Innovative Uses of Thermoplastic Polyurethane (TPU) in 3D-Printed Medical

Applications
Boubakri Hanen*1, Naoufel Ben Moussa and Mohammed Habibi

1
Mechanical, Material and Processes Laboratory (LR99ES05), ENSIT, 5, Avenue Taha Hussein 1008, University of
Tunis. hanenboubakri55@gmail.com
2
Mechanical, Material and Processes Laboratory (LR99ES05), ENSIT, 5, Avenue Taha Hussein 1008, University of Tunis.
3
Preparatory Institute for Engineering Studies of Bizerte, University of Carthage, IPEIB, 7021 Zarzouna, Tunisia.
naoufel.benmoussa@ipeib.ucar.tn
4
Department of Mechanical Engineering, University of Québec Trois-Rivières, Trois-Rivières, QC, Canada.
mohamed.habibi@uqtr.ca
*Corresponding author Email: hanenboubakri55@gmail.com

Abstract:
The rapid evolution of smart materials has become a cornerstone in materials science and engineering. Shape memory
polymers (SMPs) have gained significant attention due to their wide range of applications. Polyurethane (PU), known for
its excellent mechanical properties, flexibility, and chemical resistance, is widely used across various industries. In the field
of 3D printing, thermoplastic polyurethane (TPU) is especially valued for producing flexible and durable parts. This study
focuses on TPU 90A, aiming to improve print quality, mechanical performance, and overall reliability. Key parameters
such as extrusion temperature, print speed, bed temperature, and retraction settings are optimized through a series of
experimental tests. Building on previous research, this paper introduces a three-dimensional finite strain model for thermo-
responsive shape memory polymers (SMPs). Implemented in Abaqus, the model is used to evaluate performance in
intelligent mechanical systems and to characterize shape memory behavior. The main strength of this study lies in the
development of a Fortran UMAT code specifically designed to predict the shape memory behavior of flexible polymers.
To assess the model’s performance, several numerical tests are conducted, beginning with uniaxial tests and advancing to
a shape memory simulation. Finally, comparisons with experimental data from the literature validate the model, confirming
its potential for advanced applications in smart medical devices. This study not only advances the understanding of TPU
in additive manufacturing but also establishes a reliable computational framework for designing adaptive medical devices,
highlighting its relevance for future biomedical innovations.

Keywords: Shape memory; Additive manufacturing; Modeling; Polyurethane.

1. Introduction
The economic imperative to develop advanced materials has accelerated the emergence of smart materials capable of
dynamically responding to external stimuli, such as temperature[1], mechanical stress, humidity [2], pH, electric, or
magnetic fields. These materials are classified based on their sensitivity into thermally, chemically, photosensitive, or
magnetically responsive categories. Some smart materials, like fiber optics incorporated into composite structures [3], serve
solely as sensors, while others, such as piezoelectric ceramics, can both sense and produce mechanical actions. This study
focuses on the actuation capabilities of smart materials applied in fields where precision and adaptability are crucial, such
as aerospace, automotive, robotics, and especially biomedical devices[4]. Shape memory polymers (SMPs) are among the
most researched materials due to their shape memory properties, low density, and the flexibility to adjust activation
temperature. These materials can regain their initial shape after deformation, making them particularly relevant for orthotic
applications. The shape memory effect in SMPs is determined by their internal structure, chemical composition,

