Additive Manufacturing: Jolie Frketic, Tarik Dickens, Subramanian Ramakrishnan

Additive Manufacturing 14 (2017) 69–86
Contents lists available at ScienceDirect
Additive Manufacturing
journal homepage: www.elsevier.com/locate/addma
Review
Automated manufacturing and processing of fiber-reinforced polymer

(FRP) composites: An additive review of contemporary and modern
techniques for advanced materials manufacturing
Jolie Frketic a , Tarik Dickens a,∗ , Subramanian Ramakrishnan b
a
High Performance Materials Institute, Florida A&M University-Florida State University College of Engineering, 2525 Pottsdamer St., Tallahassee, FL 32310,
United States
b
Department of Chemical and Biomedical Engineering, Florida A&M University-Florida State University College of Engineering, 2525 Pottsdamer St.,
Tallahassee, FL 32310, United States
a r t i c l e i n f o a b s t r a c t
Article history: High throughput automated techniques are nowadays playing a key role in polymer composite manufac-
Received 14 March 2016 turing in a number of industries such as automotive and aerospace. There is a need to produce high volume
Received in revised form 13 January 2017 parts efficiently. Automated manufacturing methods such as automated tape layup and automated fiber
Accepted 22 January 2017
placement can produce composite parts efficiently, and with the advent of additive manufacturing the
Available online 25 January 2017
complexity of these components are increasing. This paper will review contemporary composite man-
ufacturing methods filament winding, automated tape layup, and automated fiber placement, and the
Keywords:
newer automation techniques of robotic pick-and-place and continuous tow shearing. It also addresses
Polymer matrix composites (PMC) (A)
Additive manufacturing
recent advances in composite additive manufacturing using vat photopolymerization, binder jetting,
Automation (E) material extrusion, sheet lamination and powder bed fusion. Methods, materials and testing results of
the manufactured components will be discussed.
© 2017 Elsevier B.V. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2. Automation of composite manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.1. Industrial automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.1.1. Filament winding (FW) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
2.1.2. Automated tape layup (ATL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.1.3. Automated fiber placement (AFP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
2.2. Lab-Scale automation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.2.1. Robotic pick and place techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.2.2. Continuous tow shearing (CTS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
2.3. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3. Additive manufacturing: polymer composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1. Vat photopolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.2. Binder jetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.3. Material extrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.3.1. Thermoplastic material extrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.3.2. Thermoset material extrusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.3.3. Material extrusion summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.4. Sheet lamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.5. Powder bed fusion (PBF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
∗ Corresponding author.
E-mail address: dickens@eng.fsu.edu (T. Dickens).
http://dx.doi.org/10.1016/j.addma.2017.01.003
2214-8604/© 2017 Elsevier B.V. All rights reserved.
70 J. Frketic et al. / Additive Manufacturing 14 (2017) 69–86
4. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
1. Introduction ination, and powder bed fusion of fiber reinforced polymer parts
[15,21–25].
Additive manufacturing (AM), or three dimensional (3D) print-
ing, has received great interest in the last few years and is known 2. Automation of composite manufacturing
for rapid prototyping. It has continued to grow and is transi-
tioning into a reliable way of creating diverse objects such as Automation is the mechanization of work to reduce human-
dresses, jet engines and turbine blades [1,2]. Additive manufac- machine interaction [26]. Historically, the Industrial Revolution
turing commonly uses plastics to create parts layer by layer, but introduced ideas of standardization and value-added processing
has progressed to using ceramics, paper, and even polymer matrix to meet increasing production volumes. The metalwork of the
composites. More recently, Oak Ridge National Laboratory and twentieth century has given rise to ever-increasing proficiency of
Cincinnati INC. printed parts to recreate the Shelby Cobra out of polymer-based reinforced composites in the twenty-first century.
carbon fiber reinforced ABS using their Big Area Additive Man- If composites could be manufactured as easily as metal parts, then
ufacturing (BAAM) machine. This achievement has expanded the polymer composite materials stand to replace traditional materials
definition of composite printing with the inclusion of reinforcing in various applications where environmental, structural and weight
material for strengthening the filament [3–5]. are significant concerns.
Composites are used for many applications in the aerospace and Though there are many methods to create composites, not all
automotive industry due to their high specific strength. Traditional are suited for high volume output automation. For example, hand
manufacturing methods for composites are time consuming and layups are useful for making small numbers of specialty parts, and
labor intensive, which leads to high costs. For manual layups, the the automation of which would not be cost effective. Injection
labor cost associated with creating composites takes a large por- molding is good for a large number of parts, but customization of
tion of the total manufacturing cost [6]. A highly trained technician the parts is costly due to the need for dedicated molds. The opti-
creates about 2.5 lbs. of composite material per hour, but even mum production capacity for composite manufacturing methods
the best can make a mistake [7]. Human error can introduce voids given the increasing level of automation as seen in Fig. 1 [8].
and irregularities into the composite part during production, which The ability to quickly and precisely manufacture metal parts
negatively affects mechanical properties, making parts unusable for comes from the ability to use computer numerically controlled
standard operation. For those reasons, manufacturers, especially (CNC) machines for subtractive manufacturing methods. The man-
in western Europe and the USA, are transitioning to automated ufacturing of composites lends itself to additive manufacturing
composite manufacturing to improve process and cost efficiency because of the use of polymers, which are easily shaped through
[8–10]. a variety of processing steps. Low-level additive manufacturing
Three popular mechanized composite manufacturing tech- techniques such as ATL, AFP, and FW use a computer-aided design
niques used in industry are automated tape layup (ATL), automated and/or computer-aided manufacturing model to build up a part, in
fiber placement (AFP), and filament winding (FW). These processes a layer by layer process, which is the basis for the additive man-
have been heavily researched, but their use is limited by the cost of ufacturing that is known today [27]. As a result, AFP, ATL and FW
the specialized machinery and constraints in fabrication of complex have become the industrial benchmarks newly developed additive
parts. However, the composite strength matches that of tradi- processes for composites are evaluated against. In summary, these
tional manufacturing techniques, making the processes beneficial high output methods are considered additive in nature because of
in speeding up production and cutting labor costs. Evidence exists their use of a layer wise process, where material is deposited in a
in the aerospace industry which has reported a 70–85% reduction layer-wise fashion using CNC machines.
in man hours due to the implementation of automated composite
tape layers [9]. 2.1. Industrial automation
Additive manufacturing is currently on the rise in industries
such as the biomedical, aerospace, and engineering fields, because Currently, the methods closest to additive manufacturing in
it offers prototyped parts quickly and inexpensively without the industrial composite production are ATL, AFP, and FW. These make
need for specialized tooling [11–14]. One of the drawbacks of use of laminated tapes or smaller laminated tows to create a part.
additive manufacturing is the low strength of the polymer and ATL and AFP are CNC machines with programmable axis move-
weakly bonded (i.e. green ceramic) parts [15,16]. A way of increas- ments, axis drives, and delivery heads that meticulously place
ing the strength is by using fillers and fiber reinforcements. Fillers composite tapes and tows to fabricate a specific part structure [28].
have been utilized as crack bridging and toughening agents in All of these processes make use of resin-impregnated continuous
polymer systems for some time. For example: Stratasys Inc., incor- fibers, also known as prepregs [7].
porates millimeter-sized glass whiskers in their specialty filament
for material extrusion printing [17]. In another example, FiberForge 2.1.1. Filament winding (FW)
Inc., a 3D printing company, created hardware to print carbon fiber FW is an automatic method to create axisymmetric as well as
filament for a 6 × 6” build tray [18]. some non-axisymmetric parts (i.e. pipe bends) by winding contin-
This review paper studies the current automated additive man- uous prepreg sheets, rovings, and monofilaments around a rotating
ufacturing techniques for composites in industry such as filament mandrel before curing [10]. For non-axisymmetric parts, 6 or 7 axis
winding (FW), automated tape layup (ATL), automated fiber place- machines, like those manufactured by CNC Technics, are necessary
ment (AFP), as well as the newer lab-scale automation techniques of [29–31]. Filament winding may use either dry fibers passed through
robotic pick and place and continuous tow shearing (CTS) [6,19,20]. a resin bath (wet winding) and prepreg materials (dry winding) for
Also discussed is additive manufacturing technology such as vat production [30]. The basic process of filament winding is displayed
photopolymerization, binder jetting, material extrusion, sheet lam- in Fig. 2. The first step in wet winding is gathering fibers from a
J. Frketic et al. / Additive Manufacturing 14 (2017) 69–86 71
Fig. 1. Ranking of composite part manufacturing processes by optimum output [8].
Fig. 2. Filament winding flow chart and diagram [30].
series of creels, grouping the fibers by passing them through a tex- different winding patterns such as hoop, helical, and polar wind-
tile board or comb, and pulling the grouped fibers through a resin ing can be formed this way. In CNC FW, the mandrel and carriage
bath. The resin bath contains liquid resin and catalyst, but may are able to move independently of each other, allowing for infinite
include additional ingredients such as pigments or UV absorbers winding possibilities; this allows production of complex shapes
[32]. Upon exiting the resin bath, the rovings are pulled through a with no axis of symmetry, such as helicopter blades [32].
wiper system, such as squeeze rollers, which controls the amount After the desired thickness of the composite part has been
of resin on the fibers. Fiber guides keep consistent tension on the reached, the mandrel is cured. After curing, the mandrel can be
fibers during the process. Once the rovings are impregnated, they removed from the part. FW can be used for parts such as large pres-
pass through a comb, straight bar, or ring and become a flat band sure vessels, as well as parts as small as pipes depending on the size
of fibers. At this point, the flat band (or prepreg material for dry of the machine [30]. For larger parts, mandrels may be collapsi-
winding) is positioned onto the mandrel, and a carriage system ble (inflatable or segmented) for easy extraction from the cured
travels back and forth to wind the fibers around the mandrel while part. For small part volumes, soluble plasters, eutectic salts, and low
the mandrel spins. The speeds of the carriage and the mandrel are melting alloys can be used to create the mandrel. Current research
adjusted to match the desired winding pattern for the part. Several is looking into the ability of using FW to join parts together, as well
A multi-axis articulating robot can be used to hold the ATL head,

