Tension

Figure 1 shows the variation in the mean values and the standard deviation for the maximum resistance to tension of all the composite samples. The presence of the coupling agent was observed to favor maximum resistance. The highest resistance, of 17.8 MPa, was registered with the 30/70 ratio and the -60/+80 mesh size; conversely, the 50/50 ration for the same -60/+80 mesh size, without a coupling agent had the least resistance (11.9 MPa), without, however, exceeding the value for plastic (31.4 MPa).

Figure 1 Maximum resistance to tension of bamboo/plastic composite materials. 

Researches by ^{Cárdenas (2012)}, ^{Lisperguer et al. (2013)}, ^{Martínez-López et al. (2014)} and ^{Samariha et al. (2015)} obtained results of 15 to 28.8 MPa for composite materials with 50 % particles. ^{Moya et al. (2012)} registered values of 17 to 38 MPa for composites with 20 to 60 % particles; however, these values did not exceed those of plastic (40 MPa). Likewise, ^{Idrus et al. (2011)} cite 19 MPa for composites with 30 % particles. Furthermore, ^{López et al. (2012)} obtained a resistance of 60 and 50 MPa for materials with 40 and 30 % jute fibers when the coupling agent was added. ^{Durowaye et al. (2014)} register values of 4 to 6 MPa with 25 % particles of sisal.

Results for the assessed test specimens are lower than those cited in the bibliography; furthermore, no formulation had values above those of plastic. This may be due to the clustering of the particles, the irregularity of the filler form, or a low interaction between the reinforcement and the matrix resulting in poor stress transfer, causing the load to act as a defect instead of a reinforcement in the matrix and therefore producing the crack more quickly (^{Liu et al., 2008}; ^{Rosa et al., 2009}; ^{Idrus et al., 2011}; ^{Cárdenas, 2012}; ^{Santos et al., 2012}; ^{Moya et al., 2012}; ^{Naghmouchi et al., 2013}; ^{Ravi et al., 2014}).

Although formulations with a coupling agent had lower values, there was an increase in the results; this indicates that the particles formed ester bonds, and therefore the composite material can bear more load (^{Rosa et al., 2009}). Figure 2 shows void spaces within the composite material, taken with a USB DIGITAL MICROSCOPE (VEHO VMS 004 spectrometer.

Figure 2 Images of void spaces (circled) in a cross-section of bamboo/plastic composite materials obtained with a -60/+80 mesh and 50/50, 40/60 and 30/70 ratios, respectively. 

According to the statistical analysis, there are no significant differences in the interaction between the mesh size and the coupling agent concentration (p = 0.5802); however, such differences were found in other variables (p<0.0001). Table 3 records the values obtained from the mean comparison test, which show that the best combination was that of the -60/+80 particle size, the 30/70 ratio and the use of MAPP.

The variation in the mean values and standard deviation of the modulus of elasticity to tension is shown in Figure 3. The MOE was observed to increase in most combinations with larger amounts of particles in the mixture (the value for polypropylene was 0.6 GPa). In the case of test specimens with an additive, the benefit of adding the coupling agent is evident, as it increased the modulus of elasticity of the composite material. The 50/50 mixing ratio with particle size -20/+40 attained the highest MOE (1.7 GPa); conversely, the 40/60 mixing ratio with the same particle size had the lowest value for MOE (1.2 GPa).

Figure 3 MOE of bamboo/plastic composite materials in tension. 

Researches by ^{Rosa et al. (2009)}, ^{Idrus et al. (2011)}, ^{López et al. (2012)}, ^{Naghmouchi et al. (2013)} and ^{Ravi et al. (2014)} indicate that the MOE values are directly proportional to the increase in the content of particles; other values cited by ^{Cárdenas (2012)}, ^{Lisperguer et al. (2013)} and ^{Samariha et al. (2015)} vary from 0.9 to 3.6 CPa for composite materials with 50 % particles. ^{Moya et al. (2012)} documented a value of 7 to 9 GPa for composites with 20 to 60 % reinforcement. ^{Idrus et al. (2011)} registered 1.1 GPa for compounds with 30 % particles; ^{López et al. (2012)} cite values of 7 and 6 GPa with 40 and 30 % jute fibers.

The present study records higher values than those of plastic; this is because the particles provide rigidity to the composite material, i.e. they act as a reinforcement (^{Rosa et al., 2009}; ^{Moya et al., 2012}; ^{Naghmouchi et al., 2013}; ^{Ravi et al. 2014}). However, most results are below those cited in the bibliography. In formulations using a coupling agent, there was an increase in the results; the 50/50 ratio had the highest value, as the coupling agent was able to bond with the hydroxyl groups and thereby improve the rigidity of the composite (^{Takatani et al., 2008}; ^{Rosa et al., 2009}; ^{López et al., 2012}; ^{Naghmouchi et al., 2013}).

The statistical analysis evidenced significant differences in all the variables (p<0.0477). Table 4 shows the multiple mean comparison; the best combination was observed to be the one with -60/+80 and -40/+60 particle sizes, the 50/50 and 40/60 ratios and the use of MAPP.

Flexion

Mixtures without an additive show a similar tendency as to the various particle sizes. In the case of test specimens with an ad-ditive, the presence of a coupling agent improves the resistance of composite materials; the highest resistance (22.7 MPa) was reached using the 30/70 mixing ratio; conversely, the -40/+60 particle and the 50/50 mixing ratio had the lowest resistance (15 MPa). These values, however, did not surpass those of plastic (25.4 MPa) (Figure 4).

Figure 4 Maximum resistance in flexion of bamboo/plastic composite materials. 

