Journal Pre-proof

3D printing of modified soybean hull fiber/polymer composites

Vamsi Krishna Balla, Jogi Ganesh Dattatreya Tadimeti, Kunal H. Kate, Jagannadh
Satyavolu

PII: S0254-0584(20)30819-1
DOI: https://doi.org/10.1016/j.matchemphys.2020.123452
Reference: MAC 123452

To appear in: Materials Chemistry and Physics

Received Date: 5 March 2020

Revised Date: 15 June 2020
Accepted Date: 17 June 2020

Please cite this article as: V.K. Balla, J.G. Dattatreya Tadimeti, K.H. Kate, J. Satyavolu, 3D printing of
modified soybean hull fiber/polymer composites, Materials Chemistry and Physics (2020), doi: https://
doi.org/10.1016/j.matchemphys.2020.123452.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2020 Published by Elsevier B.V.

 Author Statement with CRediT roles

Vamsi Krishna Balla: Investigation; Methodology; Formal analysis; Writing original draft;
Review and Editing

Jogi Ganesh Dattatreya Tadimeti: Investigation; Review and Editing;

Kunal H. Kate: Conceptualization; Methodology; Formal analysis; Supervision; Review and

Editing

Jagannadh Satyavolu: Conceptualization; Formal analysis; Review and Editing

 3D Printing of Modified Soybean Hull Fiber/Polymer Composites
Vamsi Krishna Balla1,2,3, Jogi Ganesh Dattatreya Tadimeti2, Kunal H Kate1*, Jagannadh
Satyavolu2

1
Materials Innovation Guild, Department of Mechanical Engineering, University of Louisville,
Louisville, KY 40208, USA.
2
Conn Center for Renewable Energy Research, University of Louisville, Louisville, KY 40208,
USA.
3
Bioceramics and Coating Division, CSIR-Central Glass & Ceramic Research Institute, 196 Raja
S.C. Mullick Road, Kolkata, WB 700 032, India.

* Corresponding Author
Kunal H Kate
Materials Innovation Guild
University of Louisville
2210 S Brook St. Shumaker Research Building,
Louisville, KY US 40208. E-mail: kunal.kate@louisville.edu

Page 1 of 21
Abstract
Herein, we have evaluated the soybean hull derived fibers as reinforcements to manufacture
thermoplastic copolyester (TPC) composites using 3D printing process namely fused filament
fabrication (FFF). The hulls were subjected to physical and chemical treatments to understand
their influence on the microstructural and mechanical properties of the composites. We found
strong dependence of the surface quality, printing defects and inter-bead/interlayer bonding on
the fiber treatment. The composites made using dilute acid hydrolysis treated fibers increased the
relative density of the composites to 99% and reduced the pore size from 81µm to 39µm. Defect-
free fiber-matrix interfacial characteristics in these composites enhanced the elastic modulus
from 36MPa to 54MPa. Similarly, the toughness and stress at 50% strain of these composites
were ~30% and 50% higher than the pure TPC, respectively. Our results clearly demonstrated
that the low-cost and abundantly available soybean hulls when modified using dilute acid
hydrolysis have a strong potential in the fabrication of natural fiber reinforced composites.

Keywords: Soybean hull; Natural fiber composites (NFCs); Material extrusion; Fused filament
fabrication (FFF); Polymer-matrix composites (PMCs);

1. Introduction
Significant amount of research has been carried out on the additive manufacturing (AM) or
three-dimensional printing (3DP) of variety of polymers, thermoplastic and thermosetting, for
various industrial applications [1]. AM offers several advantages such as compositional and
structural gradation, mathematically optimized complex designs, compared to conventional
manufacturing of polymer parts [2]. However, the components made using AM are still found to
be inferior to conventionally manufactured parts in terms of their mechanical and functional
performance. As a result, processing of polymer composites reinforced with appropriate fillers
that can provide improved performance is gaining attention. Considerable research and
understanding on the polymer composites processing using different AM technologies have been
reviewed [3,4]. The majority of these composites are reinforced with synthetic fillers/ fibers,
which require large amount of energy to produce compared to the natural fibers production [5].
Consequently, there has been increasingly more demand for the natural reinforcements in the
production of polymer composites [6], which provide low-cost feed stock, high strength-to-

Page 2 of 21
weight ratio (depending on the fiber), biodegradability, recyclability and overall environmental
sustainability. A wide variety of natural fibers including wood, kenaf, hemp, jute have been used
in the fabrication of polymer composites using conventional processing routes [7–11].
Manufacturing high-strength natural fiber reinforced polymer composites (NFCs) using AM is
relatively more challenging than conventional composites due to inherent characteristics of
natural fibers (thermal stability, hydrophilicity), which makes it difficult to eliminate
inhomogeneities. Some of the important problems in the AM processing of NFCs include nozzle
clogging, gas porosity, agglomeration of fibers, material flow and viscosity variations [12]. Le
Duigou et al. [13] have reported significant amount of porosity (~ 16%) in the poly(lactic acid)
(PLA) + poly(hydroxyalkanoate) (PHA) + 15 wt.% wood fiber composite filaments, which
could not be eliminated in the FFF printed parts. It was also found that the part size had strong
influence on the porosity of these NFCs and the porosity increased with increase in the part size.
In another study [14], PLA composites reinforced with wood particles (up to 50 wt.%) have been
fabricated using FFF and it was found that more than 10 wt. % wood can decrease the strength of
the composites to 30 MPa from 55 MPa. Further, these printed parts were rough with significant
amount of porosity, wood particle agglomeration and during printing nozzle clogging was also
observed. There has been very little reported literature on the AM of NFCs due to these
difficulties.

