of the beam structure. It is also interesting to note that the error for the given mesh sizing. The mesh sizing that
the stress contours are almost but not completely symmetri- minimises the error. This optimisation works particularly
cal even though this is the expectation. This can be attributed well for flat faces along the Cartesian co-ordinates but has
to the irregularities in the mesh construction of the FEA and not been fully evaluated for curved surfaces.
accuracy of the boundary conditions. In many cases, FEA Once the mesh sizing has been determined, the process
analysis is used as a guide to show potential areas of high- can now iterate through each deposition layer and generate
stress within a component and real-world validation is often the deposition path. For each layer, there is a need to identify
required to ensure that it accurately reflects the real-world the perimeter of the part. This has been achieved through the
product. In this case, the FEA is used as a guide to influence use of the visualisation tool-kit (VTK) Python library [31].
the infill design and thus, does not have to be a fully vali- A slicing plane is generated using the library and the inter-
dated model. Figure 8b demonstrates the three-dimensional sections between the plane and the STL model are identified
loading case of the bracket that is later used to explore the (Fig. 9b). A set of lines is generated from where the plane
generalisability of the proposed process. intersects the individual polygons of the STL file and it is
The results from the FEA analysis, which include the Car- then the case of identifying the connecting path between the
tesian co-ordinates of the nodes and their associated stress lines. This is achieved by taking a single line and identifying
values are exported as a csv file. Both the STL and csv file the neighbouring line by finding the end points that match.
are then imported into the script to generate the influenced The process then continues to the end of the next line and
infill design for the part. repeats until the sequences of lines returns to the starting
line. This forms a perimeter polygon for the part. There is
3.2 Mesh size and perimeter identification (stage then a check to see whether any lines remain that do not
two) form part of the chain. If so, a new chain is generated with
the remaining lines. This cycle repeats until all the lines
To generate the internal structure of the part, it is first neces- are associated to a chain. This enables the identification of
sary to establish an appropriate mesh sizing before continu- complex geometry with multiple perimeters such as a part
ing into the loop where the infill design for each deposition containing holes and/or cut-outs, such as that depicted in
layer is generated. This is a common task that all slicing Fig. 10.
routines perform ahead of generating the infill design for
the component. 3.3 Linear infill design (stage three)
To establish the mesh sizing for the print file, the process
attempts to ensure that the perimeter of the part interfaces The third stage of the process involves the generation
with a complete mesh square of the infill (as demonstrated of a set of uniformly distributed vertical and horizontal
in Fig. 9a). To achieve this, the user manually selects a mini- lines to form the linear infill design for the part. The
mum and maximum threshold for the mesh sizing, and the linear infill exists to provide a consistent structure for
process discretises this range and calculates the modulo in the FEA-influenced infill design to attach to. As the

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 153

Fig. 10  Identifying multiple

perimeter geometry during
slicing

Perimeter One
Perimeter Two
Perimeter Three

(a) Slicing of bracket geometry (b) Multi-perimeter identification

Fig. 11  Generating a linear

mesh across a deposition layer y
x

(a) Beam Ray Casting (b) Bracket Ray Casting

(c) Beam Linear Mesh (d) Bracket Mesh

FEA-influenced infill design may change with the layer 3.4 FEA‑influenced infill design (stage four)
height, the linear infill enables bridging to occur and thus,
ensures the material is deposited in the desired locations. The fourth stage involves the generation of the infill design
The linear infill generation utilises the mesh offset and that has been influenced by the FEA results. To start, the
mesh width variables defined in Stage Two, and the ray linear infill design is taken from Stage Three and extended
casting functionality of VTK. The strategy generates rays to include diagonal elements. This forms a lattice structure
along the x- and y-axes for the given layer height, and across the entire deposition layer and represents the poten-
identifies the intersections of rays against the STL model tial elements that could be deposited by the FDM machine.
of the part (Fig. 11a, b). For complex geometries, the rays To decide which elements are likely to provide the most
have the potential to intersect the model multiple times benefit for the given loading, a weighting is applied to
and thus, the lines of interest are determined by pairing each element. This weighting is calculated by taking the
intersecting points as the ray passes through the model average localised stress value from the FEA. Figure 12a
(i.e. points 1 and 2 will form a path and the same goes provides an example of the potential elements with the
for 3–4, 5–6, etc...). Fig. 11 illustrates the linear infill weighting denoted by the shade of grey.
generation for the beam and bracket.

