**3. Results and Discussion**

### *3.1. Microstructure Characteristics of Friction Stir Welding (FSW) Joint*

Figure 3 shows the optical macrostructure and microstructure of the FSW specimen. Figure 3a shows the macrostructure of the cross section of the FSWed joint according to the experimental conditions. The macrostructure in Figure 3a is enlarged. The TMAZ is shown in Figure 3b and SZ in Figure 3c. The HAZ is shown in Figure 3d, the Al base material (BM) is shown in Figure 3e, and the IMC region is shown in Figure 3f, respectively. In Figure 3a, the broken line of the magnification indicates the shape of the tool pin, and the arrow spacing of the PD indicates the spacing between the outer diameter of A390 cast Al and the shoulder of the tool. The arrow spacing of the TPD indicates the gap between the tip of the tool pin and the interlayer at which the inner diameter of A390 Al and the outer diameter of FB590 HSS meet. The cause of the cavity in Figure 3a is the geometric difference between the pipe and the plate. Akbari et al. [13] found that the contact characteristics of the tool and the workpiece were different due to the small radius of curvature of the pipe. At a PD of 0.1 mm, due to the pipe curvature, the FSW shoulder partially contacts the pipe and only the inner part makes the shoulder part contact the pipe, forming a tunnel cavity. Another cause of large cavity is a lack of material. (1) The upper aluminum material under the influence of process variables is discharged to the outside in the form of flash. (2) The upper aluminum material moves to fill the gap (gap) on the junction interface of aluminum and steel. The lack of material due to the above causes created a large internal cavity.

After FSW, there is an SZ around the center of the joint. There is a TMAZ in which magnitude of grains are increased by plastic flow on the outside of the SZ, and there is an HAZ that is heat-affected but has no plastic deformation on the outside of TMAZ. These zones were observed to have a wider region width than the retreat side on the tool's advance side (AS). To form a bond as the tool is traversed, material flows around the tool in a complex flow pattern depending on the tool geometry and process parameters, such as TRS and WS. Microstructural evolution depends on the change of the welding parameters. A suitable selection of the process parameters results in good material mixing at the joint and a sound weld.

**Figure 3.** Optical micrographs of the FSW A357 cast alloy and FB590 HSS. (**a**) Overall cross-sectional macrostructure; (**b**) microstructure of the SZ; (**c**) microstructure of the TMAZ; (**d**) microstructure of the HAZ; (**e**) microstructure of the IMC; and (**f**) microstructure of the Al base metal [16]. **Figure 3.** Optical micrographs of the FSW A357 cast alloy and FB590 HSS. (**a**) Overall cross-sectional macrostructure; (**b**) microstructure of the TMAZ; (**c**) microstructure of the SZ; (**d**) microstructure of the HAZ; (**e**) microstructure of the Albase metal; and (**f**) microstructure of the IMC [16].

After FSW, there is an SZ around the center of the joint. There is a TMAZ in which magnitude of grains are increased by plastic flow on the outside of the SZ, and there is an HAZ that is heat-affected but has no plastic deformation on the outside of TMAZ. These zones were observed to have a wider region width than the retreat side on the tool's advance side (AS). To form a bond as the tool is traversed, material flows around the tool in a complex flow pattern depending on the tool geometry and process parameters, such as TRS and WS. Microstructural evolution depends on the change of the welding parameters. A suitable selection of the process parameters results in good material mixing at the joint and a sound weld. The SZ in Figure 3c produces a fine grain structure, whereas the TMAZ in Figure 3b has a long grain structure. The SZ region of Figure 3c shows finer particles compared to the TMAZ region and the HAZ region of Figure 3d. In general, the SZ region is also referred to as a "nugget region" and is believed to be formed by dynamic recrystallization. The SZ region has a wider width at the top and decreases in the thickness direction. This is because of the difference in diameter between the shoulder and the pin. In the SZ, a finer grain was formed compared to other regions. Dynamic recrystallization produced skewed and elongated grains in TMAZ and fine grains in SZ. The grain boundary orientation is different in all three regions. The size of the SZ region in Figure 3c is smaller than that of the BM, and the crystalline grains of the BM disappear. This is in agreement with the results of Mahto et al. [21], who reported that new crystals were formed and dynamic recrystallization occurred. The microstructure of TMAZ is shown in Figure 4c. Located The SZ in Figure 3c produces a fine grain structure, whereas the TMAZ in Figure 3b has a long grain structure. The SZ region of Figure 3c shows finer particles compared to the TMAZ region and the HAZ region of Figure 3d. In general, the SZ region is also referred to as a "nugget region" and is believed to be formed by dynamic recrystallization. The SZ region has a wider width at the top and decreases in the thickness direction. This is because of the difference in diameter between the shoulder and the pin. In the SZ, a finer grain was formed compared to other regions. Dynamic recrystallization produced skewed and elongated grains in TMAZ and fine grains in SZ. The grain boundary orientation is different in all three regions. The size of the SZ region in Figure 3c is smaller than that of the BM, and the crystalline grains of the BM disappear. This is in agreement with the results of Mahto et al. [21], who reported that new crystals were formed and dynamic recrystallization occurred. The microstructure of TMAZ is shown in Figure 3b. Located between HAZ and SZ, TMAZ is characterized by a highly deformed structure. It shows an elongated non-metallic particle morphology with sub-grain coarsening. The grain structure of the HAZ region, which was not mechanically perturbed by FSW, was generated by static recrystallization and was similar to the base metal in Figure 3e HAZ is much wider at the top surface in contact with the shoulder and is narrow as the pin diameter decreases. In Su et al. [62], the grain size increases in the order of SZ, TMAZ, HAZ, and BM. Therefore, the microstructure was recrystallized in the FSWed region.

