Ultrasonic Assisted Machining of Nickel-Based Superalloy Inconel 718 Ultrasonic Assisted Machining of Nickel-Based Superalloy Inconel 718

Inconel 718 has been widely used in industries because of its excellent mechanical prop- erties. However, the machining process, particularly the turning/grinding, of Inconel 718 is still costly due to high cutting force and heavy tool damage. Fortunately, a prom - ising material removal technique, that is, ultrasonic-assisted turning/grinding (called UAT/UAG for simplicity), could potentially play a great role in the high efficiency pre - cision machining of Inconel 718 due to its excellent features such as smaller turning/ grinding force, better surface quality, longer tool working life and lower heat genera - tion. However, few attempts have been done on UAT/UAG of Inconel 718. Therefore, in this work, in order to confirm the feasibility of machining Inconel 718 by UAT/UAG, experimental apparatus/equipment has at first been constructed by installing an ultra sonic cutting-unit/spindle on a NC lathe/surface grinder for UAT/UAG operations, and then experimental investigations have been performed to elucidate the fundamental machining characteristics involving Inconel 718 workpiece including the effects of the ultrasonic vibration and the cutting/grinding speed on the work-surface finish, the machining force and temperature, the chip formation, the tool/wheel wears and so on. The obtained results show that grinding forces and surface roughness were decreased in UAT/UAG.


Introduction
Inconel 718, a nickel-based superalloy, exhibits desirable properties over a wide temperature range and is widely used in aerospace, petroleum and nuclear industries because of its excellent mechanical properties such as high fatigue strength, good corrosion resistance and strong creep resistance [1]. In power generation equipment manufacturing industries including gas turbine engines, most of Inconel 718-made products are commonly machined by turning, cutting, milling and grinding. However, the excellent mechanical properties of Inconel 718 are leading to high cutting force, severe work-surface damage and heavy tool wear in conventional machining/finishing processes [2]. Especially, the turning/grinding process (hereinafter called CT/CG for simplicity) of Inconel 718 is still costly due to the heavy and rapid wear of tools.
As a promising material removal technique, ultrasonic-assisted turning/grinding (hereinafter called UAT/UAG for simplicity) has attracted great attention for decades for the sake of its excellent features such as smaller turning/grinding force, higher material removal rate, better surface quality, longer tool working life and lower heat generation compared with those in CT/CG [3][4][5][6][7]. However, in the most of conventional UAT processes, a linear ultrasonic vibration is applied only in a single direction [8][9][10][11][12], which leads to the difficulty in the obtainment of a reasonable material removal rate especially for difficult-machine metal such as Inconel 718. On the other hand, in conventional UAG processes, most of researches have ever focused mainly on the hard-brittle materials such as crystal silicon and ceramics [13][14][15][16][17][18][19], and few attempts have been done involving Inconel 718.
Therefore, the present authors proposed a novel UAT technique in which an elliptical ultrasonic vibration generated by the synthesis of two orthogonal linear ultrasonic vibrations is applied on the tool in the base plane and experimentally confirmed its performance in the turning of Inconel 718 [20,21]. This new method is hereafter called as EUAT. In addition, the UAG of Inconel 718 in which a linear ultrasonic vibration is imposed to the grinding wheel along the wheel axis was also attempted, and some significant results were attained [22,23]. In this chapter, the processing principles of EUAT/UAG and the respective corresponding apparatuses are described. Then, the fundamental machining characteristics obtained are detailed.

Processing principle and apparatus
The processing principle of elliptical ultrasonic-assisted turning (EUAT) is as illustrated in Figure 1; an elliptical ultrasonic vibration is imposed on the cutting tool in the base plane (XOY plane) and the workpiece is rotated at a speed n w around its own axis (X-axis) in addition to a feed motion of the cutting tool in X-direction at a feed rate v f and a depth of cut a p . The elliptical ultrasonic vibration is achieved by the synthesis of two vibrations in X-and Y-directions with the respective amplitudes of A x and A y generated simultaneously by the ultrasonic vibration unit.
For realizing the processing principle, an experimental apparatus was constructed by installing a commercial ultrasonic vibration unit (UL40-A1 by Takesho Co., Ltd.) onto the tool post of a commercial CNC lathe (TAC-360 by Takizawa Co., Ltd.) as shown in Figure 2. A power supplier was used to apply an AC voltage into the ultrasonic unit for inducing the ultrasonic vibration. A commercial dynamometer was fixed below the ultrasonic unit to measure the cutting forces. As the tool, a coated carbide insert, which is installed at the tip of the ultrasonic unit, was employed. The cutting temperature was measured using an infrared camera.
