European Journal of Mechanics / A Solids 94 (2022) 104587

Contents lists available at ScienceDirect

European Journal of Mechanics / A Solids

journal homepage: www.elsevier.com/locate/ejmsol

Experimental characterization and mechanical behavior of 3D

printed CFRP
Kouider Bendine a, *, Dennis Gibhardt b, Bodo Fiedler b, Alexander Backs a
a
SPITZNER ENGINEERS GmbH, Hermann-Blohm-Str. 3 / 4. OG, 20457, Hamburg, Germany
b
Hamburg University of Technology, Institute of Polymer and Composites, Denickestrasse 15, 21073, Hamburg, Germany

A R T I C L E I N F O A B S T R A C T

Keywords: Additive manufacturing technology is pointed as the next generation production process. It not just allows the
3D printing manufacturing of high complex design but also reduces the required production time from model to part. The
Carbon fiber filament CFF main obstacle that holds 3D printing from growing up is the weak rigidity of the deliver part. To address this
Mechanical characterization
issue, continuous carbon fiber is nowadays merged and included in the printing process. The printing technology
Finite element analysis (FEA)
Microscopy
of reinforcement carbon fiber is not yet mature and needs to be explored more. Understanding the mechanical
behavior of the printed carbon fiber composites can help improving this process. The present study is a
contribution enhancing the understanding of the behavior of 3D printing parts based on carbon fiber rein­
forcement. Four types of standardization tests including tensile, interlaminar shear (ILSS) and dynamic me­
chanical thermal analysis (DMTA) have been performed for the mentioned purpose. Furthermore, first tensile-
tensile fatigue results are presented. The mechanical properties and their corresponding microstructure as
well as the effect of the fiber volume fraction have been analyzed. While the amount of pores found must still be
reduced, the printed samples already show a very promising mechanical behavior. The work also includes
different numerical models for the validation purpose.

1. Introduction generation manufacturing process. Its superiority comes from the fact
that it produces high complex geometry parts with almost no limits
Aerospace industry is a quite demanding field, which requires parts unlike the traditional processes. Moreover, it requires low investment
with the best mixture of light weight, high strength and rigidity (Irving costs and limits the time from the design to the serial production. All the
and Soutis, 2019). For this purpose, scientists have been working to previous advantages not just make the 3D printing method suitable to
propose materials that fulfill such requirements and they ended up by fully automate the CF manufacturing, but also can decrease the amount
proposing composite based long carbon fiber (CF) (Forintos and Czi­ of CF required for the processing and resulting in better manufacturing
gany, 2019). CFs possess the mentioned requirements and have in controllability. Nowadays, variety of 3D printing methods are available
addition a good fatigue and corrosion resistance which makes it a great in the market and multiple materials can be printed, from all these
substitution of metal material and can assist to better improve the techniques fused deposition modeling (FDM) also known as fused fila­
aerospace structures performance. Different types of manufacturing ment fabrication (FFF) (Abdullah et al., 2020; Adumitroaie et al., 2019;
processes have been used to fabricate CF noted among them molding, Eichenhofer et al., 2018; Hu et al., 2020) shows a great potential of
pick and place automated layup (Björnsson et al., 2018), (de Zeeuw composite long and short fiber printing.
et al., 2020), automated tape laying (Grimshaw et al., 2001) and tailored FDM uses nozzles to heat the filament and then deposits the melted
fiber placement (Nik et al., 2014). Unfortunately, the stated methods material on the building platform based on the layer-by-layer approach.
require high investment and involve a long time and also produce ma­ Two kinds of deposition techniques have been offered in the market; the
terial waste, consequently resulting in an enormous cost increase. first is by means of separate nozzles, one for the fiber and the other for
Alternative processes like 3D printing are now under investigation and the filament. The second is by coupling both using a common nozzle
have shown promising outcomes (Ngo et al., 2018). 3D printing also then making the deposition. Although, variety of thermoplastic fila­
well known as additive manufacturing (AM) is defined as the next ments dedicated to the carbon fiber printing are also available noted for

* Corresponding author.
E-mail address: bendinekouider@gmail.com (K. Bendine).