between HAZ and SZ, TMAZ is characterized by a highly deformed structure. It shows an elongated non-metallic particle morphology with sub-grain coarsening. The grain

structure of the HAZ region, which was not mechanically perturbed by FSW, was generated by static recrystallization and was similar to the parent material in Figure 3e HAZ is much wider at the top surface in contact with the shoulder and is narrow as the pin diameter decreases. In Su et al. [62], the grain size increases in the order of SZ, TMAZ, HAZ,

and BM. Therefore, the microstructure was recrystallized in the FSWed region.

**Figure 4.** SEM pictures according to TPD of experiment group A. (**a**) TPD = 0.0 mm; (**b**) TPD = 0.5 mm; and (**c**) TPD = 1.0 mm. **Figure 4.** SEM pictures according to TPD of experiment group A. (**a**) TPD = 0.0 mm; (**b**) TPD = 0.5 mm; and (**c**) TPD = 1.0 mm.

#### *3.2. IMC Thickness Characteristics of FSW Joints 3.2. IMC Thickness Characteristics of FSW Joints*

In order to investigate the change in the IMC thickness due to the TPD, the PD is fixed at 0.0 mm, and changes in the remaining process parameters are considered. First, the effect on the IMC thickness during FSW according to the TPD of group A with a WS of 0.1 rpm is investigated. Figure 4 shows the change in the IMC thickness during FSW according to the TPD in group A as an SEM photograph. Figure 4a shows the IMC average thickness when the TPD is 0.0 mm in experiment no. 1, Figure 4b shows the IMC average thickness when the TPD is 0.5 mm in experiment no. 2, and Figure 4c shows the IMC average thickness at a TPD of 1.0 mm in experiment no. 3. At a TPD of 0.0 mm, the IMC average thickness is 5.1 μm; at a TPD of 0.5 mm, the IMC average thickness is 3.9 μm; and at a TPD of 1.0 mm, the IMC average thickness is 3.1 μm. It can be seen that the IMC average thickness decreases as the TPD increases. Next, the effect of the TPD on the IMC average thickness was investigated in group B with a WS of 0.15 to 0.2 rpm during FSW. Figure 5 shows the change in the IMC average thickness during FSW according to the TPD in group B as an SEM photograph. Figure 5a–c shows the IMC average thickness of experiment no. 4 with a TPD of 0.0 mm, experiment no. 5 with TPD of 0.5 mm, and an experiment no. 6 with a TPD of 1.0 mm. When the TPD is 0.0 mm, the IMC average thickness is 9.05 μm; at a TPD of 0.5 mm, the IMC average thickness is 6.43 μm; and at a TPD of 1.0 mm, the IMC average thickness is 5.07 μm. As in group A where the WS is 0.1 rpm, it can In order to investigate the change in the IMC thickness due to the TPD, the PD is fixed at 0.0 mm, and changes in the remaining process parameters are considered. First, the effect on the IMC thickness during FSW according to the TPD of group A with a WS of 0.1 rpm is investigated. Figure 4 shows the change in the IMC thickness during FSW according to the TPD in group A as an SEM photograph. Figure 4a shows the IMC average thickness when the TPD is 0.0 mm in experiment no. 1, Figure 4b shows the IMC average thickness when the TPD is 0.5 mm in experiment no. 2, and Figure 4c shows the IMC average thickness at a TPD of 1.0 mm in experiment no. 3. At a TPD of 0.0 mm, the IMC average thickness is 5.1 µm; at a TPD of 0.5 mm, the IMC average thickness is 3.9 µm; and at a TPD of 1.0 mm, the IMC average thickness is 3.1 µm. It can be seen that the IMC average thickness decreases as the TPD increases. Next, the effect of the TPD on the IMC average thickness was investigated in group B with a WS of 0.15 to 0.2 rpm during FSW. Figure 5 shows the change in the IMC average thickness during FSW according to the TPD in group B as an SEM photograph. Figure 5a–c shows the IMC average thickness of experiment no. 4 with a TPD of 0.0 mm, experiment no. 5 with TPD of 0.5 mm, and an experiment no. 6 with a TPD of 1.0 mm. When the TPD is 0.0 mm, the IMC average thickness is 9.05 µm; at a TPD of 0.5 mm, the IMC average thickness is 6.43 µm; and at a TPD of 1.0 mm, the IMC average thickness is 5.07 µm. As in group A where the WS is 0.1 rpm, it can be seen that the IMC average thickness decreases as the TPD increases.

