Professional Documents
Culture Documents
Coconut Pla
Coconut Pla
https://doi.org/10.1007/s00170-022-08679-2
ORIGINAL ARTICLE
Received: 12 November 2021 / Accepted: 4 January 2022 / Published online: 11 January 2022
© The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2022
Abstract
This study investigates the effects of four variables during fused filament fabrication of organic biocompatible composite
material, PLA with coconut flour, at the ultimate tensile strength (UTS), and elasticity module (E) of the printed parts. The
parameter optimization uses Taguchi L18 design and regression models. The examined deposition variables are the layer
thickness, the nozzle temperature, the raster deposition angle, and filament printing speed. The effects of the above variables
on the strength of the parts are essential to enhance the mechanical response of the printed parts. The experimental outcomes
are investigated using the ANOM and ANOVA and modeled utilizing linear regression models. In addition, an independent
experiment was repeated three times at optimum parameters’ levels to evaluate the methodology, giving predictions errors
less than 3%. The observed results showed that the raster deposition angle dominates among the other variables in the studied
experimental area.
Keywords Taguchi · Optimization · Additive manufacturing · 3D printing · FFF · FDM · Mechanical strength · PLA wood ·
Eco-friendly material · Sustainability
13
Vol.:(0123456789)
4318 The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328
width, and air gap (infill density) on the strength character- LT increases the water absorption and the cross-sectional
istics of the FFF test specimen [14, 27, 28]. They concluded porosity, and the durability decreases. More, the density
that the strength of the FFF parts varies according to the of the printed PLA/W part increased as the printing speed
quality of the interlayer bond formation [27, 29]. The ther- decreased, and the tensile and flexural strength altered sig-
mal properties of the polymeric materials and the deposition nificantly by the deposition rates [36].
rates need optimization to achieve a robust bond interface The above researchers investigated the influence of the
and better strength [30]. All the investigations conclude that FFF parameters on the strength of different PLA wood flour
different printing parameters are appropriate for each other contents. In conclusion, to the authors' knowledge, it was not
material for optimizing the quality of FFF parts, such as reported any research in the literature where the influence of
mechanical response and shape accuracy [31]. four parameters, i.e., the LT, NT, raster DA, and PS, on the
In the literature, the organic biocompatible composite, mechanical response of the 3D printed PLA/W is studied.
PLA with wood compounds (PLA/W) investigated in dif- Moreover, the proposed experimental area of the selected
ferent material synthesis [12, 32–35] and FFF 3D printing parameters is investigated for the first time.
conditions [36–38]. Faludi et al. [12] reported dependable Therefore, this work is an experimental investigation of
adhesion between PLA and wood flours, and they proved how these four FFF parameters, the LT, NT, raster DA, and
that interfacial adhesion is strong in PLA/wood composites. PS, influence the ultimate tensile strength (UTS) and the
The use of para-wood powder derived from the furniture elasticity module (E) of the PLA mixed with coconut wood
industry was tested as an infilled material in the PLA matrix flour FFF parts. The results were analyzed and presented
in Chansoda et al. [35]. The authors showed that wood pow- using statistical tools such as main effect plots (MEP), anal-
ders influence composite PLA/wood filament quality and ysis of variances (ANOVA), interaction charts, and linear
mechanical properties. Zandi et al.[38] investigated the regression models. So, the optimum parameters proposed
fatigue behavior of a PLA/W (Timberfill, 8% wood fibers) to optimize the strength of the organic biocompatible com-
processed through FFF process parameters, i.e., the layer posite FFF-PLA/W material.
