3.4 Wear Behavior with Taguchi Design of Experiment
This method is an effective design concept [23, 24], usually designed to produce better quality products in a cost-effective manner and is extensively used for various industrial applications. Generally, it is used to evaluate the effect of varying parameters. The test samples were subjected to experimentation under the room temperature (27ºC) by considering Taguchi analysis of L27 orthogonal array (OA). Wear test samples were prepared based on the ASTM G-99 standard size of 8 mm dia and 35 mm of length. Figure 4 shows the wear testing equipment which was used for the present investigation. During wear tests, the test samples were held rigidly against the rotating hard disc (steel). The steel disc and test samples were cleaned thoroughly by using organic compound (acetone) to maintain the accuracy in outcomes. In the present investigation, wear behaviour is expressed as weight loss (Gms) of the test specimen. During the experimentations, the specimens were systematically cleaned by using acetone fluid, and then weighed by digital weighing apparatus to maintain the accuracy of ± 0.0001 gm. After each wear test trials, the samples were washed with acetone fluid and then the final weight of the test samples were measured precisely. Here a wear rate is calculated by difference of initial and final weights of test specimens. Test trials were performed based on the selected parameters and their levels, which is tabulated in Table 2. Orthogonal array (OA) of 27 tests and their outcomes of samples for the Al7075, mono-composites (Al7075 + 10%Al2O3) and hybrid composites (Al7075 + 10% Al2O3 + 5%SiC) are depicted in Table 3.
Table 2
Input parameters and their levels
Sl. No.
|
Process parameters
|
Level – 1
|
Level − 2
|
Level − 3
|
1
|
Load (N)
|
7.5
|
10
|
12.5
|
2
|
Sliding Speed (rpm)
|
150
|
300
|
450
|
3
|
Sliding Distance (m)
|
300
|
600
|
900
|
Table 3
Taguchi L27 orthogonal array and their response
Trial No.
|
Load (N)
|
Sliding Speed (rpm)
|
Sliding Distance (m)
|
Wear Loss (Gms)
|
(Al-7075)
|
Mono Composite
(10% Al2O3)
|
Hybrid Composite
(10% Al2O3 + 5% SiC)
|
1
|
7.5
|
150
|
300
|
0.035
|
0.015
|
0.011
|
2
|
7.5
|
150
|
600
|
0.042
|
0.020
|
0.013
|
3
|
7.5
|
150
|
900
|
0.049
|
0.030
|
0.026
|
4
|
7.5
|
300
|
300
|
0.038
|
0.017
|
0.012
|
5
|
7.5
|
300
|
600
|
0.041
|
0.018
|
0.013
|
6
|
7.5
|
300
|
900
|
0.059
|
0.030
|
0.021
|
7
|
7.5
|
450
|
300
|
0.056
|
0.025
|
0.023
|
8
|
7.5
|
450
|
600
|
0.064
|
0.032
|
0.028
|
9
|
7.5
|
450
|
900
|
0.071
|
0.038
|
0.035
|
10
|
10.0
|
150
|
300
|
0.049
|
0.020
|
0.017
|
11
|
10.0
|
150
|
600
|
0.068
|
0.041
|
0.039
|
12
|
10.0
|
150
|
900
|
0.074
|
0.050
|
0.045
|
13
|
10.0
|
300
|
300
|
0.060
|
0.042
|
0.040
|
14
|
10.0
|
300
|
600
|
0.072
|
0.051
|
0.042
|
15
|
10.0
|
300
|
900
|
0.075
|
0.055
|
0.050
|
16
|
10.0
|
450
|
300
|
0.056
|
0.030
|
0.026
|
17
|
10.0
|
450
|
600
|
0.064
|
0.038
|
0.036
|
18
|
10.0
|
450
|
900
|
0.066
|
0.045
|
0.038
|
19
|
12.5
|
150
|
300
|
0.048
|
0.020
|
0.018
|
20
|
12.5
|
150
|
600
|
0.052
|
0.039
|
0.030
|
21
|
12.5
|
150
|
900
|
0.068
|
0.040
|
0.036
|
22
|
12.5
|
300
|
300
|
0.071
|
0.055
|
0.051
|
23
|
12.5
|
300
|
600
|
0.072
|
0.056
|
0.049
|
24
|
12.5
|
300
|
900
|
0.079
|
0.060
|
0.057
|
25
|
12.5
|
450
|
300
|
0.081
|
0.065
|
0.061
|
26
|
12.5
|
450
|
600
|
0.085
|
0.070
|
0.065
|
27
|
12.5
|
450
|
900
|
0.089
|
0.073
|
0.070
|
