Materials and Manufacturing Processes

ISSN: 1042-6914 (Print) 1532-2475 (Online) Journal homepage: http://www.tandfonline.com/loi/lmmp20

Synthesis of Polishing Fluid and Novel Approach

Dilshad Ahmad Khan PhD & Sunil Jha PhD

To cite this article: Dilshad Ahmad Khan PhD & Sunil Jha PhD (2017): Synthesis of Polishing
Fluid and Novel Approach for Nanofinishing of Copper using Ball End Magnetorheological Finishing
Process, Materials and Manufacturing Processes, DOI: 10.1080/10426914.2017.1328112

To link to this article: http://dx.doi.org/10.1080/10426914.2017.1328112

Accepted author version posted online: 11

Submit your article to this journal

Article views: 7

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

Download by: [The UC San Diego Library] Date: 13 May 2017, At: 01:02
Synthesis of Polishing Fluid and Novel Approach for Nanofinishing of Copper using
Ball End Magnetorheological Finishing Process

Dilshad Ahmad Khan PhD1, Sunil Jha PhD2

Address correspondence to Dr Sunil Jha PhD, Email: suniljha@mech.iitd.ac.in

Abstract

Ball end magnetorheological (MR) finishing process utilizes the magnetically control

stiffened ball of magnetorheological fluid for finishing purpose. Copper is mechanically

soft and chemically reactive material so it is difficult to finish up to nanometer order level

by traditional and most of the advanced finishing processes. In this research work the

problems associated with ball end MR finishing of copper has been explored and a fluid

composition suitable for finishing of copper has been developed. A novel approach using

two opposite magnetic pole has been used to enhance the magnetic flux density

distribution between tool tip and copper workpiece surface. The same has been

magnetically simulated and verified experimentally. The effect of fluid composition

parameters have been analyzed by the statistical model developed by response surface.

After 30 minute of finishing time nano finished surface with very few shallow scratches

has been achieved.

KEYWORDS: Ball-end, Magnetorheological, Finishing, Copper, Nanofinishing,

1
Received 08 Dec 2016, Accepted 22 Apr 2017

INTRODUCTION

Due to the advancement of nanotechnology in manufacturing field, the demand of ultra-

finished components is very high in electronics and metal optics industries. In electronics

industries, copper is used as the interconnect material which require very high level of

planarization [1-3]. In metal optics industries, copper is considered as best material for

industrial ultra-finished mirrors due to its high natural reflectivity, good thermal

conductivity, high heat capacity and durability [4, 5]. Ultra-finished copper mirrors are

used in high power laser applications like laser cutting, laser engraving and laser welding

systems for guiding the laser beam. Copper mirrors are made in flat, concave, convex or

in freeform shapes. Copper, being a soft material, is very difficult to finish by traditional

and some advanced finishing processes. This is because these processes involve high

normal finishing force which is not suitable for ultra-finishing. In loose abrasive finishing

processes like lapping, honing, polishing, etc. the high normal force leads to abrasive

getting embedded in the soft surface and a considerable polished area is covered by

abrasives [6-9].

Advanced finishing processes also have their own limitations, in chemo mechanical

polishing (CMP) the chemicals attack the finished copper surface. Moreover there are

environmental issues associated with disposal of the chemical slurry [10]. CMP is most

suitable for flat surfaces configured in two-dimensional space and finishing of three

dimensional and freeform surfaces increases the cost of the finishing. In CMP process the

2
slurry plays a crucial role in its successful completion. Sometimes the chemical slurry

needs extra care as it causes the possible corrosion of the polishing setup [11]. CMP

process become complex in nature due to the involvement of considerable number of

parameters. The performance of CMP process mainly depends on pressure applied on

polishing pad, relative velocity of the workpiece, material to be polished, slurry

composition, chemistry of polishing slurry, abrasive type, abrasive particle size, hardness

and roughness of the polishing pad [12]. In CMP process the pressure is applied on the

polishing slurry which is in between the workpiece and the polishing pad. Usually the

abrasive particles are harder than the workpiece due to which the abrasives get embedded

in the workpiece surface up to some extent. This results in particulate contamination after

CMP. In CMP process, polishing pad needs periodic conditioning as the asperities on the

pad collapses due to pressure and gets sheared by polishing slurry. Frequent or aggressive

conditioning leads to early end of life of polishing pad [13].

