Effects of a Novel Three-Dimensional-Printed Wood–Polylactic Acid Interlayer on the Mode II Delamination of Composites

Abstract

The interlayering method effectively enhances resistance against delamination in laminated composites. However, synthesis methods for interlayers have been limited and, at times, expensive. Consequently, this study investigates the effect of innovative 3D-printed wood–PLA interlayers on the mode II interlaminar fracture toughness (ILFT) of glass/epoxy composites. These interlayers feature a geometric structure comprising rhomboidal cell shapes, enabling the filament to maintain an equal volume percentage to the resin at the delamination interface. To this end, end-notch flexure (ENF) specimens were prepared, and the mode II ILFT was determined using the compliance-based beam method. The experimental results demonstrate a substantial increase in initiation load tolerance ($\cong 32$%) due to the 3D-printed interlayer. The R-curve analysis of the specimens with interlayers reveals significant enhancement in critical delamination parameters, including the length of the fracture process zone ($\cong 23\%$), initiation ILFT ($\cong 80\%$), and propagation ILFT ($\cong 44\%$), compared to the samples without interlayers. The fracture surface analysis of the reinforced specimens with interlayers demonstrated that the interlayer positively impacts the delamination resistance of the ENF specimens. They create a larger resin-rich area and increase surface friction at the delamination interface. Also, this facilitates a crack front pinning mechanism and changes the direction of crack growth.

1. Introduction

The weak, out-of-plane properties and stress discontinuities between layers in laminated composites render these materials highly susceptible to some damage, particularly delamination [1]. Delamination, which refers to the separation and detachment of layers in composite laminates, is regarded as one of the most critical forms of damage due to its potential to significantly reduce the load-bearing capacity of these materials with minimal visible signs [2,3,4,5,6,7]. The lack of visible indications during the delamination process, which are compounded by the widespread use of laminated composite materials in industries such as aerospace, shipbuilding, packaging, and the construction of storage tanks, has prompted researchers to intensify their focus on comprehending this damage, identifying influencing factors, and devising methods to control it [8,9,10,11]. In fact, given the operational sensitivity in these industries, failure to adequately control or comprehend delamination can lead to irreparable consequences.

Numerous investigators have conducted research indicating that various factors influence the interlaminar fracture toughness (ILFT) of polymer-based composite materials [12,13,14,15,16,17,18,19,20,21,22]. These factors include the stacking sequence of layers [12,13,14], the weave pattern of fibers [15,16], fiber volume fraction [17], and the mechanical properties of the matrix phase [18,20,22]. For instance, a study by Ozdil et al. [14] revealed that a [±45]₅ stacking sequence exhibits higher ILFT than that obtained for [±30]₅ or unidirectional layers in laminated composites. Ogasawara et al. [16] found that using a five-harness satin weave significantly improves mode I ILFT due to better interaction between the fibers and the matrix, as well as improved resin flow between the layers. Akhavan-Safar et al. [18] identified that incorporating micro-cork particles within the matrix materials of glass/epoxy laminated composites significantly improves the initiation and propagation of mode II ILFT.

Another crucial factor impacting the delamination resistance of composite laminates is the insufficient adhesion at the fiber–resin interface, which presents significant challenges in this field. Various methods have been suggested to enhance the adhesion between layers in composite laminates. These methods include treating the surface of the fibers [23,24,25], using woven fibers [26,27], utilizing fibers with a higher weave index (which is determined by the structure of the fibers’ unit cell) [28,29], and using an extra secondary layer between composite layers (known as the interlayering method) [1,30,31,32]. For instance, Qian et al. [33] demonstrated that oxygen plasma treatment increased mode II ILFT by 42% in glass/epoxy laminated composites. Also, in the study conducted by Blythe et al. [1], the researchers explored the impact of a polyamide nanofiber veil on the failure mechanisms of hybrid carbon/glass fiber composites. Their findings indicated that this configuration altered the dominant failure mechanism, leading to fiber yielding through localized kinking and a reduction in instances of buckling failure.