https://doi.org/10.1016/j.euromechsol.2022.104587
Received 7 January 2022; Received in revised form 17 February 2022; Accepted 16 March 2022
Available online 17 March 2022
0997-7538/© 2022 Elsevier Masson SAS. All rights reserved.
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

normal application ABS, PLA, PC, and PS (Dudek, 2013), (Vinyas et al., have been printed using Onyx and CFF with a 0.125 mm as deposition
2019) and for more advance applications generally filaments based on layer and heat nozzle of 270 ◦ C. The CFF direction was chosen to be
high performance polymers (HPPs) such as PEEK or PEKK (Singh et al., perpendicular to the specimen cross section. The mechanical properties
2019). Unfortunately, printing the carbon fiber-based thermoplastic of both materials given by the manufacturer and used for the modeling
(3DP-CFRP) is not yet fully monitored. On one hand because it is not are shown in Table 1. All the tests have been repeated three times with
always possible to align the fibers along the principal stresses directions maintaining the same test conditions to ensure the results validity.
especially when the part has a complex geometry and on the other hand,
the final parts present high void ratios and low fiber filament bonding 3. Results and discussion
(Caminero et al., 2018). This may lead to unexpected failure during life
time. To this end, a fully understanding of the printing parts behavior 3.1. Tensile tests
under different load scenarios, no matter if it is constant or cyclic loaded,
can help to predict their performance and serve to propose solutions for Unidirectional CFRP ASTM D3039 tensile specimens with aimed
further improvements and for more emerging in the industrial sectors. In fiber ratios of 13%, 21%, 30% and >40% respectively, were printed. and
this regard, research attempts and publications have been made towards tested using a 100 kN universal testing machine (ZwickRoell GmbH &
the design and characterization of 3DP-CFRP. Matsuzaki et al. (2016) Co. KG, Ulm, Germany). The load has been applied to the specimens
used FDM based 3D printing to manufacture CFRP out of PLA and CF. with a constant displacement speed of 2 mm/min until failure. The
The authors validated their method by performing tensile tests. The complete setup of the performed test is revealed in the right side of
obtained results show that the tensile stress reached a maximum value of Fig. 1. The corresponding stress strain curves for different fiber ratio
220 MPa (noted here that no fiber volume fraction or printing fiber ratio configurations have been plotted and illustrated in the left side of Fig. 1.
has been provided by the author). Justo et al. (2018) performed stan­ The results show clearly that the specimens expose a brittle behavior
dard mechanical tests on 3D printed carbon fiber specimens. The tests with no existence of a plasticity domain. This is quite expected due to the
were performed for 40% fiber ratio and the results showed that the fact that the specimens are driven by the carbon fiber properties, as they
printed specimens reached an approximate tensile strength around 700 are aligned in loading direction. During the test campaign it became
MPa. Van de Werken et al. (van de Werken et al., 2019) evaluated clear that the tensile properties for aimed fiber ratios of more than 40%
experimentally and numerically the responses of a printed CRFP spec­ showed a high deviation for samples printed with only slight changes in
imen under mechanical and thermomechanical load cases for several CF the aimed fiber ratios (41%, 43%, and 50%). While some of the speci­
ratios. The presented results confirm the dependency of the specimen mens performed exactly like the 30% fiber ratio specimen, others
strength and failure on the specimens geometry, fiber parameters and of showed the expected increase in strength and stiffness. Analyzing cross
the infill pattern. sections using microscopy uncovered in this regard, the number of CF
The authors noticed a lack of research papers dealing with the design layers printed in the core section of the samples varied between four and
and characterization of CFRP 3D printing. In this regard, the present seven layers. This printing issue has to be considered for the printing of
work is a contribution to the understanding of the 3DP-CFRP behavior parts with high fiber ratios. Therefore, the number of CF layers is given
under different load cases. The idea is to offer the research community a additionally for the tensile samples. The average results of tensile
reliable prediction on the expected thermo-mechanical properties of strength and young’s modulus are analyzed and plotted against their
3DP-CFRP. Thus, specimens are printed for different CF ratios and tested nominal fiber ratio and number of CF layers. The results are depicted in
based on the ASTM standardization. The benchmark experimental tests Fig. 2. The nominal tensile strengths are found to be with values of 207
involved tensile, ILSS and DMTA investigations as well as an outlook ± 16, 383 ± 20, 484 ± 13 and 738 ± 24 MPa respectively, while
into the tensile-tensile fatigue behavior. Accordingly, the corresponding young’s modulus reached 17.5, 23.7, 32.1 and 47.1 GPa. The mechan­
properties are distinguished. Moreover, finite element methods (FEM) ical results are means of three specimens for each fiber ratio and number
are used to provide models for the validation purposes. Finally, a of CF layers. It can be clearly seen from the figure that the results have a
microscopically investigation was performed to check the real fiber very small standard deviation which indicates that the printed repro­
volume fraction, number of CF layers printed and the void appearance as ducibility is assured at least for a single set up. Furthermore, it can be
well as the bonding between the fibers and the filament. found from Fig. 2 that the tensile strength and young’s modulus increase
linearly over the number of CF layers and thus with the real fiber volume
2. Material and methods fraction. Unfortunately, it is not possible to relate the mechanical
properties with the aimed fiber ratios specified for the printing process.
The setup consists of a Mark Two commercial 3D printer and its But as the mass increases also linearly with the number of CF-layers (due
corresponding slicing software called Eiger both provided by Mark­ to the difference in density of the CF and the Onyx layers), the weight or
forged, Inc. The printer offers a build volume of 320 × 132 × 154 mm density of the specimen gives a possibility to recheck the CF proportion
and is equipped with two extruders, one for the nylon and the other for after the printing. Additionally, strength and stiffness are calculated
the CF filament (CFF). The nylon used is known as Onyx, which is using only the area of the CF layers in each configuration. For this
specially made for the Mark Two and reinforced with chopped CF. When
working with CFF, the printer uses a fixed deposition layer equal to
Table 1
0.125 mm and can track partially or fully reinforced paths on the
Material properties for the carbon fiber filament and Onyx filament provided by
addressed area. It also offers four types of infill patterns for the non- the manufacturer.
reinforced area. The infill parameters geometry and the fiber direction
Property Carbon fiber reinforced Onyx
are illustrated in the appendix. It is noticeable that the printer provides
excellent finishing. But unfortunately, it gives the user a limited E1 52 GPa 1.7 GPa
E1 4.0 GPa 1.7 GPa
controllability on the fibers orientations and distributions which may
G12 2.0 GPa
negatively affect the parts rigidity especially for high complex geometry. υ12 0.33 0.4
In order to address the pointed goals cited previously, four different σT1 700 MPa –
mechanical tests were performed with respect to the ASTM standardi­ σC1 − 320 MPa –
zation. The tests included ASTM D3039 (A. C. D.-30 on C. Materials, σT2 48 MPa –
2008) tensile, ASTM D2344 (ASTM, 2006) interlaminar shear tests σC2 − 100 MPa –
73 MPa
(ILSS) and dynamic mechanical thermal analysis (DMTA) as well as γ12 –
σT1 700 MPa –
tensile-tensile fatigue tests on ASTM D3039 specimen. The specimens