or it can be suspended above the tool on a gantry. When used with
a gantry, the ATL head is able to have up to 10 ◦ of freedom, with 5
axes on the head and 5 on the gantry [9]. The first step to compos-
ite production is depositing a starting amount of prepreg tape onto
the mold using a soft silicone roller. After this is done, the machine
deposits the tape according to CNC paths defining the part geome-
try. At the end of layup, the tape is cut automatically by rotating or
pinching blades. The speed at which the machine does this is vari-
able depending on the complexity of the part being created. This is
a similar process to traditional AM technologies, where material is
deposited in accordance with a defined CNC path.
For large noncomplex (flat or single curvature) parts, ATL is
an efficient way to produce composite structures [39]. To create
Fig. 3. ATL head diagram showing tape laying over a curved mold [40].
non-flat structures, a second forming step is used to shape the ATL
laminate sheet [27]. The biggest problem with this is that the lami-
nate has a tendency to warp and buckle during the forming process
as the ability to miniaturize the technology to create bench top if the curvature is too large [19,39,41]. Analysis has shown that the
sized FW machines [10,33]. The use of programmable robots shows two major components that influence buckling are the tackiness of
promise as a way to make filament winding obtainable to startup the prepreg and the tow width [39]. The way to combat buckling
companies [10,31]. Companies currently using FW include Airborne is to enforce a minimum turning radius of the machine depending
Composites BV (Netherlands) [10], MF Tech (France), Spectrum on the material and tow width being used [42]. As shown in Fig. 4,
Aeronautical LLC (USA), and Rocky Mountain Composites (USA), MTorres has been developing smaller ATL machines that are able
with the machines produced in the USA, France, Germany, the to lay down 3 inch tapes, as well as a flat laminate machine based
United Kingdom and Macedonia [8,10]. off of small robots [43]. This shows the progress towards smaller
For a quality composite part, the rovings must be in constant scale ATL projects, making it more accessible to smaller companies
tension during application to provide good fiber collimation and [44].
reduce sagging. Voids are a common defect in FW parts; Rousseau
et al. attributed the voids to the undulations that occur when a tow 2.1.3. Automated fiber placement (AFP)
goes over and then under previously laid tows [34]. Micro-cracks AFP places 2–32 individual tows or ribbons of slit tape using low
can also form because of the resin rich nature of the undulating tension and compaction pressure [45]. This method can be used
areas [34]. Voids can also come from bubbles in the resin bath with both gantry style and robotic setups that hold the AFP head
when there is improper wetting of the fibers, and there is failure which are shown in Figs. 5 and 6 [45]. The material is typically
to uniformly wipe away excess resin [35,36]. Delaminations and 0.125 in wide, and can be made from both thermoset and thermo-
fiber wrinkles are also found in FW parts. The bonding of dry layers plastic material. To deposit the thermoset-based prepreg tows, the
to previously wound layers can lead to improper adhesion where AFP machine must be equipped with a cooled storage and rollers
delamination can occur [37]. Wrinkling of the fibers due to poor that are coated with a release agent [45]. Thermoplastic tows do
winding tension and misalignment within the rovings can cause not require the cooled storage, but the cutting and placement tools
defects in the final part [38]. must be hardened to withstand the processing the harder type of
plastic [45].
The AFP process makes use of individual tow control, or the
2.1.2. Automated tape layup (ATL) ability to stop, cut and restart each tow during placement [28].
ATL was developed over 25 years ago. ATL machines lay a Individual tow control means each tow can be laid at its own speed,
prepreg tape or continuous fabric strips with widths ranging from allowing for material to be deposited along a radius. By stopping
75 to 300 mm onto a flat surface in various orientations [7,27,39]. individual tows, the width of the tape can also be changed, which
Both thermoset and thermoplastic materials may be used in ATL. allows variable width parts to be created.
Creels of prepreg tape are stored near or on the head of the layup The major limitation of possible AFP ply shape is the shortest tow
machine for use during production. The head of the ATL machine length the machine can lay down, called minimum cut length [7].
usually consists of spools of tape, a winder, winder guides, a com- A major issue of these using the AFP processes is the cost involved
paction shoe or roller, tape cutter, and a position sensor as shown in acquiring the equipment necessary. Several manufacturers are
in Fig. 3 [7,40]. moving towards using robotic arm style AFP machines to reduce
Fig. 4. a) MTorres robotic arm ATL machines b) Small gantry style ATL machine (MTorres) [43].
sists of a computer-aided design program and their robotic arm

fiber cutter machine [50].
2.2.2. Continuous tow shearing (CTS)

CTS modifies AFP technology to produce tighter curvature with-
out gaps, cuts, and overlaps in the tow placement [41]. The main
differences in CTS compared to AFP are the materials being used
during layup and the path layout [19,41].
In CTS, a specially designed head takes dry carbon fiber tows
and applies a layer of resin film to one side, as seen in Fig. 7 [19].
This semi-impregnated tow is then applied to the mold via a modi-
fied CNC ply cutting machine. This results in a material comparable
to commercially available prepreg materials, with a random void
distribution [41]. The error found in the curvature of the part com-
pared to the path was at maximum 5◦ but was on average 2◦ [41].
CTS is able to approach curvature values more than 10 times that
of conventional AFP due to the ability to place dry tows [41]. How-
ever, inaccuracy in some of the paths and slight tow bending makes
Fig. 5. Gantry style AFP [45].
it still a design in progress.
2.3. Conclusion
Composite automation is used in various industries. The

automation process necessitates simple part design but the abil-
ity to change the design quickly is not without major cost. Because
of this design complexity, composite materials cannot be used to
their full extent and benefit, which is true for any manual lay-up
and assembly. As Timothy Gutowski said, “The challenge that faces
industry now is to develop flexible methods of automation that
can significantly expand the window of shape and microstructure
available to the designer at reasonable cost” [51]. ATL and AFP are
working towards this flexibility, and there are also ways of textile
weaving that can allow for higher degrees of freedom during fabri-
cation. The fiber layup mechanism and the fiber −to −wetting- to
−solidification process are two issues that need to be resolved to
create better automated processes for functional composites. [51].
To create a process that could fully encapsulate these two issues
quickly would allow for better automated process for composites.
AFP, ATL and FW are industrial benchmarks for automated
methods to evaluate developing processes. Research on automated
methods for creating thermoset composites has been over the span
Fig. 6. Robotic style AFP [45].
of the time line, as seen in Fig. 8. The number of published papers
on AFP and ATL has steadily risen over the past few years, while
the number of papers on additive manufacturing of composites has
the cost of these systems, like the machines by Coriolis Compos- recently reached the same level as filament winding.
ites, Accudyne Systems, Electroimpact, and Automated Dynamics,
which also boasts a small scale lab set up for research purposes,
which is about the size of a CNC mill [46–49]. 3. Additive manufacturing: polymer composites
Many methods of additive manufacturing (AM) exist [52], each