be seen that the IMC average thickness decreases as the TPD increases. Mahto et al. [28] reported that the higher the heat input, the thicker the IMC layer was in the FSLW of AA 6061-T6 and AISI 304 stainless steel, each having a thickness of 1 mm. According to Ogura et al. [63], the nature of the IMC formed during welding is determined by the combined influence of base materials stirring and thermal conditions. As the TPD increases, the contact area between the tool and the material increases and the heat input tends to increase. Another trend is that the flow of material will increase due to agitation as the TPD increases. Therefore, it can be deduced that the thickness of the IMC

layer decreases because the effect of the material flow due to agitation is greater than the increase in the heat input due to the increase of the TPD. to agitation as the TPD increases. Therefore, it can be deduced that the thickness of the IMC layer decreases because the effect of the material flow due to agitation is greater than the increase in the heat input due to the increase of the TPD.

Mahto et al. [28] reported that the higher the heat input, the thicker the IMC layer was in the FSLW of AA 6061-T6 and AISI 304 stainless steel, each having a thickness of 1 mm. According to Ogura et al. [63], the nature of the IMC formed during welding is determined by the combined influence of base materials stirring and thermal conditions. As the TPD increases, the contact area between the tool and the material increases and the heat input tends to increase. Another trend is that the flow of material will increase due

*Materials* **2022**, *15*, x FOR PEER REVIEW 9 of 18

**Figure 5.** SEM pictures according to TPD of experiment group B. (**a**) TPD = 0.0 mm; (**b**) TPD = 0.5 mm; and (**c**) TPD = 1.0 mm. **Figure 5.** SEM pictures according to TPD of experiment group B. (**a**) TPD = 0.0 mm; (**b**) TPD = 0.5 mm; and (**c**) TPD = 1.0 mm.

#### *3.3. Characteristics of IMC Thickness and Tensile Shear Load (TSL) According to the Tool Penetration Depth (TPD) 3.3. Characteristics of IMC Thickness and Tensile Shear Load (TSL) According to the Tool Penetration Depth (TPD)*

Table 3 summarizes the IMC thickness and TSL according to the change in TPD and WSs of 0.1 rpm and 0.15 rpm or more. Table 3 summarizes the IMC thickness and TSL according to the change in TPD and WSs of 0.1 rpm and 0.15 rpm or more.


**Table 3.** IMC thickness range and IMC average thickness and TSL on TPD. **Table 3.** IMC thickness range and IMC average thickness and TSL on TPD.

Figure 6 compares the differences between the two groups with changes in the IMC average thickness according to TPD in group A with a WS of 0.1 rpm and group B with a WS of 0.15 to 0.2 rpm. In Figure 6, group A of experiments no. 1, no. 2, and no. 3 is indicated by a solid line to represent the IMC average thickness change according to the TPD and a WS of 0.1 rpm. It shows a gradual decrease in the IMC average thickness as TPD increases from 0.0 mm to 1.0 mm. In Figure 6, group B of experiments no. 4, no. 5, and no. Figure 6 compares the differences between the two groups with changes in the IMC average thickness according to TPD in group A with a WS of 0.1 rpm and group B with a WS of 0.15 to 0.2 rpm. In Figure 6, group A of experiments no. 1, no. 2, and no. 3 is indicated by a solid line to represent the IMC average thickness change according to the TPD and a WS of 0.1 rpm. It shows a gradual decrease in the IMC average thickness as TPD increases from 0.0 mm to 1.0 mm. In Figure 6, group B of experiments no. 4, no. 5, and no. 6 are indicated by broken lines to represent the IMC average thickness change at WS values of 0.15 and 0.2 rpm. It shows a gradual decrease in the IMC average thickness as the TPD increases from 0.0 mm to 1.0 mm. It can be seen that the IMC average thickness gradually decreased as TPD increased from 0.0 mm to 1.0 mm, regardless of the influence of WS and other process parameters.

