Tribological behavior of aluminum nanocomposites studied by application of response surface methodology

Abstract

An aluminum Al-0.8Al₂O₃ and Al − 1.6Al₂O₃ nanocomposite was prepared by a novel ultrasonic assisted stir casting method. A three-level Box-Behnken design of experiment was developed using response surface methodology. Dry sliding wear tests were performed as per the experimental design using a pin-on disc setup at room temperature. Analysis of variance (ANOVA) was applied to investigate the influence of process parameters, viz., wt.% reinforcement, load and sliding distance, and their interactions on specific wear rate and coefficient of friction. Further, a mathematical model has been formulated by applying response surface method in order to estimate the tribology characteristics such as wear and COF of the nanocomposites. The specific wear rate and coefficient of friction are significantly influenced by % of Al₂O₃ and load. The wear test parameters were optimized for minimizing specific wear rate and COF using desirability function approach. A set of optimum parameters of combination for AMMNC was identified as Al₂O₃–1.1 wt.%; load, 34 N and sliding distance, 2931 m with specific wear rate, 1.06 g/N-m; and coefficient of friction, 0.305. The AFM image of Al6061–1.1Al₂O₃ nanocomposite at optimized condition confirms the improvement in the wear surface smoothness of the nanocomposite compared to Al6061.

1 Introduction

Conventionally, aluminum alloys are extensively employed in the manufacture of components in aeronautics and automotive sector due to its low density, good corrosion resistance, high damping capacity, and formability [1,2,3]. But Al alloys in their pure form possess less wear resistance which limits their use in certain fields. However, reinforcing the matrix alloy with hard ceramic particles improves the wear resistance of aluminum. The friction and wear of AMCs depend on the reinforcement content, size, and type. The reinforcing particles can be a carbide, oxide, or nitride [4,5,6]. The reinforcement of aluminum oxide (Al₂O₃) particles in aluminum leads to their superior wear resistance combined with a low friction coefficient. The development of Al/Al₂O₃ nanocomposite is a novel material in the area of AMMNC_S with improved mechanical and tribological properties [7, 8]. The use of these novel Al/Al₂O₃ nanocomposite materials can improve energy efficiency, safety, and reliability [9,10,11,12,13,14].

Wang et al. established a novel method for dispersion of nanoparticles in molten metal by combined solidification process with ultrasonic cavitation [15]. Ezatpour et al. have successfully prepared A7075/Al₂O₃ nanocomposites with alumina powder of 0.4, 0.8, and 1.2 wt% by stir casting method followed by extrusion. The hardness, tensile, and compression strength of nanocomposite significantly improved compared to matrix alloy Al7075 [16]. Kumar and Balasubramanian developed a numerical model using RSM to forecast the abrasive wear rate of AA7075-SiC composite. The influence of volume % in reinforcement, reinforcement size, applied load, and sliding speed on the wear behavior of AA7075-SiC composite was analyzed, and it was inferred that the reinforcement size exerted the greatest effect on wear [17]. Basavarajappa et al. (2007) fabricated Al-SiC-Gr hybrid composites by liquid metallurgy route, and the results of dry sliding wear of the hybrid composite were compared with that of Al-SiC composite. Using Taguchi technique, a set of experimental data for optimization was developed in a controlled manner. He concluded that wear of Al-SiC composite and Al-SiC-Gr hybrid composites is greatly affected by the sliding distance [18]. Sahin (2010) developed Al-15 wt% SiC composite by powder metallurgy (PM) method and used analysis of variance (ANOVA) to examine the input parameters which affect the wear of the Al-15 wt% SiC composite. He stated that the abrasive wear is significantly affected by abrasive grain size followed by hardness [19]. Suresha and Sridhara (2010) developed Al-SiC-Gr hybrid composite by stir casting method and studied the tribological behavior under dry sliding condition. Design of experiment (DOE) technique was employed to study the influence of parameters like load, sliding distance, sliding speed, and % reinforcement. They concluded that load and sliding distance have positive effect on the wear of the hybrid composite [20]. In addition, various research work on the wear behavior of AMCs, and optimization has been done and reported [21,22,23].

However, the parametric studies on the dry sliding wear of AMMNCs are scarce. In this present work, an attempt is made to study the influence of process parameters, viz., load, sliding distance, % reinforcement, and their interactions on specific wear rate (SWR) and coefficient of friction (COF) of Al-Al₂O₃ nanocomposite by using RSM.

