Viscoelastic, Shape Memory, and Fracture Characteristics of 3D-Printed Photosensitive Epoxy-Based Resin Under the Effect of Hydrothermal Ageing

Abstract

Using 3D-printed (3DPd) polymers and their composites as shape memory materials in various smart engineering applications has raised the demand for such functionally graded sustainable materials. This study aims to investigate the viscoelastic, shape memory, and fracture toughness properties of the epoxy-based ultraviolet (UV)-curable resin. A UV-based DLP (Digital Light Processing) printer was employed for the 3D printing (3DPg) epoxy-based structures. The effect of the hydrothermal accelerated ageing on the various properties of the 3DPd components was examined. The viscoelastic performance in terms of glass transition temperature (T_g), storage modulus, and loss modulus was evaluated. The shape memory polymer (SMP) performance with respect to shape recovery and shape fixity (programming the shape) were calculated through dynamic mechanical thermal analysis (DMTA). DMTA is used to reveal the molecular mobility performance through three different regions, i.e., glass region, glass transition region, and rubbery region. The shape-changing region (within the glass transition region) between the T_g value from the loss modulus and the T_g value from the tan(δ) was analysed. The temperature memory behaviour was investigated for flat and circular 3DPd structures to achieve sequential deployment. The critical stress intensity factor values of the single-edge notch bending (SENB) specimens have been explored for different crack inclination angles to investigate mode I (opening) and mixed-mode I/III (opening and tearing) fracture toughness. This study can contribute to the development of highly complex shape memory 3DPd structures that can be reshaped several times with large deformation.

1. Introduction

The growing need for sustainable materials has led to an increased adoption of 3D printing techniques across various applications. These techniques offer valuable advantages, such as the ability to create complex designs, faster production on-site, high resolution, and significantly reduced waste. Additionally, they promote recyclability, making them more cost-effective compared to traditional polymer processing methods like moulding and extrusion. This shift has had a significant economic and social impact [1,2]. Furthermore, 4DP SMPs are recognised as a unique category of smart, sustainable materials. Significant knowledge is required to understand the temporary shape (shape fixity, also called shape programming) and its shape-changing region to enhance the utilisation of 3DPd structures [3,4,5]. Approaches based on temperature responsiveness as a thermal trigger for the SMPs have been extensively studied. SMPs can change their shape with the effect of thermal stimulus that, after being shaped into a “temporary, also known as programming” configuration, they can be returned into a “permanent” shape through a thermomechanical process. These smart polymers can return to their permanent shape when heated above the glass transition temperature [6,7,8]. Various mechanisms may be utilised to obtain deployment responses like temperature-responsive (thermal stimuli) SMPs. Most SMP materials have two stages: permanent shape recovery after a temporary shape through a thermomechanical procedure (programming) [6,9]. In addition, functionally graded SMP materials have attracted researchers’, manufacturers’, and supply chains’ attention as durable materials whose structure has the capability to switch to multiple shapes, thus enabling the optimisation of the characteristics for the required applications [10,11].

The efficiency of such materials in performing their function is the capability to withstand different loading conditions during their lifetime. One of these loading cases is the propagation of an existing crack. The fracture toughness of a material containing a crack is represented by the critical stress intensity factor (K_IC). It is the critical value of stress intensity at a crack tip essential for understanding the conditions that lead to catastrophic failure under simple uniaxial loading [12,13]. Mixed-mode fracture, which includes both mixed-mode I/II and mixed-mode I/III scenarios, is the most common type of fracture associated with cracked components. Research has focused on crack tilting, which involves opening and sliding for mixed-mode I/II [14,15,16] and crack twisting, characterised by opening and tearing for mixed-mode I/III [17,18,19,20]. However, compared to the testing available for mixed-mode I/II, there are fewer qualified loading configurations for mixed-mode I/III testing. Pan et al. [17] investigated the fracture parameters to show that the pristine mode I fracture, and mixed-mode (I/II, I/III and I/II/III) fractures can be realised by varying the direction of the crack angles. Also, the study showed that the critical stress intensity factor decreased when the angle between the crack and loading direction changed from mode I (crack angle parallel to the load) to mixed mode I/III. Avci et al. [21] found that the mixed-mode I/III fracture toughness behaviour of polyester resin filled with sand particles and chopped glass fibre was significantly affected by varying crack angles under the SENB test.