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

154 Progress in Additive Manufacturing (2018) 3:145–163

Fig. 12  Influenced infill design generation

Fig. 13  Strategies capability to handle a 3-dimensional stress case

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 155

It is then a case of taking these print lines and forming the

code that a FDM machine can use to manufacture the part.

3.5 G‑code generation (stage five)

The fifth and final stage of the process involves the genera-
tion of the G-Code for the FDM printer. For each deposition
layer, the perimeter path G-Code is generated, followed by
the combined printer path from the merging of the linear
Fig. 14  Print line generation and FEA-influenced infill designs. The pre- and post-amble
G-Code for the specific FDM printer is then inserted, result-
ing in the final G-Code file.
The elements are then ordered by their stress values
and the elements with least stress are removed until the 3.6 Summary
desired infill percentage is attained (Fig. 12b). Follow-
ing this, a post-process is performed on the infill design, This section has presented, in detail, a five-stage process
where the remaining elements are analysed once more to that integrates the results of FEA into the infill generation of
remove elements that contain a single degree of connec- FDM parts. The key features of the strategy are that it takes
tivity (Fig. 12b, i). This is due to the fact that an element the manufacturing constraints of FDM into account, does not
requires two degrees of connectivity for load transfer. alter the external geometry and is able to handle complex
Figure 12c shows the FEA-influenced infill design for the geometries and load cases.
beam undergoing a three-point bend test with infill density Although the process has been automated where possible,
of 50%. there are a number of settings that are manually set by the
As this stage is performed for each layer, it enables the user. These are:
process to handle complex load cases. This is demonstrated
in Fig. 13, which shows the infill for layers at 5, 30 and 80 Layer height (mm) The increment in the z-axis.
mm for the bracket where it can be seen that the infill has Maximum mesh width (mm) The largest permissible mesh
been tailored to the relative stress distribution in each layer. square for the part.
It is then the case of defining the deposition path for Minimum mesh width (mm) The smallest permissible
the layer where consideration has been made to minimise mesh square for the part.
production time and ensure no overlapping of filament that Extrusion rate The length of material to be extruded per
would lead to distortions in the final printed component. length of travel.
First, the print paths for the linear infill are calculated. The Mesh offset (mm) The permissible gap between the perim-
process takes the vertical and horizontal paths and incre- eter and infill design.
ments back and forth to minimise travel of the printer head. Infill density The percentage of the interior volume that
The FEA-influenced infill design is then mapped onto the should be consumed by material.
linear infill structure. The elements within the FEA-influ-
enced infill design are then compared to the linear infill
design to remove any potential duplicate elements. The 4 Investigating the potential
optimum print path for the remaining elements is then cal- of FEA‑influenced infill designs
culated. This is achieved by applying graph theory to detect
whether the remaining elements form a Eulerian graph. If To investigate the potential affordances of influencing the
this is the case, the process solves the Chinese Postman infill design using FEA, a beam undergoing four different
Problem to generate a route that visits each element only loading scenarios has been tested. The test cases are as fol-
once [32–34]. If the network is non-Eulerian, the process lows with further details presented in Fig. 15.
identifies the elements with odd connectivity. Travel moves
are then added between these elements to ensure that they 1. Three-point bend test.
have an even degree of connectivity and thus, lead to an 2. Off-centre three-point bend test.
overall Eulerian graph. Figure 14 provides an example of 3. Four-point bend test.
the print lines generated for a layer of the three-point bend 4. Inverted four-point bend test.
test beam.
The process then continues to loop through Stages Two, The experimental procedure is consistent with the work by
Three and Four to form the print lines for the entire part. Lužanin et al. [26] who performed a three-point bend test to

