*2.1. Materials*

In this study, the workpiece was Ni-Ti stent wire, a biomaterial commonly used in a variety of biomedical applications. The Ni-Ti stent wires were 120 mm in length and 0.5 mm in diameter, with an Ra of approximately 0.24 μm. A 2D dimension view of Ni-Ti stent wire workpiece is shown Figure 5. Figure 6 shows a photograph comparison of Ni-Ti stent wires used in this study. Tables 1 and 2 list the mechanical and chemical compositions of the Ni-Ti stent.

**Figure 5.** Two-dimensional view of Ni-Ti stent wire workpiece.

**Figure 6.** Photographic comparison of Ni-Ti stent wire material.


**Table 1.** Mechanical properties of Ni-Ti stent wire.



#### *2.2. Typical Processing Oils*

Processing oils are essential elements of the finishing or grinding process, reducing the friction and high temperatures that occur between the surface finish of the sample and the abrasive particles [21]. During the UPMAF process, the mechanical friction between the relative motion of the sample and abrasive particles can cause microcracks on the finished surface. High temperatures generated during the finishing process can increase the wear of the abrasive tools, resulting in dimensional deviation and premature failure. To reduce both friction and temperatures in the finishing process, processing oils are applied to the mixtures of magnetic abrasive tools. In previous research, industrial processing oils have been used in the magnetic abrasive finishing process, but after the finishing process, undesirable toxic substances in the industrial oil are likely to exist on the finished surface, which can also suffer from unacceptable surface roughness. To overcome these problems, we replaced the industrial processing oils commonly used in finishing or machining processes with oil associated with low environmental impacts, including olive, castor, and light oil. A comparison of the properties of the processing oil used for UPMAF process is supplied in Table 3. Light oil has the lowest viscosity and density when compared with olive oil and castor oil, but the highest surface tension (31 dyne/cm), followed by castor oil and olive oil. However, castor oil has the highest viscosity among three oils.


**Table 3.** Characteristics of processing oil used in the finishing process.

## *2.3. Experimental Conditions*

Detailed conditions for this experiment are supplied in Table 4. Ni-Ti stent wires with an Ra of 24 μm were chosen as the sample workpiece and finished at di fferent rotating speeds of the magnetic field (i.e., 500, 1000, 1500, and 2000 rpm), for 150 s of total finishing time. Nd-Fe-B permanent magnets were utilized to generate the high magnetic force. The vibration frequency of the magnetic pole was 10 Hz with an amplitude of 5 mm, and the moving feed of the workpiece was 80 mm/min. Three di fferent processing oils (light, olive, and castor oil) were applied during the process for comparison. Scanning electron microscope microimages were utilized to evaluate the changes in Ra of the Ni-Ti stent wire. For determination of surface roughness after processing, the average surface roughness (Ra) of Ni-Ti stent wire was measured at three di fferent positions every 30 s of processing time by using a surface roughness tester (Mitutoyo SJ-400) (Mitutoyo, Sakado, Japan). Figure 7 shows a measuring procedure of surface roughness Ra value for wire material using a surface roughness tester (Mitutoyo SJ-400). As shown in the photo that during the measuring process, the tip was moved along with the length of wire material and the measuring length on wire sample was 5 mm with the measuring speed was 0.5 m/s. Also, the value of removed diameter of Ni-Ti stent wire was measured every 30 s of processing time by using a laser scan micrometer (Mitutoyo LSM-6200) (Mitutoyo, Sakado, Japan). The photos of the scanning electron microscope (SEM at 120×) were used to assess the e ffect of environmentally friendly oil on the improvement in Ra of the Ni-Ti stent wire in an UPMAF process using rotating magnetic field.


**Figure 7.** Measuring procedure of surface roughness value for wire material, using a surface roughness tester (Mitutoyo SJ-400).

#### **3. Result and Discussion**

To investigate the finishing characteristics of an ultraprecision magnetic abrasive finishing with different processing oils, the electrolytic iron powder (Fe#200) and diamond abrasive particles (0.5 μm) were mixed together with light oil, olive oil, and castor oil. The effect of different processing oils on finishing characteristics of Ni-Ti stent wires at different magnetic field rotating speeds was discussed.

