Experimental optimization of machining GH4145 by atomizing discharge ablation milling

Atomizing discharge ablation milling (ADAM) technology is an efficient discharge machining technology derived from the traditional electrical discharge machining (EDM) method, which can be used to efficiently machine hard-to-machine materials such as nickel-based superalloy. In this present, the performance of machining nickel-based superalloy GH4145 by ADAM and Air near-dry EDM were compared, and the experimental results showed that the material removal rate (MRR) obtained by ADAM was nearly double that of the latter. A single-factor experiment were conducted to investigate the effect of electrode rotation speed on ADAM’s processing performance. Subsequently, an orthogonal experimental method was used to design the experiment. The signal-to-noise ratio analysis method was used to systematically study the performance characteristics of ADAM, including the influence of atomization amount, oxygen pressure, discharge current, duty ratio on MRR and tool electrode relative wear rate (TWR). The results showed that discharge current was the most influential processing parameter on MRR and TWR. Finally, the optimal combination of processing process parameters that met the requirements of various processing effect evaluation indicators were obtained and the correctness of the single objective optimization results was verified through experiments.


Introduction
Electrical discharge machining (EDM) is a relatively new process that utilizes electrical and thermal energy for machining, commonly known as discharge machining.The difference between electric discharge machining and general cutting machining is that during electric discharge machining, the tool does not come into contact with the workpiece but relies on the continuous pulse spark discharge between the tool and the workpiece, which uses the local and instantaneous high temperature generated during the discharge to erode the metal material gradually.Due to the visible sparks generated during the discharge process, it is called electric discharge machining [1,2].The machinability of the processed material mainly depends on its conductivity and thermal characteristics and is almost unrelated to its mechanical properties, such as hardness and strength.Therefore, EDM can break through the limitations of traditional mechanical cutting methods on tool materials and achieve the processing of hard and tough workpiece materials with relatively soft tool materials.Consequently, it is often used to process hard-to-machine materials with conductive properties, such as Ni-base superalloy, titanium alloys, mold steels, etc, and has received more applications and attention in the fields of aerospace, molds, and other fields [3][4][5].
Regarding the exploration and research on EDM processing of difficult to machine materials, many literature studies have been conducted on material properties, structural forms of tool electrodes, and the types of discharge medium.Somu C et al [6] changed the physical and chemical properties of the tool electrode by adding Cr metal material into the copper metal electrode.They achieved higher machining results on Inconel 718 Alloy through EDM.Kumar Rakesh et al [7] studied the EDM of Inconel 718 alloy with the ariance analysis method and found that the tool electrode rotation speed had the most significant influence on the surface roughness.The research of Tanjilul M et al [8] showed that with the increase of the depth of EDM processing superalloy Inconel 718, the processing efficiency decreased.Mainly for the effective evacuation of process debris leading to secondary discharges and resulting in increased machining time.Ishfaq Kashif et al [9] employed different cryogenically treated electrodes combined with varying media of kerosene, found that using non DCT electrodes and adding Span-20 kerosene together to process Inconel 617 alloy can achieve higher material removal rates.Jiang Yi et al [10] adopted sintered porous materials as tool electrodes for electric discharge machining, which has higher efficiency in processing titanium alloys than traditional solid electrode materials.Xu Moran et al [11] through experiments found that EDM oil can obtain higher surface quality when processing titanium alloys compared with deionized water medium.The conventional atomizing medium is mainly composed of air and liquid medium, and its processing efficiency is limited by the energy of the pulse power supply.Therefore, the atomizing medium composed of oxygen and liquid medium has received much attention and research.Oxygen can oxidize the workpiece to release tremendous chemical energy at high temperatures, improving EDM's material removal rate [12,13].In order to achieve efficient and stable discharge ablation machining of titanium alloys, Kong et al [14,15] proposed two efficient methods for machining titanium alloys by reducing oxygen concentration through mixed gas discharge ablation and mixed gas atomization discharge ablation, which significantly improved the processing efficiency compared with traditional atomizing medium.
There are also many literatures to explore and optimize the machining parameters of EDM machining hardto-machine materials by orthogonal experimental method, one parameter at a time methodology, response surface methodology and other methods.In order to improve significantly the material removal rate of HSS steel materials in EDM, Yadav VK et al [16] introduced oxygen medium into the discharge medium to achieve this.Rakesh Kumar et al [17] found that the current and pulse width had the most significant influence on the overcutting size of nickel-based superalloy processed by EDM by using the Taguchi's DOE method to study the impact of the machining parameters.Param Singh et al [18] used one parameter at a time methodology to explore and study the effect of higher ultrasonic power and tool electrode speed on material removal rate in ultrasonic vibration assisted micro electrical discharge machining of Inconel 718 alloy.Alharbi N [19] used a single-factor experimental method to investigate and found that in the machining of Inconel 718 alloy by electric discharge machining, the surface erosion resistance obtained with increased tool electrode speed will also be correspondingly improved.Shaiful H et al [20] used response surface methodology to investogate the characteristics of electrical discharge machining of certain cast irons.
According to the above literature, to improve the efficiency or quality of hard-to-machine materials in electrical discharge machining, better processing results can be obtained by changing the tool electrode material, structure, and discharge medium.In this present, the sample of nickel-based superalloy GH4145 was taken as the workpiece, and atomized discharge ablation milling technology was carried out by mixing oxygen and liquid medium to form a discharge medium.First, the effect of near-dry EDM with an air atomization medium is studied by comparison.Then the effects of gas pressure, current and other process parameters on material removal rate, electrode relative wear rate were systematically studied by orthogonal test method, and the primary and secondary factors affecting the performance indicators were analyzed, and the optimal parameter configuration was obtained.Finally, the feasibility of the obtained optimized parameters was verified.

