Assessing the Friction and Wear Behavior of AZ91-Based Hybrid Composites Reinforced with Nano hBN/Micron TiB 2 Ceramic Particles Using WASPAS and ARAS Techniques †

Presented at the International Conference on Recent Advances in Science and Engineering


Introduction
AZ91 is a magnesium alloy consisting primarily of aluminum (about 9%) and zinc (about 1%).The alloy is widely known for its lightweight properties, good mechanical strength, and low cost compared to other materials like aluminum alloys.AZ91-based composite materials combine the AZ91 alloy with other reinforcing materials to enhance specific properties for various applications [1].AZ91-based composites offer several advantages besides challenges to address, such as the susceptibility of magnesium alloys to corrosion and the difficulty of processing these materials.The applications of AZ91 magnesium alloy highlight its versatility in various industries, particularly where a balance between lightweight construction, strength, and corrosion resistance is desired in biomedical [2], aerospace, automobile, and railway [3] applications.
The friction and wear behavior of AZ91-based composites is a critical aspect to consider when evaluating their suitability for specific applications [4].The incorporation of reinforcing materials into the AZ91 matrix can significantly influence the friction, wear resistance, and overall tribological properties of the composite [5].The friction behavior of AZ91-based composites is influenced by several factors, including the type and Eng.Proc.2024, 59, 156 2 of 12 amount of reinforcement, the surface roughness of the materials in contact, and the operating conditions.In general, the addition of certain reinforcing materials can lead to reduced friction due to improved surface hardness, enhanced lubrication, and altered surface interactions [6].
Composites containing solid lubricants tend to exhibit lower wear rates due to reduced friction and improved lubrication at the sliding interface.Incorporating solid lubricants like graphite [7], MoS 2 (molybdenum disulfide), or PTFE (polytetrafluoroethylene) can reduce friction by providing a lubricating effect at the interface [8].Hard reinforcements such as ceramic particles [9,10] or fibers can promote abrasive wear on the counterface [11], potentially increasing friction [12].Composites reinforced with hard particles or fibers can enhance wear resistance by acting as load-bearing components and forming a protective layer that resists wear.However, in cases where the reinforcement is less well-bonded to the matrix, abrasive wear can be exacerbated.The wear behavior of AZ91-based composites is closely related to their composition, microstructure [13], and the specific wear mechanisms [14].Some reinforcements may improve the composite's thermal stability, reducing the likelihood of wear due to excessive temperature rise during friction.The development of coating on magnesium alloy plays a crucial role in enhancing the wear resistance [15].
Hybridizing AZ91 composites involves incorporating multiple types of reinforcing materials to take advantage of their combined benefits and improve wear resistance.Graphene and Carbon Nanotubes (CNTs) [16,17] improve wear resistance by forming a strong interface with the matrix, enhancing the composite's mechanical properties, and reducing friction [18].The choice of reinforcements depends on the specific wear mechanisms expected in the application and the desired properties of the composite [19].Hybridizing AZ91 composites allows engineers to tailor the material's properties to meet the requirements of a particular application, such as automotive components, aerospace parts, industrial machinery, or medical devices.
Multi-Criteria Decision-Making (MCDM) techniques play a crucial role in various fields and industries where complex decisions need to be made considering multiple conflicting factors.MCDM techniques provide a systematic way to consider multiple factors when making decisions related to friction and wear analysis.Analytic Hierarchy Process (AHP), Analytic Network Process (ANP), TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution), Promethee (Preference Ranking Organization Method for Enrichment Evaluation), ELECTRE (Elimination and Choice Translating Reality), Vikor, and Fuzzy Logic-Based Approaches are few of the frequently used MCDM techniques.The Attribute-Ranking with Additive Ratio Assessment (ARAS) is used for evaluating and ranking alternatives based on multiple criteria.ARAS is particularly designed to handle qualitative and quantitative criteria with different measurement scales [20].ARAS combines elements of the Analytic Hierarchy Process (AHP) and the Ratio System methods.The Weighted Aggregated Sum Product Assessment (WASPAS) is a Multi-Criteria Decision-Making (MCDM) technique used for evaluating and ranking alternatives based on multiple criteria [21].It is particularly effective when dealing with qualitative and quantitative criteria of different units or scales.
It was obvious that there are very few contributions in optimizing the tribological characteristics of AZ91-based hybrid composites with respect to their operating conditions using ARAS and WASPAS Multi-Criteria Decision-Making methods.The main objective of the present research is to optimize the amount of hBN nano particles and operating conditions such as normal load, sliding speed, and sliding distance against the response coefficient of friction, and wear loss using ARAS and WASPAS methods.Hexagonal boron nitride (hBN) is often used as a solid lubricant due to its unique properties and structure.The hypothesis of the research is that the presence of solid lubricant as an inherent part of the material system would enhance the resistance to friction and wear of the material.The wear mechanisms of hybrid composites in optimized conditions are discussed using scanning electron microscope results [22][23][24][25].