#### *3.1. E*ff*ect of Light Oil on Finishing Characteristics*

To find the most optimal magnetic field rotating speed in terms of the surface roughness, the experiment was performed at a magnetic field rotating speed (500, 1000, 1500, and 2000 rpm), 10 Hz of vibration frequency, 5 mm of amplitude, and a feed rate of 80 mm/min. The effect of light oil on improvement in surface roughness of the Ni-Ti wire with various rotation speeds of the magnetic field at (500, 1000, 1500, 2000 rpm) is shown in Figure 8. As shown in Figure 8, Ra values of the Ni-Ti wire stent were significantly improved by light oil at all rotation speeds. The greatest improvement in Ra was obtained at 1500 rpm followed by 1000, 500, and 2000 rpm. This indicates that increasing the rotation speed of the magnetic field can improve Ra of an Ni-Ti wire stent. In the case of 1500 rpm, the Ra of the stent wire decreased from 0.24 μm to 0.07 μm over 150 s of processing time. However, at 2000 rpm of magnetic field rotating speed, the centrifugal force of the magnetic abrasive tools was increased, resulting in reduced the magnetic force, and therefore the magnetic abrasive tools flew in all directions.

**Figure 8.** Correlation of surface roughness Ra vs. processing time, (light oil, diameter 0.5 μm, 80 mm/min).

#### *3.2. E*ff*ect of Olive oil on Finishing Characteristics*

Olive oil is an environmentally friendly substance that is generally utilized in the food and medical industries [22,23]. It has a viscosity of 0.0341 Pa·s, and a density of 0.857 kg/m3. In this study, 0.2 mL of olive oil was combined with 0.8 g of electrolytic iron particles and 0.3 g of diamond paste at 25 ◦C using an unbonded magnetic abrasive method. The e ffect of olive oil on improvement in Ra of the stent wire at various rotating speeds of the magnetic field is shown in Figure 9. Olive oil significantly improved the surface roughness of Ni-Ti wire at all rotating speeds. However, 1500 rpm was found to be the optimal condition at 90 s of processing time. When 1500 rpm of rotating speed was used, the original Ra value of Ni-Ti improved from 0.24 μm to 0.07 μm for 90 s, after which the Ra did not improve further, because the unevenness of a surface of Ni-Ti stent wire was completely removed by that time. The slope of 2000 rpm shows worse improvement in Ra compared with the other conditions. The result can be attributed to the increase in the centrifugal force of magnetic abrasive tools.

**Figure 9.** Correlation of surface roughness (Ra) vs. processing time, (olive oil, 0.5 μm, 80 mm/min).

#### *3.3. E*ff*ect of Castor Oil on Finishing Characteristics*

Castor oil is a colorless vegetable oil pressed from castor beans [24]. The boiling point of castor oil is 313 ◦C, its viscosity is 0.3115 Pa·s, and its density is 0.956 kg/m3. This oil is commonly utilized in cosmetic products, including creams and moisturizers. In addition, it has been utilized to improve hair conditioning in other products due to supposed antidandru ff properties. In this study, 0.2 mL of castor oil was combined with 0.8 g of electrolytic iron particles and 0.3g of diamond paste. The e ffect of castor oil on the improvement in Ra of the wire stent at various rotating speeds of the magnetic field (500, 1000, 1500, and 2000 rpm) is shown in Figure 10. As with the light oil and olive oil, the Ra of the Ni-Ti wire stent improved at all rotating speeds. The slope of 1500 rpm shows the greatest improvement in Ra, from 0.24 μm to 0.16 μm for 90 s of processing time.

**Figure 10.** Correlation of surface roughness (Ra) vs. processing time, (castor oil, 0.5 μm, 80 mm/min).

#### *3.4. Percentage Improvement in Surface Roughness (PIISR)*

The effect of different processing oils on the improvement in surface roughness of Ni-Ti stent wire is shown in Figure 11. The processing time of each workpiece was 150 s. In the bar graph, the surface roughness of the Ni-Ti wire workpiece rapidly improves from 0 s to 30 s of processing time, and then improves at a diminished rate until 150 s. The greatest improvement in Ra was obtained with light oil, followed by olive oil, and castor oil. In the case of industrial oil, the Ra value decreased from 0.24 μm to 0.07 μm. With olive and castor oils, Ra values improved from 0.24 μm to 0.08 μm and from 0.24 μm to 0.12 μm, respectively. The percentage of improvement formula for Ni-Ti wire stent Ra as a function of different processing oils can be expressed by a formula. *BUPMAF* (before ultraprecision magnetic abrasive finishing) is the Ra value before processing and *AUPMAF* (after ultraprecision magnetic abrasive finishing) is the value after 150 s of processing. *PIISR* is the percentage improvement in surface roughness, expressed as the rate of change of Ra for each set of processing conditions as a percentage of the improvement in Ra.

**Figure 11.** Correlation of surface roughness Ra vs. processing time according to processing oils, (1500 rpm, 0.5 μm, 80 mm/min).

