Additive Manufacturing 61 (2023) 103315

Contents lists available at ScienceDirect

Additive Manufacturing
journal homepage: www.elsevier.com/locate/addma

A novel method and printhead for 3D printing combined nano-/microfiber

solid structures
Yahya Kara a, *, Norbert Krisztián Kovács a, b, Péter Nagy-György c, Róbert Boros a,
Kolos Molnár a, d, **
a
Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Polymer Engineering, Műegyetem rkp. 3., H-1111 Budapest,
Hungary
b
MTA-BME Lendület Lightweight Polymer Composites Research Group, Műegyetem rkp. 3., H-1111 Budapest, Hungary
c
Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Hydrodynamic Systems, Műegyetem rkp. 3., H-1111 Budapest,
Hungary
d
ELKH–BME Research Group for Composite Science and Technology, Műegyetem rkp. 3., H-1111 Budapest, Hungary

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

Keywords: In this study, we demonstrate a novel 3D printing method and printhead capable of 3D printing objects, which
Nanofiber are combinations of nano-/microfibers and solid or infill layers. The method is a unique combination of material
Nonwoven extrusion and melt-blowing: the filament can be fused and then either deposited or turned into ultrafine fiber
Material extrusion
mats with the aid of hot pressurized air. This study introduces the operation of the prototype device, the
Airflow
Hierarchical structures
computational fluid dynamics (CFD) simulations of the airflow field, and various tests on the generated speci­
mens. We were able to make layers of ultrafine fiber mats from poly lactic acid (PLA), which adhered to the solid
deposited layers of the same PLA material. Higher air speed and smaller nozzle size resulted in smaller fibers.
Fibers with a diameter of as small as 300 nm and an average of 900 ± 100 nm were generated at supersonic air
velocities. The fiber layers enhance the crystallinity of the printed products, and can reinforce those. We can
control the position and ratio of the fibers and the solid (or infill) parts of the printed object, so even porous
objects can be generated. These hierarchical 3D printed objects have lots of possible applications, such as
scaffolds, controlled drug delivery devices, filter media, structural composites and smart textiles.

1. Introduction maintenance [7]. In 2015, the FFF or material extrusion AM was defined
by ISO/ASTM 52900 as "An additive manufacturing process in which ma­
The interdisciplinary expansion of additive manufacturing (AM) has terial is selectively dispensed through a nozzle or orifice" [8]. The FFF
led to a growing demand for suitable materials and technologies [1]. consists of melting the thermoplastic feedstock in the print head (i.e.,
Even though the COVID-19 pandemic caused a significant slowdown, extruder assembly), and then the molten polymer is pressurized through
the worldwide AM market for related products and services grew by the nozzle at a desired volumetric flow. FFF technology was invented in
7.5% in 2020 [2]. Recent reports revealed that the AM industry was 1988 by Scott and Lisa Crump. In 1989, Crump patented [9] the first FFF
expanded by 19.5% in 2021 [3]. The distinct advantages, including but method called fused deposition modeling (FDM). Crump then created
not limited to: reduction in manufacturing cost and process downtime, the software process that converts Standard Tessellation Language (STL)
less waste, straightforward and single-step production, reliability and files into another format to slice sections of the 3D model and determine
ease in developing customized engineering products, accelerated the how the layers will be printed. After Crump’s invention, today’s
extension of the AM industry and related end-products [4–6]. well-known companies like Stratasys, General Electric, HP, 3 M, 3D
Fused filament fabrication (FFF) is one of the most popular AM Systems, Markforged, Boeing, EOS, Xerox, and many others actively
methods due to its versatility, good value for money, and ease of advanced and industrialized the FFF and related 3D printed parts for

* Corresponding author.
** Corresponding author at: Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Polymer Engineering,
Műegyetem rkp. 3., H-1111 Budapest, Hungary.
E-mail addresses: [email protected] (Y. Kara), [email protected] (K. Molnár).

https://doi.org/10.1016/j.addma.2022.103315
Received 22 October 2022; Received in revised form 11 November 2022; Accepted 22 November 2022
Available online 24 November 2022
2214-8604/© 2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 1. Schematics of the developed FFF apparatus (a) A half-section view (1: heat sink, 2: nozzle, 3: air tube connector, 4: heat sink, 5: throat heat break, 6: throat
heat break, 7: air knife module, 8: heated block, 9: O-ring, 10: air tube, 11: air channel, 12: assembly connectors. 13: bolt (leveling)), the demonstration of (b)
filament fusion (FF) and (c) fiber generation (FG) mode operations.

years [2,10,11]. 3D textile structures.

To date, many efforts have been paid to advance the manufacturing To date, several attempts have been made to commercialize 3D
of complex structures via FFF 3D printing [12,13]. Nevertheless, push­ printing textiles. A UK-based company called TamiCare developed an
ing the limits toward generating ultrafine fibers and integrating them AM-based process called CosyFlex [26]. Their technology involves
with printed structures via the FFF remains a great challenge. In the last customized 3D printing with textile fibers manufactured in layers by a
two decades, many research activities have been done to develop stra­ spray jet. Another company from the USA, Electroloom, used AM and
tegies for FFF printing textile structures like nonwovens and nanofibers electrospinning methods to generate 3D-printed textiles [27]. Electro­
[14,15]. In this regard, novel methods like melt electrospinning writing loom’s approach used traditional electrospinning to generate fiber mats
(or melt electrowriting) (MEW) were developed using the fundamentals deposited onto a 3D mold (e.g., similar to RFACS). Although generating
of AM combined with electrospinning [16,17]. It is rooted in traditional fibers via AM became real in the industry quickly, these companies fell
electrospinning, which is a technique that involves stretching a polymer short of commercialization expectations. This is because of the com­
solution or melt into nanofibers under a high electric field and accu­ plexities and difficulties in implementing the proposed technology,
mulating it on the counter electrode in the form of a randomly oriented being only available for fiber making, and high operational costs.
fiber web [18,19]. In MEW, the molten polymer is electrostatically Creating complex structures via FFF consisting of nano-/microfibers
drawn into the nano-/micro fibers towards a computer-controlled col­ and 3D printed struts has followed a separate multi-step process [28,29].
lector plate in a well-confined path with fairly good resolution. How­ In these methods, a fibrous structure was produced, for example, via
ever, the need for sophisticated instrumentation and operation electrospinning and then incorporated into 3D print layers. Kozior et al.
difficulties hamper MEW’s applicability [20]. In addition, generating [30] studied a sequential spin-on method involving electrospinning in
complex structures consisting of 3D print struts and fibers via MEW is combination with FFF and material jetting (using photo-cured liquid
not achieved so far. Farer et al. [21,22] took an innovative approach to polymer resins, a.k.a. PolyJet Matrix) to develop structures for medical
fabricate seamless textiles by robotic fiber dispensing and assembly applications. In their study, 3D-printed rectangular prisms are produced
control. The method is rooted in melt blowing, for which the polymer by FFF and material jetting, and then electrospun fibers are deposited
melt is extruded through a die containing numerous small capillaries onto them. Another method comprises alternating and repeating
and then stretched via a jet of hot air [23]. The ratio of high-velocity air layer-by-layer material (e.g., hydrogel inks) dispensing combination
versus the lower velocity polymer provides a drag force that rapidly with the insertion of prepared electrospun fiber mats between the layers
attenuates the forming fibers [24,25]. The robotic fiber assembly and during the printing process [31]. Yu et al. [32] proposed a novel
control systems (RFACS) method aimed to deposit fiber mats onto a 3D approach composed of infusing electrospun nanofibers into 3D-printed
collector to form a randomly aligned fine fiber web. Farer and coworkers meshes to make biocompatible composite scaffolds. Xu et al. [33]
[22] produced PP fibers with a traditional extruder and a robotic control developed a unique bioprinting technique that uses tandem construction
system with an average diameter of around 10 µm. However, this of material jetting (a.k.a. inkjet printing) and electrospinning head.
method only allows generating fine fibers via robotic control, and it also Several research studies confirmed that incorporating fiber mat into a
requires expensive machinery and complex processing control to create 3D-printed piece enhances the mechanical performance of the

2
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 2. Optical images of FFF printer assembly (a) Velleman 8200 FFF printer with the developed apparatus, (b) close look up to the developed apparatus.