Out of all the methods mentioned, the interlayering method has emerged as one of the most effective approaches, attracting considerable attention from researchers [23]. This is because interlayering improves the adhesion between layers in composite laminates, maintaining in-plane properties and strengthening out-of-plane characteristics, thus enhancing the material’s resistance to delamination. However, Salamat-Talab and Kazemi [23] highlighted the importance of tailoring the interlayer design based on the type of damage experienced by laminated composites. They explained that if the delamination process involves an opening mode (mode I), the main focus of interlayer design should be to improve interfacial adhesion and activate mechanisms that resist delamination growth. Additionally, they suggested that if the damage mode involves shear (mode II), the interlayer should maintain these characteristics and increase surface friction at the fiber–resin interface to prevent layers from slipping against each other quickly. For instance, Jia et al. [34] reported that using a three-dimensional (3D) graphene network interlayer can increase mode I ILFT by 70% in glass/epoxy laminated composites due to their ability to facilitate proper resin flow between layers. Liu et al. [35] found that spraying multi-walled carbon nanotubes (MWCNTs) onto carbon fibers did not significantly change the mode I ILFT of carbon/epoxy laminated composites due to MWCNTs aggregation interfering with resin flow. On the other hand, this interlayer significantly improved mode II ILFT due to increased surface friction at the fiber–resin interface caused by MWCNTs. Liu et al. [36] investigated the effects of a hybrid polyether sulfone and carbon nanotubes (CNTs) interlayer on mode II ILFT. Their results showed that the best performance was achieved with an interlayer containing 1% by weight of CNTs, leading to a 317% increase in mode II ILFT. In fact, this interlayer not only enhanced interlaminar adhesion (due to the presence of the polymeric phase) but also improved surface friction, owing to the inclusion of CNTs in the interlayer structure. Kazemi et al. [30] showed that electrospinning polyvinyl alcohol nanofibers onto glass fibers increased initiation and propagation mode II ILFT of glass/epoxy composites by 177% and 36%, respectively. Also, they found that the polymeric interlayer used significantly improved the initial adhesion and the initiation value of mode II ILFT. However, because there was no substantial improvement in the fiber and resin interface friction, they did not observe a significant improvement in the propagation value of mode II ILFT. In addition to the methods mentioned above, new structures have been explored to create interlayers and improve the ILFT of laminated composites. Among these new approaches, Beylergil and Duman [37] incorporated a 3D-printed polyamide-based interlayer between the layers of carbon/epoxy composites. They reported 43% and 81% increases in mode I and mode II ILFT compared to samples without interlayers, respectively.

Based on the findings and existing literature review, it was discovered that the methods for manufacturing interlayers are generally limited to specific approaches. Therefore, in this study, we are examining the effect of a new 3D-printed interlayer, made from wood–polylactic acid (PLA) filament, which falls under the category of natural and organic materials, on the mode II ILFT in glass/epoxy laminated composites. It should be noted that to the best of our knowledge, there is no comprehensive study on the effect of a wood–PLA 3D-printed interlayer on the delamination behavior of laminated composites. The wood–PLA 3D-printed interlayer comprises rhomboidal cell shapes designed to approximately equalize the matrix-to-reinforcement volume fraction at the delamination interfaces. To assess mode II ILFT, end-notch flexure (ENF) specimens with 28 layers of glass fiber were prepared using the hand layup technique, with the interlayer placed between layers 14 and 15. The test results are analyzed based on the compliance-based beam method (CBBM). Additionally, a microscale mode II fracture surface analysis of the ENF specimens is conducted to provide a more detailed understanding of the active mechanisms during the delamination process.

