Keywords

1 Introduction

In a few decades, humans witnessed a leaping advancement in the development of technologies for the production of polymeric materials. Currently, the global production of polymeric material stands at 270 million tons per year. More than 55% of polymeric wastes are discarded and only 20% of polymeric wastes are recycled every year. Central Pollution Control Board (CPCB, Government of India) estimated that ~ 4059 tons of polymeric waste are generated in major cities of India per day, where 25,940 tons of polymer waste are generated per day in India only. The degradation of polymers through biological activities in terrestrial environments takes more than five centuries and in the marine environment takes two decades. Hence, polymers pose a high degree of pollution and hence the consequences on the environment [1, 2]. In most wood mills, sawdust is considered commercial waste. The commercial value of the sawdust waste could be increased by the possible creation of new products. It is estimated that around 4 million to 6 million tons of wood waste is generated per year, which is mostly from the furniture manufacturing, construction, and packing sectors [3, 4]. Responsible consumption and production would ensure the sustainable use of natural resources.

Also, the practice of recycling wastes would sustainable environment for us and our future generation [5]. Turning waste into composite material presents a major technical, economic, and ecological challenge and importance both internationally and nationally [6,7,8]. Fabrication of composite has been an interesting research area, as evidenced from the previous research work. Basalp et al. [9] investigated the influence of wood flour composition and processing temperature on the properties of wood-plastic composite. The decrease in the property of water absorption in the specimen indicates that there was an increase in interfacial adhesion between recycled polymer and wood flour. The study concluded that processing at 185 ℃ improves the tensile strength of the wood-plastic composite.

Hernandez et al. [10] studied the mechanical behavior of wood polymer composites that were made from polystyrene and white oak wood flour. The particle size, the ratio of the wood, and the wood content influence the mechanical and water absorption property in the specimen. Nami et al. [11] studied water absorption, fungal resistance in natural fiber (wood, bamboo), and polypropylene-based wood-plastic composite. The wood particle size (30 mesh to 100 mesh) and particle content affected the water absorption and fungal resistance of the composite specimen. Baptista et al. [12] studied the microstructure of thermoplastic composite made of sawdust, wool, and high-density polyethylene or polypropylene. A less-intensive bonding was observed between wood, wool, and high-density polyethylene or polypropylene, because of the absence of a coupling agent. Jumadi et al. [13] studied the microhardness of the composite, which was prepared with recycled material (polypropylene and polyethylene) and natural fiber (kenaf fiber). The polished specimen with 40 wt.% kenaf fiber had better visibility and a maximum microhardness of 9.2 HV.

Taghi et al. [14] investigated the fabrication of wood-plastic composite from completely recycled material (wood flour and granules of polypropylene). The results indicate that the addition of wood flour beyond 35 wt.% deteriorated the physical and mechanical properties. However, a low-water absorption percentage (0.5%) was observed in the wood-plastic composite with 60 wt.% polypropylene. Youssef et al. [15] investigated the structural and mechanical characteristics of the wood-plastic composite (high-density polyethylene waste and camphor wood waste). Maximum bending strength and Young’s modulus were observed in the wood-plastic composite that had 40% plastic content. Also, the composite exhibited better impact bending strength (30.75 MPa) than other specimens. Fabrice et al. [16] investigated mechanical and thermal properties of bio-based material to develop an insulating material using miscanthus. At 50 wt.% miscanthus and 50 wt.% binder, better compressive strength (3.5 MPa) was observed.

Beatriz et al. [17] studied the mechanical behavior and vibroacoustic behavior of rubber granules with rice husk. The composite with 75% rice husk content exhibited a maximum compressive strength of 5.951 MPa. An increase in density of stiffened the composite. Behzad et al. [18] studied the influence of immidazolium-based green solvents on the water absorption of wood-plastic composite. The wood flour was treated chemically with imidazolium-based ionic liquids and that in turn was combined with plastic (high-density polyethylene). It was observed that imidazolium-based ionic liquids increased the quality of adhesion between the polymer matrix and the cellulosic materials. Nourbakshi et al. [19] studied the properties of the wood-plastic composite made from high-density polyethylene, wood fiber (55, 70, 80 wt.%), and coupling agent (polyethylene-grafted maleic anhydride) (0, 2, 4 wt.%). The results indicated that addition of more wood fiber and coupling agent improved the mechanical and physical properties of the composite.

