Simulation and experimental investigation of inner-jet electrochemical grinding of GH4169 alloy

Hnsong LI,Shuxing FU,Qingling ZHANG,b,Shen NIU,Ningsong QU

aCollege of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China

bAVIC Jincheng Nanjing Engineering Institute of Aircraft Systems,Nanjing 211106,China

1.Introduction

With the continuing progress in developments in the aerospace industry,nickel-based superalloys have been widely used in recent years.GH4169 alloy(Ni-Cr-Fe)is one of the most commonly used nickel-based super alloys in aero engine turbine blades,as it has excellent fatigue resistance,high-temperature strength,and good resistance to corrosion and radiation.1–5In traditional machining,its excellent strength at high temperatures leads to a high cutting force and tool wear seriously.6–8At present,some non-traditional machining methods,such as electrochemical machining (ECM),9electrical discharge machining(EDM),10laser machining,11and electrochemical grinding (ECG),12are commonly used for machining GH4169 alloy and other difficult-to-machine materials.

ECG is a compound machining method that involves both ECM and conventional grinding.In ECG,electrochemical reactions and mechanical grinding work at the same time when materials are removed from a workpiece.13Due to the existence of mechanical grinding,ECG has a better machining accuracy and surface quality than those of ECM.In addition,compared to conventional grinding,ECG has a higher material removal rate and a better processing quality,especially for processing high-temperature alloys and other difficult-to machine materials.14

In ECG,a common method is to use an external nozzle to spray electrolyte into an ECG area.Tehrani et al.investigated the problem of ‘over cutting”in electrolytic grinding of 304 stainless steel,and found that the use of a pulsed power supply and an adjustment of the duty cycle could effectively reduce over cutting.15Puri et al.investigated the effects of varying the machining voltage and cutting speed on the material removal rate(MRR),surface quality,and passive film in ECG of P20 cemented carbide.16Curtis et al.used a diamond grinding wheel to machine an aero engine blade tenon by ECG.17Kozak and Skrabalak.explored the relationship between ECM and mechanical grinding in ECG of stainless steel plates by using a brazed diamond grinding wheel.18Also using a brazed diamond grinding wheel,Qu et al.studied the influences of ECG parameters on the speed and MRR of Inconel 718 alloy,where the depth of the machined slot was 0.5 mm in one pass at a feed rate of 6.6 mm·min-1.12

Zhang et al.presented inner-jet ECG of GH4169 alloy using aninnerhollowhemispherical grinding wheel,inwhichtheelectrolyte was injected into a processing area from holes in the bottom surface of the grinding wheel.When the machining depth was set to 3 mm in a single pass,a feed rate of 1.8 mm·min-1was achieved.The machined surface was an arc groove.19

A flat surface is one of the most common finished surfaces in mechanical processing.Nevertheless,few investigations have been reported on machining a flat surface using an inner-jet ECG method.To obtain a GH4169 alloy machined flat surface with high precision,two types of inner-jet ECG grinding wheels were used to make a comparison,machining results under different conditions were simulated,and a series of experiments was conducted later.Relevant conclusions of the experimental results were analyzed and reviewed.In addition,a sample was machined by one type of the grinding wheel with a better machining effect.

2.Simulation of machining process

2.1.Principle of inner-jet ECG

Fig.1 shows a schematic diagram of inner-jet ECG.In a machining method of using an external nozzle to spray electrolyte into an ECG area,when the machining depth was increased to 1 mm or deeper,it was difficult for the electrolyte to enter the machining gap,and it was easy to produce sparks due to lack of electrolyte,resulting in losing abrasive particles on the cathode and burning on the surface of a workpiece.19Inner-jet ECG could overcome the above shortcomings,and in this paper,the machining depth was set as 3 mm.During the process of inner-jet ECG,the grinding wheel moves along a prescribed trajectory while rotating at a high speed,and the electrolyte is injected into the processing area from holes in the bottom surface of the grinding wheel.At the same time,the electrolyte is connected to both the anode(workpiece)and the cathode(grinding wheel)to form a conducting circuit.The electrolyte takes away the dissolved product of the anode and dissipates the Joule heat produced by the electrochemical reaction and abrasive grinding in the machining area

For inner-jet ECG,the shape and size of the workpiece are copied from the cathode(grinding wheel).The copy precision refers to the similarity between the shape and size of the workpiece and those of the cathode.Since a machining gap appears in inner-jet ECG,in principle,the shape and size of the machined surface and those of the cathode surface will not be the same.The machining gap between the bottom of the grinding wheel and the machined surface of the workpiece is the main factor affecting the copy precision of the machined flat bottom surface.The smaller the machining gap is,the higher the copy precision of the machined surface is.In this paper,two types of inner-jet ECG grinding wheels were compared,and the electric field software COMSOL was used to simulate the machining gaps machined by the two grinding wheels.