The effect of process parameters were studied by ANOVA technique, main effect plots, regression analysis, normal probability plots and surface plots. The “smaller is better” condition is used for analysis of wear loss of the MMCs. The ANOVA outcomes are depicted in Table 4–6 for wear loss. The confidence level of P-value < 0.05 were measured as significant contribution to the performance [25, 26]. The significance of the parameters was confirmed through main effects plots as depicted in Fig. 5. ANOVA analysis for wear characteristics of Al 7075 alloy are as shown in Table 4. An ANOVA result indicates the load (36.20%) is more significant on wear loss followed by sliding speed (21.66%) & sliding distance (18.54%). Correspondingly, whereas in Al7075 + 10% Al2O3, load (48.01%), sliding speed (14.91%) and sliding distance (13.7%) have high influence on wear loss given in Table 5. However, for hybrid MMCs (Al7075 + 10% Al2O3 + 5% SiC), load (48.66%), sliding speed (16.17%) and sliding distance (10.59%) have great influences on wear loss given in Table 6. The outcomes reveal that, the load is the major significant parameter followed by other two parameters for wear loss in all the composites. From the outcomes it is confirmed that in hybrid composites (Al7075 + 10% Al2O3 + 5% SiC), exhibit higher low wear rate when compared to Al alloy (Al7075) and mono composite (Al7075 + 10% Al2O3) because of hard ceramic particulates. Similar outcomes have been observed by other researchers [27]. The ceramic particles protruding from the composites surface produced sharp asperity and formed non-uniform interaction (contact) between the samples and counter-face which tends to increase wear rate. On addition of secondary hard ceramic reinforcement in MMCs, the distances between particulates reduce affecting the presence of more reinforcing particulates. Earlier studies exhibited that the existence of hard reinforcement particulates led to enhanced hardness. The wear behaviour was related to the inversion of material hardness and wear loss [12]. From the Fig. 2(a-c), it is seen that the increase in wear parameters led to increase in wear loss. Generally, the key reasons for the obtained outcomes are due to the development of an oxide layer on the matrix surface resulting in more wear. And, also, due to increase in temperature as well as softening of the composite surface results in high wear. Whereas, in case of mono MMCs, the reinforcing of hard particulates improves wear resistance. As observed, the sliding speed generally increased the wear loss because of high strain rate and this led to delamination. The hybrid composites exhibit high wear resistance while compared to mono composites and monolithic [28, 29].
Table 4
ANOVA results of Al 7075 alloy (monolithic)
Source
|
DF
|
Seq. SS
|
Adj. SS
|
Adj. MS
|
F
|
P
|
Cont. (%)
|
Remarks
|
Load (N)
|
1
|
0.0020056
|
0.0020056
|
0.0020056
|
35.3014
|
0.0000047
|
36.20
|
Significant
|
Sliding Speed (rpm)
|
1
|
0.0012005
|
0.0012005
|
0.0012005
|
21.1310
|
0.0001271
|
21.66
|
Significant
|
Sliding Distance (m)
|
1
|
0.0010276
|
0.0010276
|
0.0010276
|
18.0868
|
0.0002998
|
18.54
|
Significant
|
Error
|
23
|
0.0013067
|
0.0013067
|
0.0000568
|
|
|
23.58
|
|
Total
|
26
|
0.0055403
|
|
|
|
|
100
|
|
R-Sq = 76.41%
|
Table 5
ANOVA results of mono composite
Source
|
DF
|
Seq. SS
|
Adj. SS
|
Adj. MS
|
F
|
P
|
Cont. (%)
|
Remarks
|
Load (N)
|
1
|
0.0035561
|
0.0035561
|
0.0035561
|
46.0132
|
0.0000006
|
48.01
|
Significant
|
Sliding Speed (rpm)
|
1
|
0.0011045
|
0.0011045
|
0.0011045
|
14.2916
|
0.0009689
|
14.91
|
Significant
|
Sliding Distance (m)
|
1
|
0.0009680
|
0.0009680
|
0.0009680
|
12.5253
|
0.0017526
|
13.07
|
Significant
|
Error
|
23
|
0.0017775
|
0.0017775
|
0.0000773
|
|
|
24.00
|
|
Total
|
26
|
0.0074061
|
|
|
|
|
100
|
|
R-Sq = 76.00%
|
Table 6
ANOVA results of hybrid composite
Source
|
DF
|
Seq. SS
|
Adj. SS
|
Adj. MS
|
F
|
P
|
Cont. (%)
|
Remarks
|