thickness (LT), the nozzle diameter, the infill density (ID),
and the printing speed (PS). They used the Taguchi L27
orthogonal array and found that the LT was the most critical 2 Materials and methods
parameter. In Kain et al. [19], two wood fiber contents (15
and 25%; Wood fibers, ARBOCEL® C100) mixed with PLA 2.1 Preparation of experiment
(Ingeo™ 3251D) tested in different raster deposition angles
(DA) during FFF (0–90°; step 15°). The wood content of According to ASTM D638 standards (see Fig. 1a), a dog-
25% gave better strength than 15%, and that the raster DA bone was designed having a thickness of 4 mm and trans-
affected FFF parts' mechanical response. The mechanical lated in STL format at the SolidWorks CAD program. Then,
performance of PLA/W parts, changing the nozzle tempera- the Craftbot Plus 3D printer (Fig. 1b) fabricated the speci-
ture (NT) between 210 and 250 °C, was also investigated in mens. Its platform is made of aluminum, thus giving the
[39]. The increase in the NT from 210 to 230 °C improves capability for printing specimens using different types of
the strength slightly, which agrees with other studies con- material. The maximum volume of the vat was 250 X 200
cerning FFF composites with PLA as a base material (see X 200 mm, and the printing speed was up to 200 mm/s.
[18, 31]. However, above 230 °C, the wood particle’s degra- The commercially available material NEEMA3D™ WOOD-
dation influences the strength and is not suggested. Ayrilmis PLUS consisted of 30% wood fibers of coconut and addi-
et al. [37] investigated the impact of LT on the water absorp- tives, and 70% pure PLA polymer was utilized. The filament
tion and strength of FFF-PLA/W specimens. Increasing the diameter was 1.75 mm. The specific gravity was 1.2 g/cc
13
The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328 4319
(ASTM D1505), the melting point between 140–150°, and in Fig. 2. The cross-head speed rate was 1 mm/min. At the
the nozzle's diameter was 0.4 mm. The minimum contended same time, force and displacement data for each test coupon
tensile strength and modulus for the NEEMA3D™ PLA were recorded through a data acquisition system and stored
series filaments is 70 MPa 1900 MPa, respectively (ASTM through BlueHill software for further treatment.
D882) as measured by the manufacturer. Figures 3a–f presents all graphs of eighteen experi-
The static tensile testing is performed on a strain-controlled ments of the L18 Taguchi orthogonal array. At first glance,
Instron 3382 universal testing machine with a load capacity of Figs. 3a–f shows that the raster DA most affects the UTS and
100 kN, as per ISO, equipped with a gripping fixture as shown E. The 0° DA presents the higher UTS and E in all graphs,
Fig. 3 Stress–Strain curves
for all specimens: (a-c) with
LT 0.1 mm, and (d-f) with LT
0.3 mm
13
4320 The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328
Table 1 Parameters with levels strands of the woven pattern. Finally, the printing speed
Parameters Units Levels determines the time between layers and affects the inter-
laminar quality [30]. All other FFF process parameters were
1 2 3
kept constant in this study to minimize the noise error.
Variable parameters This study utilizes the L18 Taguchi Orthogonal Array
Layer Thickness (LT) (mm) 0.1 0.3 - (OA). This OA consists of eighteen arrays and eight columns
Nozzle Temperature (NT) (°C) 180 200 220 [44, 45]. The eighteen arrays show the number of the experi-
Deposition Angle (DA) (o) 0 45 90 ments, 18 experiments, and the columns show the possible
Printing Speed (PS) mm/s 30 40 50 variables [41]. We have used only 4 out of 8 columns, i.e.,
Constant parameters four parameters (LT, NT, DA, and PS). The process error
Travel speed mm/s 40 (e) is estimated by considering the empty columns. So, it is
Infill density % 100 unnecessary to repeat each combination of the 18 experi-
Bed temperature o
C 60 ments. Table 2 lists the variable parameters (LT, NT, DA and
Room temperature o
C 20 PS), the measured attributes (UTS and E), and the eighteen
Room humidity % 50 (18) fractional combinatorial experimental design accord-
Outline/perimeter shells/top Number 1/1/1 ing to the Taguchi L18 (21 × 37) OA. A qualified individual
layers/bottom layers carefully did the eighteen mechanical response experiments
to eliminate noise errors.
Table 2 shows substantial max–min values between
and the 90° DA is the lowest. However, the effects of the eighteen experiments that were DA, LT, NT, and PS values
other processing parameters are not apparent, and statistical changing.
modeling should be applied to optimize the levels of the
parameters.
