An experimental investigation of wire breakage and performance optimisation of WEDM process on machining of recycled aluminium alloy metal matrix composite

The CRITIC weighting method deals with the interdependence between the criteria. The CRITIC method is more appropriate for weighing up the weights of both conventional and modern performance measures, and it comprises all the information in the assessment criteria. Moreover, the SAW method is used to compute the index score.

The weights of the criteria play an essential role in deciding the actual degree of a criterion's control. In describing the output performance of WEDM, the indicator with the maximum weight is considered as the most significant indicator, and against this conceptual background, this research used the CRITIC weighting method to establish the weights of the MR and surface roughness by using the following steps:

Step 1: The normalisation of the decision matrix is represented in Eq. (1), where Z_pq stands for the observation of MR and surface roughness: (1) Zpq¯=Zpq−ZpqminZpqmax−Zpqmin(or)Zq¯=Zpq−ZpqminZqmax−Zqmin \matrix{ \hfill {\overline {Zpq}\, = {{Zpq - {Zpq^{\min }}} \over {{Zpq^{\max }} - {Zpq^{\min }}}}} \cr \hfill {\left( {{\rm{or}}} \right)\overline {Zq}\, = {{Zpq - {Zpq^{\min }}} \over {{Zq^{\max }} - {Zq^{\min }}}}\;\;\,} \cr } (2) Zpq¯=Zpqmax−ZpqZpqmax−Zpqmin(or)Zq¯=Zpqmax−ZpqZpqmax−Zpqmin \matrix{ \hfill {\overline {Zpq}\, = {{{Zpq^{\max }} - Zpq} \over {{Zpq^{\max }} - {Zpq^{\min }}}}} \cr \hfill {\left( {{\rm{or}}} \right)\overline {Zq}\, = {{{Zpq^{\max }} - Zpq} \over {{Zpq^{\max }} - {Zpq^{\min }}}}} \cr }

Step 2: Find the standard deviation σ_q for indicator q. (3) σq=∑(zp−β)2K \sigma q = \sqrt {\sum {{{{{\left( {zp - \beta } \right)}^2}} \over K}} }

Step 3: Find the symmetrical matrix k × k with the element s_qr, which represents the linear correlation coefficient between the vectors z_q and z_r, respectively. (4) sqr=k∑n=1kqnrn−∑n=1kqn∑n=1krnA−B sqr = {{k\sum\nolimits_{n = 1}^k {qnrn} - \sum\nolimits_{n = 1}^k {qn} \sum\nolimits_{n = 1}^k {rn} } \over {A - B}} where A=k∑n=1ksn2−(∑n=1ksn)2B=k∑n=1krn2−(∑n=1krn)2 \matrix{ {A = \sqrt {k\sum\nolimits_{n = 1}^k {s{n^2}} - {{\left( {\sum\nolimits_{n = 1}^k {sn} } \right)}^2}} } \hfill \cr {B = \sqrt {k\sum\nolimits_{n = 1}^k {r{n^2}} - {{\left( {\sum\nolimits_{n = 1}^k {rn} } \right)}^2}} } \hfill \cr }

Step 4: Next, evaluate the contradiction that criterion q creates within the context of the decision scenario described by the remaining criterion, which is evaluated by the formula represented below in Eq. (5). (5) ∑r=1j(1−Sqr) \sum\nolimits_{r = 1}^j {\left( {1 - {S_{qr}}} \right)} (6) dq=σq(*)∑r=1j(1−Sqr) {d_q} = {\sigma _q}\left( * \right)\sum\nolimits_{r = 1}^j {\left( {1 - {S_{qr}}} \right)} With regard to each criterion, determine the quantity of information as well.

Finally, the following expression in Eq. (7) below represents the objective weight for indicator q: (7) wqobj=dq∑r=1jdq w_q^{obj} = {{{d_q}} \over {\sum\nolimits_{r = 1}^j {{d_q}} }} where wqobj w_q^{obj} refers to the objective indicator for the q^th indicator.

