FEASIBILITY STUDY
TO 3D PRINT A FULL SCALE CURTAIN WALL FRAME
AS A SINGLE ELEMENT
(APPLICATION OF 3D PRINTING/ ADDITIVE MANUFACTURING
TECHNOLOGIES IN THE AEC INDUSTRY)
by
Natasa Mrazovic, M.Arch.Eng., lic.arch., M.CE.
Research Partner: Permasteelisa R&D Group / Dr. Danijel Mocibob, M.CE.
Acknowledgment
This report could not be finished without the financial and intellectual support of
Permasteelisa’s R&D group. My sincere thanks to Permasteelisa’s Technical Director
Bert van de Linde that initiated this research and to Henk de Bleecker, Permasteelisa’s
R&D Group Director for their trust in my abilities to conduct this research.
Special thanks to dr. Danijel Mocibob who helped in collecting the data and processing
the results.
2
Prologue
This report is the result of collaboration between CIFE and the Permasteelisa group, a
leading firm in design and construction of building envelopes, to validate the potential of
Additive Manufacturing (AM) technologies, popularly known as 3D printing, for the
application in design and construction of curtain wall (CW) components. In general, the
AEC industry is becoming aware of the benefits that AM technologies offer, like massproduction of custom parts, just in time delivery, and production of components with
complex geometries. Permasteelisa decided to invest in a feasibility study to 3D-print
multifunctional CW segments and large-scale CW components, specifically a one-story
metallic frame, as a single element on a construction site to eliminate the cost of
extrusion, assembly, transportation, and packaging.
Research methodology designed for the feasibility study to 3D-print large-scale CW
elements brought breakthrough insights and quantified some ideas: (1) identify state of
the art in AM related to the problem and all relevant AM manufacturers to contact; (2)
gather and prepare information about conventionally manufactured Permasteelisa CW
frame; (3) gather information about the cost, time, and energy consumption from the
relevant manufacturers using a questionnaire, semi-structured interviews with AM
representatives, joint file preparation with AM designers and interviews with CEOs; (4)
calculate specific costs by using data from previous scientific research and data collected
from manufacturers; (5) comparatively analyze the specific costs to design and
manufacture CW frame using traditional manufacturing/construction methods, standard
AM metallic systems and existing hybrid large scale metal deposition systems; and (6)
recommend future research and development.
The key findings about 3D printing of highly optimized multifunctional CW metallic
segments in sizes that could fit commercially available machines are:
 Research has low risk as an investment for Permasteelisa’s R&D group because
necessary knowledge and an established research process already exist in specialized
institutes.
 Specific cost is not competitive for manufacturing traditional geometries currently
produced with conventional methods, but it is highly competitive for multifunctional
application and further research in adding new functions to specific components.
 Commercial machines with appropriately sized build chambers are available on the
market, e.g. EOS, Arcam.
 AM processes exist in arrays of different materials; additional investments are needed
in material research if unconventional materials are designed for application.
 A new optimization (simulation and FEA analysis) and customization process, as part
of AM, brings revolution in design of not only curtain walls but buildings in general.
The key findings about 3D printing a full-scale metallic CW frame as a single element
are:
3
 Technologies with small-scale platforms available on the market, from e.g. EOS or
Arcam, are unsuitable for the task because of their build chamber sizes and
disproportionally greater manufacturing costs in comparison with conventional
methods.
 A ready-to-use AM process for the fabrication of a specified full-scale CW frame as a
single element does not exist on the market (for the required size and in appropriate
material with mechanical properties that are mandatory for the building envelope)
resulting in higher investment in this type of research with a higher risk of not
producing results.
 Hybrid metal deposition technologies, from Sciaky Inc., DM3D, Optomec and
Fabrisonic, are suitable for further investigation.
 The estimated specific cost to 3D-print a predefined CW frame is approx. 5,000.00€,
which is equivalent to 900€/m2 of a building facade, where 30% of the total 3D
printing cost goes to the material cost and 70% to the printing cost, including
processing and all postprocessing.
 Printing cost is mainly influenced by processing speed.
 The necessary investment to prototype an AM metallic hybrid system in the size
required by Permasteelisa is between 1.6 and 4 mil€.
Key next steps are:
 Develop the most suitable AM process for CW frame manufacturing: (1) establish
team of experts; e.g. in collaboration with a research institute and a selected
manufacturer; (2) Develop research to decrease the fabrication cost by e.g. increasing
the processing speed; (3) establish new design process in Permasteelisa, e.g. re-design
the CW frame to be suitable for AM, apply techniques to design material, use new
software applied in automotive and aerospace industry, include optimization and
simulation, etc.; (4) monitor the change of standard design process in Permasteelisa;
(5) manufacture prototypes and test their performance, include quality management.
 If the final results of the R&D group’s research are satisfactory, prepare further
detailed feasibility studies about implementing a new design process and AM
technology as routine systems in standardized Permasteelisa’s design and
manufacturing process.
 Further investment and marketing prediction studies are required.
 3D Printing of large metallic building components is technologically almost feasible
today but it is cost prohibitive; develop a framework for practitioners to evaluate
when to chose among conventional, standard AM and hybrid metal deposition
systems to cost efficiently construct a specific metallic building component.
4
1
Table of Contents
1
Prologue ....................................................................................................................... 3
2
Table of contents ........................................................Error! Bookmark not defined.
3
Introduction ................................................................................................................. 7
3.1
State of the art in the AEC industry/ Permasteelisa ............................................. 7
3.2 State of the art of Additive Manufacturing (AM)/3D Printing (3DP)
Technologies ................................................................................................................... 8
3.2.1
Definition of Additive Manufacturing technologies ..................................... 8
3.2.2
Standardized AM approaches ..................................................................... 10
3.2.3
Manufacturers of AM technologies ............................................................ 13
3.2.4
Advantages and disadvantages of AM systems .......................................... 14
3.2.5
General conclusions about 3D Printing[13]:............................................... 16
3.2.6
Future predictions about AM technologies ................................................. 16
4 Points of departure: What is possible today? Current application of AM technologies
in the AEC industry (related to buildings and specifically building envelopes / curtain
wall systems) ..................................................................................................................... 18
5
Preliminary study to 3D print multifunctional CW segments .................................. 25
5.1
7
Preliminary study conclusions ........................................................................... 27
5.1.1
What is possible today .................................Error! Bookmark not defined.
5.1.2
R&D project ................................................................................................ 27
Requirements for AM to manufacture large scale CW components ........................ 28
7.1
Parameters of AM system technology crucial for the task................................. 31
7.1.1
Scale/Scalability of the Build Chamber ...................................................... 31
7.1.2
Process Speed.............................................................................................. 31
7.1.3
AM Materials .............................................................................................. 32
7.1.4
Automated Building Technology................................................................ 33
7.1.5
Conclusions ................................................................................................. 34
7.1.6
Conclusions after the literature review ....................................................... 35
8
Feasibility study goal and assumptions .................................................................... 35
9
Research methodology .............................................................................................. 36
10 Research methods and results ................................................................................... 36
10.1 Identify relevant manufacturers to contact ......................................................... 36
5
10.2 Prepare information about conventionally manufactured Permasteelisa CW
frame 37
10.3 Prepare files for AM manufacturers (3D models) .............................................. 39
10.4 Develop a questionnaire and distribute to all relevant manufacturers ............... 39
10.5 Calculate specific costs using scientific data ..................................................... 41
10.6 Conclusions after the first research round .......................................................... 41
10.7 Semi-structured interviews ................................................................................. 42
10.8 Final cost comparison: standard vs. 3D printed CW frame ............................... 43
10.9 Conclusions about manufacturing specific costs ............................................... 43
10.10
Description of chosen manufacturing processes ............................................ 44
Sciaky’s ..................................................................................................................... 44
DM3D ....................................................................................................................... 44
Optomec .................................................................................................................... 45
Fabrisonic .................................................................................................................. 45
11 Final conclusions ...................................................................................................... 46
12 References ................................................................................................................. 47
6
2
Introduction
2.1 State of the art in the AEC industry/ Permasteelisa
Architects today design complex geometries to stay present in a very competitive market.
The results are building forms that are aesthetically considered as world wonders but they
are source of many problems for the construction industry. For example, Foster's Apple
campus is nearly $2 billion over budget due to the complexity of its geometry of curved
glazing and special parts systems (Figure 1) [1].
Figure 1: Foster’s design for Apple campus with complex space-ship like geometry
Design and construction companies waste too much time, energy and resources to redesign, manufacture, transport, assembly and construct these complex geometries.
According to Permasteelisa group [2] (Figure 2), the world leader in design and
construction of building facades, approximately additional 20% of energy is wasted on
design and productions of additional systems and assemblies, e.g. human resources,
energy and material waste, related CO2 emissions, etc.
Figure 2: Shading elements on a Walbrook building facade in London; Foster + Partners design (Courtesy of
Permasteelisa)
7
One could say that buildings “leak” physically and virtually, where “leaks” are defined as
different failure types. The physical leaks occur at the intersection of different building
layers and elements (Figure 3), visible and measurable as specific building performances,
e.g. water, acoustics, air leakage, energy, resources, etc. The virtual leaks are related to
the financial lost, e.g. trough contracts or, as previously stated, trough energy waste due
to re-design and prototyping of different forms to accommodate architects’ complex
designs.
Figure 3: Buildings leak at the intersection of different layers and segments
2.2 State of the art of Additive Manufacturing (AM)/3D Printing (3DP)
Technologies
Additive Manufacturing is an emerging technology that could potentially eliminate
building leaks described above and successfully solve some of the problems mentioned
above. Often referred to as 3D Printing, it is a relatively recent approach to the
manufacture of end-use components that is based on creating parts and products directly
from raw material in powder, liquid, sheet or filament form and digital 3D design data.
