International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909

https://doi.org/10.1007/s12008-020-00684-7

ORIGINAL PAPER

Integrated CAD platform approach for Design for Additive

Manufacturing of high performance automotive components
Enrico Dalpadulo1 · Fabio Pini1 · Francesco Leali1

Received: 13 April 2020 / Accepted: 25 July 2020 / Published online: 3 August 2020
© Springer-Verlag France SAS, part of Springer Nature 2020

Abstract
Use of Additive Manufacturing provides great potentials to settings focused on high performance products. It allows feasibility
of sundry innovative features to completely rethink geometries and shapes and it leads to embrace new design approaches.
The enhanced design freedom can be exploited to optimize products, using techniques such as topology optimization. The
study of methods for development of optimized components to be produced by AM becomes therefore fundamental. A
framework for the methodological approach to operations to be carried out from the concept model to the printed component
has been analyzed and it is clear that issues and research efforts relapse both the global level of the workflow and the local
level of singular tasks to be performed. Problems related to management of Design for Additive Manufacturing workflow
can be solved with holistic approach, through the use of computer aided integrated tools. The aim of this work is to test the
effectiveness at local level of such tools with respect to operations for both design and industrialization optimization, working
on an automotive case study. In particular, specific tools for topology optimization, product simulation, printing preparation
and process simulation are taken as reference and results obtained with an integrated CAD platform are discussed.

Keywords Design for Additive Manufactruing · CAD based integrated platform · Powder Bed Fusion · High performance
automotive components

1 Introduction process constraints. Potentially, production can occur in less

quantity, more late (Just-In-Time), with less waste materials,
Additive Manufacturing (AM), as opposed to subtractive less energy, less goods transport and storage [4–6]. Advan-
manufacturing, is a technology to produce parts adding tages could interest also manufacturing management and
material layer upon layer. Its implementation offers great control (assembly) and industrial logistics (part count) [7,8].
potentials for the design of high performance components. The idea of re-thinking product systems and sub-systems
Implementation of AM technologies can provide many ben- archetypes, with functional or multifunctional based design,
efits, such as the possibility to produce components of every replacing parts or functional groups with pre-assembled or
form and function, minimizing design methods issues, invest- optimized and integrated ones is challenging [9].
ment costs, development time, enabling re-thinking product
development logics [1–3]. For example, small batches pro-
1.1 Design optimization
duction, customizing costs reduction, complex and perfor-
mance products become feasible removing manufacturing
Form freedom related to technology implementation let fully
re-thinking objects whose design underwent approach meth-
B Enrico Dalpadulo ods to freeze due to traditional constraint of manufacturing
[email protected]
and assembly [10]. The idea below the outstanding design
Fabio Pini freedom is that a bi-dimensional shape is easy to create
[email protected]
independently from the geometrical complexity of a tri-
Francesco Leali dimensional part. The possibility to produce combination of
[email protected]
organic elements, lattice elements, hollow elements, variable
1 “Enzo Ferrari” Department of Engineering, University of density infills, let new developments of functional design
Modena e Reggio Emilia, Modena, Italy [11]. All these techniques can be exploited and combined in

