Analysis of Surface Finish Improvement during Ultrasonic Assisted Magnetic Abrasive Finishing on Chemically treated Tungsten Substrate

Available online at www.sciencedirect.com

ScienceDirect Procedia Manufacturing 10 (2017) 136 – 146

45th SME North American Manufacturing Research Conference, NAMRC 45, LA, USA

Analysis of Surface Finish Improvement during Ultrasonic Assisted Magnetic Abrasive Finishing on chemically treated Tungsten substrate Nitesh Sihaga *, Prateek Kalab and Pulak M Pandeya b

a Department of Mechanical Engineering, IIT Delhi, Delhi, 110016, India Department of Mechanical Engineering, BITS Pilani, Pilani,333031, India

Abstract In the era of globalization, the demand of new products with advanced material and process technologies is increasing. Conventional manufacturing techniques are not capable to process the advanced engineering materials with stringent properties. This paper presents a novel approach to finish some advanced engineering materials with stringent properties, which is a challenge for existing conventional machining processes. In this study the positive outcomes of Magnetic Abrasive Finishing (MAF), Chemical-Mechanical Polishing (CMP), and ultrasonic vibrations have been integrated and a novel finishing process Chemo Ultrasonic Assisted Magnetic Abrasive Finishing (CUMAF) has been developed. The machining performance has been enhanced with the process resulting in better surface finish and reduced finishing time. In order to establish the process, an experimental study was done to analyze the influence of five different process variables on surface roughness of workpiece. The response surface methodology and analysis of variance was used to design the experiments and analyze the results respectively. A regression model was also developed and validated, to foresee the process response. Optimization of the model was carried out at the end to obtain the best performance. © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license © 2017 The Authors. Published by Elsevier B.V. (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of organizing the Scientific Committee NAMRI/SME. Peer-review under responsibility of the committee of theof 45th SME North American Manufacturing Research Conference Keywords: Hybrid magnetic abrasive finishing; surface finish; chemo mechanical polishing; ultrasonic vibration

*Nitesh Sihag is former M.Tech. student of IIT Delhi. She is currently doing PhD at BITS Pilani, Pilani. Corresponding author. Tel.: +919461245852; E-mail address: [email protected]

2351-9789 © 2017 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of the 45th SME North American Manufacturing Research Conference doi:10.1016/j.promfg.2017.07.040