3 Results and discussion
2.2 Design of experiment
Table 2 tabulates all the results from the experiments con-
In the current study, the Taguchi L18 (2 1 × 3 7) experi- ducted in this work. The UTS is the higher tension a mate-
mental approach was employed [40–42]. This methodol- rial can sustain while being stressed before breaking. At the
ogy adopts balanced experiments according to Taguchi's same time, the elasticity module (E) is the linear proportion
proposed Orthogonal Arrays (OA). Although the number of between the stress and the strain in the elastic deformation
the executed experiments is a fraction of the full combina- zone of the stress–strain curve. Both UTS and E characterize
torial design, the methodology can find the best parameter the materials' withstand properties. Table 2 also summarizes
levels and construct predictive mathematical models [20, the basic statistics of the mechanical performance (mean,
43]. Finally, three validation experiments against the best min, max, spread). The spread values (max–min) of the UTS
conditions are applied to verify the experimental design and and E are considerable concerning the data set of the 18
evaluate the experiment's spread at the optimum conditions. samples. This variability is about 100% for the UTS and 69%
The first step in the above approach is selecting the vari- for E. Therefore, modeling the process is highly significant,
able parameters that affect the aspired attributes: the UTS reducing variability considerably. Finally, process optimiza-
and the elasticity module (E). This task is critical and has tion is vital for sustainable FFF printings [46].
two main concerns: (i) to select the appropriate variable It is worth here noting that the deposited beds' cooling
parameters and levels and (ii) all the experiments of the process affects the adhesion and fusion between sequential
decided orthogonal array (OA) to be achievable (should do layers, which, in turn, are affected by the LT, NT, and PS
all experiments contained in the OA). [10, 47–50]. Therefore, controlling the FFF process PS, NT,
Table 1 summarizes the selected variables and the con- and LT is conclusively significant [38, 51–54]. In Fig. 4,
stant parameters after the literature review in the introduc- side views and broken surfaces pictures are presented by
tion section and a 'trial-and-error' procedure. According to SEM for 0.3 mm LT, showing good interlaminar adhesion
the literature review, layer thickness and nozzle temperature and bonding. Scanning electron microscopy enabled detailed
(LT and NT) influence the dimensional accuracy, surface examination of the area of failure and the failure mode of
roughness, and mechanical response of the FFF parts [10]. the fabricated specimens. The morphological shape of the
In addition, raster deposition angle (DA) is also an essen- fractured cross-sectional surface implies a brittle mode of
tial parameter [18]. Furthermore, it affects the strength of failure (Fig. 4a-f). This fact verifies the already observed
the FFF parts, as this parameter defines the direction of the failure mode optically exhibited in tensile testing with no
13
The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328 4321
necking phenomena. Figure 4a, b presents an SEM magnifi- A way for identifying how one parameter interacts with
cation of the side view of a specimen where illustrated both another is by utilizing the interaction plots [44, 53, 55–57].
the number and length of PLA sequenced layers. Moreover, For example, Fig. 6a, b shows the LT, NT, DA, and PS inter-
in Fig. 4c, d, the brittle nature of the failure is also confirmed actions concerning the UTS and E accordingly. In addition,
by the resulting surface texture. these two plots show a strong interaction between the domi-
nant parameter raster DA and the others, PS, NT, and LT.
3.1 Effects of the variable parameters on UTS and E Analytically, the interaction plots showed that: (i) LT
interacts synergistically with NT and DA, i.e., the trend lines
The diagrams that show the effects of the variable param- have similar directions, (ii) LT interacts anti-synergistically
eters on an attribute are known as ANOM diagrams (analy- with PS, i.e., the trend lines have complex directions, (iii)
sis of means) or main effect plots (MEP). Such diagrams DA interacts synergistically with NT and PS, and (iv) NT
explain the influence of each variable graphically in consid- and PS interact anti-synergistically.
eration of the attribute measure. So, we quickly observe the After the MEP and interaction plots, the ANOVA is
optimum parameter levels and the trend lines. For example, the qualitative tool to investigate the importance of each
by employing these plots, the effects of the four process variable parameter on the quality attributes (here, the UTS
parameters (LT, NT, DA, and PS) on both the UTS and E and E). Analysis of variances (ANOVA) decomposes the
can be extracted (see: Fig. 5). errors of each variable on the total error. ANOVA fits
The MEP plots showed that the dominant parameter in linear or quadratic mathematical models and studies the
the utilized experimental space for the UTS and E measures least-squares differences. The 'MEP' plots showed that
is the raster DA. The three other variables are significant at the raster DA is the dominant parameter in our case by
a lower level. According to the MEP plots, the optimized utilizing linear mathematical models for the 'UTS' and 'E'
values for all the variables are zero raster DA, 30 mm/s PS, analysis of variances (see: Tables 3 and 4, respectively).
