Investigation of the effects of machining parameters on tool life and surface roughness during the face milling of the NiTi shape memory alloy with uncoated tools

Shape memory alloys (SMAs) are increasingly used in the fields of aviation, automotive and biomedicine due to their unique properties. Nickel-Titanium (NiTi) alloy materials, which are one of the shape memory alloys, are among the most frequently used alloy materials. The shape memory and super elastic effects of NiTi alloys, high ductility and deformation hardening make it difficult to shape burr. An additional problem is the formation of a white layer during machining. In this study, surface milling operations were performed in dry cutting conditions with uncoated cutting tools with different nose radii. The processing parameters were determined based on the experience gained as a result of the preliminary tests. Tungsten carbide cutting tools with different nose radii (0.4mm and 0.8mm) were used for the milling operations. Milling was carried out at three different cutting speeds (20, 35, 50 m/min), feed rates (0.03, 0.07, 0.14 mm/tooth), and a constant axial cutting depth (0.7 mm). As a result of our experimental studies, the best tool life was found to be in 0.8 mm nose radius cutting tools at 20 m/min cutting speed and 0.03 mm/tooth feed rate (0.264 mm). The minimum average surface roughness was found after milling with 0.8 mm nose radius cutting tool at 20 m/min cutting speed and 0.03 mm/tooth feed rate (0.346 μm). It has been determined that increasing the cutting tool nose radius reduces both the flank wear over the cutting tool and the average surface roughness.


Introduction
In daily life, the demand for more functional products due to the problems encountered in the fields of medicine and industry caused scientists to improve the properties of materials and to produce new materials with superior properties. Smart materials which have made great progress in recent years have the ability to change their properties according to environmental conditions, and they are used to transform one type of energy into another one [1]. The use of smart materials in biomedical, aviation and automotive industries is gradually increasing. Shape memory alloys, which are among smart materials, are the materials that can return to their original form (shape or size) when subjected to a recall process between two transformation phases dependent on temperature or magnetic field [2].
A review of the literature showed that no studies have yet been conducted on the machining of the NiTi shape memory alloy with cutting tools with different nose radii. In the literature, most of the studies are devoted to micro-milling with carbide end mills. However, conventional milling is used for the manufacturing of large-size plates which are used to join bones in medicine. In this study, surface milling of NiTi shape memory alloy with cutting tools that have different nose radii was investigated at different cutting parameters and under dry cutting conditions. The effect of the cutting tools with two different nose radii on the tool lifetime and surface roughness of workpiece was investigated

Test Specimens
The NiTi shape memory alloy used for joining the broken bones in the field of orthopedics has been used as a workpiece. The NiTi alloy was produced using the vacuum arc melting (VAM) method and then was subjected to hot rolling. It was in the austenite phase at room temperature. The dimension of the plate was 25x100x100 mm, and it included 55.8% Ni element. The chemical composition of the NiTi shape memory alloy is given in Table 1 and the mechanical properties are given in Table 2. The phase transformation temperatures of the alloy were measured using the differential scanning calorimetry (DSC) method (see Table 3). The phase transformation temperature values in Table 3 show that NiTi alloy was in the austenite phase at room temperature before processing Table 3. Phase transformation temperatures of NiTi shape memory alloy.

Cutting Tools and Machining Parameters
In the milling of NiTi shape memory alloys, uncoated changeable tungsten carbide cutting tools with 0.8 mm nose radius of R39011T308E (NLH13A) quality and 0.4 mm nose radius of R39011T304E (NLH13A) quality produced by the Sandvik Coromant Company were used. To ensure the same conditions in all experiments, a new cutting tool was used in each test. After each test, the bench was stopped, the cutting speed and feed rate were changed, and a total of 18 experiments were carried out, with 9 experiments for each type of cutting tool with different nose radii. The geometric dimensions of the cutting tools determined in accordance with ISO 1832 are given in Figure 1. Two-blade tool holder R390-025A25-11L, suitable for changeable tools, was used in the experiments. The tool holder was selected according to ISO 5608. In order to provide constant machining parameters, the experiments were carried out by installing only one cutting tool on the tool holder. While determining the machining parameters, three different cutting speeds, feed rates and constant axial cutting depth values were determined based on the experience gained as a result of preliminary tests and considering the values recommended by both ISO 1832 and the manufacturers for the 0.4 mm and 0.8.mm nose radii for cutting tools. The machining parameters used in cutting tests are given in Table 4.