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

137

1. Introduction Increasing customer demands and market competition is forcing manufacturing industries to produce precise products with stringent design requirements. Modern manufacturing industries such as aircraft, aerospace, medical, electronics and semiconductor, tools and dies etc. widely use advanced engineering materials such as Titanium alloys, ceramic materials, tungsten, composite materials etc. These non-conventional materials possess some superior characteristics like high wear resistance, toughness, hardness, and better strength. Finishing operations of materials having high hardness values are most precarious, sometimes uncontrollable, and costly phase of entire production process. It is a challenge for modern manufacturing industries to machine these materials precisely to complex shapes with high surface quality [1][2]. Abrasive finishing is widely adapted by manufacturing industries to achieve good quality in terms of precision, accuracy, surface integrity, and form [3]. The conventional abrasive finishing processes utilize a rigid tool that imparts high normal stress to the work-piece. This may cause microcracks in the finished surface and affect strength and reliability of the machined parts. Therefore they may be less productive in terms of time, cost and quality. This has led to development of various advanced finishing processes. MAF is widely used finishing process to obtain highly finished surfaces. In MAF process, abrasive particles are forced against the target surface using magnetic forces causing removal of very fine chips from the workpiece surface. MAF has several advantages in terms of productivity and high surface finish but it is not efficient for processing hard materials like ceramics and high carbon steels [4], [5]. Several attempts have been made to explain the mechanism of MAF and to improve its performance by integrating one or more finishing processes to it. The relevant research papers have been discussed below. Some researchers [5][6][7][8][9] studied the influence of various process variables such as working gap, voltage supplied to the magnet, magnet rotation and grade of abrasive particles on performance of MAF [6]. Singh et al. [7] conducted an experimental study to explain the mechanism of material removal and wear for magnetic field assisted abrasive flow machining. They reported that the MRR and surface quality was improved due to magnetic field for non-ferromagnetic work materials. Pashmforoush and Rahimi [10] studied the material removal mechanism of MAF on BK7 optical glass. They observed that the most significant factor to affect the process was the size of magnetic and abrasive particles. Jiao et al. [11] also conducted a study to enhance the quality and integrity of sample surface by imparting the rotary motion to the magnetic abrasive brush (MAB) and concluded that improving the trajectory of MAB led to improved surface homogeneity, surface quality and significant increase in the valid finishing region. Singh and Singh [9] performed MAF on cylindrical pipes using sintered magnetic abrasives to study the process characteristics. Lee et al. [12] combined planetary motion and vibrations within magnetic abrasive polishing to improve the efficiency of the abrasives and reported significant improvement in surface quality. Mulik and Pandey [13] successfully finished an AISI 52100 sample with ultrasonic assisted MAF. They concluded that with the application of ultrasonic vibration during MAF tangential force increased, which resulted in improved surface quality. In another study by Kala et al. [14], copper alloy was polished using ultrasonic assisted double disk MAF. They reported that magnetic flux density was increased while using two magnetic disks, which consequently increased the finishing force and improved the surface finish. Zhuo et al. [15] incorporated ultrasonic vibrations with MAF for machining titanium and reported 40% improvement in the process efficiency. Kala and Pandey integrated the ultrasonic vibrations into double disk magnetic abrasive polishing process and found that higher magnetic flux density can be attained by using magnetic poles on either side of the workpiece [16]. Yin and Shinmura applied horizontal, vertical and compound vibration during processing a SUS 304 sample using MAF and reported that providing vibration along the direction of feed and perpendicular to the feed together resulted in improved surface finish with reduced finishing time [17]. Chemo Ultrasonic Polishing is another process used to finish plane surfaces using mechanical and chemical forces together. This process integrates chemical oxidation and abrasive finishing. CMP has disadvantage of low material removal rate (MRR) when applied to hard materials. Wang Z. et al. [18] studied the influence of two main parameters namely process temperature and concentration of slurry abrasive, in the removal mechanism of CMP for tungsten. The removal rate was observed to increase linearly with the process temperature and to be linear to the cubic root of the slurry abrasive concentration. Judal and Yadava [19] proposed a hybrid abrasive finishing process which combines electro-chemical reaction with magnetic abrasive machining (C-EMAM). Forsberg [20] studied the variation in MRR with change in process parameters during CMP of Silicon. He reported that the material is

138

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

removed at higher rate at increased plate speed, applied pressure, and slurry concentration. Park et al. [21] experimentally investigated the material removal characteristics of a silicon wafer and reported that the abrasive concentration of the slurry directly influences the rate of material removal. Some researchers incorporated the effect of chemical oxidation/itching to enhance the MAF performance. Yan et al. [4] incorporated electrolytic effect into MAF and observed significant improvement in finishing characteristics at higher electrolytic current. Du et al. tried to efficiently finish hard materials by combining MAF with electrolytic polishing process (EPP) [22] and reported 50% improvement in the processing efficiency with the proposed process. The performance of finishing processes is improved with the efforts made in the literature discussed above but finishing of hard materials with low magnetic permeability is still a challenge for manufacturing industries. Further, MAF is effective when used for ferromagnetic materials as they have high permeability and therefore produces high finishing forces for effective finishing. But paramagnetic/diamagnetic materials have very low magnetic permeability and hence MAF or Ultrasonic assisted magnetic abrasive finishing (UAMAF) are not efficient to finish these materials satisfactorily. Hence, there is a need to develop an efficient finishing process to finish wide range of materials. Therefore, an attempt is made to develop a novel hybrid finishing process by combining ultrasonic vibration, MAF and CMP together to enhance finishing performance and the process is tested on copper [23]–[25]. The machining performance was enhanced with this new hybrid process resulting in better surface finish with reduced finishing time. Earlier the study was performed on copper workpiece, which is soft and paramagnetic in nature. In the present study authors have tested the newly developed Chemo-Ultrasonic Assisted Magnetic Abrasive Finishing (CUMAF) process on tungsten workpiece. Tungsten was selected because it varies significantly from copper in terms of magnetic, physical, and chemical properties. Hence, the mechanism and performance also differ significantly. By testing the process on two different types of workpieces gives more comprehensive impression about the process performance on a wide range of materials. 2. CUMAF set up design and experimentation The details of experimental set-up, magnetic tool, ultrasonic generator unit and workpiece fixture, process parameters and methodology have been discussed below. 2.1. Experimental set up