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

156 Progress in Additive Manufacturing (2018) 3:145–163

Fig. 15  Illustration of test piece

and image of test case

(a) Beam geometry (b) Test machine

(c) Three-point bend test (d) Offset three-point bend test

(e) Four-point bend test (f) Inverted four-point bend test

evaluate the effect of layer thickness, deposition angle and testing equipment. For example, where the beam was no
infill percentage on the maximum flexural strength of FDM longer being supported by the rollers (Fig. 15b, i) but the
parts. For this investigation, beams of 100 mm length, 40 shoulders of the jig itself (Fig. 15b, ii). The specimens were
mm height and 10 mm in depth have been tested on a 25 kN tested to their stress–strain limits so that comparisons across
Instron machine (see, Fig. 15). The Instron machine was set the full extent of their structural behaviour can be made. The
to provide a constant linear displacement of 0.02 ­mms-1 in test was repeated five times for each case.
the direction of compression and the compression continued For each test case, the industry standard hexagonal pat-
until the specimen either destructed or interfered with the tern design formed the baseline against which the FEA-
influenced infill design was compared. To further control
the conditions of the tests, the level of extruded material
Table 1  Material extrusion details for test pieces remained constant and reflects a 32% infill density for all
Beam Value (mm) beams concerned. Table 1 provides further details on the
amount of deposited material with the maximum difference
Reference
between the reference specimen and a test specimen being
Honeycomb 4289.4
no more than 1.5% in favour of the reference specimen. The
Optimised
material used was a polylactic acid (PLA) as sold by Maker-
Three-point bend test 4251.2 (− 0.9%)
Bot Industries2 with a 1.75 mm diameter filament.
Off-centre three-point bend test 4224.6 (− 1.5%)
The slicing and generation of the infill has been based
Four-point bend test 4285.6 (+ 0.09%)
on the x-, y-plane as highlighted in Fig. 15a. Figure 16
Inverted four-point bend test 4234.3 (− 1.3%)
illustrates the FEA-influenced infill designs that have
Statistics
Mean 4257.0
Standard deviation 26.3 2
http://downloads.makerbot.com/filament/PLA+MSDS.pdf.

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 157

Fig. 16  Finite element results and associated test pieces

to carry the majority of the load and thus, there is greater

potential in increasing the strength-to-weight ratio of the
component if the non-load-carrying lattice elements are
removed.
The vertical line (1) highlights the cut-off to ensure a
32% infill density for all designs. It can be seen that the off-
centre three-point bend test has the potential to benefit most
in terms of strength-to-weight ratio as approximately 78% of
the cumulative weighting is carried by the lattice elements
that have remained in the FEA-influenced infill design. Simi-
Fig. 17  Beam stress along potential lattice elements larly, it can be asserted that the inverted four-point bend test
will see a reduced benefit as the weightings are more evenly
distributed across the test piece and the remaining lattice
been generated whilst Fig. 17 shows the Pareto chart for elements cover just 65% of the imposed stress. The remain-
the lattice elements in terms of their weightings that have ing three-point and four-point bend specimens are likely to
been calculated during Stage Four of the process. The lat- experience a similar change to the strength of the material as
tice elements have been ordered by their contribution. A 68% of total internal stresses are covered by the remaining
steeper curve indicates that fewer elements are required lattice elements.

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

158 Progress in Additive Manufacturing (2018) 3:145–163

Table 2  Max force, extension and stiffness table of results

Three-point Off-centre three-point
Honeycomb Optimised Honeycomb Optimised

Max force (N)

Mean 534.6 995.6 (1.9×) 459.1 1633.1 (3.5×)
Standard deviation 85.1 92.1 (1.0×) 63.7 225.0 (3.5×)
Max force extension (mm)
Mean 14.4 6.3 (0.4×) 13.6 13.7 (1.0×)
Standard deviation 5.1 0.6 (0.1×) 2.5 2.3 (0.9×)
Stiffness (N/mm2)
Mean 151.9 492.3 (3.2×) 144.3 298.0 (2.1×)
Standard deviation 9.7 32.1 (3.3×) 36.4 53.1 (1.5×)
Energy absorbed (N mm)
Mean 6,013,261.2 8,448,650.0 (1.4×) 5,005,817.1 15,287,407.5 (3.1×)
Standard deviation 398,381.9 2,658,821.2 (6.6×) 248,686.6 1,469,032.7 (5.9×)
Four-point Inverted four-point
Honeycomb Optimised Honeycomb Optimised