In this article, we report FFF of thermoplastic copolyester (TPC) composites reinforced with 5
and 10 wt.% of dilute acid hydrolyzed and unhydrolyzed soybean hull fibers. Soybean hull fibers
were chosen as potential reinforcement, because the soybean production has been 123 million
tons in 2018 and after processing (for oil, protein, etc.) the residual soy hulls has very low market
value as soy meal and animal feed [15]. Further, the microfibrils and microparticles derived from
the soybean hulls were found to have good thermal stability after chemical processing [16,17],
while their shear thinning behavior can be beneficial for composite processing. Therefore,
through this investigation, we intend to evaluate soybean hull fibers as reinforcing material in the
TPC composite fabrication. The primary goal of this investigation is to analyze the
microstructural and mechanical properties of TPC-soybean hull fiber composites made using
FFF and understand the influence of different physical and chemical treatments on these
properties. Detailed topological and microstructural analysis was performed in terms of porosity,

Page 3 of 21
pore size, uniformity of fiber distribution and fiber-matrix interfacial characteristics, which were
correlated with composites’ mechanical properties.

2. Materials and methods

2.1 Preparation of TPC-soybean hull fiber composites
In this work the soybean hulls (Owensboro Grain Company, Owensboro, KY, USA) were used
in different conditions to fabricate 5 and 10 wt.% TPC composites using FFF. Two sets of
soybean hulls were prepared without chemical treatment i.e. dry blending and wet shear mixing
(soy hulls: water = 1:1.75 wt.%), to reduce their size. In another group, dilute acid hydrolysis
(single-stage and double-stage) was used to improve the wettability between the soybean hull
fibers and the TPC matrix. More details about the hydrolysis treatment can be found elsewhere
[18]. These chemical treatments were aimed to remove the impurities and reduce the
hydrophilicity of the soybean hulls, which is essential to achieve good compatibility with
hydrophobic polymers [12]. The details of different composites prepared in this investigation are
presented in Table 1.

Before mixing, the TPC granules were dried in an oven at 80°C for 8h and the moisture content
of the soybean hull fibers was determined using standard moisture analyzer. The composite feed
stocks were prepared by mixing appropriate amount of soybean hull fibers (to achieve 5 and 10
wt.% reinforcement) and TPC granules (DuPont Hytrel4056) using a torque rheometer
(Intelli-Torque Plasti-Corder, C. W. Brabender Instruments, Inc. NJ, USA) at 180°C, 50 RPM
for 18 to 20 min. The composite mixtures were then crushed to small granules (1 to 5 mm),
which were then extruded in the form of ∅ 1.75 mm filaments using capillary rheometer
(Rheograph 20, GÖTTFERT Werkstoff-Prüfmaschinen GmbH, Germany). The composite
filaments were extruded at 165°C with 0.15 mm/s speed (shear rate = 30 s-1) using tungsten
carbide die. The viscosity of different composites was also measured during filament extrusion.
The extruded filaments were spooled and further used for FFF of the composite parts.

Page 4 of 21
 Table 1 Different composites and FFF parameters used in the present investigation.
Material Treatment Moisture, % FFF printing parameters
AR-Dry Dry blended as-received hulls 10.05 ± 0.37 Layer thickness 200 µm
AR-Shear Wet shear mixed as-received hulls 3.61 ± 0.26 Printing speed 30 mm/s
ST1 Single-stage hydrolyzed hulls 2.75 ± 0.52 Extrusion width 550 µm
ST2 Two-stage hydrolyzed hulls 2.96 ± 0.31 Printing temperature 220 °C
TPC Neat TPC Bed temperature 65 °C
Nozzle diameter 500 µm
Fill angle ± 45°

We have performed several experiments to print TPC-soybean hull fiber composites using a
desktop FFF machine (Printrbot) and then selected the parameters shown in Table 1 to fabricate
several test coupons for further characterization and testing. Our initial printing experiments
using extrusion temperature between 200 °C and 240 °C showed that a temperature of 220 °C
ensures consistent material flow during printing. Low extrusion temperatures resulted in severe
nozzle clogging. Good part adhesion with build plate was achieved with a bed temperature of 65
°C. All the samples were printed with 100% infill density to achieve defect free, dense parts with
high strength. Other important printing parameters are summarized in Table 1 and Simplify3D
software was used to generate G-code for printing.