Single point diamond turning (SPDT) is also an option to finish soft materials like

copper, aluminum etc. but it leaves concentric tool marks at the surface which are not

desirable in high precision optics. Also by this process only axi-symmetric surfaces can

be finished [14, 15]. Magnetic assisted finishing processes such as magnetic abrasive

finishing (MAF) [6, 16], magnetorheological finishing (MRF) [17], magnetorheological

abrasive flow finishing (MRAFF) [18] also have shape limitations and are suitable either

for flat or less complex surfaces. These magnetic field assisted finishing processes are not

preferred for more complex surfaces.

3
In magnetorheological finishing, a smart fluid known as magnetorheological polishing

(MRP) fluid is used. MRP fluid is a composition of micron sized carbonyl iron particles

(CIPs), abrasives and a carrier medium like mineral oil, silicon oil or water. In the

absence of magnetic field, CIPs are dispersed randomly in MRP fluid but under the

influence of magnetic field the CIPs get magnetically polarized and align themselves

along the magnetic flux lines to form CIPs chains [17-20]. The abrasive particles are

gripped in between these CIPs chains. The gripping strength of the CIPs come from the

applied magnetic field strength. After removal of magnetic field the CIPs depolarize and

return to their random state.

Ball end magnetorheological (MR) finishing is a recently developed nanofinishing

process which is a variant of magnetorheological finishing process [21-24]. This process

utilizes a flexible ball shaped finishing tool made of magnetorheological polishing fluid

[25]. The flexible ball adjusts its compliance according to the shape in contact. This

makes the process capable of finishing complex surfaces like concave, convex, aspheric

or freeform overcoming the shape limitations of magnetic assisted processes [26, 27]. For

gentle finishing the stiffness of the MRP fluid ball can be controlled externally. Ball end

MR finishing is a deterministic process and the final surface finish depends on the

process parameters.

MATERIALS AND METHODS

Even though MRP fluid is a mixture of abrasives and CIPs in a viscoelastic medium but

when MRP fluid comes under the influence of magnetic field the CIPs move toward the

4
higher magnetic flux density region i.e. the tool tip and the non-magnetic abrasive

particles are pushed toward the lower magnetic flux region i.e. workpiece surface. This is

due to magnetic field gradient in the working gap between tool tip and workpiece. This

phenomenon ensures that at the workpiece surface there is abrasive rich MRP fluid. The

abrasive particles those are in direct contact with the workpiece surface and perform

actual cutting action are called active abrasives. During finishing, active abrasives shear

off the roughness peaks gradually (Fig.1) and impart the desired surface finish in

multiples passes.

Experimental Conditions for Trial Experiments

MRP Fluid

Due to the scarcity of literature available on MR finishing of copper a set of trial

experiments were carried out using water and oil based MRP fluid samples. The MRP

fluid samples were prepared using silicon carbide (SiC) and alumina abrasives (Al2O3) of

1000 mesh size. In all MRP fluid samples, the composition was kept as vol 20 % CIP, vol

20% abrasive and vol 60% carrier medium. In water based MRP fluid, deionized water

was used as the carrier medium whereas in oil based MRP fluid a viscous substance

containing wt 80% heavy paraffin oil and wt 20% grease was used as the carrier medium.

Copper Samples

The experiments were performed on oxygen free high conducting (OFHC) copper

samples. The workpiece samples were prepared in the form of disc of 35 mm diameter

and 6 mm thickness. These samples were prepared by lathe operations followed by

5
conventional lapping. Conventional lapping was used to remove cutting tool marks and

unevenness on the surface. The surface roughness of the prepared samples varied from 60

nm to 70 nm. Finally the prepared workpiece samples were cleaned by distilled water and

dried by high air flow. Before ball end MR finishing, the workpiece surface was cleaned

by acetone and extreme care was taken to keep the surface uncontaminated.