3D-printed hierarchical composites [1,28,34]. However, the difficulties 2. Materials and method
in the processing (e.g., incorporating a delicate fibrous layer), high cost,
and low precision are some obstacles that hinder the applicability of 2.1. Method description
creating hierarchical 3D printed composites via FFF. The current efforts
are far from integrating these fibrous substrates’ in a cost-efficient, We invented a new approach and a FFF apparatus that generates
high-speed and -throughput production system. The other issue is that nonwoven layers of nano-/micro fibers and fabricates hierarchical
generating such textile structures via AM is either expensive or requires structures via FFF technology and patented it recently [38]. The print­
advanced hardware, and besides, processing difficulties occur that head can generate both ultrafine fibers and struts. A half-section view of
hinder their wide use. Many approaches have been reported that are the apparatus is shown in Fig. 1(a). The opening between the heated
inaccessible to everyday users as these methods require either unique block (8) and the air knife (7) can be adjusted by tightening and loos­
and expensive materials, machinery or customized hardware and soft­ ening bolts (Fig. S1). The opening between the heated block and the air
ware beyond conventional FFF printers [20,35]. Although some efforts knife allows the positioning of the nozzle inside or outside of the air
have been devoted to FFF printing textiles structures (e.g., knitted knife module, as shown in Fig. 1(b-c) (Figs. S2 and 3). A heat-resistant
meshes), the properties of those structures do not fulfill fundamental O-ring (9) avoids air leakage during the operation. When the nozzle is
textile parameters, such as being light, porous, oriented, flexible, strong, inside the air knife and pressurized air is applied, the apparatus works in
permeable, etc. [36,37]. The ability to generate fine fiber structures and the continuous fine fiber generation (FG) mode, while filament fusion
3D printed struts with a single apparatus would possibly fit in numerous (FF) mode takes place when the nozzle is placed outside. In the proto­
markets in automotive, aerospace, defense, medical, pharmaceutical, type system, the screws are set manually; however, this motion can be
textile, etc. Yet, no process or tool was reported to generate fine fibers automated. Based on the FFF printing nozzle (2) geometry and air exit
via the AM method and implement it into 3D printing with a single tool. annular orifice (Ø2.5 mm), the opening distance between the air knife
Furthermore, creating complex FFF printed 3D structures with contin­ module and the heated block was chosen as a minimum of 1 mm for
uous nano-/submicron fibers directly from the feedstock (e.g., molten smooth operation.
polymers) remains unfolded. The developed apparatus does not require too expensive hardware;
In this study, we introduced a novel FFF method that combines for example, the throat tube, brass nozzle and heat sink are commer­
continuous ultrafine fibers and 3D print struts within a single object via cially available for FFF printers. Just like in classical FFF printing, the
a single apparatus. We demonstrated the applicability of the developed filament is fed from a spool, passes through throat tubes (5,6), and
apparatus using a commercial desktop printer. We detailed the method reaches the moving heated block (8). The molten polymer is forced out
and the respective FFF technology by testing the continuous fiber of the nozzle (2). The molten polymer meets with hot air at the nozzle
manufacturing and FFF printing struts. We first analyzed the fiber for­ exit in the air knife module (7). The hot air fed through air tubes (10)
mation mechanism and detailed the airflow field using computational passes vertical air slots (11) located in the heated block and reaches the
fluid dynamics (CFD) simulations. Then, we investigated the effect of the nozzle tip (2) (Fig. 1(b-c)). With the developed apparatus, a commercial
airflow field on the fiber-generating process, the resultant fiber FFF printer with the necessary modifications can be used for both fiber
morphology, and thermal and mechanical properties. We tested novel making and filament fusion (Fig. 2, Figs. S3–6).
apparatus applicability for classical FFF 3D printing operation. Finally, The novel apparatus consists of two major parts: the air knife module
hierarchical structures comprising fine fibers and 3D-printed struts were (7) and the heated block (8). The air knife module (7) has a conic ge­
3D printed and evaluated using SEM. Results revealed that the novel ometry with a cone opening half-angle (α) that generates a high drawing
method of generating 3D print struts and nano-/micro continuous fibers force (Fig. 1). The air gap between the heated block and the nozzle
with one single printing head could complement their respective represented by the α angle affects the air outlet velocity, the air tem­
advantages. perature and the fiber attenuation rate. When the α angle is large, it
results in fiber separation or high turbulence. On the other hand, a lower
α angle yields roped fibers and loosely coiled bundles, which are un­
desirable [39,40]. Therefore, the cone opening half-angle of 35◦ and 60◦

3
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 3. Illustration of the operating modes; (a) Optical images of the FF mode (VIDEO 1), (b) printed object and SEM image of the struts and (c) optical image of FG
process with the novel apparatus and SEM image of the generated nanofibers (VIDEO 2).

(i.e., air knife) was chosen to investigate for generating continuous fi­
bers. The apparatus possesses a straightforward design that allows
switching between FF and FG modes. The opening between the air knife
module (7) and the heated block (8) can be adjusted by leveling bolts
(Fig. S1). Therefore, hierarchical structures can be generated by FF and
FG modes of operation continuously. The developed method and
apparatus facilitate generating 3D printed parts and fine fibers via FFF
technology, as shown in Fig. 3.
Supplementary material related to this article can be found online at
doi:10.1016/j.addma.2022.103315.
Unlike traditional extrusion-based fiber-making methods such as
melt-blowing technology, the developed apparatus generates fibers from
thermoplastic filament extruded through a heated short length (e.g.,
35 mm) without using an extrusion screw (Fig. 3). The apparatus uses a
FFF filament as the raw material as it is used in FFF 3D printing. The
apparatus can generate fibers without an extruder screw (i.e., traditional
extrusion-based fiber spinning methods require an extruder with a
screw), and this extinct difference makes the method novel, while it
allows FFF print structures (FF mode) from the same material used for
FG mode. This feature also allows FFF printing of structures and fine
fibers without employing expensive hardware or software (Figs. S3 and
4). Introducing a functional nano-/submicron fiber mat via FFF printing
of the filament layers can bring potential benefits, such as a more flex­
ible and cost-effective production of advanced engineering structures
via FFF. For example, a potential application can be tissue engineering, Fig. 4. The quarter model of the flow field and the environment with the
where the scaffold’s desired mechanical performance and custom ge­ corresponding boundary conditions.
ometry can be ensured by the solid parts of the body, while the cells can
adhere better to the fiber structure [41,42]. real-time experiments.