Milling Machine, Surface Roughness Measuring Instrument, and Scanning Electron Microscope (SEM) Device
For chip removal operations, a triaxial "Falco VMC 855-B" industrial CNC milling machine with a power of 10 kW was used. The general view of the milling process is shown in Figure 2. The amount of cutting tool wear and the surface roughness (Ra) of the workpiece were measured at the end of each 100 mm of cutting length of the tool during the cutting process. The measurement of the workpiece surface roughness was performed after each cutting depth removed from the surface. After each test, the cutting operation was paused, the cutting tool was removed from the tool holder and the depth of tool wear was measured using an optical microscope. During the cutting tests, a CCD camera mounted on the Vision SX45 stereo zoom optical microscope was used to measure the amount of flank wear (VB) that occurs on the cutting tools. According to the surface milling standard TS ISO 8688-1, the amount of wear on the free surface of the cutting tool is predicted to be 0.3 mm [63]. The wear mechanism on the cutting tools was examined using the FEI Quanta FEG 250 type scanning electron microscope (SEM) and energy dispersing spectroscope (EDS). The Mitutoyo Surftest SJ-310 tipped surface roughness device was used to measure surface roughness ( Figure 3). After each cutting process, the roughness values taken from three different points over the machined workpiece surface were recorded. The average surface roughness (Ra) was determined by calculating the arithmetic mean of the three values. The cut-off length and the sampling length were taken as 0.8 mm and 5.6 mm, respectively for the measurement of the workpiece surface roughness.
(ae) Radial depth of cut (mm) 15 Figure 3. The image of surface roughness measurement on three different points of NiTi shape memory alloy performed using the Mitutoyo Surftest SJ-310 tipped surface roughness device.

Change in Surface Roughness due to Cutting Speed and Feed Rate
The surface roughness values obtained in three different cutting speeds and feed values with uncoated tungsten carbide tools with two different nose radii are given in Table 5. When the surface roughness values obtained at different cutting speeds were examined, it was observed that the lowest surface roughness was in the chip removal performed at a cutting speed of 20 m/min with a cutting tool with 0.8 mm nose radius and as cutting speed increased, the surface roughness increased as well. The lowest surface roughness was obtained with the cutting tool with a nose radius of 0.8 mm and a feed rate of 0.03 mm/tooth.  Figure 4 shows the workpiece surface roughness values that emerge as a result of the machining of the NiTi shape memory alloy with 0.4 mm nose radius tungsten carbide tools at three different cutting speeds and feed rates. The highest surface roughness was obtained with the cutting speed of 50 m/min and a feed rate of 0.14 mm/tooth (0.736 μm), while the lowest surface roughness was achieved with the cutting speed of 20 m/min and a feed rate of 0.03 mm/tooth (0.424 μm). As we can see from the graph in Figure 4, the values of surface roughness are 0.530 μm at 20 m/min, 0.632 μm at 50 m/min, and 0.660 μm at 35 m/min from lowest to highest. It is known from the literature that with a higher cutting speed, the surface roughness of the workpiece can be reduced [70][71][72][73][74][75]. The reason for lower surface roughness at lower cutting speeds (less than 20 m/min) may be attributed to the BUE that is formed on the cutting tool, which increases the nose radius [38][39][40][41][42][43][44][45]. The scanning electron microscope (SEM) image of BUE that is formed on the cutting tool at 20 m/min cutting speed is shown in Figure 5a. At a certain cutting speed (35 m/s), surface roughness reaches the maximum value, and then decreases again. This surface roughness decrease can be explained by less deformation hardening, better deforming of the workpiece material around the cutting edge and nose radius due to the increasing temperature at high cutting speed, and the flow zone formed at these high temperatures [70 -74]. A 4% decrease in surface roughness (0.632 μm) was achieved with a 30% increase in cutting speed (50 m/min). As already mentioned, when cutting speed is increased from 20 m/min. to 35 m/min, surface roughness increases. This is a consequence of flank wear (Figure 5b) of cutting tool at high temperature because the friction coefficient increases with cutting speed [40][41][42][43][44][45][60][61][62][63][64][65]. An improvement in surface roughness due to the increase in cutting speeds is an expected result [65][66][67][68][69][70][71][72][73][74]. The increase in surface roughness at high cutting speed (50 m/min) and feed rate (0.14 mm/tooth) was attributed to the high tool wear on the cutting tool ( Figure 4c). A 23% and 24% change was observed in surface roughness in the cutting process performed at 20 m/min cutting speed and 0.7 mm constant depth of cut. The lowest surface roughness change was observed when the feed rate was increased from 0.03 mm/tooth to 0.07 mm/tooth (a 133% increase). The increase in surface roughness was found to be 23%. 11% and 2% changes were observed in surface roughness values when there was 133% and 100% increase in feed rate in the cutting process performed at 35 m/min cutting speed and 0.7 mm constant depth of cut. The lowest surface roughness change was observed when the feed rate was increased from 0.07 mm/tooth to 0.14 mm/tooth (a 100% increase). The increase in surface roughness was determined as 2%. 13% and 20% change was observed in surface roughness values when the feed rate was increased by 133% and 100% in the cutting process performed at 50 m/min. cutting speed and 0.7 mm constant depth of cut. The lowest surface roughness change was seen when the feed rate was increased from 0.03 mm/tooth to 0.07 mm/tooth (a 133% increase). The increase in surface roughness was determined as 13%. When the effect of cutting speed on surface roughness was examined, the minimum surface roughness was achieved at 20 m/min cutting speed (0.530 μm).
When the surface roughness values obtained based on the feed rate are examined, the minimum surface roughness was obtained at low feed rate of 0.03 mm/tooth (0.527 μm), while the maximum surface roughness was obtained at high feed rate of 0.14 mm/tooth (0.605 μm). When the surface roughness values in Figure 4 obtained based on the feed rate are examined, the (arithmetic) surface roughness values were (0.527 μm) at the lowest feed rate of 0.03 mm/tooth, (0.605 μm) at 0.07 mm/tooth, and (0.691 μm) at 0.14 mm/tooth from minimum to maximum, respectively. 24% improvement (0.527 μm) was observed in surface roughness obtained at a high feed rate (0.14 mm/tooth) (0.691 μm) when the feed rate was decreased by 79% (0.03 mm/tooth). Increase in surface roughness depending on the increase in feed rate is theoretically (Equation 1) expected and the feed rate has to be reduced to decrease surface roughness [70][71][72][73][74][75]. A higher surface roughness value at high feed rate (0.14 mm/tooth) may be attributed to tool wear (Figure 4c). The results obtained coincide with those in the literature [38][39][40][41][42][43][44][45]53]. The lowest surface roughness (0.527 μm) was obtained at 0.03 mm/tooth feed rate when the effect of feed rates on surface roughness is examined.   Figure 6 shows the workpiece surface roughness values after the machining with 0.8 mm nose radius tungsten carbide tools in three different cutting speeds and feed rates. The maximum surface roughness (0.585 μm) was achieved with the cutting speed of 50 m/min and a feed rate of 0.14 mm/tooth, while the minimum surface roughness (0.346 μm) was obtained at 20 m/min cutting speed and 0.03 mm/tooth feed rate. The factors leading to a decrease/increase in the surface roughness value obtained based on the cutting speed and feed rate with the cutting tool with the radius of 0.8 mm are similar to the surface roughness graph obtained as a result of the machining with the cutting tool with a radius of 0.4 m. The surface roughness (arithmetic mean values) was 0.421 μm at 20 m/min, 0.463 μm at 50 m/min and 0.548 μm at 35 m/min. In the chip removal process at low cutting speeds (35 m/min), a 15% decrease (0.463 μm) was observed in surface roughness with a 30% increment in the cutting speed (50 m/min). A decrease in surface roughness was observed when the cutting speed of 35 m/min was increased to 50 m/min. This may be attributed to easier deformation of the workpiece material and the formation of yield stress region as a result of increasing temperatures on the toolchip interface during cutting process [70][71][72][73][74]. 