Fig.2 shows the structures of two inner-jet ECG grinding wheels:grinding wheel A and grinding wheel B.Grinding wheel A was a hollow metal bar with an electroplated diamond abrasive layer on the outer cylindrical surface.The inside and outside diameters of the hollow metal bar were 4 mm and 6 mm,respectively,and the thickness of the metal bar bottom was 1.5 mm.The thickness of the electroplated diamond abrasive layer was 0.1 mm.6 small holes with a diameter of 1 mm were uniformly distributed on grinding wheel A,and the distance from the center of the holes to the bottom surface of grinding wheel A was 2 mm.The six holes ensured a uniform distribution of the electrolyte in the ECG machining area.Grinding wheel B was obtained by machining a conical concave on the bottom surface of grinding wheel A.In order to obtain an obvious experimental comparison and ensure the structural strength of the grinding wheel bottom,the depth of the conical concave of grinding wheel B was set to 1 mm.

2.2.Electrical simulation

ECM plays a major role in ECG.In order to study the effect of inner-jet ECG using the two grinding wheels,two electric field simulation models were established as shown in Fig.3.Model A was an electric field simulation of inner-jet ECG using grinding wheel A,and Model B using grinding wheel B.The machining process under different parameters was simulated.

The following assumptions were made

1.The process of ECG was in a state of equilibrium

2.The electrical parameters did not change with time

3.The conductivity and temperature of the electrolyte were uniform

The electric potential φ in the interelectrode gap Ω could be approximately described by Laplace’s equation20–22as follows:

Boundary conditions were as follows:

where U represents the potential between the anode workpiece and the cathode grinding wheel.n represents the normal coordinate of each point on the boundary.

Fig.1 Schematic diagram of inner-jet electrochemical grinding.

Fig.2 Structures of two inner-jet ECG grinding wheels.

According to Faraday’s law,the material removal rate vacan be described by23:

where η is the current efficiency,ω is the volume electrochemical equivalent of the material,and i is the current density.

Both models were analyzed using COMSOL software.The parameters used in the calculations are listed in Table 1.In this study,four groups of simulation conditions were designed.To study the effects of different applied voltages on the experimental results,applied voltages of 10,15,20,and 25 V were chosen.When the applied voltage was too low,the electrolytic effect was not obvious,and when the applied voltage was too high,the machined surface quality was poor.The grinding wheel feed rate ranged from 1.2 mm·min-1to 1.8 mm·min-1,equally divided into four intervals.When the grinding wheel feed rate was too slow,the rate of electrochemical material removal was too slow,and when the grinding wheel feed rate was too fast,it was unable to continue the process due to frequent short circuits.

Fig.4(a)and(b)shows the simulation results at different machining times for Model A and Model B,respectively.The conditions of the simulations were set as follows:applied voltage,25 V;grinding wheelfeed rate,1.8 mm·min-1;machining time,0,400,and 800 s.

Fig.5 shows schematic diagrams of the simulation results for Model A and Model B.The diagram shows the grinding wheel,the machined workpiece,and the machining gap.

Fig.3 Two electric field simulation models.

Table 1 Simulation conditions.

Fig.4 Simulation results at different processing times for Models A and B.

Fig.5 Schematic diagrams of simulation results.

The results show that the size of the machining gap under different simulation conditions can be calculated using COMSOL electric field simulation software.Fig.6 shows the machining gaps of Model A and Model B under the simulation conditions given in Table 1.It can be seen from Fig.6 that with an increasing grinding wheel feed rate,the size of the machining gap decreased,and with an increasing applied voltage,the size of the machining gap increased,when the other machining conditions remained unchanged.In addition,under the same grinding wheel feed rate and the same applied voltage,the machining gap was smaller in Model B than in Model A.