Max force (N)

Mean 721.1 972.7 (1.3×) 673.2 1206.2 (1.8×)
Standard deviation 154.2 131.8 (1.1×) 77.5 135.3 (1.7×)
Max force extension (mm)
Mean 14.4 6.3 (0.4×) 13.6 13.7 (1.0×)
Standard deviation 8.8 1.9 (0.2×) 2.6 2.4 (0.92×)
Stiffness (N/mm2)
Mean 187.4 345.1 (1.8×) 213.0 252.5 (1.2×)
Standard deviation 60.4 47.5 (0.8×) 56.5 76.6 (0.74×)
Energy absorbed (N mm)
Mean 7,054,588.9 5,055,034.9 (0.7×) 7,555,013.6 13,577,009.8 (1.8×)
Standard deviation 335,315.7 3,985,400.9 (11.9×) 1,039,383.9 1,348,942.1 (1.3×)

5 Results and discussion The main objective of the evaluation is to understand

the effect the infill design process has had on the structural
To evaluate the proposed process, the stress–strain curves behaviour of the component across its entire stress–strain
and the co-variance of the maximum load and displace- range and thus, the following sections present and discuss
ment at which it occurred have been plotted. In addition, the results from the test cases in relation to the:
the mean, variance and standard deviation for the maxi-
mum force and stiffness have been calculated. The stiffness 1. Maximum loading capacity;
has been calculated for the gradient between the 0–0.8 mm 2. Extension at which maximum loading occurs;
displacement for all specimens. 3. Beam stiffness;
The results from the four test scenarios have been 4. Energy absorption; and;
combined in Fig. 18 and Table 2. The graphs show the 5. Failure modes of the structures.
stress–strain and co-variance plots of the maximum load-
ing for the FEA-influenced and hexagonal infill designs. 5.1 Maximum loading capacity
The superimposed ellipses represent standard deviations
from the mean point of failure. An ideal case would be a Focusing on the maximum loading capacity of the beams
small elliptical shape, which would indicate that the speci- tested, it is immediately apparent from Fig. 18 that the FEA-
mens have a consistent and predictable failure mode. The influenced infill design can accommodate a much greater
colour of the traces corresponds to the specimen drawings loading than the hexagonal. In particular, the off-centre
within the legend. three-point bend test sees the greatest increase in loading
capacity with a three-fold increase whilst the four-point bend