WS and other process parameters.

6 are indicated by broken lines to represent the IMC average thickness change at WS values of 0.15 and 0.2 rpm. It shows a gradual decrease in the IMC average thickness as the TPD increases from 0.0 mm to 1.0 mm. It can be seen that the IMC average thickness gradually decreased as TPD increased from 0.0 mm to 1.0 mm, regardless of the influence of

**Figure 6.** IMC thickness comparison according to TPD for experiment groups A and B. **Figure 6.** IMC thickness comparison according to TPD for experiment groups A and B.

In addition, it can be seen that the IMC average thickness increases as the WS increases at the same TPD. This is consistent with the experimental results reported by Shen et al. [34]. IMC at the Fe–Al interface decreases with increasing TPD during the FSLW of 2.2-mm-thick AA5754 Al and 2.5-mm-thick DP600 dual-phase steel. However, their results did not provide quantitative results for IMC thickness increase. Mahto et al. [35] performed FSLW between 0.0 mm and 0.3 mm with a pin depth (PLD) of 0.1 mm spacing on 1-mm-thick dissimilar materials AA6061-T6 and AISI304. It is said that the IMC thickness increases as the PLD increases. This is the result of changing the TPD in the narrow range of 0–0.3 mm, and it appears to be different from the current wide region, where the TPD is changed between 0.0 and 1.0 mm. In the study reported by Liu et al. [25], it was proposed that the IMC thickness decreased with the increase of the WS and the increase of the tool offset. In this study, the influence of TPD was dominant, showing the opposite trend. From the experimental results, as the TPD increases, the material fluidity increases and the IMC thickness decreases. This is because the effect of TPD is more dominant on material flow than on heat input. Figure 7 shows graphs of the variation of the TSL with TPD. In addition, it can be seen that the IMC average thickness increases as the WS increases at the same TPD. This is consistent with the experimental results reported by Shen et al. [34]. IMC at the Fe–Al interface decreases with increasing TPD during the FSLW of 2.2-mm-thick AA5754 Al and 2.5-mm-thick DP600 dual-phase steel. However, their results did not provide quantitative results for IMC thickness increase. Mahto et al. [35] performed FSLW between 0.0 mm and 0.3 mm with a pin depth (PLD) of 0.1 mm spacing on 1-mm-thick dissimilar materials AA6061-T6 and AISI304. It is said that the IMC thickness increases as the PLD increases. This is the result of changing the TPD in the narrow range of 0–0.3 mm, and it appears to be different from the current wide region, where the TPD is changed between 0.0 and 1.0 mm. In the study reported by Liu et al. [25], it was proposed that the IMC thickness decreased with the increase of the WS and the increase of the tool offset. In this study, the influence of TPD was dominant, showing the opposite trend. From the experimental results, as the TPD increases, the material fluidity increases and the IMC thickness decreases. This is because the effect of TPD is more dominant on material flow than on heat input.

Figure 7 shows graphs of the variation of the TSL with TPD.

In Figure 7, group A of experiments no. 1, no. 2, and no. 3 is indicated by a dash-dotted line to represent the changes in the TSL with the TSL and a WS value of 0.1 rpm. As TPL increases from 0.0 mm to 1.0 mm, it can be seen that the TSL decreases up to a TPD value of 0.5 mm, and TPD increases above 0.5 mm. In Figure 7, group B of experiments no. 4, no. 5, and no. 6 are indicated by dash-double dotted lines to represent the changes in the TSL for WS values of 0.15 and 0. 2 rpm. It shows a gradual increase in the IMC average thickness as the TPD increases from 0.0 mm to 1.0 mm. It can be seen that the IMC average thickness gradually decreased as TPD increased from 0.0 mm to 1.0 mm, regardless of the influence of WS and other process parameters. As TPD increases, TSL decreases to TPD = 0.5 mm at a WS of 0.1 rpm. At the WS of 0.1 rpm, the TPD increases over 0.5 mm, and at the WS of 0.15 and 0.2 rpm, the TPD increases over the entire region. In the study of TSL according to TPD, Lee et al. [52] reported that the TSL increased as the TPD increased,

but this result is due to the difference between FSSW and this study, which has a large initial thermal effect. In order to increase the TSL according to TPD, a WS above a certain speed was not proposed. Shen et al. [34] reported that the TSL decreased and increased with increasing TPD. Contrary to the results of this study, this results from an inability to distinguish between the mutual effects of TPD and PD. but this result is due to the difference between FSSW and this study, which has a large initial thermal effect. In order to increase the TSL according to TPD, a WS above a certain speed was not proposed. Shen et al. [34] reported that the TSL decreased and increased with increasing TPD. Contrary to the results of this study, this results from an inability to distinguish between the mutual effects of TPD and PD. *Materials* **2022**, *15*, x FOR PEER REVIEW 11 of 18

**Figure 7.** Comparison of TSL with TPD for experiment groups A and B [16]. **Figure 7.** Comparison of TSL with TPD for experiment groups A and B [16]. **Figure 7.** Comparison of TSL with TPD for experiment groups A and B [16].