123
900 International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909

order to perform design optimization. Some advantages of 2 Design for Additive Manufacturing
these methods, together with innovative metallic alloys use, workflow
are the possibility to obtain extremely light structures, such
as free-form topology optimized parts. The design of parts specifically thought to be produced by
AM technologies, according to Design for Manufacturing
[10], is challenging and consequently, the development of
knowledge, skills, rules, tools, processes and methodolo-
1.2 Topology optimization gies to fully benefit of AM become fundamental, as shown
by Thompson et al. [2]. The DfAM fulcrum is both design
This kind of design-optimized parts usually has excellent and product optimization. As subset of Design for X (DfX)
performance from a structural point of view, in terms of [18], DfAM requires that the effects of design on manu-
weight/stiffness or weight/strength ratio. Topology optimiza- facturing and vice-versa have to be taken into account to
tion is a numerical technique to calculate optimal material optimize product performance and quality minimizing devel-
distributions, so that the material can be positioned only opment and production time and costs. Firstly, there are
where needed, to achieve established structural performance challenges in embracing new design approaches and in the
according to set targets and constraints. First contributes use of specific design rules and tools, preventing the users
of its theoretical definition have some tens of year and exploiting completely the potentials of the technology. Sec-
are due to Bendse and Kikuchi [12], but despite the tech- ondly, even though the use of additive technologies firmly
niques are known approaches [13] their application on design joins a CAD mathematical model and the actually printed
of real parts has been thwarted and part performance has component, the workflow from the concept to the definitive
been compromised due to both design tools, manufactur- job results in many sequential steps which may have complex
ing and technological constraints. Implementation of AM and slow relationships [19]. With respect to the development
makes these approaches more flexible since the construction of topology optimized metal parts to be printed by Powder
of a freeform organic part becomes feasible and thus, Topol- Bed Fusion (PBF) processes, the DfAM general workflow is
ogy optimization is having a great trend. In order to create depicted in Fig. 1. The basis of DfAM workflow is an itera-
optimized parts, sunder numerical methods can be applied tive refinement of the starting model of the part with respect
to compute ideal geometries [14] and some of them are to Design and Industrialization constraints to respectively
currently implemented in specific Computer Aided Engineer- improve the part shape and the associated printing process,
ing (CAE) software. Design of topology optimized parts is as the Authors firstly tried to define it in [20]. The input is
increasing especially in medical engineering, aerospace and the Product Data, which is made of the information (mod-
automotive settings, different design approaches and meth- els, analyses, objectives, constraints) that are required for the
ods are being explored in order to fully exploit the aforesaid design. The output is the Production, or rather the 3D printing
potentials induced by AM technologies [15,16]. The idea is to process and the other related operations (e.g. post processing,
re-think both products and their development methods, thus, CNC machining and testing). The whole approach relies on
the Design for Additive Manufacturing (DfAM) is receiving specific tasks listed as follow.
an increasingly attention [17].

Fig. 1 Design for Additive Manufacturing workflow

123
International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909 901

• Product optimization—Design CAM software to trig the Printing Preparation. Generally,

the part requires some modification and so some iterations
– Model Preparation
between CAD and Printing Preparation tasks occur. The out-
– Topology Optimization
put of this step is the industrialized job to be printed. In order
– Design Interpretation
to optimize the industrialization, a file containing a tessel-
– Product Simulation
lated representation of the job is created and this is imported
• Process optimization—Industrialization in a Finite Element -based CAE software so that Process Sim-
ulation can be performed (e.g. PBF AM thermomechanical
– Coarse Modelling
simulation). The results of the analysis are used to modify
– Printing Preparation
either the geometry of the parts by CAD or the Printing
– Process Simulation
Preparation tasks, so some iterations are required in order
to meet functional and technological targets. Again, if some
The Design and Industrialization phases, which constitute Industrialization constraints change, all the Industrialization
the two main elements, respectively refer to the Product Opti- workflow has to be repeated. After Process Optimization the
mization and the Process Optimization. With respect to the job is ready, the output file for the machine can be gener-
Design phase, Model Preparation is the first step that returns ated (generally through a build-processor software) and the
the boundary conditions for subsequent optimization. Def- Production can start. Except for the steps that involve CAD
inition of design space, or rather the model containing the tools (Model Preparation, Design Interpretation and Coarse
permissible and forbidden volumes for shape computation, Modelling), active steps are also referred as the local levels
is required. In order to do that, a mathematical representation of the approach. Conversely, global level activities are those
(generally the original non-optimized part) can be imported related to data management and exchange, such generation
as support, otherwise it can be directly modelled through a of STEP or IGES files for neutral representations, or STL file
CAD software. An assembly analysis is required to define the for tessellated representations.
maximum volume for the design space, according with the
operative and assembly constraints. Generally, a non-design 2.1 Integrated CAD platform approach
space is also required, so that volumes that have to be kept for
functional regions are defined. A neutral representation file As described by Huang [18], proper design tools need to
of these geometries is the input for Topology optimization. be developed for an effective implementation of DFX meth-
A Finite Element -based CAE software computes the opti- ods. In most cases, sundry specific pieces of software are
mized shape of the part. The result is usually a cloud of points used to perform the different tasks that form the described
that can be manipulated for triangularization or exported as DfAM workflow. This, can lead to issues in the workflow at
tessellated representation. In most cases, this file is used as global level, related to data management and exchange. Lav-
support for Design Interpretation, a geometry reconstruction erne et al. [21] suggest to integrate DfAM tools into PLM
to be performed in a CAD environment, which is generally to benefit enrichments of knowledge and know-how related
a highly time-consuming operation. A neutral representation to data management. A previous work by the Authors [20],
file is generated from the model of the optimized part and is shows how holistic design approach implemented by CAD
then imported in a Finite Element Analysis CAE software in based platforms, can speed up the entire workflow. In partic-
order perform the last step of Design phase. Product Simu- ular, it is possible to overcome issues at global level keeping
lation evaluates the required performance on the part. The parametricity and associativity between different CAD based
results of the analysis (e.g. structural simulation or addi- environments. Conversely, a local level evaluation is required
tional ones) are used to improve part properties working to test the effectiveness of the singular step-oriented tools as
on its geometry. At this point, some iterations in the work- well.
flow between Model Preparation and Product Simulation or
even Topology optimization, Model Preparation and Product 2.2 Local level tasks
Simulation are necessary so that part optimization can meet
design requirements. If some design constraints change, all Local level steps for Design and Industrialization phases
the Design workflow has to be repeated. After Product Opti- respectively are Topology Optimization and Product Sim-
mization the final design is ready, technical product data (e.g. ulation, and Printing Preparation and Process Simulation.
drawings and specifications) can be produced and the Indus- In both the cases, re-design loops are required: to optimize
trialization phase can start. Coarse Modelling is the first step the product with respect to the product objectives and con-
of Industrialization that creates the raw part models through straints along the Design phase; for process optimization
a CAD software. Usually, a tessellated representation of the acting on both part design and the job preparation over the
geometry is then generated and imported in a 3D printing Industrialization phase. The specific tasks to be performed