2
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

Fig. 1. Tensile test stress strain curves for different numbers of CF layers and fiber ratios (left) and tensile test rig (right).

Fig. 2. The effects of fiber ratio on the mechanical parameters of the tensile tests.

purpose, microscopy was used to determine these areas. The results are two for 13%, three for 21% and four for 30% as well as for the 43%. The
shown in Fig. 2 as well and indicate that the strength and stiffness within maximum number of CF layers found was seven for an aimed fiber
the CF layers of each sample are constant. The mean values, depicted as fraction of 41% printing another set of tensile samples. Optical analysis
dotted lines indicate the maximum strength and stiffness of samples of the cross sections furthermore uncovered that the maximum local
consisting only of UD CF layers. fiber volume content in all specimen configurations was about 35%.
As the setting options are limited for the printing process, micro­ Accordingly, it seems to be impossible to print parts or samples with this
scopy was used to evaluate the fiber content and distribution within the material/printer combination that exceed this fiber volume fraction. The
tensile specimen. Therefore, cross sections of each aimed fiber content overall fiber volume fraction over the four CF filament layers of the 40%
were prepared and analyzed. As it can be seen in Fig. 3, the specimens sample was furthermore found to be between 15% and 25% depending
are divided into a defined number of layers in dependence of the aimed on the local spot of analysis. Additionally, this value is even more
fiber fraction. For the investigated fiber ratios, the number of layers was reduced by the volume of the Onyx layer. In summary the investigation

3
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

Fig. 3. Microscopic images of the 3D printed specimen with various fiber volume fractions.

clearly shows the weak points of the CFRP printing processes due to the 21.1%. Furthermore, some pores can be seen within the CF layer as well.
challenge of comparability of the real fiber contents in comparison with Part b) of the figure indicates that the Onyx layer including the chopped
the aimed fiber contents. Fig. 4 shows in a) some details of the fiber CF has well distributed fibers and a low pore content. The interphase
distribution and fraction within a part of a 40% tensile specimen. The between the CF and the Onyx layer seems to be smooth and without an
exemplary cross-sectional analysis presents a usual fiber distribution increased amount of defects.
including some resin rich areas and a “real” fiber volume fraction of It is well known that the presence of voids in composite material is

Fig. 4. Detailed microscopic images of the 3D printed specimen with aimed 40% fiber volume fraction. In a) the fiber distribution and fraction of 21.1% within two
CF layers is shown in red. In b) the Onyx layer is shown in detail. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web
version of this article.)