S

0 10

S

N

Magnetic flux density (mT)

30.5

݊

݊

N

25

The test set up consist of three key units, i.e. magnetic disks, ultrasonic vibration generator and fixture to hold workpiece. A disk of aluminum with four symmetrical blind holes was used and rare earth magnetic disks ĭmm x 3 mm) were inserted into every hole to convert it into the magnetic tool. The schematic arrangement of magnetic poles in the aluminum disk has been presented in Figure 1. To produce higher magnetic flux density two magnetic tools having same number of magnetic disks were taken and workpiece was placed in-between them. The magnetic flux density has been measured for both single pole and double pole system and it was found that the two pole magnetic system is more efficient. Figure 2 indicates the change in maximum magnetic flux density with total number of magnets used for both the systems. The density of magnetic flux was maximum at a distance of 2 mm from the magnetic disk surface. Two pole Single pole

450 350

250

250

150

150

Fig. 1. Representation of magnetic poles in aluminum disk

150

170

180

110 60

50 10

All dimentions in mm

430 380

330

15

20

Total number of magnets used

Fig. 2. Variation of maximum magnetic flux density with total number of magnets used for both the systems

139

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

The experimental set up fabricated to study CUMAF process is presented in Figure 3(a).The magnetic disk at top was attached to the spindle of CNC vertical milling center through a collar. A fixture was specially designed to hold the lower magnet above the machine table. The magnet was independent of the machine table movement. The second unit of the experimental set up consisted of UP 1200 ultrasonic processor to produce ultrasonic vibration. This unit comprises of a horn, piezo-electric transducer, and power supply. The horn diameter is 0.75 inches and frequency of vibration used is 20 KHz. The transducer converted high frequency electric signals generated by the ultrasonic power supply into mechanical vibrations. The amplitude achieved from the transducer was very small (of the order 4-5 microns) and therefore a horn is used to amplify the amplitude. The amplified vibrations were then transmitted to the sliding tray in the longitudinal direction which was in directly connected to the ultrasonic horn. The workpiece was fixed on the sliding tray and hence got vibrations as well. A dial indicator ZLWKOHDVWFRXQWRIȝPZDVXVHGWRPHDVXUHWKHYLEUDWLRQOHYHODWWKHHGge of workpiece. The vibration was found to be in range of 5 to 12 µm. The third key component of this experimental set up was a workpiece holding fixture specially designed for this set up. Figure 3(b) shows the schematic of designed workpiece holding fixture. It was designed to serve three functions: hold the workpiece, contain slurry and facilitate freedom to workpiece for a slight longitudinal movement. For this a rectangular plastic container was fixed on a one mm thick tray of stainless steel, which facilitates reciprocating motion in the side guides. For side guides two acrylic sheets were taken and slots were made along its length. The dimensions of grooves was such that it facilitates only the longitudinal motion of the tray and restricted the motion in other two directions. Two springs were fixed in a wooden block and the front end of sliding tray was made to rest against these springs. 2.2. Parameter selection In this study five parameters (concentration of oxidizing agent, RPM, pulse on time for vibration, weight proportion of abrasive, and gap between workpiece and upper magnet) are chosen as process variables. Small rectangular tungsten samples (Hardness 440HV) with dimensions 10mm x 15 mm x 1 mm were selected as workpieces to study the process performance. The tungsten alloy was 1.9% ferromagnetic and consisted of W (92%), Ni (5%), and Fe (Balance). In the present work, H 2 O 2 is selected as an oxidizing agent. Preliminary experiments were conducted in absence and presence of oxidizing agent and results are shown in table 1. The other four parameters were fixed at gap = 2 mm, % wt. of abrasive = 20%, Pulse on time for ultrasonic vibration (t on ) = 3 sec and RPM = 150. Mesh number of Alumina was chosen to be 1200 based on some preliminary experiments with different mesh numbers. Table 2 represents the list of constant parameters. Fixture to hold lower magnet Workpiece

