Microscopic morphology, thermodynamic and mechanical properties of thermoplastic polyurethane fabricated by selective laser sintering

TPU powders (bulk density of 1.2 g cm⁻³) were obtained from laser sintering powders LUVOSINT^® Lehmann & Voss, German, and used as received. The scanning electron microscopy (SEM) image (figure 1a) shows the irregular shape but relative smooth surfaces of TPU powders with a dominant particle size of 40–80 μm and an average size of 79.1 μm (figure 1b), which are suitable for SLS manufacturing process [42].

Zoom In Zoom Out Reset image size

The effects of SLS process parameters on the mechanical properties of TPU were investigated through the designing orthogonal experiments (table 1) and the results were calculated and analyzed to find out the optimal strategy. The relevant factors were scanning power, scanning speed and layer thickness. The orthogonal table L₉(3⁴) was designed in such manner that each factor has three levels (20, 23 and 25 W), set as scanning power (factor A), scanning speeds (factor B) of 4000, 4500 and 5000 mm s⁻¹ were applied and layer thickness (factor C) of 0.1, 0.13 and 0.15 mm were used (please refer to supporting information tables S1 and S2 is available online at stacks.iop.org/MRX/7/055301/mmedia).

The TPU powders were manufactured via SLS equipment (HK S320 SLS system, Wuhan Huake 3D Technology Co. Ltd, China). A continuous wave CO₂ laser (k = 10.6 μm) was equipped, which could be adjusted from 0 to 30 W. The powder was preheated by infrared radiation heating pipes before sintering. The effect parameters on the quality of SLS fabricated components included the laser power (p), laser scanning speed (v), laser scanning spacing (s), layer thickness (h) and powder bed temperature (T_b).The energy density (ED) was generally defined as the amount of energy input in a unit volume, which can be calculated by the equation (1) [43].

$$\begin{eqnarray}&&ED=\displaystyle \frac{P}{S\times V}\end{eqnarray} \tag{ 1 }$$

Three major parameters were fluctuated at discrete time intervals, including laser power, scanning speed and laser scanning spacing. Simultaneously, the rest parameters were set as fixed values (table 2). A 0.4 mm laser beam was used and the scanning spacing was set at 0.2 mm to ensure an overlap of 0.2 mm to the previous scan line. The specific SLS processing parameters were shown in Table S2.

Fourier transform infrared spectroscopy (FTIR) implemented by using a Bruker Tensor 27 instrument was used to measure the absorption of infrared radiation with an absorbance pattern of 600 to 4000 cm⁻¹.

The crystalline characteristics of TPU powder and SLS fabricated specimens were analyzed by X-ray diffractometer technique (XRD, D8 ADVANCE, Bruker, Germany), using Cu Kα radiation in the 2θ range between 5° and 75° at sampling pitch of 0.02°, preset time of 5 seconds and scanning rate of 3 °/min.

A Zwick/Roell Z020 with a 1 kN load cell and 1 kN pneumatic clamps was used to carry out the tensile testing at room temperature according to the standard ISO 527-2 1B. Three samples ('dog bone' shape) were tested with a crosshead speed of 50 mm/min until the failure and the average value was reported. Since elastomers typically showed a extremely large elongation, the strain was calculated as the ratio of the extension of the clamps to the initial distance between the two clamps. The toughness of the samples was evaluated by the area under the stress-strain curve [44] and average value was reported after three repeated tests.

Thermal properties of SLS specimens were evaluated by Differential Scanning Calorimetry (DSC) (Perkin Elmer Instruments Ltd) under a nitrogen atmosphere. Specimens were ramped up to 250 °C at a heating rate of 10 °C/min and isothermally held for 5 mins. Afterwards the samples were ramped down to − 60 °C with a cooling rate of 5 °C/min.

The morphologies of the cross-section areas of the factured specimens (coated with platinum) after tensile tests were observed using JSM-7600F Scanning Electron Microscope (SEM) at the voltage of 10.0 kV.