With an increasing grinding wheel feed rate,the dissolution time of the workpiece was shortened,and the size of the machining gap decreased.Moreover,the higher the applied voltage was,the stronger the current density was,and a higher current density led to a faster dissolution rate of the workpiece,which resulted in a larger size of the machining gap.The bottom of grinding wheel B was a conical concave,so the ability to dissolve the workpiece declined compared to grinding wheel A;hence,the machining gap was smaller in Model B than in Model A.

The smaller the machining gap was,the better the copy precision of the machined flat bottom surface was.According to the simulation results,it can be concluded that both using a faster grinding wheel feed rate and using a grinding wheel with a conical concave bottom surface(grinding wheel B)improved the copy precision of the machined flat bottom surface.

3.Experimental procedure

3.1.Experimental system

Fig.7 shows a schematic diagram of the experimental system for inner-jet ECG.This system consisted of an X–Y–Z movement stage,an electrolyte recycling system,an electrolyte temperature control system,and a machining current acquisition system.During inner-jet ECG,the movement of the grinding wheel was controlled by a computer.The rotating speed of the grinding wheel was controlled by a rotary motor.The electrolyte was pumped out,and injected into the processing part through a rotary joint,a hollow spindle,and an inner-jet ECG grinding wheel.There were a pump,a valve,a pressure,and a filter in the electrolyte recycling system.In this system,the pressure and the flow of the electrolyte could be controlled,and the insoluble processing product in the electrolyte could be cleaned up timely.The temperature of the electrolyte was controlled by a thermostat,which ensured a constant electrolyte temperature.The power supply was a direct current power supply,and the workpiece was connected to the positive electrode of the power supply,while the grinding wheel was connected to the negative electrode by a conducting ring.In the machining current acquisition system,the processing cur-rent could be collected and sent to the computer in real time by a Hall sensor set on the circuit.

Fig.6 Machining gaps of Models A and B.

Fig.7 Schematic diagram of the experimental system for inner-jet ECG.

Fig.8 shows a photograph of the two grinding wheels.They were made of hollow stainless steel bars with nickel-based electroplated diamond.The electrical conductivity of the nickelbased electroplated diamond was good.The diamond grain sizes of the two electrodeposited diamond grinding wheels were both 80 μm.The use of small electrodeposited diamond reduced the machining gap and improved the processing accuracy.

3.2.Maximum feed rate and MRR

To improve the copy precision of the machined flat bottom surface,the size of the machining gap had to be reduced.According to the simulation results,increasing the grinding wheel feed rate reduced the size of the machining gap.In this paper,the maximum feed rates of inner-jet ECG using grinding wheels A and B under different machining conditions were investigated.With the reference of relevant literature,the machining conditions are listed in Table 2.12,19

Fig.8 Photograph of two grinding wheels.

The maximum MRRs of inner-jet ECG using the two grinding wheels were also studied.MRR is one of the most important criteria determining the machining efficiency and is defined as follows:where m is the mass removed in the ECG process,and t is the machining time.

When other machining conditions remain unchanged,the faster the grinding wheel feed rate is,the higher the MRR is.To obtain the maximum MRR of inner-jet ECG using the two grinding wheels,the maximum feed rates were studied.In inner-jet ECG,when the feed rate of a grinding wheel exceeds the maximum feed rate,the material removal rate is slower than the grinding wheel feed rate,which will lead to frequent short circuits,and machining cannot continue.In the experimental system of this paper,the machining current was collected and sent to the computer in real time by a Hall sensor set on the circuit.With an increase of the grinding wheel feed rate,once a frequent short circuit signal was monitored by the computer,it showed that the grinding wheel feed rate had exceeded the maximum feed rate.By this method,the maximum feed rate could be accurately measured.

Fig.9 shows the maximum feed rates and maximum MRRs of inner-jet ECG using grinding wheels A and B under the machining conditions in Table 2.The experiments of each group were repeated three times,the results were obtained under stable processing,and there was no obvious short circuiting.Before and after each experiment,the GH4169 alloy was weighed,and the mass removed in the ECG process was weighed using an analytical balance with an accuracy of 0.001 g.When the applied voltages were 10,15,20,and 25 V,the maximum feed rates using grinding wheel A were 1.8,2.3,2.5,and 2.6 mm·min-1,respectively,and the maximum MRRs were 0.301,0.417,0.456,and 0.493 g·min-1,respectively.When the applied voltages were 10,15,20,and 25 V,the maximum feed rates using grinding wheel B were 1.8,2.3,2.5,and 2.6 mm·min-1,respectively,and the maxi-mum MRRs were 0.288,0.372,0.411,and 0.431 g·min-1,respectively.