Materials and methods
A nickel-based superalloy is a complex multi-component alloy mainly composed of nickel, carbon, iron, chromium, and other small amounts of aluminum, copper, and other metals.It is a variety of metal materials with high-temperature strength, erosion resistance, and oxidation resistance.They are usually used for manufacturing components in high-temperature environments, such as aviation engines, gas turbines, petrochemical equipment, etc [21,22].Therefore, GH4145 material is selected as the workpiece material in this present, which is mainly a nickel-based superalloy strengthened by the aging of γ′ [Ni3 (Al, Ti, Nb)] phase.It has good resistance erosion and oxidation resistance below 980 °C, high strength below 800 °C, and good relaxation resistance below 540 °C.It also has good formability and weldability.Table 1 shows the chemical composition of the GH4145 material.Table 2 shows the physical and mechanical properties of GH4145.However, its high hardness, high thermal strength, and other characteristics are similar to other nickel-based superalloy materials, and it is prone to tool wear, thermal expansion, deformation, and other problems during mechanical cutting, resulting in the reduction of cutting quality and other problems [23,24].
Atomizing discharge ablation milling (ADAM) technology uses the atomizing medium formed by lowconcentration oxygen and deionized water as the dielectric for electrical discharge machining.It is sent into the inter-electrode discharge gap through a hollow rotating electrode to participate in the inter-electrode discharge ablation milling process.The oxygen medium in the medium reacts violently with the molten metal material formed by discharge, and a massive amount of chemical energy is released to achieve efficient erosion of the metal material.In contrast, the liquid medium improves the discharge machining process and processing efficiency [25].The schematic of the device system for ADAM is shown in figure 1(a), and the rotary atomization device designed and modified on the NH25F CNC EDM forming machine (Beijing Ninghua Co, Ltd, China) is shown in figure 1(b).This device can achieve atomization of oxygen and deionized water, and the rotation speed range of hollow electrodes is 0-1500 rpm, meeting the requirements of atomization discharge ablation milling.
The experiment used a multi-hollow copper electrode with an outer diameter of 3.5 mm, as shown in figure 1(b).The length of the milling groove was 8.5 mm, the width 3.5 mm, and the depth 3 mm.The method of layer-by-layer reciprocating milling with each layer milling of 0.5 mm is adopted.In this experiment, GH4145  flat plate with a seze of 30 mm × 15 mm × 5 mm was used as a workpiece.In order to characterize the processing effect, material removal rate (MRR), tool electrode relative wear rate (TWR), and surface roughness were used to evaluate.The MRR was obtained by dividing the volume of the removed material by the total processing time, the TWR was expressed by the volume percentage of the electrode material wear and the corresponding workpiece material removal amount [26].The workpiece removal quality and electrode removal quality was measured using a METTLER TOLEDO high precision balance (accuracy 0.1 mg).The machined surface roughness (Ra) was measured using a TR200 surface roughness tester.The surface microstructure after discharge ablation were observed using a Sigma 500 field emission scanning electron microscope (SEM).
In this present, to explore the effectiveness of ADAM technology in processing GH4145, a comparative study was conducted on the performance of ADAM and Air Near dry EDM in processing nickel-based superalloy GH4145.Then, the single factor experimental method was used to study the effect of tool electrode rotation speed on process parameters in ADAM technology.Table 3 shows the specific processing parameters for relevant selection.The so-called positive polarity machining of electric discharge refers to the connection of the workpiece being processed to the positive electrode of the pulse power supply, while the tool electrode is connected to the negative pole of the pulse power supply.
Orthogonal Design of experiments is a method to study multi factors and multi-levels.It can significantly reduce the number of tests without affecting the comprehensive understanding of the impact of many factors on performance indicators.It is an efficient, fast, and economical design of the experiment method.Therefore, in order to systematically study the processing performance of ADAM technology, this present selects MRR and TWR as the evaluation indicators for the orthogonal experiment.Based on the previous experiment, a fixed electrode speed of 500 rpm, milling thickness of 0.5 mm, voltage of 220 V, and other parameters remain unchanged.Atomization amount, oxygen pressure, current, and duty cycle are selected as influencing factors, and a 3-level 4-factor L9 (34) orthogonal experiment table is established.The factor level design is shown in table 4.