Polymeric materials can be sensitive to long-term exposure to moisture and heat. For example, the mechanical and thermal behaviour of polymer materials can be significantly affected by moisture absorption, particularly at high temperatures. This means that water can progressively penetrate the polymer matrix [22,23]. Despite that, some thermoplastic materials like PA6 could promote crystallisation with moisture diffusing during hydrothermal treatment into the amorphous region due to the reorganisation of the hydrogen bond among amide groups in the amorphous region [24]. In contrast, the exposure of thermosets like epoxy to moisture and heat for long periods may lead to damage to their structural integrity due to the degradation of chain crosslinking and chain scission. This environment can deteriorate the matrix of thermoset-based materials by affecting their chemical compound. Polymer chain crosslinking can make the polymer brittle, potentially resulting in microcracks on the material’s surface. Meanwhile, chain scission can reduce the molecular weight and weaken the thermal and mechanical properties of the polymer matrix [23,25,26].

According to the literature survey, several researchers have investigated the effects of hydrothermal ageing on the shape memory properties of the 3DPd structures. To the best of the author’s knowledge, no study has been found to report the effect of humidity and temperature on the mixed mode fracture toughness and shape memory properties of the 3DPd UV-curable resin. In this study, a DLP resin was employed for 3DPg epoxy-based components. The viscoelastic, shape memory, and fracture characteristics of the epoxy-based UV-curable resin under the effects of short and long hydrothermal accelerated ageing were investigated. The curing behaviour was explored by differential scanning calorimetry (DSC). The viscoelastic properties and shape memory (SM) behaviour were evaluated via DMTA. This investigation study can increase the 4DPg technology of thermosets-based 3DPd structures in various engineering applications.

2. Materials and Methods

Rigid 140C Black resin (referred to as 3DP-Ep) (3D Systems, Inc., Rock Hill, SC, USA) was employed as a photopolymer resin (UV light-activated resin). This resin consists of some chemical materials (

Table 1. Chemical compositions of the 3DP-Ep resin.

Chemical Name	Weight Ratio (%)
2-Propen-1-one, 1-(4-morpholinyl) (Part A)	20–30
Epoxy Resin (Part A)
Oxirane, 2,2’,2’‘-[ethylidynetris(4,1- phenyleneoxymethylene)]tris- (Part A)	20–30
Phenyl bis(2,4,6-trimethylbenzoyl)-phosphine oxide (Part A)	25–35
Isobornylacrylate (Part A)	1–10
1-Vinylimidazol (Part B, Used 1:19 as the mixing ratio with part A.	<5

). The 3DP-Ep printed sample had a melting temperature of about 285 °C. 3DP-Ep (A two-part epoxy/acrylate hybrid material) is a heat-resistant material combining high elongation and high strength for direct plastic production, tool-less, which can be used for various applications like under-the-hood and internal cabin automotive covers, clips, connectors, fasteners and housings, board connectors, and electrical latching [27].

The resin in part A bottle was shaken and mixed for 1 h for first-time use on the 3D Systems LC-3D mixer, which was also mixed for 10 min before the subsequent use. A 19:1 mix ratio of part A to part B was shaken vigorously in a mixing container for 5 min. The resulting 3DP-Ep blend resin was stirred in the printer’s tray for 30 s between every printing process. Figure 4^® 3D printer (3D Systems) was utilised to manufacture the 3DP-Ep samples with a UV light wavelength of 405 nm and a layer thickness set to 0.1 mm. The support structure of the printed samples was oriented at 90° with the platform to provide a minimum structure with good resolution and surface finish. The 3DPd objects were washed with isopropyl alcohol (IPA) for 2 min and then, after drying, UV-post-cured for 90 min as suggested by the manufacturer. A three-hour thermal post-cure at 135 °C was also used for the full-curing process, as preferred by the manufacturer. A Memmert water path machine was used to immerse the 3DPd samples in distilled water coupled with 45 °C for 0, 60, and 120 days as hydrothermal accelerated ageing.