This process works by depositing material layer by layer, without the need for molds,
tools or dies [3].
2.2.1 Definition of AM technologies
In 2009, after more than 20 years of confusing terminology, the American Society for
Testing and Materials (ASTM) International F42 committee on Additive Manufacturing
Technologies [4] defined AM as the “process of joining materials to make objects from
3D model data, usually layer upon layer, as opposed to subtractive manufacturing
methodologies.” These technologies were also called rapid prototyping, direct digital
manufacturing, solid freeform fabrication, additive fabrication, additive layer
manufacturing, and other similar technology names over the years. In the technical
8
community, an international consensus has coalesced around the use of “additive
manufacturing” whereas in the popular press the technologies are known as “3D
printing.” Every existing commercial AM machine works in a similar way (Figure 4);
first a 3D CAD file is sliced into a stack of 2-dimensional planar layers; these layers are
then built by the AM machine and stacked one after the other to build up the part.
Figure 4: A representation of a common 3D printing process: A 3D solid model of the desired component is created
with Computer Aided Design (CAD) software (1). The model is then typically translated into standard triangulation or
Stereolithography Language (STL), a standard data format that can be used by most AM machines. This describes the
surface of the object, which can then be ‘sliced’ into layers so that the part can be constructed sequentially (2). Each
layer is then sent to the machine (3) and the information is used to control the location of a printer head. The printer
head deposits a binder on a fine layer of powdered material where the layer is to be made solid (4). The machine
reconstructs a 3D object by sequentially bonding these ‘2D’ layers of material. [5]
In general, any technological process with controlled deposition of materials layer by
layer could be categorized as an AM process.
There is an obvious difference between additive and conventional manufacturing
technologies (Figure 5: Additive vs. conventional manufacturing); some technologies, called
subtractive manufacturing, like CNC machining, drilling, and sawing, operate by
removing material, while some technologies, called formative manufacturing, like
casting, molding, and forging, operate by giving shape.
Figure 5: Additive vs. conventional manufacturing
9
2.2.2 Standardized AM approaches
According to the Terminology Subcommittee of the ASTM F42 committee, today there
are seven different approaches to AM, and dozens of variants of these approaches: Vat
Photopolymerization, Material Jetting, Binder Jetting, Material Extrusion, Powder Bed
Fusion, Sheet Lamination and Directed Energy Deposition (Figure 6) [6]. Each of these
processes has many advantages and disadvantages[7]. The decision to choose a specific
system depends on a final product one wants to produce, either if it is prototype or final
product.
Figure 6: Diagrams of Seven Major AM Processes
Vat Photopolymerization (Figure 6) is an AM process in which liquid pre-deposited
photopolymers in a vat are selectively cured by light-activated polymerization.
Stereolithography, the first patented and commercialized AM process, works by scanning
a laser across the surface of a vat of photopolymer. A platform is raised to just one layer
thickness below the surface of the liquid. The laser scans the first cross-sectional layer,
attaching the layer to the platform. The platform is lowered within the vat one layer
thickness and material flows over top of the previously formed layer, or is formed over
the previous layers using a blade/deposition device. The process repeats until the part is
completed. Photopolymer vat techniques give some of the best accuracies and surface
finishes of any AM process. Some photopolymer vat technologies have been developed
to use DLP projectors to project an image of the layer on the surface of the vat, thus
cross-linking the photopolymer and converting the entire layer from a liquid to a solid
simultaneously. Photopolymer vat technologies require a support network to be built for
overhanging structures; otherwise these structures are subject to breaking or deforming.
These supports are made from the same material that the part is made from, so these
supports must be cut away after the part is completed.
10
Material jetting (Figure 6) is the use of inkjet printers or other similar techniques to
deposit droplets of build material that is selectively dispensed through a nozzle or orifice
to build up a 3-dimensional structure. Objet’s PolyJet is an example of this process. In
most cases these droplets are made up of photopolymers or wax-like materials to form
parts or investment casting patterns respectively. These processes are truly 3D-printing
machines, as they use inkjet and other “printing” techniques to build up 3-dimensional
structures. Photopolymers are useful materials for material jetting as they transform from
a liquid to a solid in the presence of light. Photopolymers can be tuned to cross-link and
harden in response to different wavelengths of light; and for AM they typically transform
in the visible or ultraviolet wavelength ranges. Material jetting techniques often use
multiple arrays of print heads to print different materials. The most common reason for
printing two materials is for one of the materials to be used as the “build” material, while
the second material is used as a “support” material. For 3D geometry that includes
channels, voids, or overhanging structures, a support must be built below any
overhanging surfaces, as droplets have to land on something to keep them in a fixed
location. When a secondary support material is used, a watersoluble material is
commonly used so that the supports can be removed by immersing the part in a waterbased liquid. Material jetting is capable of printing multi-material and gradient material
structures. Applications of multimaterial parts range from parts with controlled
hardness/flexibility to parts with differing electrical properties in various regions to tissue
engineered structures with different biological properties in different regions of the part.
In Binder Jetting (Figure 6) a liquid bonding agent is selectively deposited to join powder
materials. The commercially known ZPrinter line uses this process. This technique also
uses nozzles to print material. It starts by first depositing a thin layer of powder. A print
head is then used to print a glue pattern onto the powder, thus forming the first layer. A
new layer of powder is deposited and glue is printed again. This pattern is repeated until
the part is completed. Two benefits of binder jetting are its speed and its lack of need for
secondary support materials. Since the majority of the volume of the part is made up of
the powder material, only a small fraction of the volume of the part needs to be deposited
from the printheads. As a result, a layer can be formed very quickly, using arrays of print
heads – often in a matter of seconds. The powder which surrounds the part being formed
will naturally act as a support for any subsequent overhanging geometry, and no
secondary support materials are necessary. The only commercially available full-color
3D printing machines are binder jetting machines. A binder jetting machine can be set up
in such a way that a complete color spectrum can be printed layer-by-layer. This enables
assembled parts to be produced in the intended colors and for graphics, labels and other
visual features to be directly printed onto a part as it is being produced.
Material extrusion (Figure 6) selectively dispenses material through a nozzle or orifice in
a controlled manner to build up a structure. Stratasys’ Fused Deposition Modeling (FDM)
technology is an example of this process. The largest installed base of AM techniques is
based upon material extrusion. The build material is usually a polymer filament which is
extruded through a heated nozzle; an automated version of the hot-glue-gun used for arts
11
& crafts. After a layer of material is deposited by the nozzle onto a platform, the platform
either moves down or the nozzle moves up; and then a new layer of material is deposited.
In instances where two nozzles are installed in a machine, one of the nozzles is typically
used to deposit a water-soluble support material. Three or more nozzles are sometimes
used in machines designed for tissue engineering research, so that scaffolds and other
biologically-compatible materials can be deposited in specific regions of the implant. The
simplicity of material extrusion machines make them suitable for in-office and home
environments. Today, thousands of inexpensive extrusion machines, ranging in price
from $700-$3000, are sold each year to do it yourself enthusiasts, small companies,
educational institutions and hobbyists. This proliferation of low-cost AM machines is a
major reason for the current hype around 3D printing.
In Powder bed fusion (Figure 6) thermal energy selectively fuses regions of a powder
bed. EOS’ Direct Metal Laser Sintering (DMLS) is an example of this process. Machines
work in a manner similar to binder jetting but instead of printing glue onto a layer of
powder, thermal energy is used to fuse the powder into the desired pattern. In most
machines a laser is used to melt polymer or metal powders to build up 3-dimensional
objects. Another common variant is the use of an electron beam to melt metal powders.
In the case of polymer powders, the powder surrounding the part being built makes
possible the creation of complex 3-dimensional objects without supports. However, for
metal powders, the thermal shrinkage of metal parts during solidification causes the parts
to warp, and supports are used to attach the part to a thick metal baseplate to maintain the
accuracy of the parts and restrain them from warping. These metal supports are machined
off after the part is completed. Powder bed fusion machines are the most commonly used
AM machines for the creation of end-use parts for highly engineered products. Polymer
and metal parts made using these techniques are becoming widely used in aerospace,
defense and other highly-engineered systems.
Sheet lamination (Figure 6) techniques work by cutting and stacking sheets of material to
form an object. The Mcor Technologies printers use this type of process. This approach
has been used with paper, plastic and metal sheets to build up wood-like, plastic and
metal parts respectively. A binder is typically used to bond paper and plastic sheets;
whereas welding, either thermal brazing/welding or ultrasonic welding, or bolting of
sheets together is typically used for metals. Additionally, sheet lamination has been used
with ceramic and metal green tapes, e.g. powder held together by a polymer binder in the
form of a sheet of material, to build up structures which are later fired in a furnace to
achieve a dense part.
Directed energy deposition (Figure 6) machines melt material with a laser or other energy
source as material is being deposited. These machines work similarly to material
extrusion machines except that, instead of melting the material with a nozzle, the wire or
powder feed material is melted as it is being deposited onto a part. In order to make parts
with overhangs, directed energy deposition systems need to either use a 5-axis deposition
system, so that material can be deposited from any orientation, or a secondary support
12
material. As these systems are typically used to make metal or ceramic composite
structures, any supports that are used require machining to remove them. Directed energy
deposition processes are used primarily to add features to an existing structure, such as
adding strengthening ribs onto a plate, or for repair of damaged or worn parts. In most
cases these processes are used to build up metal structures, and thus they are commonly
referred to as metal deposition AM machines.