Upper magnet

Ultrasonic horn Work piece

Ultrasonic horn

Side guide Sliding tray Lower magnet

Workpiece holding fixture

Fig. 3. Experimental set up (a) actual photograph; (b) Schematic of workpiece holding fixture

Springs

140

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146 Table 1. Effect of chemical oxidation on surface roughness Working gap (mm)

%weight of abrasive

1.

2

20

2.

2

20

S. No.

RPM

Conc. of H 2 O 2 (%w/w)

Initial R a

Final R a

ǻ5 a

3

150

NA

0.3411

0.2880

15.56

3

150

5

0.3329

0.1717

48.42

T on (s)

Table 2. Process parameters kept constant for present study Parameter

T off (s)

Al 2 O 3 mesh number

Ultrasonic power supply (W)

Experiment time(min)

Value

2.0

1200

720

20

2.3. Experimental Procedure In present work central composite design (CCD) methodology was used because it resulted in high accuracy for smaller number of experiments. A second order equation of following form is obtained.

Y

k

k

i 1

i 1

D 0  ¦ D i X i  ¦ D ii X i2  ¦¦ D ij X i X j  H

(1)

j i

Where, Y and k represent the response variable and number of variables respectively, Į 0 Į i Į ii DQGĮ ij are constants, X i , X i 2, and X i X j are first degree, second degree and interaction terms for process variables and H is the arbitrary error. The workpiece was exposed to H 2 O 2 for 30 minutes for uniform chemical treatment of surface. The gap between upper magnetic disk and workpiece was filled with a mixture of ferro-magnetic and abrasive particles. Hence, a flexible magnetic abrasive brush (FMAB) was formed at four poles, which serves as a flexible cutting tool. The levels of process parameters were decided by conducting preliminary experiments and have been shown in Table 3. Table 3. Level and description of process parameters Factor representation

Description

Level -2

-1

0

1

2

X1

Rotational speed (RPM)

50

100

150

200

250

X2

Working gap (mm)

1

1.5

2

2.5

3

X3

Concentration of oxidizing agent (% w/w)

1

3

5

7

9

X4

Pulse on time for ultrasonic vibration, t on (s)

1

2

3

4

5

X5

Percentage weight of abrasive (%wt)

10

15

20

25

30

The process response selected to evaluate the study is ‘percentage change in surface URXJKQHVV¨5 a )’, calculated with the formula given below: οܴ௔ =

ூ௡௜௧௜௔௟ ோೌ ିி௜௡௔௟ ோೌ ூ௡௜௧௜௔௟ ோೌ

× 100

(2)

The R a of the finished samples was measured with Talysurf. The traverse length of the instrument was 120mm and the cut-off length was selected as 0.2mm. The R a was measured at three points for every workpiece before and after finishing and then average was calculated. Based on above inputs, 33 experiments were conducted and R a

141

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

values recorded. Further, regression study was performed on the data obtained to construct a statistical model for ǻ5 a and then the model was validated and optimized. 3. CUMAF set up design and experimentation 3.1. Statistical PRGHOIRUǻ5 a The statistical model obtained by regression analysis shown by equation (3). The data was further analyzed using analysis of variance (ANOVA) after eliminating the insignificant terms.

 108  0.64 * X 1  25.4 * X 2  11.9 * X 3  4.6 * X 4  14.1 * X 5  0.001 * X 12

'Ra

 5.04 * X 22  0.31 * X 32  0.083 * X 42  1.37 * X 52  3.15 * X 2 * X 3  0.011 * X 1 * X 4

(3)

Where, X 1, X 2, X 3, X 4, and X 5 represent rotational speed (RPM), working gap, concentration of oxidizing agent, pulse on time for vibration and weight proportion of abrasive respectively. 3.2. Prediction of the predictive model There may be a slight variation between the statistically predicted and experimentally obtained ǻR a .This variation can be determined with the following equation:

'Ra _ range Where models,

'Ra _ predicted r tD 2,DF * Ve

'Ra _ range

tD 2 , DF

is the range of process response,

(4)

'Ra _ predicted

is the response predicted by the respective

is the statistical t-coefficient at a certain significance level.