2 Materials and methods

2.1 Fabrication of Al6061-Al₂O₃ nanocomposites

Al6061-Al₂O₃ nanocomposite was fabricated by ultrasonic cavitation method by adding aluminum oxide (Al₂O₃) of 0.8 and 1.6% by weight basis. Figure 1 shows the ultrasonic cavitation assisted stir casting setup. For each casting, about 1 kg of Al6061 was first melted in the crucible to a temperature of 750 °C. Al₂O₃ nanoparticle was preheated to 800 °C for 1 h in a muffle furnace in order to improve the wettability [5]. When nano-Al₂O₃ were added in the matrix alloy melt, the viscosity of the Al6061 matrix alloy significantly increased. To improve the wettability of the nanoparticles, efficient ultrasonic treatment done in the melt pool and maintained casting temperature of 750 °C to ensure the flow ability inside the mold. The Al6061-Al₂O₃ nanocomposite was fabricated [6].

Fig. 1

(a) Experimental setup for fabricating nanocomposites

2.2 Characterizations of Al6061-Al₂O₃ nanocomposites

The mean particle size of the nano-Al₂O₃ particle purchased from US Research Nanomaterials, Inc., USA, was measured using Horiba SZ-100 particle analyzer. The particle size obtained by this process was measured as 60 nm. Figure 2 shows the distribution of particle size of nano-Al₂O₃ particle by particle analyzer in the form of the distribution curve.

Fig. 2

Size distribution of nano-Al₂O₃ particle

A Carl Zeiss NTS GMBH, Germany (SUPRA 55), Field Emission Scanning Electron Microscope (FESEM) images were used for the assessment of the morphology of raw materials and casted nanocomposite materials. Figure 3(a) shows the SEM image of Al6061, Fig. 3(b) shows SEM image of nano-Al₂O₃, Fig. 3(c) shows SEM image of Al-0.8Al₂O₃ nanocomposite, and Fig. 3(d) shows SEM image of Al-1.6Al₂O₃ nanocomposite. The nano-Al₂O₃ are uniformly distributed with some microclusters in Al-0.8Al₂O₃ nanocomposite. Agglomerations of nano-Al₂O₃ particles are observed in Al-1.6Al₂O₃ nanocomposite. When the particles were wetted in the metal melt, the particles will tend to sink or float to the molten melt due to the density differences between the reinforcement particles and the matrix alloy melt, so that the dispersion of the ceramic particles is not uniform and particles have high tendency for agglomeration and clustering. Wettability and distribution of reinforcement particles become more difficult when the particle size decreases to the nanoscales. This is due to the increasing surface area and surface energy of nanoparticles which cause an increasing tendency for agglomeration of reinforcement particles. Moreover, several structural defects such as porosity, particle clusters, oxide inclusions, and interfacial reactions arise from the unsatisfactory casting technology [7]. The scale bar for base material Al6061 is 20 μm. The scale bar for nano-Al₂O₃ is 200 nm to note down the nano-Al₂O₃ particle. The scale bar for Al606–0.8Al₂O₃ nanocomposite and Al6061–1.6Al₂O₃ nanocomposite is 1 μm which is in well agreement with the results of the various researchers S.A. Sajjadi et.al (2011) and HR Ezatpour et.al (2015) reported previously. The hardness of Al6061 is 55VHN, Al-0.8Al₂O₃ nanocomposite is 80VHN, and Al-1.6Al₂O₃ is 89VHN nanocomposite. There is a significant improvement in hardness of nanocomposites compared to the Al6061.This is attributed to the presence of nano-Al₂O₃ in the matrix alloy which restricts the dislocation movement [4, 6].

Fig. 3

(a) SEM of Al6061, (b) SEM of nano-Al₂O₃, (c) SEM of Al6061–0.8Al₂O₃ nanocomposite, (d) SEM of Al6061–1.6Al₂O₃ nanocomposite

X’Pert PRO (PANalytical) X-ray diffractometer with Cu Ka radiation at the voltage of 40 kV was used for carrying out XRD analysis with the current like 30 mA. Twenty data from 20° to 80° was collected by following continuous scan mode. The XRD results of Al6061, Al606–0.8Al₂O₃, and Al6061–1.6Al₂O₃ nanocomposites are shown in Fig. 4. The presence of aluminum and Al₂O₃ particles is confirmed by different peak intensities as shown in Fig. 4. The presence of major peaks indicates aluminum, and very minor peaks represent the Al₂O₃ particles. The peaks of Al₂O₃ nanocomposites seem weak because the weight fraction of Al₂O₃ (0.8 wt% i.e., 8 g, and 1.6 wt% i.e., 16 g in 1 kg of aluminum for a batch) is less than that of aluminum. It is also observed from Fig. 4 that there is no evidence of new phases (intermetallic compounds) which were attributed to no interfacial reactions and good dispersion of Al₂O₃ during casting process. Also it is clearly seen that the Al peaks marginally shift to higher angles as the weight % of the aluminum oxide is increased. The nanophase identification is in good agreement with Ravindran et al. (2013) and Li et al. (2018).