Machining Conditions

The machining parameters were taken as: magnetizing current = 7A, working gap

between tool tip and copper samples = 0.8 mm, tool rotational speed = 250 rpm and table

feed rate = 10mm/min. A length of 20 mm is finished along the diameter of the disc. The

finishing time was kept as 30 minutes.

Problems Overcome in Ball End MR Finishing of Copper

In ball end MR finishing of copper, two major problems have been encountered. The

observed problems and their individual solutions are discussed in the following sub-

sections.

Chemical Reaction at the Finishing Site

Trial finishing experiments were performed using water and oil based MRP fluids

containing SiC and alumina as the abrasives. A total of four combinations of base fluid

and abrasives were tested for ball end MR finishing of copper samples. After 30 minutes

of finishing it was observed that in all samples, irrespective of base fluid medium and

abrasives, the finished surface had undergone some chemical reaction and turned black

6
and brownish in color (Fig. 2). To investigate the chemical composition of these reacted

surfaces, energy-dispersive x-ray (EDX) analysis were done (Fig. 3) for all the finished

samples.

From the EDX analysis it was found that the surface had an appreciable amount of

oxygen present on it which is an indication of oxidation of nascent surface. It was

supposed that during finishing the nascent surface comes in contact with the atmospheric

air and reacted with the oxygen present in it. Black and brownish appearance on surface

might be because of the copper oxides as copper oxides exhibit the same color. It was

also observed that the surface roughness of the copper samples increased after the

finishing operation which also might be due to the formation of oxides.

In EDX spectra, the oxygen peaks are clearly visible and the analysis shows that the

oxygen varies from 2.6 to 11% by weight. So, prevention of copper from being oxidized

is the prime requirement in ball end MR finishing and for this purpose benzotriazole

(BTA) was used with MRP fluid.

BTA is a strong corrosion inhibitor that is widely used in chemical engineering to protect

the copper from corrosion. This is also used to control the pH of the polishing slurry in

CMP of copper. BTA was mixed in water as well as in oil based MRP fluids and it was

found that there is no chemical reaction on the copper surface for both types of fluids.

The images of copper samples after finishing with BTA mixed fluids are shown in Fig. 4.

7
Even though BTA mixed water-based MRP fluid is capable of finishing copper sample

but due to fast evaporation of water during finishing operations, the MRP fluid becomes

dry. This requires frequent replenishment of the water-based MRP fluid thus making the

process uneconomical and labor intensive.

So from this study it is concluded that BTA mixed oil-based MRP fluid is suitable for

copper finishing. Alumina being a soft abrasive is suitable for finishing of soft materials

such as copper and so that in next part of the research BTA mixed oil-based MRP fluid is

used with alumina as abrasive.

It was also observed that by using BTA in oil-based MRP fluid, finishing effect (reducing

Ra value) was very low and the process was not able to remove deep scratches from the

copper surface. Thus there was a need to enhance the finishing effect to make the process

fast and economical. In the next section the problem of low magnetic flux density has

been explored due to which low finishing effect was achieved.

Low and Uneven Magnetic Flux Density on Copper Sample

Due to diamagnetic nature of copper when it is exposed to a magnetic field it repels the

magnetic flux lines. In ball end MR finishing process the tool tip works as a magnetic

pole and the magnetic flux lines emerge from it. When copper sample is placed under the

tool tip the magnetic flux lines are repelled and get diverted from centre to outward in

working gap.

8
Without considering MRP fluid in the working gap, magneto-static simulation of the

magnetic flux density over copper sample was done and magnetic field lines are shown in

Fig. 5. All the simulations were carried out at 7 A magnetizing current and at 1mm

working gap. It is shown in Fig. 5 that magnetic field lines flare out in the working gap

and concentrate along the periphery of the tool tip in the form of a ring of comparatively

higher magnetic flux density. This may be due to the combined effect of repulsion of

magnetic flux lines by copper workpiece as well as magnetic edge effect experienced by

the tool tip.