2.2. Lab-Scale automation suited to different materials and different needs. Applications range
from printing new organs [53] and cranioplasty [54], to creating
2.2.1. Robotic pick and place techniques lightweight parts [55], fuel cells [56] and even art [57]. Parts cre-
Angerer et al. proposed a robotic pick and place technique which ated from additive manufacturing can be complex and customized,
uses a robotic arm, fitted with different end effectors, to cut and without the need for expensive mold redesign as would be needed
place prepreg or dry fabric to create composite laminates [6,20]. for injection molding. The main issue with additive manufacturing
Robotic pick and place techniques are believed to be able to replace is the lack of materials with structural capabilities. Current mate-
the previously mentioned AFL and AFP systems [6]. This method is rials typically are not robust enough to be used as an end-product
less mature than AFL and AFP, and no full scale production method or component [58]. Additionally, additive manufacturing has the
has been used [6]. The current issue with this method is the time it constraint of working “in-the-box”, and is constrained to the size
takes to perform the task, where the most time intensive process of the print area [59]. The average size of a large industrial printer
is the removal of the paper backing from the prepreg [6]. To com- is around 1 m3 , although recently a build area of 26 m3 has been
bat this problem, different end effectors must be developed that implemented at Oak Ridge National Laboratory using their Big Area
can efficiently cut and remove prepreg backing, and arrange the Additive Manufacturing machine [60].
prepreg material within the desired mold. Assyst Bullmer in the Recently, attention has been focused on the ability of additive
UK is currently selling a composite manufacturing cell that con- manufacturing to create structural parts for the use in building
Fig. 7. CTS machine set-up [19].
Fig. 8. Number of papers considering FW, AFP, ATL, and Additive Manufacturing in 4 year intervals (data from Google Scholar).
end-product components [61–63]. Composite material in partic- stereolithography (SLA). Many variations of this process are being
ular suits this process, since most AM processes already use the used today in makers such as Polyjet, as well as several other indus-
matrix polymers (i.e. thermosets and thermoplastics) that are the trial printing systems [66].
key components of fiber reinforced composites [64]. From creat- The basis of vat photopolymerization systems come from
ing thermoplastic filament preloaded with fibers [24] to mixing depositing photo-polymer by either wiping resin over the part [67],
fibers with powdered plastic base of some processes [22], research depositing resin via inkjet method [68], or incrementally submerg-
is heading towards quickly realizing a fully printable, functional ing the part into a vat of photopolymer. This additional layer is cured
composite part. The below sections (1) review the capabilities of using photo-polymerization into a solid polymer. SLA typically uses
various AM composite methods and (2) provide insight into the UV light for curing, but some systems are able to use visible light
composite micro-structure of layered prints. [12]. Projection stereolithography uses digital light processors and
is able to create slices without scanning by using a mask [69]. This
reduces the fabrication time greatly while still allowing high feature
3.1. Vat photopolymerization
resolution [70].
Most of the existing SLA systems implement a bottom up
Liquid-based additive manufacturing systems started with
approach, scanning each layer, then subsequently moving the base
Chuck Hull in the mid 1980 s, when he created a part using a
downwards and wiping an additional layer of polymer on top
laser and ultra violet light (UV)-curable materials. The company 3D
before again photo-polymerizing [11,12,69,70]. In addition, sev-
Systems used this knowledge to sell machines that could “rapidly
eral top down approaches have also been created. Fig. 9 depicts
prototype” parts [65]. This method is termed vat photopolymer-
both approaches [11]. The main criticism of top down approaches
ization by ASTM F2792 standards, but may also be referred to as
Fig. 9. Parts created using the bottom up approach of stereolithography on the left, and the top-down approach shown right [11].
Fig. 10. Rapid layer composite manufacturing system [67].
is that each time a layer is polymerized, extra stress is exerted on the fiber volume fraction of 20% is the upper limit for this method, due
part in order to detach it from the bottom of the dish. However, this to increased viscosity with increasing fiber content causing issues
stress has said to have a negligible effect on the part, and the prod- with the application of thin layers [73]. One way to overcome the
ucts made in this manner have smoother surfaces, use less material high viscosity is to deposit the resin fiber mixture with a nozzle,
and are not as oxygenated as their “bottom up” counterparts [11]. before going over the deposited layer with a wiping device that
A new method called Continuous Layer Interface Production smooths the surface [67,74,75]. Fig. 10 is an example of the method
(CLIP) was created to alleviate the problem of detaching from the that has been used with discontinuous fibers, allowing easy depo-
vat surface due to a highly oxygenated layer that inhibits resin sition of the fiber-doped resin [67]. Fig. 10(a) shows the bottom
curing [71]. Designed as a layer scanning method, resin is replen- platform of the vat moving down an incremental amount. Fig. 10(b)
ished as the cured layers are pulled up out of the vat, due to shows a nozzle depositing the fiber-doped resin on top of com-
the suction created by the part moving upwards. The dead zone pleted layers. A wiper then smooths the top of the deposited resin
(highly oxygenated zone where resin cannot cure) thickness is con- in Fig. 10(c), before a laser cures the corresponding layer pattern
trolled by the product of the photo initiator concentration (Dc0 ), the in Fig. 10(d). To incorporate a woven mat of fibers, the woven fiber
wavelength dependent absorptivity (˛PI ,), the number of incident mat is placed on a bottom layer of solid resin, then coated with resin
photons at the imaging plane per area time (˚0 ), and a proportion- as in other SLA methods. A problem with this procedure is that the
ality constant (C) following the function given in Eq. (1). wiper could shift the fiber mat during printing [72].
˚ ˛ −0.5 Fiber-reinforced micro-parts with aligned ferromagnetic fibers
0 PI can be created using a similar method, then magnetically aligning
DeadZoneThickness = C (1)
DC0 the fibers [76]. Using this method, fiber orientation can be changed
The creation of the dead zone allows for the ability to use a bot- within different layers of the part, depending on the magnetic field,
tom up approach to SLA where the resin replenishes the print zone which leads to highly customizable parts in terms of direction
automatically due to the infill of resin as the solidified layer moves of anisotropic properties like strength. However, the mechanical
upwards. The inventors claim this process is up to a hundred times properties and interfacial strength of the method have yet to be
faster than what is currently commercially available. The process tested [76].
is faster because the resin replenishing and UV exposure are con- Fig. 11 shows the vat photopolymerization composite with the
tinuous rather than a series of discrete steps, making the resin cure highest strength (72 MPa), formed with special addition of inter-
rate and viscosity limiting factors. The continuous layer creation layer bonding sites [74]. The idea behind interlayer bonding is that
process yields parts with high resolution. small sections of composite are left uncured, such that when the
Vat photopolymerization can create composite parts from next layer is deposited, some of the fibers are able to transcend the
chopped, woven, or continuous fibers. The creation of composite layer boundary, thus causing better bonding between the layers
parts consists of submerging the reinforcing fibers in the UV- [74]. It has been theorized that the ultimate strength of the com-
curable resin, then curing the resin [72]. This method to make posite is less affected by the addition of fibers than the modulus,
composites with discontinuous fibers is only slightly modified from because the weakest point in a printed composite is the interlayer
the original wiper method: a mixer is added to keep the fibers from bonding [75]. Since fibers do not typically transcend layers, the
settling during the cure process [73]. It has been reported that a strength between layers is purely due to the matrix interface.
Fig. 11. Cross section of SLA composite created with rivets [75].
Work has been done by Zak et al. to model the modulus and
strength of the composite using the distribution of the orienta-
tion of the fibers, leading to a modified rule of mixtures (ROM)
for printed composites [74]. The modified ROM takes into account
non-alignment of fibers during printing. The modulus can thus be
found via Eq. (2),
Ec = ␹1 2 vf Ef + vm Em , 0 < 1 < 1 (2)
where ␹1 is orientation efficiency factor and is estimated by Eq. (3),

␹2 (Eq. (4)) and ␹4 (Eq. (6)) are fiber length distribution correctors,
and ␹3 (Eq. (5)) is the orientation efficiency factor [74]
n
E t
k=1 k k
1 = n (3)
E1 t
k=1 k Fig. 12. Cross section of 20% GFRP fabricated via SLA [73].

m
2 = Psk (1 − tanh (nsk ) /nsk ) (4)

k=1
n
t
k=1 uk k
3 = n (5)
u1 t
k=1 k

m sc

m
4 = Pslk 1− + Pssk sc /2sk (6)
2sk
k=1 k=1
In these equations, tk is thickness of the kth lamina panel, ␾ is the

angle at which the fibers are aligned in respect to the z direction,
s is the aspect ratio of the fibers of length l, sc is the critical aspect Fig. 13. Distribution of fiber orientation in composite with respect to wipe direction
ratio for the composite, E1 is modulus for perfectly aligned fibers, [73].
E is modulus along the z direction for fibers aligned at angle , and
Ps is a weighting factor proportional to the total volume of all fibers
of length l. composite with stronger properties along the wipe direction, in
The stress of the composite is also determined using the parame- this case from 0 to 10◦ .
ters 3 and 4 as well as the fiber volume fraction f , fiber ultimate Overall, the vat polymerization method, though functional with
strength fu , the matrix volume fraction m , and matrix ultimate non-reinforced polymers, has issues when fibers were added due to
strength m *, as shown in Eq. (7). the increase in viscosity of the resin [75]. Combating this issue led
∗ to the creation of the deposit method, where the resin fiber mixture
c = 3 4 vf fu + vm m (7)
was extruded by nozzle in a more precise manner, and then lev-
Cheah et al. [73] reported good correlation between the above eled off using a wiper [67,74,75]. Finally, in order to create stronger
equations and their results, but Shofner et al. [23] reported that composites, more thought must go into ways to create interlayer
using rule of mixtures was orders of magnitude higher than what bonding such that stress concentrations will not build up at the
was found experimentally. Cheah et al. utilized cross sectional layer boundaries. The majority of the methods to print composites
images of SLA composites, as shown in Fig. 12, to determine the using SLA yields composites that have 1–5% the strength and modu-
distribution of fibers in the composite. From the images it was lus of a typical continuous composite material made with standard
determined that the fibers are distributed within the composites methods, and about half the strength of chopped fiber composites.
at various degrees near parallel to the direction of the resin wip- The strongest printed composites are those that have strong inter-
ing during the layer deposition. Fig. 13 shows the fiber orientation facial bonding between layers, use mats of woven material, and use
distribution [73]. This kind of distribution creates an anisotropic a higher laser power during fabrication.
Table 1
Strength and modulus of composites created from vat photopolymerization.
Material Test (Standards) Fiber Content Print method Modulus (GPa) Strength (MPa) REF
Glass Fiber- SL5170 resin Tensile (ASTM Standard ∼13 NA ∼1.75 ∼59 [75]
Method D638-91a, Type M-III) ∼16 NA ∼2.5 ∼65
∼22.5 NA ∼2.6 ∼66
20 Uncured rivets 1.96 44
20 No rivets 1.57 44
Glass Fiber- urethane acrylic Tensile (ASTM Standard Test 20 Laser Power 70 mW ∼2.6 ∼27.5 [73]
photo polymer Method D638) 100 mW ∼2.8 ∼27.5
130 mW ∼3 ∼28
160 mW ∼3.2 ∼29
Glass Fiber- SL5170 resin Tensile (dogbone) 16 Large interlayer bonding sites 2.48 71.8 [74]
17 Small interlayer bonding sites 2.49 72.0
E-glass Fiber-Acrylic resin Tensile (dogbone) Non-woven mat (7 g/m2 ) NA 2.23 43.6 [72]
Non-woven mat (17 g/m2 ) NA 2.85 55.2
Carbon Fiber-Acrylic resin Tensile (dogbone) Non-woven mat (7 g/m2 ) NA 1.81 43.6
Aramid fiber-acrylic resin Tensile (dogbone) Non-woven mat (17 g/m2 ) NA 2.07 29.9
E-glass −epoxy resin Tensile (dogbone) Non-woven mat (34 g/m2 ) NA 2.76 42.2
E-glass- Epoxy Tensile (dogbone) Non-woven mat (7 g/m2 ) NA 2.55 50.2 [78]
E-glass − polyurethane Tensile (dogbone) Non-woven mat (30 g/m2 ) NA 2.96 48.7
The vat photopolymerization process has several problems that of the structure is built in to the fabrication, so printing overhangs
occur due to 1) the mechanics of fiber addition, 2) fiber settling, 3) on parts take less time than while using a liquid-based manufactur-
formation of bubbles (voids), 4) increase in resin viscosity (causes ing system [80]. This process also takes place at room temperature,
poor mixing, leading to poor interfacial properties [77]), and 5) laser reducing complexity of needed equipment. The problem encoun-
diffraction (long cure time) caused by diffraction of laser energy tered using binder jetting to create composites is the low strength of
by the liquid and particulates[72,73]. However, vat photopolymer- the green part before the final processing occurs. The strength of the
ization can be used to created glass fiber reinforced samples with green part solely depends on the binder adsorption and mechanical
increased modulus and strength, as compared to neat resin [72,73]. interlocking of the powder [81]. Powder can also become trapped
For example, Karalekas et al. found that the addition of woven inside the body of the part. The liquid chosen as the ink of the pro-
glass fiber mats increased the resin modulus from 1.90 GPa for cess needs to be able to have the material properties necessary for
acrylic-based resin to 2.85 GPa, and the strength increased from fusion; not many materials can be used in this process.
37.1 MPa to 55.2 MPa. Additional results from this and other papers Little has been done in terms of fabricating composites using
are summarized in Table 1. The highest moduli in Table 1 are from binder jetting techniques because of the difficulty of jetting fiber or
Karalekas et al., who made use of non-woven mats in their compos- wetting of a fiber-polymer blend. Polymer fibers and glass fibers,
ites, and Cheah et al., who found that higher laser power and lower to improve the green strength of the parts, were successfully intro-
laser pitch can increase final composite strength [73,78]. However, duced into a ceramic composite using binder jetting [15]. The fibers
vat photopolymerization composites still do not reach standard were sieved, then mixed in with the powder bed, and a binder was
strength (100–300 MPa) and stiffness (3–20 GPa) of chopped fiber jetted on to make the parts. Unfortunately, most of the parts were
composites manufactured using means such as vacuum bagging or too weak to be removed from the build chamber without an infil-
autoclave processing [32,79]. Even with the use of fiber mats, com- tration of polymer. In another experiment, carbon nanofibers were
posites created by vat photopolymerization have only a fraction of dispersed in epoxy to increase conductivity [82]. The nanofibers
the modulus and strength of typical composites. were small enough (i.e. 100 nm diameter and 50–200 ␮m long) that
the fibers were mixed in the resin and infilled after the binder jet-
3.2. Binder jetting ting process [82]. Both papers showed that the created composite
was too weak without adding an additional binding step. To date,
Powder-based printing that uses binder jetting fuses a pow- no mechanical testing of reinforced polymer composites created by
der together by a sprayed on binder material or glue. This method powder bed processing with liquid binder has been attempted.
was initially termed 3D printing, but has changed to binder jet-
ting with the adoption of ASTM standard F2792. The steps involved
can be seen in Fig. 14 [80]. First, a layer of powdered material is
spread by a roller into a thin layer on a moveable platform. A sol-
vent, or binder liquid is then jetted onto the powder layer, bonding 3.3. Material extrusion
together the particles that make up the powder bed. The platform
then moves down a specified amount and an additional layer of Material extrusion composite additive manufacturing can be
powder is deposited. This process repeats until the part is finished. divided into two major subsections, depending on the feed mate-
The binder material can either be a liquid that reacts with the pow- rial used: (1) thermoplastic material and (2) thermosetting resin.
der to solidify the part, or a liquid binder that adheres the powder For thermoplastic materials, a heated nozzle melts the polymer-
to itself. fiber blend and deposits it onto a surface. Thermosetting resin is
Like powder bed fusion, binder jetting is able to make use of less commonly used as an additive manufacturing material, but
a wide range of powdered materials such as plastics, metals and is rapidly gaining favor in additive manufacturing because of the
ceramics. The ability to print with powder means that the support resins’ strength and toughness [16].
Fig. 14. The steps of 3D printing 1) powder is rolled onto the build area 2) the powder is smoothed and slightly compressed 3) a nozzle deposits the binder fluid onto the
bed of powder 4) the build station moves down while the powder bed piston moves up allowing the process to begin again [80].
3.3.1. Thermoplastic material extrusion