2.2 Physical characterization

The printed parts were characterized for their relative density to assess porosity and printing
induced defects. Initially, the density (g/cm3) of soybean hull fiber and TPC granules was
determined using a gas pycnometer. These density values were used to calculate the theoretical
density (ρt, g/cm3) of the composites following rule of mixtures. The density of 3D printed
composite parts was experimentally measured using Archimedes principle (ρc, g/cm3). Then the
relative density (%) of the 3D printed composites was calculated as: (ρc/ρt)×100. The surface
roughness of the composite samples was measured with portable surface roughness tester
(Surftest SJ-210, Mitutoyo America Corporation, OH, USA) using a scan length of 5 mm at 0.5
mm/s speed. At least 12 scans were made on each sample (X-Y surface – top surface, X-Z
surface – build direction) and an average roughness with standard deviation is reported. Further,
the composites were examined using scanning electron microscope (SEM, TESCAN USA, Inc.,

Page 5 of 21
Warrendale, PA, USA) for surface topography, the presence of printing induced defects, porosity
and exposed soybean hull fibers.

2.3 Microstructural and mechanical properties

All composite samples, including pure TPC, were sectioned along and across the build direction
followed by a cold mounting in acrylic resin. Then the mounted samples were ground on series
of SiC emery papers and final polishing was performed on velvet cloth using 1 µm Al2O3
suspension. These polished sections of composite samples were analyzed for microstructural
features such as fiber distribution, porosity, printing defects and fiber-matrix interface using
SEM.

The tensile mechanical properties of the printed composite test coupons, in terms of Young’s
modulus, stress at 5% and 50% strain (as none of the samples failed within the cross-head span
of the tensile testing machine) and toughness (area under the stress-strain curve) were
determined using universal testing machine at 100 mm/s cross head speed (ASTM D638). The
tensile testing of the printed samples (minimum of three samples for each composition) was
carried out in normal to build direction. The mechanical properties were presented as mean ±
standard deviation and Student's t-test was used to perform statistical analysis where p < 0.05
was considered statistically significant. To understand the failure of TPC-soybean hull fiber
composites made using FFF, selected tensile tested samples’ surfaces were examined using
SEM.

3. Results and discussion

3.1 Surface topography, roughness and density of 3D printed composites
Initially the surface morphology of as-printed TPC-soybean hull fiber composite parts was
examined using SEM to understand the influence of soybean hull reinforcement on material’s
flow during printing and resultant changes in the surface morphology. Fig. 1 shows typical top
surface morphology of pure TPC and its composite parts. As expected, the pure TPC parts
showed relatively smooth surface morphology compared to the soybean hull fiber reinforced
composites. All composite parts revealed miniature surface undulations and some exposed
soybean hull fibers. Further, among these composites, the surface of the AR-Dry composites

Page 6 of 21
appears to be roughest, which can be attributable to the coarse fraction of untreated soybean hull
fibers. With wet shear mixing and hydrolysis the reinforcements became finer, which resulted in
smoother surface comparable to that of the pure TPC parts. Under present printing conditions
some printing induced defects, primarily between the perimeter and infill, were observed, as
indicated by arrows in Fig. 1. However, the bonding between the adjacent beads appears to be
very good (circles in the insets in Fig. 1). At the same time, the bead-perimeter interfacial region
and the interface between the adjacent beads showed some material build up/over flow in the
pure TPC parts, which was relatively less in the composite parts. This is attributable to the high
viscosity of pure TPC (2700 Pa.s) compared to the composites (1775 to 2100 Pa.s) at identical
printing temperature of 220°C. The high viscosity of pure TPC also resulted in debonding or
improper bonding between beads at some locations, Fig. 1. Although some geometrically
difficult to fill areas (triangular areas in Fig. 1, 10AR-Shear and 10ST2) were observed in the
composite parts, the interface between the adjacent beads or layers appears to be smooth
suggesting complete bonding (insets of Fig. 1).

Fig. 1 SEM images showing the typical top surface (X-Y direction) topographical features of
TPC and TPC-soybean hull fiber composites fabricated using FFF. Arrows indicate the interface
between the beads and some defects at these regions. Circles in the in-set microstructures show
complete bonding between adjacent beads.
Page 7 of 21
SEM microstructures showing typical surface morphology of FFF fabricated TPC and TPC-
soybean hull composite parts along the build direction are presented in Fig. 2. All parts revealed
good interlayer bonding with isolated defects. The interface between the layers was found to be
straight in the TPC parts due to its high viscosity compared to the composites, which exhibited
uneven interface (easy lateral flow due to low viscosity). The high viscosity of TPC also resulted
in occasional debonding or unbonded regions between the layers, as shown in Fig. 2, which can
be eliminated by increasing the printing temperature. All the composite samples showed some
exposed soybean hull fibers on the surface, which is unavoidable. However, the interface
between the layers found to be tight and diffuse in the composites, as shown in the insets of Fig.
2, and therefore these parts are expected to have strong bonding. Relatively low viscosity of the
soybean hull fiber reinforced composites enables better flow of the material on previously
printed layers and wetting with the same. As a result, the layer-to-layer and bead-to-dead
bonding found to be better in the composites than in the pure TPC parts, which exhibited a sharp
interface between the layers (Fig. 2).