123
902 International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909

in the four local levels, are detailed below. In order to com- assessment approach for local level steps. Chapter 4 provides
pute a topology optimization, firstly it is necessary to import the results on the case study achieved with respect to each
a geometrical model and to define the design space and of the specific tasks, that are lastly highlighted in the final
non-design space areas. It is then required to setup a finite discussion in the Chapter 5.
element model, composed by a mesh of the part and the sets
of restraints and loads to replicate part physical behavior.
The targets and constraints for the optimization need to be
specified, a geometrical result must be obtained from com- 3 Case study, tools and assessment approach
putation and hopefully a validation of this should be carried
out. Product Simulation is especially performed in that sense, In order to implement the described method on the integrated
a generic geometrical model has to be imported and the finite tool and validate it at local level, a case study is represented by
element model must be set in the same way, particular simula- a Formula SAE wheel knuckle. This component is a key part
tion modes should be possible as well as material definition for vehicle dynamics performance in racing cars, since non
could be deepened. A key point is the result analysis with suspended masses have to be minimized in order to guarantee
post-processing features. Printing Preparation requires the cornering performance, due to the better tarmac-tyre contact.
machine definition and the possibility to manipulate input Dumbre et al. [22] analyzed wheel knuckle behavior in criti-
geometries (such as STL). Tasks such as part positioning, cal working conditions and defined a modeling approach for
part orientation and support generation are necessary, as simulations. Actually, sundry examples of design optimiza-
well as slicing computation is mandatory, moreover, tools tion of this component have been explored. Generally, this
for slicing analysis are useful. Consequently, it is necessary is done embracing traditional technologies and, for exam-
to export the data for build processors or in printing format ple, Bhardwaj et al. worked on the design optimization for
if this is integrated in the software. In order to perform the a part to be produced by CNC milling [23], while Harzheim
Process Simulation, an industrialized job has to be imported et al. achieved the optimization for a part to be produced by
and material temperature-dependent data and process-related casting [24]. Conversely, few examples of DfAM including
machine parameters must be specified. A mesh of the job and topology optimization of a wheel knuckle to be produced by
its connections have to be created and computation settings PBF AM process exist, such as the Electron Beam Melting
which reflects the process have to be defined. Finally, some manufactured one from Walton et al. [25].
post-processing features are required to analyze the result and The case study part, originally made by traditional sub-
tools to manipulate it should be significant. The described tractive technology, had been re-designed requiring the use
tasks require express implementation in tools and since they of many specific software for the different tasks, and then
affect the workflow at local level, Topology Optimization, printed by Selective Laser Melting (SLM) PBF AM process
Product Simulation, Printing Preparation and Process Simu- with benefit of weight reduction. According to the work-
lation have to be tested. flow above described (see Fig. 1), the steps and the related
Therefore, the aim of this work is to implement the general results for benchmarked part are depicted in Fig. 2. The
method, understanding the behavior of the integrated tool at tools involved for the benchmarked part operations are listed
local level. The next chapter presents the integrated CAD as follow. Dassault Systmes Solidworks had been used for
based platform that has been tested for Design and Industri- CAD modelling and re-design operations, Altair Optistruct
alization phases with respect to a selected case study and the for topology optimization, MSC Marc for Product Simula-
tion, Materialise Magics for Printing Preparation, PTC Creo