Omoniyi et al. [20] studied the influence of saw dust (Gmelina Arborea) and plastic (waste water bottles) content on the mechanical behavior of wood-plastic composite. An increase in wood fiber decreased the tensile strength of the composite. The mechanical characteristics of composites were significantly affected by an increase in the sawdust-plastic ratio. Faith et al. [21] investigated the mechanical properties of wood-plastic composite that was fabricated pine wood fiber and thermoplastic. It was observed that the high-density and polypropylene-based composite had a maximum tensile strength of 37.02 MPa. Polypropylene-based composite exhibited a maximum flexural strength of 35.82 MPa. Harish et al. [22] studied the influence of polymer waste (high-density polyethylene, low-density polyethylene) mixed with different natural filler materials (concrete, clay, natural residue, rice husk), and coupling agents (maleic anhydride polypropylene, natural oil, silane, and alkaline treatment) in the properties of wood-plastic composite. The wood-plastic composite with 70% rice husk and 30% high-density polyethylene had a high-tensile strength of 36 MPa. Qiang et al. [23] investigated the characteristics of wood-plastic composite made from polyethylene with maple wood fiber utilizing twin-screw extruder method. The inclusion of maleic anhydride-grafted polyolefin (0.9 wt.%) to the composite increased the amount of adhesion between the wood fiber and the polyolefin matrix. The fractography confirmed the fracture of the cellulose fiber and lesser voids that in turn indicated improved mechanical performance. Prakash et al. [24] studied the composite material (coconut fiber, sansevieria cylindrica fiber, artificial fibers) that would be suitable for tsunami-resistant design. The mechanical strength of composite was evaluated and which was reinforced with coconut fiber, sansevieria cylindrica fiber, glass fiber with resin (isophthalic polyester resin) was evaluated. The composite material made from coconut fiber exhibited an impact strength of 0.022 J/mm2 and moisture percentage comparable to that of glass fiber.

In this study, wood polymer composite was fabricated through powder metallurgy processing technique. The microstructure, tensile strength, compressive strength, water absorption, and thickness swelling were studied for various wood polymer composites.

2 Materials and Method

2.1 Materials

The high-density polyethylene (HDPE) bottles were collected from the domestic waste. The bottles were shredded using a mechanical shredder. The shredded HDPE bottles were termed waste polymer. The wood husk was collected from a local sawmill. The average particle size of wood husk was ~1500 μm. The wood husk was dried in shade at room temperature for 72 h to eliminate the moisture content.

2.2 Experimental Layout

The composite specimen was prepared by varying the composition of wood husk and waste polymer. The composition of the fabricated composite specimen is shown in Table 1.

Table 1 Composition of wood and polymer used for making composite

2.3 Composite Preparation

The wood husk and waste polymer were dried under direct sunlight for 24 h. The dried contents were weighed in the specified ratio, as given in Table 1, using a high-precision weighing balance with a readability of 0.0001 g. The weighed contents were homogeneously blended using a mechanical ball milling unit at 250 rpm for 6 h. The blended contents were molten in a graphite crucible using a resistance-heating furnace at a temperature range of 160 ℃ to 180 ℃. The agglomeration and segregation were avoided by an automatic stirring setup. The melt was poured into a rectangular mold of dimension 100 × 100 × 25 mm3. Before pouring, mineral oil was applied to the walls of the rectangular mold to lessen the adhesive effects. In the course of cooling, the specimens were compacted using a hydraulic compaction machine. The specimens were compacted such that the volume reduction of the composite at the end of compaction was 10%. Subsequently, the specimen was cured in the atmosphere to obtain the fabricated specimen.

2.4 Microhardness

The composite specimens were polished using silicon carbide paper to reduce surface uniformity. The Vicker’s microhardness of the composite specimen was measured as per the standard ASTM E 384 [25]. The specimen was indented with a diamond in an axial load of 10 g for a dwell time of 15 s. An average of five measurements was reported.

2.5 Water Absorption Test

The water absorption of the specimens was determined as per the standard ASTM D 5790 [26]. The mass of the specimens was measured using a precision weighing balance with the readability of 0.0001 g. The specimens were immersed in water at room temperature. The specimens were removed after 2 h and 24 h of immersion. The specimens were wiped to remove the water on the surface and weighed immediately. The mass gain of the specimen was measured, and water absorption (%) was calculated using the Eq. (1) [26].

$${\text{Water}}\;{\text{Absorption }}\left( {\text{\% }} \right) = \frac{Mc - Mo}{{Mo}} \times 100$$
(1)

where \(Mc\) is the mass of the specimen after immersion (g) and \(Mo\) is the mass of the specimen before immersion (g).

2.6 Thickness Swell Test

The swelling of the specimens after the immersion test was measured. The thickness of the specimen before and after the immersion test was measured using a digital vernier caliper with a least count of 0.01 mm. The swelling thickness was calculated using Eq. (2) [26].

$${\text{Thickness}}\;{\text{Swell }}\left( {\text{\% }} \right) = \frac{Tc - To}{{To}} \times 100$$
(2)

where \(Tc\) is the thickness of the specimen after immersion and \(To\) is the thickness of the specimen before immersion.

