Abstract
This research investigated the use of reclaimed rubber (RR) from waste tires to partially replace the rubber compound (RC) when making wood fiber-rubber composites (WRCs). Ninety panels of WRC containing RR were manufactured with RR contents of 0% to 40%, mixing times of 6 min to 14 min, and vulcanizing temperatures of 150 °C to 170 °C. There were three steps, which were the fiber-rubber mixing, tabletting, and vulcanization molding processes. Four regression equations for the tensile strength, elongation at break, hardness, and rebound resilience as functions of the RR content, mixing time, and vulcanizing temperature were derived, and a nonlinear programming model was developed to obtain the optimum panel properties. It was found that when the RR content was within 20%, the wood fibers were well encapsulated and embedded in the RC/RR blends, and the processability of the WRCs were improved by adding RR. The incorporation of RR into the WRCs increased the average tensile strength and hardness by 33.9% and 2.3%, respectively, while the swelling ratio in toluene and 24-h water absorption were reduced by 13% and 42%, respectively.
Download PDF
Full Article
Fabrication of Wood Fiber-rubber Composites with Reclaimed Rubber
Dongwei Shao,a,b Min Xu,a,* Liping Cai,c and Sheldon Q. Shi c
This research investigated the use of reclaimed rubber (RR) from waste tires to partially replace the rubber compound (RC) when making wood fiber-rubber composites (WRCs). Ninety panels of WRC containing RR were manufactured with RR contents of 0% to 40%, mixing times of 6 min to 14 min, and vulcanizing temperatures of 150 °C to 170 °C. There were three steps, which were the fiber-rubber mixing, tabletting, and vulcanization molding processes. Four regression equations for the tensile strength, elongation at break, hardness, and rebound resilience as functions of the RR content, mixing time, and vulcanizing temperature were derived, and a nonlinear programming model was developed to obtain the optimum panel properties. It was found that when the RR content was within 20%, the wood fibers were well encapsulated and embedded in the RC/RR blends, and the processability of the WRCs were improved by adding RR. The incorporation of RR into the WRCs increased the average tensile strength and hardness by 33.9% and 2.3%, respectively, while the swelling ratio in toluene and 24-h water absorption were reduced by 13% and 42%, respectively.
Keywords: Wood rubber composites (WRCs); Reclaimed rubber (RR); Panel properties; Nonlinear programming model
Contact information: a: Key Laboratory of Bio-Based Material Science and Technology (Northeast Forestry University), Ministry of Education, 150040, Harbin, China; b: College of Mechanical Engineering, Jiamusi University, 154007, Jiamusi, China; c: Mechanical and Energy Engineering department, University of North Texas, Denton 76201, Texas, USA;
* Corresponding author: dl-xumin@nefu.edu.cn
INTRODUCTION
In recent years, the recycling and disposal of the enormous quantity of difficult to degrade waste rubber (e.g., waste tires) discarded each year has garnered a lot of attention because of environmental concerns (De et al. 2006; Ayrilmis et al. 2009a; Terzi et al. 2009; Hassan 2015). In the past two decades, the rubber industry has been facing a major challenge to find a viable way to recycle the enormous quantity of waste rubber (Adhikari et al. 2000; Hernandez-Olivares et al. 2002; Chen and Qian 2003; Rooj et al. 2011). Driven by green solutions for industry products and the life cycle of consumers using waste tires, an increasing amount of studies have focused on the development of wood-rubber composites (WRCs) during the past decade (Li et al. 2005; Zhao et al. 2008; Ayrilmis et al. 2009a; Zhao et al. 2010; Xu and Li 2012; Xu et al. 2014). Previous literature indicated that the water resistance, dimensional stability, sound insulation, and energy absorption of WRCs are better compared with neat wood composites.
Wood-rubber composites couple the advantages of both wood and rubber. Wood has a lower density and a better biodegradability compared with inorganic materials (Zhao et al. 2008, 2010). Rubber has a low moisture absorption, high compressive performance, and good sound insulation properties (Ayrilmis et al. 2009a; Yang et al. 2016). By taking advantage of both materials, WRCs could have multi-functional properties and excellent potential for extended applications. A number of previous studies have been conducted on the manufacturing of WRCs using wood, adhesives, and waste rubber powder with hot-pressing technology (Xu et al. 2014; Ding et al. 2015; Xia et al. 2015). Waste rubber powder is used as a filler and the adhesive is used as a bonding agent. This method has a high production cost and there are concerns about the toxicity and volatility (Shi and Wang 1997; Shi et al. 1999). Limited information is available on using rubber compounds (RCs) as a matrix.