2.1. DSC, Viscoelastic and Water Uptake Test Procedure

In the DSC measurements, 7 mg samples was heated from −10 to 230 °C at a rate of 5 °C/min, followed by cooling back to −10 °C at the same rate. The DSC measurements were accomplished under the nitrogen atmosphere. DMTA was conducted using a TA Instruments machine Q800 (New Castle, DE, USA). In this test, a single cantilever mode was utilised with an oscillation frequency of 1 Hz and a strain of 0.02%. The temperature was increased from 25 °C to 210 °C at a heating rate of 3 °C/min. The free length, width, and thickness of the specimen for the DMTA test were 12 mm × 17.5 mm × 3.2 mm, respectively. T_g was determined from the tan(δ) and loss modulus peaks.

The water uptake was calculated according to ASTM D570 from the change in the weight of the 3DPd specimens (C) as follows:

C = (W_t − W_o)/W_o × 100%

(1)

where W_t is the weight of the specimen after a specific immersion time of up to 408 h (17 days) at 45 °C, and W_o is the weight of the sample before ageing and after drying in an oven for three h at 50 °C. The specimens were weighed to the nearest 0.001 g before and after a fixed interval of hydrothermal ageing. The rectangular 3DPd specimens with dimensions of 100 mm × 17 mm × 2 mm were immersed in distilled water using a Memmert water path machine at 45 °C. The average values were obtained from the four rectangular specimens.

2.2. SMP Test Procedure

Two procedures were used to understand the SMP behaviour of the 3DPd epoxy-based objects. The first SMP procedure was the thermomechanical SMP cycle test (Figure 1), which was used to investigate the shape fixity and shape recovery using the DMTA single cantilever mode. This technique is usually used to analyse the fixity/recovery behaviour of SMPs [28,29]. T_g was considered a reference temperature when designing the thermomechanical procedure. This procedure is performed as follows: (1) heating the sample above T_g with a rate of 5 °C min⁻¹, then isothermal for 25 min; (2) deforming the sample with a ramped force at a rate of 0.5 N min⁻¹; (3) cooling down the sample to 25 °C withholding the force constant; (4) releasing the force; (5) heating back the sample above T_g. Then, we repeated the process from step (2).

The shape recovery ($R_r$) and shape fixity ($R_f$) values were determined by the following equations:

$$R_f\left(\%\right)={\displaystyle \frac{{\epsilon }_2}{{\epsilon }_1}}\times 100\%$$

(2)

$$R_r\left(\%\right)={\displaystyle \frac{{\epsilon }_2-{\epsilon }_3}{{\epsilon }_1}}\times 100\%$$

(3)

where ε₁ is the strain induced by an external load above T_g. ε₂ is the strain after releasing the external load at room temperature. ε₃ is the strain when heating above T_g again.

The second SMP procedure was conducted to explore the sequential deployment of the flat and circular 3DPd samples [30]. Figure 2a,b represent a schematic diagram and photographs for the programming and recovery process of the 3DPd samples. The size of the 3DPd flat sample was 50 mm × 7 mm× 1 mm, and it was the same weight as the circular sample when it was programmed under the shape fixity process. The SM performance for the flat and circular samples included two processes: programming shape (temporary) occurred during the fixity shape due to heating the sample in sunflower oil, with a temperature around T_g (160 °C), then it cooled down to the room temperature (20 °C). The flat and circular samples were put in a silicon mould, and a flat metal plate was put on the sample to maintain their programming shape. The second process, i.e., the recovery shape (permanent), was achieved by heating the sample again in a sunflower oil from 0 °C to 160 °C. The test was repeated 5 times. A high-resolution camera was positioned in front of the sample to capture the recovery process. A thermometer was used to monitor the temperature near the sample inside a glass beaker. The height (h) (see Figure 2a) was measured using digital imaging analysis software (ImageJ, version 1.54d).

This procedure was utilised to calculate the shape fixity ($R_f$) and shape recovery ($R_r$) values using the following equations:

$$R_f\left(\%\right)={\displaystyle \frac{h_o-h_u}{h_o-h_t}}\times 100\%$$

(4)

$$R_r\left(\%\right)={\displaystyle \frac{h-h_t}{h_o-h_t}}\times 100\%$$

(5)

where h_t is the temporary height, h_o is the as-printed height, h_u is the height after load release, and h is the actual height, as shown in Figure 2a.