Fig. 2 Typical build direction surface morphology of TPC and its composites. Arrows indicate
interlayer interface/defects and exposed soybean hull fibers. In-sets show bonding between the
layers at high magnification.
Page 8 of 21
The influence of soybean hull fiber treatment and their concentration on the average roughness
of the 3D printed composite parts in different directions is shown in Fig. 3. The top surface
roughness (Ra) of pure TPC parts was found to be 7.7 ± 1.9 µm and the composite parts
exhibited a roughness between 7.0 ± 1.6 µm and 11.9 ± 2.3 µm, depending on the treatment and
concentration of the soybean hull fiber. It can be seen from the Fig. 3a, that there is no
significant difference in the average top surface roughness among the parts under present
printing conditions. However, the observed variations can be directly ascribed to the reduction in
melt viscosity and reinforcement size (80-120 µm for AR-Dry and AR-Shear, and 10-100 µm for
ST1 and ST2) in the composites prepared using chemically treated fibers (ST1 and ST2),
compared to untreated fibers (AR-Dry and AR-Shear). We expected that increasing the
concentration of soybean hull fiber in the composites increases the surface roughness due to
increased number of exposed fibers on the surface. However, it appears that the easy material
flow, due to the decrease in the viscosity during printing, would have worked against the
anticipated increase in surface roughness with increased soybean hull fiber concentration. This
could be beneficial during tensile testing as eliminating such potential weak regions (exposed
hull fiber) can improve mechanical properties. The reduction in the exposed fibers is also
expected to improve moisture resistance of these NFCs. The roughness of the parts along the
build direction, shown in Fig. 3b, was almost double that of the top surface. For example, the
roughness of pure TPC parts in the build direction was 20.9 ± 1.5 µm compared their top surface
roughness of 7.7 ± 1.9 µm. Similarly, the composite parts showed build direction roughness
between 17.0 ± 2.6 µm and 20.9 ± 2.1 µm. As observed with top surface roughness, the
hydrolysis of soybean hull fibers can help in reducing the surface roughness marginally in the
build direction under present printing conditions. Further improvements in the surface quality of
these composites parts can be achieved by fine-tuning printing parameters as well as using high-
end printers, which is beyond the scope of present investigation.

Page 9 of 21
Fig. 3 Average surface roughness of pure TPC and TPC-soybean hull fiber composites (a) Top
surface (X-Y direction), (b) Build direction (X-Z direction).

The relative density of 3D printed parts is presented in Fig. 4, which clearly shows the beneficial
effect of dilute acid hydrolysis of soybean hull fibers in achieving high density parts (low
amount of total porosity). The composite parts showed a relative density between 89 ± 0.9 % and
99 ± 0.5 %, while the density of the pure TPC part was 89 ± 1 %. The low density of pure TPC
parts could be due to printing induced defects and lack of proper bonding between the beads as a
result of its high viscosity under present printing conditions (Table 1), as shown in Figs. 1 and
2. Among the composites, the ST2 composites exhibited the highest relative density. The
composites made using untreated fibers showed significant decrease in the density (97 to 89%)
with increasing soybean hull fiber concentration from 5 to 10 wt.%. However, the chemical
treatment of soybean hulls found to reduce or eliminate the detrimental effect of increased
soybean hull fiber concentration on the density of 3D printed composite parts. It can be seen
from the Table 1, that the moisture content of ST1 and ST2 samples is significantly less than the
other two samples (p < 0.01 to 0.0001), as the dilute acid hydrolysis can reduce the moisture
content, lignin and other impurities from these fibers [12,19,20], which can create porosity due to
evaporation during composite processing. Therefore, the reduction in the porosity of the
composites made with hydrolyzed soybean hull fibers is attributed to their low moisture content

Page 10 of 21
and hydrophilicity [12] leading to better bonding with the TPC matrix. Overall, the reduction in
the porosity of these 3D printed composites can improve their mechanical properties.

Fig. 4 Influence of soybean hull treatment and concentration on the relative density of 3D printed
TPC composites.