10% and 41% change was observed in surface roughness values when the feed rate was increased by 133% and 100% in the cutting process performed at 20 m/min cutting speed and 0.7 mm constant cutting depth. The minimum surface roughness change was observed when the feed rate was increased from 0.03 mm/tooth to 0.07 mm/tooth (a 133% increase). The increase in surface roughness was determined as 10%. 7% and 3% change was observed in surface roughness values when the feed rate was increased by 133% and 100% in the cutting process performed at 35 m/min cutting speed and 0.7 mm constant cutting depth. The minimum surface roughness change was observed when the feed rate was increased from 0.07 mm/tooth to 0.14 mm/tooth (a 100% increase). The increase in surface roughness was determined as 3%. 9% and 39% change was observed in surface roughness values when the feed rate was increased by 133% and 100% in the cutting process performed at 50 m/min cutting speed and 0.7 mm constant cutting depth. The minimum surface roughness change was observed when the feed rate was increased from 0.03 mm/tooth to 0.07 mm/tooth (a 133% increase). The increase in surface roughness was determined as 9%. When the effect of cutting speed on surface roughness is examined, the lowest surface roughness was obtained at a cutting speed of 20 m/min (0.421 μm).
The minimum surface roughness was obtained at a low feed rate (0.03 mm/tooth) (0.416 μm) and the maximum surface roughness was obtained at high feed rate (0.14 m / tooth) (0.565 μm) when the surface roughness obtained based on the feed rate was examined. When the (arithmetic) surface roughness values obtained depending on the feed rate ( Figure 6) are examined, the surface roughness values obtained in the selected feed rate were found as (0.416 μm) at 0.03 mm/tooth, (0.451 μm) at 0.07 mm/tooth, and (0.565 μm) at 0.14 mm/tooth, respectively from minimum to maximum. The increase in the feed rate increased the surface roughness. The increase in the amount of removed chip with the increase in the feed rate increases vibration, thereby increasing surface roughness values by accelerating the formation of wear in the cutting tool (Figure 6c) [42-45, 53, 70-72]. An improvement of 26% (0.416 μm) was observed in surface roughness (0.565 μm) obtained at a high feed rate (0.14 mm/tooth) with the reduction of the feed rate by 79% (0.03 mm/tooth). The minimum surface roughness was obtained at a feed rate of 0.03 mm/tooth (0.416 μm) when the feed rate effect on the surface roughness was examined.  When surface roughness obtained depending on the feed rate was examined (Figure 8), it was found that surface roughness achieved with a 0.4 mm nose radius cutting tool was 0.527 μm at 0.03 mm/tooth, 0.605 μm at 0.07 mm/tooth and 0.691 μm at 0.14 mm/tooth. When the nose radius was chosen as 0.8 mm, the change in obtained surface roughness was 0.416 µm at 0.03 mm/tooth, 0.451 μm at 0.07 mm/tooth, and 0.565 μm at 0.014 mm/tooth from lowest to highest. When the effect of feed rates on surface roughness was examined, it was found that the lowest surface roughness was achieved with 0.8 mm nose radius cutting tools at 0.03 mm/tooth feed rate, and a 21% decrease was observed (0.416 μm). It was determined that the surface roughness value increased as the feed rate of both cutting tools increased. These findings coincide with those in the literature [24,42,[70][71][72][73][74]. The decrease in surface roughness depending on the increase in the nose radius is an expected result theoretically (Equation 1) and increasing the nose radius to obtain lower surface roughness value is in line with the literature [38-45, 52-55, 70-72]. Maximum surface roughness was obtained with 0.4 mm nose radius cutting tools, while the minimum surface roughness was obtained with tools with a nose radius of 0.8 mm.