2.3. Mixed Mode I/III Fracture Test Procedure

Mode I (opening) fracture is the most common fracture in bodies with cracks. Crack propagation can occur when the stress intensity (strain energy release rate) at the crack tip reaches a critical value. The stress intensity factor values of the SENB specimens resulting from conducting the three-point bending (TPB) test in accordance with the linear elastic fracture mechanics (LEFM) theory were used by several studies in the literature [9,21,31]. Figure 3 represents a schematic diagram of the geometrical configuration of the SENB specimen. The mode I critical stress intensity factor can be determined via the initial notch depth technique by the equations below:

$$K_A=\left({\displaystyle \frac{P}{{BW}^{\frac{1}{2}}}}\right)f\left({\displaystyle \frac{a}{W}}\right)$$

(6)

$$f\left({\displaystyle \frac{a}{W}}\right)={\displaystyle \frac{3s{A}^{1/2}}{2W}}{\displaystyle \frac{1.99-A\left(1-A\right)\left(2.15-3.93A+2.7A^2\right)}{\left(1+2A\right){\left(1-A\right)}^{\frac{3}{2}}}}$$

(7)

where p is the fracture load for crack propagation, B is the sample width, and s is the span length. $f\left({\displaystyle \frac{a}{W}}\right)$ is the correction parameter.

For an inclined notch of the SENB test, several researchers determined the critical stress intensity factor for mixed-mode I/III [13,21] and specified that the apparent stress intensity factor (K_A) can be used to calculate the critical stress intensity factors for mode I (K_Ic) and mode III (K_IIIc) by substitution K_A into Equations (8) and (9) below:

$$K_{IC}=K_A{sin}^2\theta$$

(8)

$$K_{IIIC}=K_{A}sin\theta cos\theta$$

(9)

The K_Ic and K_IIIc were calculated from Equations (6), (8), and (9), as follows:

$$K_{IC}=\left({\displaystyle \frac{P}{{BW}^{\frac{1}{2}}}}\right)f\left({\displaystyle \frac{a}{W}}\right){sin}^2\theta$$

(10)

$$K_{IIIC}=\left({\displaystyle \frac{P}{{BW}^{\frac{1}{2}}}}\right)f\left({\displaystyle \frac{a}{W}}\right)sin\theta cos\theta$$

(11)

3. Results

3.1. DSC and Water Uptake Behaviour

The DSC measurement is a reliable tool for measuring and visualising the endotherm reaction of epoxy thermosets. The DSC test was conducted on the thermal post-cured sample. Figure 4a shows the DSC second heating and second cooling curves for the 3DPd epoxy-based sample before thermal post-curing. The DSC curve shows the amorphous nature and crosslinking behaviour of the 3DPd samples. The glass transition is one of the most important properties of amorphous and semi-crystalline materials. The glass transition temperature of the fully crosslinked sample is determined based on the second heating because it is shifted to a higher temperature in the second heating [32]. The same behaviour was noticed in the literature [33]. The heat flow of the 3DPd sample shows an exotherm peak in the curve at approximately 140 °C, which refers to the glass transition temperature.

Figure 4b displays the change in humidity content for the samples as a function of the square root of the hydrothermal ageing time. In this figure, the curve can be divided into two steps. At the beginning of step 1, the experimental values of the humidity content for all the studied samples increased proportionally to the square root of ageing time. The sharply rising slope indicated quick humidity absorption processes in the samples. The same behaviour was seen in the literature on epoxy-based composites [23,34,35,36,37]. Water absorption after 24 h was 1.47%, calculated in accordance with ASTM D570. According to the literature [38,39], epoxy matrices alone, for instance, can absorb up to 5 wt% of their weight in water, which affects their mechanical performance. After achieving certain humidity contents in stage 1, the moisture uptakes of the samples slowed down until reaching the humidity quasi-equilibrium plateau with a maximum water uptake of approximately 7.5 %.

3.2. Viscoelastic Behaviour

DMTA was used to determine the possible changes in the viscoelastic behaviour by exposing the 3DP-Ep components to a broad range of temperature values under an oscillating load. The damping factor (tan(δ)), storage modulus (E’), and loss modulus (E”) behaviour of the 3DPd epoxy-based samples before and after short and long hydrothermal ageing are presented in Figure 5a, Figure 5b, and Figure 5c, respectively.

Table 2. T_g, E’, and E” properties of the 3DP-Ep before and after the hydrothermal ageing.