Hybrid and direct write AM technologies combine either multiple AM techniques within
the same machine or AM is combined with subtractive techniques, such as CNC milling
or laser cutting. For example, by combining a simple horizontal milling head into a
material jetting machine, Solidscape has created a popular line of high-precision wax
printing machines, with layer thicknesses of 0.0005”, that are heavily used in the jewelry
industry to create wax patterns for custom jewelry. Several AM techniques have been
modified to work at a small scale to deposit passive electronic structures, e.g. conductors,
insulators, resistors, antennas, etc. These techniques are often known as direct write
techniques and, for example, use electronic “inks” that contain nano-particles or other
additives that result in electronic properties after drying, thermal decomposition or other
post-treatment. By combining direct write techniques with other AM techniques it
becomes possible to create multi-functional 3D-embedded electronic structures on a
layer-by-layer basis that combine structural, thermal, electronic and other functions into a
single component.
2.2.3 Manufacturers of AM technologies
Today’s prices of 3D printers range from approx. $1,000.00 for low cost printers to $1,5
million for high-end printers, and the prices for some hybrid systems are even higher. The
major high-end manufacturers and service providers in the industry are: Stratasys, EOS,
Makerboot, Optomec, ExOne, Organovo, Materialise, etc. On the other side, almost
every country has at least one low cost printer manufacturer, e.g. Type A Machine, 3D
Kits, Cubify, BatBot, Ultimaker, TrinityLabs, etc.
According to one Lux research [8] the automotive, aerospace and medical industries take
84% of the market. Figure 7 shows some product examples from those industries.
13
Figure 7: Examples of products manufactured using AM technology from automotive, aerospace and medical
industries; above from left: BMW’s Formula One steering wheel, Airbus’ wind bracket (metal prototype);
Computational design optimization for wing bracket (metal prototype); below: bionic arm (polymer prototype with
printed sensors), Metal hip implant (Various online sources)
2.2.4 Advantages and disadvantages of AM systems
Main benefits that 3DP is bringing to the manufacturing world are complexities of
geometries that couldn’t be produced with any other technology and customization of
products according to specific users. Further benefits are possibility of parts
consolidation, assembly processes or tools, molds or dies are not required, new unique
materials are designed and manufactured without any waste and with controllable
microstructures, e.g. multi-materials and gradients of new products with embedded
electronics or integrated functional new structures. As a result special geometries and
functions are discovered that were not considered before.
According to some experts 3DP already brought new revolution into design process
because, in theory, human intervention is not required in transformation of a digital to a
physical model (Figure 8).
Figure 8: Digital Trinity brings revolution in a design process [9]
Main general disadvantages of AM systems are: limited palette of available materials to
be processed, low process productivity, poor surface finishing, dimensional accuracy for
large size, great production cost at medium/large volumes, process innovations are
needed, faster build speed, etc. According to experts, commercialization of new
functionality products is expected to be 10-15 years away, substantial technological
hurdles need to be overcome for the technology to reach the peak of its productivity.
14
3DP is also bringing new business models [10][11][12], has impact on logistics of
production and as a result an impact on social spheres. It enables business models used
for 2D printing to be applied in 3D. For example, one could either print at home, at local
FedEx, Kinkos, through Shapeways or at local store. As a result, it removes the low-cost
labor. New horizons for entrepreneurship are opened as the patents expire, new software
tools are developed and the number of service providers is growing. User-changeable
web content plus a network of AM producers already enables new entrepreneurial
opportunities, e.g. Shapeways.com, Freedom of Creation, FigurePrints, Spore, etc. It
eliminates drivers to concentrate production where “design and manufacture anywhere”
is now possible, changing “Just-in-time Delivery” to “Manufactured-on-Location Just-InTime”, which makes local manufacturing of products normative. As a result, small
business can compete with multi-national corporation to produce goods for local
consumption and parts produced closer to home cost the same as those made elsewhere.
Furthermore, minimizing shipping/transportation drives regional production, resulting in
a reverse increasing urbanization of society, e.g. move to a “big city” is not needed if one
can design his/her products and produce anywhere. It is also visible that the creativity in
design becomes more important than labor costs for companies to be successful.
Figure 9: Application of AM technologies for customized furniture and clothing (the design process starts with 3D
scanning of human body); above: 3D printed lounge chair, Neri Oxman, MIT; below left: 3d printed topologically
optimized car seat; Toyota in collaboration with Materialise, Belgium; Below right: Nervous Systems created an online
shopping process where a customer can order a dress perfectly adjusted to his/her body after sending the data of body
scanning
15
2.2.5 General conclusions about 3D Printing[13]:






Technology for 3DP of products is highly accessible and a very useful tool in
education.
3DP provides more than one path how to design and produce, it has and will have in
the future a great impact on manufacturing industry, e.g. new ways of thinking and
making business, e.g. inspect and validate with optimization algorithms as you go; it
allows on site production; hence the global scale impact on economy, law, DRM, etc.
3DP will further flourish in special areas where digital technologies are developing,
e.g. San Francisco Bay Area.
In general, the market and developments depend on the money flow but, according to
experts, for AM technology application in the medical industry markets have
relatively certain future investment because of its benefits like mass-customization,
nano-scaled controlled structures, bio-printing, etc.
Substantial research is still required in size, speed, density and reproducibility of
material/s properties, surface finish, and quantification of the potential in AM
material recycling, etc.
Momentum for product mass-customization has probably passed; it is not feasible
that every person will design and 3D print different products, but talented designers
will prevail.
2.2.6 Economic predictions about AM technologies
The economic predictions are conflicting (Figure 10) but relevant experts agree that
further growth is expected if the current compound annual growth rate of around 30% per
annum remains.
Figure 10: Economic predictions for 3D Printing: Wohlers/Deutsche Bank claims that industry growth has accelerated;
Canalys, Citi Lux show conflicting growth expectations [14]
16
According to Gartner[15][16], every technology goes over time trough a hype cycle;
from the innovation trigger to the plateau of productivity. Figure 11 is showing the latest
hype cycle for 3d printing in 2014. Scientists experts consider this prediction speculative,
but the diagram does give an overview on how specific aspects of 3DP are going to
develop over time and gives a relatively accurate image of what are consumer needs and
where and when to invest. For example, macro 3DP, an interesting topic for the
construction industry, is still in the area of innovation trigger and it will take more than
10 years to reach the plateau of productivity. On the lower end of the technological
spectrum, low cost systems aimed directly at consumers are already available. According
to diagram, consumer 3DP just passed the peak of inflated expectations. These systems
feature limited capabilities and are useful for making early prototypes, decorative items
or simple parts for non-demanding applications. Although such machines triggered the
current hype surrounding 3DP, they make up only a relatively small part of the overall
3DP industry and are responsible for less than 10% of annual industry turnover.
Figure 11: Hype Cycle Curve for 3D printing in 2014[16]
17
3
Points of departure: What is possible today? Current application of AM
technologies in the AEC industry, related to buildings and specifically to
CW systems
The tendencies in several scientific research groups around world are to scale down the
operating principle of 3DP for very small applications (Figure 12)[17]. It is expected that
this could be highly beneficial for novel types of medical and biological applications,
new types of sensing equipment and the downsizing of electronic structures.
Figure 12: On the left: 3D printed electronics; on the right: 3D printed microstructures [18]
For example, at the EPSRC Centre for Additive Manufacturing [19] at the University of
Nottingham, which is one of the UK’s leading 3D Printing research group, scientists are
working on processes capable of building up functional structures smaller than one
micrometer. These systems employ advanced technologies such as two-photon
polymerization and optical tweezer technology (Figure 13) capable of manipulating very
small structures using only laser beams.
s
Figure 13: Research into micro-scale 3D Printing at the University of Nottingham
For the application in the construction industry, preliminary studies [9][20][5] showed
that, besides the reduction of problems on construction site related to human factors due
18
to partial automation, there are three key issues that affect how AM application impacts
traditional construction methods: cost (Figure 14 13), time (Figure 14) and value added
(Figure 15).
Figure 14 : Traditional and 3D-printed wall construction cost comparison[20]; To identify the principal cost issues
associated with using additive processes for building large structural components a hypothetical wall structure was
devised, based on a typical domestic housing application. The wall is comprised of 13 mm of internal plaster finish on a
100 mm concrete block, with a 50 mm cavity and 100 mm external facing brick. Fixings, brick ties, insulation, etc. are
neglected. The wall is assumed to be 5 m long by 3 m high. Z Corporation's 3D printing process and powdered gypsum
as the build material were used to hypothetically calculate the cost of such as 3D-printed wall.
Figure 15 : Comparison of time to completion [20]. The same wall as in Error! Reference source not found. was
evaluated in terms of build time to completion. The diagram compares building duration using a traditional process and
the constant build rate of 43 mm/h associated with the 810 machine (Z Corporation's 3D printing process). The steps in
the traditional methods come from having to leave every ∼1 m height in brickwork overnight for the mortar to cure
(maximum weight on wet mortar). Assuming no operational efficiency in the labor allocation (continuous work) and
neglecting the set-up time for the machine, the 3D printing is comparable in build time to traditional methods.
19
Figure 16 : Thermal conductivity/density characteristics for building materials and 3D printed test panels[20][21]. The
thermal conductivity of primary building materials is important and the requirements for greater levels of insulation are
only going to increase. Using a primary build material with a low thermal conductivity is therefore beneficial.