There are several different ways that solid materials are used
as printing materials in additive manufacturing. The most well-
known is material extrusion, which is also termed fused deposition
modeling (FDM) when used with thermoplastic polymers. In this
process raw polymer filament is fed through a heated nozzle until
it is liquefied, and then layered on a heated platform. This system
typically uses a thermoplastic material as the structural material
along with a support material that is used to support pieces that
may overhang the previous layers. The support material is removed
during post processing. Thermoplastic material extrusion is able to
achieve high resolution printing (0.0254 mm) in the x, y and z direc-
tions. The ease of use and the low toxicity of the printed material
gives this method the ability to be used in office settings. However, Fig. 15. MarkOne print path discontinuities [88].
because of the way the material is extruded, the printed material is
confined to those materials which can be melted by the nozzle and
then cooled and solidified on the print surface. The round nozzle thus the fibers have a lower average length than the original fiber
shape used in most material extrusion printers makes it difficult to composition [24,85].
achieve sharp external corners in parts [65]. Recently, several projects in industry and academia have
The material feed rate sets the speed of the part production: demonstrated reinforced thermoplastic material extrusion using
if a process is able to increase the melting and depositing of the continuous carbon fiber filaments [86,87]. Two commercial print-
material, the scan rate can be increased. Another way that systems ers from MarkForged, have demonstrated the printing of nylon
improve the speed of creation is by using an infill pattern, such reinforced with glass, carbon, and Kevlar fiber (i.e. MarkOne and
as honeycomb. This infill pattern means the part is mostly hollow MarkTwo) [86]. A study conducted by F. vand der Klift et al. [88],
and helps by saving material and time while printing. In order to showed that although the printer does print continuous fibers, the
achieve high strength of the parts, the path of printing has to be path planning and slicing algorithm on the MarkOne allows for cuts
changed during layers, since strength between deposited layers is within the part in order to not become “stuck”, by running into the
weaker than that along the print direction. previous deposited layer, while printing. For instance, to create a
Thermoplastics used in printing fiber reinforced composites square piece, the printer must move inward once while completing
using material extrusion include, polyetheretherketon (PEEK), the first pass around the perimeter of the part. This allows for a dis-
polycarbonate, Poly(methyl methacrylate) (PMMA), polylactic acid continuity in the part at the area where the print head moved inside
(PLA) and ABS [23,83,84]. Chopped fiber is most often used in mate- this perimeter, as shown in Fig. 15. Another disadvantage found
rial extrusion, where the fibers are mixed in with a thermoplastic, with printing continuous fibers stems from the need to include a
then extruded, creating coils of a fiber reinforced plastic. This rein- layer of pure thermoplastic on the bottom and top layers. These
forced plastic feedstock is then used within the rapid prototyping non-reinforced layers serve to help the fiber reinforcement stick to
machine by feeding through a heated nozzle (205◦ C for ABS doped the platform as well as keep the fibers from disintegrating when the
with carbon fiber) to soften and depositing on a platform (heated to part is removed. The fiber volume fraction was found to be around
85◦ C for ABS doped with carbon fiber) where the material cools and 34.5% in the samples printed from the MarkOne [88].
solidifies. Chopped fibers are typically the reinforcement of choice, One way to print a continuous fiber composite is by using a
be it glass or carbon fiber. It has been noticed in several cases that continuous fiber coated in thermoplastic as filament; another is to
breaking of the fibers occurs during processing and mixing, and merge a tow of fiber with a neat polymer filament during the extru-
Fig. 16. Comparison between compression molded (CM) composites and printed composites. a) CM dogbone with 0% carbon fiber (CF), b) 10% CF, c) 20% CF, d) 30% CF. e) a
printed thermoplastic dogbone with 0% CF, f) 10% CF, g) 20% CF and h) 30% CF [24].
sion process [86,87]. Work by Matsuzaki et al. [89], has recently to 58.60 MPa [93,94]), Tekinalp et al. showed that the inclusion of
created a continuous fiber printing system where carbon and jute fibers initially decreases the interlayer bonding by increasing the
continuous fibers are mixed with polylactic acid within the nozzle amount of voids, compared to that of a compression molded sample
before being printed. When printing with carbon fiber, a preheat- [24]. Fig. 16shows the comparison between the samples [24]. How-
ing element was used to heat the fibers before the consolidation ever, as the volume fraction of fibers increase, these voids begin to
within the nozzle, whereas the jute fibers were unheated in order decrease in size, increasing the interlayer bonding. It is suspect the
to not damage the fibers, as high temperatures during extrusion has increased bonding is due to fibers transcending the layers, in agree-
shown to cause decomposition of the jute resulting in voids[90]. ment with Zhong et al.’s work [24,93]. The coefficient of thermal
Additionally, there is currently a 3D printer in development that expansion of the carbon fiber helps in this situation, reducing the
prints continuous carbon fiber filaments with epoxy resin, by wet- die-swell of the bead as it exits the nozzle, in turn leading to less
ting the fiber before extrusion and using a UV laser to solidify the inter-laminar gaps during processing (Fig. 16(e) & (f)) [24]. How-
resin during printing [91]. Additionally, Gardner et al. [92] have ever, the voids within the material can increase with at higher fiber
recently printed composites using Ultem 1010 and carbon nan- loading (Fig. 16(f)–(h)) [24]. Even though voids slightly increase as
otube (CNT) yarn. One of the issues with printing continuous fibers the percentage of fiber reinforcement increases, the strength also
is the adhesion of the fiber to the print bed to begin the deposition increases. The thought is that the increase of fibers increases bridg-
process. To achieve this, the nozzle was advanced along the printing between deposited layers [93]. Fiber-filled printed composites
ing surface until adhesion occurred, and printing commenced. Parts show a tendency to have stronger material properties along the
made from CNT yarn are conductive as well as structural, creating print direction, due to the fibers aligning while being forced through
multifunctional parts. The structural testing showed the CNT com- the print nozzle [16,23,24,95].
posites have a strength around 125 MPa and a Young’s modulus of Another downfall of material extrusion fiber reinforced parts is a
3 GPa, within the expected range given by The Rule of Mixtures. By poor interface between the matrix and the small-scale fibers, which
using CNT yarn, 180◦ turns were achievable during printing with- lowers the strength of the material due to fiber pullout [23,93].
out gaps in the composite. Even greater mechanical results could Fig. 17 is an image of vapor grown carbon fiber (VGCF)-doped ABS
be achieved with better fiber wetting [92]. filament that was created by Shofner et al. [23]. The fibers are highly
A problem that plagues thermoplastic material extrusion com- aligned in the axial direction, indicated by the double headed arrow
posite printing is the voids created in the composite when in the image. Poor fiber/matrix adhesion from poor wetting of the
printing [23,24]. Though the inclusion of fibers within the filament fibers is also observed. The fibers are clean and the surface of the
improves the strength of the printing filament (from 24.5 MPa up polymer does not show damage around the fibers, indicating a high
Fig. 17. VGCF extruded via MATERIAL EXTRUSION aligned along the print direction,
with a high concentration of voids verifying poor interfacial properties [23].
Fig. 18. Epoxy-based composite fracture surface [101].

