The experiments were carried out in order to investigate the effects of different turning parameters and flow rates on chip shape and surface integrity. The arithmetic average surface roughness \({(R}_{a})\), the maximum height of the profile\(\left({R}_{t}\right)\), the axial surface residual stress \({(\sigma }_{a})\), and the hoop surface residual stress \({(\sigma }_{h})\) were measured to examine the interactions among the turning parameters including cutting speed, feed rate, and depth of cut and MQL with three flow rates. Each measurement was carried out twice and the values were averaged in order to obtain more accurate results.
3.1. Interaction between MQL flow rate and cutting speed
Test Nos. 1, 3, and 7, with a fixed feed rate of 0.19 \(mm/rev\) and a fixed depth of cut 1.5 \(mm\), were carried out to study the influence of various flow rates of MQL and cutting speed on chip shape and surface integrity. Table 3 contains a comparison of chip shapes for various flow rates and cutting speed. According to ISO 3685 − 1977, the chips produced in metal machining processes are classified based on their sizes and shapes [18]. As seen in this table, short washer-type helical chips were produced at the cutting speeds of 145 and 650 \(m/min\) (LST) for all the MQL flow rates, whereas snarled ribbon chips were generated at the cutting speed of 1155 \(m/min\) (HST). Hence, it is observed that chip shape changed from LST to HST, while it remained unchanged for all the flow rates. The latter is in agreement with the results found by Davim et al. [13], where chip shape produced by turning of brasses did not change with MQL flow rate.
Figure 4(a) shows that the cutting speed of 145 \(m/min\) provided the lowest \({\sigma }_{a}\) for all the flow rates. Furthermore, the largest value of \({\sigma }_{a}\), equal to 120 \(MPa\), was found at the cutting speed of 1155 \(m/min\) in MQL1 mode, whereas the lowest value of \({\sigma }_{a}\), equal to 64 \(MPa\), was obtained at 145 \(m/min\) for MQL2 mode. Therefore, turning with MQL2 at the cutting speed of 145 \(m/min\) is recommended to obtain small values of \({\sigma }_{a}\).
Figure 4(b) displays that the cutting speed of 145 \(m/min\) provided the lowest \({\sigma }_{h}\) for all the flow rates, similar to \({\sigma }_{a}\). Moreover, the highest value of \({\sigma }_{h}\), equal to 161 \(MPa\), was achieved at the cutting speed of 1155 \(m/min\) in MQL1 mode, whereas by contrast the smallest value of \({\sigma }_{h}\) around 38 \(MPa\) was obtained at 145 \(m/min\) for MQL1 and MQL3 modes. As a result, MQL1 and MQL3 turning at 145 \(m/min\) are suggested to capture small values of \({\sigma }_{h}\).
It is worth mentioning that the dissimilar behavior of \({\sigma }_{a}\) for the three flow rates can be attributed to interactions among machining parameters [19, 20]. This is known as interaction effects, when the influence of one variable (machining parameter) on the response (\({\sigma }_{a}\)) depends on the level of another variable(s) [20, 21]. The interaction effect can be seen in Fig. 4(a), in which the dashed lines are not parallel.
In a turning process, a rise in cutting speed increases both the plastic work and the frictional work [22, 23], resulting in higher amounts of the generated heat, and therefore, an increase in temperature and residual stresses [19, 24]. In contrast, an increment in cutting speed increases material removal rate (MRR), which augments the heat evacuation, and as a result, can decrease temperature [25] and residual stresses [19, 22–24]. The competition between these two mechanisms reveals the level of residual stresses. For the first range of cutting speed, the first mechanism took place for all the examined flow rates, where higher residual stresses were measured with increasing cutting speed. In the second range of cutting speed, the two phenomena had almost the same contributions to the generation of residual stresses for MQL2 and MQL3, while the first phenomenon dominated in MQL1.