Table 2 Machining conditions 1.

Under the different applied voltages in Table 2,the maximum feed rate of inner-jet ECG using grinding wheel A was the same as that using grinding wheel B,the maximum MRR of inner-jet ECG using grinding wheel A was higher than that using grinding wheel B at the same applied voltage,and the maximum MRR could be increased by increasing the applied voltage in the range of 10–25 V.With an increasing applied voltage,the intensity of electrolysis was strengthened,and the MRR was increased.

3.3.Copy precision of machined bottom surface

Inner-jet ECG of GH4169 was studied using the two grinding wheels at the maximum feed rates measured.Table 3 lists the machining conditions.The machining path was a straight line,and the other machining conditions were the same as those given in Table 2.The experiments for each group were repeated three times,the results were obtained under stable processing,and there was no obvious short circuiting.In this experiment,the maximum machining gaps and the standard deviations of the gaps machined using the two grinding wheels for each group were investigated.

Fig.10 shows the specimen sections machined by grinding wheels A and B under the processing conditions of each group.These figures were captured by a three-dimensional pro filometer(DVM5000,Leica,Germany).As it can be seen from these figures,there was a more obvious concave in the middle of the bottom surface machined by grinding wheel A,compared to that by grinding wheel B.

Fig.11 shows the electrolytic reaction area of grinding wheel A.Because the side and bottom of grinding wheel A were both conductive,the machined bottom surface was electrolyzed repeatedly after ECG.Since the bottom of grinding wheel A was very close to the machined bottom surface,it played an important role in the formation of the concave in the middle of the machined bottom surface.

Fig.12 shows a schematic diagram of the machined bottom surface.An imaginary yellow line was set on the surface of the machined bottom,as well as 5 evenly distributed points on it(A0,B0,C0,D0,and E0).The electrolysis times of the 5 points were different when the grinding wheel moved from position 1 to position 2 and passed the yellow line.Since the 5 points were mainly dissolved by the bottom of the grinding wheel,the dis-solution times of the 5 points were proportional to the length of the line segment corresponding to each point(A0,B0,C0,D0,and E0corresponding to A0,B1-B2,C1-C2,D1-D2,and E0,respectively).Therefore,the midpoint of the yellow line had the longest dissolution time,and the endpoints of the line had the shortest dissolution time.

Fig.9 Maximum feed rates and maximum MRRs using grinding wheels A and B.

Table 3 Machining conditions 2.

Fig.10 Specimen sections machined by grinding wheels A and B.

Fig.13(a)and(b)shows the current density distribution of the machined bottom surface during machining using grinding wheels A and B,respectively.The current density distribution was simulated using COMSOL software,in which the applied voltage was set as 25 V.According to Faraday’s law,the higher the current density is,the faster the MRR is.In Fig.13,the left circular area was the location of the bottom of the grinding wheel,and because the current density of the other area was close to zero,the circle was the main dissolution area of the machined bottom surface.The current density distribution of the circular dissolution area was uniform during machining using grinding wheel A.When using grinding wheel B,the current density was decreased from the circumference to the center of the circular dissolution area.Compared to grinding wheel A,the average current density of the circular dissolution area was lower using grinding wheel B.

Fig.11 Electrolytic reaction area of grinding wheel A.

When grinding wheel A was used for machining,because the current density distribution of the circular dissolution area was uniform and the middle of the machined bottom surface had the longest dissolution time,the amount of material removal was the most in the middle and the least on both sides of the machined bottom surface,so a concave was produced in the middle of the bottom.

When grinding wheel B was used for machining,because the bottom of grinding wheel B was a conical concave,the distance between the bottom of grinding wheel B and the machined bottom surface was increased,the current density was decreased from the circumference to the center of the circular dissolution area.Compared to grinding wheel A,the average current density of the circular dissolution area was lower using grinding wheel B;therefore,there was less difference in the amount of material removal at different locations at the bottom,and a flatter machined bottom surface could be obtained.