Change in Tool Life Based on Cutting Speed and Feed Rate
The wear of cutting tools after machining the NiTi shape memory alloy with 0.4 mm nose radius tungsten carbide tools at three different cutting speeds and feed rates are shown by the OM images in Figure 9. It is evident from the OM images that the high deformation hardening property of NiTi shape memory alloy caused abrasive wear, while the ductility of this alloy caused adhesive wear.  Figure 10d). This form of wear could be due to the low strength of cutting edge, the mechanical fatigue and high temperature [72]. BUE was observed at a cutting speed of 35 m/min due to the low temperature at the tool-chip interface (Figure 10b). The wear types of cutting edge at high cutting speed (50 m/min) and feed rate (0.14 mm/tooth) were identified as BUE and fracture tool wear ( Figure   10). We assume that the fracture tool wear was caused by the mechanical and thermal fatigue that occurred because of discontinuous cutting (such as milling) [32,43,[69][70][71][72][73][74].  obtained depending on the feed rate were attained as (0.584 mm) at 0.03 mm/tooth, (0.727 mm) at 0.07 mm/tooth, and (0.887 mm) at 0.14 mm/tooth, respectively from minimum to maximum. A 34% improvement (0.584 mm) was achieved in flank wear obtained at a high feed rate (0.14 mm/tooth) (0.887 mm) as the feed rate was decreased by 79% (0.03 mm/tooth). As the feed rate increases, the amount of heat generated in environment also increases. Since NiTi shape memory alloys have lower thermal conductivity, heat could not be removed and flank wear increased due to the effect of high temperature and pressure (see Figure 12) [40][41][42][43][44][45][71][72][73][74]. The minimum flank wear (0.584 mm) was achieved at a feed rate of 0.03 mm/tooth.  At elevated temperature, a chemical reaction occurred between the carbide cutting tool and the workpiece material. Therefore, with increased temperature, flank wear on the cutting tool increases [38][39][40][41][42][43][44][45][68][69][70][71][72]. Thus the highest flank wear value (0.796 mm) was observed at low cutting speed (50 m/min.) and feed rate (0.14 mm/tooth) (Figure 13i). Low flank wear value at low cutting speed (20 m/min) and low feed rate (0.03 mm/tooth) caused the formation of BUE on the cutting tool ( Figure   13a). This can be explained by the low temperature that formed at the tool-chip interface [70][71][72][73][74].  Figure 14 shows the SEM images of the cutting edge after the machining of NiTi shape memory alloy with 0.8 mm nose radius tungsten carbide tools at three different cutting speeds and feed rates.
The reason for the formation of BUE on the cutting tool is the chemical structure of the workpiece with the low temperature caused by friction between the cutting tool and the workpiece, especially at low cutting speeds. The parts are cut off from the cutting tool with the breaking of the BUEs that formed on the cutting tool and hardened in time. Therefore, breaks occur in the tool by means of the adhesion mechanism [42-48, 53, 70-73]. The SEM image in Figure 14f shows what happens when the cutting speed is 50 m/min and the feed rate is 0.03 mm/tooth. On these images, two different types of wear can be identified: mechanical fatigue fracture and crater wear. The crater wear on the rake face wear is due to the elevated temperature resulting from high cutting speed [40][41][42][43]72]. When the cutting speed is 50 m/min and the feed rate is 0.14 mm/tooth, the predominant wears are BUE, flank wear and the fracture in cutting edge (see Figure 14i). The damage in the cutting edge resulting from fracture is due to cyclical loads (thermal and mechanical shocks) on the cutting edge in the cutting process. [69][70][71][72][73][74].  was caused by the hard carbide particles in the workpiece material [65-67, 69.70]. BUE formation has been attributed to low temperature formation on tool-chip interface during the machining of the NiTi shape memory alloy [38][39][40][41][42][43][69][70][71][72][73]. An elemental analysis of BUE was performed with EDX and its image is shown in Figure 15d. The EDX analysis results show that carbide base material and chips were oxidized. in the cutting process performed at 20 m/min cutting speed and 0.7 mm constant depth of cut. The minimum flank wear change was seen when the feed rate was increased from 0.03 mm /tooth to 0.07mm/tooth (a 133% increase). A 7% and a 3% change were observed in the value of flank wear when the feed rate was increased by 133% and 100% in the cutting process performed at 35 m/min cutting speed and 0.7 mm constant depth of cut. The minimum flank wear was observed when the feed rate was increased from 0.07 mm/tooth to 0.14 mm/tooth (a 100% increase). A 9% and a 39% change were observed in the value of flank wear when the feed rate was increased by 133% and 100%

Preprints
in the cutting process performed at 50 m/min cutting speed and 0.7 mm constant depth of cut. The minimum flank wear change was seen when the feed rate was increased from 0.03 mm/tooth to 0.07 mm/tooth (a 133% increase). When the effect of the cutting speed on flank wear was examined, the minimum flank wear was achieved at 20 m/min cutting speed (0.415 mm).