2.2.1. Meshing
2.2. Ansys airflow field simulations
Fig. 4 depicts the CFD model’s geometry data and boundary condi­
tions. One-quarter of the whole geometry was modeled using symmetry
Several attempts have been devoted to analyzing the free jet using
simplification to reduce the computational cost. The structured mesh
various methods such as CFD [43], numerical stability analysis [44,45]
had hexahedra elements in the inflation layer near the wall and tetra­
and experimental measurements [46]. In this study, we chose the CFD
hedra cells inside the flow domain (Fig. S7). The inflation layer was
technique, which is validated by experimental measurements due to its
0.2 mm in width and contained 5 layers with a growth rate of 1.2. The
general applicability and high accuracy [47–50]. We investigated the
mesh is refined inside the apparatus and in the vicinity of the apparatus
airflow field formed in FG mode by CFD simulations performed with
orifice.
Ansys CFX 2019R3 software. The same material properties, boundary
To eliminate the discretization error, mesh independence analysis
conditions and numerical settings were used as they were in the

4
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Table 1 2.2.4. Air temperature and flow rate measurement

The summary of the boundary conditions. To verify the airflow field CFD model, we measured and predicted
Boundary Validation of Validation of Experimental setup the effects of air knife design and air pressure on velocity and temper­
condition pressure loss temperature field parameters ature distributions. We conducted experiments to measure the airflow
Inlet Tin = 20 ◦ C Tin = 250 ◦ C Tin = 250 ◦ C rate at the apparatus inlet and the air temperature after air exits the
pin = 0.4–2.4 bar pin = 1;1.5 pin = 1;1.5;2 bar heated block. We measured the temperature in-line, with a K-type bare
Apparatus Adiabatic thermocouple and a digital thermometer (Testo 830-T2, Germany). The
inside walls No-slip thermocouple was placed parallel to the apparatus to read the air
Apparatus Twall = 20 ◦ C Twall = 200 ◦ C Twall = 200 ◦ C No-
outside walls No-slip No-slip slip
centerline temperature. With this aim, the thermocouple was set and
Fiber collector Tplane = 20 ◦ C fixed to variable distances (5–40 mm) from the apparatus during the
plane No-slip experiments. The temperature was recorded when the read value
Environment Tenv = 20 ◦ C reached a stable region, e.g., ± 1 ◦ C. A thermal camera (A325sc, FLIR
penv = 0 bar
Systems Inc., Wilsonville, OR, USA) was also used to monitor apparatus
temperature during the operation. A digital flowmeter (flow switch,
was performed with three different mesh sizes. We tested 1456k, 2998k, PFMB721-C8-F, SMC, Japan) was used to measure the air volume flow
and 6782k elements for the air knife with a half-cone opening angle of rate. The measurement setup used is shown in Fig. S8.
60◦ , and we compared the obtained air flow rate values to the mea­ We measured the volume flow rate and the temperature field sepa­
surement data. The results show that the relative difference in the ob­ rately, because the flowmeter used was unsuitable for operating at high
tained values was less than 1% in the case of these three meshes. temperatures. Firstly the air volume flow rate without heating was
Considering simulation time and the small difference between the measured at room temperature. The inlet air pressure was set in the
simulation and experimental results, the mesh with 2998k cells was 0.4–2.4 bar range using a pressure regulator (MS2-LFR, Festo, Ger­
chosen for further simulations. many), and the corresponding flow rates were recorded. According to
the datasheet of the flowmeter, the recorded volume flow rate values
2.2.2. Boundary conditions correspond to the standard condition ( T = 15 ◦ C, p = 101.3 kPa), where
The heated and pressurized air flows into the apparatus at the inlet the density is ρairstd = 1.225 kg/m3. The recorded air volume rate value
described with a pin relative pressure and Tin temperature. It is assumed was converted to the mass flow rate ṁmeas using Eq. (1).
that there is only a short time for heat exchange due to the high veloc­ ṁmeas = qairstd ρairstd (1)
ities; therefore, the heat transfer inside the nozzle was neglected. The
apparatus inside walls were modeled with adiabatic no-slip walls. Based where qairstd is the measured volume flow rate corresponding to the
on the empirical observations, the outer wall temperature was assumed standard conditions. The relative error between the CFD simulations and
to be Twall = 200 ◦ C and Twall = 20 ◦ C for the heated and non-heated the measured flow rate was calculated using Eq. (2).
cases, respectively. The environment was represented by the opening
boundary condition with Tenv = 20 ◦ C temperature and penv = 0 Pa ε=
|ṁmeas − ṁCFD |
(2)
entrainment pressure. The fiber collector plane was considered with a ṁmeas
20 ◦ C, no-slip wall.
where ṁCFD is the mass flow rate calculated via CFD computations.
For validation purposes, two independent cases were considered. In
the first case, the volume flow rate was calculated when the inlet tem­
2.3. Materials
perature was equal to ambient, Tin = Tenv = 20 ◦ C, and the pressure
varied between 0.4 and 2.4 bars. In that case, the outer wall of the
A PLLA homopolymer (Total Carbion, Luminy® L105, MFI: 70 g/
apparatus was Twall = 20 ◦ C. In the second case, the temperature dis­
10 min at 190 ◦ C, 2.16 kg) granules were used to prepare filaments for
tribution was determined when Tin = 250 ◦ C and pin= 1 and 1.5 bar. For
FFF 3D printing. The PLLA granules was dried in a WGL 45B oven
that, the outer apparatus wall temperature was Twall = 200 ◦ C. Finally,
(Huanghua Faithful Instrument, China) at 100 ◦ C for 4 h before
these two cases were combined to validate experimental observations,
processing.
for which pin = 1;1.5;2 bar, Tin = 250 ◦ C, and Twall = 200 ◦ C values
were applied.
Table 1 summarizes the applied boundary conditions. 2.4. FFF filament production

2.2.3. Material properties and turbulence model We made filaments from PLLA granules mentioned above to generate
The air was considered to be an ideal gas with μ = 1.831⸱10− 5 Pa⸱s fibers and 3D print structures with the same device. PLLA filaments
dynamic viscosity, λ = 0.0261 W/(mK) thermal conductivity and cp (filament diameter 1.75 mm) were produced with a Precision 450 type
= 1004.4 J/(kgK) specific heat (at constant pressure) at 25 ◦ C reference filament maker (3devo B.V., The Netherlands). The desktop extruder
temperature. Because the computational domain contains narrow (in­ consists of 4 heating zones. The filament was extruded through the
side the apparatus) and large far-field (outside the apparatus) regions, nozzle and then cooled by multiple ventilators after the nozzle. The
the Shear Stress Transport (SST) turbulence model was applied with an filament diameter was computer controlled by DevoVision App provided
automatic wall function. As a result, a set of equations for the continuity, by the desktop extruder producer. The filament diameter was measured
energy, momentum, turbulent eddy frequency and turbulence kinetic by an optical sensor before it was wound up onto the spool. The tem­
energy were solved. perature profile was set to 185–210–195–185 ◦ C (from hopper to die).
Since the real application is time-independent, steady-state simula­ The extruder screw rotation speed was set to 2.4 rpm.
tions were performed with high-resolution advection and first-order
turbulence schemes. The convergence criteria were set to 10− 5, while 2.5. Fused filament fabrication (FFF) device and experimental setup
the volume flow rate was monitored. The simulation stopped when the
convergence criteria and the steady-state solution of the volume flow A Velleman K8200 desktop FFF printer (Velleman vn., Belgium) was
rate were reached. used to test the developed apparatus (Fig. 2 and Fig. S3). Two K-type
thermocouples, one mounted to the air knife module (7) and the other
mounted to the heated block (8), and a PID control unit was used to heat
up and control the apparatus temperature, respectively. We used the