Property	Ep0h-Pre	Ep0h	Ep600h	* Dec. (%)	Ep1800h	Dec. (%)
Tg (°C) (tan (δ))	105.3 ± 3.5	166.7 ± 4.9	159.1 ± 5. 3	5	142.5 ± 4.2	15
Tg (°C) (E”)	76.8 ± 2.6	128.9 ± 3.7	117.5 ± 2.4	9	96.2 ± 3.8	25
E’ (MPa), T = 25 °C	2184 ± 30	2159 ± 27	1679 ± 46	22	1368 ± 32	37
E’’ (MPa), T = 25 °C	55.8 ± 2.3	53.6 ± 2.9	52.3 ± 1.8	2	48.9 ± 2.1	9

* Dec. = decrement.

shows the T_g, E’, and E” properties of the 3DP-Ep components before and after the hydrothermal ageing. EP0h-Pre represented the 3DPd sample before the thermal post-curing for three hours at 135 °C. The EP0h-Pre 3DPd sample undergoes the glass transition state (rubbery) early at about 70 °C. This is typically seen where polymer entanglement or crosslinking occurs due to molecular chain mobility, and the chains can slide past each other [30,40]. The T_g was noticed at about 166 °C and 128 °C from the tan(δ) and E’, respectively, while it was seen at 140 °C from the DSC curve (Section 3.1). The value of the T_g from the loss modulus was noticed to be close to the value from the supplier datasheet (124 °C) [27].

The temperature and humidity of the long-term ageing in distilled water significantly deteriorated the viscoelastic performance in terms of storage modulus and T_g. It was observed that after the hydrothermal ageing in the distilled water at 45 °C, the curves of T_g, E’, and E” were shifted to the left, and all the results were reduced after the 600 h and 1800 h hydrothermal degradation. As seen from Table 2, the reduction in the T_g from (tan (δ)) curves are 5 % and 15 %, i.e., after 600 h and 1800 h of the hydrothermal ageing, respectively. On the other hand, the E’ was reduced by about 22 % and 37 %, respectively. Therefore, the deformation can occur earlier at high temperatures due to the reduction in the T_g after long-term ageing, which led to higher molecular chain mobility for the hydrothermally aged samples. This behaviour, for example, can play a vital role in the behaviour of the SMP in that the materials enter the rubbery region from the glass transition region early.

3.3. SMP Behaviour

The SMP performance of the 3DPd epoxy-based objects undertaking repeated 4-step DMTA thermomechanical cycles is demonstrated in Figure 6.

Table 3. Strain, recovery, and shape fixity obtained from the 4-step DMA thermomechanical cyclic test.

Property (%)	Ep-135 °C	Me-100 °C	Ep0h	Ep600h	Ep1800h
Rf	89.2 ± 1.1	97.3 ± 1.3	98.6 ± 0.6	97.4 ± 1.4	96.1 ± 1.5
Rr	84.6 ± 1.0	90.4 ± 0.8	92.5 ± 0.5	91.2 ± 1.0	93.8 ± 0.7

displays the shape fixity and shape recovery values calculated by Equations (2) and (3) in Section 2.3. Figure 6a represents the 4-step thermomechanical SMP cycle for the 3DP-Ep at the temperatures of (135 °C and 170 °C) and the methacrylate-based resin (Tough-Blk-20 resin, 3D Systems, referred to as Me-100 °C). The SMP cycle of the 3DPd Ep-135 °C and Ep-170 °C samples was compared with the Me-100 °C to show the shape memory behaviour of 3D-printed polymers that have various glass transition temperatures. The 3DP-Ep samples were heated to 135 °C and 170 °C with the applied loads of 10 N and 2.5 N, respectively, (Figure 6). Hence, molecular chain mobility increased, and shape changing became easier in the end glass transition region and rubbery region. Thus, the strain result for the Ep-135 °C was less than Ep-170 °C (Ep0h). This indicated that the molecular mobility level was low and still under the glass transition region for the Ep-135 °C, while the Ep-170 °C showed a high molecular mobility level under the rubbery region. Therefore, as shown in Table 3, the shape fixity and shape recovery values for the Ep-135 °C are less than 90%.

On the other hand, the 4-step thermomechanical cycle of the Ep-170 °C showed similar behaviour to the Me-100 °C (T_g was 83.2 °C, from tan (δ) results), both of which were heated above T_g within the rubbery region of the 3DPd material.