In Figure 7, group A of experiments no. 1, no. 2, and no. 3 is indicated by a dashdotted line to represent the changes in the TSL with the TSL and a WS value of 0.1 rpm. As TPL increases from 0.0 mm to 1.0 mm, it can be seen that the TSL decreases up to a TPD value of 0.5 mm, and TPD increases above 0.5 mm. In Figure 7, group B of experiments no. 4, no. 5, and no. 6 are indicated by dash-double dotted lines to represent the changes in the TSL for WS values of 0.15 and 0. 2 rpm. It shows a gradual increase in the IMC average thickness as the TPD increases from 0.0 mm to 1.0 mm. It can be seen that the IMC average thickness gradually decreased as TPD increased from 0.0 mm to 1.0 mm, regardless of the influence of WS and other process parameters. As TPD increases, TSL decreases to TPD = 0.5 mm at a WS of 0.1 rpm. At the WS of 0.1 rpm, the TPD increases over 0.5 mm, and at the WS of 0.15 and 0.2 rpm, the TPD increases over the entire region. In the study of TSL according to TPD, Lee et al. [52] reported that the TSL increased as the TPD increased, but this result is due to the difference between FSSW and this study, which has a large initial thermal effect. In order to increase the TSL according to TPD, a WS above a certain speed was not proposed. Shen et al. [34] reported that the TSL decreased and increased with increasing TPD. Contrary to the results of this study, this results from an inability to distinguish between the mutual effects of TPD and PD. In Figures 6 and 7, as TPD increases at a WS of 0.1 rpm, the IMC average thickness In Figures 6 and 7, as TPD increases at a WS of 0.1 rpm, the IMC average thickness decreases and TSL decreases until TPD = 0.5 mm. In Figure 8, it can be seen that the IMC average thickness decreases and the TSL increases at TPD = 0.5 mm or more as the TPD increases at the WS of 0.1 rpm. When WS varies from 0.15–0.2 rpm, the IMC average thickness decreases and TSL increases as TPD increases in all regions. Picot et al. [54] used the process parameters of the rotational speed and translational speed of the tool in FSLW of stainless steel 316L and Al alloy 5083 and reported that the TSL decreased with increasing IMC average thickness. Helal et al. [55] used the process parameters of TRS, WS, and tool offset in FSLW of 6061-T6 Al alloy and ultra-low carbon steel and reported that as the IMC average thickness increases, TSL decreases. Kimapong et al. [53] reported that TSL decreased with increasing IMC average thickness using welding parameters such as rotational speed, traversing speed of 0.4 to 1.43 mm/s, and a pin depth in FSLW of A5083 Al alloy and SS400 steel. In previous studies, it was reported that the TSL increased with decreasing IMC average thickness. As WS varies from 0.15–0.2 rpm, it was consistent with the research results of previous researchers. However, at the low WS of 0.1 rpm, for a TPD value up to 0.5 mm, the IMC average thickness decreased as the TPD increased, but the TSL decreased. In Figures 6 and 7, as TPD increases at a WS of 0.1 rpm, the IMC average thickness decreases and TSL decreases until TPD = 0.5 mm. In Figure 8, it can be seen that the IMC average thickness decreases and the TSL increases at TPD = 0.5 mm or more as the TPD increases at the WS of 0.1 rpm. When WS varies from 0.15–0.2 rpm, the IMC average thickness decreases and TSL increases as TPD increases in all regions. Picot et al. [54] used the process parameters of the rotational speed and translational speed of the tool in FSLW of stainless steel 316L and Al alloy 5083 and reported that the TSL decreased with increasing IMC average thickness. Helal et al. [55] used the process parameters of TRS, WS, and tool offset in FSLW of 6061-T6 Al alloy and ultra-low carbon steel and reported that as the IMC average thickness increases, TSL decreases. Kimapong et al. [53] reported that TSL decreased with increasing IMC average thickness using welding parameters such as rotational speed, traversing speed of 0.4 to 1.43 mm/s, and a pin depth in FSLW of A5083 Al alloy and SS400 steel. In previous studies, it was reported that the TSL increased with decreasing IMC average thickness. As WS varies from 0.15–0.2 rpm, it was consistent with the research results of previous researchers. However, at the low WS of 0.1 rpm, for a TPD value up to 0.5 mm, the IMC average thickness decreased as the TPD increased, but the TSL decreased.