'Ra _ range

'Ra _ predicted r 5.92

(5)

Further, some validation experiments were conducted and results are tabulated in Table 5. It is witnessed that the experimental results are within the limit predicted by the statistical model. Table 4. Analysis of variance for tungsten Source

DF

Seq SS

MS

F

P

R2

Remark

Regression

12

3626.74

302.228

36.98

0

93.10%

= 2.27 ‫ܨ‬൫௦௧௔௡ௗ௔௥ௗ ଴.଴ହ,ଵଶ,ଶ଴൯

Linear

5

2634.10

526.820

64.47

0

Square

5

696.18

139.236

17.04

0

Interaction

2

296.46

148.230

18.14

0

Residual error

20

163.43

8.172

Lack-of-Fit

14

59.53

4.252

0.25

0.986

Pure error

6

103.90

17.317

Total

32

3790.17

Table 5. Results of experimental runs conducted for validation of statistical model

‫ ܨ‬௥௘௚௥௘௦௦௜௢௡ > ‫ܨ‬൫௦௧௔௡ௗ௔௥ௗ ଴.଴ହ,ଵଶ,ଶ଴൯ ௦௧௔௡ௗ௔௥ௗ ‫(ܨ‬଴.଴ହ,ଵସ,ଶ଴) = 2.22 ௦௧௔௡ௗ௔௥ௗ ‫ܨ‬௟௔௖௞ି௢௙ି௙௜௧ < ‫(ܨ‬଴.଴ହ,ଵସ,ଶ଴)

model is acceptable and lack of fit is not significant

142

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

Trail

ǻ5 a

Process variables X1

X2

X3

X4

X5

Obtained by eqn. (3)

Obtained by experiment

1

200

1.5

3

2

20

35.05±5.92

37.38

2

150

2

7

4

15

64.52±5.92

62.18

3

100

2.5

9

3

25

46.23±5.92

42.26

3.3. Optimization of objective function The objective function obtained in equation (3) was optimized with Genetic Algorithm toolbox available in MatLab. The algorithm was run for 100 generations with default values set as population size = 50, scattered crossover type, adaptive feasible type of mutation method. Table 6. Optimization results for tungsten X1

X2

X3

X4

X5

ǻ5 a (%)

132

1

9

5

18

86.29

The optimization results obtained are validated by experiment and 86.30% change in surface roughness was obtained, which is in acceptable range. 4. Results and Discussion 4.1. Influence of process variables RQǻ5 a The main effect plot for process response has been shown in Figure 4 ǻR a is increasing with RPM up to a certain point because rate of shearing of workpiece surface peaks is proportional to RPM. However, above 160 RPM, ǻR a is reduced because at higher RPM, the available magnetic flux density may not be adequate to keep the magnetic abrasive particles (MAP’s) together. This may cause the abrasive particles to topple without indenting into the workpiece surface resulting in reduced finish. Also tungsten being a hard material (440Hv), requires higher indentation force for effective finishing, which can be obtained with higher magnetic flux density at lower working gap. As the working gap between upper magnet and workpiece increases, the total available magnetic force responsible for indentation decreases and hence ǻR a also decreases. Further, ǻR a increases with concentration of H 2 O 2 because at lower concentration of H 2 O 2 , the available OH- ions are not enough to react with all the tungsten atoms on top layer of workpiece sample. This results in partial oxidation over the surface and hence non uniform removal of material takes place generating a rough surface. However, as the concentration of H 2 O 2 increases, the entire layer of tungsten atoms on the surface is oxidized to form a uniform oxide layer [26]. This layer is then subsequently removed uniformly by FMAB and a flat surface with good finish is generated. Increasing t on causes an increase in the duration of impact of abrasives and results better finish [27]. Also increase in weight proportion of abrasives increases ǻR a upto a certain point. This can be explained as, at higher percentage of abrasive particles, more cutting edges will be available for shearing the surface texture peaks. However, beyond a certain limit, the iron particles in the mixture are so less that the magnetic force produced is not sufficient to hold the chain of magnetic abrasive particles (MAP). It reduces the strength of FMAB and causes reduction in ǻR a [13]. 4.2. Effect of interactLRQWHUPVRQǻ5 a for tungsten Figure 6 and 7 show the surface plots for interaction terms and variation of ǻR a . It can be noted that higher ǻR a is achieved at higher concentration of H 2 O 2 and lower working gap. This is because tungsten being a hard material requires high normal force to produce optimum indentation, which was possible at lower working gap. Also as