2. Manufacturing and Experimental Setup

2.1. Three-Dimensional-Printed Interlayer

For the 3D printing of the desired interlayer, wood–PLA filament from eSUN Company with a 1.75 mm diameter was used. The 3D model of the interlayer with rhomboidal cell shapes was created in ANSYS Workbench 2021 Design Modeller and then converted into G-code using Ultimaker Cura 5.5.0 software. The interlayer thickness was designed to be 200 μm. Figure 1 provides a schematic of the 3D-printed interlayer preparation process and the geometric parameters of the rhomboidal cell shapes. The interlayer design aimed to ensure an approximately equal volume fraction of resin to wood–PLA filament at the delamination surfaces. It should be noted that the orientation of the rhombic cell shape is set at 45 degrees to the pre-crack front, as illustrated in Figure 1.

During the 3D printing process, the desired interlayer had a 100% infill density. Also, the 3D-printed interlayer’s dimensional accuracy in the x, y, and z directions was 15, 15, and 5 μm, respectively. Additional parameters of the 3D printer used in the sample printing process are provided in

Table 1. Three-dimensional printer parameters during printing of wood–PLA 3D-printed interlayer.

3D-Printer Parameter	Value	3D-Printer Parameter	Value
Extruder temperature (°C)	220	Build platform temperature (°C)	60
Nozzle diameter (mm)	0.4	Printing speed (${\displaystyle \raisebox{1ex}{$\mathrm{mm}$}\!\left/ \!\raisebox{-1ex}{$\mathrm{s}$}\right.}$)	40
Colling fan speed (%)	100

.

2.2. ENF Specimens

Upon the completion of the wood–PLA 3D-printed interlayer with rhomboidal cell shapes, the ENF composite samples were meticulously fabricated using LR620 epoxy resin, which was sourced from Iran Composite Kavian Co., Iran, in conjunction with woven glass fibers of 200 μm thickness and 200 g/m² aerial density. The ENF samples were created with 28 layers of glass fibers and a resin-to-hardener weight ratio of 5:1. Notably, the interlayers were positioned between layers 14 and 15 of each sample, and a 17-micron-thick Teflon tape was utilized for pre-cracking. Subsequent to the hand layup fabricating, the prepared samples underwent a 24 h curing process at room temperature and pressure. Also, to complete the post-curing process, the samples were heated for two hours at each temperature, 80 °C and 100 °C, according to the resin manufacturer’s instructions. To ensure result repeatability, three specimens were prepared for each of the two conditions investigated in this study, i.e., samples with and without a 3D-printed interlayer.

2.3. Mode II Interlaminar Fracture Toughness Test

After the curing process and the preparation of ENF specimens, the specimens were polished using fine-grit sandpaper. The ENF specimens prepared have dimensions of 6 mm thickness, 25 mm width, and 160 mm length. A Santam machine (STM 250) equipped with a highly accurate load cell with a capacity of 2000 kgf was used to conduct the mode II ILFT test at a loading rate and span length (S) of 0.5 mm/min and 100 mm, respectively. Figure 2 shows a schematic of the ENF sample and the mode II ILFT test setup.

Following ASTM D7905 [38], which is the standard method for determining the mode II ILFT of fiber-reinforced polymer matrix composites, and the compliance-based beam method, the mode II ILFT was calculated. It is important to note that the data reduction method specified in the ASTM D7905 standard is the compliance calibration method (CCM). This method requires visually recording the instantaneous crack length to calculate the mode II ILFT. However, other studies have indicated that it is possible to assess and calculate the mode II ILFT without relying on the instantaneous crack length by using the compliance-based beam method [39]. Thus, in this study, the CBBM approach has been utilized due to the presence of the 3D-printed interlayer, which complicates the visual registration of the crack length. In fact, this method allows for more effective analysis in situations in which direct observation of crack length is not feasible. According to this method, the mode II ILFT ($G_{IIc}^{CBBM}$) is calculated using Equation (1) [39].

$$G_{IIc}^{CBBM}={\displaystyle \frac{9P^2{a_{eq}}^2}{16B^2{t}^3{E}_f}}$$

(1)

In this context, the parameters P, B, and t correspond to the applied load, width, and half of the thickness of ENF specimens, respectively. Additionally, the parameters $a_{eq}$ and $E_f$ denote the equivalent crack length and the equivalent elastic modulus, respectively, which are calculated using Equations (2) and (3) [39].