4
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

responsible for micro cracks and early damage initiation which leads to thicknesses were selected to be 0.125 mm to match that of the printed
a significant reduction of the global performance, especially when parts specimen. The stacking sequences were integrated as a unidirectional
are subjected to fatigue loading. For the case of aerospace application, multilayer with angel of [0◦ ]. The analysis has been performed using a
the acceptable void ratio must not exceed (<1%) (Eichenhofer et al., nonlinear solver in “Optistruct”. The obtained simulation results are
2018). Thus, as an effort to investigate the microstructure of the tensile compared with those of the experimental tests and both are plotted and
test specimens, a microscopic study focusing on the pores has been depicted in Fig. 6. It is clearly shown from the figure that the results are
performed as well. The occurrence of voids is evident and can be seen in in good match which indicated the usability of the proposed numerical
all microscopy figures mentioned before. The overall void distribution model.
can be seen more clearly in Fig. 3 for all fiber volume fractions. The void
content has been analyzed in terms of 2D micrograph analysis. There­
fore, the void content can only be addressed in terms of a 2D fraction of 3.2. In plane shear stress tests
the specimen cross section. Within this analysis it was found that the
pore fraction is about 5% within the inner fiber layer structure of the Interlaminar shear strength tests have been performed for printed
samples. At the edges and the transition between Onyx and CF layers, unidirectional CFRP specimens according to ASTM D2344 standard. The
single pores of huge dimension were found. These pores increase the test setup consists of a short beam placed on two cylindrical supports
overall content significantly. The pore structure can be categorized in at and bended using a half cylindrical indenter (Fig. 7). The setup helps to
least three categories. The first contains huge pores of dimensions up to reduce the beam rotation which allows the obtaining of a pure shear
more than 100 μm depth and 200 μm–500 μm length and width. These state. The tested beam printing was also performed layer by layer
pores appear mainly at the specimen edges between the Onyx and the CF through the thickness with a value of 0.125 mm for each. The proposed
filament or between single CF filaments. The second category consists of samples geometries and boundary conditions are depicted in the Ap­
pores up to a depth of 50 μm. These pores are regularly found along the pendix. The tests have been performed for 14%, 24%, 31% and 44%
boundaries of CF layers. The last category contains small pores of depth aimed fiber contents respectively. A displacement speed of 1 mm/min
of about 5 μm–25 μm. These can be found also within single CF layers. was used. Three specimens per test for each fiber ratio have been tested
For the sake of brevity only the case of nominal 40% fiber content is to ensure representative results. Fig. 7 also shows the results of the
presented within this study in more detail. Therefore, a 3D microscopic bending force versus the displacements. As can be noted from the graphs
image of the 3D printed sample is presented in Fig. 5. Especially the the maximum bending force is proportional to the fiber ratio, which is
geometry of smaller intra and inter layer pores can be seen due to the quite obvious due to the increase in the CF contents. The maximum force
three-dimensional analysis. The pore dimensions are mainly around 40 has been reached 1200 N for the case of 44% fibers, while for 31%, 24%
μm in diameter and 15 μm in depth (measured from the cutting plane). It and 14% the maximum forces have been found to be 1000 N, 600 N and
is worthy to be mentioned here that the main factor that controls the 250 N respectively. To illustrate the effect of number of tests on the
void appearance is the printing extrusion process itself. In the future, average value of the shear stress, the means for each fiber ratio are
printing parameter studies should help to better control the pore content presented in Fig. 8 using box representations. The predominant failure
in relation to the fiber volume fraction. mechanism found is a shear failure within the fiber rich central area of
In the present work, a 2D FE Model has been prepared based on the the sample. In order to check the effect of the fiber ratio on the speci­
geometry of the ASTM D3039 tensile specimen. The 3D design of the mens’ responses, the maximum bending and the shear stress are plotted
specimen has been created using Catia V5. The meshing process was against the fiber ratio and presented in Fig. 9. According to the results,
done using Hypermesh. Since the sample used is composite based, a the responses are linearly dependent on the fiber ratio and therefore the
decision has been made to perform the analysis using 2D shell elements. real fiber volume fraction.
Thus, the 3D geometry has been converted to a 2D model using the In order to evaluate the experimental test results numerically, a FE
midsurface option in Hypermesh. The model has then been meshed model test bench consists of a composite short beam and an indenter has
using 1200 quadratic elements with one side fixed and the other exposed been created and meshed using Hypermesh. The beam layers were
to an enforced load. The composite layers have been designed using a modeled the same way as presented in the previous section and subject
layer wise approach to build the exact layer distribution of the printed to a simply supported boundary condition in the left and right side. The
model. In order to obtain a good representative model, the layers indenter was modeled as a rigid body having only one degree of freedom
around the axis of the applied force. Surface to surface and frictionless