Fig. 6 shows the simulation results of magnetic flux density distribution in tool tip, MRP

fluid in the gap and over the top surface of the copper workpiece. From the simulation

results it is clear that the MRP fluid has the low magnetic flux density below the tool tip

and comparatively higher magnetic flux density along the periphery of the tool tip in the

form of a ring (Fig. 6b). In ball end MR finishing of copper, only one magnetic pole

develops at the tool tip because copper is non-magnetic in nature. This is contrary to the

two opposite poles developed while finishing ferromagnetic material with ball end MR

process. In ball end MR finishing of ferromagnetic materials one pole develops at the tool

tip and it induces the other pole (opposite) at the workpiece. Due to single magnetic pole

in ball end MR finishing of copper the overall magnetic field achieved over the copper

surface is very low and insignificant for effective finishing (Fig. 6b).

It is observed that if a permanent magnet (PM) is used below the copper workpiece, it

enhances the magnetic flux density in MRP fluid as well as on the copper surface. This

9
gives uniform magnetic flux density by avoiding the formation of high flux density ring

along the periphery of the tool tip. The results of magnetic simulations for copper placed

over a permanent magnet are shown in Fig. 7. For magnetic simulation, a permanent

magnet having the same pole strength as that of the tool tip is used. Fig. 7a shows that

magnetic flux density is uniform under the tool tip without any ring of higher magnetic

flux. Fig. 7b shows that using permanent magnet below the copper sample enhances the

magnetic flux density at the copper surface too which is significant for the finishing

action.

To validate the simulation results, the magnetic flux densities at different currents (0 to

10 ampere) were measured over a copper sample without and with permanent magnet

(PM) as base (Fig. 8). The magnetic flux densities were measured with the help of a

Gauss meter, keeping a fixed gap of 1.2 mm between tool tip and workpiece. The results

show that placing a PM below the copper sample increases the magnetic flux densities

significantly at all current levels. Since the magnetic flux density depends on the

magnetizing current, so increase in magnetizing current increases the magnetic flux

density in the working gap. The increase in magnetic flux density was observed only up

to a certain level of magnetizing current. Beyond this there was not much increment in

magnetic flux density due to the magnetic saturation of the tool material as shown in Fig.

8.

The copper samples finished without and with permanent magnet base are shown in Fig.

9. The experimental conditions were kept same as that in magnetic simulations (current

7A, gap 1 mm). The experiments were performed at 300 rpm of tool speed.

10
Fig. 9a shows two finishing regions; a central region denoted by “A” has no finishing

effect whereas another region “B” with a ring shaped area of insignificant finish. Fig. 9b

shows that when copper is placed over a permanent magnet it is finished uniformly.

Difference in finishing effect on both the samples can be seen easily by the light patterns

available at the samples surfaces.

Experimentation

With the above solutions it is now possible to finish copper by ball end MR finishing

process. Strength of the ferromagnetic particles chains (yield stress) depends on the

induced dipole moment which is directly proportional to the diameter of the

ferromagnetic particle.

d 3 µp µ f
md H0 … …………………………………...(1)
2 µ p 2µ f

Where:

md= induced dipole moment

d= Particle diameter

µp= particle permeability

µf= fluid permeability

So, to increase the fluid strength further, electrolytic iron powder (EIP) of bigger mesh

size (#300) was used in place of CIP.

11
The experiments were conducted on copper disc of 35 mm in diameter and 6 mm

thickness. For experiments, the copper sample was placed over a permanent magnet of

0.5 Tesla to enhance the magnetic flux in the working gap. Other fixed process

parameters are summarized in table 1.

Abrasive and iron powder concentration in MRP fluid affect the finishing performance of

ball end MR finishing process. To analyze the effect of these two parameters, a two

factors central composite design (CCD) technique was used for the experiments. The

abrasive concentration, denoted by A, was varied from 10 to 20 volume % and EIP

concentration, denoted by B, was varied from 15 to 25 volume %. These limits were

chosen to maintain the viscosity of the MRP fluid, in which the total solid concentration

(volume of abrasive and Iron powder) should not exceed beyond 45 volume %. The

coded levels of the parameters and their actual values are shown in table 2 and 3

respectively. Percentage reduction in surface roughness (%ΔRa) is considered as the

performance measure of the process.