level of fiber pullout. Even with these defects in the extruded fil-
ament, the VGCF-doped ABS had higher strength (∼35 MPa) than throughout the curing cycle. The curing process takes place in a
the un-doped specimen (∼25 MPa) [23]. heated oven for 15 h for the composite to cure completely, and is
Ning et al. found that adding carbon fiber into thermoplastic then allowed to cool for 2 h before removing. The extruded fiber
filament lead to higher tensile strength and greater Young’s modu- doped ink typically shows fiber alignment, so long as the extru-
lus, but decreased toughness, yield strength, and ductility. In these sion rate of the slurry is greater than or equal to the print speed
studies, 5 wt% carbon fiber had the highest tensile strength, flexural [101,102]. It was found that the higher the pump speed compared
stress, flexural modulus, and flexural toughness [96]. Samples con- to the write speed, the more oriented the fibers became [102].
sisting of 7.5 wt% had the highest Young’s modulus. Higher weight Compton et al. [16], used this method to create cellular compos-
percentages experienced issues with porosity which led to poor ites to mimic the performance of balsa wood. The printer ink has
mechanical properties [97]. a pot life of 30 days, which allows for the printing and the curing
In later studies comparing processing conditions, it was deter- process to take place separately. Separating these processes allows
mined that having a raster angle of [0, 90] provides higher tensile layers to merge during the printing process, yielding better bond-
strength, Young’s modulus, and yield strength than raster angles ing between layers. The alignment of the fibers during the extrusion
of [45,−45]. Higher infill speeds led to less interaction and bond- process also allows spatial orientation of the reinforcement during
ing between printed paths giving lower mechanical properties, and printing, giving strength along the print direction. This capability
thinner layers decreased the number of voids and increased the ten- to align reinforcement allows for the potential to spatially opti-
sile properties. Finally, a nozzle temperature of 220◦ C allowed for mize the strength of an engineering design [16]. Nozzle size has
well-bonded interlayers, and thus gave the best properties over- no notable influence on this orientation when tested with nozzles
all. Higher temperatures yielded pores within the samples which between 200 and 610 ␮m [16].
caused lower tensile properties [98]. Figs. 18 and 19 show the fracture surface of an epoxy ink printed
One of the advantages that additive manufacturing has over tra- composite [16,101]. The fibers are oriented in the same direction,
ditional composite manufacturing is the ability to create preforms and fiber pull out and voids can be seen throughout the fracture
with greater complexity, without the added effort. This is demon- surface. The strength of printed composites using the epoxy ink
strated by a study by Quan et al. in 3D composite preforms [99], in method have yet to reach the full strength of typical epoxy compos-
which chopped carbon fiber and ABS were used to create preforms ites. This is possibly because of the poor wettability of the fibers, and
that were infused with silicon to create high strength composites. the increase of voids caused by the printing process [16,101]. The
However, this study again demonstrated the potential complex- boundaries between consecutive layers are unnoticeable in both
ity of printing composites, the poor interfacial bonding between figures, showing that the deposited layers fully coalesce, strength-
matrix and reinforcement [99,100]. ening the composites.
Current research trends are looking into the ability to orient
3.3.2. Thermoset material extrusion fibers within the extrusion process in a direction that differs from
Thermoset resins have been used in material extrusion sys- the print direction. Holmes et al. [104] manipulated nanofillers
tems to create fiber reinforced composite specimens in a process such as aluminum using their Field-Aided Laminar Composite sys-
previously titled solid freeform fabrication which is now referred tem (FALCom). FALCom works by using electric fields to align and
to as a branch of material extrusion by ASTM F2792 standards orient particles within a polymer system during the printing pro-
[16,101,102]. The most common method of accomplishing a ther- cess [104]. Collino et al. [105] demonstrated acoustic-field-assisted
moset print ink is by additives such as nano-clay particles or silica deposition of composites created from a shear thinning epoxy ink,
which are added to make the resin into a viscoelastic fluid that holds combined with SiC fibers, glass, and BaTiO3 microspheres. This
its shape upon printing [16,101]. The resin was diluted with diox- method allowed for higher volume fractions compared to other
ane (a solvent), with the expectation that after printing on a 100 ◦ C methods because of the ability to manipulate particles within a
hotplate, the dioxane would evaporate and create a hardened part small stream during deposition. Kokkinis et al. [106], was able to
that could then be moved to the oven to cure overnight at 100 ◦ C magnetically orient alumina platelets in a soft composite in order
[103]. By altering epoxy resin using additives to create a thixotropic to control the microstructure of printed soft composites. Adding
material that thins during extrusion through the nozzle, layers of this additional ability to change the microstructure of the compos-
reinforced resin were produced that were able to maintain shape ite during printing gives extended capability that is unattainable
Fig. 19. Extruded thermoset composite also shows propensity to fibers aligning along print direction a) shows the surface perpendicular to print, and b) shows the surface
parallel to the print direction [16].
Fig. 20. Distribution of strength of composites printed with material extrusion.

Fig. 21. Composite layers at 0◦ and 90◦ , created by LOM [108].
in traditional composite manufacturing methods, and allows for

tailorable strength in multifunctional printed parts.
lamination is able to embed wires into the layers during manufac-
3.3.3. Material extrusion summary turing [65]. Tension tests of fiber reinforced specimens showed that
The strength of composites printed with material extrusion is the ultimate strength was 70% that of the manufacturer’s supplied
summarized in Fig. 20 and tabulated in Table 2. It can be noted ultimate tensile strength for the material, thus highlighting a prob-
from Fig. 20, with few exceptions, the tensile strength of thermoset lem in creating the samples that leads to parasitic properties [108].
material extrusion has only reached between 20% −75% that of There are specific methods that allow fabrication of curved parts,
a typical chopped glass or carbon fiber composite. The reasoning reducing waste by utilizing a curved mandrel surface during the
behind this could range from use of discontinuous fibers to poor print process [25]. A curved mandrel reduces stress concentrations
interfacial strength between the matrix and fiber or poor interlayer in the composite, because fiber continuity is kept throughout the
bonding during printing. The highest elastic modulus thus far for process [25,109]. Curing of the composite parts usually requires
chopped fiber composites is by Compton et al. [16] at 24.54 GPa. another step, as sheet lamination machines are unable to provide
This could be due to the high fiber volume content at 34.8%, which the heat and pressure needed for a full cure [108].
is also the highest up to now for this method. Crockett et al. created Sheet lamination creates composites from laminated prepreg
the strongest composite with a carbon fiber-PEEK composite with sheets of fiber reinforced polymers, making it the most similar
30% fiber volume; low void content is expected to have contributed to stacking by dry preforms. The material it uses must be in a
to their success [95]. The use of single walled carbon nanotubes and sheet form that can be rolled, pressed, and cut as required. The
VGCF in a printed polymer composite had the lowest strength and issues exhibited when making composites pressed during fabri-
modulus because of poor interfacial properties, and considerable cation are: (1) shrinkage, (2) warping, (3) delamination, and (4)
voids and pullout [94]. shape deformation. The pressing process helps by creating parts
that are comparable to that made by conventional processes and
3.4. Sheet lamination have a void content below 5% (compared to the 2% rule of thumb
for conventionally manufactured composites) [108].
Most notably used for creating objects using sheets of fused Fig. 21 shows a micrograph of a unidirectional composite
paper or plastic, sheet lamination is one of the quickest additive formed by sheet lamination. The smooth surface is indicative of
manufacturing methods to create a composite part. Sheet lamina- a well-bonded composite and is similar to failure modes seen in
tion is useful when creating parts from prepreg sheets. The process traditionally processed composites [110].
consists of a machine that deposits the sheets and then heats, com- Table 3 details the results of composite production using sheet
presses, and cuts the layers to form a single part [25,107,108]. Sheet lamination technology.
Table 2
Material properties of composites printed using material extrusion.
Material Test (Standards) Fiber Content (%) Orientation to Young’s Modulus Strength (MPa) Ref.
print direction (GPa)
(degrees)
CarbonFiber-ABS Tensile (ASTM D638 type-V) 10 0 ∼8 ∼52 [24]

20 0 ∼11 ∼60
30 0 ∼14 ∼65
40 0 ∼14 ∼68
SWNT-ABS Tensile (ASTM D638 Type V) 5 NA ∼0.3 ∼22 [94]
Tensile (Fiber) ∼1.75 ∼33
VGCF-ABS Tensile (ASTM D638 Type V) 5 NA ∼0.25 ∼20 [94]
Tensile (Fiber) ∼1.3 ∼27
VGCF-ABS Tensile (ASTM D638 Type V) 10 0 0.79 37.4 [23]
Tensile (dogbone) 10 10/90 NA 24.4
Glass Tensile (none-printed 10.2–13 0 NA 38.93–58.60 [93]
fiber-ABS-LLDPE-PE square frames) 90 NA 0.72–11.15
Silicon Carbide Tensile (tensile bar 32 0 16.10 96.6 [16]
whiskers − Epoxy −none) 32 90 10.61 69.8
Carbon fiber/Silicon Tensile (tensile 34.8 0 24.54 66.2
carbide whiskers − bar-none) 34.8 90 8.06 43.9
Carbon fiber − PEEK
Epoxy Tensile (ASTM D638) 30 0 9.4 257 [95]
30 90 3.6 124
Glass fiber − Tensile (ASTM D638) 30 0 7 130
Polycarbonate 30 90 7 130
*Carbon fiber reinforced-PLA Tensile (dogbone) 1 0 5.8 90.0 [87]
Continuous Carbon Fiber − Nylon Tensile (tensile bar) 34.5 0 35.7 466.4 [88]
Table 3
LOM composite material properties.
Material Test (Standards) Fiber Content (%) Orientation Young’s Modulus (GPa) Strength (MPa) Ref.
Glass Fiber-Epoxy Tension (ASTM-d-3039) 55 0 NA 713 [108]