2.7 Tensile Test

The tensile test specimens were prepared as per standard ASTM D638 [27]. The specimens were tested for tensile strength and elongation at room temperature using a tensile testing machine (Make: Tinius Olsen) at a crosshead speed of 1 mm/min.

2.8 Compressive Test

The compression strength of the specimens was determined as per the standard ASTM D 3410 [28]. The compressive test of the specimens was carried under room temperature using a universal testing machine at the crosshead speed of 5.6 mm/min.

2.9 Fracture Morphology Analysis

The fracture morphology of the tensile-fractured specimen was observed under a field emission scanning electron microscope (Make: Zeiss Sigma). The fractography were captured at EHT of 5 kV and various magnifications.

3 Results and Discussion

3.1 Fabricated Specimen

The wood husk and waste polymer of specified proportionate composition were ball milled to a homogeneous mixture. The mixture was melted in a graphite crucible induction furnace and was poured into a rectangular mold. Subsequently, the semi-solid melt was compacted using a hydraulic press. The properties of the composite specimens are described below.

3.2 Microhardness Test

Figure 1 shows the Vicker’s microhardness of the fabricated composite specimen. The polished specimens had a visible indentation mark, which enabled easy and effective determination of microhardness. The microhardness of specimen A (9.1 HV) was higher than specimen B (6.4 HV). In turn, the microhardness of specimen C (4.3 HV) was lesser than that of specimen A and specimen B. A linear decreasing trend was observed in hardness with a decrease in waste polymer content of the composite. The results explicited the significant influence of waste polymer content on the microhardness of the composite specimen.

Fig. 1
A bar graph shows the microhardness of the composite. The bars plotted along with their values are, specimen A at 9 H V, specimen B at 6.5 H V, specimen C at 4.5 H V. Values are estimated.

Microhardness of the composite

3.3 Water Absorption Test

Figure 2 shows the water absorption of the specimens based on various content of wood husk content after 2 h and 24 h of immersion in water. Specimen A had 10% of water absorption after 24 h of immersion test. Specimen B had 13.9% of water absorption after 24 h of immersion, which is 3.9% higher than specimen A. Specimen C with an equal proportion of wood husk and waste polymer had 16.3% of water absorption after 24 h of immersion. The water absorption of specimen C was 6.3% higher than specimen A. The water absorption of the specimens increased with an increase in immersion time. Also, it was observed that the water absorption of the specimens increased with an increase in wood husk content. The reason could be attributed to the natural hydrophilic phenomenon of wood husk with water. Wood in turn is a hydrophilic porous composite that consists of cellulose, lignin, and hemicellulose polymers. These polymers are rich in functional groups such as hydroxyls, which readily interact with water molecules by hydrogen bonding. Hence, the specimens with more wood husk content have the potentiality to absorb water. However, the reason for low-water absorption in the specimens with high-waste polymer could be attributed to the hydrophobic character of HDPE. Also, poor joint interface bonding between waste polymer and wood husk enabled the formation of voids. The voids acted as a medium for stored the absorbed water. Therefore, the water absorption increases with an increase in immersion time from 2 to 24 h, as well as with the proportionate increase in wood content increase in water absorption.

Fig. 2
A double bar graph shows the water absorption of specimens A, B, and C. The first bar shows after two hours and the second one after twenty-four hours. Y-axis shows water absorption in percentages up to eighteen. Maximum water absorption is done by specimen C.

Water absorption of the specimen

3.4 Thickness Swell Test

Figure 3 shows the thickness swelling of the specimens based on various content of wood husk content after 2 h and 24 h of immersion in water. Specimen A had 4.9% of thickness swelling after 24 h of immersion test. Specimen B had 5.2% of thickness swelling after 24 h of immersion, which is 0.3% higher than specimen A. Specimen C with an equal proportion of wood husk and waste polymer had 5.8% of thickness swelling after 24 h of immersion. The thickness swelling of specimen C was 0.9% higher than specimen A. The thickness swelling of the specimens had a similar tendency as that of water absorption. The results indicated that the specimens with more fractions of wood husk had higher thickness swelling. Also, the results illustrated that the specimens with more fractions of waste polymer had lower thickness swelling. The primary reason for thickness swelling lies with the hydrophilic nature of wood husk and the hydrophobic nature of waste polymer (HDPE). The wood has a critical surface energy between 40 mJ/m2 and 60 mJ/mm2 which is higher than waste polymer. The substantial difference in surface energy between wood and waste polymer might make waste polymer to be water repellent or hydrophobic (16).