1
macromolecular arrangement, and the process for programming temporary shapes. Common shape-fixing methods include
heating-loading-cooling cycles or cold drawing[5] [6], with the former widely applied in polyurethane-based polymers [7].
Despite some challenges related to high-temperature programming[8], alternative approaches, such as phase-transition
modeling[9], [10], [11] and linear viscoelastic models[12], [13], [14], have been explored to enhance SMP performance.
In this study, inspired by the work of Barot et al. (2008)[9], we propose a three-dimensional large-strain model for
thermally responsive SMPs, implemented in ABAQUS software through a UMAT subroutine to assess their behavior in
smart orthotic devices. The model simulates shape memory cycles, such as loading-unloading and shape recovery, applied
to an innovative 3D/4D printed orthotic made from polyurethane-based SMP. This thermoplastic, semi-crystalline
elastomer offers benefits like high elasticity and recovery capacity, with a deformation resilience up to 200% and a recovery
temperature close to body temperature, making it promising for biomedical applications. In summary, this study integrates
TPU's material optimization with SMP modeling, highlighting TPU 905A’s mechanical versatility and potential for
orthopedic devices. The paper details the constitutive equations, development of the UMAT subroutine, finite element
modeling, and analysis of TPU-based orthotic applications, concluding with insights into future improvements and
applications for smart, responsive materials in biomedical fields.
2. UMAT subroutine and numerical simulation
a. Choice of Model
In this study, we have chosen the phase transition approach, where the material becomes softer at higher temperatures and
retains a softer state at lower temperatures. This method effectively captures the dual-phase behavior, distinguishing
between a rubbery phase and a glassy state. By employing internal variables and constraints, as demonstrated in prior
research [9], [15], [16], this approach allows us to model the transition between phases accurately. This choice provides a
robust framework for representing the thermal and mechanical responses essential to our application in shape memory
polymers.
. σ = f(ε, T) = σg + σr (1)

where: 𝜎 is the total stress in the material is the strain and T is the temperature. 𝜎𝑟 and 𝜎𝑔 are the stresses in the rubbery
and glassy phases, respectively.
The equation of the stress for cooling phase of the cycle can be written as

𝜎 = −𝑝𝐼 + (1 − 𝑍𝑔 )𝜇𝑎 𝐵𝑘𝑎 + 𝜇𝑔 𝑍𝑔 𝐵𝑘𝑔 (2)

where 𝜇𝑔 is the shear modulus of the glassy phase and 𝑍𝑔 is the glassy volume fraction. , 𝐵𝑘𝑔 is the left Cauchy stretch
tensor, which has been a relation with deformation gradient in glassy phase

The progression of phase fractions between the glassy and rubbery states is typically governed by a phase fraction equation
that is often dependent on temperature.
𝑍𝑟 (𝑇) = 1 − 𝑍𝑔 (𝑇) (3)

where: 𝑍𝑔 (𝑇) is the fraction of the material in the glassy phase and 𝑍𝑟 (𝑇) is the fraction in the rubbery phase. According
to the previous research [9], a function describing the following evolution for the volume fraction of the glassy phase Zg
is introduced into the constitutive model to explain the changes of polymer material stress and strain during the process of
shape memory which is expressed as:

𝛳
𝑍𝑔 = 𝐺𝐺 ∗ (𝛼) ∗ (1 − ( ) − 𝜇𝑎 ∗ (𝐼𝐵𝑎 − 3) + 𝜇𝑐 ∗ (𝐼𝐵𝑐 − 3) 𝑖𝑓 𝛳 ≤ 𝛳𝑡 (4)
𝛳𝑡

𝛳
𝑍𝑔 = 𝐺 ∗ (𝛼0 − 𝛼) ∗ (1 − ( ) − 𝜇𝑎 ∗ (𝐼𝐵𝑎 − 3) 𝑖𝑓 𝛳𝑡 ≤ ϴ
𝛳𝑡

2
where the constant αo represents the maximum crystallinity, G is a constant, and ϴ and 𝛳𝑡 are the temperature and transition temperature
respectively.

3. Experimental validation
a. Material

The material selected for this study is TPU, chosen for its elastic properties and compatibility with medical applications.
The tensile test and shape memory behavior of the selected TPU were analyzed using properties outlined in Table 1, which
includes key material characteristics derived from the work of [17] . Tensile simulations in Abaqus focused on evaluating
the TPU’s Young’s modulus and tensile strength, providing insights into its performance under load. Additionally, the
shape memory behavior was studied by simulating thermal and mechanical recovery processes, where the TPU’s ability to
return to its original shape was assessed under various temperature and deformation conditions. These analyses highlighted
the material’s elasticity and adaptability, confirming its suitability for medical orthotic applications.