Table 4 shows sum of squares and % contribution of input parameters on response, which

are derived by analysis of variance (ANOVA). On the basis of sequential sum of squares

and lack of fit test it was observed that quadratic model gives good conformation with the

experimental results. Thus, to incorporate nonlinearity of the response parameter, the

following quadratic equation is derived from the experimental results:

In coded form:

% Ra 43.65 0.84*A 6.45*B 8.62*A 2 3.53*B2 2.58*A *B (2)

In actual form:

12
% Ra 319.22 24.57*A 16.20*B 0.68 A 2 0.28*B2 0.20*A *B (3)

The percentage change in surface roughness (%ΔRa) calculated on the basis of measured

values of surface roughness have reasonably good coefficient of correlation (R2 = 94.39

%). The results in terms of effect of volume % of abrasive and volume % of EIP on the

response have been computed using quadratic model (Eq 2 or 3) derived from the

analysis of variance.

RESULTS AND DISCUSSION

The effect of volume % of Abrasive and volume % of EIP on the response (%ΔRa) are

discussed in the following sub-sections:

Effect of Abrasive Concentration

The effect of abrasive concentration on the response %ΔRa has been studied at

differentEIP concentrations. From the experimental results it is observed that initially by

increasing the abrasive concentration in MRP fluid, the percentage reduction in surface

roughness (%ΔRa) increases. However beyond a certain limit of abrasive concentration,

%ΔRa decreases (Fig. 10). The reason behind this may be that increased level of abrasive

concentration in MRP fluid contains more number of abrasive particles, these increased

number of abrasive particles remove more material during finishing and thus %ΔRa

increases. But when the abrasive concentration is increased beyond a certain limit, it

decreases the magnetic permeability of the MRP fluid due to increase of non magnetic

particles. The EIP chains in the fluid of low magnetic permeability are thin, weak and

discontinuous in nature. The discontinuous chains grip the abrasive particles loosely.

13
These loosely gripped abrasive particles remove less material in finishing due to which

the percentage reduction in surface roughness (%ΔRa) decreases.

Effect of Electrolytic Iron Powder Concentration

The effect of Electrolytic iron powder concentration on %ΔRa at different abrasive

concentrations has been shown in Fig. 11. The figure shows that for abrasive

concentrations from vol. 10 to 15% as the EIP concentration is increased, %ΔRa

increases. The increase of %ΔRa with increasing EIP concentration may be due to the fact

that increasing the EIP concentration increases the magnetic permeability of the MRP

fluid. At higher EIP concentration, due to higher magnetic permeability, the EIP chains

become continuous and grow in width. These strong EIP chains aggregate in the form of

a thick and strong MRP column and result in firmly gripping of abrasive particles in

between EIP chains. The firmly gripped abrasives perform more cutting during finishing

and increase the %ΔRa. Beyond vol. 23% EIP concentration the %ΔRa stabilizes and

not much change was observed. This is due to the fact that up to 23 vol % of EIP

concentration all the available abrasives are gripped sufficiently and adding more EIP

only aggregates outside the EIP chains already gripping the abrasive.

Optimization Study

In this optimization study the goal is to maximize the percentage reduction in surface

roughness (%ΔRa). Previously developed quadratic model of %ΔRa is optimized

subjected to lower and higher values of abrasive and EIP concentrations in the design

space.

14
Maximize:

% Ra 319.22 24.57 *A 16.20*B 0.68 A 2 0.28*B2 0.20*A *B

Subjected to:

10≤A≤20

15≤B≤25

For optimization of these two fluid parameters Design Expert software was used to

calculate the optimum level of the independent parameters for the response.

The optimum values of abrasive and EIP concentrations were found at 14 vol% and 23

vol% respectively. For the validation of opmization result, one experiment was conducted

at the optimized values of abrasive and EIP concentration. The experimental value of

percentage reduction in surface roughness (%ΔRa) was found 42.34 in place of 46.59

obtained theoretically. Hence there seems to be approx 10% error. The predicted and

experimental values of the %ΔRa at optimum parametric conditions are given in table 5.

The surface roughness profile of initial copper surface and after finishing with optimum

fluid conditions are shown in Fig. 12. It is observed that the larger rouhness peaks seen

on the initial roughness profile (Fig. 12a) are sheared off significantly for 30 min of

finishing time (Fig. 12b).

