Enhancing shape memory properties of multi-layered and multi-material polymer composites in 4D printing

2.1.1. Design and theory.

As shown in figure 2(a), a multi-layered SMPC consists of two SMP layers on the sides and an elastomer layer in the middle. The governing equilibrium equation for a programmed composite is

$$\begin{equation}\int\limits_0^t {\sigma _x}w{\text{ d}}z = 0\end{equation} \tag{ 1 }$$

Zoom In Zoom Out Reset image size

where ${\sigma _x}$ is the stress in the axial direction, and $t$ and $w$ are the stripe thickness and width, respectively. We aim to program a three-layer flat composite to have a curved shape, when programmed (i.e. in its temporary state). When the composite is programmed (figure 2(b)), two SMP layers maintain their programmed shape, while the elastomer layer tends to repossess its original shape. In the programmed state of SMPC, there is a static equilibrium in the structure. Therefore:

$$\begin{equation}2{M_{\text{S}}} - {M_{\text{E}}} = 0\end{equation} \tag{ 2 }$$

where ${M_{\text{E}}}$ is the moment exerted on the elastomer layer by the SMP layers and ${M_{\text{S}}}$ is the moment applied on each SMP layer by the elastomer.

2.1.2. Production of lattice plates.

PLA and PLA-TPE lattice structures (M-L) with PLA to TPE thickness ratios of 5:1, 4:2, and 3:3 were produced by fused deposition modeling (FDM) AM machine INDU (TOP 3DP, Iran). To investigate the effect of the layered structure of lattices on shape and force recovery, a two-layered PLA lattice with equal ratios (M-L PLA 3:3) is produced and glued together to be compared to the PLA lattice. The reason behind using the FDM method is its straightforward functioning, low production time, and low cost. On the other hand, employing innovative materials [36, 37] and reinforced composites [38, 39] was noticed with FDM 3D printers in recent years. Since initial printing conditions can affect the mechanical properties and shape memory behavior of SMPC, all specimens are printed under identical conditions. The parameters used for 3D printing multi-layered specimens are presented in table 1. The TPE layer is in the middle part of all the multi-layered structures, and PLA is placed symmetrically on its sides (figure 3(a)).

Zoom In Zoom Out Reset image size

Table 1. Summary of 3D printer configurations used for manufacturing the lattice plates.

Printer setting	Value	Printer setting	Value
Layer height	0.2 mm	TPE nozzle temperature	240 °C
Shell thickness	0.4 mm	PLA-TPE nozzle temperature	200 °C
Fill density	100%	Print-bed temperature	50 °C
Print speed	20 mm s−1	PLA filament diameter	1.75 mm
PLA nozzle size	0.4 mm	TPE filament diameter	1.75 mm
TPE nozzle size	0.4 mm	PLA-TPE filament diameter	1.75 $\pm$ 0.1 mm
PLA nozzle temperature	200 °C

2.2.1. Design and theory.

From a microscopic point of view, thermoplastic materials like PLA have a disoriented molecular structure which makes them amorphous. Alternatively, elastomeric materials such as TPE usually have an oriented molecular structure that can potentially be exploited to improve the crystallinity of thermoplastic materials, as schematically demonstrated in figure 4. This increase in crystallinity would enhance SME in the resulting material. The length of polymeric chains in amorphous materials is shorter than crystalline materials, making amorphous polymers more vulnerable to damage during deformations compared to crystalline polymers. Blending a crystalline polymer with an amorphous polymer will lower damage to the chains under large deformations. Thus, SMP and elastomer blending enable us to tune the SME of the mixture [40].

Zoom In Zoom Out Reset image size

2.2.2. Production of multi-material SMPC filaments.

Several multi-material filaments composed of different volumetric ratios of PLA and FLEX TPE (YouSu, China) were produced. First, PLA and TPE filaments with PLA to TPE volumetric ratios of 5:1, 4:2, and 3:3 are chosen and shredded using a pelletizing machine. A mixing machine (Brabender, Germany) is used to merge filaments at 200 °C in 5 min. The mixing machine measures and displays the instantaneous torque resistance of the material. When the measured torque becomes stable, it implies that the mixture has become homogenous (figure 5(a)). Then, after being blended employing a hot press machine, the materials are formed into thin sheets (figure 5(b)). Using the hot press machine, the materials are heated up to 200 °C and pressurized up to 90 bar. In the end, materials are transformed into granules for the filament production process (figure 5(c)).