Figure 6b shows the effect of the hydrothermal degradation on the 4-step DMTA thermomechanical cycles after (0 h, 600 h, and 1800 h) hydrothermal accelerated ageing in distilled water. Compared to the unaged sample, the strain values were reduced after ageing by about 10% and 29% for the Ep600h and Ep800h, respectively. However, the deformation occurred earlier at high temperatures due to the reduction in the T_g after 600 h and 1800 h hydrothermal ageing, as described in the Viscoelastic Behaviour Section that the difference increased between T_g and heating up to (170) leads to higher molecular chain mobility for the hydrothermal aged samples. In spite of the long-term temperature and humidity ageing, the shape fixity and shape recovery values were still higher than 90%.

The SMP behaviour of the sequential deployment and shape changing for the flat and circular 3DPd samples is due to thermal heating to assess the shape programming (shape fixing) and shape recovery behaviour. The 3DP-Ep sample is thermally heated in sunflower oil to a temperature close to the T_g (160 °C). The average shape fixity and shape recovery values were calculated by Equations (4) and (5) in Section 2.3. The average shape fixity and shape recovery values were 98% and 92% for the flat sample, respectively; these average values were 96% and 88% for the circular 3DPd sample, respectively. Therefore, this 3DP-Ep material can be used for various SMP applications like actuators and thermal sensors. Figure 7 shows the programmed (temporary) and recovery shapes by heating above T_g and cooling the 3DPd epoxy-based structures to room temperature.

3.4. Mixed Mode I/III Fracture Toughness Behaviour

The fracture toughness of the 3DP-Ep beams with different crack inclination angles was investigated for the mode I (opening) and mixed-mode I/III (opening and tearing) loading conditions.

Table 4. Fracture loads of the SENB specimens under the TPB test.

Ageing (h)	Notch Angle, θ (degree)	Fracture Load (N)
0	90°	322.3 ± 6.1
	60°	340.9 ± 10.1
	30°	407.8 ± 8.9
600	90°	313.6 ± 6.1
	60°	321.9 ± 10.1
	30°	385.8 ± 8.9
1800	90°	217.6 ± 7.4
	60°	251.9 ± 10.8
	30°	286.3 ± 14.3

represents the fracture loads of the SENB specimens resulting from the TPB test. The fracture load of SENB specimens of the unaged samples reached its maximum value (407.8 N) at θ = 30° and gradually reduced when the crack inclination angle was increased to reach 322.3 N at θ = 90° with a fracture load reduction of about 21%. These values were reduced under 600 h and 1800 h of the hydrothermal accelerated ageing in the distilled water to 286.3 N and 217.6 N at angles 30° and 90°, respectively.

Figure 8 shows the mode I and mixed mode I/III critical stress intensity factors, K_Ic and K_IIIc, respectively, with various crack inclination angles of the epoxy-based 3DPd beams with the effect of short and long hydrothermal ageing intervals. This figure shows that when the crack inclination angle changed from 30° to 90°, the K_Ic increased to reach the highest values at θ = 90° as 1.18 MPa√m of unaged samples. In contrast, K_IIIc reached maximum values at θ = 30°. As expected, at θ = 90°, the K_IIIc values reached zero, which means no tearing loading when the crack is in the same loading direction. The K_Ic and K_IIIc values are expected to be equal at the crack inclination of 45° due to equal loading conditions.

The increment in the K_IIIc values was attributed to the increase in surface area at the notch section, owing to the various notch inclination angles, which increased the crack propagate resistance and, thus, the fracture toughness of the cracked material.

The K_Ic and K_IIIc values of the 3DPd beams were influenced by the immersion periods in distilled water. Hence, the K_Ic and K_IIIc values were reduced after 600 h and 1800 h hydrothermal ageing, respectively. The reduction in the K_Ic values was about 2%, 6%, and 4% for the crack angles 90°, 60°, and 30° after 600 h of ageing, while its reduction values after 1800 h of long ageing were 32%, 26%, and 29%. On the other hand, the reduction in the K_IIIc values for the 600 h ageing interval was about 3% and 7% for the crack angles 60° and 30°, while its reduction values after 1800 h of ageing were 24% and 30%. This behaviour was attributed to the degradation caused by the temperature and moisture of the hydrothermal ageing, which weakened the interfacial adhesion between the constituents and led to increasing the stress concentration areas [14,41,42]. Therefore, the long-ageing phenomenon generated remarkable degradation that significantly reduced the fracture toughness of the investigated epoxy-base 3DPd beams.