Fig. 4

(a) XRD of Al6061, (b) XRD of Al6061–0.8Al₂O₃ nanocomposite, (c) XRD of Al6061–1.6 Al₂O₃ nanocomposite

2.3 Wear test

The tribological studies were carried out on a computer integrated monitor (TR-20-PHM-M1 DUCOM) with an inbuilt load cell to measure the frictional force. For wear test, pins are machined to 10 mm diameter and 20 mm height. Disc surface of 25 mm diameter is maintained as sliding path. The tribo-testing was carried out at three different normal loads (20 N, 30 N, and 40 N). The sliding distance of 1000 m, 2000 m, and 3000 m and sliding velocity of 0.5 m/s were used. The COF is directly recorded by a computer system connected to the wear test machine. The wear test procedure followed according to ASTM G99 standard has been discussed in our earlier work [6]. The three-dimensional topography of worn surface was analyzed by XE 70, Park Systems, a Korean-made atomic force microscope (AFM).

3 Dry sliding wear studies by RSM method

3.1 Experimental design

RSM is used to find the relation between a set of input parameters and its output response. Design Expert-16 software is used to design the experiment and to study the effect of input parameters on SWR and COF of aluminum nanocomposite. In this study, three factors with three levels as shown in Table. 1 are used. Reinforcement % of Al₂O₃, sliding distance, and load are the input factors considered in this study. Table. 2 shows the number of experiments to be conducted as per the experimental design matrix.

Table 1 Input levels of sliding wear parameters

Table 2 Design factors and responses in coded form

4 Results and discussion

4.1 Mathematical model for SWR and COF

For analysis of SWR and COF, the quadratic model suggested from the fit summary is statistically significant. ANOVA (Table 3 and Table 4) contains the results of the quadratic models. When the value of R² approaches unity, the difference between actual and predicted data is very less which means that the response model fits the actual data. Further, if adequate precision (AP) is greater than 4, the predicted value can be compared with the average prediction error at the design point. Signal to noise ratio discriminates the adequate model if the value is greater than 4. These developed models have higher R² and AP values. The values obtained are as follows: R² = 0.9392 and AP = 18.290 for SWR and R² = 0.9664 and AP = 31.66 for COF. Subsequently, these mathematical models developed for SWR and COF are considered to be significant. The lack of fit test must be insignificant. The insignificant terms are removed by backward elimination process to fit in to the quadratic models [24]. After backward elimination process, the response equations of SWR and COF of the final quadratic models are presented below.

$$ \mathrm{Specific}\ \mathrm{wear}\ \mathrm{rate}\ \left(\mathrm{g}/\mathrm{N}-\mathrm{m}\right)\kern0.5em =+1.52-0.66\ast \mathrm{A}-0.28\ast \mathrm{B}-0.42\mathrm{A}\ast \mathrm{C}+0.69\ast \mathrm{A}\ast \mathrm{A} $$

$$ \mathrm{C}\mathrm{oefficient}\ \mathrm{of}\ \mathrm{friction}\left(\upmu \right)\kern0.5em =+0.34-0.12\ast \mathrm{A}-0.048\ast \mathrm{A}\ast \mathrm{C}+0.12\ast \mathrm{A}2 $$

Table 3 ANOVA for response surface quadratic model of specific

Table 4 ANOVA for response surface quadratic model of coefficient

4.2 Effect of dry sliding parameters on SWR

The surface graph in Fig. 5 shows the influence of % of Al₂O₃ on SWR of Al6061-Al₂O₃ nanocomposites for the applied load of 20–40 N and sliding distance of 1000–3000 m. It is observed that when the applied load increased, the SWR decreased due to the development of mechanically mixed layer (MML layer). On comparing all the nanocomposites, when the Al₂O₃% is increased, SWR decreased up to 1.2% and increases for 1.6%. The increase in SWR is due to more agglomeration at 1.6% of Al₂O₃._. For all applied load, the SWR decreased due to significant improvement in hardness of Al-Al₂O₃ nanocomposites. This is attributed to hard nano-Al₂O₃ particles on reinforcement with Al6061 matrix alloy that creates good interfacial bonding of the nanocomposite which improves the mechanical properties hardness and strength. During sliding, the Al₂O₃ particles act as load bearers and protect the soft matrix alloy from wear. The wear trend of the present work is consistent with Mosleh-shirazi and Akhlaghi [4]. The SWR is significantly influenced by % Al₂O₃ followed by load.