Zoom In Zoom Out Reset image size

Producing filament with the desired quality, diameter, and extrudability has always been challenging [41]. This research, produces filaments with decent quality using an extruder developed by the School of Mechanical Engineering of the University of Tehran (figure 5(d)). First, the temperature of the extruder nozzle should be set to an appropriate level to result in a consistent filament diameter. To this end, we use a trial-and-error approach for the extruder temperature in the temperature interval of 160 °C–230 °C. Afterward, PLA-TPE granules are added to the hopper. The motor's rotational speed is set to 35 rpm. After leaving the nozzle, the extruded filaments are cooled down using a fan. By this method, filaments made of PLA-TPE with volumetric ratios of 4:2 and 3:3 are produced at the extruder temperature intervals of 180 °C–190 °C and 190 °C–200 °C, respectively. However, no optimal extruder temperature could produce 5:1 volumetric ratio PLA-TPE filaments with a consistent diameter. FESEM images in the section 3.2 further investigate this phenomenon.

2.2.3. Production of lattice plates.

To measure ${T_{\text{g}}}$ of each material, DMTA was performed. The results are essential for thermomechanical shape memory tests which are essential for calibrating smart materials constitutive models. DMTA is carried out by applying a fluctuating sinusoidal load with a frequency of ${ }1{\text{ Hz}}$ on each specimen. Various properties such as storage modules ($E{^{{\prime}}}$), loss modules ($E{^{{\prime} ^{\prime}}}$), tan($\delta$), and glass transition temperature (${T_{\text{g}}}$) are obtained by this method. The glass transition temperature was obtained by $E{^{^{\prime}}}$ onset method [42]. To obtain the ${T_{\text{g}}}$ of the PLA filaments as well as the fabricated PLA-TPE filaments with different volumetric ratios, DMTA is performed in the temperature intervals of $- 20$ °C to $90$ °C and $- 80$ °C to $150$ °C, respectively. It is expected that adding TPE decrease ${T_{\text{g}}}$ of the resulting material. Therefore, the temperature interval for the fabricated PLA-TPE specimens is chosen to be in a broader range.

In this section, conditions of three-point bending are discussed. Figure 6 demonstrates a schematic of complete shape recovery and force recovery processes in a three-point bending test [26]. The SANTAM device measured the required force for a specific displacement for each specimen. To obtain the shape and force recovery, specimens are placed in respectively unconstrained and constrained configurations to measure the displacement and force due to shape and force recovery. For programming of the specimens, a displacement of 4.5 mm ⁵ (which leads to 10% principal strain in the bulk material) was applied on all of them at the programming temperature (${T_{\text{p}}}$). Liquid water is used to control the temperature throughout tests. The advantage of using water in thermomechanical tests is that the temperature of water propagates to all the regions of the specimen uniformly, and thus, the specimen possesses an isothermal condition.

Zoom In Zoom Out Reset image size

The thermomechanical shape recovery test in three-point bending mode for SMPCs is schematically shown in figure 6(a). As mentioned in section 3, water is used for the thermomechanical tests and the temperature is controlled by hot- and cold-water flows. Each specimen is placed separately in a vessel containing room-temperature water and then heated to a higher temperature of ${T_{\text{h}}}$ (step 1). A displacement of 4.5 mm is applied to the specimen using a pin attached to the load cell at the higher temperature (step 2). Then, a cold-water flow is fed to the vessel by a pump, and a drain pump depletes the hot water (step 3). After cooling the vessel, the specimen maintains its programmed shape, after which the pin is removed from the loading point (step 4). The specimen may demonstrate an insignificant elastic strain. In the last step for the shape recovery, the water temperature is arisen to ${T_{\text{h}}}$ and the specimen recovers its initial state (step 5i). Alternatively, in the force recovery case, the load cell pin is placed at the middle of the specimen to restrain its shape recovery. The force exerted by the specimen, is measured during its recovery process triggered by reheating to ${T_{\text{h}}}$ (step 5ii). Force recovery ratio is defined by the ratio of the maximum force recovery to the pre-force spent on the programming of the specimen.

Equation (3) is used to calculate the shape recovery ratio in the shape memory cycle [41]. In this equation, the applied deflection of the beam is 4.5 mm and ${X_{{\text{unrecovered}}}}$ is the unrecovered portion of the specimen deflection at 64 °C

$$\begin{equation}R = \left( {\frac{{{X_{{\text{deformed}}}} - {X_{{\text{unrecovered}}}}}}{{{X_{{\text{deformed}}}}}}} \right) \times 100.\end{equation} \tag{ 3 }$$

It should be noted that microscopic camera and image processing techniques are used to measure the beam deflection shape recovery of the specimens.