Compression (ASTM-d-695) 55 0 NA 896
Flexure (ASTM-d-790) 55 0 NA 1190
Interlaminar shear (ASTM D-2344) 55 0/90 NA 42.6
3.5. Powder bed fusion (PBF) with un-sintered powder during fabrication, harming composite
properties. Additionally, sanding, lacquering, painting or other post
Powder-based systems have the unique attribute of being able processing procedures must be performed to get a good finish, due
to manufacture plastics, polymers and metal parts. In powder bed to the grittiness of powder-based parts [66].
fusion (PBF), a laser beam is used to consolidate powder together PBF is being used for fiber reinforced composite fabrication to
to form a layer. The build platform moves downward the thick- create small projects like unmanned aerial vehicles [114]. Typi-
ness of one layer, and powder is spread over the sintered layer. cally, the reinforcement fibers are mixed with the powdered matrix
The chamber where the powder is kept is typically kept at a tem- material. The laser sinters together the material, and post process-
perature just below the melting point of the material for plastic ing is performed. Post processing can consist of first placing the part
parts, so the laser needs to heat the material only slightly to initi- in an oven to convert binders, and then infilling the part with resin
ate melting [13]. Typically, this method of creating parts is useful to further strengthen the composite and reduce voids [56].
because of its ability to use a mixture of powders to create a new Glass fiber reinforced composite materials made by 3D Systems
hybrid or composite material, for example metal matrix, ceramic have been tested using ASTM standard E647 [21]. It was found
matrix, and polymer blends [111,112]. J.P. Kruth et al. proposed that the fibers were effective in stopping crack propagation dur-
a method of distinguishing the different ways that powders can ing fatigue testing and the mode of failure for these composites
be fused together. The 4 main ways cited are solid state sintering, was weak interfacial bonding. The Young’s modulus and tensile
chemically induced binding, liquid phase sintering partial melting strength were slightly higher than that of non-reinforced ABS pro-
and full melting [113]. cessed by injection molding (E = 1.72 GPA, ␴ = 45.1 MPa), as shown
In solid state sintering, the material is fused together by neck for- in Table 4 [21]. After this infiltration, the part is cured at 80 ◦ C for
mation between particles. The parts formed by powder bed fusion 1 h. The inclusion of fibers leads to problems with consolidation of
usually have high material properties because the material is fused the polymer [116].
together by heat. Chemically induced binding is caused from the PBF carbon fiber reinforced composites were tested for ten-
laser disintegrating the material, which then causes the bonding sile modulus following ASTM standard D3039 [22]. The created
between particles, as with SiC ceramics. For composite systems composite was an orthotropic material, seen by the visible align-
formed by PBF, the most likely form of consolidation would come ment of the fibers in the xy plane (Fig. 22(A) & (B)). However,
from liquid phase sintering, where one material remains solid dur- these composites showed significant fiber pull out and breakage
ing processing while the matrix material liquefies to become the during testing, indicative of poor interfacial properties [110]. Flex-
binder. No supports are needed to during the creation of parts ural tests done as per standard ISO178-1993(E) with carbon fiber
because of the powders’ ability to support overhanging structures reinforced polyamide-12 found enhanced flexural strength and
during manufacturing, requiring less material for parts with many modulus [115]. The fracture surface of the specimen, seen in Fig. 23,
overhangs. However, a major drawback of this method is the energy shows matrix deformation occurred as well as tensile fracture of
cost of keeping the material just below melting temperature for the fibers, indicating good interfacial bonding between the matrix
the duration of manufacturing. Cavities within parts can be filled and the fiber reinforcement [110]. This is similar to the properties
Table 4
PBF composite material properties.
Material Test (Standards) Fiber Content (%) Orientation to print direction Modulus (GPa) Strength (MPa) Ref.
Glass fiber− Fatigue @ 23 ◦ C (ASTM E647) 25 y-z plane 2.71 43.7 [21]
Polyamide 12 Fatigue @ −50 ◦ C (ASTM E647) 25 y-z plane 6.6 73
Carbon Tensile (ASTM D3039) NA NA Ex = 6.3 NA [22]
Fiber-Polyamide 12 Ey = 3.54
Ez = 2.70
Carbon fiber- Flexural 30 0 ∼2.7 ∼75 [115]
polyamide 12 (surface (ISO178-1993(E)) 40 0 ∼3.2 ∼95
treatment) 40 0 ∼4.75 ∼115
Fig. 22. Alignment of fibers in the A) xz plane and the B) xy plane [22].
Fig. 23. Fracture surface of fiber reinforced PA-12 a) panoramic view b) view of fiber bridging and pullout along with matrix deformation [21].
observed in injection molded composites of similar composition. Figs. 24 and 25 compare the modulus and strength for the differ-
Furthermore, by treating the fibers with oxidation, a greater interfa- ent additive manufacturing methods. Overall, material extrusion
cial bond was created between the polymer and the fiber, leading to is used most often, whereas sheet lamination exhibits the best
increased strength [115]. By coating the fibers with polymer using quality composites. As shown in Fig. 24, composites made with
chemical precipitation before PBF, the wettability of the composite material extrusion along the print direction had the highest modu-
increased due to higher interfacial properties [115]. lus thus providing the ability to tailor stiffness in specific directions.
Fig. 24 also shows a corresponding lack of stiffness in the direction
transverse of printing for material extrusion composites as seen
4. Conclusion
in material extrusion 90◦ data points. This is probably due to the
weakness along the layered direction of a print, where boundaries
There is a need to quickly produce high quality composite parts
between layers yield voids and discontinuities. In terms of fiber
in a cost effective manner. While AFP, ATL, and FW all have the capa-
percentage, the highest fiber volumes are achieved using material
bilities of quickly creating parts accurately, the expense in buying
extrusion and PBF, where the percentage of reinforcement loading
the necessary specialized equipment often keeps these technolo-
is not limited due to the process. PBF does not yet have the Young’s
gies out of the reach of small-to-medium manufacturers. Additive
modulus achieved by material extrusion due to limitations with
manufacturing has the ability to quickly transverse from idea to
voids during the sintering process.
part without the waste in material, tooling, and time seen in the
The strength of printed fiber reinforced composites does not
automated methods. Parts can be customized for a number of appli-
have such clear delineations as does the elasticity, but material
cations and this can occur as quickly as needed, which makes
extrusion does have the highest strength of the different method-
the technology ideal for prototyping and individualization. Current
ologies, as seen in Fig. 25. This may be due to the higher loading
methods for composite AM have the capability of producing strong
percentages of chopped fiber within these printed composites.
parts, but full capability has yet to be realized with current systems.
The direction of force versus the direction of print does not show
Sheet lamination, for example, has the ability to create parts out of
much of a difference in strength due to poor interfacial properties
woven composite material that is pre-impregnated with resin, but
between the matrix and reinforcement material and fiber pullout
as of yet has not been used in an industrial setting.
Fig. 24. Modulus of composites made from AM technology.
Fig. 25. Strength of Composites made from AM technology.
(which leads to low load transfer to the fiber component of the properties must be better understood. In the future, the difference
composite). between parts made with traditional composite manufacturing
Compared to traditional manufacturing, AM has the ability to technology and AM will continue to decrease. The merging of the
create composites with shorter cycle times and less waste due manufacturing methods will yield higher quality parts, with more
to near net shape printing. Current trends in composite additive customization and less cost and variability in the manufactured
manufacturing are moving towards the ability to include continu- part.
ous fiber as a filament, in order to obtain material properties that
are comparable with traditional manufacturing methods. Also, the
ability to reorient the reinforcing material during the printing pro-
cess is currently gaining traction due to its ability to customize
part strength. AM contains the capacity to create a fully optimized
Acknowledgements
product, with the ability to align fibers multiple ways as well as con-
trol the fiber volume throughout the part − processes that would
J. Frketic is supported by the Florida State University Adelaide
be unfeasible with traditional composite manufacturing methods.
Wilson graduate fellowship. S. Ramakrishnan was supported by
The technology is able to create customizable products without
AMSRD-ARLRO-SI proposal number: 62885-MS-REP, agreement
having to invest in tooling for each product iteration. However,
number: W911NF-13-1-0132 from the Department of Defense
to compete with the strength of traditional methods, interfacial
(Army Research Office).
References [38] W. Polini, L. Sorrentino, Influence of winding speed and winding trajectory
on tension in robotized filament winding of full section parts, Compos. Sci.
[1] Heavy Metal: Three-dimensional Printing May Help Entrench the World’s Technol. 65 (10) (2005) 1574–1581.
Engineering Giants, The Economist, Berlin, 2014. [39] A. Beakou, M. Cano, J.B. Le Cam, V. Verney, Modelling slit tape buckling
[2] M. LaMonica, 3-D Printing Will Soon Become a Routine Manufacturing Tool | during automated prepreg manufacturing: a local approach, Compos. Struct.
MIT Technology Review, Technology Review, 2015. 93 (10) (2011) 2628–2635.
[3] ORNL Revealed the 3D-Printed Shelby Cobra, 2015 http://web.ornl.gov/sci/ [40] W. Goldsworthy, Geodesic path length compensator for composite-tape
manufacturing/media/news/detroit-show/index.shtml. placement method, Patent US 3810 805.
[4] O.R.N. Laboratory. http://www.ornl.gov/. [41] B.C. Kim, P.M. Weaver, K. Potter, Manufacturing characteristics of the
[5] P. Zelinski, IMTS’s 3D Printed Car Will Be Made of Carbon-Fiber-Filled continuous tow shearing method for manufacturing of variable angle tow
Plastic, 2014 http://www.mmsonline.com/blog/post/imtss-3d-printed-car- composites, Compos. A: Appl. Sci. Manuf. 61 (0) (2014) 141–151.
will-be-made-of-carbon-fiber-filled-plastic. [42] S. Nagendra, K. Srinivas, D. Jonathan, V. Parthasarathy, Optimization of tow
[6] J.E. Lindbäck, A. Björnsson, K. Johansen, New Automated Composite fiber paths for composite design, 36th Structures, Structural Dynamics and
Manufacturing Process: is it possible to find a cost effective manufacturing Materials Conference, American Institute of Aeronautics and Astronautics
method with the use of robotic equipment? Proceedings of the 5th (1995).
International Swedish Production Symposium (2012) 523–531. [43] Mtorres, 2014. http://www.mtorres.com/.
[7] ATL & AFP: Defining the megatrends in composite aerostructures, High [44] C. Grant, Composites automation: Trending smaller and robotic:
Perform. Compos. 16 (4) (2008) 68–71. CompositesWorld, High Performance Composites, 2014.
[8] A. Mafeld, The Automation of Polymer Composites Manufacturing, [45] G. Marsh, Automating aerospace composites production with fibre
Creaprod, Paris, France, 2010. placement, Reinf. Plast. 55 (3) (2011) 32–37.
[9] M.N. Grimshaw, J. Manuel, L. Diaz, Advanced technology tape laying for [46] Automated Dynamics, 2014. http://www.automateddynamics.com/.
affordable manufacturing of large composite structures, 2016. [47] Accudyne Systems, Inc, 2014. http://www.accudyne.com/.
[10] M.L. Skinner, Trends, advances and innovations in filament winding, Reinf. [48] Fiber Placement Robot − Coriolis Composites, (2014).
Plast. (2006). [49] Electroimpact, 2014. https://www.electroimpact.com/.
[11] F.P.W. Melchels, J. Feijen, D.W. Grijpma, A review on stereolithography and [50] Assyst Bullmer Composites, 2014, http://assystbullmer.co.uk/industries/
its applications in biomedical engineering, Biomaterials (2010). composites/.
[12] H. Lin, D. Zhang, P.G. Alexander, G. Yang, J. Tan, Application of visible [51] T.G.P. Gutowski, Advanced Composites Manufacturing, John Wiley & Sons,
light-based projection stereolithography for live cell-scaffold fabrication 1997.
with designed architecture, Biomaterials (2013). [52] K.V. Wong, A. Hernandez, A review of additive manufacturing, ISRN Mech.
[13] A. Mazzoli, Selective laser sintering in biomedical engineering, Med. Biol. Eng. 2012 (2012).
Eng. Comput. (2013). [53] F.P.W. Melchels, M.A.N. Domingos, T.J. Klein, J. Malda, P.J. Bartolo, D.W.
[14] L.S. Dimas, M.J. Buehler, Modeling and additive manufacturing of Hutmacher, Additive manufacturing of tissues and organs, Prog. Polym. Sci.
bio-inspired composites with tunable fracture mechanical properties, Soft 37 (8) (2012) 1079–1104.
Matter. 10 (25) (2014) 4436–4442. [54] G. Wurm, B. Tomancok, K. Holl, J. Trenkler, Prospective study on cranioplasty
[15] S. Christ, M. Schnabel, E. Vorndran, J. Groll, U. Gbureck, Fiber reinforcement with individual carbon fiber reinforced polymere (CFRP) implants produced
during 3D printing, Mater. Lett. 139 (2015) 165–168. by means of stereolithography, Surg. Neurol. 62 (6) (2004) 510–521.
[16] B.G. Compton, J.A. Lewis, 3D-Printing of lightweight cellular composites, [55] C. Emmelmann, P. Sander, J. Kranz, E. Wycisk, Laser additive manufacturing
Adv. Mater. (2014). and bionics: redefining lightweight design, Phys. Procedia (2011).