^{Cárdenas (2012)}, ^{Lisperguer et al. (2013)}, ^{Martínez-López et al. (2014)}, and ^{Samariha et al. (2015)} register values of 17.5 to 60 MPa for composite materials with 50 % particles; ^{Bahari and Krause (2016)} register a value of 59.6 MPa for composites with 25 % particles; ^{Stark and Rowland (2003)} cite values of 38.7 to 42.6 MPa for composites with 40 % particles. Likewise, ^{Idrus et al. (2011)} point out that resistance increases in direct proportion to the number of fibers; results shown in Figure 4 do not coincide.

Most values obtained in the present study are lower than those reported in the bibliography; furthermore, no formulation surpassed the value for plastic. This may be due to the presence of void spaces between the fiber and the matrix, which result in a poor interfacial union, as well as to the clustering of particles, which leads to a scarce dispersion in the matrix (^{Ravi et al., 2014}). All of this caused a poor stress transfer, which in turn produced the flaw more rapidly. As in the outcome for maximum resistance in tension, the composite materials made with a coupling agent increased their resistance, which is an indication that the particles formed ester bonds that favored their ability to bear more load (^{Rosa et al., 2009}).

There were no significant differences in the interaction of the mixing ratio and the coupling agent concentration (p=0.0552); however, some such differences did occur for the other variables (p<0.0001). Table 5 summarizes the results of the multiple mean comparison. The best combination was obtained with the -60/+80 particle size, the 30/70 and 40/60 ratios and the use of MAPP.

The variations in the modulus of elasticity means and standard deviation for flexion are shown in Figure 5; an increase in the MOE is observed in most combination with an increased number of particles in the mixture (the value for polypropylene was 0.6 GPa). The test specimens with an additive displayed a similar tendency for the various formulations; the highest MOE (0.9 GPa) was attained using the 40/60 mixing ratio with the -60/+80 particle size, while the 30/70 mixing ratio with the -20/+40 particle size had the lowest MOE (0.7 GPa).

Figure 5 MOE of the bamboo/plastic composite materials in flexion. 

In their analysis of composites with fibers and polypropylene, ^{Idrus et al. (2011)}, ^{Mattos et al. (2014)} and ^{Ravi et al. (2014)} point out that the MOE increases in direct proportion to the number of fibers. On the other hand, ^{Cárdenas (2012)}, ^{Lisperguer et al. (2013)},^{Chen et al. (2014)}, ^{Samariha et al. (2015)}, and ^{Bahari and Krause (2016)} register values ranging between 0.9 and 5.2 GPa for composites with 50 % particles, while ^{Stark and Rowland (2003)} and ^{Liu et al. (2008)} report values of 2.1 to 3.2 GPa for composites with 30 to 40 % reinforcement.

The results of the present research, in most formulations, exceed the values for plastic. This is because the particles confer rigidity to the composite material (^{Ravi et al., 2014}; ^{Bahari and Krause 2016}). Nevertheless, these values are below those cited in the literature due to the poor union between the materials that produces void spaces, particle clustering as a result of poor dispersion in the matrix (^{Stark and Rowland, 2003}; ^{Ravi Kumar et al., 2014}; ^{Bahari and Krause, 2016}).

The statistical analysis indicated that there are no significant differences between the double interactions of the coupling agent concentration with the mesh size (p = 0.7565) and mixing ratio (p = 0.5995). Table 6 shows the values for the multiple mean comparison; the best combination resulted from the use of a -60/+80 particle size, 50/50 and 40/60 mixing ratios, and MAPP.

Impact

Figure 6 represents the variation in the mean values and the standard deviations of the impact trial for all the samples of composite materials. No marked tendency in regard to the energy absorbed by the composite materials is observed; they all appear to be numerically similar. As for the test specimens with an additive, there is a light increase in the values of energy absorbed by the composite materials. The mixture with a -40/+60 particle size and 30/70 mixing rate reached the highest value (0.44 J), and the lowest value (0.33 J) was for the -20/+40 particle size with the 30/70 ratio; however, the values for plastic (2 J) were higher.

Figure 6 Resistance of bamboo/plastic composite materials to impact. 

^{Faruk et al. (2012)} point out that the impact trial is the ability of the material to resist a fracture after stress is applied at a high speed. Kumar et al. (2014) emphasize the importance of the fibers in transmitting the stress to the matrix. A study by ^{Kinochita et al. (2009)} shows the values of impact to be directly proportional to the increase in particle size. ^{Cárdenas (2012)} registers values of 660 to 682 K m^-1, while ^{Lisperguer et al. (2013)} obtained values of 6 to 7 J m^-1 for composites with 50 % reinforcement. ^{Bahari and Krause (2016)} record a value of 3.8 kJ m^-² for composites with 25 % reinforcement; both these values are lower than the 20 kJ m^-² obtained for plastic.

^{Durowaye et al. (2014)} document a value of 1.9 to 1.2 for composites with 25 % sisal particles. The values for the assessed test specimens are lower, and no formulation exceeded the value for plastic. This is due to an inadequate adhesion of the materials, which generates microspaces between the particles and the matrix; furthermore, as the number of particles increases, these may cluster due to poor dispersion in the matrix (^{Stark and Rowland, 2003};^{Ravi et al., 2014}; ^{Bahari and Krause, 2016}).

The statistical analysis evidenced the presence of significant differences in the coupling agent concentrations (p = 0.0001) and in the double interaction between mesh size and mixing ratio (p<0.0001) and coupling agent concentration (p = 0.0187), but not in regard to other variables (p>0.0814). The multiple mean comparison indicates that the best combination resulted from the use of a -60/+80 particle size, the 50/50 and 40/60 mixing ratios and MAPP (Table 7).

In most formulations with MAPP there was an increase in the mechanical properties due to the increased interfacial adhesion obtained with the use of a coupling agent (^{Lisperguer et al. 2013}).