decreases and TSL decreases until TPD = 0.5 mm. In Figure 8, it can be seen that the IMC average thickness decreases and the TSL increases at TPD = 0.5 mm or more as the TPD increases at the WS of 0.1 rpm. When WS varies from 0.15–0.2 rpm, the IMC average thickness decreases and TSL increases as TPD increases in all regions. Picot et al. [54] used the process parameters of the rotational speed and translational speed of the tool in FSLW of stainless steel 316L and Al alloy 5083 and reported that the TSL decreased with increasing IMC average thickness. Helal et al. [55] used the process parameters of TRS, WS, and tool offset in FSLW of 6061-T6 Al alloy and ultra-low carbon steel and reported that as the IMC

creased with increasing IMC average thickness using welding parameters such as rotational speed, traversing speed of 0.4 to 1.43 mm/s, and a pin depth in FSLW of A5083 Al alloy and SS400 steel. In previous studies, it was reported that the TSL increased with decreasing IMC average thickness. As WS varies from 0.15–0.2 rpm, it was consistent with the research results of previous researchers. However, at the low WS of 0.1 rpm, for a TPD value up to 0.5 mm, the IMC average thickness decreased as the TPD increased, but the

**Figure 8.** Comparison of IMC thickness for different TRSs of experiment groups A and B. **Figure 8.** Comparison of IMC thickness for different TRSs of experiment groups A and B.

#### *3.4. Characteristics of IMC Thickness and TSL According to TRS 3.4. Characteristics of IMC Thickness and TSL According to TRS*

TSL decreased.

Table 4 summarizes the IMC thickness and TSL according to the change in the TPD and for WS values of 0.1 rpm and 0.15 rpm or more. Figure 8 shows the relationship between the IMC thickness and TSL according to the change in the TRS. Table 4 summarizes the IMC thickness and TSL according to the change in the TPD and for WS values of 0.1 rpm and 0.15 rpm or more. Figure 8 shows the relationship between the IMC thickness and TSL according to the change in the TRS.

**Table 4.** IMC thickness range and thickness average and TSL on TRS.


**Experiment no. 2 1 3 6 5 4 Table 4.** IMC thickness range and thickness average and TSL on TRS.

WS of 0.15 to 0.2 rpm. In Figure 8, group A of experiments no. 1, no. 2, and no. 3 is indicated by a solid line to represent the changes in the IMC average thickness change according to TRS and a WS of 0.1 rpm. As TRS increases, it can be seen that the IMC average Figure 8 compares the differences between the two groups in terms of the change in the IMC thickness according to TRS in group A with a WS of 0.1 rpm and group B with a WS of 0.15 to 0.2 rpm. In Figure 8, group A of experiments no. 1, no. 2, and no. 3 is indicated by a solid line to represent the changes in the IMC average thickness change according to TRS and a WS of 0.1 rpm. As TRS increases, it can be seen that the IMC average thickness increases up to TRS of 1800 mm and decreases when TRS exceeds 1800 mm. In group B of experiments no. 4, no. 5, and no. 6 in Figure 8, at WS values of 0.15 and 0.2 rpm, the variation in the IMC average thickness according to TRS is indicated by broken lines. It shows a gradual increase in the IMC average thickness as TRS increases from 1700–1900 rpm. When TRS is increased from 1700–1900 rpm, the tendency of the IMC average thickness change is different according to the effect of WS. As TRS increased, the IMC average thickness decreased for a WS value of 0.1 rpm and when TRS of 1800 rpm or more. When WS was 0.1 rpm, TRS was 1800 rpm or less, and when WS was 0.15 and 0.2 rpm, it showed an increase over the entire range of TRS.

the IMC thickness according to TRS in group A with a WS of 0.1 rpm and group B with a

In the study by Kundu et al. [51], the IMC thickness increased with increasing TRS. These results do not indicate conditions for WS above a certain speed. Wan et al. [25] and Das et al. [51] simultaneously studied the increase in the IMC thickness according to the increase of TRS, and the decrease in the IMC thickness according to the increase in WS. However, their study also did not report that the IMC thickness increased with increasing TRS at WS above a certain speed. In the study by Kundu et al. [51], the IMC thickness increased with increasing TRS. These results do not indicate conditions for WS above a certain speed. Wan et al. [25] and Das et al. [51] simultaneously studied the increase in the IMC thickness according to the increase of TRS, and the decrease in the IMC thickness according to the increase in WS. However, their study also did not report that the IMC thickness increased with increasing TRS at WS above a certain speed.