In this research, storage modules ($E{^{^{\prime}}}$) versus temperature diagrams are used to measure the glass transition temperature ${T_{\text{g}}}$. For PLA it was found to be ${ }6{\text{4 }}^\circ {\text{C}}$ (figure 7(a)). In developing multi-material specimens, ${T_{\text{g}}}$ was decreased around ${\text{8 }}^\circ {\text{C}}$ (figure 7(b)) by adding a minor fraction of TPE to PLA (1:5 ratio). Further increase in the TPE volume fraction resulted in an only minor reduction of ${T_{\text{g}}}$, as depicted in figure 7(c). Ultimately, adding more TPE to PLA resulting in equal volumetric ratios, lowered ${T_{\text{g}}}$ about $1{\text{0 }}^\circ {\text{C}}$ as compared to pure PLA (figure 7(d)). The reduction of ${T_{\text{g}}}$ is a significant improvement in SMPC characteristics which can potentially increase their usage in many applications particularly in biomedical applications.

Zoom In Zoom Out Reset image size

In the material design stage, FESEM images of the developed materials are utilized to examine the effect of the volumetric ratio on the miscibility of PLA-TPE in filament form. The three multi-material filament specimens with different volumetric ratios of PLA-TPE were initially cooled down in liquid nitrogen to provide the possibility of a brittle fracture. FESEM images of the fractured cross-sections of the filament specimens are presented in four scales in figure 8. Higher TPE fractions have led to a better mixture between two constituents, forcing PLA polymeric chains to adapt TPE molecular structures. Hence, porosity and voids of the multi-material specimens have been decreased in higher levels of TPE fraction. On the other hand, as a thermoplastic material, PLA tends to develop more voids when heated, which is abundant in the fabricated M-M 5:1 filament. As one may observe from figure 8, the M-M 5:1 mixture has led to dissatisfaction and, therefore, varying filament diameter due to the phase separation at low TPE fraction. Thus, M-M 5:1 is not considered in the following tests.

Zoom In Zoom Out Reset image size

XRD analysis is based on x-ray beams and Bragg's law which is usually utilized to study materials with crystalline structures [43]. Since the diffracted spectrum of each produced material is unique, XRD is utilized to study the crystallinity of the specimens. Figure 9 shows the XRD result of four PLA, TPE, M-M 4:2 and 3:3 specimens. Due to the phase separation in M-M 5:1 and varying filament diameter, an XRD test was not conducted for this ratio. As illustrated in figures 9(a) and (b), PLA is amorphous due to the low and mild peak in the spectrum, while TPE has an intense peak with a large magnitude which means it can be considered a crystalline material. Figure 9(c) shows an intense peak in M-M 4:2 meaning it must have a highly crystalline structure. By adding more TPE to the mixture to obtain M-M 3:3, the crystallinity of the PLA-TPE specimen is decreased, as demonstrated in figure 9(d). In other words, adding TPE to PLA in lower ratios can highly increase the crystallinity of the mixture, while higher ratios of TPE disorient the molecular structures resulting in less crystallinity in the mixture.

Zoom In Zoom Out Reset image size

Reheating during the shape recovery process in the three-point bending test (step 5i in figure 6) is shown for the specimens in figure 10. In the PLA and multi-layered lattice cases, the temperature has been raised to $6{\text{4 }}^\circ {\text{C}}$ (${T_{\text{g}}}$ of PLA). Since reducing the programming temperatures broadens the range of applications of an SMPC material, a lower programming temperature of ${T_{\text{p}}} \cong 5{\text{0 }}^\circ {\text{C}}$ (close to their ${T_{\text{g}}}$) was considered for the multi-material specimens. This can justify our purpose of expanding the applications of SMPs in different areas, as discussed in section 4.4. In addition, PLA has an amorphous structure which makes it vulnerable to damage at $T < {T_{\text{g}}}$, therefore, the programming step for PLA and multi-layered specimens cannot be carried out in lower temperatures.

Zoom In Zoom Out Reset image size

Shape recovery versus temperature diagrams are presented in figure 11, revealing that the PLA lattice has the lowest shape recovery extent. Furthermore, the layered structure of PLA lattice does not affect much the shape recovery of the samples. As depicted in figure 11(a), the slope of the multi-layered specimens during the shape recovery process is higher than those in the PLA and multi-material lattices. In other words, the shape recovery speed of the multi-material specimens is slower than that of the multi-layered specimens. Technically, the recovery process of multi-material specimens is started in lower temperatures compared to multi-layered specimens. This does not mean that multi-material specimens have a faster shape recovery, since the recovery speed is the slope of the recovery curve. As shown in figure 11(a), the slope of multi-layered specimens is larger than the slope of multi-material counterparts. From a macroscopic point of view, the multi-material specimens are composed of a single material of PLA-TPE, while multi-layered specimens are comprised of two different materials of PLA and TPE. Therefore, the elastic characteristic of TPE in multi-layered specimens expedites the recovery process and enhances the maximum shape recovery, once the PLA phase transforms from glassy to rubbery.