5
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 5. Comparison of the CFD and experimental air temperature data at the applied air pressure of (a) 1 bar and (b) 1.5 bar.

open-source Printrun 2. X software (Pronterface, GNU General Public glass transition temperature (midpoint temperature) samples of samples
License) for controlling the FFF printer via computer. according to ASTM E1356–08 [52].
In the FG mode, the apparatus temperature was set to a constant
200 ◦ C. The desired air temperature (250 ◦ C) was provided by an inline 2.8. Producing 3D printed samples with the novel apparatus
heater (AHPF-082, Omega, UK) and a K-type inline nozzle thermocouple
and a PID controlling unit. The distance between the apparatus and the We tested the developed apparatus’s capability to create hierarchical
printing bed was set constant at 100 mm for fiber generating, while for structures, for which we produced FFF printed samples using FF and FG
FF mode it was 50 µm. The filament feeding speed was set to 1 mm/min. modes. The accuracy between the CAD model and printed parts was also
The nozzle diameter of 0.2, 0.4 and 0.5 mm were tested. The influence of investigated [53,54]. We used SolidWorks 16 (Dassault Systèmes,
three air pressure levels (1, 1.5 and 2 bar) and two air knife angles France) to design the 3D model sample with dimensions of 80 × 10×1.5
(35◦ and 60 ◦ ) on the fiber properties were tested. [mm] (length x width x thickness). The 3D model was converted to the ".
STL" file format using SolidWorks 16. Then, CraftWare 1.21 slicing
2.6. Tensile tests software (Craftbot, Hungary) was used to construct the model based on
the settings of the 3D printer (Fig. S9). The same printer and software
We prepared rectangular fiber mat samples in 40 mm × 10 mm size. described in Chapter 2.5 were used to 3D print samples. The apparatus
The fiber mat length and width were measured using a micrometer and printing bed temperatures were set to 200 ◦ C and 60 ◦ C, respec­
(Louis Schopper Leipzig, Germany) with 0.01 mm precision. We tively. The nozzle diameter and filament drawing speed were 0.2 mm
weighed and recorded the PLA fiber mats using a Sartorius Quintix and 60 mm/s, respectively. Infill density was set at 25%. The 3D printed
125D–1CEU (Sartorius, Germany) semi-micro scale. We calculated the samples were compared in terms of geometry and precision with the 3D
fiber mat area by using Eq. (3). CAD model. We produced 5 samples and measured the 3D-printed
mfm sample’s length, thickness and width with a precision of 0.01 mm
A= (3) using a digital caliper (Fowler Promax, USA).
lfm ρPolymer

where, mfm is the fiber mat mass, lfm is the length of the fiber mat and 2.9. Scanning electron microscopy (SEM)
ρPolymer is the polymer bulk density (e.g., PLA=1.24 g/cm3).
The tensile properties of the fiber mats were tested at room tem­ The samples’ morphology was observed using scanning electron
perature with a Zwick Z005 (Zwick, Germany) type universal tensile microscopy (SEM; JEOL 6380 LA, Japan). Samples were pasted onto
tester equipped with a 20 N load cell. The gauge length and tensile speed metallic studs with double-sided conductive tape. The sample surface
were 20 mm and 10 mm/min, respectively. The testing routine was was finely coated using a JEOL JFC-1200 (Jeol Ltd., Japan) fine coater
performed 6 times for each sample group. with gold (Au) in order to avoid their charging. We measured 100 fibers
for each sample to analyze the fiber diameter distributions. We used
ImageJ 1.51k software for this measurement.
2.7. Differential scanning calorimetry (DSC) tests

3. Results and discussion

The thermal properties of the fiber mats and the filament were
studied by differential scanning calorimetry (DSC) with a Q2000 DSC
3.1. Analyzing the airflow characteristics
(TA Instruments, USA) device. The heat-cool-heat tests were performed
in an inert atmosphere (N2; 50 ml/min purge flow rate) in a temperature
3.1.1. Validation of the CFD model
range of 0–250 ◦ C with a heating and cooling rate of 5 ◦ C/min. The
Since the developed method possesses aerodynamically-driven fiber
degree of crystallinity (χ ) of the samples was determined based on Eq.
stretching, the motion of the polymeric jet and the fiber characteristics
(4). Based on ref. [51] 97 J/g was taken as the heat of fusion of the 100%
are assumed to be rooted in the aerodynamics of the airflow field
crystalline (ΔH0m ) PLA in the analysis (Eq. (4)):
(temperature, velocity, turbulence intensity, etc.) [47,55]. Therefore,
ΔHm − ΔHcc computational fluid dynamics (CFD) simulations were employed in the
χ= × 100[%] (4)
ΔH 0m FG mode to model the airflow field at the air outlet. We also conducted
experimental air temperature and airflow rate measurements to corre­
where, ΔHm is the experimental heat of fusion and ΔHcc is the experi­ late those with CFD simulation results. The CFD and experimental air
mental cold crystallization obtained by the DSC scans. We analyzed the temperature data are shown in Fig. 5. We measured 7 points at 5–40 mm

6
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Table 2 no air heating was applied. We analyzed the air’s mass flow rate ob­
Air mass flow rate data obtained from experiments and CFD simulations. tained by experimental and CFD analysis. The results are given in
Air pressure Experimental air flow rate ṁcalculated ṁCFD Relative error Table 2. Our findings showed that the relative error between the CFD
[bar] [L/min] [g/s] [g/s] [%] and mass flow rate data determined by experimental ways was fairly
0.39 59 0.91 1.07 14.7 small (<10%). We investigated the pressure loss characteristics between
0.67 82 1.26 1.37 7.9 the CFD and the experimental data. The results showed that the CFD
1 105 1.60 1.73 6.9 mass flow rate regime also correlates with the experimental data and
1.35 128 1.96 2.11 7.2 expresses fair agreement (Table 2 and Fig. S11). In conclusion, CFD
1.68 150 2.27 2.47 8.1
2.02 172 2.58 2.84 9.2
simulations can be used to estimate airflow characteristics with
2.38 196 2.92 3.24 9.8 reasonable accuracy.

3.1.2. Results of the application-oriented CFD model

distance from the tool tip with the thermocouple. We also monitored the We investigated the effect of air pressure (1, 1.5 and 2 bars) and the
apparatus temperature while the hot pressurized air (250 ◦ C) was cone opening angle (35◦ and 60◦ was chosen) on the air temperature
flowing. In addition, the apparatus temperature was continuously (Figs. 6 and 7) and the air velocity field. The results implied that the air
monitored via the printer interface (with an additional K-type thermo­ temperature drops faster near the apparatus when the air pressure in­
couple attached to the apparatus) and with a thermal camera (Fig. S10). creases; hence, the temperature gradient close to the die affects the final
The results implied that the difference between the CFD and experi­ fiber properties (e.g., fiber-to-fiber contacts, fiber mat thermal and me­
mental data was reasonable; hence, we concluded that CFD results could chanical properties, fiber crystallization and attenuation rate) dramati­
be used to model the airflow field and related process parameters. cally. For example, when the air cools down fast, it provides a quicker
We also measured the airflow rate (in L/min) with a flow meter when

Fig. 6. Variation of the air temperature upon exiting the die for the air knife with half-cone opening angles of 35◦ (blue-colored dashed lines) and 60◦ (green-colored
continuous lines) at various pressure levels.

Fig. 7. Counter plots of air temperature profiles for the air knife with half-cone opening angles of 60◦ and 35◦ at the air pressure of 2 bar.