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 159

Fig. 18  Stress–strain and maxi-

mum load distribution of test
specimens

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

160 Progress in Additive Manufacturing (2018) 3:145–163

test sees the least gain of 30%. This relative level of benefit repeated. This provides an indication that the initial response
is consistent with the relationship between the proportion behaviour of FDM beams is consistent within the elastic
of stress contained by the remaining paths in the FEA-influ- deformation region and it is the plastic region where devia-
enced infill designs (Fig. 17). tions in the stress–strain behaviour of parts with the same
Although the off-centre three-point beam shows the great- infill design occurs.
est increase in maximum loading capacity, it also shows the
greatest variance in loading capacity. This suggests that the 5.4 Energy absorption
more defined the stress profile, the more susceptible the infill
design is to variations in the testing of the beam. It is also the The final aspect of the results presented in Table 2 are the
case that the variance of the optimised beams is consistently energy absorption characteristics. All but one case dem-
greater than the variance of the hexagonal beams across the onstrated an increase in energy absorption for the opti-
four test cases. mised beam with the off-centre three-point beam showing
the greatest increase and four-point showing a decrease in
5.2 Maximum force extension energy absorption. This can be attributed to the premature
failure of the four-point beam whilst the honeycomb design
Considering the extension at which the maximum force was able to sustain a much greater extension. This further
occurs, it can be seen from Table 2 that the maximum force confirms that components with a well-defined load profile
occurs at either an equivalent or lower strain for the FEA- benefits most from an FEA-influenced infill design.
influenced infill when compared to the hexagonal. Although, Focusing on the standard deviation across the test cases,
it is interesting to note that the beam that experienced the the FEA-influenced infill designs reveal a greater deviation
least benefit in maximum force (four-point bend test) sees in their energy absorption capacity. Thus, even though there
the greatest reduction in the strain at which the force occurs. is a general increase in the energy absorption capacity of
In contrast, the off-centre three-point test that exhibited the the component for the given loading scenario, it is more dif-
greatest increase in maximum force saw comparably very ficult to determine and predict the absorption capacity for
little change in strain at which the maximum occurred. Thus, the optimised beam.
it appears there is a degree of negative correlation between
the maximum stress and strain values for the FEA-influenced 5.5 Failure modes of the structures
infill designs.
In contrast to the maximum loading condition, the vari- When considering the failure modes, there are some interest-
ance of the strain level at which the maximum loading ing buckling phenomena that are exhibited across both the
occurs is more consistent for the optimised beams than for FEA-influenced and hexagonal infill designs. In increasing
the hexagonal beams. This could be an important feature if the stiffness of the beam through the FEA-influenced design,
one were to use these designs for particular failure modes the brittleness of the beam also increases. With each element
and require a component to fail at a particular strain. This of the FEA-influenced structure contributing significantly to
observation holds true across all four test cases, where the the load-carrying capacity of the beam, it can be seen from
variance in strain appears consistent and provides further Fig. 18a that any buckling or element failure leads to a sub-
evidence of the potential utility for using the infill design for stantial decrease in the load-carrying performance. This is
designing components that require specific operating strains. particularly apparent in the three-point and off-centre three-
point bend tests (Fig. 18a, c).
5.3 Beam stiffness There are also consistent buckling events that occur in the
early stages of the loading of the FEA-influenced beams and
The average and standard deviations for the stiffness for the this is where the pathways connecting the infill design and
beams is also presented in Table 2. In all four test cases, perimeter buckle. The rollers of the jig then settle within the
the FEA-influenced infill designs consistently outperformed optimised structure where the load-carrying capacity starts
the hexagonal infill designs exhibiting up to three times the to increase once again (Fig. 19b). The FEA-influenced infill
stiffness with the maximum gain being achieved during the then completely fails after reaching the maximum loading
three-point bend test. This also highlights the significance condition (Fig, 19c). The consistency of the buckling points
of the loading scenario when investigating topology optimi- for the FEA-influenced beams could be used to support
sations as the greatest changes in maximal loading, strain planned and predictable failure modes for parts, which could
at which maximal occurs and stiffness have been measured also help in post-failure analysis.
during different loading scenarios. In contrast, the hexagonal beams begin to plateau at a
In contrast to the maximum force and strain metrics, the steady-state loading condition. The beams would then stead-
stiffness exhibited by the beams is relatively consistent when ily deform whilst maintaining this load before failing or

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 161

Fig. 19  Buckling and failure

modes of the beams designs

the extension at which the maximum loading occurred was

reduced for the FEA-influenced infill. This leads onto the
third finding where the FEA-influenced infill are stiffer than
the hexagonal design. The fourth finding is that the FEA
influenced infill designs generally saw a reduction in the
co-variance at which the maximum loading point occurred.
This is an indicator that the FEA-influenced infill is more
predictable in its loading behaviour.
The fifth finding is the predictability of the loading behav-
iour has also been seen in how the parts begin to buckle.
The FEA-influenced infill design saw a consistent perimeter
buckling event, which commonly occurred at the same level
of displacement. Embodying this predictability of buckling
Fig. 20  Comparison of key loading features from the four test cases could enable designers to use the buckling behaviour as pre-
emptive warning signs for particular load limits.

reaching a stage where the hexagonal structure began form-

ing layers of deformed material (Fig. 19d). In this case, the 6 Future work
loading would often increase as the beam starts to form an
almost solid structure. This paper has demonstrated the potential of using FEA
to influence the infill design process of FDM-printed parts
5.6 Summary where an increase loading capacity of four beams in different
loading conditions has been observed. It has also explored
From the four test cases, five key findings are evidenced with the generalisability of the proposed process through its abil-
Fig. 20 highlighting the key results from each of the tests. ity to handle complex geometries such as the bracket (c.f.
The first is that in all cases, the FEA-influenced infill design Fig. 13). In the creation of the proposed process, a number
outperforms the hexagonal structure in supporting higher of new research challenges have been exposed.
loads. Second, the FEA-influenced infill provided substantial First, the process has only been evaluated using two-
increases in strength in situations where the stress profile dimensional loading cases and there is a need to under-
is less distributed across the part to be printed. In addition, stand the potential of the strategy for more complex