Fig. 3
A double bar graph represents the thickness swelling of specimens A, B, and C. The first bar represents strength after two hours and the second one after twenty-four hours. The vertical axis depicts thickness swelling in percentages up to seven. Maximum thickness swelling is by specimen C at 3.9 and 6 for each bar. Values are estimated.

Thickness swelling of the specimen

3.5 Tensile Strength

The tensile test results are as shown in Fig. 4. Specimen A had a peak load of 350.52 N and tensile strength of 7.0104 MPa. Specimen B had a peak load of 252.5 N and tensile strength of 5.05 MPa, which was lower than specimen A. The tensile strength of specimen C was 3.233 MPa, with a peak load of 161.68 N, which was lesser than specimen A. The tensile results indicate that the specimen with low-wood husk content had higher tensile strength than the specimen with high-wood husk content.

Fig. 4
A bar graph represents the tensile strength of W P C of specimens A, B, and C. The horizontal axis has values of Tensile Strength in Mega Pascals up to eight. The first bar represents strength at graph point 7, the second represents at graph point 5, and the third represents at graph point 3. Values are estimated.

Tensile strength of WPC

3.6 Fractography Analysis

The fractograph of the specimen A is shown in Fig. 5. The transition region from matrix to fiber (cusp) and cracks was observed in Fig. 5. Figure 5 shows the pores in matrix. The pore formation could be attributed to chemical interaction of the waste polymer with the external environment in the course of processing. Figure 6 shows the fractography of the specimen B. Fiber (wood husk) pulled out region and deformation zone are extensively observed in specimen B. This was attributed to the poor adherence of incompatible wood husk and the waste polymer matrix. More pores and cracks were observed in the fractograph of the specimen C, as shown in Fig. 7. Also, the fractographs confirmed the presence of individual fibers in the matrix. Some wood husks could be seen partially embedded in the matrix, demonstrating poor adherence. The bonding is not complete since some debonding may be seen on the interface of the wood husk and the waste polymer matrix. The strength decreases because of the inadequate interfacial interaction between waste polymer and wood husk. That in turn hindered the effective stress transfer between the materials, despite the fact that natural fibers have a greater modulus than polymeric materials. Also, the binding capability of the waste polymer used appeared to be decreasing over time. Besides, the homogeneity of lignocellulosic materials disseminated in the polymeric matrix has a significant impact on the mechanical performance of wood polymer composite.

Fig. 5
Images show the fractograph of 40 percent W F and 60 percent H D P E of specimen A. Cusp, crack, and pore are marked in the images.

Fractograph of 40% WF and 60% HDPE (specimen A)

Fig. 6
Images show the fractograph of 45 percent W F and 55 percent H D P E of specimen B. Cusps, fiber pull-out region, and deformation zone is marked.

Fractograph of 45% WF and 55% HDPE (specimen B)

Fig. 7
Images show the fractograph of 50 percent W F and 50 percent H D P E of specimen C. Crack and pore are marked in the images.

Fractograph of 50% WF and 50% HDPE (specimen C)

3.7 Compressive Strength

Specimen A had a compressive strength of 11.57 MPa, whereas specimen B with 45 wt.% wood husk exhibited a compressive strength of 7.33 MPa. With a further increase in the wood husk to 50 wt.% in specimen C, compressive strength of 5.50 MPa was achieved. The decreasing trend of the compressive strength was similar to that of the decreasing trend of tensile strength of the composite, as shown in Fig. 8. With an increase in wood husk content, the compressive strength decreased. The decrease in compressive strength could also be attributed to the ineffective interfacial interaction between waste polymer and wood husk. That in turn hindered the efficient stress transfer between the materials. In macroscopic point of view, an increase of wood husk percentage might have increased the void fraction, causing a decrease in compressive strength of the specimen.

Fig. 8
A bar graph shows the compressive strength of W P C of specimens A, B, and C. The y-axis has values of Compressive Strength in Mega Pascals from 0 to 12 in increments of 2. The values are as follows; Specimen A, 11.8; Specimen B, 7.5; Specimen C, 5.6. All values are approximate.

Compressive strength of WPC

4 Conclusion

The technology for the fabrication of a composite material from wood husk and polymer was demonstrated. The prepared composite exhibited good physio-mechanical properties. The results demonstrated the following:

  • The wood husk and waste polymer synergistically influence the microstructure and physio-mechanical properties of the wood polymer composite.

  • The specimen A with of 50% wood husk and 50% waste polymer exhibited a maximum tensile strength of 7.01 MPa, compressive strength of 11.57 MPa, microhardness of 9.1 HV, and least water absorption.