Engineering Manufacturing Letters, 2:1 (2024) 45-52 Received: 10 May, 2024

ISSN 2795-5168 Accepted: 14 July, 2024

DOI: 10.24840/2795-5168_002-001_2675 Published: 11 October, 2024

Challenges on extrusion-based additive manufacturing of

thermoplastic polyurethane
Manuel Sardinha
IDMEC, Instituto Superior Técnico, Universidade de Lisboa, Portugal
([email protected]) ORCID 0000-0003-2124-8569
Luís Ferreira
IST, Instituto Superior Técnico, Universidade de Lisboa, Portugal
([email protected]) ORCID 0009-0004-1157-1623
Tânia Ramos
CEGIST, Instituto Superior Técnico, Universidade de Lisboa, Portugal
([email protected]) ORCID 0000-0002-2321-2431
Luís Reis
IDMEC, Instituto Superior Técnico, Universidade de Lisboa, Portugal
([email protected]) ORCID 0000-0001-9848-9569
M. Fátima Vaz
IDMEC, Instituto Superior Técnico, Universidade de Lisboa, Portugal
([email protected]) ORCID 0000-0003-1629-523X

Author Keywords Abstract

Thermoplastic elastomer, Fused filament fabrication (FFF) offers rapid production

capabilities with user-friendly operation and cost-
fused filament fabrication, defects.
effectiveness, yet challenges in printing with thermoplastic
polyurethane (TPU) persist. This study identifies causes of
Type: Rapid communication print defects in FFF-TPUs and their relation to process
parameters, by visually analysing samples produced with
Open Access various TPU materials. Extrusion temperature, printing
speeds, filament cooling, overlap, and extrusion flow are
Peer Reviewed
subject to evaluation. Common issues include filament
CC BY entanglements, under-extrusion, stringing, or geometric
deviations. Among other approaches, results highlight the
importance of slow, constant print speeds and careful
optimization of extrusion temperature.

1. Introduction
Fused filament fabrication (FFF) has become one of the most widely used additive
manufacturing techniques, due to its ability to rapidly produce complex shapes, user-friendly
operation, cost-effectiveness, and potential environmentally friendly features (Awasthi and
Banerjee 2021). FFF is a layer-based process in which material is extruded through a nozzle in
a semi-molten state. To convert a 3D CAD model into machine code, a slicer software is used,
where various process parameters can be tuned to each material’s specific characteristics.
Besides the nozzle, an extrusion system is typically composed of a feeder, a hot end, and
cooling fans (Figure 1(a)). The feeder mechanism, which is responsible for directing the
filament into the hot end, is either mounted on the extrusion head (direct drive extrusion) or
connected to the hot end by a flexible tube that guides the filament (Bowden extrusion), as
shown in Figure 1 (b-c), respectively.

45
Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

In FFF, the mechanical properties of parts are significantly influenced by parameters such as
extrusion temperature, printing speed, infill type, layer thickness, and build orientation (Xu et
al. 2020). Recently, the process has found various applications across industries such as
architecture, aerospace, and the medical sector, but certain drawbacks still need to be
addressed (Altıparmak et al. 2022). Namely, components can exhibit weak interlayer strength,
low dimensional accuracy, size limitations, and poor surface quality (Wickramasinghe et al.
2020).
Classified as a thermoplastic elastomer (TPE), thermoplastic polyurethane (TPU) is a
multiphase block copolymer formed by rigid (diisocyanate) and flexible (polyol) domains,
which allow for rubber-like elasticity and the processability of thermoplastics (Hohimer et al.
2017).

Figure 1: Fused filament fabrication process: (a) schematic illustration of

relevant process parameters; (b) a direct drive extrusion machine, Prusa
Mk4; (c) a Bowden-based extrusion machine, Ultimaker S5.