Fig. 5

Interaction plot for specific wear rate 3D plot

4.3 Effect of dry sliding parameters on COF

Figure 6 shows the effect of load, sliding distance, and % of Al₂O₃ on COF of Al-Al₂O₃ nanocomposite. It is noted that COF decrease when load is increased from 20 to 40 N. Due to the formation of mechanically mixed layer in Al-Al₂O₃ nanocomposites at high load, friction between disc and pin get reduced resulting in lowered COF compared to Al6061 matrix alloy. The existence of a stable mechanically mixed layer at the point of contact is the key factor in reducing the COF of nanocomposite independent of sliding distance. The COF value decreases as the % of Al₂O₃ is increased up to 1.2%, but the COF increases beyond 1.2% of Al₂O₃. The COF is significantly influenced by % Al₂O₃ followed by load.

Fig. 6

Interaction plot for coefficient of friction 3D plot

4.4 Worn surface analysis for various nanocomposites

The morphology of the wear surface of Al6061 matrix and Al606–0.8Al₂O₃ nanocomposite pin tested at load of 20 N–40 N is shown in Fig. 7(a–e). The SEM image provides valuable information about the wear mechanism of alloy and nanocomposite. On increasing the load to 40 N, abrasive wear decreases, and adhesive wear increases, which is clearly identified by big size wear debris as shown in Fig. 7(a). The dominating wear mechanisms of Al6061 are abrasive and adhesive wear.

Fig. 7

SEM of worn surfaces of (a) Al6061 D-2000 m L-40 N, (b) Al-0.8 Al₂O₃ D-3000 L-20 N, (c) Al-0.8Al₂O₃ D-3000 L-40 N, (d) Al-1.6Al₂O₃ D-2000 L-20 N, (e) Al-1.6Al₂O₃ D-2000 L-40 N. L load, D sliding distance

Figure 7(b) shows the worn surface of Al-0.8Al₂O₃ nanocomposite at the applied load of 20 N. The SEM image shows parallel and less deep grooves due to abrasive wear. Figure 7(c) shows the worn surface of Al-0.8Al₂O₃ nanocomposite at the applied load of 40 N. It is observed that deep craters are formed on the nanocomposite worn surface which indicates that abrasive wear is dominant at high load. The worn surfaces of Al6061–0.8Al₂O₃ nanocomposite possess less plastic deformation when comparing to the wear surface of Al6061. From Fig. 7(d) and Fig. 7(e), it is observed that the wear surface appears more rough compared to Fig. 7(b) and Fig. 7(c). The most important factor is the presence of hard Al₂O₃ particles in the soft Al6061, which prevent the Al6061 from sliding wear and strengthen the Al6061. Moreover, these Al₂O₃ particles restrict the dispersion and cutting of the disc in to nanocomposite surface, thereby preventing delamination. In addition, the wear debris formed during dry sliding gets oxidized and forms a protective layer called mechanically mixed layer (MML) on the surface of nanocomposites. The worn surface of 0.8Al₂O₃ nanocomposite appears smooth, compared to Al6061 matrix alloy. The MML layer findings are consistent with Mosleh-Shirazi and Akhlaghi [4].

Figure 8 shows the AFM images of worn surfaces of (a) Al6061 D-2000 m L-40 N (b) Al-0.8Al₂O₃ D-3000 m L-20 N (c) Al-0.8Al₂O₃ D-3000 m L-40 N (d) Al-1.6Al₂O₃ D-3000 m L-20 N and (e) Al-1.6Al₂O₃ D-3000 m L-40 N (L load, D sliding distance). The results of AFM analysis match with wear examinations and the above said morphologies of worn surfaces. The AFM image of Al-0.8Al₂O₃ D-3000 m L-40 N exhibits relatively smooth surface after wear compared to matrix alloy. The worn surface of Al6061 indicates rough surface compared to nanocomposites.

Fig. 8

AFM (3D) of (a) Al6061 D-2000 m L-40 N, (b) Al-0.8Al₂O₃ D-3000 L-20 N, (c) Al-0.8 Al₂O₃ D-3000 L-40 N, (d) Al-1.6Al₂O₃ D-2000 L-20 N, (e) Al-1.6Al₂O₃ D-2000 L-40 N. L load, D sliding distance

4.5 EDS of wear worn surfaces

Figure 9(a–c) indicates the EDS of worn surface of alloy, nanocomposite after wear test under 40 N load. Figure 9(a) shows the EDS of the worn surface of alloy. It majorly contains aluminum which reveals that the alloy wear was high compared to the disc surface. The MML layer formation owing to the oxidation of iron and aluminum debris of nanocomposite is confirmed by the EDS analysis of Al6061–0.8Al₂O₃ nanocomposite, and Al6061–1.6Al₂O₃ nanocomposite was shown in Fig. 9(b) and Fig. 9(c). The hard MML lowers the SWR of the nanocomposite compared to Al6061 alloy [4].