$$a_{eq}={\left[{\displaystyle \frac{C_c}{C_{0c}}}{a_0}^3+{\displaystyle \frac{2}{3}}\left({\displaystyle \frac{C_c}{C_{0c}}}-1\right)\left({\displaystyle \raisebox{1ex}{$S^3$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}\right)\right]}^{{\displaystyle \raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{$3$}\right.}}$$

(2)

$$E_f={\displaystyle \frac{3{a_0}^3+2\left(\raisebox{1ex}{$S^3$}\!\left/ \!\raisebox{-1ex}{$2$}\right.\right)}{8B{t}^3}}\times {\left(C_{0c}\right)}^{-1}$$

(3)

In these equations, the parameter $a_0$ represents the initial crack length of the mode II ILFT test (30 mm). Additionally, the parameters $C_{0c}$ and $C_c$ correspond to the initial equivalent compliance and the equivalent compliance associated with the load-displacement curve, which are calculated using Equations (4) and (5) [39].

$$C_{0c}=C_0-{\displaystyle \frac{3\left(\raisebox{1ex}{$S$}\!\left/ \!\raisebox{-1ex}{$2$}\right.\right)}{10Bt{G}_{13}}}$$

(4)

$$C_c=C-{\displaystyle \frac{3\left(\raisebox{1ex}{$S$}\!\left/ \!\raisebox{-1ex}{$2$}\right.\right)}{10Bt{G}_{13}}}$$

(5)

where $C_0$, $C$, and $G_{13}$ represent the initial compliance, the compliance corresponding to the load-displacement curve for each specimen, and the out-of-plane shear modulus, respectively. It should be noted that in this method, as mentioned in the literature [39,40,41], the values of $G_{13}$ and $E_f$ have a negligible impact on the mode II ILFT of the specimens. Therefore, similar values from similar research literature can be utilized for $G_{13}$ and $E_f$ without the need for specific testing. It is important to note that in this study, the mode II ILFT was calculated using the parameters $G_{13}$ and $E_f$, with values of 4 and 19.5 GPa, respectively [42].

3. Result and Discussion

3.1. Mode II Interlaminar Fracture Toughness

The mode II ILFT test was performed on the ENF specimens, and the resulting load-displacement curves for ENF composite samples with 3D-printed wood–PLA interlayers (reinforced samples) and without interlayers (reference samples) are depicted in Figure 3. By analyzing this figure, we can conclude valuable information about the behavior of the ENF specimens when they undergo delamination. The comparison of load-displacement curves between the reference and reinforced samples reveals a noteworthy increase in load-bearing capacity at the point of crack initiation ($P^{ini.}$) in the reinforced samples, compared to the reference samples, as indicated by the black circles in Figure 3. This observation can be attributed to mechanisms that counteract crack propagation in the reinforced specimens. Furthermore, the results depicted in Figure 3 demonstrate a significant increase in the maximum load-bearing capacity ($P^{Max.}$) in the reinforced samples, suggesting that using the interlayer has substantially enhanced the mechanical strength.

Also, the research findings indicate that in the samples reinforced with the 3D-printed interlayer, there is a slight increase in displacement at the point of crack initiation and at the maximum load, compared to the reference samples. This phenomenon is likely associated with the presence of resin-rich areas in the cellular structure of the interlayer and the reinforcing wood–PLA filaments, which collectively contribute to the enhanced load-bearing capacity and displacement.

To provide a more comprehensive comparison of the load-bearing capacity between the ENF reference and reinforced samples, the data from the load-displacement results depicted in Figure 3 have been detailed in

Table 2. Comparison of load-bearing limit corresponding to the examined ENF specimens.