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

162 Progress in Additive Manufacturing (2018) 3:145–163

loading scenarios. This includes three-dimensional load 7 Conclusion

cases, which involve varying levels of compression, ten-
sion, torsion and/or a mixture of modes. In addition, Additive manufacturing (AM) has and continues to expe-
cyclic loading and/or fatigue testing could be considered. rience considerable market and technological growth with
Being able to standardise these tests would also enable many forecasting a tripling in market value in the next dec-
comparison between a wider range of alternative infill ade. A key affordance of these technologies has been the
designs. increased design freedom in both the external form and
Second, for the tests presented, the volume of material internal structure. This paper has taken advantage of this
was controlled but it has been shown that the best per- affordance and uses the results from FEA to influence the
forming design occurred where the number of structural infill design of FDM parts.
paths covered the majority of the internal stress distri- This process has been discussed in depth alongside two
bution within the beam (i.e. the off-centre three-point cases that of a beam and bracket undergoing contrived load-
beam test). It would be interesting to see how an improved ing scenarios. A key feature of the proposed process is its
strength-to-weight ratio can be determined by the stress ability to generate FEA-influenced infills for complex parts
profile of the beam alone and to evaluate the optimisation with multiple perimeters and three-dimensional loading
across a range of infill densities. conditions.
Third, although the FEA-influenced design has been The proposed process has been evaluated experimentally
shown to improve the loading capacity of the beam, for a beam in four different loading cases and the designs
there is currently no computational method to predict the generated were compared to the industry standard honey-
improvement. In addition, the current FEA model is used comb infill. These tests confirmed a significant performance
solely to guide the design of the infill and does not reflect increase of the beam under loading, where the tests showed
the actual stresses the internal geometry is likely to expe- a:
rience. Thus, future research could post-process the print
file to predict the loading and buckling profile of the part • 3.5× increase in the loading capacity of the beam, par-
ahead of printing. This could also lead to an iterative pro- ticularly for highly localised stress profiles;
cedure where incremental changes to the lattice structure • 3× increase in the stiffness of the beam;
could be made and evaluated computationally to arrive at • reduction in the co-variance of the maximum loading
further improved internal structure. case; and,
Fourth, the optimisation has only considered a lattice • more predictable buckling mode.
design and there exists many other configurations that
could be used to generate the optimised infill design. In creating the proposed process, four research areas were
Such designs could potentially use the honeycomb as exposed and are summarised as:
the base pattern and/or adapt meshing techniques used
in FEA and CFD and structures developed in the natural 1. further evaluations involving three-dimensional, com-
world. In addition, the strategy does not take the mate- pressive, tensile, torsional and cyclic load cases leading
rial into account and one of the key areas of development to a standardised testing platform for print parts;
in FDM is the ability to print multiple materials. Future 2. further optimisation of the beam structuring through
processes could, therefore, seek to determine the optimum self-determination of the optimum infill percentage;
placement of combinations of materials to achieve the 3. post-processing of part geometry to print to predict load-
desired mechanical properties for a component. carrying capacity and loading/buckling profile; and,
Related to the aforementioned area is the challenge of 4. further analysis of potential infill designs including the
determining the attributes of the infill design that contrib- use of multiple materials within the part.
ute to the overall performance gains in the component.
Future work could investigate whether the performance
gains were due to: Acknowledgements The work reported in this paper has been under-
taken as part of the Language of Collaborative Manufacturing Project
at the University of Bath and University of Bristol, which is funded by
• higher material concentrations in the stressed regions the Engineering and Physical Sciences Research Council (EPSRC),
of the component; Grant reference EP/K014196/2. Underlying data are openly available
• the orientation of the lattice elements; and/or, from the University of Bath Research Data Archive at https://doi.
• the anisotropic behvaiour of the material. org/10.15125/BATH-00420 and a demonstration of the process can be
found at https://nbviewer.jupyter.org/github/JamesGopsill/PythonNote-
books/blob/master/FDM%20Force%20Path%20Methods/Generat-
ing%20infill%20designs%20based%20on%20forces%20through%20
a%20part.ipynb.