The scanning electron micrographs and atomic force microscopy images of the initial and

finished surfaces finished with optimum polishing fluid composition are shown in Fig. 13

15
and 14 respectively. The SEM images show complete removal of shallow scratches while

the deep scratches present on the initial surface have turned into shallow scratches after

the finishing operation. The AFM images show the correction in the unevenness of the

initial profile to result in a better surface profile after finishing.

CONCLUSION

On the basis of experimental and theoretical work presented, the following conclusions

may be drawn:

In ball end MR finishing copper reacts with the oxygen and the surface turns

black and brownish in color due to oxide formation irrespective of type of base fluid. To

overcome this problem, BTA mixed MRP fluid has been synthesized for MR finishing of

copper. This synthesized fluid contains bigger particle sized EIP (instead of CIP) to

enhance the finishing effect on copper. Bigger size magnetizable particle (EIP) perform

better in terms of scratches removed as compared to CIP.

While finishing copper using BEMRF process a ring of comparatively higher

magnetic flux density was observed only along the periphery of the tool tip and the

overall magnetic flux density was lower at the copper surface. Two opposite magnetic

pole approach has been proposed to enhance the magnetic flux density in the working

gap. Using this novel approach a uniform and significantly higher magnetic flux density

was obtained when copper was placed over a permanent magnet with the opposite pole

facing the tool tip. With this arrangement the magnetic flux density was enhanced from

0.35 to 0.85 Tesla at 10 A current for a gap of 1.2 mm.

16
 The parametric analysis of the MRP fluid shows that increasing the abrasive

concentration upto a certain limit increase the percent change in surface roughness

(%ΔRa) but further adding more amout of abrasive decreases the magnetic permeability

of the MRP fluid due to which %ΔRa decreases. As the EIP concnetration was increased

the %ΔRa increases but beyond a certain limit (around 23 vol%) the results stabilizes due

to magnetic saturation of the fluid.

Optimization study shows that the optimum values of abrasive and EIP

concentrations for the copper finishing are 14 vol % and 23 vol % respectively. Using

optimum polishing fluid composition, in 30 min of finishing, the surface roughness was

reduced from 65.90 nm to 38 nm with 42.34% improvement.

REFERENCES

[1] Li, Y. (Ed) Microelectronic Applications of Chemical Mechanical Planarization;

Wiley-Interscience: Hoboken, NJ, 2008; 9 pp.

[2] Murarka, S.P.; Verner, I.V.; Gutman R.J. Copper—Fundamental Mechanisms for

Microelectronic Applications; John Wiley & Sons, Inc., New York, 2000; 337pp.

[3] Ein-EliY.; Starosvetsky, D. Review on copper chemical–mechanical polishing

(CMP) and post-CMP cleaning in ultra large system integrated (ULSI)—An

electrochemical perspective. Electrochimica Acta 2007, 52, 1825-1838.

[4] Wagal, SS. Diamond coated copper optics. Patent No. WO 1996024488 A1, Aug.

15, 1996.

[5] Ahmad , A. Optomechanical Engineering Handbook; CRC Press LLC: Florida,

USA, 1999; Chapter 4.

17
[6] Givi, M.; Fadaei, A.; Mohammadi, A. Polishing of the aluminum sheets with

magnetic abrasive finishing method. International Journal of Manufacturing technology

2012, 61, 989-998.

[7] Da Silva, M.F.; Shimizu, K.; Kobayashi, K.; Skeldon, P.; Thompson, G.E.;

Wood, G.C. On the nature of the mechanically polished aluminium surface. Corrosion

Science 1995, 37(9), 1511-1514

[8] Barletta, M.; Guarino, S.; Rubino, G.; Tagliaferri, V. Progress in fluidized bed

assisted abrasive jet machining ( FB-AJM ) internal polishing of aluminium tubes.

International Journal of Machine Tools and Manufacture 2007, 47, 483–495.

[9] Xie, Y.; Bhushan, B. Effect of particle size, polishing pad and contact pressure in

free abrasive polishing. Wear 1996, 200 (1-2), 281-295.

[10] Ahn, Y.; Yoon, J.; Baek, C.; Kim, Y. Chemical mechanical polishing by colloidal

silica-based slurry for micro-scratch reduction. Wear 2004, 257, 785–789.