In this study, the WRCs were fabricated using rubber processing technology, which includes fiber-rubber mixing, tabletting, and vulcanization molding processes. Instead of adhesives and waste rubber powder, a RC (i.e., un-vulcanized mixing rubber, mainly containing natural rubber (NR), butadiene rubber (BR), butadiene styrene rubber (SBR), and additives) was used to make the composite. As a common material in the tire rubber industry, tire RCs can be used in tough environments and climate conditions because of their ability to withstand high and low temperatures, and high anti-rotting properties (Zhao et al. 2011).
To reduce the composite cost, as well as to accentuate the recycling of waste rubber (such as discarded tires), reclaimed rubber (RR) was used to partially replace the RC as part of the rubber in WRCs. Lamminmäki et al. (2006) determined that RR presents much better properties than fillers when used as a part of the rubber.
With a plastic characteristic and viscosity, RR can be re-vulcanized (Adhikari et al. 2000). However, the drawbacks of RR, such as poor elongation at break (Eb) and rebound resilience (Rr) compared to the raw rubber, could affect the performance of rubber products (Fang et al. 2001; Zhang et al. 2009). Therefore, it is desirable to determine the optimal RR content in WRCs.
The objective of this research is to investigate the effect of the incorporation of RR into wood fiber (WF) and an RC matrix on the properties of WRCs. A further goal was to determine the optimal RR content and processing parameters.
EXPERIMENTAL
Materials
The WF used in the experiments had a length to width ratio of 5:45, moisture content of 3% to 5%, and was provided by Wooden Forest Products Co., Ltd. (Harbin, China). The RC was supplied by Xingda Rubber Factory (Harbin, China), and contained 30% NR, 6% SBR, 24% BR, 30% carbon black (N330), 1% sulfur, 2.5% zinc oxide, 2% stearic acid, 3% spindle oil, and 1.5% toluene (S.D.). The RR was made from rubber tires that were from light trucks and passenger cars and processed using a continuous shear flow reaction treatment (Xingda Rubber Factory, Harbin, China).
Sample preparation
The WF was initially oven-dried at 103 °C ± 2 °C for 24 h. After the fibers were cooled to room temperature, they were placed in sealed plastic bags until they were used to manufacture the WRCs containing RR (RR-WRCs). Sheets of RC and RR were cut into small pieces (approximately 5 mm3) so they could mix evenly.
The RR-WRCs were manufactured using rubber processing technology, which includes mixing in a twin rotor mixer, tabletting in an open mill, and sulfide formation in a plate vulcanizing machine. The RR-WRC panels had the dimensions 260 mm × 260 mm × 2 mm (length × width × thickness) and were fabricated at a target density of 1.0 g/cm3 with RR contents of 0%, 10%, 20%, 30%, and 40% (referred to as R0, R1, R2, R3, and R4, respectively), a mixing time that ranged from 6 min to 14 min, and a vulcanizing temperature that ranged from 150 °C to 170 °C. Six replicate panels were made for each type of composite, which resulted in a total of ninety panels. The control samples (R0) were fabricated using 70% RC and 30% WF.
The RC and RR were premixed at a mixing temperature of 60 °C for 3 min in a twin rotor mixer (XH-409, Zhuosheng Mechanical Equipment Co., Ltd., Dongguan, China). The main rotor speed was 25 rpm, and the speed ratio was 1.3. Then, the WF was gradually added into the mixer and blended for 3 min, 5 min, 7 min, 9 min, and 11 min.
The tabletting process of the mixture was performed using a laboratory two-roller mill (XH-401A, Zhuosheng Mechanical Equipment Co., Ltd., Dongguan, China) for 3 min. The two-roller mill consists of two parallel rollers with different rotation speeds. The speed ratio between the rollers was 1.2, and the gap was 2 mm. The mixing speed, gap between the two rollers, mill roller speed ratio, and sequence of adding the ingredients were kept constant for all of the sample preparations. The sheets from the two-roller mill were conditioned at a temperature of 23 °C ± 2 °C for 24 h in a closed container before vulcanization. Vulcanization was performed in a plate vulcanizing machine (XH-406B, Zhuosheng Mechanical Equipment Co., Ltd., Dongguan, China) at 150 °C, 155 °C, 160 °C, 165 °C, and 170 °C under a pressure of 15 MPa for an optimum curing time (t90), which was also called the vulcanization molding time.
Methods
Characterizations
Tensile specimens that are shown in Fig. 1 were 2-mm-thick, 116-mm-long, 25-mm-wide, and 6-mm-wide at the narrow portion. According to the ISO 37 (2011) standard, a universal testing machine (Instron 4505, Boston, MA, USA) was used determining the tensile strength (Ts) and Eb. The shore type-A durometer (JZ-LX-A, Jingzhuo Company, Yangzhou, China) was utilized for examining the hardness (Ha) of the composite panels according to the standard ISO 7619 (2010). According to the standard ISO 4662 (2009), the elastic impact tester (JZ-6022, Jingzhuo Company, Yangzhou, China) was used to determine Rr with a pendulum of 0.5 J potential energy.