3.2 Microstructural analysis

The microstructural analysis of 3D printed parts was performed on polished sections to
understand printing induced defects, pore size, pore shape and soybean hull fiber distribution.
Initially, those regions of the parts (tensile test coupons) that are geometrically difficult to print
were examined and a comparison of low-magnification light microstructures of pure TPC and
TPC-soybean hull fibers composite parts is shown in Fig. 5. It can be clearly seen that the pure
TPC parts have large printing induced defects such as debonding or lack of bonding between the
beads and incomplete fill. However, such defects appear to be absent in the composites due to
their better flowability and wettability as a result of their low melt viscosity compared to the pure
TPC. The better flowability of the composites is reflected in their lower viscosity between 1775
and 2100 Pa.s compared to that of pure TPC with 2700 Pa.s. The bonding between the beads was
also found to be sound across the composite part surface. However, the composites made using
untreated soybean hull fibers (AR-Dry and AR-Shear) exhibited large fiber bundles along the
parts’ perimeters and some agglomerates as well. On the other hand, the fibers treated with dilute
acid hydrolysis resulted in more uniform distribution of the fibers within the deposited beads
with excellent inter-bead bonding. SEM microstructures, shown in Fig. 6, provide clear
Page 11 of 21
differences in the defects observed in the present 3D printed parts. The lack of bonding between
the beads, in each layer, leading to the formation of characteristic inter-bead voids can be seen in
the pure TPC parts. Similar voids were also observed between the perimeters (inset of Fig. 6),
which resulted in relatively low density in the pure TPC parts (Fig. 4). Such voids were absent in
the composite parts reinforced with treated soybean hull fibers, as shown in Fig. 6 (10ST2),
while some small inter-bead voids were observed in 10AR-Shear parts. As discussed above, the
better flowability and wettability of the composite melt (with previous and adjacent deposits)
eliminated the formation of such gross voids. However, all composite parts revealed fine
spherical or elliptical gas porosity due to water vapor liberated from the natural fibers during
printing at 220°C. The amount of porosity found to decrease with chemical treatment of fibers
due to decrease in their moisture content, Table 1, compared to physically treated fibers.

Fig. 5 Typical light microstructures of 3D printed TPC and TPC-soybean hull fiber composites
showing potential printing induced defects at geometrically difficult to fill areas i.e., part corners
and transition region between the grip and the gauge length in tensile test coupons.

Page 12 of 21
Fig. 6 SEM microstructures showing the distribution of porosity, soybean hull fibers and printing
defects in different 3D printed parts. Arrows indicate pores in the composites.

The composite parts were further examined for gas porosity characteristics using SEM at high
magnification, Fig. 7. Correlating with the results of relative density, shown in Fig. 4, the dilute
acid hydrolysis of soybean hull fiber significantly reduced the amount of porosity and pore size
in these composites. Since the relative density of 10 wt. % composites is lower than that of
composites with 5 wt. % soybean hull fibers, we have quantitatively measured the pore size in
these composites to understand the differences. The experimental data revealed 81 ± 40 µm
pores in 10AR-Dry composites, which was reduced to 58 ± 21 µm after wet shear mixing of
fibers and finally to 39 ± 10 µm in the composites made using ST2 fibers. The pore size data and
the images of these composites, shown in Fig. 7, indicate large scatter in the pore size of AR-Dry
and AR-Shear composites. This is due to large variation in their size and non-uniform
distribution within the TPC matrix during printing. After dilute acid hydrolysis the soybean hull
fibers become finer and resulted in more uniform distribution in the matrix. Additional reduction
in the moisture content, Table 1, and hydrophilicity of chemically treated fibers [12] resulted in
significant reduction in the porosity and pore size after 3D printing. Both hydrolysis treatments

Page 13 of 21
(ST1 and ST2) found to provide comparable pore size at 10 wt.%. It is also important to note that
no visible interface between the successive beads could be seen in the present composite
microstructures, Fig. 7, which demonstrates complete and strong inter-bead bonding.

Fig. 7 Microstructures showing the influence of soybean hull fiber treatment on the porosity and
fiber distribution in FFF fabricated TPC-soybean hull fiber composites. Pores/voids are marked
with arrows and circles shows fibers or fiber bundles.

Although the printing induced voids and natural fiber induced gas porosity can influence the
mechanical properties of these composites, fiber-matrix interfacial characteristics play decisive
role, which depends on the hydrophilic and hydrophobic nature of fiber and matrix, respectively
[21,22]. For improved fiber-matrix bonding different physical and chemical surface modification
approaches have been used [22]. As a result of improved fiber-matrix bonding the load transfer
between the fibers and the matrix will be effective and therefore better mechanical properties can
be expected in such NFCs. Therefore, the fiber-matrix interfacial characteristics in the present
3D printed soybean hull fiber composites were analyzed and the microstructures are presented in

Page 14 of 21
Fig. 8. Composites made using AR-Dry fibers revealed visible gap between the fiber and the
TPC matrix. After wet shear mixing of as-received soybean hulls, the interface appears to be
improved. It can be seen from the inset of 10AR-Shear microstructure, shown in Fig. 8, that the
interfacial gap (indicated by arrow) is significantly reduced compared to that observed in the
10AR-Dry composites. The fiber-matrix interface in 10ST2 composites was found to be compact
with no visible gap. Complete and sound interfacial bonding can be clearly seen, indicated by
arrows, from the inset of 10ST2 in Fig. 8. These observed improvements in the interfacial
compatibility between the hydrophobic TPC matrix and the hydrophilic soybean hull fibers is
due to the reduction in their surface hydrophilicity [20] as a result of dilute acid hydrolysis. It is
well-established that during chemical treatments, such as the one used in this investigation, the
number of hydroxyl groups on the surface of natural fibers decreases significantly due to reaction
with these chemicals [12,19,20]. In addition, these treatments can remove lignin and other
insoluble waxes and impurities from the surface of these fibers. Therefore, the natural
hydrophilicity of the fibers decreases after chemical treatments leading to good bonding with the
hydrophobic polymer matrix [12]. The improved fiber-matrix interfacial characteristics in ST1
and ST2 composites expected to have positive influence on their mechanical properties
compared to AR-Dry and AR-Shear composites.