To optimise the MR and surface roughness, this paper proposes to use the SAW method, as given in Eq. (8): (8) Bp=∑q=1jwq(zpq)¯ {B_p} = \sum\nolimits_{q = 1}^j {{w_q}\overline {\left( {{z_{pq}}} \right)} } where w_q refers to the weighted value for the indicator q and zpq¯ \overline {zpq} refers to the normalised values of MR and surface roughness.

The greatest value B_p denotes the highest ranking for MR and surface roughness.

Figure 3 shows the effect of voltage on time for the first incidence of wire breakage while machining the slot in MMC. It is evident from the graph that at the parameter level of 30 V, 7 mm/min, 30 A, 120 μs (ON_T) and 70 μs, the first incidence of wire breakage has occurred at 304 s, before the completion of the slot. At this instance, the slot dimension for the height is measured using a Vernier height gauge and marked height of 9.4 mm. It is clear that there is a shortage in the height by - 0.6 mm, thereby not allowing the completion of the slot. At a lower voltage level, there is a greater availability in the pulse on-current for machining, and accordingly the rapid movement of material is introduced in order to ensure slot completion. Increasing the voltage from 40 V to 70 V increases the production of gas bubbles, together with the bowing effect of the wire triggering the breaking of the wire. Moreover, in the WEDM process, the wire gap condition is also an important factor that decides the efficiency of machining. During the WEDM process, the presence of reinforcement varies the gap based on the agglomeration of reinforcement. During the higher voltage levels, the presence of clusters of reinforcements can be attributed to bulk removal material and sharp reinforcement existing in the delaminated areas, and, consequent to exposure to wire, these are the areas that are the first to initiate the damage; and as a result of this effect, the wire electrode experiences breakage for the higher voltage levels. The graph trend of Figure 3 shows that at 30 V, it took a long time for the breakage of wire to occur, and with an increase in the voltage level, a frequent failure in wire was noticed. Figures 4A and 4B show the SEM images of the machined surface of the completed slot. The complete machined surface shows a few occurrences of the Al₂O₃ and re-solidified areas. The sparse presence of Al₂O₃ prevents any encounter between the material used as reinforcement and the wire, thereby facilitating the completion of the slot.

Fig. 3

Fig. 4

Figures 5A and 5B show the SEM picture of wire before and after machining, and it is evident that no significant changes are observable, except for a small quantity of erosion and re-solidified surfaces. The condition of the wire confirms efficient machining for the parametric combination of 30 V, 7 mm/min, 30 A, 120 μs (ON_T) and 70 μs.

Fig. 5

The primary parameter that significantly influences the WEDM process is the wire feed rate. The effect of wire feed rate on the machinability of a slot in AMMC is shown in Figure 6. As can be clearly observed from the graph, we have confirmation that usage of the discussed experimental combination of input parameters—i.e., 30 V, 3 mm/min, 30 A, 120 μs (ON_T) and 70 μs—facilitates the completion of the slot without breakage of the wire, and the time taken to machine the slot is found to be 435 s. It is evident from the graph that the lower wire feed rate ensures the completion of machining of the slot without wire breakage. During the lower wire feed rate, high current passes through the wire, which allows the creation of sparks when it gets close to the conductive workpiece. Each spark is prolonged for few seconds, resulting in excess temperature, which is high enough to melt and evaporate the workpiece material. The higher electrical parameters and lower wire feed rate contributed to non-breakage of the wire. From this graph, it is evident that the wire breakage frequency increases with an increase in wire feed (F_w). The increase in wire feed rate from 3 mm/min to 4 mm/min shows a notable change in the wire breakage, and a further increase from 4 mm/min to 7 mm/min increases the frequency of the wire breakage. Figures 7A and 7B show the SEM image of the eroded and broken wire of the top and bottom portions, at the experimental condition of 70 V, 7 mm/min, 30 A, 120 μs and 70 μs. It is clear from the figure that the erosion and breakage of the wire is not instantaneous, and that a cone shape is formed on either portion of the wire. It is noted that the centre portion of the wire shows no evidence of a melted zone and occurrence of breakage is envisaged due to weakening of mechanical properties. The presence of protruding sharp-edged Al₂O₃ on the workpiece surface after melting of matrix material creates the shearing effect on the wire.