220 °C, and 0.3 mm LT; UTS and E are edified as 'the maxi- The ANOVA showed that the raster DA variable affects
mum the best' attributes. 80.1% and 92.6% on UTS and E, respectively. The three
13
4322 The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328
other variables (PS, NT, and LT) are insignificant for UTS and E) increases by decreasing the raster deposition angle.
and E, respectively. The F values are below two, and the P It means physically that the 0° aligned filament shows the
values are higher than 0.05. highest UTS and E values.
The outcomes occurring from Figs. 5 and 6 as well as of The other three parameters, NT, PS, and LT, are less
Tables 3 and 4 conclude in the following: important for the utilized experimental space. The F values
The raster DA parameter dominates both the UTS and are smaller than 2, and P values are higher than 0.05 (F < 2
E values; 80.1% and 92.6%. The F values are higher than and P > 0.05). The F value for the LT concerning the UTS
four (F > 4) and P values smaller than 0.05 (P < 0.05), which attribute is between 2 and 4 (2 < F < 4), which statistically
means this conclusion is statistically significant. The best means that it is more significant than the NT and PS. These
value for the raster DA is 0°, and the worst the 90°. The statistical values show that the PS, NT, and LT affect the
interactions of raster DA with the other parameters are optimal thermal bonding condition in an anti-synergistic
slightly synergistic (Fig. 6). The mechanical response (UTS way (see interaction charts, Fig. 6).
13
The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328 4323
Fig. 6 Interaction plots between DA, PS, NT, and LT according to the: (a) UTS and (b) E
Considering previous experimental work and knowing PS leads to a rise in UTS and E for 0.1 mm LT, which fol-
that the interlaminar bonding conditions are affected ini- lows the finding of a similar investigation for the same range
tially by the PS and NT, and secondly, by the LT [58], we of PS values, 30–50 mm/s [59].
decided to construct two 'MEP' plots for 0.1 and 0.3 mm LT The NT has different trend lines for 0.1 and 0.3 LT. In
(see Fig. 7). These MEP plots decompose better the influ- the case of 0.1 LT, when the NT increases, the UTS and E
ence of raster DA, NT, and PS according to each LT level. increase (Fig. 7a, b). On the other hand, for 0.3 LT, the NT
Figure 7a–d shows that the NT is a significant parameter for is insignificant for both UTS and E (Fig. 7c, d).
0.1 mm LT and minor for 0.3 mm LT, denoting that the inter- Finally, concerning the PS parameter, Fig. 7a–d shows a
laminar conditions are different in the case of 0.3 and 0.1 LT. significant influence for both UTS and E attributes for each
Specifically, the deposited strands cooling process is of the two LT (0.1 and 0.3 correspondingly). By increasing
affected more by the PS than the NT for 0.3 LT, while for 0.1 the PS, the UTS and E decrease for the 0.1 mm LT while
LT is affected evenly by NT and PS. Furthermore, Coogan rising for the 0.3 mm LT. In the total experimental space
and Kazmer [29] proved that the longer the material stays at (Fig. 5), the trend line of the PS is similar to that of the
a higher temperature than its glass transition level, the better 0.1 LT (Fig. 7a, b). The ANOVA shows that the PS is an
the bond becomes. Therefore, increasing NT or decreasing insignificant parameter (F < 2 and P > 0.05; Tables 3 and
13
4324 The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328
4) following Zandi et al. [38]. Remark that Fig. 7a–d is 4 is the raster DA having F-values higher than four and
employed to analyze MEP plots for the two different LTs P-values lower than 0.05.