Materials and Methods
AZ91 magnesium alloy was used as a matrix to prepare composite.The composition of AZ91 magnesium alloy is given in Table 1.The composites were initially prepared using micron-sized TiB 2 particles.Based on a pilot study, it was identified that 6 wt.% TiB 2 offered optimum mechanical properties.Hence, in the present study, hybridization with hBN nano particles was carried out by keeping 6 wt.% of TiB 2 as constant content in all the composites.The nano-sized hexagonal boron nitride in multilayer platelet structure as shown in Figure 1 was used to hybridize the composite.the material.The wear mechanisms of hybrid composites in optimized conditions are discussed using scanning electron microscope results [22][23][24][25].

Materials and Methods
AZ91 magnesium alloy was used as a matrix to prepare composite.The composition of AZ91 magnesium alloy is given in Table 1.The composites were initially prepared using micron-sized TiB2 particles.Based on a pilot study, it was identified that 6 wt.% TiB2 offered optimum mechanical properties.Hence, in the present study, hybridization with hBN nano particles was carried out by keeping 6 wt.% of TiB2 as constant content in all the composites.The nano-sized hexagonal boron nitride in multilayer platelet structure as shown in Figure 1 was used to hybridize the composite.AZ91 billets were cut into pieces and placed in the furnace.They were gradually heated to 1000 °C at a controlled rate of 10 °C per minute.The temperature and duration of liquification play a crucial role in atom diffusion, facilitating the creation of a singular solid solution.It is important to strike a balance, as excessively high temperatures and prolonged liquification could lead to the undesirable formation of harmful compounds.To find this balance, the liquification temperature was upheld for a duration of 30 min.Subsequently, the molten metals were meticulously blended to achieve a uniform amalgamation.The TiB2 particles of 6 wt.% were added to the melt and stirred using a mechanical graphite stirrer for a duration of 10 min.Later, the nano-sized hBN particles were incorporated in the melt.An ultrasonic probe was used to distribute the nano particles uniformly in the melt.The pulsation of ultrasonication was carried out for a duration of 15 min.Finally, the melt was poured into the preheated die and solidified.The microstructure of the prepared composite is shown in Figure 2. Figure 2 shows the equiaxed grain structure with the reinforcements embedded in the grain boundaries, represented by white spots.AZ91 billets were cut into pieces and placed in the furnace.They were gradually heated to 1000 • C at a controlled rate of 10 • C per minute.The temperature and duration of liquification play a crucial role in atom diffusion, facilitating the creation of a singular solid solution.It is important to strike a balance, as excessively high temperatures and prolonged liquification could lead to the undesirable formation of harmful compounds.To find this balance, the liquification temperature was upheld for a duration of 30 min.Subsequently, the molten metals were meticulously blended to achieve a uniform amalgamation.The TiB 2 particles of 6 wt.% were added to the melt and stirred using a mechanical graphite stirrer for a duration of 10 min.Later, the nano-sized hBN particles were incorporated in the melt.An ultrasonic probe was used to distribute the nano particles uniformly in the melt.The pulsation of ultrasonication was carried out for a duration of 15 min.Finally, the melt was poured into the preheated die and solidified.The microstructure of the prepared composite is shown in Figure 2. Figure 2 shows the equiaxed grain structure with the reinforcements embedded in the grain boundaries, represented by white spots.
Taguchi's L16 orthogonal array was used to develop the design of the experiment for the factors and levels considered in the study.The details of factors and levels are shown in Table 2.The range of levels was chosen based on the response of the material in the pilot study conducted.The developed design of the experiment is shown in Table 3. Taguchi's L16 orthogonal array was used to develop the design of the experiment for the factors and levels considered in the study.The details of factors and levels are shown in Table 2.The range of levels was chosen based on the response of the material in the pilot study conducted.The developed design of the experiment is shown in Table 3.
Nano hBN (Wt.%) 0, 0.5, 1, 1.5 The pin-on-disc wear test was carried out as per the plan of experimentation in Table 3 following standard ASTM G99.A tribometer made by DUCOM instruments was used  The pin-on-disc wear test was carried out as per the plan of experimentation in Table 3 following standard ASTM G99.A tribometer made by DUCOM instruments was used to conduct the tests.The tests were conducted in dry conditions and at room temperature.EN31 high carbon steel was employed as a counterface during the wear process.The surface roughness of the mating surfaces was maintained below 10 µm.The friction force of the experiments was recorded using a load cell attached to the test equipment.The mass loss of the samples was measured using an electronic weighting machine with 0.01 mg resolution.