Insulating concrete blocks are a common solution for the wall construction. The experiment considered a number of
potential wall panel designs. Using the Z Corporation 3D printing process, the panels were constructed from gypsum.
The internal geometry of the panels was designed to test the extent to which the material geometry could be used to
improve thermal conductivity. Two of the 3D printed panels are depicted on the left side of the Error! Reference source
not found.. Each panel, including a panel of solid material, was tested on a UKAS accredited EN 12667 guarded hot
plate apparatus. Although this does not generate the industry accepted k-value (thermal conductivity, W/mK), the
results give a good indication of performance. The graph shows approximate regions of k-values for typical building
materials, plotted against material density. The test panels demonstrate a performance at least as good as aerated
concrete. The performance is achieved through maximizing the resistance of the conduction path. The panel in the left
had a k-value of 0.112 W/mK, nearly half that of the right hand panel.
In theory, once the AM layer-by-layer approach is considered for construction, the
possibilities for system integration and reduction of material interfaces are striking [5]
(Error! Reference source not found.):
 Integration of mechanical and electrical services within the structure means reduced
amounts of wasteful and time-consuming work by builders;
 Better control over the deposition of build material will result in better internal and
external finishes, completed as the structure is built;
 Being able to consider the structure as a homogenous unit will negate the need for
difficult interface detailing, reducing the chance for error and hence costly remedial
work;
 The coupling of digitally controlled process with solid modeling techniques will mean
greater design freedom at no or little extra cost, etc.
20
Table 1: Possibilities for processes for creating freeform structures [5]
Three most famous research projects to 3D print large components and even entire
buildings (Figure 17, Figure 18, Figure 19) have goals to create novel types of highperformance architecture which may be highly customized and energy efficient and to
realize improved construction supply chains, thereby reducing the overall environmental
footprint of construction activity. A mega-scale 3DP process for concrete structures
developed at the University of Southern California is aiming to significantly reduce the
time and cost of construction projects, especially in inaccessible locations, while
maintaining high quality of construction.
Figure 17: The Canal house from DUS architects in Amsterdam[22][23], where each room is a separate 3D printed
unit, material is photopolymer, each unit is printed and tested on site; building will be afterwards assembled on site.
21
Figure 18: On the right, Mobius strip/ Landscape house by Dutch architect Janjaap Ruijssenaars in collaboration with
Italian roboticist Enrico Dini [24]. The slabs and structural walls will be 3D printed in segments and assembled on site.
Material is ceramics found on site.
Figure 19: Mega-scale 3D Printing system[25][26], Courtesy of Behrokh Khoshnevis, University of Southern
California
It is unlikely that 3DP technology will take over the construction industry in the near
future[27]. Complete automation and digital control of a construction site might be
possible on Mars or Moon [28][29], but here on the Earth the construction will probably
remain as a hybrid system composed of different technologies due to its complexity. It
would be reckless to force a new production technology where a traditional system is
more efficient but also not to harvest the benefits the 3DP is offering.
In comparison with traditional manufacturing, research on the 3DP of metal parts for
complex building envelopes at the Technical University Delft has shown that
enhancements are possible on many levels, including smaller stockholding requirements,
greater manufacturing efficiency, easier installation and higher product performance [30].
Effectively, 3DP promises “just in time” facade manufacturing, removing the necessity
for production in advance and storage of large numbers of parts. Moreover, it avoids
large investments in production tooling and permits frequent design improvements.
22
Figure 20: Semi-finished 3D printed corner cleats[30] Working on the semi-finished part level means to optimize the
individual product. Corner cleats for frames and window profiles, from the system portfolio of Kawneer-Alcoa, from a
current production series are used to stiffen profile corners as well as a connecting piece for gluing and grouting the
profiles. All parts are evaluated in terms of their potential for optimization, the result of the optimization attempts and
their ranking after completion of the research project. The part was used as a starting point. Optimization potential is
given by the fact that the angle (from 30° to 180°) has to be adjusted manually. A higher degree of stiffness of the
corner connection could be achieved if the part was printed with ‘digital’ angles. The angles of a window system could
be digitalized and transferred to the necessary CAD files for production. The corner cleats could then be printed
individually for one entire window or facade system, in the exact quantities needed for a particular project. The cleats
would be more rigid and stiff due to the fact that the loose pivoting joint is eliminated. Material savings are achieved by
implementing a lightweight structure into the massive area of the cleat. The potential for an immediate application is
obvious, supported by the feasible size of the parts. The application and realization with DMF is possible, savings can
be achieved because there is no more tied up capital for infrequently used products.
23
Figure 21: Semi-finished 3D printed T-connector [30] offers good potential for optimization because it has different
designs to fulfil the free-form task of the AA-100 system. The digital integration of angles and the possibility to save
material and enhance performance show how the system benefits from AM improvement. Material savings of 25 %
were achieved compared to the orthogonal standard connector by digitally ‘cutting off’ the material where it was not
needed. The mounting system remained the same. Therefore all tools and accessories of the standard AA-100 system
can be used. The advanced connector is the first part of the project that was ‘printed’ in stainless steel. It shows the
possibilities and the change in design and performance, even. If it lacks further engineering, improvements in the
performance characteristics could be achieved by sequential digital simulation and evaluation using a FE analysis.
(Source: Holger Strauß: AM Envelope: The potential of Additive Manufacturing for façade construction, PhD Thesis at
Delft University of Technology, Faculty of Architecture, Architectural Engineering + Technology department, 2013)
Figure 22: An optimized nodal point in a curtain wall[30] The “Nematox” node is the first printed facade node for a 1:1
mock-up. Hybrid constructions from system components and individualized AM parts show a realistic path that could
be followed as a first step. To produce this type of node, it takes a lot of effort and understanding of both the
engineering aspects of facade systems and the great range of CAD tools to sketch and script the node. For the two
versions of the nodal point 120h of CAD engineering were needed to generate a print-proof *.stl-file: 60h were needed
for the first attempt with all joints and dimensions to fit the standard aluminum profile. This was followed by the
decision to reduce the costs for the prototype by changing the dimensions of the profiles to a smaller size. These
changes took another 50h of CAD work. 10h of computation and translating were needed to finalize the *stl. file, and
another 2h to place the part in the virtual building chamber of the AM system software. And finally, it took 76.5h of
processing time to ‘print’ the nodal point in aluminum as a 1:1 prototype with lasercusing (by conceptlaser). Upon
completion of the build job, the part was finished and post-processed in another 4h of labor. The team that designed
parts was as hybrid as the facade itself. The nodal points show a way to realize free-formed facades with a standard
facade system combined with parametrically planned components. The optimization potential stretches across different
areas of facade manufacturing and assembly. (Source: Holger Strauß: AM Envelope: The potential of Additive
Manufacturing for façade construction, PhD Thesis at Delft University of Technology, Faculty of Architecture,
Architectural Engineering + Technology department, 2013)
24
4
Preliminary study to 3D print multifunctional CW segments
Previously described research at the TU Delft [30] showed the potential of optimized
customized and 3D printed metal CW frame segments. Furthermore, Arup, in
collaboration with EOS, WithinLab and AM partner CRDM/3D Systems developed
design method [31][32] that allows 3DP techniques to be used in production of critical
structural steel elements for use on complex projects.
Figure 23: A conventionally manufactured steel node (that is at the moment still cheaper to produce, but it is expected
that this will change in the short term) vs. 3D printed optimized node for the same loads (Online source:
http://www.arup.com/News/2014_06_June/05_June_Construction_steelwork_makes_3D_printing_premiere.aspx)
It would be feasible for Permasteelisa’s R&D group to explore further the design of
highly optimized and customized multifunctional CW segments and develop a new
design process that could potentially revolutionize standardized design techniques in the
company, but also have an impact on design and manufacture practice of building
envelopes. Furthermore, 3D printers to manufacture these complex CW segments are
available on the market. This research idea could be further developed by following the
objectives and the results of the Final Report of ATKINS project in Additive
Manufacturing (A Low Carbon Footprint) [33] :
 Investigation of waste minimization during production: ensuring optimized and
repeatable AM production systems to substantially reduce or eliminate waste materials;
 Exploration of process efficiency gains: using AM processes to replace inefficient and
wasteful conventional manufacturing processes;
 Reducing transportation: Using digital supply chains and AM technologies to
significantly reduce logistical requirements by shortening the supply chain and
minimizing the need for waste material disposal or recycling;
 Product design for whole life cycle impact: Exploiting AM design freedoms to
minimize weight for significant reductions in greenhouse gas emissions over the whole
product life cycle;
25
 Product design for optimized performance: manufacturing truly optimized products
that are more efficient in their application compared to traditional parts constrained by
the standard process “Design for Manufacture”.
Suggested by dr. Danijel Mocibob, a member of Permasteelisa’s R&D group, first we
explored the usage of AM technology to improve performance of a bracket, i.e. a
multifunctional CW component. The bracket is a plate that transfers the loads from the
CW system into the primary slab structure. Its performance was improved by adding new
lightweight structures, e.g., lattices that could be manufactured only with AM
technologies and by optimizing its geometry to resist wind loads, absorb the blast energy,
dissipate the earthquake load, damp the extreme wind, disperse impact load from gondola
and large-missile (Figure 23).
Figure 24: Potential research project: Optimized and customized multifunctional bracket
Specific costs[34][35][36][37] to manufacture different sizes of multifunctional brackets
(Figure 25) with two metallic AM technologies, Direct Metal Laser Sintering (DMLS)
and Electro-Beam Melting (EBM) were calculated (
Table 2).