Zoom In Zoom Out Reset image size

Note that the reported values in figure 11(b) are obtained by calculating the shape recovery ratio (i.e. $R$ in equation (3)) at both $5{\text{0 }}^\circ {\text{C}}$ and $6{\text{4 }}^\circ {\text{C}}$ for each specimen. In other words, even though ${T_{\text{g}}}$ for multi-material specimens is lower than that in PLA and multi-layered specimens, the shape recovery process of multi-material specimens is also reported for temperatures up to $6{\text{4 }}^\circ {\text{C}}$. As speculated in section 2, the addition of elastomer material (TPE) has enhanced the SME of SMPs. In general, the thermomechanical test results (figure 11) unveil that the shape recovery of multi-layered and multi-material lattice plates has remarkable improvements compared to the pure PLA lattice.

The shape recovery of PLA and the multi-layered lattices started at $4{\text{6 }}^\circ {\text{C}}$, while the shape recovery process in multi-material specimens started at lower temperatures (i.e. ∼$3{\text{5 }}^\circ {\text{C}}$) (figure 11). Among multi-layered specimens, the maximum shape recovery belonged to M-L 4:2 lattice, while the maximum shape recovery in multi-material specimens was recorded in M-M 4:2 case (figure 11(b)). The multi-material lattices were almost fully recovered at $5{\text{0 }}^\circ {\text{C}}$, while at that temperature, PLA and multi-layered lattices were just about to start their recovery process. As expected, raising the temperature of the multi-material specimens to $6{\text{4 }}^\circ {\text{C}}$ did not changed their shape recovery response. This expedition in recovery process of multi-material lattices can potentially make the PLA-TPE specimens more beneficial to practical applications. As discussed in XRD test results, M-M 4:2 has a higher crystallinity and thus longer polymeric chains and more oriented micro-molecular structures than M-M 3:3, which results in a more significant shape recovery for M-M 4:2.

The results of force recovery in constrained thermomechanical tests are illustrated in figure 12. Figure 12(a) plots the maximum force recorded for PLA and multi-layered specimens at $6{\text{4 }}^\circ {\text{C}}$ and for the multi-material specimens at $5{\text{0 }}^\circ {\text{C}}$. The cooling step was performed after applying the required displacement at the designated temperatures, which led to a zero force for PLA and multi-layered specimens. However, the measured force after the cooling step for the multi-material specimens did not approach zero as the increased crystallinity in the specimens had enhanced their stiffness.

Zoom In Zoom Out Reset image size

The peaks in figure 12(b) show each specimen's maximum force recovery value during the reheating step. At the beginning of this process (up to $T = 3{\text{0 }}^\circ {\text{C}}$), the force recovery value is negligible. By further increasing the temperature, the force recovery of PLA and multi-layered lattices started at a temperature interval of $35{ }^\circ {\text{C}}$−$4{\text{0 }}^\circ {\text{C}}$. On the other hand, the force recovery process for the multi-material specimens started from the very beginning of the reheating process. The force recovery speeds of the multi-material specimens are higher than those of PLA and multi-layered specimens. This is because the temperature interval for the force recovery process of multi-material specimens is smaller than that of PLA and multi-layered specimens. It should be noted that the structure of the M-L PLA compared to pure PLA, has some constraints due to the printing process occurring between layers which makes the structure have a higher stored energy and therefore have a higher recovered force.

As shown in figure 12(c), the maximum force recovery in the multi-material lattices is much greater than in PLA and multi-layered lattices due to the increased crystallinity of multi-material specimens. The recovery ratio for the PLA specimen was the highest recorded one, while the multi-material specimens had the lowest recovery ratio (figure 12(c)). It should be noted that, in many applications, the magnitude of pre-force in the programming stage is not considered as a limiting factor. Under this consideration, the maximum force recovery of multi-material specimens is the highest recorded value.

Dashed-black contours indicate the relaxation step of PLA and multi-material specimens in figure 12(b), and it is elaborated with a force-time diagram in figure 12(d). Relaxation is calculated from the maximum force to the final force in the force-time diagram, shown in figure 12(d). The final force exerted by the PLA specimen approached zero, while this value for multi-material specimens approached non-zero values (about 2.5 N for M-M 4:2 and 4 N for M-M 3:3), which are desirable in many applications.

SMPs have various applications in self-shape-shifting structures which their performance can be improved by higher shape and force recovery. For instance, self-folding origami structures can provide a remote and automatic assembly to minimize cost, time, and packaging efforts compared to conventional machining methods. For this purpose, self-folding SMP hinges and heating circuits are developed to generate the required torque for rotating hinges which can be used to self-assemble an origami-based robot [44, 45]. Self-folding structures can be built as a composite consisting of SMPsactivated by uniform heating [46]. These self-folding mechanisms can have applications in micro-scaled systems for various purposes such as antennas [47] and machine assembly [48]. Shape-shifting behavior can be manipulated by implementing a hyperelastic layer in bi- or multi-layered SMP strips, which yield self-twisting/self-rolling or self-folding origami structures [49].