TPU is known for its flexibility, low elastic modulus, and the ability to undergo substantial
elongation before fracture (Vidakis et al. 2021). Fused filament fabricated thermoplastic
polyurethanes (FFF-TPUs) have been proposed for applications such as cushioning, wearables,
footwear, and medical devices (Geng et al. 2023; Awasthi and Banerjee 2021). Moreover,
current developments propose to expand their utility into promising areas such as flexible
electronics (Papazetis and Vosniakos 2019), pneumatic actuators (Salem et al. 2018), and non-
pneumatic tires (Wang et al. 2020; Sardinha et al. 2022). Nevertheless, the inherent softness
of these filaments poses significant challenges in controlling their extrusion flow, and can
cause unpredictable viscoelastic deformations, which then compromise the manufacturability
of geometrically intricate structures (Mian et al. 2024; Perry et al. 2022). Commonly reported
issues of FFF-TPUs are related to material oozing and stringing effects, and solutions proposed
are related to the fine-tuning of extrusion temperature and print speed or limiting reverse
movements of the feeder motor that pull the filament away from the outlet nozzle (retraction)
(León-Calero et al. 2021; Awasthi and Banerjee 2021). Even so, parameters that promote
manufacturability often conflict with those needed for adequate mechanical integrity, and
trade-offs are often required. Notably, TPU filament materials can vary widely in composition,
making it difficult to establish consistent processing parameters for achieving reliable
mechanical properties. Moreover, available information often lacks methodology support and
there is still a lack of consensus among sources regarding some issues and potential solutions.
For instance, various research suggests extrusion temperatures for FFF-TPUs ranging from
190oC to 250oC (Arifvianto et al. 2021; Kasmi et al. 2022). This variability, coupled with the fact
that different polymer chain compositions have distinct thermal activation and dissociation
requirements, and that thermal degradation can occur between 110°C and 270°C (Zia et al.

Engineering Manufacturing Letters 2:1 (2024) 45-52 46

Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

2007), highlights the inconsistencies and emphasizes the need for further studies on the
subject.
In this study, the authors investigate how FFF process variables affect the manufacturability
of TPU parts. The experiments involve testing various combinations of machines and
materials, followed by a qualitative visual analysis to identify common issues and types of
macroscopic defects observed in FFF-TPUs. Particular attention is given to extrusion
temperature, printing speed, cooling fan speed, deposition overlap, retraction, and extrusion
flow amount. During the analysis, the authors explore parameter adjustments to optimize TPU
part production conditions, aiming at facilitating parameter pre-selection for future
mechanical characterization studies of these materials.
2. Materials and Methods
For the experiments performed in this work, two different FFF machines were used, a
Bowden-based Ultimaker S5 that uses 2.85 mm filaments, and a Prusa MK4 which is equipped
with a direct drive extrusion and uses 1.75 mm filaments. Concerning tested materials, TPU
95A from Ultimaker®, and Filaflex 95A were printed in the Ultimaker S5, and Flexfill TPU 92A
and 98A, as well as Smartfil TPU 93A were printed in the Prusa MK4. According to the 3D
printer, the slicer software used to prepare the machine files was either the Ultimaker Cura
4.11.0 or the PrusaSlicer 2.7.1. Among the tested parameters, extrusion temperature was
varied according to each material, but always between 200oC and 240oC. Printing speeds of
up to 100 mm/s were attempted, but when evaluating the effect of other process parameters,
a constant speed of 25 mm/s was used. The filament cooling was also assessed, varying the
cooling fan speed on three different levels (0%; 20% and 100%). Besides these, deposition
overlap up to 100%, and extrusion flows up to 190% were also subject to evaluation.
3. Results and Discussion
In FFF, the filament inside the heating unit of the extrusion head should be viscous to a point
that it can be continuously deposited, yet sufficiently cooled above this region so that it can
be pushed by a driving force. Achieving this delicate balance is particularly challenging when
working with flexible polymers which, in the case of the experiments performed in this work,
often resulted in nozzle clogging issues. In addition to nozzle clogging, filament entanglement-
related issues, like the example shown in Figure 2(a), are also very common. This study
supports common understanding that avoiding retraction, a standard practice for stiffer
materials, can significantly reduce print failures associated with filament entanglement and
nozzle clogging. Even without the use of retraction to withdraw material from the nozzle
outlet, excess material can still drip from it, potentially leading to over-deposition issues.
When this occurs on the outer surface of a part, the phenomenon is commonly referred to as
stringing (Figure 2(b)). The macroscopic defects observed in Figure 2(c), often referred to as
blobs, represent another common occurrence of excess filament on the outer surface of
produced parts. In the example of Figure 2(d), a tensile specimen produced with a concentric
infill pattern accumulates excess material in its narrow section, affecting the study of material
properties. Furthermore, in patter-based designs as the part in Figure 2(e), excess material
can create unintended features and may even result in a distinct pattern, difficult to
distinguish from the structure itself.