7
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 8. Counter plots of air velocity profiles for the air knife with half-cone opening angles of 60◦ and 35◦ at the air pressure of 2 bar.

Fig. 9. Variation of the air velocity upon exiting the die for the air knife with half-cone opening angles of 35◦ (blue-colored dashed lines) and 60◦ (green-colored
continuous lines) at various pressure levels.

solidification of the fibers, giving poor fiber bonding. However, air ve­ the 35◦ air knife. This is related to the air velocity, which is slower in the
locity also plays a vital role in the fiber structure development via the case of the 35◦ half-cone opening angle. The 60◦ air knife exhibited
drawing it provides [24]. lower air temperatures and higher air velocities than the 35◦ one for
The results showed that the air cools down slower after exiting from each pressure level. In this regard, we observed a region where the air

Fig. 10. Variation of fiber diameter versus air pressure and nozzle diameter; (a) fibers produced with air knife module with cone opening angle of 60◦ and (b) 35◦ .

8
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 11. SEM images of fibers. Effect of nozzle diameters of (a) 0.5 mm, (b) 0.4 mm, (c) 0.2 mm at 1.5 bar air pressure, and effect of air pressure (a nozzle diameter of
0.2 mm) of (d) 1 bar, (e) 1.5 bar, (f) 2 bar on fiber mat morphology produced.

jets converge for higher pressures, as shown in Figs. 6–9. This phe­ respect to the nozzle diameter and the air pressure, while the fiber
nomenon results in air velocity and temperature fluctuation near this diameter distribution histograms are given in Figs. S14 and S15.
location. When the air velocity reached over 450 m/s for high inlet We found that a smaller nozzle diameter resulted in thinner fibers
pressures (e.g., at 2 bar), the air jets near the apparatus converged and a with fewer defects, as also shown in Fig. 11(a-c). Decreasing the nozzle
velocity and temperature fluctuation appeared. The CFD simulations diameter from 0.5 mm to 0.2 mm decreased the fiber diameter by up to
implied that air velocity for the 60◦ tool could reach as high as 1.62 2.6-fold and 2.25-fold for the 60◦ and the 35◦ air knives, respectively.
Mach (~600 m/s), while the 35◦ tool had a max of 1.41 Mach (~550 m/ This is because the same drag force provided by the pressurized hot air
s) when applied air pressure was 2 bar (Figs. 8 and 9, Fig. S12). jet was applied to a higher polymer mass. Besides, higher shear rates are
This fluctuation is associated with high turbulence kinetic energy favorable for attenuating the polymeric jet into fine fibers. The smaller
[48,56]. Increasing the inlet air pressure resulted in high turbulence capillary diameter generates higher shear rates on the melt. In the case
intensity, so the air velocity and temperature fluctuated over the dis­ of thermoplastic polymers, high shear rates are desirable for the polymer
tance (Fig. S13). For the 60◦ angle, we observed two maxima (i.e., melt to be drawn into fine fibers because these are typically shear-
negative and positive) in the region where the air jets converged for the thinning materials. The fiber attenuation rate was higher in the 60◦
air velocity field (Figs. 8 and 9). This might increase the fiber attenua­ air knife case than in the 35◦ . This is because the air velocity, so the
tion rate. On the other hand, the results implied that the 60◦ air knife drawing force, was higher in the 60◦ air knife case (see Fig. 9).
possesses faster air cooling due to its high-velocity airflow. Results showed that a smaller nozzle diameter and high air pressure
are favorable for generating continuous, evenly distributed, defect-free,
thin fibers with the novel apparatus (Fig. 11(d-f)). Higher air velocity
3.2. Fiber mat morphology translates to larger attenuation rates. The air exerts a higher forwarding
drag force and stress that draws the polymeric jet into a fine fiber. The
We generated fine fibers with the apparatus in the FG mode and average fiber diameter produced with the 60◦ air knife was 7–12%
investigated the influence of nozzle diameter and air pressure on the thinner than that produced by the 35◦ one. We found that increasing air
fiber morphology. Fig. 10 shows the variation of the fiber diameter with

Fig. 12. Tensile properties of the fiber mats produced by cone opening angles of 60◦ and 35◦ (a) specific strength versus applied air pressure and (b) stain at σmax
versus applied air pressure.

9
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 13. Fiber mat’s DSC thermograms; (a) 1st heating and (b) cooling cycles.

findings showed that the fibers’ glass transition temperature and crys­
Table 3
tallinity are 4 ◦ C and 6% points higher, respectively than those of the
DSC 1st heating data of the fiber mats and filament.
filament. These improvements are attributed to the shear-induced
α Applied Glass Cold Crystalline Degree of crystallization and increased molecular orientation along the thread
air transition crystallization melting crystallinity,
line [25,57,58]. That is an advantage in keeping the structural integrity
pressure temperature, enthalpy, enthaply χ
Tg ΔHcc1 + ΔHcc2 ΔHm of the fibers when the 3D printer adds the next layer of dispensed molten
[ ]
◦
[bar] [◦ C] [J/g] [J/g] [%] material during the process.
60 2 58.0 25.5 58.1 33.7
Results showed that the fibers generated by the 60◦ air knife showed
1.5 58.8 28.7 58.7 31.0 a higher degree of crystallinity for the same air pressure applied than the
1 58.4 29.7 56.6 27.7 35◦ one, which is related to the higher attenuation rates (i.e., air ve­
35 2 58.3 26.9 55.8 29.8 locity). Findings implied that the fibers generated at high air pressure
1.5 58.6 31.9 60.2 29.1
also showed a higher degree of crystallinity. In general, a higher degree
1 58.3 28.8 59.0 31.1
filament 54.6 32.2 59.3 28.0 of crystallinity translates to better mechanical properties. This is not the
(reference) case in the case of fiber mats produced via the novel method. The DSC
data represents microstructural development (e.g., molecular orienta­
tion, crystallinity) in fiber making. On the other hand, the conducted
pressure from 1 to 2 bar decreased the average fiber diameter by tensile tests represent the macroscale properties of the fiber mat. DSC
50–110% for the 60◦ air knife. On the other hand, the average fiber cooling results revealed that the PLLA fibers showed different crystal­
diameter decreased by 51–75%, increasing air pressure from 1 to 2 bar lization kinetics due to the high shearing applied. This is attributed to
for the 35◦ air knife. shear-induced nuclei formed during the FG mode. A similar finding has
These results are in-line with the CFD simulation findings. The air also been reported by Du et al. [57] in which they observed improved
velocity and the temperature fields for the different cone opening angles crystallization peak temperature for high-shearing applied PLLA in the
and inlet air pressures resulted in different fiber and fiber mat proper­ DSC cooling curve. They concluded that a shish-like oriented structure
ties. These findings prove that the novel apparatus provides straight­ due to high-shearing melt allows crystal growth at higher temperatures.
forward manipulation of the fiber characteristics in a wide range. These findings suggest that the PLLA fibers generated at different air
pressures and half-cone opening angles had different nucleation kinetics
subjected to different shear and melting conditions, resulting in different
3.3. Evaluating fiber mat’s thermal and mechanical properties
crystallization peak temperatures during DSC cooling cycles.
Overall, the DSC findings implied that the fibers had a different
The fiber mats produced via the novel apparatus with a nozzle
microstructure development compared to the filament. This also in­
diameter of 0.2 mm showed thin fibers with fine morphology. Therefore,
dicates that the FG mode is suitable for making fine polymeric with
we chose the nozzle diameter of 0.2 mm to investigate the fiber mat’s
enhanced thermal stability.
thermal and mechanical properties. Results showed that increasing
The simulation study showed that the 60◦ cone opening angle pos­
pressure increased the fiber mat’s specific strength. This is because of
sesses a higher air velocity (i.e., high attenuation rate) than the 35◦ one
increasing fiber packing density and thinner fibers generated with
and cools down the air faster. A high attenuation rate improves the
increasing air pressure, as shown in Fig. 12 (Fig. S16 and Table S1). A
crystallinity, but high air velocity causes a faster cooling of both the air
mechanically consolidated fiber mat structure accompanies the
and the forming fiber. This, in turn, gives poor fiber-to-fiber contacts.
improved specific strength. On the other hand, increasing air pressure
The newly formed fiber’s ability to penetrate those fibers collected at the
resulted in more fiber entanglements and lowered the fiber mat’s strain
printing bed affects the fiber mat’s load-bearing capability. Higher
at break and maximum stress.
attenuation rates (e.g., higher air pressures) could compensate for the
The specific strength of fiber mats produced with a cone opening
loss in mechanical properties with improved crystallinity and thinner
angle of 60◦ was lower than that of produced with the 35◦ . This differ­
fibers with an increased number of fiber entanglements.
ence is related to the cooling of the fiber. Fine fibers’ microstructure
development and crystallinity are mainly related to jet cooling and
attenuation rate. Therefore, we conducted DSC tests for fiber mats and 3.4. Analyzing the novel apparatus FFF printing ability: Tests on the
filaments to analyze their thermal properties. DSC thermograms are generic 3D samples
shown in Fig. 13. The 1st heating data obtained are given in Table 3,
while the cooling and 2nd heating data are given in Table S2. DSC The developed apparatus can be used to generate continuous fibers