26
Figure 25: Two sets of dimensions of multifunctional bracket to calculate AM production costs
Table 2: Costs to manufacture multifunctional brackets from Figure 25: Two sets of dimensions of multifunctional
bracket to calculate AM production costs
4.1 Preliminary study conclusions
The key findings about 3DP of highly optimized multifunctional CW metallic segments
in sizes that could fit commercially available machines are:
 Research has low risk as an investment for Permasteelisa’s R&D group because
necessary knowledge and an established research process already exist in specialized
institutes.
 Specific cost is not competitive for manufacturing traditional geometries currently
produced with conventional methods, but it is highly competitive for multifunctional
application and further research in adding new functions to specific components.
 Commercial machines with appropriately sized build chambers are available on the
market, e.g. EOS, Arcam.
 AM processes exist in arrays of different materials; additional investments are needed
in material research if unconventional materials are designed for application.
 A new optimization (simulation and FEA analysis) and customization process, as part
of AM, brings revolution in design of not only curtain walls but buildings in general.
4.1.1 R&D project
Direction 1:
 Develop (design/optimize/customize, 3D print, test the performance) one specific
multifunctional CW component (e.g. bracket);
 Predicted cost of the research project is approx. 0.5 mil€;
27




Expected duration of the project is 1 year;
Low risk project regarding the investment;
Suggested partners for the project are research institutes, e.g. Additive
Manufacturing and 3DP Research Group at the University of Nottingham, UK;
Project is suitable for Permasteelisa’s internal research.
Direction 2: For further investigation of other 3D printed multifunctional CW
components allocate the resources and funding, e.g. HORIZON 2020.
28
5
Requirements for AM to manufacture large scale CW components
Buildings are complex large-scale products composed of numerous smaller parts, for
example, of different CW systems. Based on experience of the other industries where
different highly optimized small-scale products were successfully digitally designed and
3D printed, in theory it should be possible to develop a process and a methodology, as
well as identify and/or develop a computational technology to prototype a full scale CW
frame as a single innovative high performance large-scale facade element customized,
optimized and calibrated to chosen environmental constraints using AM technology as
the final output. The 3D printed prototypes would be manufactured and tested in a
controlled environment on- or offsite.
Currently there are a few experimental projects where large scale products, of the scale of
one story CW frame, use a type of AM technologies as a segment of a production
process. One example is the work of the American artist Ioan Florea[38][39]. In
collaboration with Voxeljet[40] Florea 3D prints designed molds/segments of his
sculptures (Figure 26: Ioan Florea “3D printed” car shell and wall; taken by the author at the Inside 3D Printing
conference, New York, April 2014 which are afterwards assembled in a continuous large scale
surface, e.g. wall. The manufacturing process that Florea is using for his art is not
applicable for CW production because the final product does not have performance
mandatory for building envelopes; e.g. structural, thermal, resistant to weather
conditions, acoustic, durable, manufacturable, constructible, economically feasible, etc.
Furthermore, the AM technology used here is only one manufacturing step in a complex
manufacturing chain that involves human intervention.
Figure 26: Ioan Florea “3D printed” car shell and wall; taken by the author at the Inside 3D Printing conference, New
York, April 2014
Permasteelisa was using multistep manufacturing technologies to produce complex
geometries in previous projects, for example, in manufacturing of shading elements on
the Walbrook Bank in London [41](Figure 27Figure 28). Inspired by wings of airplane
these shading elements were designed by Foster and Partners. After numerous redesigns,
tests and trials the shading elements were manufactured from composite Glass Fiber
Reinforced Polymers (GRP). The composite material and the final product were
rigorously tested and the conclusions were that GRP´s are easier to use for free form
geometries than cast aluminum which is expensive and offers only medium or low
surface quality. Manufacturing and testing of elements was performed in three different
29
factories in Europe and Russia. The quality/performance of the final product was tested in
different German institutes. The main disadvantages of this manufacturing process is its
complexity by having many processing steps and geographical dislocation of specific
manufacturing steps. The aim of this feasibility study is to explore AM technologies that
could localize the production and quality management under “one roof“.
Figure 27: Complex manufacturing process of shading elements designed for Walbrook bank, London
30
Figure 28: Walbrook, London, Shading elements; Production sites and quality management workflow[41]
31
5.1 Parameters of AM systems crucial for the task
5.1.1 Scale/scalability of the build chamber
The obvious discrepancy between building technology and AM is the relatively small
size of the process chambers offered by the systems currently available, especially for the
AM technologies that could be applied for the CW frame production. Table 3 gives an
overview of AM processes, available on the market that could be potentially used for
manufacturing large-scale CW components.
Table 3: Relevant manufacturers/service providers for 3D printing large scale building components
The most suitable AM technologies for the CW fabrication have small build chamber size
and vice-versa, if the build chamber size exists, then the other characteristics of the
system, e.g. mechanical properties of materials in the process are unsuitable for CW
frames. Nevertheless, the author decided to contact all above numbered manufacturers to
inquire about the state of the art of a specific technology.
5.1.2 Process speed
The AM methods cannot achieve fast enough processing times that could be comparable
with conventional manufacturing processes yet. At the beginning of the development, the
focus lay on comparing conventional prototyping with Rapid Prototyping. The fact that
AM technologies do not require tools led to an advantage, particularly in terms of the
time needed to turn the initial idea into the finished prototype. Another benefit of AM is
the fact, that the product can be developed quickly in collaboration with the customer or
user, and that initial improvements can be integrated during the developmental stage.
Since making new tools is extremely expensive, this option simply does not exist with
conventional methods.
Another advantage is that with AM the production of small series or even single pieces is
economical; it is irrelevant whether 100 equal or 100 unique parts are 3D printed which
32
allows mass-customization: production of adjusted products to specific customer’s
demands. Complexity of customized products play minor roles in measuring time, but if
we draw a direct comparison between AM and conventional fabrication processes, this
advantage is lost and the processing time turns into a disadvantage. High speed milling or
CNC controlled processing centers feature disparately higher yields because the products
are generated from semi-finished parts. An AM manufacturing cycle for each part is the
same. Currently, only the production of large quantities of small parts yields satisfactory
results.
If AM technology should be applied in building technology, where a large number of
building parts must be available in a relatively short period of time, the processes must be
included in construction scheduling and planning, e.g. in a curved facade designed to be
produced with AM nodes, one would require several hundred nodes; it might take 24
hours to produce one component at a high risk of defects, but at the same time AM
technology does provide better solutions for complex CW interfaces and joints [30].
5.1.3 AM materials
Only a few of the materials developed for AM technologies are perfectly suited to be
transferred to architecture. Therefore the available materials need to be further developed.
There is a noticeable discrepancy when considering a direct transfer of the manufacturing
principles. Current process chambers are dimensioned for components, not for entire
building elements. The system technology is small scale oriented and the results are
therefore fine and precise. Material-specific issues must be considered if the know-how
of these systems is transferred to large-scale equipment to manufacture building parts or
housing.
To transfer AM to facade technology we must not only consider the material properties of
the final product but also a change in system technology caused by the different size of
the components.
The main criteria are [30]:
5.1.3.1 Material viscosity
The challenge of manufacturing larger structures is to ensure high quality. Problematic
areas are the hardening behavior as well as the stability of the structure during
production. Furthermore, when using AM materials, care must be taken to have sufficient
material available, and that the base material is formulated in a manner that the individual
layers bond to one another and cure quickly into a monolithic form.
5.1.3.2 Method of application
Technologies using extrusion nozzles designed for the millimeter range do not
necessarily function as well when employed for larger sizes. Material properties change
significantly from being applied in a thin capillary tube to a large diameter tube. Thus,
changing material properties must be examined in terms of controllability and
33
homogeneity of the printed structure. Minimum and maximum achievable resolution for
large structure details must also be observed. It is critical to choose the appropriate
material for large areas or small details (cement, aggregate, grain size, material mix). The
larger the extruded material quantity results in the lower the resolution of the final
product.
5.1.3.3 Material behavior of composite parts
Different melting temperatures, curing behavior and curing times need to be considered
when using different materials in one component. Process temperatures and process
speeds can vary as well, depending on the chosen materials. Appropriate software
solutions for simulating deformation and tension during and after a building process are
necessary to achieve consistent material quality[42].
The incorporation of all factors mentioned above in the development of new materials
results in very specialized materials. The end use of the desired products determines the
development of the various material groups.
A development for direct application in the building technology does not yet exist.
Currently, only the Contour Crafting method and the methods for direct fabrication of
metal parts employ materials in raw material form. Functional Gradient Materials (FGM)
and metals offer the most potential for direct application in the building industry.
Conceivable products could include electric lines, products with different material
densities or hard as well as flexible areas, and the use of different materials in one
manufacturing process. In order to achieve this, the material properties as well as the
manufacturing systems must be further developed. In general, mechanical behavior of 3D
printed materials[43] is predictable based on the traditional understanding of
microstructure and processing, porosity has a strong influence on the mechanical
behavior, anisotropy is not an issue if parts are built with low porosity and good layer
interface and polymer produced using best practice have isotropic strength and
anisotropic ductility. A universal 3D printer that could print at the same time different
materials does not exist yet.
5.1.4 Automated Building Technology
It is not sufficient to simply scale up available AM technologies to the dimensions
necessary for a building. This step requires several other considerations in terms of
equipment technology as well as material selection. Several concepts following this
approach have been developed and they provide answers to how current AM technologies
can be modified to meet the demands of large structures. One example is previously
mentioned Contour Crafting process.