thickness increases up to TRS of 1800 mm and decreases when TRS exceeds 1800 mm. In group B of experiments no. 4, no. 5, and no. 6 in Figure 8, at WS values of 0.15 and 0.2 rpm, the variation in the IMC average thickness according to TRS is indicated by broken lines. It shows a gradual increase in the IMC average thickness as TRS increases from 1700–1900 rpm. When TRS is increased from 1700–1900 rpm, the tendency of the IMC average thickness change is different according to the effect of WS. As TRS increased, the IMC average thickness decreased for a WS value of 0.1 rpm and when TRS of 1800 rpm or more. When WS was 0.1 rpm, TRS was 1800 rpm or less, and when WS was 0.15 and 0.2

*Materials* **2022**, *15*, x FOR PEER REVIEW 13 of 18

rpm, it showed an increase over the entire range of TRS.

Figure 9 is a graph showing the relationship between the TSL and TRS. In the figure, group A of experiments no. 2, no. 1, and no. 3 is indicated by a dash-dotted line to represent the TSL change according to TRS at WS of 0.1 rpm. Figure 9 is a graph showing the relationship between the TSL and TRS. In the figure, group A of experiments no. 2, no. 1, and no. 3 is indicated by a dash-dotted line to represent the TSL change according to TRS at WS of 0.1 rpm.

**Figure 9.** Comparison of TSL with TRS for experiment groups A and B [16]. **Figure 9.** Comparison of TSL with TRS for experiment groups A and B [16].

As TRS increases from 1700 rpm to 1900 rpm, it can be seen that the TSL increases up to a TRS of 1800 rpm, and TSL decreases for a TRS of 1800 or more. In Figure 9, group B of experiments no. 4, no. 5, and no. 6 is indicated by dash-double dotted lines to represent the change in the TSL according to TRS at WS values of 0.15 and 0.2 rpm. As the TRS increases from 1700 rpm to 1900 rpm, the TSL shows a gradual decrease. When TRS is increased from 1700 to 1900 rpm, the tendency for the change in TSL differs according to the effect of WS. As TRS increased at the WS of 0.1 rpm, the TSL increased up to TRS of 1800 rpm. It shows that as TRS increases at the WS of 0.1 rpm, the TSL decreases over the entire range of TRS when TRS is over 1800 rpm and at WS values of 0.15 and 0.2 rpm. As TRS increases from 1700 rpm to 1900 rpm, it can be seen that the TSL increases up to a TRS of 1800 rpm, and TSL decreases for a TRS of 1800 or more. In Figure 9, group B of experiments no. 4, no. 5, and no. 6 is indicated by dash-double dotted lines to represent the change in the TSL according to TRS at WS values of 0.15 and 0.2 rpm. As the TRS increases from 1700 rpm to 1900 rpm, the TSL shows a gradual decrease. When TRS is increased from 1700 to 1900 rpm, the tendency for the change in TSL differs according to the effect of WS. As TRS increased at the WS of 0.1 rpm, the TSL increased up to TRS of 1800 rpm. It shows that as TRS increases at the WS of 0.1 rpm, the TSL decreases over the entire range of TRS when TRS is over 1800 rpm and at WS values of 0.15 and 0.2 rpm.

With respect to the TSL and TRS, Kimapong et al. [53] stated that the TSL decreased with increasing TRS. Their study also did not consider the effect of WS on specific regions. In this study, as in the study by Kimapong et al. [53], the TSL decreased as TRS increased at a WS of 0.15 to 0.2 rpm. However, it was found that the TSL increased as TRS increased at the WS of 0.1 rpm and when TRS was 1800 rpm or less.

In Figures 8 and 9, at a WS of 0.1 rpm and below TRS of 1800 rpm, the IMC average thickness increases and the TSL increases as TRS increases. At a WS of 0.1 rpm, if TRS is 1800 rpm or more, the IMC average thickness decreases and the TSL decreases as TRS increases. When WS ranges from 0.15 to 0.2 rpm, the IMC average thickness decreases and TSL decreases as TRS increases in all regions.

Picot et al. [54] used the process parameters of the rotational speed and translational speed of the tool in the FSLW of stainless steel 316L and Al alloy 5083 and reported that the shear strength decreased with increasing IMC average thickness. Helal et al. [55] used the process parameters of TRS, WS, and tool offset in FSLW of 6061-T6 Al alloy and ultra-low carbon steel and reported that as the IMC average thickness increases, TSL decreases. Kimapong et al. [53] reported that the shear strength decreased with increasing IMC thickness using welding parameters such as the rotational speed, traversing speed of 0.4 to 1.43 mm/s, and a pin depth in FSLW of A5083 Al alloy and SS400 steel.