[11] Luo, Q.; Campbell, D.R.; Babu, S.V. Chemical–mechanical polishing of copper in

alkaline media. Thin Solid Films 1997, 311, 177–182.

[12] Steigerwald, J. M.; Murarka, S.P.; Guttmann R.J. Chemical Mechanical

Planarization of Microelectronical Materials; John Wiley & Sons, Inc., New York, 1997;

36-47 pp.

[13] Han, R.; Sampurno, Y.; Philipossian, A. Fractional In Situ Pad Conditioning in

Chemical Mechanical Planarization. Tribol Letters 2017, 65(21), DOI 10.1007/s11249-

016-0803-7.

[14] Zhang, X.; Zhang, Y. Study on the surface quality of a diamond-turned reflector

used in a high-power CO 2 laser. Optical Engineering 1997, 36 (3), 825–830.

18
[15] Mahajan, K. A.; Sadaiah, M.; Gawande, S.H. Experimental investigations of

surface roughness on OFHC copper by diamond turning machine. International Journal

of Engineering Science and Technology 2010, 2(10), 5215–5220.

[16] Kala, P.; Kumar, S.; Pandey, P. M. Polishing of Copper Alloy Using Double Disk

Ultrasonic Assisted Magnetic Abrasive Polishing. Materials and Manufacturing

Processes 2013, 28, 200–206.

[17] Hirris, D.C. Hisory of Magnetorheological finishing, in Proceedings of SPIE

8016, 80160N, 2011 doi: 10.1117/12.882557

[18] Jha, S.; Jain, V.K. Design and development of magnetorheological abrasive flow

finishing process. International Journal of Machine Tools and Manufacture 2004, 44

(10), 1019–1029.

[19] Sidpara, A.; Jain, V.K. Nano–level finishing of single crystal silicon blank using

magnetorheological finishing process. Tribology International 2012, 47, 159-166.

[20] Song, W.L.; Choi, S.B.; Cai, Q.C.; Choi, J.Y.; Lee, C.H. Finishing Performance

of Magneto-Rheological Fluid under Magnetic Field. Mechanics of Advanced Materials

and Structures 2013, 20, 529-535.

[21] Singh, A.K.; Jha, S.; Pandey, P.M. Magnetorheological ball end finishing process.

Material and Manufacturing Processes 2012, 27(4), 389-394.

[22] Singh, A.K.; Jha, S.; Pandey, P.M. Performance analysis of ball end

magnetorheological finishing process with MR polishing fluid. Material and

Manufacturing Processes 2015, 30, 1482-1489.

19
[23] Singh, G.; Singh, A.K.; Garg, P. Development of magnetorheological finishing

process for external cylindrical surfaces. Material and Manufacturing Processes 2015,

DOI: 10.1080/10426914.2016.1221082

[24] Maan, S.; Singh, G; Singh, A.K. Nano-surface-finishing of permanent mold

punch using magnetorheological fluid-based finishing processes. Material and

Manufacturing Processes 2016, DOI: 10.1080/10426914.2016.1232823

[25] Saraswathamma, K.; Jha, S.; Rao, P.V. Rheological characterization of MR

polishing fluid used for silicon polishing in BEMRF process. Material and

Manufacturing Processes 2015, DOI: 10.1080/10426914.2014.994767

[26] Singh, A.K.; Jha. S.; Pandey, P.M. Design and development of nanofinishing

process for 3D surfaces using ball end MR finishing. International Journal of Machine

tool and manufacture 2011, 51, 142-151.

[27] Singh, A.K.; Jha, S.; Pandey, P.M. Nanofinishing of a typical 3D ferromagnetic

workpiece using ball end magnetorheological finishing process. International Journal of

Machine Tool and Manufacture 2012, 63, 21-31.