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Progress in Additive Manufacturing (2018) 3:145–163 163

Compliance with ethical standards 15. Gregorian A, Elliot B, Navarro R, Ochoa F, Singh H, Monge E,
Foyos J, Noorani R, Fritz B, Jayanthi S (2001) Accuracy improve-
ment in rapid prototyping machine (FDM-1650). In: Solid Free-
Conflict of interest On behalf of all authors, the corresponding author
form Fabrication Proceedings, pp 77–84
states that there is no conflict of interest.
16. Zhang Y, Chou K (2008) A parametric study of part distortions in
fused deposition modelling using three-dimensional finite element
Open Access This article is distributed under the terms of the Creative analysis. Proc Inst Mech Eng Part B J Eng Manuf 222(8):959
Commons Attribution 4.0 International License (http://creativecom- 17. Pandey P, Reddy N, Dhande S (2003) Real time adaptive slicing
mons.org/licenses/by/4.0/), which permits unrestricted use, distribu- for fused deposition modelling. Int J Mach Tools Manuf 43(1):61.
tion, and reproduction in any medium, provided you give appropriate https://doi.org/10.1016/S0890-6955(02)00164-5
credit to the original author(s) and the source, provide a link to the 18. Leary M, Merli L, Torti F, Mazur M, Brandt M (2014) Optimal
Creative Commons license, and indicate if changes were made. topology for additive manufacture: a method for enabling additive
manufacture of support-free optimal structures. Mater Des 63:678.
https://doi.org/10.1016/j.matdes.2014.06.015
19. Rezaie R, Badrossamay M, Ghaie A, Moosavi H (2013) Topol-
References ogy optimization for fused deposition modeling process. Procedia
CIRP 6:521. https://doi.org/10.1016/j.procir.2013.03.098
1. Gonigam D (2013) Printed to life: the tipping point. http://ago- 20. Ziemian C, Sharma M, Ziemian S (2012) Anisotropic mechanical
rafinancial.com/2013/11/12/printed-to-life-the-tipping-point/. properties of ABS parts fabricated by fused deposition modelling,
Accessed 01 Aug 2016 Mechanical Engineering. In: Gokcek M (ed) InTech. https://doi.
2. Palermo E (2013) Fused deposition modeling: most common 3d org/10.5772/34233
printing method. http://www.livescience.com/39810-fused-depo- 21. Es-Said OS, Foyos J, Noorani R, Mendelson M, Marloth R, Preg-
sition-modeling.html. Accessed 01 Aug 2016 ger BA (2000) Effect of layer orientation on mechanical proper-
3. Rozvany G (2009) A critical review of established methods ties of rapid prototyped samples. Mater Manuf Process 15(1):107.
of structural topology optimization. Struct Multidiscip Optim https://doi.org/10.1080/10426910008912976
37(3):217. https://doi.org/10.1007/s00158-007-0217-0 22. Ghorpade A, Karunakaran K, Tiwari M (2007) Selection of opti-
4. Sigmund O, Maute K (2013) Topology optimization approaches. mal part orientation in fused deposition modelling using swarm
Struct Multidiscip Optim 48(6):1031. https://doi.org/10.1007/ intelligence. Proc Inst Mech Eng Part B J Eng Manuf 221(7):1209
s00158-013-0978-6 23. Sood AK, Ohdar R, Mahapatra S (2010) Parametric appraisal of
5. Deaton JD, Grandhi RV (2014) A survey of structural and mul- mechanical property of fused deposition modelling processed parts.
tidisciplinary continuum topology optimization: post 2000. Mater Des 31(1):287. https://doi.org/10.1016/j.matdes.2009.06.016
Struct Multidiscip Optim 49(1):1. https://doi.org/10.1007/ 24. Jones RO, Iravani P, Bowyer A (2012) Rapid manufacturing of
s00158-013-0956-z functional engineering components. Other, Bath, UK. http://opus.
6. Gopsill JA, Hicks BJ (2016) Deriving infill design of fused depo- bath.ac.uk/30281/