	${\mathit{P}}^{\mathit{i}\mathit{n}\mathit{i}.}$ (N)	${\mathit{P}}^{\mathit{M}\mathit{a}\mathit{x}.}$ (N)
Reference Sample	$217.9 \pm 1.3$	$477.8 \pm 30.5$
Reinforced Sample	$288.5 \pm 24.9$	$565.3 \pm 23.9$

. It is noteworthy that the $P^{ini.}$ for the reinforced samples exhibited a notable increase of 32.4% in comparison to the reference samples. It is predicted that this increasing trend may also be observed in the crack growth resistance curve (R-curve) for ENF samples, which will be further discussed. Furthermore, the comparison of the $P^{Max.}$ data in

Table 3. Comparison of $G_{IIc}^{ini.}$, $G_{IIc}^{prop.}$ and $L_{FPZ}$ values of each ENF specimen.

	${\mathit{G}}_{\mathit{I}\mathit{I}\mathit{c}}^{\mathit{i}\mathit{n}\mathit{i}.} (\mathbf{J}/{\mathbf{m}}^2)$	${\mathit{G}}_{\mathit{I}\mathit{I}\mathit{c}}^{\mathit{p}\mathit{r}\mathit{o}\mathit{p}.} (\mathbf{J}/{\mathbf{m}}^2)$	${\mathit{L}}_{\mathit{F}\mathit{P}\mathit{Z}} (\mathbf{J}/{\mathbf{m}}^2)$
Reference Sample	124.7 ± 7.9	845.7 ± 35.7	6.4 ± 0.3
Reinforces Sample	224.9 ± 32.3	1213.5 ± 50.2	7.9 ± 0.2
Gain in parameter, (%)	80.3	43.5	22.9

shows that this value for the reinforced samples also increased by 18.3% compared to the reference samples.

For a more comprehensive analysis of the impact of the 3D-printed wood–PLA filament interlayer on the delamination behavior of the ENF samples being studied, the crack growth resistance curve for both the reference and reinforced samples is depicted in Figure 4a,b, respectively. When comparing the two R-curves in Figure 4, significant changes in mode II initiation and propagation ILFT in the samples reinforced with the interlayer, as opposed to the reference samples, are evident. As outlined in Table 3, the mode II initiation ILFT for the reinforced samples increased from 124.7 J/m² to 224.8 J/m², indicating an 80.3% rise. This demonstrates that the use of the interlayer, along with the presence of resin-rich regions within the rhomboidal cell shape structures, significantly improved the initial adhesion and resin flow between the layers of the glass/epoxy laminated composites. Additionally, there was also an observable increase in mode II LFT in the reinforced samples, with a 43.5% rise compared to the reference samples. This improvement can be attributed to the effective presence of rhomboidal cell shapes and wood–PLA reinforcing filaments, which impede rapid crack growth and create mechanisms against crack propagation. It should be noted that additional extensive fractographic analyses of the fracture surfaces are necessary to explore these mechanisms further, which will be discussed in more detail.

In Figure 4, it is essential to note that the length of the fracture process zone ($L_{FPZ}$) in the reinforced samples is longer compared to the reference samples. It should be noted that the $L_{FPZ}$ is one of the parameters used to analyze the delamination behavior of laminated composites. It refers to the difference in crack length from the onset of crack growth to the point where the delamination growth becomes stable [43,44]. According to Table 3, the presence of the 3D-printed interlayer in the glass/epoxy laminate composites led to an increase in this parameter from 6.4 to 7.9 mm, representing a 22.9% increase. This indicates that when encountering the rhomboidal cell shapes in the interlayer, the crack front faces more counteracting mechanisms, slowing its rapid and sudden propagation and leading to more gradual growth in the crack front in the reinforced sample.

In order to thoroughly analyze the impact of incorporating the 3D-printed interlayer into the glass/epoxy laminated composites being studied, we have presented the changes in all main parameters during the delamination process, i.e., $P^{ini.}$, $G_{IIc}^{ini.}$, $G_{IIc}^{prop.}$, and $L_{FPZ}$, using a radar diagram in Figure 5. It is important to note that for better clarity, the data have been scaled to ensure that all values range between 0 and 100.