Fig. 2 DfAM of the benchmarked part

123
International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909 903

for CAD industrialization re-design operations, MSC Simu- Table 1 Loadcases definition
fact Additive for Process Simulation. Conversely, the CAD Loadcase Element Value [unit]
based platform selected for the local level evaluation is the
Dassault Systmes 3DExperience. The CAD based platform Bump Acceleration 5.6 G [m/s2 ]
fundamentally integrates the design environment of Catia Vertical force 4400 [N]
suite, with CAX engineering tools of Simulia and Delmia Cornering Acceleration 2.5 G [m/s2 ]
and PLM management of Enovia. Drift force 4400 [N]
Vertical force 2810 [N]
3.1 Local level evaluation approach Drift moment 961 [Nm]
Reaction moment 809 [Nm]
For each of the core tasks of Design and Industrialization Braking Acceleration 3 G [m/s2 ]
phases, output of the standalone tool at the state of art has Longitudinal force 2700 [N]
been identified as reference. This becomes the target for Vertical force 1805 [N]
the operations performed in the CAD platform through the Braking moment 885 [Nm]
respectively use of specific apps, summarized in Fig. 3). With Transport moment 616 [Nm]
respect to the first step, the output of Altair Optistruct is taken
as reference. The 3DExperience application for this purpose
is Catia Functional Generative Design, which integrates in
a CAD environment Tosca solver to perform topology opti- performance. In particular, the original CNC-machined part
mization and compute the shapes computed. The aim is to had some stress concentrations over 300 MPa.
replicate the set up of three different loadcases and discover During part utilization and stress this behavior could lead
if the tools, starting from the same input data, will provide to crack trigger on the material and eventually part fail-
similar results despite of different features and computational ure. During the re-design operations, exploiting the branched
algorithms. In details, bump is the first loadcase; it simulates organic shape, particular attention had been put to the limita-
the maximum load related to a vertical acceleration. The cor- tion of part stress at 200 MPa and the product simulation
nering loadcase is to simulate the load of a corner at high upheld that part behavior. The output results are the Von
speed with the maximum lateral acceleration (tyre limit con- Mises stress distribution maps. The 3DExperience applica-
dition). Braking is the loadcase that simulates the load due tion for structural validation is Simulias Linear Structural
to the maximum negative longitudinal acceleration. Table 1 Validation, which involves Abaqus computation code. The
summarizes the loadcases definition. With respect to the sec- output of Materialise Magics is taken as reference for the
ond step, the output of MSC Mark is taken as reference. The third step. The job had been created introducing the part in the
wheel knuckle prototype had been validated through finite machine context, computing a precise orientation, designing
element (static and buckling) analysis simulations, in order special support structures and performing slicing analysis.
to check, despite the mass saving, the adequate structural The benchmarked job preparation had required part orienta-

Fig. 3 Standalone tools and platform apps involved in Design and Industrialization core tasks