143

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

concentration of H 2 O 2 increases, an even oxide layer forms over the surface, which can be easily removed at relatively low force. But at higher working gap, even for higher concentration of H 2 O 2 ǻR a is not found to change significantly. This happens because at higher working gap, very low magnetic flux density is generated. The results in less indentation force which makes it harder to efficiently remove the material. Pulse on time Gap Conc. RPM^2 RPM Error % Weight Gap*Conc. RPM*%Weight % weight^2 Pulse on time^2 Conc.^2 Gap^2

0DLQ(IIHFWV3ORWIRUƩ5D Data Means

Gap

%Wt. of abrasive

Conc.

60 50 40

Mean

30 1.0

1.5 2.0 2.5 Pulse on time

3.0

10

15

20 RPM

25

30

1

3

5

7

9

60 50 40

0.00

21.99% 17.36% 15.08% 13.51% 10.86% 4.31% 4.21% 4.18% 3.64% 2.49% 1.05% 0.75% 0.57% 5.00

10.00 15.00 20.00 25.00

30 1

2

3

4

5

50

100

150

200

250

Fig. 4. Main effect plots of process parameters for tungsten during CUMAF

Fig. 5. Percentage contributions of significant factors on ǻR a for tungsten

In figure 7 it can be seen that at lower RPM, process response improves with increase in abrasive weight. However, after a particular value of RPM, ǻR a starts to decrease with increase in abrasive weight. This happens because at higher RPM, the normal force to keep MAPs together is insufficient and hence the MAPs tend to move outward. This leads to lack of indentation and ineffective cutting. Working Gap

(a)

(b) 100

70

ǻ5a

60

80

50 40

60

ǻRa

30

40

8

20

6

Conc. of H2O2

3

4

0

2.5 2

2

1.5 5 1

0

Working gap

5 10 Concentration of H2O2 (%w/w)

Fig. 6. Interaction effect of working gap and concentration of H 2 O 2 (a) Response surface of ǻR a ; (b) Variation of ǻR a with gap and concentration of H 2 O 2 (% weight of abrasive = 20, RPM = 150 and t on = 3 sec )

(a)

(b) ǻ5a

60

60 50

40 20

40

ǻRa

0 30

25

at 50 RPM at 100 RPM at 150 RPM at 200 RPM at 250 RPM

20 20

%wt of abrasive

30

250 15

10

200

0

150 10

100 50

RPM

5

15

25

35

Percentage weight of abrasive

144

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146 Fig. 7. Interaction effect of % weight of abrasive and RPM (a) Response surface of ǻR a ; (b) Variation of ǻR a with abrasive percentage and RPM (Gap = 2 mm, concentration of H 2 O 2 = 5% w/w and t on = 3 sec )

4.3. Analysis of surface quality obtained with CUMAF The atomic force microscope (AFM), and Scanning electron microscope (SEM) image profiles of sample before and after finishing as per the conditions given in Table 7 are shown in figure 8(a) and, Figures 8(b-d) respectively. It can be noted that for the rough sample the crests and valleys vary from -1.4µm to 1.2µm and after finishing with optimum conditions it vary from -0.2µm to 0.2µm resulting in 86.30% reduction in surface roughness. SEM micrographs shows that grinding marks, pits and digs on rough surface were removed after finishing with CUMAF. Table 7. Detail of each parameter selected for R a profiles shown in figure 8 Fig. No.