as explained in the text to construct more accurate linear Considering the ANOVA (Tables 3 and 4) and the MEP
regression models and not replace Fig. 5a, b. Process opti- plots (Figs. 5 and 7) as well as the interactions charts
mization employs the L18 OA and not Fig. 7a–d. (Fig. 6), linear regression models were developed for both
the UTS and E for the 0.1 and 0.3 mm LT, respectively
3.2 Modeling and validation (Eqs. 1–4).
UTSLT=0.1mm =2.84 − 0.0889 ∗ DA − 0.0839 ∗ PS
In the case of our optimization problem, it is evident that (1)
the raster DA affects the most in the proposed experimental + 0.0797 ∗ NT ± e
area and that the other three parameters have minor effects.
ANOVA (see Tables 3 and 4) validates these results, where ELT=0.1mm = 549 − 2.570 ∗ DA − 3.62 ∗ PS + 1.04 ∗ NT ± e
the F-values of LT, PS, and NT are all below the number 4, (2)
which statistically means that they have minor significance. UTSLT=0.3mm =15.89 − 0.06615 ∗ DA + 0.0186 ∗ PS
The only significant parameter according to Tables 3 and (3)
− 0.0056 ∗ NT ± e
13
The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328 4325
ELT=0.3mm = 444 − 2.379 ∗ DA + 2.12 ∗ PS + 0.392 ∗ NT ± e that the models accurately predict both the UTS and the E
(4) attributes.
Note that the validation experiment is repeated three
Table 5 summarizes the model's predictions (Eqs. 1–4)
times at the optimum levels for the UTS and E according to
and error percentages. Finally, three evaluation experiments
the MEP plot for all 18 experiments (Fig. 5; raster DA = 0°,
are executed to validate the developed regression models
PS = 30 mm/s, NT = 220°, and LT = 0.3 mm). Therefore,
(Table 5; Fig. 8). The specimens for the validation experi-
Fig. 5 should be considered for the process optimization as
ments were built with the optimized parameters according
it has the orthogonality property and not Fig. 7a–d. Perceive
to the MEP in Fig. 5 (zero raster DA, 0.3 mm LT, 30 mm/s
that the orthogonality property is described in ref [44] and
PS and 220 °C NT). All three experiments give accurate
means that all combinations between two columns exist in
predictions with an accuracy lower than ± 5%, which means
an orthogonal array and appear evenly.
Fig. 8 Validation experiments
and strain–stress curves for the
optimized parameter values
(0.3mm LT, 220 °C NT, 0 o DA,
and 30 mm/s PS)
13
4326 The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328
13
The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328 4327
8. Vidakis N, Petousis M, Savvakis K et al (2019) A comprehensive 25. Afrose MF, Masood SH, Iovenitti P et al (2016) Effects of part
investigation of the mechanical behavior and the dielectrics of build orientations on fatigue behaviour of FDM-processed PLA
pure polylactic acid (PLA) and PLA with graphene (GnP) in fused material. Prog Addit Manuf 1:21–28. https://doi.org/10.1007/
deposition modeling (FDM). Int J Plast Technol 23:195–206. s40964-015-0002-3
https://doi.org/10.1007/s12588-019-09248-1 26. Vidakis N, Vairis A, Petousis M et al (2016) Fused deposition
9. Vidakis N, Petousis M, Velidakis E et al (2020) Three-dimensional modelling parts tensile strength characterisation. Academic J
printed antimicrobial objects of polylactic acid (PLA)-Silver nano- Manuf Eng 14
particle nanocomposite filaments produced by an in-situ reduction 27. Gurrala PK, Regalla SP (2014) Multi-objective optimisation of
reactive melt mixing process. Biomimetics 5:42. https://doi.org/10. strength and volumetric shrinkage of FDM parts. Virtual Phys Pro-
3390/biomimetics5030042 totyp 9:127–138. https://doi.org/10.1080/17452759.2014.898851
10. Chaidas D, Kechagias JD (2021) An investigation of PLA/W parts 28. Srivastava M, Rathee S (2018) Optimisation of FDM process
quality fabricated by FFF. Mater Manuf Process. https://doi.org/ parameters by Taguchi method for imparting customised proper-
10.1080/10426914.2021.1944193 ties to components. Virtual Phys Prototyp 13:203–210. https://d oi.