Fig. 5. Detailed 3D microscopic image of the pore dimensions within the 3D printed specimen with an aimed 40% fiber volume fraction.

5
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

Fig. 6. Numerical validation of the tensile test.

Fig. 7. In-plane test shear stress vs displacement curves for different fiber ratio.

contact has been used for modeling the contact interface between the a relatively low number of four CF layers has been tested in tensile-
beam and the indenter. In order to check the validity of the proposed FE tensile fatigue tests with a loading ratio of R = 0.1 and a testing fre­
model, a nonlinear analysis using “Optistruct” was carried out, it is quency of 3.5 Hz. Therefore, samples were tested on load levels which
worthy to mention here that the idea is just to provide a model that can equal to about 80%, 77%, 75% and 71% of the ultimate tensile strength
predict well the behavior of the specimens in the elastic region. There­ (UTS) of the four-layer specimen (480 MPa). As the number of tested
fore, no damage or plasticity criteria has been included during the specimens is low so far, the results should only be treated as a first es­
analysis. The obtained results are compared with the experimental tests timate on how the printed specimen performs under fatigue loading. In
and presented in Fig. 10. Both results are only slightly different, because Fig. 11 the fatigue results are presented in terms of a typical Wöhler
the effect of voids was not considered when performing the simulation. diagram showing the maximum applied tensile stress per cycle over the
As a solution of this issue, the volume element representation approach number of cycles to failure. The tests indicate so far that the curve
can be used. proceeds very flat and that the range between about 101 and 106 cycles
to failure will be found within less than 10% of stress reduction. The
overall fatigue stress level (from 80% to 70% of UTS) is comparable with
3.3. Fatigue investigation
non-printed unidirectional CFRPs and is thus very promising. The
typical failure patterns of the samples are also shown in Fig. 11. For all
For the fatigue investigation 250 mm long and 15 mm width speci­
specimens except the specimen tested at 340 MPa load level, the failure
mens of a nominal thickness of 1 mm were tested using an Instron servo
hydraulic testing machine equipped with a 100 kN load cell. The lifetime was within the free gauge length in terms of stepwise bundle failures.
The sample at the 340 MPa load level showed some splitting and failed
of the printed specimen with an aimed fiber volume fraction of 50% but

6
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

Fig. 8. The effects of the fiber ratio on the shear stress.

Fig. 9. The effects of fiber ratio on the mechanical parameters for the in-plane shear stress.

due to a machine boundary error. Even if the sample was not completely samples DMTA analysis was performed on samples of two different
broken, it was twisted and plastically deformed after the test, so that this fiber volume fractions. As can be seen in Fig. 12, the specimens were
sample was not tested again. Taking the actual quite high void level into tested in a temperature range between 23 ◦ C and 140 ◦ C. Due to the
account, the first fatigue results show in particular a good potential of limited applicability of parts above Tg, only the Tg region is investigated
the 3DP-CFRP material in terms of fatigue properties, which even might in detail. The melting temperature of PA6.6 is known to be around
be enhanced by reducing the void content and increasing the maximum 260 ◦ C. The glass transition temperature was measured in terms of the
fiber volume fraction in the future. onset temperature, which was calculated as point of intersection be­
tween the tangent at the low temperature regime and the tangent
3.4. DMTA results through the mid-temperature within the region of highest loss in storage
modulus. Both samples have a similar Tg of 85.7 ◦ C and 86.7 ◦ C
To characterize the thermo-mechanical behavior of the 3DP-CFRP respectively. Accordingly, the fiber volume content has no influence on

7
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

Fig. 10. Comparison between the experimental and numerical results of the ILSS test.