X1

X2

X3

X4

X5

ǻ5 a (%)

10(b)

250

2

5

3

20

31.45

10(c)

100

1.5

7

4

25

68.59

10 (d)

132

1

9

5

18

86.29

5. Conclusions The set up to conduct CUMAF experiments has been designed and fabricated to finish tungsten. The ultrasonic vibrations are found to increase the interaction of abrasive cutting edges with the crests on the surface and chemical oxidizer is useful for development of a uniform oxide layer over workpiece surface. The integration of two processes lead to improved surface finish with negligible defects. The influence of five process variables on the process response has been studied. It is reported that pulse on time of ultrasonic vibrations has the maximum effect (21.99%) on surface finish improvement. Surface finish was improved by 86.30% while finishing tungsten workpiece with CUMAF under optimum conditions. Two interaction terms have been found to be significant. Interaction effect between the working gap and concentration of H 2 O 2 has 4.18% contribution in improvement of process response whereas interaction effect between weight proportion of abrasive and RPM has 3.64% contribution in the same. Surface integrity of the surfaces had been observed with help of SEM images and AFM images. It has been shown that very fine scratches are left on the surface after finishing with CUMAF.

145

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146

X 2.000µm/div Z 700nm/div (a) Ra = 0.3351µm

X 2.000µm/div Z 700nm/div (b) Ra = 0.2297µm

X 2.000µm/div Z 700nm/div (c) Ra = 0.1052µm

X 2.000µm/div Z 700nm/div (d) Ra = 0.0459µm

Fig. 8. R a profiles, SEM images, and AFM profiles of (a) rough sample; (b-d) sample finished with conditions given in table 7

References [1] [2]

V. K. Jain, Advanced machining processes. Kanpur: Allied Publishers Private Limited, 2004. J. A. McGeough, Advanced Methods of Machining. London: Chapman and Hall Ltd, 1988.

146

Nitesh Sihag et al. / Procedia Manufacturing 10 (2017) 136 – 146 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27]