123
904 International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909

tion to be optimized in order to minimize the need of supports, – RBE2 on the holes #4 with X and Z displacement on the
so that much less powder will be required (material, time and control node at the center of the joint.
costs reduced) and the post processing step will be simpli- – RBE2 on the holes #5 with Z displacement on the control
fied. For support definition, both traditional vectorial support node coincident with the pivot axis.
from Magics library and CAD-designed solid supports had
been introduced. Solid supports provide material saving and In the Functional Generative Design application, the fol-
let reduction of support anchoring on part surfaces. A total lowing setup has been created:
amount of 205 supports between block and cone type had
been used both vertical ones and angled ones, with the aim – Spherical joint on the axis of the holes #1 positioned on
of reducing part surface support anchor. The aim is to repli- the center of the joint, with X, Y and Z displacement.
cate the operations that brought the industrialized additive – Forbidden displacement on holes #4 with control node
manufacturing job and verify if the Delmias Powder Bed positioned between the two axis and X and Z displace-
Fabrication application on 3DExperience, despite of differ- ment.
ent interfaces and features, could lead to a similar result. – Forbidden displacement on holes #5 with control node
Finally, the last step refers to the output of Simufact. Process positioned between the two axis and Z displacement.
simulation had been performed starting from the industrial-
ized job made in Magics, exported as STL. Voxel mesh and First and third cases return identical solutions, whereas
connections had been created, thermo-mechanical AlSi10Mg the second case provides a comparable solution, except for
material properties had been set, process parameters and tem- the loss of a small lever arm. The benchmarked loads setup
peratures had been created, thermal and structural analyses was made of a RBE3 element connected with the surfaces
had been performed considering the phases from building to of #3, with control node positioned at the wheel center.
cooling. The simulation tool had been calibrated with pro- On this element all the loads (force and torque) related
cess parameters of the SLM 280 HL machine used for part to the three different loadcases had been applied. In the
production and via experimental optimization. The aim is to Functional Generative Design application, feature for remote
run a preliminary process simulation, completing the opera- load and remote torque have been used, connected with the
tion to achieve the same type of result, which is in particular a surfaces of #3 applied on a point positioned at the wheel
displacement map and a deformed model. The 3DExperience center, as depicted in Fig. 4b. The computation targets are
application is Simulias Additive Manufacturing Scenario. In the results from the bump, cornering and braking loadcases
this case the expected output should not be exactly compa- above described. With respect to the last two shape com-
rable, since simulation will require then special additional putations, the introduction of the brake caliper is required,
tuning. since it contribute to the knuckle stiffness. A simulacrum
is created and fit in assembly with the wheel knuckle, the
mesh of both parts are created and they are connected intro-
ducing infinitely-stiff elements on the holes #2. In addition,
4 Results
for the braking loadcase, forces acting on the braking pads
are applied on the part. Topology optimization is set with a
4.1 Topology optimization
stiffness maximization target (minimization of element com-
pliance) and a target percentage mass value as constraint.
A tetrahedron linear mesh with cell dimension of 2 mm has
Figure 5 depicts respectively the results from Optistruct and
been created in order to balance an accurate solution and
3DExperience, with respect to the Bump (Fig. 5a), Cornering
computational cost. The structural model has been developed
(Fig. 5b) and Braking (Fig. 5c) loadcases.
in order to replicate the physical behavior of the suspension
mechanism. Figure 4a depicts the reference model of the
4.2 Product simulation
component with the main geometrical features used to setup
the optimization. Specifically, the holes for fastening the
The operations in 3DExperience can be performed on Func-
brackets for uniballs of upper and lower wishbones, respec-
tional Generative Design itself (if platform workflow is
tively #1 and #4; holes for fastening the bracket that connects
preserved) or anyway in Linear Structural Validation applica-
the steering rod, #5; housing of the wheel bearings, #3; holes
tion. For the finite element analysis, a truss rod link between
for the screws that bolt the brake caliper, #2.
the holes for the fastening of the brake caliper has been intro-
The benchmarked restraints setup was made of:
duced in order to replace the part taking into account its
stiffness (Fig. 4b). The three loadcases (Bump, Cornering
– RBE2 on the holes #1 with X Y and Z displacement on and Braking) have been simulated using the Linear Struc-
the control node at the center of the joint. tural Validation application, keeping the setting of the finite

123
International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909 905

Fig. 4 Component reference model with main geometrical features (a) and finite element model (b)