For quantitatively confirming the generation of elliptical vibration, the vibrations of tool cutting edge in X-and Y-directions were simultaneously measured by two laser Doppler vibrometers (LV-1610; Ono Sokki Co., Ltd.). The frequency of the AC voltage was set at 40.9 kHz, based on the given specification of the ultrasonic unit. Figure 3 plots the vibration amplitudes of the tool cutting edge, A x in X-direction and A y in Y-direction, at different power supplying levels. It is evident that as the power supplying increases, the values of A x and A y increase linearly to 2.35 and 2.05 μm, respectively, at the power level of 40%. Using the vibrations measured in both directions, the trajectory of the tool cutting edge was captured by an oscilloscope (WaveJet 314; LeCroy Co., Ltd), and the obtained trajectories for different power supplying levels were also exhibited in Figure 3. Clearly, the tool cutting edge follows a clockwise elliptic motion, and the   shape of ellipse remains essentially unchanged with a declined angle of 168° even the power supplying level varied, but the size of trajectory increased as the power supplying level escalated. Table 1 summarizes the machining conditions in experiments. As the workpiece, an Inconel 718 rod with the dimension of 40 mm in diameter and 100 mm in length was used. The cutting speed v c was hence changed from 75 m/min to 125 m/min when the workpiece rotational speed n w was adjusted from 600 to 1000 rpm. During machining, the cutting forces and the cutting temperature were measured using the three-component dynamometer and the infrared camera, respectively. After machining, the work-surface roughness was evaluated by using a surface profiler (Talysurf Intra by Taylor Hobson Inc.), and the surface morphology of workpiece was captured by using a laser microscope (VK-8710 by Keyence Co., Ltd). In addition, for comparison, the conventional turning (CT) operations were also conducted on the same experimental apparatus but turned off the power supplier to set the power supplying level at 0%, that is, A x = A y = 0.

Fundamental machining characteristics
Cutting force and temperature. In order to confirm that the ultrasonic vibration-reducing effect of cutting force (UREF) still exists in the high speed cutting of Inconel 718 when EUAT is performed in the base plane, the effect of cutting speed on cutting force in CT and EUAT was studied by experiments. The results are presented in Figure 4, demonstrating that in comparison with CT with EUAT, when the cutting speed was 75 m/min, the feed force F x , the radial force F y and the cutting force F z were reduced by 38%, 46% and 25%, respectively. Similarly, even the cutting speed was increased to 87.5 m/min, then 100 m/min, further 112.5 m/min and finally 125 m/min, similar reduction percentages were observed on the F x , F y and F z with those at 75 m/min.  because the ultrasonic separation exists between the tool nose and the workpiece in X-and Y-directions in the base plane, respectively, in EUAT. In addition, in both CT and EUAT, the largest force was F y , followed by F x and F z in descending order [24].
The cutting temperatures at different cutting speeds in CT and EUAT were also obtained as shown in Figure 5. The results demonstrated that either in CT or in EUAT, as the cutting speed increased, the cutting temperature rose monotonously; however, the increase rate in EUAT seemed consistently lower than that in CT, indicating the ultrasonic vibration significantly contributed to the reduction of cutting temperature in turning Inconel 718.
Chip morphology. Figure 6 shows that the chip width in EUAT is smaller than that in CT, indicating that the chip morphology is affected by the cutting forces; the increase in cutting force adds the transverse flow of the material, leading to an increase in the chip width [25]. Moreover, the adhesion of microchips can be observed evidently on the chip surface in CT, which can hardly be seen in EUAT. These phenomena may be connected to the formation of BUE (built-up edge) during cutting, which will be discussed in the "Tool Wear" section.
Work-surface finish. The surface morphologies of workpieces machined by CT and EUAT at three different cutting speeds of 75, 100 and 125 m/min were captured by the laser microscope.