Fig. 8 SEM micrographs of the composites showing the fibers/fiber bundles – matrix interfacial
characteristics.

Page 15 of 21
3.3 Mechanical properties
Fig. 9 shows experimentally determined tensile mechanical properties of FFF fabricated pure
TPC and TPC-soybean hull fiber reinforced composites. It can be seen that the addition of
hydrolyzed soybean hull fiber to TPC can increase stress (at different strains), elastic modulus
and toughness, depending on the type of reinforcement and its concentration. To understand the
strengthening effect of different reinforcements, we have compared the tensile stress at 5% and
50% strain of the samples, as none of the samples fractured during tensile testing with tensile
elongation between 750 and 800%. The data shown in Fig. 9, indicate that the stress at 5% strain
of pure TPC increased from 1.42 ± 0.07 to 2.21 ± 0.22 MPa with the addition of hydrolyzed
soybean hull fibers (ST1 and ST2), which is an improvement up to 56% depending on the
concentration of fiber. With an increase in the tensile strain to 50%, the stress also increased to
11.13 ± 0.55 MPa (8.18 ± 0.55 for pure TPC). However, the improvement appears to be low (up
to 36%) compared to the improvement achieved at low strain (5%). Interestingly, the
strengthening effect was retained by the composites up to 50% strain with 5 wt.% hydrolyzed
soybean hull fibers. Under present 3D printing conditions, pure TPC exhibited an elastic
modulus of 36.3 ± 1.6 MPa and that of composite parts varied between 34 and 56 MPa.
However, the statistical analysis using Student's t-test revealed an improvement between 41 to
54% in the elastic modulus of 5ST1, 5ST2 and 10ST2. Highest elastic modulus between 54 and
56 MPa was recorded with ST2 composites. Similarly, the improvement in the toughness (area
under the stress-strain curve) of ST2 composites with 5 wt.% and 10 wt.% reinforcement was
found to be 30% and 15%, respectively. The drop in the toughness of 10ST2 composites can be
attributable to small decrease in the total elongation of these composites. Overall, these results
shows that the dilute acid hydrolysis of soybean hull fibers can significantly improve mechanical
properties of 3D printed TPC composites, compared to untreated fibers. Among the treatments,
double-stage hydrolysis (ST2) found to be best in improving the mechanical properties of these
composites. As shown in Fig. 8, the composites printed using hydrolyzed fibers (ST1 and ST2)
have compact/diffuse fiber-matrix interface (no sharp interface/gap), which enable effective load
transfer between the fiber and the matrix leading to observed improvement in the mechanical
properties. A further, significant reduction in the level of porosity in these composites, Fig. 7,
and pore size also assisted in achieving high mechanical performance in these composites.
Hydrolysis treatment reduced the hydrophilicity [12] and moisture content of soybean hull fibers,

Page 16 of 21
which improved the microstructural characteristics and thereby the mechanical properties.
Increasing the fiber concentration, from 5 wt.% to 10 wt.%, found to have marginal influence on
the mechanical properties and therefore further studies are required to assess the effect of fiber
concentration > 10 wt.%.

Fig. 9 Comparison of tensile mechanical properties of pure TPC with TPC-soybean hull fiber
composites showing the influence of fiber treatment and concentration. *p < 0.05 compared to
pure TPC.

The surface of 3D printed composites was examined after tensile testing for any damage,
especially the delamination, shear failure, etc., and typical surface morphologies are presented in
Fig. 10. Pure TPC parts showed intact layers and no delamination was observed. However, AR-
Dry composites revealed delamination of layers at several places, as shown in Fig. 10. The
bonding between the perimeter and infill was also failed in these composites. On the other hand,
it can be seen from Fig. 10, that all layers are intact and delamination or shear between the layers
was absent in the composites reinforced with hydrolyzed soybean hull fibers. The surface
morphology is very similar to that found in as-printed condition, Fig. 2. These features clearly

Page 17 of 21
demonstrate that the bonding between the layers and between the beads is very strong. Weak
interfacial bonding could lead to debonding during tensile testing and as a result the stress
experienced by each layer would be different. High-magnification SEM images, shown in the
insets of Fig. 10, also did not reveal any delamination when treated fibers were used. However,
due to high tensile stretching, some of the exposed fibers on the surface of these composite
samples were torn off the matrix. From these observations it is concluded that the 3D printed
composites have strong bonding between successive layers and adjacent beads, especially ST1
and ST2 composites.

Fig. 10 Typical surface features (along the build direction) observed on the composite parts after
tensile testing.