10
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Fig. 14. Generic objects printed via novel printhead.

Fig. 15. Testing FF and FG modes; (a) optical image of the samples and used tools, (b) SEM images of FFF printed layer and fiber mat, (c) microstructure of a FFF
printed hierarchical structure (red rectangles show FFF printed struts) and (d) cross-sections of the hierarchical microstructure consist of fibers and 3D printed struts.

11
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

Table 4 curation, Writing-Original draft preparation, Writing-Review & Editing,

Comparison of the FFF 3D printed sample geometry with the CAD model. Visualization, Investigation, Formal analysis. Norbert Krisztián
length relative thickness relative width relative Kovács: Resources, Funding acquisition, Writing-Review & Editing.
error error error Péter Nagy György: Software, Writing-Reviewing and Editing, Vali­
[mm] [%] [mm] [%] [mm] [%] dation. Boros Robert: Validation, Writing-Review & Editing. Kolos
1 79.73 0.34 1.49 0.67 9.96 0.40 Molnár: Validation, Writing-Reviewing and Editing.
2 79.37 0.79 1.52 1.33 9.97 0.30
3 79.42 0.73 1.55 3.33 10.06 0.60
4 79.35 0.82 1.54 2.67 10.01 0.10
Declaration of Competing Interest
5 79.57 0.54 1.54 2.67 10.09 0.90
CAD 80.00 1.50 10.00
model 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.
and 3D print struts. The method makes it possible to continuously pro­
duce hierarchical FFF printed structures comprising nano-/submicron Data availability
fibers and extruded filament using a single apparatus, as shown in
Fig. 14 and Fig. 15 (Fig. S6). In this regard, we analyzed the geometry of Data will be made available on request.
the 3D printed sample and the CAD model (Fig. S9). Findings showed
that the relative error between the 3D model and 3D printed samples is
Acknowledgments
fairly small, as shown in Table 4. Findings implied that the developed
apparatus could successfully generate objects based on a geometrical
The authors thank Total Corbion PLA bv and Dr. G. Gobius du Sart,
representation given by the CAD model.
for supplying the PLA resins used in this work. The research reported in
Fig. 15(b-d) shows that the apparatus can successfully generate hi­
this paper is part of project no. BME-NVA-02, implemented with the
erarchical structures consisting of fibers and 3D printed struts. In Fig. 15
support provided by the Ministry of Innovation and Technology of
(b), an SEM image shows fibers laid onto a 3D printed layer, while an
Hungary from the National Research, Development and Innovation
SEM image of the 3D printed struts onto fiber mats is shown in Fig. 15
Fund, financed under the TKP2021 funding scheme. The research re­
(c). The fibers and struts made of the same polymer (e.g., PLLA filament)
ported in this paper was supported by the National Research, Develop­
had different melting characteristics observed in DSC tests. These find­
ment and Innovation Office (FK 138501, FK 134336 and K 138472). This
ings prove that generated fibers and 3D printed struts had different
paper was also supported by the János Bolyai Research Scholarship of
microstructures, allowing them to adhere perfectly but not deteriorate
the Hungarian Academy of Sciences (MTA).
during the FFF printing. Furthermore, fibers generated at FG mode stay
intact with the 3D printed struts, and fibers fill the gap between adjacent
struts, as shown in Fig. 15(d). This property would be an efficient so­ Appendix A. Supporting information
lution for enhancing mechanical performance in the building direction
(Fig. S6). Supplementary data associated with this article can be found in the
online version at doi:10.1016/j.addma.2022.103315.
4. Conclusion
References
We introduced a novel 3D printing method that is capable of
[1] H. He, K. Molnár, Fabrication of 3D printed nanocomposites with electrospun
generating hierarchical structures of ultrafine fibers and solid layers of nanofiber interleaves, Addit. Manuf. 46 (2021), 102030, https://doi.org/10.1016/
deposited material with a single printhead. It was shown that the j.addma.2021.102030.
printhead can be easily integrated into a commercial FFF printer without [2] T.W. Terry, C. Ian, D. Olaf, H. Ray, K. Joseph, M. Noah, Wohlers Report 2021: 3D
Printing and Additive Manufacturing Global State of the Industry, (2021).
using expensive hardware and software. The device operates in two [3] T.W. Terry, C. Ian, D. Olaf, H. Ray, K. Joseph, M. Noah, Wohlers report 2022: 3D
modes: filament fusion (FF) and fiber generation (FG) mode with simple printing and additive manufacturing global state of the industry, (2022).
switching between these. In the latter case, hot pressurized air blows and [4] V. Mohanavel, K.S. Ashraff Ali, K. Ranganathan, J. Allen Jeffrey, M.M. Ravikumar,
S. Rajkumar, The roles and applications of additive manufacturing in the aerospace
stretches the polymer melt into fine fibers. We proved that we could
and automobile sector, Mater. Today.: Proc. 47 (2021) 405–409, https://doi.org/
create hierarchical structures composed of solid (or even infill) parts and 10.1016/j.matpr.2021.04.596.
porous nano-/microfiber layers within a single 3D-printed object. [5] A.N. Solodov, J. Shayimova, D. Balkaev, A.S. Nizamutdinov, K. Zimin, A.
We can control the fiber mat structure in several ways. We simulated G. Kiiamov, R.R. Amirov, A.M. Dimiev, High-throughput, low-cost and “green”
production method for highly stable polypropylene/perovskite composites,
the airflow field by CFD and we proved that the air knife angle (35◦ vs. applicable in 3D printing, Addit. Manuf. 59 (2022), 103094, https://doi.org/
60◦ ) has a significant influence on the fiber strength. Decreasing the 10.1016/j.addma.2022.103094.
nozzle size and increasing the air pressure or air velocity decreases the [6] M. Javaid, A. Haleem, R.P. Singh, R. Suman, S. Rab, Role of additive
manufacturing applications towards environmental sustainability, Adv. Ind. Eng.
fiber diameter considerably. Increasing hot air speed decreased the fiber Polym. Res. 4 (4) (2021) 312–322, https://doi.org/10.1016/j.aiepr.2021.07.005.
diameter by up to 2-fold, while the fiber mat’s specific strength was [7] T.Q. Tran, F.L. Ng, J.T.Y. Kai, S. Feih, M.L.S. Nai, Tensile strength enhancement of
enhanced by up to 133%. With the aid of CFD simulations, we demon­ fused filament fabrication printed parts: a review of process improvement
approaches and respective impact, Addit. Manuf. 54 (2022), 102724, https://doi.
strated that the air reaches supersonic velocities during the continuous org/10.1016/j.addma.2022.102724.
FG mode. We showed that the fibers’ glass transition temperature and [8] ISO/ASTM 52900:2015 Additive manufacturing — General principles —
crystallinity are 4 ◦ C and 6% points higher than the feedstock filament. Terminology.
[9] S.S. Crump, Apparatus and method for creating three-dimensional objects, Patent,
The ability to easily generate and combine porous ultrafine fibers and US5121329A, USA, (1989).
solid parts within a single 3D printed component makes it possible to fit [10] S. Yuan, S. Li, J. Zhu, Y. Tang, Additive manufacturing of polymeric composites
in numerous applications, including but not limited to automotive, from material processing to structural design, Compos. Part B: Eng. 219 (2021),
108903, https://doi.org/10.1016/j.compositesb.2021.108903.
aerospace, defense, medical, pharmaceutical, textile and many more.
[11] A. Cano-Vicent, M.M. Tambuwala, S.S. Hassan, D. Barh, A.A.A. Aljabali, M. Birkett,
A. Arjunan, Á. Serrano-Aroca, Fused deposition modelling: Current status,
CRediT authorship contribution statement methodology, applications and future prospects, Addit. Manuf. 47 (2021), 102378,
https://doi.org/10.1016/j.addma.2021.102378.
[12] W. Johnston, B. Sharma, Additive manufacturing of fibrous sound absorbers, Addit.
Yahya Kara: Conceptualization, Methodology, Software, Data Manuf. 41 (2021), 101984, https://doi.org/10.1016/j.addma.2021.101984.