Load (N)
|
1
|
0.0036125
|
0.0036125
|
0.0036125
|
45.5658
|
0.0000007
|
48.66
|
Significant
|
Sliding Speed (rpm)
|
1
|
0.0012005
|
0.0012005
|
0.0012005
|
15.1423
|
0.0007366
|
16.17
|
Significant
|
Sliding Distance (m)
|
1
|
0.0007867
|
0.0007867
|
0.0007867
|
9.9232
|
0.0044813
|
10.59
|
Significant
|
Error
|
23
|
0.0018235
|
0.0018235
|
0.0000793
|
|
|
24.56
|
|
Total
|
26
|
0.0074232
|
|
|
|
|
100
|
|
R-Sq = 75.44%
|
The response for mean has been evaluated based on the rank allotted to the mean values as shown in Table 7–9. The factors are statistically significant and also it could be seen that the load is an important factor as the delta of means ranked it as 1 followed by the speed and sliding distance for Al alloy and composite materials.
Table 7
Response data for means for monolithic
Levels
|
Load (N)
|
Sliding speed (rpm)
|
Sliding distance (m)
|
1
|
0.05056
|
0.05389
|
0.05489
|
2
|
0.06489
|
0.06300
|
0.06222
|
3
|
0.07167
|
0.07022
|
0.07000
|
Delta
|
0.02111
|
0.01633
|
0.01511
|
Rank
|
1
|
2
|
3
|
Table 8
Response data for means for mono composite
Levels
|
Load (N)
|
Sliding speed (rpm)
|
Sliding distance (m)
|
1
|
0.02500
|
0.03056
|
0.03211
|
2
|
0.04133
|
0.04267
|
0.04056
|
3
|
0.05311
|
0.04622
|
0.04678
|
Delta
|
0.02811
|
0.01567
|
0.01467
|
Rank
|
1
|
2
|
3
|
Table 9
Response data for means for hybrid composite
Levels
|
Load (N)
|
Sliding speed (rpm)
|
Sliding distance (m)
|
1
|
0.02022
|
0.02611
|
0.02878
|
2
|
0.03700
|
0.03722
|
0.03500
|
3
|
0.04856
|
0.04244
|
0.04200
|
Delta
|
0.02833
|
0.01633
|
0.01322
|
Rank
|
1
|
2
|
3
|
Regression analysis exhibits the relationship among two / more predictor variables by using a linear regression equation. A regression equation establishes the correlation among the wear parameters and their interactions. The regression equations for monolithic, mono composite and hybrid composite are shown in Eqs. (1), (2) and (3) respectively.
Wear loss of Al7075 alloy = -0.011 + 0.004 Load + 5.44e-005 Sliding Speed + 2.51e-005 Sliding Distance
|
(1)
|
Wear loss of mono composite = -0.04 + 0.005 Load + 5.22e-005Sliding Speed + 2.4e-005 Sliding Distance
|
(2)
|
Wear loss of hybrid composite = -0.050 + 0.0056 Load + 5.4e-005 Sliding Speed + 2.20e-005 Sliding Distance
|
(3)
|
Generally, regression equations have been used to study the reactions within the parameters. To check the accuracy of predicted values, the test trials have been conducted and the comparison among the experimentation and predicted values is depicted by graphical illustrations. The response of predicted and experimental values for wear behavior of Al7075, mono composites and hybrid MMCs is shown in Fig. 6.
The contour plots which are executed by the regression model and drawn to show the combined effects of parameters used in the present investigation. Generally, these plots are used to identify the interactions between the two parameters. By studying these plots, the optimized values of the each parameter could be predicted [30]. The contour plots for wear loss versus the independent parameters of all the materials are shown in Fig. 7.
Figure 7 exhibits the outcomes of wear loss in variation of wear parameters with material compositions. It is observed that wear loss increases by increasing in load, speed and sliding distance. High friction was detected at higher load and speed. Due to this, the temperature developed in the test specimen surface has increased. Therefore, the hardness of materials has reduced due to brittleness and wear loss on the test specimen has increased. As the temperature increases, the bonding among matrix and reinforcement gradually reduces and the material becomes soft [31].