Automation in construction is not a new term; since 1960–s there were attempts to
automate a single or group of construction works. Based on good experiences with
manufacturing robots in the Japanese automotive industry, research and development of
robotics for use in the building sector started in the late 1970s. Two types of robots could
34
be differentiated: systems that handle entire process steps of the building construction, for
example to create the shell construction, or robots for smaller specialized tasks such as
welding steel carriers or assembling dry walls. In 1983, the first robot designed for
flameproof coating of steel components was presented to the public. And from 1991 until
1993, the first system that built an entire building was realized by Shimizu Corp. in the
city of Nagoya, Japan. The invented robotic systems are linked to a fix scaffold. Highrise buildings are constructed in layers, just like the models created with AM methods.
There are two basic systems: either each story is manufactured on the ground and moved
underneath the previously assembled stories or a ‘climbing’ system stacks individual
stories on top of one another. The initial goal of building robots was to increase the
productivity and reduce the cost for high-rise building projects. These goals were not
realized because the systems were too inflexible, and because high-rise building projects
are unique in appearance and material choice. The main benefit of automation, as it is
used by the automotive industry for repetitive process steps, for example, does not apply
to building technology. Therefore none of the above mentioned systems are in use today.
The only benefit from applying these technologies in combination with pre-manufactured
elements was a reduction of on-site material waste of 70% compared to traditional
building methods.
Research at the Faculty of Architecture at the Swiss Federal Institute of Technology
Zurich (ETHZ) [44][45] is conducted on possible use of robots for digitally controlled
production of building parts and the programmability of parts with the resulting level of
freedom.
5.1.5 Conclusions
The introduction of the different methods illustrates that the performance of AM facade
components can be reduced to the material properties, and that it is not mainly influenced
by the AM technologies themselves. All technologies have advantages and
disadvantages, but their fundamental functionality is similar. Therefore, the initial
decision to be made in terms of a possible application in the facade is to decide whether
one wants to use only proven materials, e.g. metals, or if one is open to use yet to be
developed materials like plastics, specifically designed for a particular purpose [46].
The differences between the different AM materials groups are:
 The need to process metal powders in a protective atmosphere in order to avoid
exothermally reactions and uncontrolled deflagration, when coming in contact with
oxygen.
 The dependency on exclusively available materials from the system suppliers, in order
to guarantee a certain component quality and secure warranty claim for the AM system.
Only a few Direct Metal Fabrication (DMF) systems allow the use of industry standard
powder without loosing the warranty claim.
 The limitation of the building chamber for the AM technologies for plastics.
35
Powder fed DMF processes are the best choice when trying to extend the size of the
process chamber because it allows for a combination with robot arms or large process
platforms. For plastics, only 3DP methods offer the possibility of significantly larger
process chambers.
5.1.6 Conclusions after the literature review
Since the systems are very expensive they need to produce high quality parts in order to
be economically efficient. The parts must feature obvious advantages over conventionally
produced parts to make the expensive production technology feasible. To guarantee a
successful use as manufacturing methods in real-life applications, production
management, quality management and a reliable standardization of norms and parameters
must be established. In addition, testing methods need to be established that evaluate
parts whose geometry changes constantly.
It was a strategic decision to focus research in this feasibility study on metal deposition
systems to reduce investment in a new material research. Furthermore, mechanical
properties of 3D printed metals are equivalent to metals produced with conventional
methods, if not even better[43]. Special attention is placed on hybrid systems that
combine an AM metal deposition process like EBM with a post-processing system like
CNC-milling.
6
Feasibility study goal and assumptions
The purpose of this feasibility study is to gather the preliminary information about the
state of the art, necessary future developments and cost of a specific AM process to 3D
print large scale components. The main goal is to investigate possibilities to 3D print a
full scale one story high CW frame as a single element on a construction site to eliminate
the cost of extrusion, assembly, transportation and packaging.
Feasibility study assumptions:
 3D printed frame should have the equivalent volume and the equivalent stiffens of a
conventionally manufactured CW frame;
 Investigate scaling up of existing desktop size technologies; scaling up of these
systems is technologically possible;
 Investigate hybrid solutions of existing technologies;
 The research focuses on metals: aluminum, stainless steel, titanium or any other metal
combination available on the market.
If an AM technology proves to be feasible regarding above mentioned issues, the plan
would be to manufacture and test these frames in collaboration with a specific
manufacturer. If tests prove positive, according to the technical director, Permasteelisa
would be interested in purchasing one of these feasible machines for future testing and
fabrication of CW frames.
36
From previous research I found that manufacturing complex structures with
improved/new functions using AM technologies does not influence the production cost or
the build rate but it is influenced only by quality of design, as a result of
optimization/simulation processes. Quantification of these benefits will be the topic of
further research.
7
Research methodology
My proposed research process:
1. Identify all relevant manufacturers to contact; literature and market review,
2. Gather and prepare CAD files and cost information about the supply chain of
conventionally manufactured Permasteelisa CW frame,
3. Gather information about cost, build time and energy consumption from relevant
manufacturers using questionnaire, semi-structured interviews, joint work efforts with
AM designers,
4. Calculate specific costs using data from scientific articles and information collected
from manufacturers,
5. Compare the specific cost of the CW frame manufactured using traditional technology
with the cost of hypothetic 3D printed frame with the same geometrical characteristics as
the traditional frame.
Due to unavailability of information from relevant manufacturers, the research process
had three rounds; comprehensive results are described in detail in the following chapter.
8
Research methods and results
8.1 Identify relevant manufacturers to contact
Through literature and market review, I identified all AM technologies with the potential
to manufacture large-scale CW segments with build platform sizes up to 4x1,5m. Table 4
shows a systemized overview of these processes. My conclusion after the first overview
was that metallic 3D printing technology with platform size 4x1,5 m does not exist.
37
Table 4 : AM technologies with the potential to be applied for manufacturing of large-scale CW components
8.2 Prepare information about conventionally manufactured Permasteelisa CW
frame
The first step in comparison of supply chain costs between traditional and 3D printed
frame with the same geometry was to gather information about two exemplary
conventionally manufactured and installed Permasteelisa’s CW frames. Dr. Danijel
Mocibob, a member of Permasteelisa’s R&D group provided the AutoCad drawings of
two CW examples: Series 190.dwg and Tower Bridge.dwg (Figure 29) and the supply
chain cost information (Table 5).
38
Figure 29: Permasteelisa’s AutoCad 2D Drawings of CW frames manufactured using standard technologies; Series 190
and BridgeTower/Shard
Table 5: Supply chain cost comparison of two conventionally manufactured Permasteelisa’s CW frames
39
8.3 Prepare files for AM manufacturers (3D models)
Permasteelisa’s original detailed drawings (Figure 29) were two-dimensional and created
in AutoCAD. AM manufacturers work only with three-dimensional digital models, so for
the purpose of getting their response about the manufacturing costs, I prepared 3D
models (Figure 30) in Rhinoceros software [47].
4000
1500
Figure 30: 3D Rhino model of Series 190 frame
8.4 Develop a questionnaire and distribute to all relevant manufacturers
For my initial approach, I created and emailed a questionnaire with customized questions
to all manufacturers specified in Table 3. Table 6 shows an exemplary questionnaire
prepared for the Optomec representative. Unfortunately, not one of the contacted
manufacturers filled out the questionnaire due to time restriction, only Optomec sent
detailed specifications about their LENS’ system [48]. I decided to simplify and reduce
the number of questions, use only one CW frame 3D model (Figure 30) and conduct
semi-structured interviews via telephone/Skype. In general, all were prepared to talk
about the investments necessary to develop their existing systems for Permasteelisa’s
requirements.
Table 6: Customized questionnaire for Optomec
Please answer following questions (for the answers that you do not have exact data, please give an approximate value,
specify that it is so and, if possible, explain why)
Question
Answer
Example 1, Full scale frame: Series 190
1.
Would it be possible to manufacture the full scale frame example 1, Series 190, as
a single element using your publicly available machine?
40
2.
3.
4.
5.
6.
7.
1.
2.
3.
4.
Please do not consider in this question other feasibility/efficiency methods like
stocking up elements and arrangements of segments on a platform
If answer is YES on Q1, please specify:
Approx. time (in hours) to re/design the example for a specific manufacturing
process (e.g. additional supports due to the AM technology) (if applicable)
Approx. time (in hours) to prepare the STL file
(if applicable)
Average build time rate (cm3/h) and the build time of the whole frame
Average cost rate ($/cm3), cost in total or, if applicable, of specific steps
Average energy used (MJ/cm3), energy used in total or of specific steps
If answer is YES on Q1, please further elaborate if you believe that there is a
more efficient way for the frame production (e.g. using some other AM
technology, consider assembly systems etc.):
What would be in your opinion the largest segment size of the frame for the most
efficient and feasible use of the process or the machine?
Explain why a)
Which AM technology (or a hybrid system) would you recommend for the
specific task?
If answer is NO on Q1, please elaborate:
What is the largest segment size you could manufacture as a single element?
Explain why a)
Any other comment in your opinion about the specific task
Which steps are necessary to scale up the specific AM system (e.g. scientific
research, technology development, etc.)?
Approximately the necessary investment for d)
Pros and cons of your 850R for the task
Pros and cons of another recommended technology by you for the task (if
applicable)
Any additional comment
Question
Example 2, Full scale frame: Tower Bridge
Would it be possible to manufacture the full scale frame, example 2, Tower
Bridge, as a single element using your publicly available machine?