When flank wear based on the feed rate is examined, it was seen that the maximum flank wear (0.744 mm) was obtained at high feed rate of 0.14 mm/tooth, while the minimum flank wear (0.300 mm) was obtained at low feed rate of 0.03 mm/tooth. The flank wear values obtained based on feed rate were (0.300 mm) at the lowest feed rate of 0.03 mm/tooth, (0.482 mm) at the feed rate of 0.07 mm/tooth, and (0.744 mm) at the feed rate of 0.14 mm/tooth, respectively from minimum to maximum. An increase in flank wear was observed due to the feed rate [68][69][70][71][72][73][74]. This situation causes an increase in the surface roughness of the workpiece (Figure 6). A 60% improvement (0.300 mm) was observed in flank wear obtained at a high feed rate (0.14 mm/tooth) (0.744 mm) when the feed rate was decreased by 79%. An increase in flank wear was detected at three different cutting speeds (20,35 and 50 m/min) and high feed rate (0.14 mm/tooth) as it can be seen in the graph in Figure 16 ( Figure 13c, Figure 13f, Figure 13i). It has been determined that worn cutting tools caused poor quality in the machined surface and accordingly an increase in surface roughness ( Figure 6).   Figure 18 shows that the flank wear of cutting tool with a 0.4 mm nose radius was 0.584 mm at 0.03 mm/tooth, 0.727 mm at 0.07 mm/tooth, and 0.887 mm at 0.14 mm/tooth. When the nose radius was 0.8 mm, flank wear decreased. These results confirm that the nose radius of the cutting tool is an important factor during the chip removal process. The cutting tool with higher nose radius creates a larger contact area with the workpiece, and a larger contact area means higher friction and thus an increase in the workpiece surface temperature. As a result, the workpiece material softens, while the cutting tools can keep their hardness at high temperatures [28][29][30][31][32][33]40]. Less tool wear and better tool life of the cutting tool with higher nose radius have been attributed to high strength of the cutting edge [45][46][47][48][70][71][72][73][74][75]. The maximum flank wear as a result of the cutting process of the NiTi shape memory alloy was obtained when the cutting tools had a nose radius of 0.4 mm, while the minimum flank wear was obtained when the cutting tools had a nose radius of 0.8 mm.

Conclusions
In this study, the average surface roughness values of workpiece were investigated during the surface milling of the NiTi shape memory alloy using the cutting tools of different nose radii at dry cutting conditions. The results of the performance tests are as follows: 1. The types of tool wear and damages observed on cutting tools during the cutting tests were found to be flank wear, BUE formation and fracture caused by mechanical fatigue.
2. Flank wear was taken into consideration to compare the lifetime of cutting tools at different machining parameters.
3. It was determined that increasing cutting speed and feed rate in general increases flank wear on the cutting tool.
4. We analysed the cutting tools with two different nose radii (0.4 mm and 0.8 mm) and we found that the lifetime of cutting tools with nose radius of 0.8 mm was approximately 34% higher at 20 m/min cutting speed and 0.03 mm/tooth feed rate.
5. The correlation between feed rate and average surface roughness is linear, and the average surface roughness increases with increasing feed rate. 6. Cutting speed and average surface roughness are inversely proportional. Increasing the cutting speed enhances surface roughness. 7. We found that the nose radius of cutting tool and feed rate have a significant effect on surface roughness.
8. Flank wear on the cutting tool adversely affects surface roughness. 9. Minimum surface roughness (0.346 µm) was achived at 20 m/min cutting speed and 0.03 mm/tooth feed rate using the cutting tools with 0.8 mm nose radius.