All statistical analysis of the assay results was performed by variance (ANOVA) using GraphPad Prosm 6.0 (GraphPad Software, La Jolla, CA, USA). The multiple comparisions were performed by one-way ANOVA with a criterion α = 0.05.

The effect of three factors on mechanical properties of TPU specimens was studied by orthogonal tests. The orthogonal table L₉(3⁴) is shown in table S1. The tensile properties of the specimens were listed in table 3. The range analysis of orthogonal experiments was shown in table 4, k₁, k₂ and k₃ are the mean values of the tensile strength and elongation at the same level, which can be calculated by equation (2).

$$\begin{eqnarray}&&K=\displaystyle \frac{{S}_{1}+{S}_{2}+{S}_{3}}{3}\end{eqnarray} \tag{ 2 }$$

Where S₁, S₂ and S₃ are the average tensile strength of scanning power (factor A), scanning speed (factor B) and layer thickness (factor C), respectively. For the same factor, range R is the average value difference between the maximum and the minimum values of K at the different levels.

Without any interaction between three factors, the result of R_C > R_B > R_A is shown in table 4, indicating that the most remarkable factor for the tensile strength and elongation is layer thickness, then scanning speed and at last the laser power. For laser power (factor A), k₃ > k₂ > k₁, which means as scanning power increases, the tensile strength and elongation of testing samples increase. For scanning speed (factor B), k₁ > k₂ > k₃ indicates that as the scanning speed increases, the tensile strength and elongation of testing samples decrease. For layer thickness (factor C), the trend of k₁ > k₂ > k₃ indicates that when layer thickness increases, the tensile strength and elongation decrease. When the scanning speed reaches the intermediate value, the tensile strength and elongation are the highest. The best combination of three factors should be A3-B1-C1, namely laser power of 25 W, scanning speed of 4000 mm s⁻¹ and layer thickness of 0.1 mm. However, the layer thickness should be larger than particle size and the average particle size of the powder was 79.1 μm. In that case, the minimum layer thickness is 0.1 mm.

In this research, the specimen produced at the parameters of 20 W, 4500 mm s⁻¹ and 0.15 mm displayed the lowest tensile strength, which was chosen as the reference to compare with the specimen fabricated at the optimisation conditions.

The maximum laser power of the equipment is 25 W. Hence, the laser scanning power of 25 W, layer thickness of 0.1 mm, and five different laser scanning speed values of 3000, 3250, 3500, 3750 and 4000 mm s⁻¹ were used to further optimize the processing parameters and investigate the effect of laser scanning speed on the tensile properties of SLS fabricated TPU products (table 5). Four samples are made for each set of parameters.

The peak at 3332 cm⁻¹ specifies the N–H group in urethane (–NHCOO–), and the peaks at 2955 and 2850 cm⁻¹ belong to the asymmetric and symmetric vibration of the –CH₂ group (figure 2). The transmittance of TPU specimens decreases after sintering, and there is a strong similarity about the results of TPU specimens produced under different parameters.

Zoom In Zoom Out Reset image size

The diffraction peak at 2θ = 20° is the characteristic of the hard segment of TPU crystalline structure [45]. Compared with TPU raw powder, the diffraction intensity of SLS specimen (3750 and 4000 mm s⁻¹) decreases (figure 3). FTIR and XRD results showed that the TPU powders were printed successfully without chemical reactions occurred during sintering because of no extra chemical bonds.