Fig.12 Schematic diagram of the machined bottom surface.

Fig.13 Current density distribution of the machined bottom surface during machining.

The size of the maximum machining gap and the standard deviation of the machining gap are important factors that re flect the copy precision of the machined flat bottom surface.A larger machining gap and a higher standard deviation of the machining gap both lead to a lower copy precision.Fig.14 shows a schematic diagram of the sections of machined specimens.The total depth was measured using a three-dimensional pro filometer(DVM5000,Leica,Germany),and the machining depth was set as 3 mm.The size of the machining gap is the difference between the total depth and the machining depth.

Fig.15 shows the maximum machining gap of each group.When using grinding wheel A,the maximum machining gaps of groups 5,6,7,and 8 were 371,591,695,and 846 μm,respectively,and when using grinding wheel B,they were 334,483,601,and 716 μm,respectively.A narrower machining gap was obtained using grinding wheel B than that using grinding wheel A under the same machining conditions.With an increasing applied voltage,the maximum machining gap increased and the copy precision of the machined flat bottom surface decreased.

Fig.16 shows the standard deviation of the machining gap.The way to obtain the standard deviation of the machining gap was as follows:20 points were chosen evenly on the machined bottom outline of the cross-section of each machined specimen,and the machining depth of each point was measured respectively using a three-dimensional profilometer(DVM5000,Leica,Germany).Then the machining gap of each point could be calculated,and the standard deviation of the machining gap was obtained by mathematical calculation.When using grinding wheel A,the standard deviations of the machining gaps of groups 5,6,7,and 8 were 25.07,32.98,87.66,and 145.13 μm,respectively,and when using grinding wheel B,they were 16.51,19.28,50.94,and 93.84 μm,respectively.It can be seen that the distribution uniformity of the machining gap was better when using grinding wheel B than that using grinding wheel A under the same machining conditions.With an increasing applied voltage,the distribution uniformity of the machining gap decreased and the copy precision of the machined flat bottom surface also decreased.

3.4.Sample

Fig.15 Maximum machining gap of each group.

Fig.16 Standard deviations of machining gaps.

Fig.14 Schematic diagram of the sections of machined specimens.

Fig.17 Sample machined by grinding wheel B.

Since a higher-precision flat bottom surface was machined by grinding wheel B,compared to that by grinding wheel A,a GH4169 alloy sample was machined by grinding wheel B,as shown in Fig.17.The applied voltage was 25 V,and the feed rate was 2.6 mm/min,while the other machining conditions were the same as in Table 2.The roughness of the machined bottom surface was 1.749 μm,measured by a surface roughness measuring instrument(Perthometer M1,Mahr,Germany).

4.Conclusions

(1)The copy precision of the machined flat bottom surface was improved when using grinding wheel B(the bottom of grinding wheel B was a conical concave with a depth of 1 mm)compared to that using grinding wheel A(the bottom of grinding wheel A was a flat surface),under the same machining conditions.

(2)The copy precision of the machined flat bottom surface was improved with an increasing grinding wheel feed rate under the same applied voltage,but it deteriorated with an increasing applied voltage(10–25 V)when the other machining conditions were unchanged.

(3)The maximum grinding wheel feed rate and the maximum MRR increased with an increasing applied voltage(10–25 V)when the other machining conditions were unchanged.The maximum grinding wheel feed rates obtained using grinding wheels A and B were both 2.6 mm·min-1,and the maximum MRRs obtained using grinding wheels A and B were 0.493 g·min-1and 0.431 g·min-1,respectively.

Acknowledgements

This work was co-supported by the National Natural Science Foundation of China(No.51323008)and the Funding of Jiangsu Innovation Program for Graduate Education of China(No.KYLX16_0316).

1.Xue C,Chen W.Adhering layer formation and its effect on the wear of coated carbide tools during turning of a nickel-based alloy.Wear 2011;270(11–12):895–902.

2.Ren XD,Zhan QB,Yuan SQ,Zhou JZ,Wang Y,Ren NF.A finite element analysis of thermal relaxation of residual stress in laser shock processing Ni-based alloy GH4169.Mater Design 2014;54(2):708–11.

3.Li X,Ma S,Meng FJ.Surface integrity of GH4169 affected by cantilever finish grinding and the application in aero-engine blades.Chin J Aeronaut 2015;28(5):1539–45.