sition modelled parts from predicted stress profiles In: Proceed- 25. Rayegani F, Onwubolu GC (2014) Fused deposition modelling
ings of the ASME 2016 International Design Engineering Tech- (FDM) process parameter prediction and optimization using group
nical Conferences & Computers and Information in Engineering method for data handling (GMDH) and differential evolution
Conference (Charlotte, North Carolina, USA, 2016) (DE). Int J Adv Manuf Technol 73(1–4):509
7. Lu L, Sharf A, Zhao H, Wei Y, Fan Q, Chen X, Savoye Y, Tu 26. Lužanin O, Movrin D, Plančak M (2014) Effect of layer thickness,
C, Cohen-Or D, Chen B (2014) Build-to-last: strength to weight deposition angle, and infill on maximum flexural force in FDM-
3D printed objects. ACM Trans Gr 33(4):97:1. https://doi. built specimens. J Technol Plast 39(1):49–58
org/10.1145/2601097.2601168 27. Onwubolu GC, Rayegani F (2014) Characterization and optimiza-
8. Tomlin M, Meyer J (2011) Topology optimization of an additive tion of mechanical properties of ABS parts manufactured by the
layer manufactured (ALM) aerospace part. In: Proceeding of the fused deposition modelling process. Int J Manuf Eng 2014. https://
7th Altair CAE technology conference, pp 1–9 doi.org/10.1155/2014/598531
9. Meisel N, Williams C (2015) An investigation of key design for 28. Lei N, Yao X, Moon SK, Bi G (2016) An additive manufacturing
additive manufacturing constraints in multimaterial three-dimen- process model for product family design. J Eng Des 27(11):751.
sional printing. J Mech Des 137(11):111406 https://doi.org/10.1080/09544828.2016.1228101
10. Ranscombe C, Hicks B, Mullineux G, Singh B (2012) Visually 29. Yang X, Xie Y, Steven G (2005) Evolutionary methods for topol-
decomposing vehicle images: exploring the influence of different aes- ogy optimisation of continuous structures with design dependent
thetic features on consumer perception of brand. Des Stud 33(4):319 loads. Comput Struct 83(12):956
11. Ahn SH, Montero M, Odell D, Roundy S, Wright PK 30. Eschenauer HA, Olhoff N (2001) Topology optimization of con-
(2002) Anisotropic material properties of fused deposi- tinuum structures: a review. Appl Mech Rev 54(4):331
tion modeling ABS. Rapid Prototyp J 8(4):248. https://doi. 31. Schroeder WJ, Lorensen B, Martin K (2004) The visualization
org/10.1108/13552540210441166 toolkit, 4th edn. Kitware
12. Sood AK, Ohdar R, Mahapatra S (2009) Improving dimensional 32. Algorithms and Theory of Computation Handbook, CRC Press
accuracy of fused deposition modelling processed part using grey LLC, 1999, “Chinese postman problem”, in Dictionary of Algo-
taguchi method. Mater Des 30(10):4243. https://doi.org/10.1016/j. rithms and Data Structures [online], Pieterse V and Black PE,
matdes.2009.04.030 (eds.) 2 Sept 2014. Available from: https://www.nist.gov/dads/
13. Lanzotti A, Del Giudice DM, Lepore A, Staiano G, Martorelli M HTML/chinesePostman.html. Accessed 10 June 2016
(2015) On the geometric accuracy of RepRap open-source three- 33. Kwan MK (1962) Graphic programming using odd or even points.
dimensional printer. J Mech Des 137(10):101703 Chin Math 1(273–277):110
14. Pennington RC, Hoekstra NL, Newcomer JL (2005) Significant 34. Hagberg AA, Schult DA, Swart PJ (2008) Exploring network
factors in the dimensional accuracy of fused deposition modelling. structure, dynamics, and function using NetworkX. In: Proceed-
Proc Inst Mech Eng Part E J Process Mech Eng 219(1):89. https:// ings of the 7th Python in Science Conference (SciPy2008), Pasa-
doi.org/10.1243/095440805X6964 dena, CA USA, pp 11–15

13

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

[email protected]