4. Conclusions
Fused filament fabrication (FFF) was used to print thermoplastic copolyester (TPC) composite
parts reinforced with soybean hull fibers, in different conditions, and their microstructural,
mechanical properties were compared with pure TPC parts. The surface morphological features
of printed parts were found to depend on the material flow during printing and gross defects were
observed in pure TPC parts due to its higher viscosity than that of the composites reinforced with
soybean hull fibers. The composite parts also exhibited better layer-to-layer and bead-to-bead
bonding. Dilute acid hydrolysis of soybean hull fibers significantly reduced the surface
roughness of the 3D printed composite parts. Moreover, the use of hydrolyzed fibers resulted in
considerable reduction in the gas porosity (from 11 to 1 %) and their size (from 81 µm to 39
µm). High-magnification microstructural analysis of the composites made with hydrolyzed fiber

Page 18 of 21
revealed compact and diffuse fiber-matrix interfaces. These microstructural characteristics found
to improve the elastic modulus (up to 54%) and stress at 50% strain (up to 38%) of TPC when
reinforced with 10 wt.% of two-stage hydrolyzed soybean hull fibers. Our results demonstrate
that the appropriate chemical treatment of soybean hull fibers makes them as a potential
reinforcement to manufacture variety of natural fiber polymer composites using 3D printing.

Declarations of interest: none

Acknowledgements
Authors acknowledge the financial support from United Soybean Board, MO, USA (Contract
No. USB#1940-362-0703-E).

References:
[1] A. Bandyopadhyay, K.D. Traxel, C. Koski, S. Bose, Additive Manufacturing of Polymers,
in: A. Bandyopadhyay, S. Bose (Eds.), Addit. Manuf., 2nd ed., CRC Press, 2019: pp. 25–
42. doi:10.1201/9780429466236-2.
[2] S. AMIT BANDYOPADHYAY, VAMSI K. BALLA, SHELDON A. BERNARD, BOSE,
MICRO-LAYERED MANUFACTURING, in: M. Koc, T. Özel (Eds.), MICRO-
MANUFACTURING Des. Manuf. Micro-Products, John Wiley & Sons, Inc, 2011: pp.
97–158.
[3] X. Wang, M. Jiang, Z. Zhou, J. Gou, D. Hui, 3D printing of polymer matrix composites: A
review and prospective, Compos. Part B Eng. 110 (2017) 442–458.
doi:10.1016/j.compositesb.2016.11.034.
[4] P. Parandoush, D. Lin, A review on additive manufacturing of polymer-fiber composites,
Compos. Struct. 182 (2017) 36–53. doi:10.1016/j.compstruct.2017.08.088.
[5] S. V. Joshi, L.T. Drzal, A.K. Mohanty, S. Arora, Are natural fiber composites
environmentally superior to glass fiber reinforced composites?, Compos. Part A Appl. Sci.
Manuf. 35 (2004) 371–376. doi:10.1016/j.compositesa.2003.09.016.
[6] Grand View Research, Natural Fiber Composites (NFC) Market Size, Share & Trends
Analysis Report By Raw Material, By Matrix, By Technology (Injection Molding,
Compression Molding, Pultrusion), By Application, And Segment Forecasts, 2018 – 2024,
n.d. https://www.grandviewresearch.com/industry-analysis/natural-fiber-composites-
market.
[7] C. Merlini, V. Soldi, G.M.O. Barra, Influence of fiber surface treatment and length on
physico-chemical properties of short random banana fiber-reinforced castor oil
polyurethane composites, Polym. Test. 30 (2011) 833–840.
doi:10.1016/j.polymertesting.2011.08.008.
[8] M. Jawaid, H.P.S.A. Khalil, A. Hassan, R. Dungani, A. Hadiyane, Effect of jute fibre
loading on the mechanical and thermal properties of oil palm-epoxy composites, Compos.
Part B Eng. 45 (2013) 619–624. doi:10.1177/0021998312450305.