V. K. Jain, “Abrasive-Based Nano-Finishing Techniques: an Overview,” Mach. Sci. Technol., vol. 12, no. 3, pp. 257–294, 2008. B. H. Yan, G. W. Chang, T. J. Cheng, and R. T. Hsu, “Electrolytic magnetic abrasive finishing,” Int. J. Mach. Tools Manuf., vol. 43, no. 13, pp. 1355–1366, 2003. T. a. El-Taweel, “Modelling and analysis of hybrid electrochemical turning-magnetic abrasive finishing of 6061 Al/Al2O3 composite,” Int. J. Adv. Manuf. Technol., vol. 37, no. 7–8, pp. 705–714, 2008. T. C. Kanish, P. Kuppan, S. Narayanan, and S. D. Ashok, “Experimental investigations and parametric analysis of magnetic field assisted abrasive finishing of SS316L,” Int. J. Manuf. Technol. Manag., vol. 29, no. 1–2, pp. 78–95, 2015. S. Singh, H. S. Shan, and P. Kumar, “Wear behavior of materials in magnetically assisted abrasive flow machining,” J. Mater. Process. Technol., vol. 128, no. 1–3, pp. 155–161, 2002. F. Pashmforoush and A. Rahimi, “Numerical-experimental study on the mechanisms of material removal during magnetic abrasive finishing of brittle materials using extended finite element method,” J. Mech. Eng. Sci., pp. 1–13, 2015. P. Singh and L. Singh, “Optimization of Magnetic Abrasive Finishing Parameters with Response Surface Methodology,” in Proceedings of the International Conference on Research and Innovations in Mechanical Engineering, 2014, pp. 273–286. F. Pashmforoush and A. Rahimi, “Numerical-experimental study on the mechanisms of material removal during magnetic abrasive finishing of brittle materials using extended finite element method,” vol. 0, no. 0, pp. 1–13, 2015. a. Y. Jiao, H. J. Quan, Z. Z. Li, and Y. H. Zou, “Study on improving the trajectory to elevate the surface quality of plane magnetic abrasive finishing,” Int. J. Adv. Manuf. Technol., vol. 80, no. 9–12, pp. 1613–1623, 2015. Y.-H. Lee, K.-L. Wu, C.-T. Bai, C.-Y. Liao, and B.-H. Yan, “Planetary motion combined with two-dimensional vibration-assisted magnetic abrasive finishing,” Int. J. Adv. Manuf. Technol., vol. 76, no. 9–12, pp. 1865–1877, 2015. R. S. Mulik and P. M. Pandey, “Ultrasonic assisted magnetic abrasive finishing of hardened AISI 52100 steel using unbonded SiC abrasives,” Int. J. Refract. Met. Hard Mater., vol. 29, no. 1, pp. 68–77, 2011. P. Kala, S. Kumar, and P. M. Pandey, “Polishing of Copper Alloy Using Double Disk Ultrasonic Assisted Magnetic Abrasive Polishing,” Mater. Manuf. Process., vol. 28, no. 2, pp. 200–206, 2013. K. Zhou, Y. Chen, Z. W. Du, and F. L. Niu, “Surface integrity of titanium part by ultrasonic magnetic abrasive finishing,” Int J Adv Manuf Technol, pp. 997–1005, 2015. P. Kala and P. M. Pandey, “Comparison of finishing characteristics of two paramagnetic materials using double disc magnetic abrasive finishing,” J. Manuf. Process., vol. 17, pp. 63–77, 2015. S. Yin and T. Shinmura, “A comparative study: Polishing characteristics and its mechanisms of three vibration modes in vibration-assisted magnetic abrasive polishing,” Int. J. Mach. Tools Manuf., vol. 44, pp. 383–390, 2004. Z. Wang, R. Peng, S. Xia, T. Kitajima, and and S. Tsai, “Removal Mechanism of Tungsten CMP Process: Effect of Slurry Abrasive Concentration and Process Temperature,” ECS Trans., vol. 41, no. 43, pp. 103–111, 1997. K. B. Judal and V. Yadava, “Modeling and simulation of cylindrical electro-chemical magnetic abrasive machining of AISI-420 magnetic steel,” J. Mater. Process. Technol., vol. 213, no. 12, pp. 2089–2100, 2013. M. Forsberg, “Effect of process parameters on material removal rate in chemical mechanical polishing of Si(100),” Microelectron. Eng., vol. 77, no. 3–4, pp. 319–326, 2005. B. Park, S. Jeong, H. Lee, H. Kim, H. Jeong, and D. A. Dornfeld, “Experimental Investigation of Material Removal Characteristics in Silicon Chemical Mechanical Polishing,” J. Appl. Phys., vol. 48, no. 11R, 2009. Z. W. Du, Y. Chen, K. Zhou, and C. Li, “Research on the electrolytic-magnetic abrasive finishing of nickel-based superalloy GH4169,” Int. J. Adv. Manuf. Technol., 2015. N. Sihag, P. Kala, and P. M. Pandey, “Chemo Assisted Magnetic Abrasive Finishing: Experimental Investigations,” Procedia CIRP, vol. 26, pp. 539–543, 2015. N. Sihag, P. Kala, P. M. Pandey, M. T. Student, I. I. T. Delhi, and I. I. T. Delhi, “Chemo-Ultrasonic Assisted Magnetic Abrasive )LQLVKLQJௗ([SHULPHQWDO,QYHVWLJDWLRQV´QR$LPWGUSS–6, 2014. N. Sihag, P. Kala, and P. M. Pandey, “Experimental investigations of chemo-ultrasonic assisted magnetic abrasive finishing process,” Int. J. Precis. Technol., vol. 5, no. 3–4, 2015. E. Lassner and W.-D. Schubert, Tungsten: Properties, Chemistry, Technology of the Element, Alloys, and Chemical Compounds Erik Lassner and Wolf-Dieter Schubert. 1999. R. S. Mulik and P. M. Pandey, “Experimental investigations and optimization of ultrasonic assisted magnetic abrasive finishing process,” J. Eng. Manuf., vol. 225, pp. 1347–1362, 2010.