Comparative experimental results
Figure 2 shows the experimental results using the conditions in table 3. It can be seen that the MRR of ADAM has increased by 91% compared to Air Near-dry EDM, which is due to the relatively high content of oxygen in the combustion supporting medium of ADAM's atomization medium, which is prone to violent oxidation reactions with high-temperature molten materials in the inter-electrode discharge area, releasing more chemical heat energy and promoting a large amount of material erosion between the electrodes.Analysis suggests that the energy released by the oxidation reaction between a unit of oxygen medium and a unit of metal is sufficient to erode many metal materials [25], significantly improving the MRR of ADAM.However, due to limitations such as the servo system function of the electric discharge forming machine tool, the advantages of ADAM technology cannot be fully utilized.
Figure 3 shows the comparison of tool electrodes for processing.It can be seen that the electrode deformation after ADAM processing is relatively large, mainly due to the introduction of chemical energy between ADAM electrodes, which induces the increase of MRR as well as the increase of electrode wear.However, due to the relatively high MRR of ADAM, the TWR of ADAM is lower than that of Air Near-dry EDM.Electrical discharge machining results from the accumulation of a large number of single pulse discharges.Introducing an inter-electrode oxidation reaction increases the volume of single pulse discharge erosion.It induces the formation of large etch pits on the surface of the workpiece to be machined.Therefore, ADAM's Ra is more larger than Air Near-dry EDM's.From the surface micromorphology of the two processing methods shown in figure 4, it can be seen that ADAM forms relatively large etching pits due to the influence of interelectrode oxidation reactions.Still, the final micro morphology characteristics of the two methods are not significantly different.
Figure 5 shows that the increase in electrode rotation speed does not have a prominent improvement effect on MRR, nor does it have a significant impact on TWR.Analysis suggests that the liquid phase particles between the electrodes are subjected to the dual energy effects of the inter-electrode discharge channel and oxidation reaction [27,28], which will rapidly vaporize and expand, causing work to be done on the inter-electrode erosion particles and timely discharge the erosion products from the discharge electrodes gap.At this time, the role of electrode rotation in improving the inter-electrode performance cannot be highlighted.Furthermore, the electrode rotation speed range used in this experiment is relatively small, and within a small range of rotational speed changes, electrode rotation does not significantly improve the inter-electrode discharge state.
The signal-to-noise ratio analysis method [29-31] is adopted, and the corresponding signal-to-noise ratio is taken as the analysis indicator for the effectiveness of ADAM.It is possible to fully consider the influence of noise factors while analyzing controllable factors, which is conducive to finding an ideal combination of machining process parameters that meet the requirements of machining effectiveness.The research purpose of this article is to search for the optimal combination of processing parameters through the Taguchi method and convert the experimental results of MRR into signal-to-noise ratio based on the case of the 'Larger the Better' of the signalto-noise ratio method in order to maximize MRR; Convert the experimental results of TWR into signal-to-noise ratio based on the case of 'Smaller the Better' of the signal-to-noise ratio method, in order to minimize TWR [32].The experimental run order and measured responses are illustrated in table 5 and attained MRR and TWR values with correlating S/N ratio.