Results and Discussion
3.1.Effect of Normal Load, Sliding Speed, Sliding Distance, and hBN wt.% on Wear Loss and COF The results of the experiments conducted as per the design of the experiments are given in Table 4.The minimum value of wear loss was recorded in the experiment run 8 (normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, hBN (wt.%)-0.5),whereas the lowest value of the coefficient of friction was noted for the experiment run 2 (normal load-5 N, sliding speed-1 m/s, sliding distance-750 m, hBN (wt.%)-0.5).The maximum value of wear loss and COF were noticed in the experiment run 12 (normal load-25 N, sliding speed-3 m/s, sliding distance-750 m, hBN (wt.%)-0) and [16 (normal load-30 N, sliding speed-3 m/s, sliding distance-500 m, hBN (wt.%)-1, 7 (normal load-10 N, sliding speed-2.5 m/s, sliding distance-3000 m, hBN (wt.%)-0)] respectively.From the above observation, it is very clear that the presence of hBN up to the 0.5 wt.% in the composition benefited the material to yield minimum values of wear loss and coefficient of friction.On the other hand, the composite without hBN showed comparatively higher values.Taguchi analysis was performed on the results to identify the influence of factors on the response.The results of the signal-to-noise ratio is shown in Figure 3 and the ranking of factors is given in Table 5. Figure 3 shows that the sliding speed influences the responses followed by normal load, sliding distance, and nano hBN (wt.%) sequentially.A similar trend was also confirmed by the results of the main effect plots of the means and response table, in Figure 4

Data Means
Signal-to-noise: Smaller is better

ARAS Method
The ARAS (Additive Ratio Assessment) is a Multi-Criteria Decision-Making (MCDM) method [20], used to rank alternatives based on multiple criteria.In this case, based on the wear loss and coefficient of friction (COF) data, the experiments were ranked using the ARAS method.The criteria for ranking were taken as follows: wear loss-the lower the better, COF-the lower the better.
Step 1: Establish the decision matrix of criteria and alternatives.
A decision matrix, often referred to as a matrix of alternatives and criteria (m and n) (i × j), is utilized for encapsulating all the fields in Equation ( 1) -the ideal value of the j-th criterion-can be decided based on the desired minimum or maximum criteria of responses.In the present study, the minimum value of wear loss is 0.002, and COF is 0.15.
Step 2: Normalize the values in the decision matrix.
The values in the decision matrix are normalized using Equation (2).
xij-reciprocal of the individual values of the criteria in each experiment, ̅  .
Step 3: Calculate the weighted normalized values.
The weighted normalized values were calculated using Equation (3).The normalized values were multiplied by the weights assigned for each criterion.In the present study,