Fig. 11. Wöhler curve of 3D printed specimen with four CF layers (aimed fiber volume fraction of 50%). The results are shown in terms of maximum tensile stress as
well as normalized to the nominal strength of four CF layer specimens with 480 MPa.

the temperature application range of the printed specimens. Compared strength of 27.5 MPa for 44% fiber content. The tensile properties of the
with other CFRP material combinations, the printed thermoplastic printed samples reached a maximum tensile strength of 738 MPa and
composites show a competitive operating temperature of up to 70 ◦ C. young’s modulus of 47.1 GPa. With a Tg about 86 ◦ C the 3DP-CFRP
specimen shows the potential to become a real alternative for complex
4. Conclusions parts in various applications. Indeed, the results are promising but
cannot yet compete with those of traditional manufacturing processes.
Mechanical properties characterization of 3D printed composite with Especially the fiber volume content is still limited which restricts most of
long carbon fiber reinforcement has been performed experimentally and all the stiffness of the printed parts. In the future, the fiber content
numerically in the presented study. Unidirectional tensile as well as ILSS within the CF layers has to be enlarged and the printing operation has to
and DMTA tests with respect to ASTM standardization have been carried guarantee a reproducible CF layer distribution also for aimed ratios
out for different fiber ratios. The results show that the printed samples above 40%. As the specimen mass was found to increase linearly with
demonstrate a brittle behavior and relatively low interlaminar shear the number of CF layers in the tested samples, a weight or density
strength, which also indicates a relatively low bound between the fiber measurement is also a possibility for quality control on printed parts,
and the filament. The ILSS tests demonstrated a maximum shear which allows a non-destructive estimation of the printed parts

8
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

Fig. 12. DMTA results for 3D printed specimen with nominal fraction of 13% and 41% of CF. Tg was measured as Tg,onset.

properties. Microscopic analysis illustrates furthermore a high presence editing. Alexander Backs: Conceptualization, Project administration,
of voids due to the printing process itself. In this regard we highly Funding acquisition.
require a consolidation procedure after the printing to improve the
interfacial bonding and reduce the void content. As pores are known to
negatively affect foremost compressive and fatigue properties of CFRP Declaration of competing interest
future work has to concentrate on these load cases as well. As far as the
void content can be reduced by process optimization, the mechanical The authors declare that they have no known competing financial
properties of the 3D printed CFRP parts might at least compete with interests or personal relationships that could have appeared to influence
those of GFRP manufactured with classical production methods in a first the work reported in this paper.
step, by having a significantly lower density.
Acknowledgments
Author statement
This research work has been carried out within the project “Evalu­
Kouider Bendine: Conceptualization, Methodology, Tests, Software, ation of a multiparametric model for the prediction and optimization of
Investigation, Data curation, Writing - original draft, Visualization. 3D printed CRFP for aviation applications (3D-SE), funded by the Eu­
Dennis Gibhardt: Conceptualization, Writing - review & editing, Tests, ropean SME innovation Associate Horizon 2020 grant agreement No.
Data curation Visualization. Bodo Fiedler: Supervising, review & 861850.

Appendix A

As mentioned before, the printing was performed using a Markforged commercial printer. Each carbon fiber reinforced filament is following an
isotropic form with zero orientation in X direction as shown in Fig. A.1. a. The non-reinforced areas are filled with onyx using rectangular features. The
infill parameters as the fiber deposit direction are presented in Fig. A.1. The readers should note that the printing speed is adjusted and optimized by
the printer software and the authors has no control on it. Also, the slicing software doesn’t offer a possibility to enter a fix value of the fiber volume
fraction. However, it just gives the user the option to choose the area of fiber deposition as well as the infill patterns.

Fig. A.1. Slicing and infill parameters.