From the microscopic images shown in Figure 7(a), it is found that regardless of the cutting speed, obvious parallel cutting traces and randomly distributed pits can be observed on the work-surface in CT; the phenomena may be attributed to the formation of BUE on the cutting tool and chip winding during cutting. By contrast, in EUAT (Figure 7(b)), knitted work-like texture was formed on the work-surface owing to the tool vibration, and the work-surface integrity became better. This is probably attributed to the fact that the chips were no longer wrapped around workpiece and easily broken for the sake of the elliptical vibration of cutting tool in the base plane. The work-surface roughness was measured in the feed direction using the surface profiler for both CT and EUAT. The evaluation length was set to 5 mm. In each test, the surface roughness parameters R a and R z were measured eight times, and their average values were calculated and regarded as the roughness. Figure 8 shows the effect of the cutting speed on the work-surface roughness in CT and EUAT at three different cutting speeds of 75, 100 and 125 m/min. The surface roughness R a and R z in EUAT seemed both larger than those in CT. In EUAT with the increase of cutting speed, the R a and R z slightly decreased, while in CT, the R a and R z tended to increase slightly which is caused by the formation of built-up edge (BUE) on tool cutting edge [20].
Tool wear. Figure 9(a) and (b) shows the SEM images of the cutting edge of tool employed for 4.2 min in CT and EUAT, respectively. Evidently, in CT, the BUE appeared on the cutting   tool, whereas in EUAT, it was absent. Comparing Figure 9(a) with Figure 9(b) shows that the cutting tool adopted in CT experienced considerably heavier flank wear than that adopted in EUAT. The maximum flank wears in CT and EUAT were 0.23 and 0.19 mm, respectively. This indicates that the cutting tool was worn down more rapidly in CT than in EUAT, possibly because the tool vibration in EUAT reduces the friction between the tool and the workpiece.
In EUAT, the ultrasonic vibration of tool occurs simultaneously in X-and Y-directions and  results in a clockwise elliptical motion in the base plane (see Figure 1). During cutting, owing to the elliptic vibration the tool repeats the cutting-in and cutting-out action periodically. This action reduces the formation of BUE and eventually retains the sharpness of the tool cutting edge. Therefore, the chips and the work-surface in EUAT are almost free of chip particles.

Processing principle and equipment
The processing principle of UAG is illustrated in Figure 10; an axial ultrasonic vibration is imposed on the grinding wheel in x-direction and the workpiece is fixed on a work-holder under which a z-stage is located for determining the wheel depth of cut Δ. The peripheral speed of grinding wheel is V c and a feed motion is given to the workpiece in y-direction. For comparison, a conventional grinding (CG) was also performed by releasing spindle from ultrasonic vibration (Figure 11).   For realizing the processing principle, an experimental apparatus was constructed by installing a commercial ultrasonic vibration spindle (URT40 by Takesho Co., Ltd., Japan) onto a commercial NC grinder (GRIND-X IGM15EX by Okamoto Machine Tool Co., Ltd., Japan). A commercial dynamometer (9256A by Kistler Co., Ltd., Switzerland) was positioned under the ultrasonic spindle. On the lower end of the spindle, a metal-bonded #140 cBN grinding wheel with a diameter of d s = 8 mm was fixed and an Inconel 718 specimen (L48 mm×W36 mm×T3 mm) was used as the workpiece. The abrasive grains on the working surface of the wheel employed were observed by a three-dimensional (3D) SEM (ERA-8900 by ELIONIX. Co., Ltd., Japan), showing that most of the grains were cone shaped and their vertical angles and average diameter were around 120° and 105 μm, respectively. Table 2 exhibits the grinding conditions. As the purpose of this work is predominantly to reveal the fundamental machining characteristics in UAG of Inconel 718 including the effects of the ultrasonic vibration and the wheel peripheral speed on the grinding force and specific

Superalloys for Industry Applications
grinding energy, the work-surface finish, the chip formation (chip size and geometry), the material removal rate and the grinding wheel wear, the grinding operations were performed at different ultrasonic vibration amplitudes and wheel peripheral speeds but the constant workpiece feed rate and wheel depth of cut.
Typically, cBN abrasives require a cutting speed of over 50 m/min for grinding Inconel 718 [16]. Therefore, the value of wheel peripheral speed V c was set at 100.5-138.2 m/min (wheel rotational speed n g = 4000-5500 rpm). In order to ascertain the benefits from application of the ultrasonic vibration to precision grinding, the workpiece feed rate was set at a low value of 5 mm/min. Correspondingly, a large value of depth of cut is required to ensure material removal rate; hence, the value of Δ was set at 80 μm. The actual vibration amplitude A p-p of the wheel was measured using the Laser Doppler vibrometer (LV-1610 by On Sokki Co., Ltd., Japan), revealing that the value of A p-p was varied in a range of 0-9.4 μm as the power supplying level rises. In addition, dry grinding operations were performed without coolant supplying.