ARAS Method
The ARAS (Additive Ratio Assessment) is a Multi-Criteria Decision-Making (MCDM) method [20], used to rank alternatives based on multiple criteria.In this case, based on the wear loss and coefficient of friction (COF) data, the experiments were ranked using the ARAS method.The criteria for ranking were taken as follows: wear loss-the lower the better, COF-the lower the better.
Step 1: Establish the decision matrix of criteria and alternatives.
A decision matrix, often referred to as a matrix of alternatives and criteria (m and n) (i × j), is utilized for encapsulating all the fields in Equation ( 1) x 0j -the ideal value of the j-th criterion-can be decided based on the desired minimum or maximum criteria of responses.In the present study, the minimum value of wear loss is 0.002, and COF is 0.15.
Step 2: Normalize the values in the decision matrix.
The values in the decision matrix are normalized using Equation (2).
x ij -reciprocal of the individual values of the criteria in each experiment, x ij − normalized values.
Step 3: Calculate the weighted normalized values.
The weighted normalized values were calculated using Equation (3).The normalized values were multiplied by the weights assigned for each criterion.In the present study, wear loss and COF were equally important, as friction and wear are inter-related.Hence, equal weights were assigned for both criteria as 0.5.
Step 4: Compute the optimality function values.
The optimality function values were determined by summing the values of the criteria for each experiment as represented by Equation ( 4).

Optimality f unction, S
Step 5: Determine the utility degree of alternative.
The utility degree of alternative is mainly used to identify the experiment which offered desirable values of response based on the criteria that was chosen.The utility degree was calculated using Equation ( 5).

Utility degree o f alternative, K
S 0 -maximum value of utility degree.The maximum value of utility degree gets the first rank, which is the best experimental run.

WASPAS
The WASPAS (Weighted Aggregated Sum Product Assessment) is a Multi-Criteria Decision-Making (MCDM) technique [21], used to rank alternatives based on multiple criteria while considering weights for the criteria.The WASPAS technique was used in this study to validate the ranks obtained through the ARAS method.The weights assigned for the wear loss and COF were 0.5, maintained as used in the ARAS method.The same decision matrix established in the ARAS method was used here as well.
Step 1: Normalize the values in the decision matrix.
The values in the decision matrix were normalized using Equation (1).In the case of the current study, the wear loss and COF need to be minimum, which is desirable.The minimum-better case is considered as a non-beneficial criterion in this technique.
For non-beneficial criteria, Step 2: Calculating the weighted sum and weighted product of values.
The weighted sum and weighted product values were computed using Equations ( 7) and ( 8): w j − weights o f criteria Step 3: Determine the overall relative index.
The overall relative index values were calculated using Equation (9).The equation combines the weighted sum and weighted product values.The ranking in the WASPAS Eng.Proc.2024, 59, 156 9 of 12 method is performed using the Qi values.The higher Qi value get the first rank, which indicates the best experiment run for the responses.
The ranks obtained in the ARAS and WASPAS methods are shown in Figure 5.Both methods suggested experiment run 8 as an optimal run to offer the best combination of wear loss and COF.Experiment run 8 was conducted with normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, and hBN wt.%-0.5.The effect of hBN was further confirmed to influence the tribological characteristics of hybrid AZ91 composites to a significant extent.The optimal amount of nano hBN particles that can be added to hybridize the composite was observed as 0.5 wt.%.
The overall relative index values were calculated using Equation (9).The equation combines the weighted sum and weighted product values.The ranking in the WASPAS method is performed using the Qi values.The higher Qi value get the first rank, which indicates the best experiment run for the responses.
The ranks obtained in the ARAS and WASPAS methods are shown in Figure 5.Both methods suggested experiment run 8 as an optimal run to offer the best combination of wear loss and COF.Experiment run 8 was conducted with normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, and hBN wt.%-0.5.The effect of hBN was further confirmed to influence the tribological characteristics of hybrid AZ91 composites to a significant extent.The optimal amount of nano hBN particles that can be added to hybridize the composite was observed as 0.5 wt.%.