[17] Stratasys. http://www.stratasys.com/. [56] D.L. Bourell, M.C. Leu, K. Chakravarthy, N. Guo, K. Alayavalli, Graphite-based
[18] Fiberforge. http://www.fiberforge.com/. indirect laser sintered fuel cell bipolar plates containing carbon fiber
[19] B.C. Kim, K. Potter, P.M. Weaver, Continuous tow shearing for additions, CIRP Ann. Manuf. Technol. 60 (1) (2011) 275–278.
manufacturing variable angle tow composites, Compos. A Appl. Sci. Manuf. [57] W.H. Wai, RP in art and conceptual design, Rapid Prototyping Journal 7 (4)
43 (8) (2012) 1347–1356. (2001) 217–219.
[20] A. Angerer, C. Ehinger, A. Hoffmann, W. Reif, G. Reinhart, G. Strasser, [58] N. Oxman, Variable property rapid prototyping: inspired by nature, where
Automated cutting and handling of carbon fiber fabrics in aerospace form is characterized by heterogeneous compositions, the paper presents a
industries, IEEE Conf. Autom. Sci. Eng. 2010 (2010) 861–866. novel approach to layered manufacturing entitled variable property rapid
[21] A. Salazar, A. Rico, J. Rodríguez, J. Segurado Escudero, R. Seltzer, Martin de la prototyping, Virtual Phys. Prototyp. 6 (1) (2011) 3–31.
Escalera Cutillas, Fatigue crack growth of SLS polyamide 12: Effect of [59] S. Jokic, P. Novikov, S. Maggs, D. Sadan, S. Jin, C. Nan, Robotic positioning
reinforcement and temperature, Compos. Part B Eng. 59 (0) (2014) 285–292. device for three-dimensional printing, arXiv preprint arXiv 3400 (2014).
[22] R.B. Floersheim, G. Hou, K. Firestone, CFPC material characteristics and SLS [60] O.R.N. Laboratories, ORNL Revealed the 3D Printed Shelby Cobra, 2015
prototyping process, Rapid Prototyping J. 15 (5) (2009) 339–345. http://web.ornl.gov/sci/manufacturing/media/news/detroit-show/.
[23] M.L. Shofner, K. Lozano, F.J. Rodriguez-Macias, E.V. Barrera, [61] D.D. Gu, W. Meiners, K. Wissenbach, R. Poprawe, Laser additive
Nanofiber-reinforced polymers prepared by fused deposition modeling, J. manufacturing of metallic components: materials, processes and
Appl. Polym. Sci. 89 (11) (2003) 3081–3090. mechanisms, Int. Mater. Rev. 57 (3) (2012) 133–164.
[24] H.L. Tekinalp, V. Kunc, G.M. Velez-Garcia, C.E. Duty, L.J. Love, A.K. Naskar, [62] S. Lim, R.A. Buswell, T.T. Le, S.A. Austin, A.G.F. Gibb, T. Thorpe, Developments
C.A. Blue, S. Ozcan, Highly oriented carbon fiber–polymer composites via in construction-scale additive manufacturing processes, Autom. Constr. 21
additive manufacturing, Compos. Sci. Technol. 105 (2014) 144–150. (0) (2012) 262–268.
[25] D. Klosterman, R.P. Chartoff, Curved layer LOM of ceramics and composites, [63] E. Atzeni, A. Salmi, Economics of additive manufacturing for end-usable
Solid Freeform (1998). metal parts, Int. J. Adv. Manuf. Technol. 62 (9-12) (2012) 1147–1155.
[26] M.P. Groover, Automatio production systems, and computer-integrated [64] J.P. Kruth, M.C. Leu, T. Nakagawa, Progress in additive manufacturing and
manufacturing, Prentice Hall Press, 2007. rapid prototyping, CIRP Ann. Manuf. Technol. 47 (2) (1998) 525–540.
[27] D.H.J.A. Lukaszewicz, C. Ward, K.D. Potter, The engineering aspects of [65] I. Gibson, D.W. Rosen, B. Stucker, Additive Manufacturing Technologies:
automated prepreg layup: history, present and future, Compos. Part B-Eng. Rapid Prototyping to Direct Digital Manufacturing, Springer Science +
43 (3) (2012) 997–1009. Business Media, New York, 2010.
[28] C. Grant, Automated processes for composite aircraft structure, Industrial [66] C.K. Chua, K.F. Leong, 3D Printing and Additive Manufacturing: Principles
Robot: An International Journal (2006). and Applications, 4 ed., World Scientific, 2015.
[29] C. Technics. http://www.cnctechnics.com/Technology.html. [67] M. Haberer, G. Zak, C.B. Park, B. Benhabib, Design of a slot-coater-based
[30] F.H. Abdalla, S.A. Mutasher, Y.A. Khalid, S.M. Sapuan, A.M.S. Hamouda, B.B. layered-composites manufacturing system, J. Manuf. Sci. Eng. 125 (3) (2003)
Sahari, M.M. Hamdan, Design and fabrication of low cost filament winding 564–576.
machine, Mater. Des. 28 (1) (2007) 234–239. [68] A.M. Elliott, O.S. Ivanova, C.B. Williams, T.A. Campbell, Inkjet printing of
[31] C. Laval, CADWIND 2006-20 years of filament winding experience, Reinf. quantum dots in photopolymer for use in additive manufacturing of
Plast. 50 (2) (2006) 34–37. nanocomposites, Adv. Eng. Mater. 15 (10) (2013) 903–907.
[32] P.K. Mallick, Fiber-reinforced Composites: Materials, Manufacturing, and [69] Y. Lu, G. Mapili, G. Suhali, S. Chen, K. Roy, A digital micro-mirror
Design, CRC press, 1993. device-based system for the microfabrication of complex, spatially
[33] J. Fleischer, J. Schaedel, Joining automotive space frame structures by patterned tissue engineering scaffolds, J. Biomed. Mater. Res. A 77 (2) (2006)
filament winding, CIRP J. Manuf. Sci. Technol. 6 (2) (2013) 98–101. 396–405.
[34] J. Rousseau, D. Perreux, N. Verdière, The influence of winding patterns on [70] R. Liska, M. Schuster, R. Inführ, C. Turecek, C. Fritscher, V. Seidl, S.V.J.
the damage behaviour of filament-wound pipes, Compos. Sci. Technol. 59 Stampfl, Photopolymers for rapid prototyping, J. Coat. Technol. Res. 4 (4)
(9) (1999) 1439–1449. (2007) 505–510.
[35] J.T. Paul Jr, Method of making low-void filament wound structures, Google [71] J.R. Tumbleston, D. Shirvanyants, N. Ermoshkin, R. Janusziewicz, A.R.
Patents, 1969. Johnson, D. Kelly, K. Chen, R. Pinschmidt, J.P. Rolland, A. Ermoshkin, E.T.
[36] C.M. Hayes, E.J. Latos, J.L. Mclarty, Method of making self-lubricating Samulski, J.M. DeSimone, Continuous liquid interface production of 3D
filament wound tube, Google Patents, 1971. objects, Science 347 (6228) (2015) 1349–1352.
[37] B. Williams, E. Shehata, S.H. Rizkalla, Filament-wound glass fiber reinforced [72] D.E. Karalekas, Study of the mechanical properties of nonwoven fibre mat
polymer bridge deck modules, J. Compos. Constr. 7 (3) (2003) 266–273. reinforced photopolymers used in rapid prototyping, Mater. Des. 24 (8)
(2003) 665–670.
[73] C.M. Cheah, J.Y.H. Fuh, A.Y.C. Nee, L. Lu, Mechanical characteristics of [95] R.S. Crockett, J. O’Kelly, P.D. Calvert, B.D. Fabes, K. Stuffle, P. Creegan, R.
fiber-filled photo-polymer used in stereolithography, Rapid Prototyp. J. 5 (3) Hoffman, Predicting and controlling resolution and surface finish of ceramic
(1999) 112–119. objects produced by stereodeposition processes, Proceedings Solid Freeform
[74] G. Zak, M. Haberer, C.B. Park, B. Benhabib, Mechanical properties of Fabrication Symposium, Univ. Texas Austin (1995) 17–24.
short-fibre layered composites: prediction and experiment, Rapid Prototyp. [96] F.D. Ning W.L. Cong J.J. Qiu J.H. Wei S.R. Wang Additive Manufacturing of
J. 6 (2) (2000) 107–118. Carbon Fiber Reinforced Thermoplastic Composites using Fused Deposition
[75] G. Zak, M.N. Sela, V. Yevko, C.B. Park, B. Benhabib, Layered-manufacturing of Modeling Composites Part B: Engineering.
fiber-reinforced composites, J. Manuf. Sci. Eng. 121 (3) (1999) 448–456. [97] F. Ning, W. Cong, J. Qiu, J. Wei, S. Wang, Additive manufacturing of carbon
[76] T. Nakamoto, S. Kojima, Layered thin film micro parts reinforced with fiber reinforced thermoplastic composites using fused deposition modeling,
aligned short fibers in laser stereolithography by applying magnetic field, J. Compos. Part B Eng. 80 (2015) 369–378.
Adv. Mech. Des. Syst. Manuf. 6 (6) (2012) 849–858. [98] F. Ning, W. Cong, Y. Hu, H. Wang, Additive manufacturing of carbon
[77] T. Vaneker, E. Hofland, Additive manufacturing with additives: improving fiber-reinforced plastic composites using fused deposition modeling: effects
the properties of products produced with mask stereolithography, 2014. of process parameters on tensile properties, J. Compos. Mater. (2016).
[78] D. Karalekas, K. Antoniou, Composite rapid prototyping: overcoming the [99] Z. Quan, Z. Larimore, A. Wu, J. Yu, X. Qin, M. Mirotznik, J. Suhr, J.-H. Byun, Y.
drawback of poor mechanical properties, J. Mater. Process. Technol. 153 Oh, T.-W. Chou, Microstructural design and additive manufacturing and
(2004) 526–530. characterization of 3D orthogonal short carbon
[79] Y. Lu, Mechanical Properties of Random Discontinuous Fiber Composites fiber/acrylonitrile-butadiene-styrene preform and composite, Compos. Sci.
Manufactured from Wetlay Process, Virginia Polytechnic Institute and State Technol. 126 (2016) 139–148.
University, 2002. [100] Z. Quan, A. Wu, M. Keefe, X. Qin, J. Yu, J. Suhr, J.-H. Byun, B.-S. Kim, T.-W.
[80] A. Butscher, M. Bohner, S. Hofmann, L. Gauckler, R. Muller, Structural and Chou, Additive manufacturing of multi-directional preforms for composites:
material approaches to bone tissue engineering in powder-based opportunities and challenges, Mater. Today 18 (9) (2015) 503–512.
three-dimensional printing, Acta Biomater. 7 (2011). [101] P. Calvert, T.L. Lin, H. Martin, Extrusion freeform fabrication of
[81] S.A. Uhland, R.K. Holman, S. Morissette, M.J. Cima, E.M. Sachs, Strength of chopped-fibre reinforced composites, High Perform. Polym. (1997).
green ceramics with low binder content, J. Am. Ceram. Soc. 84 (12) (2001). [102] J. Peng, T.L. Lin, P. Calvert, Orientation effects in freeformed short-fiber
[82] J. Czyżewski, P. Burzyński, K. Gaweł, J. Meisner, Rapid prototyping of composites, Compos. A Appl. Sci. Manuf. 30 (2) (1999) 133–138.
electrically conductive components using 3D printing technology, J. Mater. [103] P. Calvert, G. George, L. Rintoul, Monitoring of cure and water uptake in a
Process. Technol. 209 (12) (2009) 5281–5285. freeformed epoxy resin by an embedded optical fiber, Chem. Mater. 8 (6)
[83] S. Kumar, J.P. Kruth, Composites by rapid prototyping technology, Mater. (1996) 1298–1301.
Des. 31 (2) (2010) 850–856. [104] L.R. Holmes, J.C. Riddick, Research summary of an additive manufacturing
[84] J. Torres, J. Cotelo, J. Karl, A. Gordon, Mechanical property optimization of technology for the fabrication of 3D composites with tailored internal
FDM PLA in shear with multiple objectives, JOM 67 (5) (2015) 1183–1193. structure, JOM 66 (2) (2014) 270–274.
[85] S.Y. Fu, B. Lauke, E. Mäder, C.Y. Yue, X. Hu, Tensile properties of [105] R.R. Collino, T.R., Ray, R.C. Fleming, J.D. Cornell, B.G. Compton, M.R. Begley,
short-glass-fiber- and short-carbon-fiber-reinforced polypropylene Deposition of ordered two-phase materials using microfluidic print nozzles
composites, Compos. A Appl. Sci. Manuf. 31 (10) (2000) 1117–1125. with acoustic focusing, Extreme Mechanics Letters.
[86] markforged FAQs, 2014. markforged.com. [106] D. Kokkinis, M. Schaffner, A.R. Studart, Multimaterial magnetically assisted
[87] M. Namiki, M. Ueda, A. Todoroki, Y. Hirano, R. Matsuzaki, 3D printing of 3D printing of composite materials, Nat. Commun. 6 (2015).
continuous fiber reinforced plastic, SAMPE Tech Seattle 2014 Conference, [107] D. Klosterman, R. Chartoff, N. Osborne, G. Graves, Composites via Laminated
June 2, 2014 − June 5, 2014, Soc. for the Advancement of Material and Object Manufacturing (LaM), in: Seventh International Conference on Rapid
Process Engineering, 800 Convention Place, Seattle, WA 98101-2350, United Prototyping, San Francisco, 1997.
states, 2014, p. SAMPE’s Carolinas Chapter; SAMPE’s Seattle Chapter. [108] D. Klosterman, R. Chartoff, G. Graves, N. Osborne, B. Priore, Interfacial
[88] F. Van Der Klift, Y. Koga, A. Todoroki, M. Ueda, Y. Hirano, R. Matsuzaki, 3D characteristics of composites fabricated by laminated object manufacturing,
printing of continuous carbon fibre reinforced thermo-plastic (CFRTP) Compos. A Appl. Sci. Manuf. 29 (9) (1998) 1165–1174.
tensile test specimens, Open J. Compos. Mater. 6 (01) (2015) 18. [109] D. Klosterman, R. Chartoff, N. Osborne, G. Graves, A. Lightman, G. Han, A.
[89] R. Matsuzaki, M. Ueda, M. Namiki, T.-K. Jeong, H. Asahara, K. Horiguchi, T. Bezeredi, S. Rodrigues, Development of a curved layer LOM process for
Nakamura, A. Todoroki, Y. Hirano, Three-dimensional printing of monolithic ceramics and ceramic matrix composites, Rapid Prototyp. J. 5 (2)
continuous-fiber composites by in-nozzle impregnation, Sci. Rep. 6 (2016) (1999) 61–71.
23058. [110] C.R. Brooks, A. Choudhury, Failure Analysis of Engineering Materials,
[90] A.R. Torrado, C.M. Shemelya, J.D. English, Y. Lin, R.B. Wicker, D.A. Roberson, McGraw-Hill New, 2016, pp. York2002.
Characterizing the effect of additives to ABS on the mechanical property [111] A. Simchi, H. Pohl, Direct laser sintering of iron?graphite powder mixture,
anisotropy of specimens fabricated by material extrusion 3D printing, Addit. Mater. Sci. Eng. 383 (2) (2004) 191–200.
Manuf. 6 (2015) 16–29. [112] K.K.B. Hon, T.J. Gill, Selective laser sintering of SiC/polyamide composites,
[91] K. Tyler, Method and apparatus for continuous composite CIRP Ann. Manuf. Technol. 52 (1) (2003) 173–176.
three-dimensional printing, Google Patents, USA, 2014, p. 11. [113] J.P. Kruth, P. Mercelis, J.V. Vaerenbergh, L. Froyen, M. Rombouts, Binding
[92] J.M. Gardner, G. Sauti, J.-W. Kim, R.J. Cano, R.A. Wincheski, C.J. Stelter, B.W. mechanisms in selective laser sintering and selective laser melting, Rapid
Grimsley, D.C. Working, E.J. Siochi, Additive Manufacturing of Prototyping Journal 11 (1) (2005) 26–36.
Multifunctional Components Using High Density Carbon Nanotube Yarn [114] D. systems. http://www.3dsystems.com/.
Filaments, 2016. [115] C. Yan, L. Hao, L. Xu, Y. Shi, Preparation, characterisation and processing of
[93] W. Zhong, F. Li, Z. Zhang, L. Song, Z. Li, Short fiber reinforced composites for carbon fibre/polyamide-12 composites for selective laser sintering, Compos.
fused deposition modeling, Mater. Sci. Eng. A 301 (2) (2001) 125130. Sci. Technol. 71 (16) (2011) 1834–1841.
[94] M.L. Shofner, F.J. Rodrıı́guez-Macıı́as, R. Vaidyanathan, E.V. Barrera, Single [116] J.J. Beaman, J.W. Barlow, D.L. Bourell, R.H. Crawford, H.L. Marcus, K.P.
wall nanotube and vapor grown carbon fiber reinforced polymers processed McAlea, Solid Freeform Fabrication: a New Direction in Manufacturing,
by extrusion freeform fabrication, Compos. A Appl. Sci. Manuf. 34 (12) 2061, Kluwer Academic Publishers, Norwell, MA, 1997, pp. 25–49.
(2003) 1207–1217.