3.2. Fractography

In order to understand the mechanisms involved in interlaminar crack propagation, microscale imaging of the fracture surfaces of the ENF reference and reinforced samples was conducted using optical microscopy with an Insize Digital Measuring Microscope ISM-PM200SB (Figure 6). As shown in Figure 6a, interlaminar crack growth in the reference glass/epoxy composite resulted in a smooth fracture surface. In simpler terms, the crack propagated along the resin–fiber interface without significant deviation, leading to adhesive failure behavior. In other words, in the optical microscope images in Figure 6a, one side of the fracture surface is resin-free, while the other side contains visible resin. This can be attributed to factors such as insufficient bonding between the fibers and resin, as well as the absence of mechanisms to resist crack propagation [45,46].

On the other hand, upon analyzing the fracture surfaces of the samples reinforced with the 3D-printed wood–PLA interlayer, it was observed that the crack growth mechanism and fracture behavior in these samples differed significantly from those of the reference samples (Figure 6b). Specifically, the fracture surface of the reinforced samples exhibited crack deflection and a mixed-mode failure behavior involving both adhesive and cohesive failure. As a result, resin-free and resin-rich regions were visible on both sides of the fracture surfaces of the reinforced samples, indicated by yellow and red circles in Figure 6b, respectively. Another important observation from Figure 6b is the presence of a combined adhesive and cohesive failure mechanism at the start of crack growth. This indicates strong initial adhesion between the layers of the laminate composite reinforced with the interlayer. This is consistent with the increased initiation fracture toughness discussed above (highlighted in the pink rectangular, Figure 6b). Furthermore, as depicted in Figure 6b, rhomboidal cell shapes in reinforced ENF laminated composites are another effective way to prevent crack propagation and improve mode II ILFT. In fact, these cell shapes create a crack front pinning mechanism in the reinforced samples. In other words, the presence of rhomboidal cell shapes in the 3D-printed interlayer structure causes the crack front to become pinning during interlaminar crack growth. As a result, when more load is applied, the crack must either change direction or break through the 200-micrometer-thick rhomboidal-shaped cells to continue propagating (highlighted in the black rectangular, Figure 6b).

4. Conclusions

This study investigated the impact of a novel 3D-printed interlayer made from wood–PLA filament on mode II ILFT in glass/epoxy composites. It should be noted that the desired 3D-printed interlayer consists of rhomboidal-shaped cellular structures. To evaluate the mode II ILFT, the ENF specimens using 28 layers of glass fiber were prepared using the hand layup technique based on the ASTM D7905 standard. Additionally, a microscale fracture surface analysis was conducted using optical microscopy to understand better the mechanisms active during interlaminar crack propagation in the tested ENF samples. The key experimental results and a summary of the fracture surface analyses are as follows:

• The load-bearing capacity and displacement experienced in the samples with the interlayer significantly improved compared to the samples without the interlayer.
• Using the wood–PLA 3D-printed interlayer led to an 80.3% and 43.5% improvement in mode II initiation and propagation ILFT compared to the samples without an interlayer.
• The presence of the desired interlayer resulted in a 22.9% increase in the $L_{FPZ}$ and a more gradual crack growth in the samples with the interlayer.
• The application of the 3D-printed interlayer changed the fracture behavior from adhesive failure to a combination of adhesive and cohesive failure in samples with the interlayer, compared to those without.
• The resin-rich region was significantly enhanced by the presence of rhomboidal cell-shaped structures within the 3D-printed interlayer. This enhancement improves interfacial adhesion between the fibers and resin and increases surface friction at the delamination interfaces of samples with the interlayer.
• The presence of wood–PLA filament introduces mechanisms such as crack front pinning and crack path deflection during the delamination process in the reinforced samples.

Finally, the relevant literature and experimental results from this research indicate that utilizing 3D printing technology to create interlayers is an innovative method for enhancing ILFT in laminated composites. Therefore, continuing this line of inquiry by exploring various tunable parameters in the 3D-printed interlayer structures could yield a deeper understanding of the composite community. For instance, some key parameters for consideration include the shapes and dimensions of different cells within the 3D-printed interlayers and the volume fraction of the resin-rich areas at the delamination interface.