Figure 5(a) illustrates that the cutting speed of 650 \(m/min\) provided the lowest \({R}_{a}\) for MQL3 lubrication mode. Moreover, the values of \({R}_{a}\) were almost the same for all the flow rates at the highest cutting speed. Therefore, flow rate had almost no effect on \({R}_{a}\) at HST. In addition, the largest value of \({R}_{a}\) , equal to 1.653 \(\mu m\), was observed at the cutting speed of 145 \(m/min\) in MQL2. Consequently, MQL turning with the flow rate of 15 \(ml/min\) and the cutting speed of 650 \(m/min\) is recommended to reach small values of \({R}_{a}\).
Figure 5(b) portrays that the values of \({R}_{t}\) were almost the same for all the flow rates at the highest cutting speed. As a result, similar to \({R}_{a}\), flow rate did not affect \({R}_{t}\) at HST. It can therefore be concluded that in HST, flow rate had no impact on surface roughness parameters. In addition, the highest value of \({R}_{t}\), equal to 6.869\(\mu m\), was obtained at the cutting speed of 145 \(m/min\) in MQL2 mode, while the smallest value of \({R}_{t}\), equal to 5.008 \(\mu m,\) was obtained at 650 \(m/min\) for MQL3. Therefore, similar to \({R}_{a}\), MQL turning with the flow rate of 15 \(ml/min\) and the cutting speed of 650 \(m/min\) is recommended to reduce \({R}_{t}\). As mentioned earlier, the interaction effects between cutting speed and flow rate are clearly observed.
3.2. Interaction between MQL flow rate and feed rate
The influence of lubrication flow rates and feed rate on surface integrity was analyzed for Test Nos. 1, 2, and 4, whose cutting speed and depth of cut were kept fixed at 650 \(m/min\) and 1.5 \(mm\), respectively. The changes in chip shape with flow rate and feed rate are presented in Table 4. As observed in this table, for all the three flow rates, snarled ribbon chips were produced at the feed rate of 0.07 \(mm/rev\), short washer-type helical chips formed at the feed rate of 0.19 \(mm/rev\), and connected arc chips were generated at the feed rate of 0.31 \(mm/rev\). It is therefore concluded that chip shape changed with feed rate, but remained the same with the variation of flow rate. These results are in agreement with the results reported in [13] for turning of brass.
Figure 6(a) shows that the highest value of \({\sigma }_{a}\), around 166 \(MPa,\) occurred at the feed rate of 0.31 \(mm/rev\) in MQL3 mode, while the smallest value of \({\sigma }_{a}\), equal to -20 \(MPa\), took place at the feed rate of 0.07 \(mm/rev\) in MQL1 mode. Therefore, MQL turning with the flow rate of 3.5 \(ml/min\) at the feed rate of 0.07 \(mm/rev\) is proposed to obtain small values of \({\sigma }_{a}\).
Figure 6(b) displays that the largest \({\sigma }_{h}\), equal to 153 \(MPa\), occurred at the feed rate of 0.19 \(mm/rev\) in MQL3 mode, whereas the smallest \({\sigma }_{h}\) around 35 \(MPa\) was obtained at the lowest feed rate, equal to 0.07 \(mm/rev\), for MQL1 mode. Thus, MQL turning with the flow rate of 3.5 \(ml/min\) at the feed rate of 0.07 \(mm/rev\) is suggested to obtain small values of \({\sigma }_{h}\).
The above findings could be analyzed in terms of heat generated at tool-chip contact surface interactions under different machining parameters. In fact, in a turning process, a rise in feed rate raises the tool-chip-workpiece contact surface and the frictional heat, which augments temperature [26] and residual stresses [22, 27]. In contrast, an increment in feed rate increases MRR, which depletes the heat with the sliding chip from the cutting zone, and thus, can diminish temperature and residual stresses [1, 19]. As a result, in a metal cutting operation, the level of residual stresses is dependent considerably on the contest between the two mechanisms [23]. For the interval of feed rate under study, the first mechanism dominated, in which ASRS rose with feed rate. Likewise, Capello [28] and Leppert and Peng [29] found that axial residual stresses rose with feed rate in turning of steels. For HSRS, the two mechanisms had different contributions to the generation of residual stresses in the two intervals of feed rate, leading to different variations of HSRS with flow rate.