Fig. 1. Morphology of RR-WRCs specimens of dumbbell-shaped
The fracture surface was sputter-coated with gold powder (BAL-TEC, Balzers, Liechtenstein) and then characterized by scanning electron microscopy (SEM; Quanta 200, FEI Company, Hillsboro, USA). The curing characteristics, including the minimum torque (ML), maximum torque (MH), scorch time (ts2), and t90, were measured with a no-rotor rheometer (JZ-6029, Jingzhuo Company, Yangzhou, China) according to the standard GB/T 16584 (1996). The authors used the methods of 72-h toluene swelling (De et al. 2013) and 24-h water immersion according to the Chinese standard GB/T 11718 (2009) to measure the swelling and water absorption of composites at room temperature. Cut from the vulcanized specimens and dried overnight in a vacuum desiccator, circular samples with a diameter of 10-mm and square samples with a size of 50 mm × 50 mm × 2 mm were prepared. After taking the weights (m0), these circular and square samples were placed into in toluene and water in bottles for 72 h and 24 h, respectively. After removing the liquids from the sample surfaces with filter paper, the weights (m1) of these samples were immediately measured. Using the following formula, the absorptions in both the toluene and water (Wa) liquids were obtained,
Wa (%) = (m1 – m0) × 100% / m0 (1)
where m0 is the weight of the sample before treatment (g) and m1 is the weight of the sample after treatment (g).
RESULTS AND DISCUSSION
Effect of the RR Content
The properties of the panels prepared with different RR contents at a target panel density of 1 g/cm3, mixing time of 8 min, and vulcanizing temperature of 160 °C are presented in Table 1. Figures 2 and 3 show that the relationships of the RR content and the Ts, Eb, Ha, and Rr were polynomial.
Fig. 2. Ts and Eb as functions of the RR content
Fig. 3. Ha and Rr as functions of the RR content
The addition of lower amounts of RR (< 20%) improved the mechanical properties of the composites. As the RR content increased from 0% to 20%, Ts and Ha of the composites increased. The increasing trend in Ts was similar to that obtained by De et al. (2013). At an RR content of 20%, the maximum Tsand Ha of the composites were obtained, and were 33.9% and 2.3% better compared with the control, respectively. It has been reported that RR has a good plasticity and could be used as a reinforcing agent (De et al. 2013). However, if too much RR was added (> 20%), the vulcanization of the rubber and the processing of the composites became difficult because of the lack of curing agents (sulfur) and the increase of inorganic impurity. Some residual cross-linked networks and inorganic impurity in the RR hindered the mobility of the unsaturated segments of the WRCs, which caused the properties of the composites to be reduced.
Table 1. Mechanical Properties of the RR-WRCs with Different RR Contents
1Groups with the same letters in each column indicate that there was no statistical difference (p < 0.05) between the samples according to the Duncan’s multiple range test; values in parentheses are the standard deviations
Effect of the Mixing Time
The properties of the panels prepared with different mixing times at a 160 °C vulcanizing temperature, target panel density of 1 g/cm3, and RR content of 20% are presented in Table 2. The polynomial relationships between the mixing time and the Ts, Eb, Ha, and Rr are illustrated in Figs. 4 and 5.
Table 2. Properties of the RR-WRCs with Different Mixing Times
1 Groups with the same letters in each column indicate that there was no statistical difference (p < 0.05) between the samples according to the Duncan’s multiple range test; values in parentheses are the standard deviations
With an increase in the mixing time from 6 min to 14 min, the mechanical properties of the composites increased at first, and then they decreased. The mechanical properties were lowest at a mixing time of 6 min, possibly because the mixing time was too short for sufficient contact between the rubber and WF to occur. The Ts reached a maximum at 8 min, the Rr reached a maximum at 11 min, and the Eband Ha reached maximum values at 10 min. However, when the mixing time was too long (over 11 min), the mixture temperature increased too fast, and created a radial temperature gradient in the mixing chamber that caused scorching. As a result, the Ts, Eb, Ha, and Rr decreased when the mixing time was longer than 11 min.
Fig. 4. Ts and Eb as functions of the mixing time
Fig. 5. Ha and Rr as functions of the mixing time
Effect of the Vulcanizing Temperature
The properties of the panels prepared at a target panel density of 1 g/cm3 with different vulcanizing temperatures, an RR content of 20%, and mixing time of 8 min are presented in Table 3. The polynomial relationships between the vulcanizing temperature and the Ts, Eb, Ha, and Rr are illustrated in Figs. 6 and 7.