Wear Mechanism
The study suggested that the hybrid composites with 0.5 wt.% hBN offered enhanced tribological characteristics of the AZ91 hybrid composites.The mechanism of wear for such composites was studied by analysing the worn-out surface with a scanning electron microscope.The SEM images of worn surfaces are shown in Figure 6. Figure 6a shows the worn surface of a sample from experiment run 2 (normal load-5 N, sliding speed-1 m/s, sliding distance-750 m, and hBN (wt.%)-0.5).Fine abrasive grooves were noted on the surface.This shows that the abrasive wear was active under such mild operating conditions of normal load and sliding speed.When the normal load was increased to 10 N and the sliding speed to 3 m/s, the hBN nano particles were pulled out of composites and located themselves at the interface of the mating surfaces.The presence of the hBN acted as a solid lubricating agent and reduced the friction and wear loss of the material.The rise in sliding speed also increased the interfacial temperature and softened the matrix material.This can be witnessed from the smooth adhesive worn surface of the sample from experiment run 8 (normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, and hBN (wt.%)-0.5)as shown in Figure 6b.An increase in normal load of more than 20 N predominantly influenced the wear mechanism, which can be seen from Figure 6c,d, representing experiment run 11 (normal load-25 N, sliding speed-2.5 m/s, sliding distance-500 m, and hBN (wt.%)-0.5)and experiment run 13 (normal load-30 N, sliding speed-0.5 m/s, sliding distance-3000 m, and hBN (wt.%)-0.5),respectively.Plastic deformation and delamination of the wear surface was noted in such composites.This could

Wear Mechanism
The study suggested that the hybrid composites with 0.5 wt.% hBN offered enhanced tribological characteristics of the AZ91 hybrid composites.The mechanism of wear for such composites was studied by analysing the worn-out surface with a scanning electron microscope.The SEM images of worn surfaces are shown in Figure 6. Figure 6a shows the worn surface of a sample from experiment run 2 (normal load-5 N, sliding speed-1 m/s, sliding distance-750 m, and hBN (wt.%)-0.5).Fine abrasive grooves were noted on the surface.This shows that the abrasive wear was active under such mild operating conditions of normal load and sliding speed.When the normal load was increased to 10 N and the sliding speed to 3 m/s, the hBN nano particles were pulled out of composites and located themselves at the interface of the mating surfaces.The presence of the hBN acted as a solid lubricating agent and reduced the friction and wear loss of the material.The rise in sliding speed also increased the interfacial temperature and softened the matrix material.This can be witnessed from the smooth adhesive worn surface of the sample from experiment run 8 (normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, and hBN (wt.%)-0.5)as shown in Figure 6b.An increase in normal load of more than 20 N predominantly influenced the wear mechanism, which can be seen from Figure 6c,d, representing experiment run 11 (normal load-25 N, sliding speed-2.5 m/s, sliding distance-500 m, and hBN (wt.%)-0.5)and experiment run 13 (normal load-30 N, sliding speed-0.5 m/s, sliding distance-3000 m, and hBN (wt.%)-0.5),respectively.Plastic deformation and delamination of the wear surface was noted in such composites.This could be predicted due to the third body wear of pulled-out TiB 2 hard ceramic particles, which cut the surface of the samples.The surfaces also showed a slightly oxidative mode of wear, which can be witnessed by white spots on the surface.
be predicted due to the third body wear of pulled-out TiB2 hard ceramic particles, which cut the surface of the samples.The surfaces also showed a slightly oxidative mode of wear, which can be witnessed by white spots on the surface.