The material removal rate Q was also experimentally obtained using the relationship of Q = bΔ r V w , where Δ r is the actually measured work-depth removed. Furthermore, the ground work-surface roughness was measured with the surface profiler (Talysurf Intra by Taylor Hobson Inc.) to indicate the work-surface finish. For further investigation, the chips used for the SEM observation were those escaped out of the grinding zone during dry grinding and collected by a piece of conductive plastic (4⨯4 mm), which was located nearby the grinding zone. The working surface condition of grinding wheel, that is, the chips adhesion, the wheel wear behavior like grain releasing/fracture were also examined by SEM observation.

Fundamental grinding characteristics
Grinding force, specific grinding energy and material removal rate. Figure 12(a) and (b) shows the grinding forces measured under various values of A p-p and V c , respectively. Evidently, either the tangential force F y or the normal one F z monotonously decreased as the A p-p increased (Figure 12(a)), and almost linearly decreased with the increasing V c (Figure 12(b)). It should be noticed, for example, that at V c = 138.2 m/min, once the ultrasonic vibration at A p-p = 9.4 μm has been applied, the values of F y and F z were dropped by 51.9% and 38.6%, respectively, compared with that without ultrasonic vibration (A p-p = 0 μm). Additionally, considered that the F z affects tremendously, the formation of micro fracture defects on chip which is characterized by the knife chip, a lower F z in UAG might lead to a smaller number of knife chips (see Figures 18 and 19) compared with that in CG.
In addition, the grinding force ratio of F z /F y was obtained as exhibited in Figure 12, showing that the ratio increased with the increasing A p-p (Figure 12(a)), whereas little effect of the V c can be observed on the ratio (Figure 12(b)) either in CG or in UAG. As revealed by Dong Kun Zhang [26], the lower friction ploughing and smaller chip deformation lead to the increase of grinding force ratio. Therefore, the results demonstrated that ultrasonic vibration contributed to the reduction of friction ploughing and chip deformation. Further, a larger vibration amplitude seemed to increase the material removal ratae, as shown in Figure 13. This confirmed that the ultrasonic vibration can enhance the material removal owing to the decrease in grinding forces.
Furthermore, the specific grinding energy, u, in UAG was compared with that in CG. As the u is defined as the energy per unit volume of material removed [27], it can be expressed in Eq.
where F x (N) is the grinding force in x-direction due to the wheel ultrasonic vibration and V x (m/min) is the wheel ultrasonic vibration speed determined by the equation of Both of them are varied periodically at the frequency of f during grinding.
In the current work, the values of V c , b, Δ and V w have been already known as exhibited in Table 2 and that of F y is also the already known one as shown in Figure 12. The measured results of F x and the calculated value of V x are given in Table 3. The F x V x changed from 0 to the respective peek values for different A p-p , while the F y V c decreased with the increase of A p-p . Figure 14 shows the relation between u and A p-p . It can be seen from this figure that although the u varied between u L and u H at the given A p-p for the sake of the ultrasonic vibration, the u intended to decrease as the A p-p increases, confirming that the ultrasonic vibration benefits the reduction in the specific grinding energy significantly.
Work-surface finish. Given that the effective cutting edge distribution density affects the work-surface roughness in grinding processes and that this density is affected by the ultrasonic vibration [22], it was also investigated how the work-surface finish varied with vibration amplitude A p-p . Figure 15 shows the obtained results, demonstrating that the ultrasonic Figure 14. Effect of ultrasonic vibration on specific grinding energy.   vibration indeed contributed to the work-surface finish improvement; the larger the A p-p was, the smaller the work-surface roughness became. The optical microscopic images of worksurfaces in CG and UAG were also compared in the same figure, showing that on the worksurface by CG only parallel grain cutting marks were formed, whereas on that by UAG, many ultrasonic-induced knitting-patterned cutting marks overlap with the parallel ones.
Chips formed. An electron scanning microscope (SEM) with 3D measurement/observation functions (ERA-8900 by ELIONIX Co., Ltd.) was employed to measure the 3D chip size and observe the chip geometry. Figure 16(a) and (b) shows the 3D SEM images of chips formed in UAG (A P-P = 9.4 μm) and CG at V c = 138.2 m/min, respectively. The measured results of mean length and cross section area of chips are shown in Figure 16 (c) and (d). It is evident in Figure 16(c) that the ultrasonic vibration leaded to the reduction of mean length and mean cross section area of chips. Especially, when the amplitude was A p-p = 9.4 μm, the mean length was reduced by 36.3% and the cross section area was reduced by 64.3%. Figure 16 (d) shows the effect of the wheel peripheral speed V c , demonstrating that as the V c increases the cross section area intends to reduce either with or without ultrasonic vibration, but the effect on the chip length was not obvious.