Page 19 of 21
[9] X.V. Cao, H. Ismail, A.A. Rashid, T. Takeichi, T. Vo-Huu, Maleated Natural Rubber as a
Coupling Agent for Recycled High Density Polyethylene/Natural Rubber/Kenaf Powder
Biocomposites, Polym. - Plast. Technol. Eng. 51 (2012) 904–910.
doi:10.1080/03602559.2012.671425.
[10] R.A. Majid, H. Ismail, R.M. Taib, Effects of Polyethylene-g-maleic Anhydride on
Properties of Low Density Polyethylene/Thermoplastic Sago Starch Reinforced Kenaf
Fibre Composites, Iran. Polym. J. 19 (2010) 501–510.
[11] M. Jacob, B. Francis, S. Thomas, K.T. Varughese, Dynamical mechanical analysis of
sisal/oil palm hybrid fiber-reinforced natural rubber composites, Polym. Compos. 27
(2006) 671–680. doi:10.1002/pc.20250.
[12] V.K. Balla, K.H. Kate, J. Satyavolu, P. Singh, J.G. Dattatreya Tadimeti, Additive
manufacturing of natural fiber reinforced polymer composites: Processing and prospects,
Compos. Part B Eng. (2019) 106956. doi:10.1016/J.COMPOSITESB.2019.106956.
[13] A. Le Duigou, M. Castro, R. Bevan, N. Martin, 3D printing of wood fibre biocomposites:
From mechanical to actuation functionality, Mater. Des. 96 (2016) 106–114.
doi:10.1016/j.matdes.2016.02.018.
[14] M. Kariz, M. Sernek, M. Obućina, M.K. Kuzman, Effect of wood content in FDM
filament on properties of 3D printed parts, Mater. Today Commun. 14 (2018) 135–140.
doi:10.1016/j.mtcomm.2017.12.016.
[15] A. Alemdar, M. Sain, Isolation and characterization of nanofibers from agricultural
residues - Wheat straw and soy hulls, Bioresour. Technol. 99 (2008) 1664–1671.
doi:10.1016/j.biortech.2007.04.029.
[16] A. Ferrer, C. Salas, O.J. Rojas, Physical, thermal, chemical and rheological
characterization of cellulosic microfibrils and microparticles produced from soybean hulls,
Ind. Crops Prod. 84 (2016) 337–343. doi:10.1016/J.INDCROP.2016.02.014.
[17] A. Ferrer, C. Salas, O.J. Rojas, Dewatering of MNFC containing microfibrils and
microparticles from soybean hulls: mechanical and transport properties of hybrid films,
Cellulose. 22 (2015) 3919–3928. doi:10.1007/s10570-015-0768-y.
[18] D.A. Fonseca, R. Lupitskyy, D. Timmons, M. Gupta, J. Satyavolu, Towards integrated
biorefinery from dried distillers grains: Selective extraction of pentoses using dilute acid
hydrolysis, Biomass and Bioenergy. 71 (2014) 178–186.
doi:10.1016/J.BIOMBIOE.2014.10.008.
[19] J. George, M.S. Sreekala, S. Thomas, A review on interface modification and
characterization of natural fiber reinforced plastic composites, Polym. Eng. Sci. 41 (2001)
1471–1485. doi:10.1002/pen.10846.
[20] K.L. Pickering, M.G.A. Efendy, T.M. Le, A review of recent developments in natural
fibre composites and their mechanical performance, Compos. Part A Appl. Sci. Manuf. 83
(2016) 98–112. doi:10.1016/j.compositesa.2015.08.038.
[21] A.K. Mohanty, M. Misra, G. Hinrichsen, Biofibres, biodegradable polymers and
biocomposites: An overview, Macromol. Mater. Eng. 276–277 (2000) 1–24.
doi:10.1002/(SICI)1439-2054(20000301)276:1<1::AID-MAME1>3.0.CO;2-W.
[22] G. Bogoeva-Gaceva, M. Avella, M. Malinconico, A. Buzarovska, A. Grozdanov, G.
Gentile, et al., Natural fiber eco-composites, Polym. Compos. 28 (2007) 98–107.
doi:10.1002/pc.20270.

Page 20 of 21
 LIST OF TABLE CAPTIONS
Table 1 Different composites and FFF parameters used in the present investigation.

LIST OF FIGURE CPATIONS

Fig. 1 SEM images showing the typical top surface (X-Y direction) topographical features of
TPC and TPC-soybean hull fiber composites fabricated using FFF. Arrows indicate the interface
between the beads and some defects at these regions. Circles in the in-set microstructures show
complete bonding between adjacent beads.
Fig. 2 Typical build direction surface morphology of TPC and its composites. Arrows indicate
interlayer interface/defects and exposed soybean hull fibers. In-sets show bonding between the
layers at high magnification.
Fig. 3 Average surface roughness of pure TPC and TPC-soybean hull fiber composites (a) Top
surface (X-Y direction), (b) Build direction (X-Z direction).
Fig. 4 Influence of soybean hull treatment and concentration on the relative density of 3D printed
TPC composites.
Fig. 5 Typical light microstructures of 3D printed TPC and TPC-soybean hull fiber composites
showing potential printing induced defects at geometrically difficult to fill areas i.e., part corners
and transition region between the grip and the gauge length in tensile test coupons.
Fig. 6 SEM microstructures showing distribution of porosity, soybean hull fibers and printing
defects in different 3D printed parts. Arrows indicate pores in the composites.
Fig. 7 Microstructures showing the influence of soybean hull fiber treatment on the porosity and
fiber distribution in FFF fabricated TPC-soybean hull fiber composites. Pores/voids are marked
with arrows and circles shows fibers or fiber bundles.
Fig. 8 SEM micrographs of the composites showing the fibers/fiber bundles – matrix interfacial
characteristics.
Fig. 9 Comparison of tensile mechanical properties of pure TPC with TPC-soybean hull fiber
composites showing the influence of fiber treatment and concentration. *p < 0.05 compared to
pure TPC.
Fig. 10 Typical surface features (along the build direction) observed on the composite parts after
tensile testing.

Page 21 of 21
 Research Highlights

• Soybean hull derived fibers’ potential as reinforcements in polymers is evaluated.

• Thermoplastic copolyester composites were made using fused filament fabrication.

• Surface quality, print defects and layer bonding depend on the fiber treatment.

• Hydrolyzed fibers reduced the porosity and printing defects in the composites.

• Diffuse fiber-matrix interface characteristics improved the mechanical properties.

Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests:

Vamsi Krishna Balla Jogi Ganesh Dattatreya Tadimeti

Kunal H Kate Jagannadh Satyavolu