Table 1 Model parameters assumed for the comparison with experimental results by Volk et al.(2011)

Symbol Value Unit

Er 19 MPa
Eg 1636 MPa
νr 0.35 -
νg 0.35 -
θt 336 K

b. Validation of simulation results

To validate the model, numerical results were compared with experimental data from [17] through a free-recovery test
on a polyurethane-based SMP (Fig.1). The test involved shape-fixing at high temperatures, followed by unconstrained
shape recovery upon heating. Fig. 2 displays a close match between experimental and numerical data for stress-extension
behavior, with a linear response during loading and the shape memory cycle. Using the Neo-Hookean model enhanced the
accuracy of the unloading curve relative to the experimental data. However, minor differences appeared in stress and
extension versus temperature plots, as the experimental stress path varies between heating and cooling. The proposed model
links stress evolution to the glassy phase volume fraction, which behaves differently in these thermal cycles. Additionally,
Fig. 3 shows the trend of the volumetric fraction of the glassy phase, Zg, as a function of temperature. It can be observed
that the material's behavior is influenced by Zg; notably, the trend is symmetric with respect to the transition temperature,
θt=336 K.

3
 .

Fig. 1 Shape memory cycle: (a) initial sample (b) Loading step(c) cooling step (d) unloading step (e) heating step.

25

20
Stress(Mpa)

15

10

0
0,00E+00 5,00E+00 1,00E+01 1,50E+01
-5
Strain(%)

Simulation results Experimental results

Fig. 2 Stress-extension curves

simulation results experimental results

1,40E+01
1,20E+01
1,00E+01
8,00E+00
Strain(%)

6,00E+00
4,00E+00
2,00E+00
0,00E+00
-2,00E+00 290 310 330 350 370 390
Temperature(K)

Fig. 3 Extension–temperature curves

4
 1,2

Frozen Volume Fraction

1

0,8

0,6

0,4

0,2

0
320 330 340 350 360 370
Temperature(K)

Fig. 4 Trend of the volume fraction of glassy phase with temperature

4. Application
This study presents an experimental investigation into the suitability of TPU (Thermoplastic Polyurethane) with hardness
levels 90A for medical orthotic applications. A series of mechanical tests was conducted, starting with tensile testing to
determine Young's modulus, tensile strength, and elongation at break, followed by shape memory tests to evaluate TPU's
ability to recover its original shape under thermal conditions. In this paper, a shape memory application for 3D-printed
TPU finger orthoses (Fig.5) is developed to replace traditional methods and better adapt to finger deformities. By
leveraging, the shape memory properties of thermoplastic polyurethane (TPU), these orthoses provide customized support
that conforms precisely to the patient’s anatomy. Unlike conventional orthotics, which often require manual adjustment
and can be cumbersome, the TPU finger orthoses utilize temperature-dependent memory effects. This allows them to be
molded easily to the unique shape of each finger and later adjusted through heat application as the patient's needs evolve.
This approach offers enhanced comfort and functionality, making it a promising alternative in personalized orthotic care.
Results indicate that TPU demonstrates strong mechanical properties, flexibility, and adaptability, supporting its use in
patient-specific orthotic applications.

Fig. 5 Geometry of the finger orthoses

5. Conclusion
This study explores the application of thermoplastic polyurethane (TPU) in 3D-printed medical devices, focusing on the
development of shape memory finger orthoses that adapts to individual deformities. Shape memory polymers (SMPs) like
TPU can return to their original shape with heat, making them ideal for personalized orthotic solutions. Since Abaqus lacks
built-in models for SMP behavior, a custom UMAT code was developed to simulate the unique thermomechanical and
viscoelastic properties of these materials. This UMAT code enables realistic modeling of the shape memory cycle,
enhancing the precision and adaptability of TPU-based medical devices.