Additive Manufacturing: Jolie Frketic, Tarik Dickens, Subramanian Ramakrishnan

Uploaded by

Copyright:

Available Formats

Additive Manufacturing: Jolie Frketic, Tarik Dickens, Subramanian Ramakrishnan

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Additive Manufacturing: Jolie Frketic, Tarik Dickens, Subramanian Ramakrishnan

Uploaded by

Copyright:

Available Formats

Additive Manufacturing 14 (2017) 69–86

Contents lists available at ScienceDirect

Automated manufacturing and processing of ﬁber-reinforced polymer

Fig. 1. Ranking of composite part manufacturing processes by optimum output [8].

Fig. 2. Filament winding ﬂow chart and diagram [30].

A multi-axis articulating robot can be used to hold the ATL head,

sists of a computer-aided design program and their robotic arm

2.2.2. Continuous tow shearing (CTS)

Composite automation is used in various industries. The

Many methods of additive manufacturing (AM) exist [52], each

Fig. 7. CTS machine set-up [19].

Fig. 10. Rapid layer composite manufacturing system [67].

Ec = ␹1 2 vf Ef + vm Em , 0 < 1 < 1 (2)

where ␹1 is orientation efﬁciency factor and is estimated by Eq. (3),

2 = Psk (1 − tanh (nsk ) /nsk ) (4)

In these equations, tk is thickness of the kth lamina panel, ␾ is the

3.3.1. Thermoplastic material extrusion

Fig. 18. Epoxy-based composite fracture surface [101].

Fig. 20. Distribution of strength of composites printed with material extrusion.

in traditional composite manufacturing methods, and allows for

CarbonFiber-ABS Tensile (ASTM D638 type-V) 10 0 ∼8 ∼52 [24]

Glass Fiber-Epoxy Tension (ASTM-d-3039) 55 0 NA 713 [108]

Fig. 24. Modulus of composites made from AM technology.

Fig. 25. Strength of Composites made from AM technology.

You might also like