Fig. 5 Topology optimization results

element model used for the topology optimization (Fig. 4b), have set the self-supporting angle, selecting the setup of the
including loads and constraints, except for the mesh size. optimization to minimize supports. For support generation,
Figure 6 shows the stress distribution analysis obtained vectorial supports from libraries are analogue (see Fig. 7c),
respectively with Marc (Fig. 6a) and 3DExperience (Fig. 6b) but only Wired and Cones can be angled. The design of solid
for the Bump loadcase and the finite element model setup on supports is facilitated by the platform environment integra-
3DExperience to perform product simulation. tion, but the actual use of them is compromised since the
interposing of elements between those and the part is possi-
4.3 Printing preparation ble only for Cone and Tree types. Finally, about 157 supports
and a mixed of cones, wired and volume, both vertical and
On 3DExperience Powder Bed Fabrication application, angled have been introduced. Since solid supports are able
firstly setup of the virtual machine environment is required, to support Cones but not Wired, they should be replaced by
defining several building parameters related to the manu- Wired support type. Additional slicing analysis is required
facturing process (build volume, build tray, scanning type, in order to check effectiveness of supporting structures and
slicing step, recoating direction, recoating speed etc.). Part their refinement. Moreover, the slicing and scan path can
orientation can be computed in the same way as Magics, once be defined, with information about the recoater for powder

123
906 International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909

Fig. 6 Product Simulation results

Fig. 7 Printing Preparation results

Fig. 8 Process Simulation results

123
International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909 907

Table 2 Platform local level

Local level Steps Lacks Advantages
evaluation
Topology Optimization Preprocessing features to define Solid models as output of shape
elements of the finite element model computation
Re-design facilitated
Product Simulation Preprocessing features to define Model settings can be retrieved
elements of the finite element model from Topology Optimization
Printing Preparation Solid supports not completely Re-design extremely facilitated
effective, not every type can be
angled
Process Simulation Not found Volume and surface meshing
features
Industrialization optimization
facilitated

deposition, the inert gas flow and scanning strategies, pat- alignment with the original CAD model, showing respec-
terns and sequences. The generated path can be analyzed tively Simufact and 3DExperience results.
so that low melting and roughness points of each slice can
be displayed in order to optimize the printing preparation.
Figure 7 depicts the industrialized job obtained with Magics
(Fig. 7a) and 3DExperience (Fig. 7b). 5 Discussion

An evaluation on the basis of obtained results for the each of

4.4 Process simulation the local level steps is now presented. The aim is to under-
stand if the several outputs from the local level CAD platform
A finite element model is created for part, supports and build applications can be comparable to the reference ones.
tray. The voxel approach is the most suitable processing an Topology Optimization Models obtained as output of the
STL representation, like the benchmarked one. Specific type computation made in the 3DExperience show features similar
of volume and surface mesh are instead the most appro- to those made in Optistruct referring to the Bump, Cornering
priate working on a model developed inside the platform. and Braking loadcases. For each of them some different ele-
Connections between the elements are created, and physical ments unavoidably exist, due to the preprocessing interfaces
constraints are applied. AlSi10Mg alloy have to be deeply (e.g. RBE elements definition) and the implemented solvers
characterized introducing temperature-dependent data for algorithms. Therefore, despite the need of few differences for
elasticity, expansion, conductivity, specific heat, latent heat, model setup, the use of the tool can guarantee likely com-
and a plastic law (e.g. Johnson-Cook). A thermal analysis parable results. It has to be remarked that the output from
and a structural one have to be set. Process data like the 3DExperience can be a solid model instead of a cloud of
slicing and scanning path are retrieved from Powder Bed points and this is easier manipulable for part re-design and
Fabrication application. Process temperatures (both for ther- optimization.
mal and structural case) of parts, supports and build tray, Product Simulation Results related to the model static and
material deposition information and cooling data (e.g. con- buckling structural validation performed with 3DExperience
vection and radiation) have to be specified. Lastly, thermal and Marc are identical. In the same way as for the topol-
and structural analysis require proper computation step set- ogy optimization, differences in setup interfaces exist. The
tings and they finally can be launched. The displacement biggest advantage of the integrated tool could be the possi-
output (see Fig. 8b) can be exported creating a vector field bility to keep all the settings used to create the finite element
displacement file, that can be used to obtain a deformed shape model in the topology optimization step.
starting from the original model using the Virtual To Real Printing Preparation Component industrialization performed
Shape Morphing application. This can be analyzed through with 3DExperience, considering the features differences in
the Digitized Shape Preparation application to evaluate geo- the software, could be of course acceptable. Main drawback
metrical displacement on the part after the printing process or, is the CAD designed solid supports usage which is compro-
using a best-fit alignment, to measure effective part warping mised. Moreover, unlike Magics, the volume support type
compared to the ideal geometry. Figure 8a depicts the com- cannot be angled. Conversely, a benefit is in the re-designs,
parison between the deviation analyses performed on a 3D with the possibility to update model shape and printing prepa-
Point Clouds software (GOM inspect), based on a best-fit ration keeping information.