Fig. 9

EDS results of wear worn surface for various composites under 40 N load (a) Al6061, (b) Al6061–0.8Al₂O₃, (c) Al6061–1.6Al₂O₃

4.6 Multiresponse optimization of Al6061-Al₂O₃ nanocomposite

It is important to identify the optimum parameters for any process. It is difficult to select the optimum condition for SWR and COF of Al6061-Al₂O₃ nanocomposite during dry sliding since it involves lots of process variables. Multiresponse tribological parameters are optimized using grey relational analysis, artificial neural networks with genetic algorithms, Taguchi’s method, and desirability function approach. The desirability function approach is extensively used for optimization of multiresponse parameter in industry [24,25,26,27]. The value of desirability was evaluated using Design Expert software. RSM develops a set of data depending on input parameter to optimize the SWR and COF. The aim of optimization is to find the input parameters which minimize SWR and COF.

The various process parameters, its goal, upper and lower limit, and its optimum value of input parameters (load, distance, % of Al₂O₃) and output parameters (SWR and COF) are listed in Table 1. Out of various sets of input conditions, the one with highest desirability is chosen as optimum condition, and it is shown in Figs. 10 and 11. A dot on each ramp indicates response prediction for that parameter. The overall desirability function of SWR and COF is shown in the bar graph of desirability. The range of desirability value is from 0 to 1. When the desirability value is close to 1, it indicates the closeness of the target. The set of parameters for optimization is as follows: Al₂O₃–1.1 wt.%; load-34 N and sliding distance-2931 m; with specific wear rate 1.06 g/N-m; and coefficient of friction, 0.305.

Fig. 10

Ramp function graph of desirability

Fig. 11

Bar graph of desirability

4.7 SEM and AFM image for optimization confirmation

The SEM and AFM analysis of the worn surface of Al6061–1.1Al₂O₃ at optimum condition is used to check the accuracy of the model. Figure 12(a) shows the worn surface at optimum conditions: load, 34 N, and sliding distance, 2931 m. It is observed that only small scratches are seen at few places of the worn surface, which is an indication of low SWR. It confirms the accuracy of the model. Figure 12(b) shows the 3D AFM image of the Al-Al₂O₃ nanocomposite. The height of the groove indicates the roughness of the surface. The lower the height, the lower will be the surface roughness. It is clear from the AFM image that at the optimum % of Al₂O₃ particles in the nanocomposite, there is decrease in groove height indicating the smoothness of the worn surface.

Fig. 12

(a) SEM of Al-1.1Al₂O₃ worn surface (b) AFM of Al-1.1Al₂O₃ worn surface

5 Conclusion

Aluminum metal matrix composite reinforced with Al₂O₃ nanoparticles was fabricated effectively employing ultrasonic cavitation-based solidification process.

The dry sliding wear of Al-Al₂O₃ nanocomposites were successfully carried out using Box-Behnken design of experiments of RSM by conducting 17 experiments for three factors at three levels.

The influence of process parameters on SWR and COF of Al-Al₂O₃ nanocomposites was investigated. A mathematical model was developed to predict the SWR and COF of Al-Al₂O₃ nanocomposites incorporating the effects of % reinforcement Al₂O₃, applied load, and sliding distance. The predicted values match the experimental values reasonably well with R² of SWR and COF.

ANOVA was used to check the adequacy of the model. The test results show minimum deviation between actual value and predicted value which confirms the accuracy of the developed model. The SWR and COF are significantly influenced by % Al₂O₃ followed by load and sliding distance.

The parameters were optimized using desirability-based multiresponse optimization technique to minimize the SWR and COF. The optimum parameters of combination setting for Al-Al₂O₃ nanocomposites obtained are Al₂O₃ of 1.1 wt.%, load of 34 N, and sliding distance of 2931 m for minimizing SWR and COF.

The AFM image of Al6061–1.1Al₂O₃ nanocomposite at optimized condition confirms the improvement in the wear surface smoothness of the nanocomposite compared to Al6061.

Nano-Al₂O₃ reinforced composites exhibited improved performance due to the formation of mechanically mixed layer (MML) at higher levels.