11. Vidakis N, Petousis M, Maniadi A et al (2020) Sustainable addi- org/10.1080/17452759.2018.1440722
tive manufacturing: mechanical response of acrylonitrile-butadiene- 29. Coogan TJ, Kazmer DO (2017) Bond and part strength in fused
styrene over multiple recycling processes. Sustainability 12:3568. deposition modeling. Rapid Prototyp J 23:414–422. https://doi.
https://doi.org/10.3390/su12093568 org/10.1108/RPJ-03-2016-0050
12. Faludi G, Dora G, Renner K et al (2013) Improving interfacial 30. Vanaei HR, Shirinbayan M, Deligant M et al (2021) In-Process
adhesion in pla/wood biocomposites. Compos Sci Technol 89:77– Monitoring of Temperature Evolution during Fused Filament Fab-
82. https://doi.org/10.1016/j.compscitech.2013.09.009 rication: A Journey from Numerical to Experimental Approaches.
13. Ćwikła G, Grabowik C, Kalinowski K et al (2017) The influence Thermo 1:332–360. https://doi.org/10.3390/thermo1030021
of printing parameters on selected mechanical properties of FDM/ 31. Chen K, Yu L, Cui Y et al (2021) Optimization of printing param-
FFF 3D-printed parts. IOP Conf Ser: Mater Sci Eng 227:012033. eters of 3D-printed continuous glass fiber reinforced polylactic
https://doi.org/10.1088/1757-899X/227/1/012033 acid composites. Thin-Walled Struct 164:107717. https://doi.org/
14. Mohan N, Senthil P, Vinodh S, Jayanth N (2017) A review on 10.1016/j.tws.2021.107717
composite materials and process parameters optimisation for 32. Motas JG, Gorji NE, Nedelcu D et al (2021) XPS, SEM, DSC and
the fused deposition modelling process. Virtual Phys Prototyp Nanoindentation Characterization of Silver Nanoparticle-Coated
12:47–59. https://doi.org/10.1080/17452759.2016.1274490 Biopolymer Pellets. Appl Sci 11:7706. https://doi.org/10.3390/
15. Bikas H, Stavropoulos P, Chryssolouris G (2016) Additive manu- app11167706
facturing methods and modelling approaches: a critical review. 33. Bhagia S, Lowden RR, Erdman D et al (2020) Tensile properties of
Int J Adv Manuf Technol 83:389–405. https://doi.org/10.1007/ 3D-printed wood-filled PLA materials using poplar trees. Appl Mater
s00170-015-7576-2 Today 21:100832. https://doi.org/10.1016/j.apmt.2020.100832
16. Vidakis N, Petousis M, Velidakis E et al (2020) On the strain rate 34. Bulanda K, Oleksy M, Oliwa R et al (2020) Biodegradable poly-
sensitivity of fused filament fabrication (FFF) processed PLA, mer composites based on polylactide used in selected 3D tech-
ABS, PETG, PA6, and PP thermoplastic polymers. Polymers nologies. Polimery 65:557–562. https://d oi.o rg/1 0.1 4314/polimery.
12:2924. https://doi.org/10.3390/polym12122924 2020.7.8
17. Durgun I, Ertan R (2014) Experimental investigation of FDM 35. Chansoda K, Suwanjamrat C, Chookaew W (2020) Study on
process for improvement of mechanical properties and produc- processability and mechanical properties of parawood-powder
tion cost. Rapid Prototyp J 20:228–235. https://doi.org/10.1108/ filled PLA for 3D printing material. IOP Conf Ser: Mater Sci Eng
RPJ-10-2012-0091 773:012053. https://doi.org/10.1088/1757-899X/773/1/012053
18. Kechagias JD, Vidakis N, Petousis M (2021) Parameter effects 36. Yang T-C, Yeh C-H (2020) Morphology and Mechanical Properties
and process modeling of FFF-TPU mechanical response. Mater of 3D Printed Wood Fiber/Polylactic Acid Composite Parts Using
Manuf Process. https://doi.org/10.1080/10426914.2021.2001523 Fused Deposition Modeling (FDM): The Effects of Printing Speed.