123
908 International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909

Process Simulation It should be remarked that further inves- erations, and constraints. CIRP Ann. Manuf. Technol. 65, 737–760
tigations will be carried out in order to improve the process (2016). https://doi.org/10.1016/j.cirp.2016.05.004
3. Ford, S., Despeisse, M.: Additive manufacturing and sustainability:
simulation with the 3DExperience platform. Standard data an exploratory study of the advantages and challenges. J. Clean.
and general values for parameters will have to be converted Prod. 137, 1573–1587 (2016). https://doi.org/10.1016/j.jclepro.
into specific ones and experimental phases will be required. 2016.04.150
Regardless the accuracy of the result, the types of output 4. Holmström, J., Partanen, J., Tuomi, J., Walter, M.: Rapid manu-
facturing in the spare parts supply chain: alternative approaches
are of course consistent and comparable. A plus is model to capacity deployment. J. Manuf. Technol. Manag. 21, 687–697
meshing type features, a great advantage is the possibility to (2010). https://doi.org/10.1108/17410381011063996
modify both part shape and printing preparation settings and 5. Chen, D., Heyer, S., Ibbotson, S., Salonitis, K., Steingrmsson, J.,
effectively optimize the printing process. Thiede, S.: Direct digital manufacturing: definition, evolution, and
sustainability implications. J. Clean. Prod. 107, 615–625 (2015).
Potential lacks or advantages are highlighted and summa-
https://doi.org/10.1016/j.jclepro.2015.05.009
rized in Table 2. 6. Faludi, J., Bayley, C., Bhogal, S., Iribarne, M.: Comparing environ-
mental impacts of additive manufacturing vs traditional machining
via life-cycle assessment. Rapid Prototyp. J. 21, 14–33 (2015).
https://doi.org/10.1108/RPJ-07-2013-0067
6 Conclusions 7. Thomas, D.: Costs, benefits, and adoption of additive manufactur-
ing: a supply chain perspective. Int. J. Adv. Manuf. Technol. 85,
A Design for Additive Manufacturing workflow to produce 18571876 (2016). https://doi.org/10.1007/s00170-015-7973-6
8. Gebler, M., Schoot-Uiterkamp, A., Visser, C.: A global sustain-
topology optimized parts has been described, concerning ability perspective on 3d printing technologies. Energy Policy 74,
both design and industrialization optimization. It has been 158–167 (2014). https://doi.org/10.1016/j.enpol.2014.08.033
remarked that the use of integrated CAD platforms can 9. Gibson, I., Rosen, D., Stucker, B.: Additive Manufacturing Tech-
improve the method effectiveness speeding up the workflow nologies: 3D Printing, Rapid Prototyping, and Direct Digital
Manufacturing. Springer, New York (2015). https://doi.org/10.
at global level, but a local level evaluation on specific tasks 1007/978-1-4939-2113-3
was required as well. Results show the possibility to replay 10. Boothroyd, G., Dewhurst, P., Knight, W.: Product Design For Man-
the core steps that compose the workflow and can be per- ufacture And Assembly. CRC Press, Boca Raton (2011)
formed at the state of art, eventually either avoiding few 11. Milewski, J.: Additive Manufacturing of Metals: From Fundamen-
tal Technology to Rocket Nozzles, Medical Implants, and Custom
specific features or revising the adopted solution. This is due Jewelry. Springer, New York (2017). https://doi.org/10.1007/978-
to the differences between the implemented tools. Results 3-319-58205-4
are in some case exactly analogue, otherwise they can be 12. Bendsøe, M., Kikuchi, N.: Generating optimal topologies in struc-
believed anyway acceptable. The advantage offered by the tural design using a homogenization method. Comput. Methods
Appl. Mech. Eng. 71, 197–224 (1988). https://doi.org/10.1016/
integrated tools is to connect the environments, not only for 0045-7825(88)90086-2
file interchange and data manipulation, but also linking the 13. Bendsøe, M., Sigmund, O.: Topology Optimization: Theory, Meth-
features between the applications, keeping the information ods and Applications. Springer, Berlin (2004)
with associativity. This is valid not only for re-design opera- 14. Rozvany, G.I.N.: A critical review of established methods of
structural topology optimization. Struct. Multidiscip. Optim. 37,
tions initially investigated, but also for simulations, keeping 217–237 (2009). https://doi.org/10.1007/s00158-007-0217-0
settings and recovering and exploiting their results. To con- 15. Plocher, J., Panesar, A.: Review on design and structural opti-
clude, there is no advantage in replacing a stand-alone tool misation in additive manufacturing: towards next-generation
with an application of an integrated platform, rather imple- lightweight structures. Mater. Des. 183, 108164 (2019). https://
doi.org/10.1016/j.matdes.2019.108164
menting the whole method and the workflow into a platform. 16. Dalpadulo, E., Pini, F., Leali, F.: Design for additive manufacturing
Alternatively, adopting the platform for just one of design or of a topology optimized brake caliper through cad-platform-based
industrialization optimization could be a benefit. For further systematic approach. In: The International Joint Conference on
developments, it could be interesting to structure an objective Mechanics, Design Engineering and Advanced Manufacturing
(JCM 2020) (2020)
evaluation system to validate the tools. 17. Kumke, M., Watschke, H., Vietor, T.: A new methodological frame-
work for design for additive manufacturing. Virtual Phys. Prototyp.
11, 3–19 (2016). https://doi.org/10.1080/17452759.2016.1139377
18. Huang, G.: Design for x: Concurrent Engineering Imperatives.
References Chapman & Hall, London (1996)
19. Wiberg, A., Persson, J., Ölvander, J.: Design for additive manufac-
1. Gao, W., Zhang, Y., Ramanujan, D., Ramani, K., Chen, Y., turing a review of available design methods and software. Rapid
Williams, C., Wang, C., Shin, Y., Zhang, S., Zavattieri, P.: The Prototyp. J. 25, 1080–1094 (2019). https://doi.org/10.1108/RPJ-
status, challenges, and future of additive manufacturing in engi- 10-2018-0262
neering. CAD Comput. Aided Des. 66, 65–89 (2015). https://doi. 20. Dalpadulo, E., Pini, F., Leali, F.: Assessment of design for additive
org/10.1016/j.cad.2015.04.001 manufacturing based on cad platforms. In: Rizzi, C., Andrisano,
2. Thompson, M., Moroni, G., Vaneker, T., Fadel, G., Campbell, R., A., Leali, F., Gherardini, F., Pini, F., Vergnano, A. (eds.) Design
Gibson, I., Bernard, A., Schulz, J., Graf, P., Ahuja, B., Martina, F.: Tools and Methods in Industrial Engineering. ADM 2019. Lecture
Design for additive manufacturing: trends, opportunities, consid- Notes in Mechanical Engineering. Springer, Cham (2019)