Figure 9 presents the broken surfaces of the failed SENB samples under the 3-point bending (3PB) test with various crack angles. Generally, the crack was propagated continuously along the preliminary crack when θ = 90°. On the other hand, for θ < 90°, the crack was almost propagated at the same angle as the preliminary crack. Several researchers in the literature noticed the same behaviour that the crack was propagated by the twisted path for θ < 90° [14,17,21]. The bright crack surface (Figure 9a) showed the brittle behaviour of the mode I fracture. When the crack inclination angle was less than 90°, crack propagation was turned from a nearly straight path of the crack (Figure 9a) to curved-path crack propagation (Figure 9b,c).

The SEM surface morphology fractographies of the SENB samples for mode I and mixed-mode I/III are presented in Figure 10. These images obviously show the direction of the crack propagation becoming more complex in the mixed mode fracture from the uneven surface that increased the fracture surface and thus, crack resistance.

4. Conclusions and Future Scope

This experimental study investigated the influence of hydrothermal accelerated ageing at 45 °C on the viscoelastic, shape memory, and mixed mode I/III fracture toughness of 3D-printed epoxy-based resin. A UV-based DLP 3D printer was used to manufacture the samples, followed by UV and thermal post-curing. The heat flow of the 3DPd sample from the DSC test showed the glass transition temperature at about 140 °C. The 3DP-Ep objects exhibited a maximum water uptake of 11.5 %. The temperature and humidity of the long-term ageing in distilled water significantly deteriorated the viscoelastic performance in terms of storage modulus and T_g. The reduction in the storage modulus and T_g (tan (δ)) were 11.7% and 23%, respectively, after 1800 h ageing.

The calculated shape fixity and shape recovery from the 4-step thermomechanical cycles were less than 90% for the samples heated lower than T_g from tan(δ). The shape-changing region between the T_g value from the loss modulus and the T_g value from the tan(δ) is considered a glass transition region due to about 80% to 99% of the shape being recovered in this region. The strain values were reduced by 29% after long-term ageing compared to the unaged sample. Nevertheless, the deformation occurred earlier at high temperatures due to the reduction in the T_g after long-term ageing, which led to higher molecular chain mobility for the hydrothermally aged samples. Despite the long-term temperature and humidity ageing, the shape fixity and shape recovery values were still higher than 90%. The sequential deployment and shape changing for the flat and circular 3DPd structures due to thermal trigger exhibited good shape programming (higher than 88%) and shape recovery behaviour (higher than 96%).

The fracture toughness of the 3DP-Ep beams with various crack inclination angles in mode I and mixed-mode I/III loading conditions was investigated using the SENB test. The SENB load values were reduced after the 600 h and 1800 h hydrothermal ageing. The long ageing phenomenon generated remarkable degradation that significantly reduced the fracture toughness values (K_Ic and K_IIIc) for the crack angles 90°, 60°, and 30° of the investigated epoxy-based 3DPd beams. After 1800 h of ageing, the maximum reduction in the K_Ic values was 32%, 26%, and 29%, while the decrement in the K_IIIc values was 24% and 30% due to the degradation caused by the temperature and moisture, which weakened the interfacial adhesion between the constituents and led to increased stress concentration areas.

This study provides concepts for developing durable functionally graded epoxy-based structures showing shape memory properties to expand the use of 4DPg technology in various applications such as soft robotics, actuators, thermal sensors, marines, and deployable structures. In 4DPg, in which ‘time’ serves as the additional dimension, a fundamental desire is the ability to have a controlled thermal response. Therefore, the SM behaviour, including free recovery at diverse programming temperatures, could be a subject of future study to offer a fundamental understanding of the influence of the programming temperatures on the thermal responses of the 3DPd SMPs. Also, the temperature–force–strain SM behaviour of the thermomechanical cycle can be included with respect to various heating and cooling rates as further examined. Finally, the effect of various accelerating weathering ageing conditions on SM behaviour can be considered in future work.