19. Kain S, Ecker JV, Haider A et al (2020) Effects of the infill pat- Polymers 12:1334. https://doi.org/10.3390/polym12061334
tern on mechanical properties of fused layer modeling (FLM) 3D 37. Ayrilmis N, Kariz M, Kwon JH, Kitek Kuzman M (2019) Effect
printed wood/polylactic acid (PLA) composites. Eur J Wood Wood of printing layer thickness on water absorption and mechanical
Prod 78:65–74. https://doi.org/10.1007/s00107-019-01473-0 properties of 3D-printed wood/PLA composite materials. Int J
20. Fountas NA, Kechagias JD, Manolakos DE, Vaxevanidis NM Adv Manuf Technol 102:2195–2200. https://doi.org/10.1007/
(2020) Single and multi-objective optimization of FDM-based s00170-019-03299-9
additive manufacturing using metaheuristic algorithms. Procedia 38. Zandi MD, Jerez-Mesa R, Lluma-Fuentes J et al (2020) Experi-
Manuf 51:740–747. https://d oi.o rg/1 0.1 016/j.p romfg.2 020.1 0.1 04 mental analysis of manufacturing parameters’ effect on the flex-
21. Fountas NA, Kitsakis K, Aslani K-E et al (2021) An experimental ural properties of wood-PLA composite parts built through FFF.
investigation of surface roughness in 3D-printed PLA items using Int J Adv Manuf Technol 106:3985–3998. https://d oi.o rg/1 0.1 007/
design of experiments. Proc Inst Mech Eng J: J Eng Tribol. https:// s00170-019-04907-4
doi.org/10.1177/13506501211059306 39. Guessasma S, Belhabib S, Nouri H (2019) Microstructure and
22. Kechagias J, Kitsakis K, Zacharias A et al (2021) Direct 3D Printing mechanical performance of 3D printed wood-PLA/PHA using
of a hand splint using Reverse Engineering. IOP Conf Ser: Mater Sci fused deposition modelling: Effect of printing temperature. Poly-
Eng 1037:012019. https://doi.org/10.1088/1757-899X/1037/1/012019 mers 11:1778. https://doi.org/10.3390/polym11111778
23. Chaidas D, Kitsakis K, Kechagias J, Maropoulos S (2016) The 40. Tsiolikas A, Mikrou T, Vakouftsi F et al (2019) Robust design
impact of temperature changing on surface roughness of FFF pro- application for optimizing ABS fused filament fabrication process:
cess. IOP Conf Ser: Mater Sci Eng 161:012033. https://doi.org/ A case study. IOP Conf Ser: Mater Sci Eng 564:012021. https://
10.1088/1757-899X/161/1/012033 doi.org/10.1088/1757-899X/564/1/012021
24. Aslani K-E, Chaidas D, Kechagias J et al (2020) Quality per- 41. Kechagias J, Petropoulos G, Vaxevanidis N (2012) Application
formance evaluation of thin walled PLA 3D printed parts using of Taguchi design for quality characterization of abrasive water
the taguchi method and grey relational analysis. J Manuf Mater jet machining of TRIP sheet steels. Int J Adv Manuf Technol
Process. https://doi.org/10.3390/jmmp4020047 62:635–643. https://doi.org/10.1007/s00170-011-3815-3
13
4328 The International Journal of Advanced Manufacturing Technology (2022) 119:4317–4328
42. Zhang JZ, Chen JC, Kirby ED (2007) Surface roughness optimiza- 52. Vyavahare S, Kumar S, Panghal D (2020) Experimental study of
tion in an end-milling operation using the Taguchi design method. surface roughness, dimensional accuracy and time of fabrication
J Mater Process Technol 184:233–239. https://doi.org/10.1016/j. of parts produced by fused deposition modelling. Rapid Prototyp
jmatprotec.2006.11.029 J 26:1535–1554. https://doi.org/10.1108/RPJ-12-2019-0315
43. Kechagias JD, Aslani K-E, Fountas NA et al (2020) A comparative 53. Fountas NA, Papantoniou I, Kechagias JD et al (2021) Experimen-
investigation of Taguchi and full factorial design for machina- tal investigation on flexural properties of FDM-processed PET-G
bility prediction in turning of a titanium alloy. Measurement specimen using response surface methodology. MATEC Web Conf
151:107213. https://d oi.org/1 0.1 016/j.m
easurement.2019.1 07213 349:01008. https://doi.org/10.1051/matecconf/202134901008
44. Phadke MS (1989) Quality engineering using robust design. Pren- 54. Vinoth Babu N, Venkateshwaran N, Rajini N et al (2021) Influ-
tice Hall, Englewood Cliffs, New Jersey, p 07632 ence of slicing parameters on surface quality and mechanical
45. Gopinath C, Lakshmanan P, Palani S (2021) Fiber laser microcut- properties of 3D-printed CF/PLA composites fabricated by FDM
ting on duplex steel: parameter optimization by TOPSIS. Mater technique. Mater Technol. https://doi.org/10.1080/10667857.