Effect of machining parameters on MRR
Response plots for means of S/N ratios of MRR are shown in figure 6.As the atomization amount increases, the MRR first increases and then decreases, indicating an optimal atomization amount.With the increase of atomization amount, the number of liquid particles that undergo vaporization phenomenon after being heated between the electrodes increases, which generates a significant chip removal force on the particles eroded between the electrodes, thereby promoting the improvement of MRR; However, excessive liquid particles [33,34] mean not only a decrease in the oxygen medium entering the inter-electrode but also an increase in the liquid phase medium that cannot be wholly vaporized between the electrode, resulting in a decrease in MRR.It can also be seen that MRR increases with the increase of oxygen pressure, discharge current, and duty ratio.It can also be noticed that the discharge current has a more significant impact on the variation of MRR.It is concluded that an increase in oxygen pressure will promote an increase in oxygen flow rate per unit time entering the inter-electrode, inducing more oxygen media to undergo chemical ablation reactions with the inter-electrode molten material, thereby increasing MRR.In EDM, the erosion amount of material volume depends on the discharge energy provided by the pulse power supply between the two electrodes.As shown in Formula (1) [35], the current and pulse width increase the erosion energy between the electrodes, which is conducive to forming more molten metal materials between the electrodes to improve the MRR.At the same time, it also increases the molten metal materials involved in the oxidation reaction to induce the increase of chemical energy between the electrodes, thereby improving the MRR.The increase in duty ratio means an increase in pulse width and a decrease in pulse interval.The reduction in pulse interval is not conducive to deionizing the discharge channel and the smooth discharge of products from inter-electrode erosion.Therefore, a larger duty ratio does not necessarily mean that MRR has been consistently increased.Accordingly, as the duty ratio increases, it first increases and then slightly decreases.

e e 0 e
Where U e -power supply voltage, t e -pulse width, I e -peak current.In the process of EDM, when the pulse width is fixed, as the current increases, more electrons will form in the discharge channel to bombard the positive electrode, and deeper erosion pits will be created on the surface of the positive electrode workpiece [36,37] (as shown in figure 7).Therefore, under the action of the discharge   channel, more high-temperature molten materials are formed that participate in the inter-electrode oxidation reaction [38], thereby increasing the chemical reaction energy between the electrodes, and increasing the energy acting on the workpiece leads to the improvement of MRR.
In order to verify that an increase in current leads to an increase in surface erosion pits on the workpiece, three-dimensional surface morphology measurements were conducted on the machined surfaces under different currents, and the results are shown in figure 8.These images were measured using the three-   dimensional surface topography instruments (model number NANOPS50).The research shows that under the action of high current, large erosion pits will occur on the workpiece surface, resulting in higher surface roughness [39].This indicates that high current not only improves MRR, but also indirectly reduces surface quality.
The influence of various factors on MRR was studied through the S/N ratio response table.It is used to discover the process parameter that has the most MRR for the value of 'delta,' as per table 6.From table 6, it can be drawn out that the most affecting input parameters include discharge current (Ip) followed by duty ratio (i), oxygen pressure (P), and atomization quantity (L).Every parameter's optimum level is determined through Taguchi's 'larger the better' approach.According to S/N ratios, a higher value of MRR is reached for the finest (optimal) condition when L = 40 ml min −1 , P = 0.5 MPa, Ip = 18A, i = 0.8, optimal conditions for individual parameters are 'A2B3C3D3'.Minitab 18.0 software was used to calculate the corresponding signal-to-noise Ratio for the above processing parameter levels, and the predicted S/N = 36.738was finally obtained, which was greater than all the S/N ratio of MRR in table 5, indicating that the highest MRR could be obtained under this combination of levels.In order to verify the correctness of this conclusion, the discharge processing condition was set to 'A2B3C3D3' for the ADAM test, and the average MRR obtained from the three processing tests was 82.65 mm 3 min −1 , which was greater than all MRR values in table 5, indicating that the single-objective optimization results were reliable.