12
Y. Kara et al. Additive Manufacturing 61 (2023) 103315

[13] Kamyab, G. Ghasemi, Fereidoon Khonakdar, Investigation into the shape memory [35] K.F. Eichholz, I. Gonçalves, X. Barceló, A.S. Federici, D.A. Hoey, D.J. Kelly, How to
behavior of peanut-pattern auxetic structures, Express Polym. Lett. 16 (2022) design, develop and build a fully-integrated melt electrowriting 3D printer, Addit.
679–693, https://doi.org/10.3144/expresspolymlett.2022.50. Manuf. 58 (2022), 102998, https://doi.org/10.1016/j.addma.2022.102998.
[14] T. Spahiu, E. Canaj, E. Shehi, 3D printing for clothing production, [36] D.B. Sitotaw, D. Ahrendt, Y. Kyosev, A.K. Kabish, Additive manufacturing and
1558925020948216, J. Eng. Fibers Fabr. 15 (2020), https://doi.org/10.1177/ textiles—state-of-the-art, Appl. Sci. 10 (15) (2020) 5033, https://doi.org/10.3390/
1558925020948216. app10155033.
[15] K. Chatterjee, T.K. Ghosh, 3D printing of textiles: potential roadmap to printing [37] H.C. Koch, D. Schmelzeisen, T. Gries, 4D textiles made by additive manufacturing
with fibers, Adv. Mater. 32 (4) (2020) 1902086, https://doi.org/10.1002/ on pre-stressed textiles—An overview, Actuators 10 (2) (2021) 31, https://doi.org/
adma.201902086. 10.3390/act10020031.
[16] Z. Zhang, H. He, W. Fu, D. Ji, S. Ramakrishna, Electro-hydrodynamic direct-writing [38] Y. Kara, N. Krisztián Kovács, K. Molnár, (In Hungarian) An apparatus for a fiber-
technology toward patterned ultra-thin fibers: advances, materials and forming 3D printer, a 3D printer incorporating such an apparatus, and a method of
applications, Nano Today 35 (2020), 100942, https://doi.org/10.1016/j. making a 3D printed polymer composite Patent, P2200146, Hungary, (2022).
nantod.2020.100942. [39] H.M. Krutka, R.L. Shambaugh, D.V. Papavassiliou, Analysis of a melt-blowing die:
[17] F. Zhang, K. Cao, A. Zaeri, R. Zgeib, R.C. Chang, Design, fabrication, and analysis of comparison of CFD and experiments, Ind. Eng. Chem. Res. 41 (20) (2002)
spatially heterogeneous scaffold by melt electrospinning writing of poly 5125–5138, https://doi.org/10.1021/ie020366f.
(ε-Caprolactone), J. Appl. Polym. Sci. 139 (22) (2022) 52235, https://doi.org/ [40] X. Hao, Y. Zeng, A review on the studies of air flow field and fiber formation
10.1002/app.52235. process during melt blowing, Ind. Eng. Chem. Res. 58 (27) (2019) 11624–11637,
[18] Y. Kara, H. He, K. Molnár, Shear-aided high-throughput electrospinning: A https://doi.org/10.1021/acs.iecr.9b01694.
needleless method with enhanced jet formation, J. Appl. Polym. Sci. 137 (37) [41] S.V. Murphy, A. Atala, 3D bioprinting of tissues and organs, Nat. Biotechnol. 32 (8)
(2020) 49104, https://doi.org/10.1002/app.49104. (2014) 773–785, https://doi.org/10.1038/nbt.2958.
[19] Botta Lopresti, C. La, Attinasi Settanni, Garofalo Gaglio, Physical and antibacterial [42] H.-W. Kang, S.J. Lee, I.K. Ko, C. Kengla, J.J. Yoo, A. Atala, A 3D bioprinting system
properties of PLA electrospun mats loaded with carvacrol and nisin, Express Polym. to produce human-scale tissue constructs with structural integrity, Nat. Biotechnol.
Lett. 16 (2022) 1083–1098, https://doi.org/10.3144/expresspolymlett.2022.79. 34 (3) (2016) 312–319, https://doi.org/10.1038/nbt.3413.
[20] F. Zhang, K. Cao, A. Zaeri, R. Zgeib, R.C. Chang, Effects of scaffold design [43] A. Szabó, P.T. Nagy, G. Paál, Numerical simulation of an acoustically excited plane
parameters on the printing accuracy for melt electrowriting, J. Manuf. Process. 81 jet in an incompressible framework and comparison with experimental data,
(2022) 177–190, https://doi.org/10.1016/j.jmapro.2022.06.070. J. Acoust. Soc. Am. 147 (5) (2020) 3429–3443, https://doi.org/10.1121/
[21] R. Farer, A.M. Seyam, T.K. Ghosh, S.K. Batra, E. Grant, G. Lee, Forming shaped/ 10.0001256.
molded structures by integrating meltblowing and robotic technologies, Text. Res. [44] P.T. Nagy, A. Szabó, G. Paál, A feedback model of the edge tone, using the adjoint
J. 73 (1) (2003) 15–21, https://doi.org/10.1177/004051750307300103. Orr–Sommerfeld equation, J. Fluid Mech. 915 (2021) A13, https://doi.org/
[22] R. Farer, A.M. Seyam, T.K. Ghosh, E. Grant, S.K. Batra, Meltblown Structures 10.1017/jfm.2021.47.
Formed by a Robotic and Meltblowing Integrated System: Impact of Process [45] P. Tamás Nagy, G. Paál, On the sensitivity of planar jets, Int. J. Heat. Fluid Flow. 62
Parameters on Fiber Orientation and Diameter Distribution, Text. Res. J. 72 (12) (2016) 114–123, https://doi.org/10.1016/j.ijheatfluidflow.2016.09.017.
(2002) 1033–1040, https://doi.org/10.1177/004051750207201201. [46] K.G. Schuchard, A. Pawar, B. Anderson, B. Pourdeyhimi, R.A. Shirwaiker,
[23] Y. Kara, K. Molnár, A review of processing strategies to generate melt-blown nano/ Multiphase CFD modeling and experimental validation of polymer and attenuating
microfiber mats for high-efficiency filtration applications, J. Ind. Text. 51 (1_suppl) air jet interactions in nonwoven annular melt blowing, Ind. Eng. Chem. Res.