Engineering Manufacturing Letters 2:1 (2024) 45-52 47

Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

Figure 2: Examples of most common issues found when 3D printing TPUs:

(a) Filament entanglement inside a direct drive extrusion system of a Prusa
Mk4; (b) Stringing effect using Flexfill TPU 92A; (c) Material blobs in a part
made of Filaflex 95A; (d) Accumulation of excess material in a tensile
specimen produced with a concentric infill pattern; (e) Excess material
deposited with a pattern when producing a cellular structure.

Alongside retraction, printing speed, extrusion temperature, and extrusion flow are also
deeply linked with printing defects. The top view of specimens seen in Figure 3(a-c)
demonstrates that insufficient extrusion temperature can result in thin line widths, reducing
the bonding capacity of the material. At lower temperatures, viscosity may not be adequately
reduced, making it challenging to effectively push the material out of the nozzle. This can lead
to incomplete or uneven material deposition, emphasizing the significant influence of
parameters such as temperature on TPU flow behaviour. Furthermore, despite 200°C being
reported as a suitable working condition for similar materials (Arifvianto et al. 2021; Hohimer
et al. 2017), this temperature consistently resulted in a clogged nozzle.
Figure 3(d-g) show the layers of a specimen and illustrate the effects of slight variations in
filament cooling fan speed (0-20%) and extrusion temperature (235°C to 240°C). Increasing
the filament cooling had a limited impact on promoting a defect-free surface, while reducing
the temperature noticeably decreased interlayer irregularities. This may suggest that, for
Flexfill TPU 92A printed at 240°C, the material may begin to degrade, impacting the bonding
between layers. The top view of samples in Figure 3(h-j) shows the importance of slow printing
speeds when using elastomers, which is aligned with previous literature (Kasmi et al. 2022).
Moreover, this example also shows that, under certain production conditions, excessive
filament cooling might not be beneficial for material bonding.

Engineering Manufacturing Letters 2:1 (2024) 45-52 48

Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

Figure 3: Amplified views of specimens produced in a Prusa Mk4 with: (a)-(c)

Smartfil TPU 93A; (d)-(g) Flexfill TPU 92A; (h)-(j) Flexfill TPU 98A.
Figure 4(a-e) illustrate how adjustments in extrusion flow and deposition overlap can
compensate for problems related to under-extrusion. The top view of specimens in Figure 4(a-
c) reveals examples of under-extrusion issues. Figure 4(b) demonstrates how gaps between
deposited infill and walls can be corrected using an overlap, a commonly available slicer
parameter. Figure 4(c-e) show how the extrusion flow parameter can be adjusted to regulate
filament feeding, offering a valuable option when temperature, speed, and cooling settings
have already been carefully evaluated.

Figure 4: Samples of tensile specimens produced with TPU 95A in the Ultimaker S5:
(a) Representation of lack of material generating gaps between infill and walls; (b)
Amplified view of a specimen produced with overlap between infill and walls; (c)-(e)
Amplified view of equivalent samples produced with variable extrusion flow.