In previous studies, it was reported that the TSL increased with decreasing IMC average thickness. At the WS of 0.15–0.2 rpm, it was consistent with the research results of previous researchers. However, as TRS increased in all regions of TRS at a low WS of 0.1 rpm, the TSL increased as the IMC average thickness increased, and the TSL decreased as the IMC average thickness decreased. Many studies have been done to determine the relationship between the other process parameters and the IMC thickness. A study by Liu et al. [25] indicated that the IMC thickness decreased as the WS increased and the tool offset increased. However, their study results also reported a decrease in the IMC thickness with an increase in the tool offset but did not provide a condition for the WS over a specific speed. Dehghani et al. [29] showed that the IMC thickness increased with increasing PD.

In addition, when the WS was constant, the effect of the TPD was greater than the rotational speed of the tool. This is consistent with the fact that according to the results of this study, the effect of the TPD is dominant. Many studies have also been done to assess the maximum size criterion of the IMC thickness for strength [25,36,46,47]. Jamshidi et al. [36] recommended a critically reactive layer of IMC as thin as 1–2 µm for strong bond strength, as a thicker layer of IMC lowers the TSL and increases the hardness while reducing the bond ductility. However, even if it is thinner than this dimension, the mechanical properties of the joint are said to be improved. Liu et al. [25] reported that the maximum thickness of the IMC at the butt joint of AA6061-T6 and 780/800 trip steel was 1 µm, which was low enough for good weld strength. Movahedi et al. [46] reported that the presence of IMC in FSLW between Al-5083 and St-12 improves the weld strength when the thickness is less than 2 µm. Meanwhile, Sepold et al. [47] showed that excellent mechanical properties could be obtained by minimizing the thickness of the IMC layer to 10 µm or less. In this study, a region with an average IMC thickness of 3.42 to 9.05 µm was obtained over the entire region. In general, the type, size, and amount of IMC formation depends on the input heat, which is controlled by the welding process parameters. An increase in rotational speed and friction time is expected to increase the heat input. Jamshidi et al. [36] also reported that a larger amount of IMC should be formed because the amount of heat input increases with TPD. Dehghani et al. [29] reported that as the PD increased, the interaction between the shoulder and the workpiece increased and the frictional heat was generated more so that the material was sufficiently plasticized, and the material movement around the tool pin was promoted, resulting in higher heat generation.

This study differs from previous studies in that the IMC thickness increases because of the increase in heat input as the PD and TPD increase. It is estimated that the increase in the friction amount with the increase of the TPD leads to an increase in the amount of heat input. However, in the mechanism according to TPD, flow phenomena other than the heat input appear to act as a more dominant factor than the increase in heat input according to TPD. Therefore, it can be estimated that the thickness of the IMC is reduced. Therefore, these results differ from the experimental results of the one factor at a time, in which the effects of PD and TPD were not tested together. Previous studies have shown that the thickness of the IMC increases with increasing TRS. In this study, it was confirmed that the thickness of IMC decreased as TRS increased at a WS value of 0.1 rpm and a TRS of 1800 rpm or higher. In a previous study, the IMC thickness decreased with increasing WS. In this study, the IMC thickness increased with increasing WS. Based on the results

obtained, it can be seen that the effect of TPD is a significant factor in the reduction of the IMC thickness.

#### **4. Conclusions**

In this study, the following major results were obtained by studying the characteristics of the TSL and IMC thickness according to TPD and TRS in FSLW for dissimilar Al and steel pipe.


These results indicate that TPD, IMC thickness, and TSL were directly related, except for in certain areas with low WS. It is assumed that the increase or decrease of TSL in this region is determined by the relative superiority of heat input and material flow.

Therefore, in order to increase the mechanical properties in the FSLW of aluminum and steel for high-rigidity parts, the TPD should be deepened in the weld area where the WS is not high. A suitable TPD should therefore be considered along with tool wear, and continuous material development and the selection of an optimized TPD will play an important role in future dissimilar FSLW.

**Author Contributions:** L.C. performed the experiments, data analysis, and proofreading and wrote part of paper; M.K. provided basic guidance and supervision of FSW; and D.J. reviewed the paper and supervised the thesis. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2018R1D1A1B0705130214). And The APC was funded by the Korea Institute of Energy Technology Evaluation and Planning(KETEP) grant funded by the Korea government(MOTIE) (20206310100050, A development of remanufacturing technology on the grinding system of complex-function for high-precision).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