In conventional grinding of Inconel 718, the formed chip could be classified into six types: flow, shear, rip, knife, slice and melt [28]. In this study, all the six types occurred; however, dominant types were shear, knife and flow either in CG or UAG as shown in Figure 17.
Further, the number percentages of each type under different conditions were experimentally investigated, and the obtained results (Figure 18) show that the majority of chips in CG were shear type, whereas most of them in UAG were flow type especially at larger vibration amplitude (A p-p ≥ 4.1 μm). This indicates that the UAG of Inconel 718 is potentially avoiding the formation of shear chips and prefers the flow chips especially at larger amplitude.
As for the effect of the wheel peripheral speed V c , it can be found from Figure 19 that in CG when the V c changed from 100.5 m/min to 113 m/min, the percentage of the flow chip increased from 4 to 63% and the change in the percentage of knife chip was very small, while that of the shear chips reduced from 86 to 30%. As the increase of V c the number percentage of three types of chips were not obvious. In UAG, the increase of V c was not main effect to the change of number percentage of chips.
Working surface condition of grinding wheel. Through SEM observation, the working surface morphologies of wheel with/without ultrasonic vibration were captured. From the SEM images shown in Figure 20, it is found that both chips adhesion and grains releasing/fracture occurred during grinding. Comparing the right sides of Figure 20(a) and (b) revealed distinctly that the ultrasonic vibration reduced both the chips adhesion and grains releasing/ fracture significantly.
Furthermore, the chips adhesion area was filtered, extracted and binarized by using Image-Pro Plus for quantitative analysis [22]. For this purpose, an observation zone with size of 1.13 mm × 6 mm on grinding wheel working surface was selected to measure the percentage of chips adhesion area in the total wheel working surface area, as shown in Figure 21.
Obviously, increasing the ultrasonic amplitude leads to the reduction of the percentage of chips adhesion area. Especially, at A P-P = 9.4 μm, a decrease by 72.4% was observed.     At last, the releasing/fracture of abrasive grains was investigated by comparing the percentages of grain releasing/fracture number in total wheel working surface in CG and UAG, as shown in Figure 22. It is noticed that in CG (A p-p = 0 μm), the percentage reached 44%. Once an ultrasonic vibration with a small amplitude of A p-p = 2.2 μm has been imposed, the percentage turned to be less than 20%, demonstrating that the presence of ultrasonic vibration restraints  the grain releasing/fracture phenomenon considerably. In addition, when compared the CG result with the UAG one at A p-p = 9.4 μm, it is figured out that the percentage decreased by 40%. This phenomenon is attributed to the ultrasonic vibration-induced vertical inertia force which acts on chips. This vertical inertia force encourages chips to leave cutting edge and hence increases cutting capacity of grain.

Summary
As a step toward the development of a novel technique for the machining of superalloy Inconel 718, at first the elliptic ultrasonic-assisted turning (EUAT) method was proposed and its fundamental performance was experimentally confirmed, then the ultrasonic-assisted grinding of Inconel 718 was attempted and successfully elucidated the fundamental grinding characteristics. The obtained results can be summarized as following.
In EUAT, (1) the cutting force was considerably smaller than that in conventional turning (CT) and strengthening the ultrasonic vibration decreased the cutting force; (2) although ultrasonic vibration marks were formed on the work-surface along the cutting direction, the surface finish was improved, whereas scratches and pits obviously existed on the work-surface in CT and (3) the ultrasonic vibration significantly restrained the formation of built-up edge on cutting tool and decreased the flank wear of tool.
In UAG, (1) imposing the ultrasonic vibration to the grinding wheel decreased the grinding forces and increased the material removal rate significantly. Strengthening the wheel ultrasonic vibration decreased the specific grinding energy, demonstrating that the ultrasonic vibration benefits the reduction in the energy consumption. (2) The ultrasonic vibration contributed to the work-surface finish improvement; the larger the A p-p was, the work-surface roughness became better. (3) The chip size, that is, cross-section area and length, was distinctly affected by the ultrasonic vibration but little effect of wheel peripheral speed is observed. (4) The ultrasonic vibration in grinding is avoiding the formation of shear chips and prefers the flow chips especially at larger amplitude. (5) The grinding wheel wear was dominantly attributed to chips adhesion, grains releasing/fracture which are reduced significantly by the ultrasonic vibration. The percentage of chips adhesion area and the number percentage of grains released/fractured decreased as the vibration amplitude increased.