Confirmation test is the main purpose of studying the confirmatory trial was to evaluate the optimal levels of varying process parameters selected. Confirmation trials were conducted based on the optimized values from the MEP (Fig. 4(a-c)). The selected levels of parameters are depicted in Table 10. The confirmatory experimental trails were carried out and the outcomes were compared with OA experimental values and are tabulated in Table 11. The result indicates that the calculated errors are less than 10% for all the composite materials. This is within acceptable limit.
Table 10
Confirmatory test parameters for different compositions
Process parameters
|
Load (N)
|
Sliding speed (rpm)
|
Sliding distance (m)
|
Optimized values of all the composition
|
7.5
|
150
|
300
|
Table 11
Confirmation test results for all the composition along with regression values
Composition
|
Process parameters
|
Confirmation test results
|
OA experimental values
|
Error %
|
Base alloy
(Al-7075)
|
Load (N) : 7.5
Sliding speed (rpm) : 150
Sliding distance (m) : 300
|
0.036
|
0.035
|
2.85
|
Mono Composite
(10% Al2O3)
|
0.014
|
0.015
|
6.66
|
Hybrid Composite
(10% Al2O3 + 5% SiC)
|
0.012
|
0.011
|
9.09
|
To evaluate the wear characteristics of the MMCs, SEM investigation on the wornout composites samples was conducted. Usually, the wear behaviour will be influenced by the characteristics of the wornout surface of the MMCs. Figure 8 depicts the SEM image of wornout surface of monolithic, mono and hybrid MMCs tested at load of 12.5 N, speed of 450 rpm and sliding distance of 900 m. The SEM images clearly exhibit how the wear track produced on the surface of Al7075, mono-composites and hybrid composites. Figure 8(a) shows wornout surface of base alloy Al7075. The image reveals unstiffening of the monolithic at interface temperature which creates plastic deformation. At higher level of load, speed and sliding distance, adhesive wear mechanism was seen in Al7075. The image clearly shows that the wear intensity is high without the presence of reinforcement. It concludes that the matrix without reinforcements generally, undergoes wide plastic deformation. Therefore, more losses in material are seen in wear surface. Figure 8(b) depicts the SEM image of Al7075 + 10% Al2O3 composite with more number of shallow grooves. Due to presence of hard reinforcement, generally, the wear resistance will be high. And also, the wornout surface observed is rough because of ceramic particulates exposed in resisting the wear loss of the composite during sliding on the steel disc. The pull out of the hard particulates leds to abrasion on the composite surface causing in the plastic deformation of particulates. It reveals that the addition of Al2O3 particulates led to minimum wear loss in mono composite compared to monolithic. Figure 8(c) indicates the SEM image of the Al7075 + 10% Al2O3 + 5%SiC composite. Here, wear resistance increases due to presence of Al2O3 and SiC particulates. Image shows that hybrid composite has much rougher surface compared to mono composite and monolithic. The composite exhibits large deep grooves and cavities on the wear surfaces. The wear resistance is high in case of high wt. % of reinforcement. Wear mechanism has been studied for the effect of adding of hard ceramic particulates which gives many causes for the high wear resistance of hybrid MMCs. Development of Mechanical Mixed Layer (M M L) when the temperature raises, due to the hard reinforcement of SiCp undergoes a chemical interactions at the time of sliding and generally, which acts as a lubricant, particularly at high sliding speeds. The protection provided by MML is witnessed to increase by increasing the reinforcement content. Similar outcomes were observed by other researchers [32–35].
EDS (Energy Dispersive Spectroscopy) study of monolithic, mono-composite and hybrid MMCs is shown in Fig. 9. EDS study made on the wornout surface of mono composite shown in Fig. 9(a), revealed the presence of oxygen (“O” peak) due to the presence of an oxidized layer, which suggests the existence of Al2O3 (aluminum oxide) content in the composite. In hybrid composite (Fig. 9(b)), in addition to the mono reinforcement, the existence of “Si” peak was also observed. This shows the presence of SiC particles in the hybrid composite. These carbide particulates greatly influence the wear behavior of composite, since in the presence of hard particulates, the wear loss is reduced compared to mono composite and monolithic. Similar outcomes have been reported by other researchers [36–39].