123
International Journal on Interactive Design and Manufacturing (IJIDeM) (2020) 14:899–909 909

21. Laverne, F., Segonds, F., Anwer, N., Le Coq, M.: DFAM in the https://doi.org/10.1007/s00158-005-0554-9
design process: a proposal of classification tofoster early design 25. Walton, D., Moztarzadeh, H.: Design and development of an addi-
stages. In: CON-FERE, Sibenik, Croatia (2014) tive manufactured component by topology optimization. Procedia
22. Dumbre, P., Mishra, A., Aher, V.: Structural analysis of steering CIRP 60, 205–210 (2017). https://doi.org/10.1016/j.procir.2017.
knuckle for weight reduction. Int. J. Emerg. Technol. Adv. Eng. 4, 03.027
557 (2014)
23. Bhardwaj, S., Ashok, B., Lath, U., Agarwal, A.: Design and opti-
mization of steering upright to reduce the weight using FEA. SAE
Publisher’s Note Springer Nature remains neutral with regard to juris-
Technical Papers 2018, (2018). https://doi.org/10.4271/2018-28-
dictional claims in published maps and institutional affiliations.
0081
24. Harzheim, L., Graf, G.: A review of optimization of cast parts using
topology optimization-II topology optimization with manufactur-
ing constraints. Struct. Multidiscip. Optim. 31, 388399 (2006).

123