Zoom In Zoom Out Reset image size

The tensile strength of the TPU samples made by different processing parameters are shown in figure 4. It can be found that as the scanning speed increases, the tensile strength of samples first increases, and reaches the maximum value at the scanning speed of around 3500 mm s⁻¹. Then, the tensile strength begins to decrease, with the increasing of scanning speed. Generally, the higher the scanning speed, the lower the tensile strength. The scanning speed determines the heating time of TPU powder taken from the CO₂ laser beam. At a fixed laser power, an enhancement of the scanning speed shortens the heating time on polymer powders, resulting in weaker bonding between powder particles therefore deteriorating the mechanical properties. When laser energy density is too low, the TPU powder cannot be sufficiently melted, resulting in a lower mechanical strength. The melt viscosity of TPU powder decreases, and the laser energy density increase, and thus the product shows increased mechanical strength. However, if the laser energy density is too high, TPU powder would be over melted and decomposed, leading to a decrease in the mechanical strength. Based on the results of the statistical analysis in figure 4, it can be found that there is a significant difference between the tensile strength of the specimnes manufactured under the scanning speed of 4000 mm s⁻¹ and 3500 mm s⁻¹ (P < 0.05). It is also proved that the incomplete sintering has a negative effect on the mechanical properties of materials which is caused by high scanning speed.

Zoom In Zoom Out Reset image size

Based on the results discussed above, the main SLS processing parameters for the TPU powder can be optimized as follows: the laser power of 25 W, scanning speed of 3500 mm s⁻¹ and layer thickness of 0.1 mm.

The stress–strain curves of the SLS specimens fabricated under different parameters are shown in figure 5. The mechanical properties of the TPU samples made by the different parameters are summarized in table 6. The SLS specimens exhibited the best tensile strength of 20.02 MPa and toughness of 26 631 J mm⁻³. The elongation of the specimens decreases, when scanning speed increases. And the specimens manufactured under the scanning speed of 3000 mm s⁻¹ showed a significant statistical difference compared with the specimens manufacutred at 3500 mm s⁻¹ in elongation. It shows that materials aging which is induced by oversintering can exert a negative effect on elongation.

Zoom In Zoom Out Reset image size

The results indicated that the internal voids in TPU specimens have a negative effect on the toughness and ductility, and the inevitable holes in SLS sample accelerate the initiation and failure of cracks. Nonetheless, the elongation of SLS components of TPU is still better than that of commercially SLS polymers including PA, PEEK and PS [46–48].

The DSC curves of TPU powder and SLS specimens made under different parameters are shown in figure 6. Table 7 shows glass transition temperature (T_g), melting enthalpy (ΔH) and melting temperature (T_m) of TPU powders and SLS specimens. The increasing glass transition temperature was caused by a homogeneous mixture due to the partially dissolving of hard segments into soft phase [49]. Intermolecular hydrogen bonding forms an impermeable hard-segment domain, which may enhance the mechanical properties of TPU. When the scanning speed was decreased from 3750 to 3500 mm s⁻¹, the mobility of TPU chains increases. This might increase the probability of forming hard segment domains. The integrated endothermic peak of 'hard segments domains' increased from 14.11 to 14.82 J g⁻¹ between 105 ℃ and 170 ℃. Once the scanning speed decreases, the glass transition temperature of TPU specimen shifted to higher temperature due to the decrease of mobility of TPU soft segments.

Zoom In Zoom Out Reset image size

The cross-section areas of TPU parts after tensile testing are shown in figure 7. The typical failure mode is cleavage cracks accompanied with some smooth dimples. Some dimples and flowing stripes are clearly shown in the figures. The intended morphology can be found as the typical characteristic due to over-sintering [50].

Zoom In Zoom Out Reset image size

When the scanning speed reaches up to 3500 mm s⁻¹ (figure 7c), the highest tensile strength was obtained. Figure 7 (a–b) show the fracture surfaces of samples with different scanning speeds (3000 and 3250 mm s⁻¹, respectively). High energy density owing to over-sintering will lead a negative impact on the mechanical properties. When scanning speed increases, enlarged dimples and unmelted particles may appear because of insufficient energy density. These unmelted cores make the material behave as a microcomposite, which showed a pinning effect. As a result, the substantial increment in elongation at the break can be improved [19].