Due to their inherent flexibility, tall and slender structures are particularly challenging to
produce using TPUs. Figure 5(a-b) demonstrate the common occurrence of geometrical
inaccuracies caused by lateral forces imposed on the tip (current layer) of a part by the
extrusion head. In general, successful strategies for addressing the challenges of producing
tall and slender FFF-TPU structures involve increasing the contact area between parts and the
build plate and reducing printing speed and acceleration. Using anchors or additional support
structures that increase the second moment of area of parts can also be effective. Figure 5(c-
d) illustrate how the problem of wobbling encountered in test specimens was resolved by

Engineering Manufacturing Letters 2:1 (2024) 45-52 49

Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

creating a very thin grid connecting multiple specimens. The multiplications of connected
specimens stabilize the structure during fabrication, and the grid can be easily removed after
printing.

Figure 5: Examples of the instability while printing tall and slender

structures: (a) Tower of Smartfil TPU 93A with differences in geometrical
accuracy caused by wobbling; (b) Test specimen of Flexfill TPU 98A; (c) CAD
model of the very thin sacrificial grid connecting specimens; (d) Five
connected specimens of Flexfill TPU 92A.

4. Conclusions
In this study, various causes for printing defects in FFF-TPUs are identified, and their
relationship to process parameters is discussed. Among the identified issues, under- and over-
extrusion are highlighted as the most common. To address this type of issue, the authors
advise using very slow and constant print speeds, along with iterating through small extrusion
temperature intervals, as temperature had a significant influence on sample quality regardless
of the material used. Additionally, considering design for manufacturing aspects such as
optimizing part positioning to minimize retraction or travel movements can significantly
reduce visible defects in FFF-TPUs.
References
Altıparmak, S. C., V. A. Yardley, Z. Shi, and J. Lin. 2022. “Extrusion-based additive
manufacturing technologies: State of the art and future perspectives”. Journal of
Manufacturing Processes 83: 607-636. https://doi.org/10.1016/j.jmapro.2022.09.032.
Arifvianto, B., T. N. Iman, B. T. Prayoga, R. Dharmastiti, U. A. Salim, and M. Mahardika.
2021. “Tensile properties of the FFF-processed thermoplastic polyurethane (TPU)
elastomer”. The International Journal of Advanced Manufacturing Technology 117:
1709-1719. https://doi.org/10.1007/s00170-021-07712-0.
Awasthi, P., and S. S. Banerjee. 2021. “Fused deposition modeling of thermoplastic
elastomeric materials: Challenges and opportunities”. Additive Manufacturing 46:
102177. https://doi.org/10.1016/j.addma.2021.102177.
Geng, T., H. C. Xiao, X. C. Wang, C. T. Liu, L. Wu, Y. G. Guo, B. B. Dong, and L. S. Turng.
2023. “The study on the morphology and compression properties of microcellular
TPU/nanoclay tissue scaffolds for potential tissue engineering applications”. Polymers
15, no. 17: 3647. https://doi.org/10.3390/polym15173647.