Figures 7(a) and (b) portray that \({R}_{a}\) and \({R}_{t}\) were highly dependent on feed rate and rose with it for all the flow rates. Furthermore, the roughness values were slightly affected by the MQL flow rates. Consequently, the lowest values of \({R}_{a}\) and \({R}_{t}\) were achieved at the smallest value of feed rate for all the flow rates. This can be attributed to the fact that smaller feed rates generate less distinct feed marks at the machined surface and lower machining forces, resulting in smaller values of surface roughness parameters [1, 30].
3.3. Interaction between MQL flow rate and depth of cut
The impact of MQL flow rate and depth of cut on surface integrity was analyzed for Test Nos. 1, 5, and 6, whose cutting speed and feed rate were selected as 650 \(m/min\) and 0.19 \(mm/rev\), respectively. Table 5 displays the plots of the variation of chip shape with flow rate and depth of cut. As seen in this table, long washer-type helical chips were produced at the depth of cut of 0.66 \(mm\) for all the MQL flow rates, short washer-type helical chips formed at the depth of cut of 1.5 \(mm\) for all the flow rates, and snarled tubular chips were produced at the depth of cut of 2.34 \(mm\) for all the flow rates. As a result, it is observed that chip shape changed with depth of cut, but remained unaffected with flow rate. Similar results were reported by [13] for turning of brass.
Figure 8(a) illustrates that the largest \({\sigma }_{a}\) equal to 120 \(MPa\) was achieved at the depth of cuts of 1.5 in MQL3 mode, while the lowest \({\sigma }_{a}\), equal to 54 \(MPa\), was obtained at the depth of cut of 0.66 \(mm\) for MQL1 mode. As a result, MQL turning at the flow rate of 3.5 \(ml/min\) and the depth of cut of 0.66\(mm\) is proposed to capture low values of \({\sigma }_{a}\).
Figure 8(b) displays that the largest \({\sigma }_{h}\), equal to 160 \(MPa\), was obtained at the highest depth of cut equal to 2.34 \(mm\) in MQL2 mode, while the smallest \({\sigma }_{h}\) equal to 77 \(MPa\) was achieved at the depth of cut of 0.66 for MQL2 mode. Therefore, MQL turning at the flow rate of 10 \(ml/min\) and the depth of cut of 0.66 \(mm\) is suggested to achieve low values of \({\sigma }_{h}\).
In a turning process, an increment in depth of cut augments the tool-chip contact area and the frictional heat in the cutting zone, which raises temperature and residual stresses [23]. Arunachalam et al. [31] reported that an increase in the depth of cut increased the tensile character of residual stresses. In contrast, increasing depth of cut raises the heat evacuation due to augmenting MRR, and as a result, can decrease temperature [32–34] and residual stresses [1, 28]. As earlier stated, the value of residual stresses changes significantly based on this contest between these two mechanisms.
Figure 9(a) illustrates the largest value of \({R}_{a}\), equal to 1.528 \(\mu m\), was observed at the depth of cut of 2.34 \(mm\) in MQL1 mode, whereas the lowest value of \({R}_{a}\), equal to 1.188 \(\mu m\), was obtained at the depth of cut of 0.66 \(mm\) for MQL2 mode. Consequently, MQL turning at the depth of cut 0.66 \(mm\) with the flow rate of 10 \(ml/min\) is proposed to obtain small values of \({R}_{a}\).
Figure 9(b) shows that the highest value of \({R}_{t}\), equal to 6.578 \(\mu m\), was obtained at the depth of cut of 2.34 \(mm\) in MQL1 mode, while the smallest value of \({R}_{t}\), equal to 5.008 \(\mu m\), was seen at the depth of cut of 1.5 \(mm\) for MQL3 mode. Thus, MQL turning at depth of cut 1.5 \(mm\) with the flow rate of 15 \(ml/min\) is recommended to reach small values of \({R}_{t}\). Similar to the previous results of roughness parameters, the interaction effects between depth of cut and flow rate are obviously seen.