The mechanical properties of the composites were relatively low at a vulcanizing temperature of 150 °C, possibly because of poor curing conditions and insufficient contact between the rubber and WF. The mechanical properties increased with the vulcanizing temperature at first. The Rr reached a maximum value at the vulcanizing temperature of approximately 157 °C, the Ts and Ha reached maximum values at 160 °C, and the Eb reached a maximum value at the vulcanizing temperature of 162 °C. However, higher vulcanizing temperatures (over 162 °C) probably dried out the wood cell antrum, which made the WFs fragile (Zhao et al. 2008). As a result, the Ts, Eb, Ha, and Rr decreased when the vulcanizing temperature was above 162 °C.
Table 3. Properties of the RR-WRCs with Different Vulcanizing Temperatures
Groups with the same letters in each column indicate that there was no statistical difference (p < 0.05) between the samples according to the Duncan’s multiple range test; values in parentheses are the standard deviations
Fig. 6. Ts and Eb as functions of the vulcanizing temperature
Fig. 7. Ha and Rr as functions of the vulcanizing temperature
Nonlinear Regression Model
Based on the experimental data, four regression equations for the Ts, Eb, Ha, and Rr (as functions of the RR content, mixing time, and vulcanizing temperature) were obtained and are shown below,
Ts = -932.6514 + 0.1511x1 – 0.0042x12 + 3.9714x2 – 0.1960x22 + 11.5206x3 – 0.0361x32 R2 = 0.9286 (2)
Eb = -10001 – 3.5948x1 – 0.00873x12 + 150.5329x2 – 7.4321x22 + 118.1859x3 – 0.3662x32 R2 = 0.9013 (3)
Ha = -3023.3 – 0.1518x1 + 0.0039x12 + 31.8171x2 – 1.6394x22 + 37.0365x3 – 0.1158x32 R2 = 0.8929 (4)
Rr = -2460.1 + 0.1299x1 – 0.0114x12 + 20.2650x2 – 0.9892x22 +
30.68x3 – 0.0980x32 R2 = 0.8907 (5)
where x1 is the RR content (%), x2 is the mixing time (min), x3 is the vulcanizing temperature (°C), and R2 is the coefficient of determination of each regression equation.
Using these four equations, the Ts, Eb, Ha, and Rr of a panel can be predicted based on the production conditions (RR content, mixing time, and vulcanizing temperature).
Optimization
When mill personnel want to alter the production parameters for a special application, a nonlinear programming model can be used as a guideline to set the production parameters. In the case of flooring for children, the panel material needs the highest Ts, and proper Ha and elasticity to prevent children from being hurt. In cases where a maximum Ts is desired for the panels, the standard for rubber and plastic floor covering materials (Chinese standard HG/T 3747.1 (2011)) requires minimum Ts, Eb, Ha, and Rr values of 0.3 MPa, 40%, 75 Shore A, and 38%, respectively. The nonlinear programming model (Eqs. 6 to 9) was found to be,
F(x) = -932.6514 + 0.1511x1 – 0.0042x12 + 3.9714x2 – 0.1960x22 + 11.5206x3 – 0.0361x32 (6)
Eq. 6 was subject to:
-10001 – 3.5948x1 – 0.00873x12 + 150.5329x2 – 7.4321x22 + 118.1859x3 – 0.3662x32 ≥ 40% (7)
-3023.3 – 0.1518x1 + 0.0039x12 + 31.8171x2 – 1.6394x22 + 37.0365x3 – 0.1158x32 ≥ 75 (8)
-2460.1 + 0.1299x1 – 0.0114x12 + 20.2650x2 – 0.9892x22 + 30.68x3 – 0.0980x32 ≥ 38 (9)
and the experimental condition constraints were as follows:
0 ≤ x1 ≤ 40
6 ≤ x2 ≤ 14
150 ≤ x3 ≤ 170
The optimal values for x1, x2, and x3 were 18%, 10 min, and 160 °C, respectively. The results indicated that a maximum Ts of 7.96 MPa for the panel was obtained with a RR content of 18%, vulcanizing temperature of 160 °C, and target panel density of 1 g/cm3. This resulted in the Eb, Ha, and Rr for the panel being 228.61%, 91.82 Shore A, and 42.27%, respectively, which were greater than the requirements stipulated by the standard HG/T 3747.1 (2011).
Micro-morphology of the RR-WRCs
The SEM photos of the cryogenically fractured surface parallel to the thickness of the RC, R0, R1, R2, R3, and R4 samples are shown in Fig. 8. A homogeneous structure is shown in Fig. 8a for the pure RC sample. Based on Figs. 8b to 8f of the WRCs, the WFs were easily identified and embedded in the RC/RR matrices. In Figs. 8b to 8d, wetted fibers with better interfacial bonding were seen between the WFs and RC/RR matrices, which confirmed there was an optimum reinforcing effect caused by the addition of RR.