Please do not consider in this question other feasibility/efficiency methods like
stocking up elements and arrangements of segments on a platform
If answer is YES on Q1, please specify:
Approx. time (in hours) to design the example for a specific manufacturing
process (e.g. additional supports due to the AM technology) (if applicable)
Approx. time (in hours) to prepare the STL file
(if applicable)
Average build time rate (cm3/h) and the build time of the whole frame
Average cost rate ($/cm3), cost in total or, if applicable, of specific steps
Average energy used (MJ/cm3), energy used in total or of specific steps
If answer is YES on Q1, please further elaborate if you believe that there is a
more efficient way for the frame production (e.g. using some other AM
technologies, consider assembly systems etc.):
What would be in your opinion the largest segment size of the frame for the most
efficient and feasible use of the process or the machine?
Explain why a)
Which AM technology (or a hybrid system) would you recommend for the
specific task?
If answer is NO on Q1, please elaborate:
What is the largest segment size you could manufacture as a single element?
Explain why a)
Any other comment in your opinion about the specific task
41
Answer
5.
6.
7.
Which steps are necessary to scale up the specific AM system (e.g. scientific
research, technology development, investments, etc.)?
Approximately the necessary investment for d)
Pros and cons of your 850R for the task
Pros and cons of another recommended technology by you for the task (if
applicable)
Any additional comment
8.5 Calculate specific costs using scientific data
While waiting for the results of the questionnaire, I conducted a detailed scientific
literature review on how to calculate specific costs for additive manufacturing
[35][34][36][35][49] and then calculated the costs using scientific data (Table 7). For the
simplicity of calculation, I assumed that desktop size AM technologies like DMLS or
EBM could be extended to the size required for our specified CW frame.
Table 7: Cost comparison of standard vs. 3D printed CW frame after the first round of research
8.6 Conclusions after the first research round
After the first round of research, I concluded that AM technology is cost prohibitive, and
a suitable manufacturing process in full scale and appropriate material for CW frame
manufacturing does not exist.
R&D project:
Direction 1:
42
 Further development of most suitable existing manufacturing process;
 Approx. 1mil€ to develop from existing system;
 Approx. 1mil€ to develop facade and performance testing;
 Min two years of research and development;
 High risk of investment.
Direction 2:
 Development of a new process;
 Approx. 5mil€ to develop a new process;
 Approx. 1mil€ to prototype facade elements and test their performance;
 Project duration 3-5 years;
 High risk of investment;
 Possible partners: AM technology manufacturers, material and process developers,
research institutes;
 Possible source of funding is HORIZON 2020.
8.7 Semi-structured interviews
Since the questionnaire did not produce results, successful semi-structured interviews
were conducted. Table 8 shows exemplary notes taken during the interview with a Sciaky
representative. Previously calculated specific costs were updated and some specific costs
were calculated based on interview results from experts in the field. Table 9 shows final
cost comparison between standard and 3D printed CW frame.
Table 8: Exemplary notes taken during the interview with a Sciaky representative













Each layer deposition thickness: 1-10mm
height: 1-1,5mm
density of Al x volume= 38kg
extra metal for finishing to get to net shape – 1.5x38= 57kg
material cost: 25 dollars/pound ---- 125pounda
high production speed 10 lbs/h – 12.5 h for deposition
cost of deposition 3200
for material 620 dollars for wire
4000 dollars before any processing
possibly heat treatments in the oven – if more specimens in the oven: cheaper
1000 dollars /one treatment – in production somewhere around 500 to relief some stresses in metal... a few
$100,
altogether 7000 dollars
it would be designed differently – assuming that we are creating a design that is possibly to make with
Sciaky system
43
8.8 Final cost comparison: standard vs. 3D printed CW frame
Table 9: Cost Comparison: Standard CW vs. 3D Printed Frame
8.9 Conclusions about manufacturing specific costs
 Technologies with small scale platforms are unsuitable for further investigation
(manufacturers like EOS, Arcam).
 Hybrid technologies with large scale platforms are suitable for further investigation
(manufacturers like Sciaky Inc., DM3D, Optomec and Fabrisonic).
 The specific cost to 3D print CW frame (with the geometry of standard Series 190 CW
frame) is on average approx. 5,000.00€ (900€/m2).
 30% of the cost is material cost and 70% is the processing/printing cost (all
postprocessing steps included).
 Printing cost is mainly influenced by processing speed.
 Exemplary advice by AM manufacturer: Increase laser size to reduce processing speed
and printing cost.
Selected manufacturers of hybrid technologies (Figure 31) were further contacted to
gather information about the necessary investment to purchase and develop their systems
for Permasteelisa’s needs.
44
Figure 31: Most promising technologies/manufacturers to further discuss necessary investments to purchase and
develop their system for Permasteelisa’s needs
8.10 Description of chosen manufacturing processes
Sciaky’s [50](Figure 31, Figure 32) Electron Beam Additive Manufacturing (EBAM)
technology starts by building a 3D model from a CAD program. Afterwards, a fullyarticulated, moving electron beam gun deposits metal via wire feedstock, layer by layer,
until the part is built and ready for minor finish machining. Deposition rates typically
range from 7 to 20 lbs/h, depending upon part geometry and the material selected. The
EBAM package provides a controlled beam geometry that produces energy distribution
on the melt pool and the wire for great repeatable performance with preform fabrication.
Requiring little maintenance, the EBAM filaments can be changed out in 10 minutes at
the end or beginning of any chamber cycle. The largest build chamber available on the
market is 5x2,3x1,8m, which makes Sciaky’s technology the only candidate that matches
Permasteelisa’s requirements in terms of platform size and material.
DM3D’s [51] (Figure 31, Figure 32) laser-based AM technology called Direct Metal
Deposition (DMD) can create fully dense parts that are not sintered from various metals
and alloys. Additive DMDCAM software allows precision and quick toolpathing for part
build up. Five-axis deposition capability allows building features/flanges/bosses on
existing parts. Patented closed loop control results in better part quality, and no postdeposition sintering or metal infiltration is needed. DMD demonstrates significantly
higher build rate than Powder Bed Fusion (PBF) technologies in a wide variety of
materials, including steels, stainless steels, Stellite, Inconel, Titanium alloys, etc. Today
the technology is used mostly for manufacturing high value aerospace components, e.g.
45
turbine casings and components, structural components, seals, lattice structures, etc. and
for feature additions on existing components, such as flanges, bosses, etc. on casings. The
largest build platform currently available on the market is 3,6x 3,2m with the possibility
to rotate 360° around the z axis. According to the company’s CEO, development of this
system to meet Permasteelisa’s requirements would take approximately six months.
Optomec’s [52] (Figure 31, Figure 32) LENS systems use a high-power laser, 500W to
4kW, to fuse powdered metals into fully dense 3D structures. The LENS 3D printer uses
the geometric information contained in a CAD solid model to automatically drive the
LENS process as it builds up a component layer by layer. Additional software and
closed-loop process control ensures the geometric and mechanical integrity of the
completed part. It is housed in a hermetically-sealed chamber which is purged with argon
so that the oxygen and moisture levels stay below 10 parts per million. This keeps the
part clean, preventing oxidation. The metal powder feedstock is delivered to the material
deposition head by Optomec’s proprietary powder-feed system, which is able to precisely
regulate mass flow. Once a single layer has been deposited, the material deposition head
moves on to the next layer. By building up successive layers, the whole part is
constructed. When complete, the component is removed and can be heat-treated, HotIsostatic-Pressed, machined, or finished in any customary manner. LENS 850-R system
has the largest build chamber in size1,6mx0,9mx0,9m. To create a building chamber of
the required size to manufacture the Permasteelisa’s CW frame as a single element,
Optomec’s representative proposed development of a hybrid solution: their modular
LENS Print Engine technology integration with a robotic system/arm. LENS Print Engine
technology could be integrated with other metal working platforms such as mills, lathes,
robots, custom gantries, or table system. It simplifies metal fabrication applications such
as net shape rapid prototyping, hybrid manufacturing, full production, in-situ repair,
manufacturing rework and more.
Fabrisonic [53] (Figure 31, Figure 32) provides 3D metal printing services in a wide
range of metals through low temperature Ultrasonic Additive Manufacturing (UAM)
technology. Harnessed sound waves merge layers of metal foil in a process that requires
no melting. Solid state welding technology enables printing of dissimilar metal laminates
in one part without changing the metal properties/cladding and printing custom designed
metal matrix composites. Low-temperature process enables embedding electronics and
sensors in 100% dense metal structures and integration of smart materials for reactive
structures. The largest available build chamber that this system provides is 1,8x1,8x0,9m,
but according to the representative, the build chamber’s size could easily be expanded to
the size Permasteelisa requires.
46
Figure 32: Hybrid technologies of selected manufacturers and system’s estimated costs (according to manufacturers)
9
Final conclusions
 Today’s hybrid systems with large-scale platforms demonstrate potential for
fabrication of full scale CW frame as a single element.
 Manufacturing is still too expensive in comparison to conventional methods.
 Processing speed is still too slow.
 Sciaky’s hybrid technology is the only ready-to-use additive manufacturing process
capable of manufacturing a full scale CW frame as a single element.