**3URFHVVLQJWLPHWVHF**

A *PIISR* graph of the improvement in surface roughness Ra before and after processing according to processing oil within 150 s of processing time is shown in Figure 12. The results show that improvement in Ra for each processing oil was greater than 45%. Light oil was associated with an improvement of 70.83%, while olive and castor oil resulted in improvements of 66.67% and 45.833%, respectively. Light oil is therefore the preferred oil, followed by olive oil and castor oil. This result can be attributed to the viscosity of the processing oils. As shown in Table 3, the viscosity values of light oil, olive oil, and castor oil 0.3115, 0.0341, and 0.005, respectively. From this it can be concluded that low-viscosity oil produced greater improvements in Ra compared with oils of higher viscosity. Light oil, which has the lowest viscosity and density, reduced the temperature and friction generated during finishing and resulted in a high-quality Ra value. According to the lowest viscosity of light oil (0.005 Pa·s), the Ra value obtained by this condition should be much better than olive oil. However, the Ra value is difficult to improve to 0.07 μm. This is probably due to the effect of another parameter on Ra improvement, such as 0.5 μm of abrasive, which cannot enhance the Ra value less than 0.07 μm.

**Figure 12.** Correlation of percentage improvement in surface roughness (*PIISR*) vs. processing time under optimal conditons, (1500 rpm, 150 s, 0.5 μm 10 Hz).

We can conclude that environmentally friendly processing oils can be used for precision finishing of Ni-Ti wire stents via an UPMAF process. Olive oil was chosen as the processing oil for precision finishing of Ni-Ti wire material at 500, 1000, 1500, and 2000 rpm of a rotating magnetic field for 150 s. The effect of olive oil on the removed diameter of Ni-Ti against processing time is shown in Figure 13. The results show that the removed diameter (RD) of Ni-Ti wires can be significantly increased in all conditions of a rotating magnetic field. In terms of the RD, the diameters of the wire materials removed at 500, 1000, 1500, and 2000 rpm were 0.00130, 0.00212, 0.00281, and 0.00192 mm, respectively. As with Ra improvement, 1500 rpm was associated with the largest reduction in diameter. This can be explained by the fact that, at 2000 rpm of rotational speed, the unbonded magnetic abrasive particles have enough time for removing the irregular scratches from the Ni-Ti stent wire's surface. Therefore, it can be confirmed that, when increasing the rotating speed to a certain level, the best result can be received.

**Figure 13.** Correlation of removed diameter vs. processing time (olive oil, 0.5 μm, 80 mm/min).

SEM microimages of Ni-Ti stent wires before and after processing by an UPMAF process (magnified 120 times) are shown in Figure 14. The surface conditions of Ni-Ti stent wires before finishing are shown in Figure 14a. Initial scratches and unevenness were found throughout the surface of the stent wire, and the initial Ra was 0.24 μm. The surface condition of the wire after processing by light oil, olive oil, and castor oil can be found in the Figure 14b–d), respectively. It was confirmed that the finished surfaces of the Ni-Ti wire were smoother than the surface before processing. The surfaces finished with light oil and olive oil were significantly smoother than before processing, with Ra values of 0.07 μm and 0.08 μm, respectively. However, the surfaces finished with castor oil were not as smooth as those finished with light oil and olive oil. As can be seen in Figure 14d, the original scratches and irregular asperities remain on the surface. Energy-dispersive X-ray spectroscopy (EDX) analysis of a wire stent produced by ultraprecision magnetic abrasive finishing with olive oil is shown in Figure 15. The chemical composition of the stent wire before processing is shown in Figure 15a. The analysis result of EDX test shows that 44.60% Ti and 55.40% Ni were detected at the Ni-Ti stent wire surface. After processing with olive oil, 44.69% Ti, and 55.31% Ni were detected at the surface, as seen in Figure 15b. In addition, EDX analysis revealed no toxic substances on the finished surfaces of the Ni-Ti wire.

**Figure 14.** Surface of the workpiece before and after finishing. (**a**) Before finishing, Ra = 0.24 μm; (**b**) Finished with light oil, Ra = 0.07 μm; (**c**) Finished with olive oil Ra = 0.08 μm; (**d**) Finished with castor oil Ra = 0.12 μm.

**Figure 15.** Energy-dispersive X-ray spectroscopy (EDX) test results of Ni-Ti wire before and after processing by ultraprecision magnetic abrasive. (**a**) Components of the workpiece before processing (Ra: 0.24 μm, processing time: 0 s); (**b**) Components of the workpiece after processing (light oil, Ra: 0.08 μm, processing time: 120 s).