Fig. 6

Fig. 7

In WEDM, zinc-coated wire electrode occupies a considerable percentage of the machining cost. Therefore, to attain stable machining without wire breakage, it is required to set a low wire feed rate and a higher level of electrical parameters.

Figure 8A shows the SEM picture of the machined surface at the parametric combination of 70 V, 3 mm/min, 30 A, 120 μs (ON_T) and 70 μs. The higher-level values of voltage, current and pulse on/off time create the high-density sparks that in turn create craters and pits. Figure 8B shows the rapidly melted and solidified regions on the surface of the slot. Hence, the low feed rate ensures stable machining.

Fig. 8

Figure 9 demonstrates that if we consider the applicable wire breakage as the one corresponding to the lowest value of peak current (30 A), 727 s is obtained as the time taken to machine the slot up to a depth of 9.8 mm. Increase in peak current increases the rate of heat energy. Melting and vaporisation of the wire occur at a rapid rate during the continuous increment of peak current, thereby leading to wire breakage during machining. The peak current governs the maximum amount of amperage for machining the workpiece. Roughing operations are possible with the flow of high current, but they may lead to the creation of cavities. Continuous increment of peak current improves MR but reduces the surface roughness. In AMMC, the continuous increment of peak current leads to increasing frequency of wire breakage.

Fig. 9

Figure 10 indicates the effect of pulse on time on the wire breakage during the machining of MMCs. The two slots were successfully completed without wire breakage at 1,115 s and 847 s. Further increase in the pulse on time results in wire breakage. It is evident from the graph that no wire breakage has occurred at lower pulse on time; and with increase in the pulse on time, the frequency of breakage increases. It is due to the fact that increase in pulse on time increases the discharge rate. The high frequency discharge increases the erosion and breaking of the wire. Another important factor for the cause of wire breakage is short circuit. Short circuits take place due to longer pulse on time, and the material removed from the work-piece creates the conductive bridge, resulting in unwanted sparking. Hence, an increased frequency of sparking contributes towards frequent breaking of wires.

Fig. 10

On comparing the other parameter, the pulse off time (OFF_T) has not attained complete machining, as shown in Figure 11. In AMMCs workpiece, reinforcements are used to increase the strength of the material, and this amalgamation of particles causes clustering of particles and weak matrix-reinforcement bonding. Therefore, when a proper gap is not maintained during wire travel, this results in frequent wire breakages owing to the protrusion of reinforcement particles and the resultant sharp corners. SEM image of the incomplete slot machined at the parameter combination of 70 V, 7 mm/min, 310 A, 120 μs (ON_T) and 50 μs is shown in Figure 12, and the time for the first wire breakage is found to be 120 μs (ON_T). The presence of protruded Al₂O₃ is noticed on the wire travel zone.

Fig. 11

Fig. 12

As per Eqs (1)–(8), the weights for the criteria (MR and R_a) were calculated as 0.4549 and 0.5450, respectively, and are presented in Table 7. Table 8 shows the normalised values of MR and R_a determined based on Eqs (1)–(8). Based on the CRITIC and SAW methods, the parameter combination of 30 V, 7 mm/min, 30 A, 120 μs (ON_T) and 70 μs (OFF_T) is the first ranked for higher MR and lower R_a. Additionally, 30 V, 7 mm/min, 20 A, 110 μs (ON_T) and 50 μs (OFF_T) is the second ranked combination. The results of the ANOVA are represented in Table 9. The ANOVA for MR and R_a shows that wire feed rate and voltage contribute 47.82% and 21.23%, respectively, for MR; and pulse on time shows 23.06% for surface roughness.