4.Deng GJ,Tu ST,Zhang XC,Wang QQ,Qin CH.Small fatigue crack initiation and growth mechanisms of nickel-based superalloy GH4169 at 650 degrees C in air.Eng Fract Mech 2016;153:35–49.

5.Sun P,Sun J,Chen W,Shen X.Investigation of cutting force and tool wear in high speed milling of GH4169 using ceramic tools.Mater Sci Forum 2014;800–801:42–7.

6.Yang X,Li W,Li J,Xiao B,Ma T,Huang Z.Finite element modeling of the linear friction welding of GH4169 superalloy.Mater Design 2015;87:215–30.

7.Akhtar W,Sun J,Chen W.Effect of machining parameters on surface integrity in high speed milling of super alloy GH4169/Inconel 718.Mater Manuf Process 2016;31(5):620–7.

8.Kong X,Yang L,Zhang H,Zhou K,Wang Y.Cutting performance and coated tool wear mechanisms in laser-assisted milling K24 nickel-based superalloy.Int J Adv Manuf Technol 2015;77(9–12):2151–63.

9.Wang DY,Zhu ZW,Wang HR,Zhu D.Convex shaping process simulation during counter-rotating electrochemical machining by using the finite element method.Chin J Aeronaut 2016;29(2):534–41.

10.Xu H,Gu L,Chen J,Hu J,Zhao W.Machining characteristics of nickel-based alloy with positive polarity blasting erosion arc machining.Int J Adv Manuf Technol 2015;79(5–8):937–47.

11.Wang YL,Zhu LZ,Chen CY,Liu ZC,Ren WX,Wang L.Machining and characterization of deep micro holes on super alloy processed by millisecond pulsed laser. Key Eng Mat 2016;703:34–8.

12.Qu NS,Zhang QL,Fang XL,Ye EK,Zhu D.Experimental investigation on electrochemical grinding of in conel 718.Procedia CIRP 2015;35:16–9.

13.Lau wers B,Klocke F,Klink A,Tekkaya AE,Neugebauer R,Mcintosh D.Hybrid processes in manufacturing.CIRP Ann 2014;63(2):561–83.

14.Goswami RN,Mitra S,Sarkar S.Experimental investigation on electrochemical grinding(ECG)of alumina-aluminum interpenetrating phase composite.Int J Adv Manuf Technol 2009;40(7–8):729–41.

15.Tehrani AF,Atkinson J.A study of pulsed electrochemical grinding.In:Van Griethuy sen J,Kiritsis D,editors.International symposium for electro machining,proceedings:International Symposium for Electro Machining(ISEM-XI);1995 Apr 17-21;Lausanne,Switzerland;1995.p.543–52.

16.Puri AB,Banerjee S.Multiple-response optimisation of electrochemical grinding characteristics through response surface methodology.Int J Adv Manuf Technol 2013;64(5–8):715–25.

17.Curtis DT,Soo SL,Aspinwall DK,Sage C.Electrochemical superabrasive machining of a nickel-based aeroengine alloy using mounted grinding points.CIRP Ann 2009;58(1):173–6.

18.Kozak J,Skrabalak G.Investigations on abrasive electrochemical grinding process(AECG)Transactions on engineering technologies.Netherlands:Springer;2015.p.341–54.

19.Zhang QL,Wang H,Qu NS,Zhu D.Experimental investigation on GH4169 by electrochemical grinding.Electromach Mould 2016;4:21–5[Chinese].

20.Zhu D,Wang K,Yang JM.Design of electrode pro file in electrochemical manufacturing process.CIRP Ann 2003;52(1):169–72.

21.Hocheng H,Sun YH,Lin SC,Kao PS.A material removal analysis of electrochemical machining using flat-end cathode.J Mater Process Technol 2003;140(1–3):264–8.

22.Fang XL,Qu NS,Zhang YD,Xu ZY,Zhu D.Improvement of hole exit accuracy in electrochemical drilling by applying a potential difference between an auxiliary electrode and the anode.J Mater Process Technol 2014;214(3):556–64.

23.Qian SQ,Zhu D,Qu NS,Li HS,Yan DS.Generating microdimples array on the hard chrome-coated surface by modified through mask electrochemical micromachining.Int J Adv Manuf Technol 2010;47(9–12):1121–7.