20
Table1. Fixed experimental parameters and conditions

Parameters Conditions

Abrasive and size Alumina, mesh size 1000

Electrolytic Iron powder (EIP) size 300 mesh size

Base fluid A composition of wt80% heavy paraffin and wt

20% grease

Corrosion inhibitor Benzotriazole

Magnetizing current 6A

Working gap 1.2 mm

Rotational speed 300 rpm

Table feed rate 10 mm/ min

Finishing time 30 min

Finishing length 20 mm along the diameter

21
Table 2. Coded levels and actual values of parameters

Parameters Levels

-1.414 -1 0 +1 +1.414

Abrasive concentration (vol%)-A 10 11.50 15 18.50 20

EIP Concentration (vol%)-B 15 16.50 20 23.50 25

22
Table 3. Plan of experiments and response

Std. Run Input parameters Initial Ra Final Ra %Δ

order order Coded levels Actual values value value Ra

(nm) (nm)
Abrasive EIP Abrasive EIP

(A) (B) (A) (B)

1 5 -1 -1 11.5 16.5 79.70 59.60 25.22

2 9 +1 -1 18.5 16.5 65.50 47.40 27.63

3 10 -1 +1 11.5 23.5 66.50 39.30 40.90

4 11 +1 +1 18.5 23.5 68.50 45.90 32.99

5 2 -1.414 0 10 20 71.60 52.50 26.67

6 3 +1.414 0 20 20 74.80 55.50 25.80

7 12 0 -1.414 15 15 61.70 45.90 25.607

8 13 0 +1.414 15 25 72.00 38.00 47.22

9 7 0 0 15 20 72.40 41.60 42.54

10 6 0 0 15 20 76.00 43.40 42.89

11 1 0 0 15 20 79.10 43.80 44.63

12 8 0 0 15 20 72.30 40.90 43.43

13 4 0 0 15 20 70.60 39.00 44.76

23
Table 4. Sum of squares (SS) and percentage contribution of process parameters on

%ΔRa

Source Sum of Squares % Contribution

A 5.66 0.60

B 332.94 35.40

A2 470.71 50.05

B2 86.57 9.21

A×B 26.64 2.83

Pure error 17.9 1.90

24
Table 5. Validation of optimization results

Parameters Predicted %Δ Ra by Experimental % Error in

quadretic model %Δ Ra predicted value

A B

14 vol% 23 vol% 46.59 42.34 -10.05

25
Figure 1. Material removal by active abrasive in BEMRF (a) active abrasive gripped in

CIP approaching the roughness peak (b) active abrasive shearing off the roughness peaks

in a single pass (c) active abrasive gradually sheared off the roughness peaks in multiple

passes

26
Figure 2. Copper samples finished by (a) SiC-water based MRP fluid (b) Al2O3-water

based MRP fluid (c) SiC-oil based MRP fluid (d) Al2O3-oil based MRP fluid

27
Figure 3. EDX spectra of copper samples finished with (a) SiC-water based MRP fluid

(b) Al2O3-water based MRP fluid (c) SiC-oil based MRP fluid (d) Al2O3-oil based MRP

fluid

28
Figure 4. Copper samples finished by BTA mixed (a) SiC-water based MRP fluid (b)

Al2O3-water based MRP fluid (c) SiC-oil based MRP fluid (d) Al2O3-oil based MRP fluid

29
Figure 5. Magneto-static simulation over copper samples showing magnetic field lines

30
Figure 6. Magnetic flux density distribution for copper (a) in MRP fluid (b) on top

surface of copper workpiece

31
Figure 7. Magnetic flux density distribution for copper workpiece on a permanent

magnet (a) in MRP fluid (b) on top surface of workpiece

32
Figure 8. Magnetic flux densities at copper worpiece at different currents without and

with a permanent magnet base

33
Figure 9. Copper samples (a) uneven and low finishing without permanent magnet base

(b) uniform finishing using permanent magnet as the base

34
Figure 10. Effect of abrasive concentration on percentage reduction in surface roughness

(%ΔRa).

35
Figure 11. Effect EIP concentration on percentage reduction in surface roughness (%Δ

Ra)

36
Figure 12. Surface roughness profile of (a) Initial surface (b) finished surface with

optimal fluid composition

37
Figure 13. Scanning electron micrograph of (a) initial surface (b) finished surface of the

copper sample

38
Figure 14. Atomic force microscopy image of (a) initial surface (b) finished surface of

the copper sample

39