Effect of machining parameters on TWR
The orthogonal experimental results of the TWR and the signal-to-noise ratio of each group of experiments are shown in table 5.The mean value of the TWR signal-to-noise ratio at each processing parameter level is shown in table 7. The relationship between the influence of various factors on the mean value of the TWR signal-to-noise ratio is shown in figure 9.
As shown in figure 9, with the increase of atomization amount, oxygen pressure, and discharge current, TWR increases but decreases with the increase of the duty cycle.It can also be seen that the discharge current has a more significant impact on the variation of TWR.Analysis shows that when the atomization amount is relatively small, the content of oxygen medium between the electrodes is relatively high, Which increases the  proportion of energy generated between the electrodes acting on the tool electrode, resulting in relatively large electrode wear.However, as the atomization amount increases, the energy generated by the oxidation chemical reaction between the electrodes decreases, indirectly leading to a decrease in the energy acting on the tool electrode, resulting in a decrease in electrode volume loss and a decrease in TWR.The increase in oxygen pressure will increase the chip removal force of the inter-electrode erosion products [40], indirectly or indirectly reducing the 'adhesion effect' of the erosion products on the electrode surface, reducing the protective effect on the electrode and thus increasing the electrode wear.The increase in discharge current means that the energy generated by the inter-electrode discharge channel increases.At the same time, according to equation (1), the proportion of energy acting on the tool electrode also increases, so the electrode wear will also increase typically.with the increase of the duty cycle, the electrode wear decreases, which is contrary to the increase of the electrode wear with the increase of the pulse width in the relevant literature [23].At present, based on the conservation of energy between electrodes, a reasonable explanation cannot be provided.It is necessary to carry out the single pulse discharge with the pulse width change and the electrode wear mechanism analysis to explore the internal reasons.
Figure 10 shows the surface energy spectrum of the tool electrode after processing in running sequences 1 and 8, respectively.It can be seen that the surface of tool electrodes under both conditions contains elements such as Ni/Cr/O, indicating that the inter-electrode melting erosion products form a protective layer on the electrode surface.However, compared to high-duty cycle conditions, there are relatively fewer elements such as Ni/Cr/O on the electrode surface, indicating that during the discharge ablation process, there is relatively less sputtering of electrode melting material on the tool electrode surface, which has a relatively weak protective effect on the tool electrode surface, resulting in higher electrode wear.
In table 7, the 'delta' values corresponding to the process parameters that obtain the highest MRR can be found.From table 7, it can be drawn out that the most affecting input parameters include discharge current (Ip) followed by oxygen pressure (P), atomization quantity (L), and duty ratio (i).Taguchi's 'smaller the better' approach determines every parameter's optimum level.According to S/N ratios, a lesser value of TWR is reached for the finest (optimal) condition when L = 30 ml min −1 , P = 0.3 MPa, Ip = 8A, i = 0.8, optimal conditions for individual parameters are 'A1B1C1D3'.Minitab 18.0 software was used to calculate the corresponding signal-to-noise Ratio for the above parameter levels.Finally, the predicted S/N = −0.034was obtained, greater than all the S/N ratios of TWR in table 5, indicating that combining these factors could obtain the minimum electrode wear.To verify the correctness of this conclusion, the discharge machining condition was set as 'A1B1C1D3' for the ADAM test.The average TWR obtained during the three times of tests was 1.02%, which was lower than all the TWR values in table 5, indicating that the single-objective optimization results were reliable.

Conclusion
In this paper, the machining performance of GH4145 were investigated by ADAM technology.The optimal processing parameters of ADAM technology were obtained through comparative experiments with air near-dry EDM and signal-to-noise ratio analysis of orthogonal experiments.The specific conclusions are as follows: Under the same discharge machining conditions, the MRR of GH4145 processed by ADAM increased by 91% compared to Air Near-dry EDM, and the TWR was relatively low.However, due to the introduction of intense oxidation reactions between electrodes, the machined surface roughness of the workpiece was relatively high, forming relatively large erosion pits on the surface of GH4145 alloy workpiece.The single-factor experimental results found that within the selected electrode rotation speed range in this paper, the MRR and TWR of ADAM processing GH4145 remained almost unchanged with changes in electrode ratio speed.
As the atomization amount increases, the MRR of ADAM shows a trend of first increasing and then decreasing.The MRR increases with the increase of oxygen pressure and discharge current, and first increases and then slightly decreases with the increase of the duty cycle.The optimal combination of control factors obtained for maximizing single objective MRR was identified as 'A2B3C3D3'.I p and I remain utmost affecting input factors affecting MRR.TWR increases with the discharge current, oxygen pressure, and the atomization amount but decreases with the duty cycle increase.The optimal combination of control factors for the minimum optimization of TWR was identified as 'A1B1C1D3'.I p and P remain utmost affecting input factors affecting TWR.
In the future, it may be possible to conduct further investigations on the dimensional accuracy, surface roughness amount, and thickness of the surface recast layer of ADAM-machined GH4145 for practical engineering purposes.

Figure 1 .
Figure 1.(a) Schematic diagram of the ADAM setup.(b) Physical photos of the testing machine.

Figure 2 .
Figure 2. Comparison of results of different processing methods.

Figure 4 .
Figure 4. Surface of workpieces processed by different methods.

Figure 5 .
Figure 5.The effect of electrode rotational speed.

Figure 7 .
Figure 7.The influence of current on discharge channel and erosion pit.

Figure 8 .
Figure 8.The effect of different currents on the surface morphology of workpieces.

Figure 10 .
Figure 10.EDX images of the electrode after machining.

Table 2 .
Physical and mechanical properties of GH4145.

Table 4 .
Form of factor and level.

Table 5 .
Experimental plan and measured process responses.

Table 6 .
S/N ratio response table of MRR.

Table 7 .
S/N ratio response table of TWR.