9
K. Bendine et al. European Journal of Mechanics / A Solids 94 (2022) 104587

References Grimshaw, M.N., Grant, C.G., Diaz, J.M.L., 2001. Advanced technology tape laying for
affordable manufacturing of large composite structures. In: International Sampe
Symposium and Exhibition, pp. 2484–2494.
Abdullah, F., Okuyama, K., Morimitsu, A., Yamagata, N., 2020. Effects of thermal cycle
Hu, J., Zhu, S., Wang, B., Liu, A., Ma, L., Wu, L., Zhou, Z., 2020. Fabrication and
and ultraviolet radiation on 3D printed carbon fiber/polyether ether ketone ablator.
compression properties of continuous carbon fiber reinforced polyether ether ketone
Aerospace, Bd. 7 (7), 95.
thermoplastic composite sandwich structures with lattice cores. J. Sandw. Struct.
Adumitroaie, A., Antonov, F., Khaziev, A., Azarov, A., Golubev, M., Vasiliev, V.V., 2019.
Mater. 23 (6), 1099636220909949.
Novel continuous fiber Bi-matrix composite 3-D printing technology. Materials, Bd.
Irving, P.E., Soutis, C., 2019. Polymer Composites in the Aerospace Industry. Woodhead
12 (18), 3011.
Publishing.
ASTM, D., 2006. 2344/D 2344M. Standard Test Method for Short-Beam Strength of
Justo, J., Távara, L., García-Guzmán, L., París, F., 2018. Characterization of 3D printed
Polymer Matrix Composite Materials and Their Laminates. Annu Book ASTM Stand.
long fibre reinforced composites. Compos. Struct., Bd. 185, 537–548.
Björnsson, A., Jonsson, M., Johansen, K., 2018. Automated material handling in
Matsuzaki, R., Ueda, M., Namiki, M., Jeong, T.K., Asahara, H., Horiguchi, K.,
composite manufacturing using pick-and-place systems–a review. Robot. Comput.-
Nakamura, T., Todoroki, A., Hirano, Y., 2016. Three-dimensional printing of
Integr. Manuf., Bd. 51, 222–229.
continuous-fiber composites by in-nozzle impregnation. Sci. Rep., Bd. 6, 23058.
Caminero, M.A., Chacón, J.M., García-Moreno, I., Reverte, J.M., 2018. Interlaminar
Ngo, T.D., Kashani, A., Imbalzano, G., Nguyen, K.T., Hui, D., 2018. Additive
bonding performance of 3D printed continuous fibre reinforced thermoplastic
manufacturing (3D printing): a review of materials, methods, applications and
composites using fused deposition modelling. Polym. Test., Bd. 68, 415–423.
challenges. Compos. Part B Eng., Bd. 143, 172–196.
A. C. D.-30 on C. Materials, 2008. Standard Test Method for Tensile Properties of
Nik, M.A., Fayazbakhsh, K., Pasini, D., Lessard, L., 2014. Optimization of variable
Polymer Matrix Composite Materials. ASTM International.
stiffness composites with embedded defects induced by automated fiber placement.
de Zeeuw, C.M., Peeters, D.M., Bergsma, O.K., Benedictus, R., 2020. Strategies for swift
Compos. Struct., Bd. 107, 160–166.
automated pick-and-place operations of multiple large-sized layers of reinforcement-
Singh, R., Singh, G., Singh, J., Kumar, R., 2019. Investigations for tensile, compressive
a critical review. Adv. Manuf. Polym. Compos. Sci. 1–15.
and morphological properties of 3D printed functional prototypes of PLA-PEKK-HAp-
Dudek, P., 2013. FDM 3D printing technology in manufacturing composite elements.
CS. J. Thermoplast. Compos. Mater. 0892705719870595.
Arch. Metall. Mater., Bd. 58 (4), 1415–1418.
van de Werken, N., Hurley, J., Khanbolouki, P., Sarvestani, A.N., Tamijani, A.Y.,
Eichenhofer, M., Wong, J.C., Ermanni, P., 2018. Exploiting cyclic softening in continuous
Tehrani, M., 2019. Design considerations and modeling of fiber reinforced 3D
lattice fabrication for the additive manufacturing of high performance fibre-
printed parts. Compos. Part B Eng., Bd. 160, 684–692.
reinforced thermoplastic composite materials. Compos. Sci. Technol., Bd. 164,
Vinyas, M., Athul, S.J., Harursampath, D., Thoi, T.N., 2019. Experimental evaluation of
248–259.
the mechanical and thermal properties of 3D printed PLA and its composites. Mater.
Forintos, N., Czigany, T., 2019. Multifunctional application of carbon fiber reinforced
Res. Express, Bd. 6 (11), 115301.
polymer composites: electrical properties of the reinforcing carbon fibers–a short
review. Compos. Part B Eng., Bd. 162, 331–343.

10