Further research and development recommendations’ depending on the Permasteelisa
Board’s decision about the type and the amount of investment in the future research the
key next steps are:
 Develop the most suitable AM process for CW frame manufacturing:
(1) Establish team of experts (research institute, e.g. University of Nottingham,
selected manufacturer, etc.);
(2) decrease the fabrication cost by e.g. increasing the processing speed;
(3) Establish new design process (re-design the CW frame to be suitable for additive
manufacturing, material design, new software, optimization and simulation
included in the process, etc.);
(4) monitor the design process/change the standard routine of design process in
Permasteelisa;
(5) manufacture prototypes and test their performance with quality management;
47
 if the final results of the R&D group’s research are satisfactory, prepare further
detailed feasibility studies about implementing a new design process and AM
technology as routine systems in standardized Permasteelisa’s design and
manufacturing process;
 further investment and marketing prediction studies are required;
 3D Printing of large metallic building components is technologically almost feasible
today but it is cost prohibitive; develop a framework for practitioners to evaluate
when to chose among conventional, standard AM and hybrid metal deposition
systems to cost efficiently construct a specific metallic building component.
10 References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
“Apple Spaceship Campus Reportedly $2 Billion Over Budget Before Construction Has
Begun,”
Architizer.
[Online].
Available:
http://www.architizer.com/en_us/blog/dyn/82146/apple-spaceship-campus-reportedly-2billion-overbudget-before-construction-hasbegun/?fb_action_ids=164541013705307&fb_action_types=og.likes&fb_source=aggregati
on&fb_aggregation_id=288381481237582#.UXCHC8oRWRd. [Accessed: 18-Apr-2013].
“Permasteelisa Group.” [Online]. Available: http://www.permasteelisagroup.com/.
[Accessed: 30-Mar-2014].
dr. Martin Baumers and dr. Chris Tuck, “3D Printing A Differentiation Engine.” .
“Committee F42 on Additive Manufacturing Technologies.” [Online]. Available:
http://www.astm.org/COMMITTEE/F42.htm. [Accessed: 16-May-2015].
R. A. Buswell, A. Thorpe, R. C. Soar, and A. G. F. Gibb, “Design, data and process issues
for mega-scale rapid manufacturing machines used for construction,” Autom. Constr., vol.
17, no. 8, pp. 923–929, Nov. 2008.
J. Newman, “The Art of Simplicity: Standardized Process Terminology,” Rapid Ready
Technology. [Online]. Available: http://www.rapidreadytech.com/2012/06/the-art-ofsimplicity-standardized-process-terminology/. [Accessed: 06-Feb-2015].
I. Gibson, D. W. Rosen, and B. Stucker, Additive Manufacturing Technologies: Rapid
Prototyping to Direct Digital Manufacturing, 2010 edition. New York: Springer, 2009.
“3D Printing Market to Quadruple to $12 Billion in 2025.” [Online]. Available:
http://www.luxresearchinc.com/news-and-events/press-releases/read/3d-printing-marketquadruple-12-billion-2025. [Accessed: 16-May-2015].
Rupert Soar and Farid Fouchal, “Construction Processes for the Digital ‘Trinity,’” The 1 st
International Conference On Industrialised, In tegrated, Intelligent Construction. 14-16
May 2008.
John Hornick, “3D Printing and the future (or Demise) or Intellectual Property
Inside3DPrinting_johnhornick,” 19:07:48 UTC.
Chuck Alexander, “Industrial Scale Additive manufacturing Technologies Inside3 d
printing_chuckalexander,” 16:25:49 UTC.
Prof. Brent Stucker, “Reshaping Manufacturing: Understanding 3D Printing Processes
Inside3DPrinting_BrentStuckerWorkshop,” 18:30:30 UTC.
“A Framework for Bringing 3D Printing into the Construction Industry | Center for
Integrated
Facility
Engineering.”
[Online].
Available:
http://cife.stanford.edu/Frameworkfor3DPrinting. [Accessed: 31-Mar-2014].
Sunny Web, “Digital Supply Chain Inside3DPrinting_SunnyWebb,” 18:30:36 UTC.
“Technology Research | Gartner Inc.” .
48
[16] “Gartner’s Hyper Cycle For 3D Printing Predicts Enterprise & Medical Application
Growth,”
Forbes.
[Online].
Available:
http://www.forbes.com/sites/louiscolumbus/2014/08/19/gartners-hyper-cycle-for-3dprinting-predicts-enterprise-medical-application-growth/. [Accessed: 08-Feb-2015].
[17] N. Mrazovic and M. Baumers, “3D printing opens up new horizons in manufacturing,”
UKSPA Innovation into Success. [Online]. Available: http://www.scienceparks.co.uk/.
[Accessed: 08-Feb-2015].
[18] “Nanoscale 3D-printed Microstructures by Nanoscribe,” Simotron. .
[19] “EPSRC Centre for Innovative Manufacturing in Additive Manufacturing - EPSRC.” .
[20] R. A. Buswell, R. C. Soar, A. G. F. Gibb, and A. Thorpe, “Freeform Construction: Megascale Rapid Manufacturing for construction,” Autom. Constr., vol. 16, no. 2, pp. 224–231,
Mar. 2007.
[21] R. Soar, “Freeform Construction: the next phase of digital integration or a bridge too far?,”
DPR, 11-Nov-2013.
[22] “3DPRINTCANALHOUSE
by
DUS
Architects.”
[Online].
Available:
http://3dprintcanalhouse.com/. [Accessed: 30-Mar-2014].
[23] 3D Print Canal House. 2013.
[24] “Will This Möbius Strip House Be The First 3-D Printed Building? | Co.Design | business +
design.” [Online]. Available: http://www.fastcodesign.com/1671713/will-this-m-bius-striphouse-be-the-first-3-d-printed-building. [Accessed: 30-Mar-2014].
[25] Contour Crafting: Automated Construction: Behrokh Khoshnevis at TEDxOjai. 2012.
[26] B. Khoshnevis, “AUTOMATED CONSTRUCTION BY CONTOUR CRAFTING –
RELATED ROBOTICS AND INFORMATION T ECHNOLOGIES,” J. Autom. Constr. –
Spec. Issue, Jan. 2004.
[27] “A Framework for Bringing 3D Printing into the Construction Industry | Center for
Integrated Facility Engineering.” .
[28] Behrok Khoshnevis, “ISRU-based Robotic Construction Technologies for Lunar and
Martian Infrastructures,” CEE Stanford, Jan-2014.
[29] “SpiderFab_SpringNIAC_v6.pptx - SpiderFab_SpringNIAC_v6.pdf.” .
[30] “AM Envelope. The potential of Additive Manufacturing for facade constructions | Strauss |
A+BE | Architecture and the Built Environment.” .
[31] “Arup uses 3D printing to create structural steel components.” [Online]. Available:
http://www.gizmag.com/arup-laser-sintering-construction/32457/. [Accessed: 08-Feb2015].
[32] “Construction steelwork makes its 3D printing premiere | Arup | A global firm of consulting
engineers, designers, planners and project managers.” .
[33] “ATKINS PROJECT: Additive Manufacturing; A Low Carbon Footprint,” May 2012.
[34] M. Baumers, C. Tuck, P. M. Dickens, and R. Hague, “How can Material Jetting
SystemsbBe Upgraded for more Efficient Multi-Material Additive Manufacturing?”
[35] M. Baumers, “Production cost, machine productivity and the emergence of an Additive
Manufacturing industry,” Publ. Yet.
[36] M. Baumers, C. Tuck, R. Hague, I. Ashcroft, and R. Wildman, “A Comparative Study Of
Metallic Additive Manufacturing Power Consumption,” 2010.
[37] M. Baumers, C. Tuck, R. Wildman, I. Ashcroft, and R. Hague, “Energy Inputs To Additive
Manufacturing: Does Capacity Utilization Matter?,” 2011.
[38] “Florea, Designer of the 3D Printed House & Gran Torino has More up his Sleeve – Is a
Submarine Next? - 3DPrint.com.” .
[39] “QuietYellTM » A Designed World: 3D Printed Sculptural Painting Wall Art by Ioan
Florea.” .
[40] “Design from a 3D printer.” .
49
[41] Permasteelisa, “Members Made of GRP as Architectural Feature in Modern Curtain
Walling for Shading Elements on Walbrook Bank in London,” 2009.
[42] Prof. Brent Stucker, “Simulation of Additive Manufacturing Technologies: Enabling a 3D
‘Print Preview’ Inside3DPrinting_BrentStucker,” 18:30:26 UTC.
[43] Prof. D.L.Bourell, “Materials for 3D Printing Inside3DPrinting_DavidBourell,” 15:40:40
UTC.
[44] “Gramazio Kohler Architects ETH SIA BSA.” .
[45] “Prof Fabio GRAMAZIO | Future Cities Laboratory.” .
[46] H. Strauss, “AM Envelope. The potential of Additive Manufacturing for facade
constructions,” BE Archit. Built Environ., vol. 3, no. 1, pp. 1–270, Jan. 2013.
[47] “Rhinoceros.” [Online]. Available: https://www.rhino3d.com/. [Accessed: 14-May-2015].
[48] “Microsoft
Word
Manufacturing
Engineering-LENS.doc
Optomec_Manufacturing_Engineering_LENS.pdf.” .
[49] M. Baumers, “Economic aspects of additive manufacturing: benefits, costs and energy
consumption,” Loughborough University, UK.
[50] “Sciaky, Inc. | Electron Beam Welding | Additive Manufacturing | 3D Printing.” [Online].
Available: http://www.sciaky.com/. [Accessed: 11-Feb-2015].
[51] “DMD3D Technology.” [Online]. Available: http://www.dm3dtech.com/. [Accessed: 11Feb-2015].
[52] “3D Printing Electronics Laser Additive Manufacturing Systems.” [Online]. Available:
http://www.optomec.com/. [Accessed: 11-Feb-2015].
[53] “Fabrisonic | Sound 3D Printing.” [Online]. Available: http://fabrisonic.com/. [Accessed:
11-Feb-2015].
50