Engineering Manufacturing Letters 2:1 (2024) 45-52 50

Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

Hohimer, C., J. Christ, N. Aliheidari, C. Mo, and A. Ameli. 2017. “3D printed thermoplastic
polyurethane with isotropic material properties.” In Behavior and Mechanics of
Multifunctional Materials and Composites 2017, 1016511.
https://doi.org/10.1117/12.2259810.
Kasmi, S., G. Ginoux, E. Labbé, and S. Alix. 2022. “Multi-physics properties of
thermoplastic polyurethane at various fused filament fabrication parameters”. Rapid
Prototyping Journal 28, no. 5: 895-906. https://doi.org/10.1108/RPJ-08-2021-0214.
León-Calero, M., S. C. R. Valés, Á. Marcos-Fernández, and J. Rodríguez-Hernandez. 2021.
“3D printing of thermoplastic elastomers: Role of the chemical composition and
printing parameters in the production of parts with controlled energy absorption and
damping capacity”. Polymers 13, no. 20: 3551. https://doi.org/10.3390/polym13203551.
Mian, S. H., E. A. Nasr, K. Moiduddin, M. Saleh, and H. Alkhalefah. 2024. “An insight into
the characteristics of 3D printed polymer materials for orthoses applications:
Experimental study”. Polymers 16, no. 3: 403. https://doi.org/10.3390/polym16030403.
Papazetis, G., and G. C. Vosniakos. 2019. “Mapping of deposition-stable and defect-free
additive manufacturing via material extrusion from minimal experiments”.
International Journal of Advanced Manufacturing Technology 100, no. 9-12: 2207-
2219. https://doi.org/10.1007/s00170-018-2820-1.
Perry, S., V. Huayamave, B. Gonzalez, Z. Nadeau, and R. Rodriguez. 2022. “3D printing
material testing and applications in biomaterial modeling for pediatric medical
trainers”. In Volume 4: Biomedical and Biotechnology; Design, Systems, and
Complexity. American Society of Mechanical Engineers.
https://doi.org/10.1115/IMECE2022-94352.
Salem, Mohamed E. M., Qiang Wang, Ruoshi Wen, and Ma Xiang. 2018. “Design and
Characterization of Soft Pneumatic Actuator for Universal Robot Gripper”. In
International Conference on Control and Robots (ICCR), 6–10.
https://doi.org/10.1109/ICCR.2018.8534483.
Sardinha, M., L. Reis, T. Ramos, and M. F. Vaz. 2022. “Non-pneumatic tire designs suitable
for fused filament fabrication: An overvie”. Procedia Structural Integrity 42: 1098-
1105. https://doi.org/10.1016/j.prostr.2022.12.140.
Vidakis, N., M. Petousis, A. Korlos, E. Velidakis, N. Mountakis, C. Charou, and A. Myftari.
2021. “Strain rate sensitivity of polycarbonate and thermoplastic polyurethane for
various 3D printing temperatures and layer heights”. Polymers 13, no.16: 2752.
https://doi.org/10.3390/polym13162752.
Wang, J., B. Yang, X. Lin, L. Gao, T. Liu, Y. Lu, and R. Wang. 2020. “Research of TPU
materials for 3D printing aiming at non-pneumatic tires by FDM method”. Polymers
12, no. 11: 1-19. https://doi.org/10.3390/polym12112492.
Wickramasinghe, S., T. Do, and P. Tran. 2020. “FDM-Based 3D printing of polymer and
associated composite: A review on mechanical properties, defects and treatments”.
Polymers. 12, no. 7: 1529. https://doi.org/10.3390/polym12071529.
Xu, T., W. Shen, X. Lin, and Y. M. Xie. 2020. “Mechanical properties of additively
manufactured thermoplastic polyurethane (TPU) material affected by various
processing parameters”. Polymers 12, no. 12: 1-16.
https://doi.org/10.3390/polym12123010.

Engineering Manufacturing Letters 2:1 (2024) 45-52 51

Challenges on extrusion-based additive manufacturing of thermoplastic polyurethane
M. Sardinha, L. Ferreira, T. Ramos, L. Reis, M. Fatima Vaz

Zia, K. M., H. N. Bhatti, and I. A. Bhatti. 2007. “Methods for polyurethane and
polyurethane composites, recycling and recovery: A review”. Reactive and Functional
Polymers 67 no. 8: 675-692. https://doi.org/10.1016/j.reactfunctpolym.2007.05.004.

Acknowledgments
This work was supported by Fundação para a Ciência e a Tecnologia (FCT), through IDMEC, under
LAETA Base Funding (https://doi.org/10.54499/UIDP/50022/2020), and through CEGIST
(https://doi.org/10.54499/UIDB/00097/2020). Manuel Sardinha acknowledges FCT, for his PhD
research grant, https://doi.org/10.54499/2021.04919.BD.

Engineering Manufacturing Letters 2:1 (2024) 45-52 52