(2022) 137S–180S, https://doi.org/10.1177/15280837211019488. (2022), https://doi.org/10.1021/acs.iecr.2c01710.
[24] Y. Kara, K. Molnár, Revealing of process–structure–property relationships of fine [47] Y. Yang, Y. Zeng, Measurement and comparison of melt-blowing airflow fields:
polypropylene fiber mats generated via melt blowing, Polym. Adv. Technol. 32 (6) nozzle modifications to reduce turbulence and fibre whipping, Polymers 13 (5)
(2021) 2416–2432, https://doi.org/10.1002/pat.5270. (2021) 719, https://doi.org/10.3390/polym13050719.
[25] Y. Kara, K. Molnár, Development of single-polypropylene composites interleaved [48] S. Xie, W. Han, G. Jiang, C. Chen, Turbulent air flow field in slot-die melt blowing
with MWCNT-doped melt-blown fine fiber mats, Polym. Compos. 43 (8) (2022) for manufacturing microfibrous nonwoven materials, J. Mater. Sci. 53 (9) (2018)
5208–5221, https://doi.org/10.1002/pc.26812. 6991–7003, https://doi.org/10.1007/s10853-018-2008-y.
[26] TamiCare, Speedy Additive Manufacturing of Fabrics, 2021. 〈https://www.tamicar [49] J. Schmidt, S. Shenvi Usgaonkar, S. Kumar, K. Lozano, C.J. Ellison, Advances in
e.com/manufacture〉. (Accessed July 2022). melt blowing process simulations, Ind. Eng. Chem. Res. 61 (1) (2021) 65–85.
[27] Electroloom, The World’s First 3D Fabric Printer, 2016. 〈https://www.kickstarter. [50] D. Guo, H. Fan, Z. Zhu, K. Xiao, Numerical investigation into the effect of
com/projects/electroloom/electroloom-the-worlds-first-3d-fabric-printer〉. geometric shape of slot on air flow field in melt-blown for polymer fiber, J. Text.
(Accessed July 2022). Inst. 113 (6) (2022) 1133–1141, https://doi.org/10.1080/
[28] P. Romero-Araya, V. Pino, A. Nenen, V. Cárdenas, F. Pavicic, P. Ehrenfeld, 00405000.2021.1916222.
G. Serandour, J.G. Lisoni, I. Moreno-Villoslada, M.E. Flores, Combining materials [51] T. Tábi, S. Hajba, Cross effect of natural rubber and annealing on the properties of
obtained by 3d-printing and electrospinning from commercial polylactide filament poly (Lactic Acid), Period. Polytech. -Mech. Eng. 64 (4) (2019) 270–277, https://
to produce biocompatible composites, Polymers 13 (21) (2021) 3806, https://doi. doi.org/10.3311/PPme.12825.
org/10.3390/polym13213806. [52] ASTM E1356–08(2014):Standard Test Method for Assignment of the Glass
[29] H. He, M. Gao, B. Illés, K. Molnar, 3D printed and electrospun, transparent, Transition Temperatures by Differential Scanning Calorimetry.
hierarchical polylactic acid mask nanoporous filter, Int. J. Bioprinting 6 (4) (2020), [53] T. Kim, S. Lee, G.B. Kim, D. Hong, J. Kwon, J.-W. Park, N. Kim, Accuracy of a
https://doi.org/10.18063/ijb.v6i4.278. simplified 3D-printed implant surgical guide, e2. J. Prosthet. Dent. 124 (2) (2020)
[30] T. Kozior, A. Mamun, M. Trabelsi, L. Sabantina, Comparative analysis of polymer 195–201, https://doi.org/10.1016/j.prosdent.2019.06.006.
composites produced by FFF and PJM 3D printing and electrospinning technologies [54] M.M. Hanon, L. Zsidai, Q. Ma, Accuracy investigation of 3D printed PLA with
for possible filter applications, Coatings 12 (1) (2022) 48. various process parameters and different colors, Mater. Today.: Proc. 42 (2021)
[31] Y. Yoon, C.H. Kim, J.E. Lee, J. Yoon, N.K. Lee, T.H. Kim, S.-H. Park, 3D bioprinted 3089–3096, https://doi.org/10.1016/j.matpr.2020.12.1246.
complex constructs reinforced by hybrid multilayers of electrospun nanofiber [55] Y. Yang, H. Huang, Y. Zeng, Turbulence of melt-blowing airflow field: comparison
sheets, Biofabrication 11 (2) (2019), 025015, https://doi.org/10.1088/1758- of a convergent jet and a typical free jet, Phys. Fluids 33 (7) (2021), 075107,
5090/ab08c2. https://doi.org/10.1063/5.0055571.
[32] Y. Yu, S. Hua, M. Yang, Z. Fu, S. Teng, K. Niu, Q. Zhao, C. Yi, Fabrication and [56] H.M. Krutka, R.L. Shambaugh, D.V. Papavassiliou, Effects of die geometry on the
characterization of electrospinning/3D printing bone tissue engineering scaffold, flow field of the melt-blowing process, Ind. Eng. Chem. Res. 42 (22) (2003)
RSC Adv. 6 (112) (2016) 110557–110565, https://doi.org/10.1039/C6RA17718B. 5541–5553, https://doi.org/10.1021/ie030457s.
[33] T. Xu, K.W. Binder, M.Z. Albanna, D. Dice, W. Zhao, J.J. Yoo, A. Atala, Hybrid [57] M. Du, K. Jariyavidyanont, R. Zhang, A.M. Rhoades, C. Schick, R. Androsch,
printing of mechanically and biologically improved constructs for cartilage tissue Thermal stability of shear-induced crystal nuclei of Poly(l-lactic acid, ACS Appl.
engineering applications, Biofabrication 5 (1) (2012), 015001. Polym. Mater. 4 (10) (2022) 7823–7833, https://doi.org/10.1021/
[34] J.A. Smith, E. Mele, Electrospinning and additive manufacturing: adding three- acsapm.2c01365.
dimensionality to electrospun scaffolds for tissue engineering, 674738-674738, [58] A. Jalali, M.A. Huneault, M. Nofar, P.C. Lee, C.B. Park, Effect of branching on flow-
Front Bioeng. Biotechnol. 9 (2021), https://doi.org/10.3389/fbioe.2021.674738. induced crystallization of poly (lactic acid, Eur. Polym. J. 119 (2019) 410–420,
https://doi.org/10.1016/j.eurpolymj.2019.07.045.

13