5
 Reference
[1] B. Zhou, S. C. Zhao, H. Qi, et L. M. Zhou, « Numerical Simulations on Shape Memory Effect of Shape Memory
Polymer », KEM, vol. 452‑453, p. 465‑468, nov. 2010, doi: 10.4028/www.scientific.net/KEM.452-453.465.
[2] S. Chen, J. Hu, C. Yuen, et L. Chan, « Novel moisture-sensitive shape memory polyurethanes containing pyridine
moieties », Polymer, vol. 50, no 19, p. 4424‑4428, sept. 2009, doi: 10.1016/j.polymer.2009.07.031.
[3] Z. Ding, C. Yuan, X. Peng, T. Wang, H. J. Qi, et M. L. Dunn, « Direct 4D printing via active composite materials »,
Sci. Adv., vol. 3, no 4, p. e1602890, avr. 2017, doi: 10.1126/sciadv.1602890.
[4] R. Liu, S. Xu, X. Luo, et Z. Liu, « Theoretical and Numerical Analysis of Mechanical Behaviors of a Metamaterial-
Based Shape Memory Polymer Stent », Polymers, vol. 12, no 8, p. 1784, août 2020, doi: 10.3390/polym12081784.
[5] A. Lendlein et S. Kelch, « Shape-Memory Polymers », Angewandte Chemie International Edition, vol. 41, no 12, p.
2034‑2057, 2002, doi: 10.1002/1521-3773(20020617)41:12<2034::AID-ANIE2034>3.0.CO;2-M.
[6] A. Lendlein et V. P. Shastri, « Stimuli-Sensitive Polymers », Advanced Materials, vol. 22, no 31, p. 3344‑3347, 2010,
doi: 10.1002/adma.201002520.
[7] J. N. Rodriguez et al., « In vivo response to an implanted shape memory polyurethane foam in a porcine aneurysm
model », J Biomedical Materials Res, vol. 102, no 5, p. 1231‑1242, mai 2014, doi: 10.1002/jbm.a.34782.
[8] G. Li et W. Xu, « Thermomechanical behavior of thermoset shape memory polymer programmed by cold-
compression: Testing and constitutive modeling », Journal of the Mechanics and Physics of Solids, vol. 59, no 6, p.
1231‑1250, juin 2011, doi: 10.1016/j.jmps.2011.03.001.
[9] G. Barot, I. J. Rao, et K. R. Rajagopal, « A thermodynamic framework for the modeling of crystallizable shape
memory polymers », International Journal of Engineering Science, vol. 46, no 4, p. 325‑351, avr. 2008, doi:
10.1016/j.ijengsci.2007.11.008.
[10] Y.-C. Chen et D. C. Lagoudas, « A constitutive theory for shape memory polymers. Part I: Large deformations »,
Journal of the Mechanics and Physics of Solids, vol. 56, no 5, p. 1752‑1765, mai 2008, doi:
10.1016/j.jmps.2007.12.005.
[11] J. H. Kim, T. J. Kang, et W.-R. Yu, « Thermo-mechanical constitutive modeling of shape memory polyurethanes
using a phenomenological approach », International Journal of Plasticity, vol. 26, no 2, p. 204‑218, févr. 2010, doi:
10.1016/j.ijplas.2009.06.006.
[12] J. Diani, Y. Liu, et K. Gall, « Finite strain 3D thermoviscoelastic constitutive model for shape memory polymers »,
Polymer Engineering & Sci, vol. 46, no 4, p. 486‑492, avr. 2006, doi: 10.1002/pen.20497.
[13] H. Tobushi, T. Hashimoto, S. Hayashi, et E. Yamada, « Thermomechanical Constitutive Modeling in Shape Memory
Polymer of Polyurethane Series », Journal of Intelligent Material Systems and Structures, vol. 8, no 8, p. 711‑718,
août 1997, doi: 10.1177/1045389X9700800808.
[14] R. Xiao, J. Choi, N. Lakhera, C. M. Yakacki, C. P. Frick, et T. D. Nguyen, « Modeling the glass transition of
amorphous networks for shape-memory behavior », Journal of the Mechanics and Physics of Solids, vol. 61, no 7, p.
1612‑1635, juill. 2013, doi: 10.1016/j.jmps.2013.02.005.
[15] Y.-C. Chen et D. C. Lagoudas, « A constitutive theory for shape memory polymers. Part II: A linearized model for
small deformations », Journal of the Mechanics and Physics of Solids, vol. 56, no 5, p. 1766‑1778, mai 2008, doi:
10.1016/j.jmps.2007.12.004.
[16] J. H. Kim, T. J. Kang, et W.-R. Yu, « Thermo-mechanical constitutive modeling of shape memory polyurethanes
using a phenomenological approach », International Journal of Plasticity, vol. 26, no 2, p. 204‑218, févr. 2010, doi:
10.1016/j.ijplas.2009.06.006.
[17] B. L. Volk, D. C. Lagoudas, et D. J. Maitland, « Characterizing and modeling the free recovery and constrained
recovery behavior of a polyurethane shape memory polymer », Smart Mater. Struct., vol. 20, no 9, p. 094004, sept.
2011, doi: 10.1088/0964-1726/20/9/094004.