Conclusions
AZ91-based hybrid composites were fabricated using stir casting assisted by the ultrasonication method.The hard ceramic reinforcement TiB2 and solid lubricant hBN were used as fillers in the composite.The experiments were designed using Taguchi's method and pin-on-disc tests were carried out to characterize the friction and wear behavior of the hybrid composites by optimizing the hBN wt.% and operating conditions.The study revealed the following conclusions.The minimum value of wear loss was recorded in the experiment run 8 (normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, hBN (wt.%)-0.5),whereas the lowest value of the coefficient of friction was noted for experiment run 2 (normal load-5 N, sliding speed-1 m/s, sliding distance-750 m, hBN (wt.%)-0.5).The ARAS and WASPAS Multi-Criteria Decision-Making methods suggested experiment run 8 to offer the best results for the responses to wear loss and COF considered in the study.The presence of the hBN acted as a solid lubricating agent and reduced the friction and wear loss of the hybrid composite material.The hBN showed its effect efficiently up to the addition of 0.5 wt.% in the composition.Abrasive wear was noted to be active in lower operating conditions of normal load and sliding speed.Adhesive wear was found to be active in the optimal experiment run 8. Delamination and

Conclusions
AZ91-based hybrid composites were fabricated using stir casting assisted by the ultrasonication method.The hard ceramic reinforcement TiB 2 and solid lubricant hBN were used as fillers in the composite.The experiments were designed using Taguchi's method and pin-on-disc tests were carried out to characterize the friction and wear behavior of the hybrid composites by optimizing the hBN wt.% and operating conditions.The study revealed the following conclusions.The minimum value of wear loss was recorded in the experiment run 8 (normal load-10 N, sliding speed-3 m/s, sliding distance-1500 m, hBN (wt.%)-0.5),whereas the lowest value of the coefficient of friction was noted for experiment run 2 (normal load-5 N, sliding speed-1 m/s, sliding distance-750 m, hBN (wt.%)-0.5).The ARAS and WASPAS Multi-Criteria Decision-Making methods suggested experiment run 8 to offer the best results for the responses to wear loss and COF considered in the study.The presence of the hBN acted as a solid lubricating agent and reduced the friction and wear loss of the hybrid composite material.The hBN showed its effect efficiently up to the addition of 0.5 wt.% in the composition.Abrasive wear was noted to be active in lower operating conditions of normal load and sliding speed.Adhesive wear was found to be active in the optimal experiment run 8. Delamination and surface cutting due to third body wear by TiB 2 particles were noted in higher normal loads (more than 20 N) and longer sliding distances (more than 1000 m) in the current study.

Figure 1 .
Figure 1.Scanning electron microscope image of hBN nano particles.

Figure 1 .
Figure 1.Scanning electron microscope image of hBN nano particles.

Figure 3 .
Figure 3. Main effect plots for the signal-to-noise ratio of factors and levels.

Figure 3 .
Figure 3. Main effect plots for the signal-to-noise ratio of factors and levels.

Figure 4 .
Figure 4. Main effect plots for means of factors and levels.

Figure 4 .
Figure 4. Main effect plots for means of factors and levels.

Figure 5 .
Figure 5. Ranks of experiments from the ARAS and WASPAS methods.

Figure 5 .
Figure 5. Ranks of experiments from the ARAS and WASPAS methods.

Table 2 .
Details of factors and levels.

Table 3 .
Design of the experiment.

Table 2 .
Details of factors and levels.

Table 3 .
Design of the experiment.

Table 4 .
Results of wear loss and COF for the experiments.

Table 5 .
Response table for signal-to-noise ratios of factors and levels.Criteria-smaller is better.

Table 6 .
Response table for means of factors and levels.

Table 5 .
Response table for signal-to-noise ratios of factors and levels.Criteria-smaller is better.

Table 6 .
Response table for means of factors and levels.