Manuf Process. https://doi.org/10.1080/10426914.2021.1981939 2021.1915056
46. Saxena P, Stavropoulos P, Kechagias J, Salonitis K (2020) Sus- 55. Kechagias JD, Ninikas K, Petousis M et al (2021) An investigation
tainability assessment for manufacturing operations. Energies. of surface quality characteristics of 3D printed PLA plates cut
https://doi.org/10.3390/en13112730 by CO2 laser using experimental design. Mater Manuf Process
47. Harris M, Potgieter J, Archer R, Arif KM (2019) In-process thermal 36:1544–1553. https://doi.org/10.1080/10426914.2021.1906892
treatment of polylactic acid in fused deposition modelling. Mater 56. Kechagias JD, Fountas NA, Ninikas K et al (2021) Surface char-
Manuf Process 34:701–713. https://doi.org/10.1080/10426914. acteristics investigation of 3D-printed PET-G plates during CO2
2019.1566611 laser cutting. Mater Manuf Process. https://doi.org/10.1080/
48. Tamburrino F, Graziosi S, Bordegoni M (2019) The influence of 10426914.2021.1981933
slicing parameters on the multi-material adhesion mechanisms of 57. Kechagias J, Ninikas K, Petousis M, Vidakis N (2021) Laser
FDM printed parts: an exploratory study. Virtual Phys Prototyp cutting of 3D printed acrylonitrile butadiene styrene plates for
14:316–332. https://doi.org/10.1080/17452759.2019.1607758 dimensional and surface roughness optimization. Int J Adv Manuf
49. Hashemi Sanatgar R, Campagne C, Nierstrasz V (2017) Investiga- Technol
tion of the adhesion properties of direct 3D printing of polymers 58. Savvakis K, Petousis M, Vairis A et al (2014) Experimental deter-
and nanocomposites on textiles: Effect of FDM printing process mination of the tensile strength of fused deposition modeling
parameters. Appl Surf Sci 403:551–563. https://doi.org/10.1016/j. parts. In: IMECE2014. ASME, Montreal, Quebec, Canada
apsusc.2017.01.112 59. Christiyan KGJ, Chandrasekhar U, Venkateswarlu K (2016) A study
50. Deisenroth DC, Moradi R, Shooshtari AH et al (2018) Review on the influence of process parameters on the Mechanical Proper-
of heat exchangers enabled by polymer and polymer composite ties of 3D printed ABS composite. IOP Conf Ser: Mater Sci Eng
additive manufacturing. Heat Transf Eng 39:1648–1664. https:// 114:012109. https://doi.org/10.1088/1757-899X/114/1/012109
doi.org/10.1080/01457632.2017.1384280
51. Melenka GW, Schofield JS, Dawson MR, Carey JP (2015) Evalu- Publisher's Note Springer Nature remains neutral with regard to
ation of dimensional accuracy and material properties of the Mak- jurisdictional claims in published maps and institutional affiliations.
erBot 3D desktop printer. Rapid Prototyp J 21:618–627. https://
doi.org/10.1108/RPJ-09-2013-0093
13