6

CHAPTER II

LITERATURE REVIEW

Abrasive flow machining (AFM), also known as extrude honing[4], is a method of

smoothing and polishing internal surfaces and producing controlled radii. A one-way

or two-way flow of an abrasive media is extruded through a workpiece, smoothing

and finishing rough surfaces. One-way systems flow the media through the

workpiece, then it exits from the part. In two-way flow, two vertically opposed

cylinders flow the abrasive media back and forth[5].The process was first patented

by the Extrude Hone Corporation in 1970[6].

The most time consuming and labour intensive segment of the manufacturing

process in today’s industry is the final finishing of complex and precision

components. This consumes as much as 5 – 15% expenditure of the overall

manufacturing process[3]. The manufacture of precision parts emphasizes final finish

machining operations, which may account for as much as 15% of the total

manufacturing costs [1,2]. Abrasive flow machining (AFM) has the potential to

provide high precision and economical means of finishing parts. Inaccessible areas

and complex internal passages can be finished economically and productively [1,2].

To finish an external surface, additional tooling is generally required to ensure that

the flow gap between the external surface and the tooling is sufficiently tight for

adequate abrasive action [2]. AFM has been likened to a semi-solid flowing file; and

perhaps its greatest advantage lies in its ability to finish (deburr, polish and radius)

complex internal passages or areas that are inaccessible to more traditional methods

such as mechanical honing.

The process is particularly useful for difficult to reach internal passages, bends,

cavities, and edges. Particular advantages are the directional finishes produced,

which optimize the flow coefficient of certain work-pieces, and the perfectly-formed,

tangentially-blended radii formed on edges, which improves high- and low-cycle

fatigue strength of certain work-pieces. In AFM, a semisolid media consisting of a

polymer based carrier and abrasives in specific proportion is extruded under

pressure through or across the surface to be machined [7,8]. The media acts as a

flexible cutting tool whenever it is subjected to any restriction as shown in Figure

2.1(a).

The complex holes, used in the punching or injection moulds, are easily finished by

WEDM process. However, the machined surfaces are full of micro cracks and craters

due to heat erosion and these defects lead to bad quality of the products in the

punching or the injection process. Although some methods have been proposed to

remove the defects[9−12], they either take long time to work or limit the machined

7

shape. Abrasive flow machining(AFM) is an effective method for the purpose of

deburring, polishing and removing the recasted layers[13−16].

Today, AFM enjoys the status of one of the best processes for finish machining of

inaccessible contours on difficult to machine components of a wide range of metallic

materials. AFM can produce surface finishes of the order of 0.05 µm. Holes as small

as 0.2 mm and edge radius from 0.025 mm to 1.50 mm can be successfully finish

machined with this process [17].

Fig.2.1 (a) Abrasive Flow Finishing medium adapts the shape of the work-piece (i)

Octogonal, (ii) Concave, (iii) Convex, (iv) Turbulated hole, and 2.1 (b) Schematic

diagram of material removal by ploughing [18]

Ravi Sankar et al[19] suggested that the chip size in AFM is far smaller (µ-chips) than

the ones obtained during machining with tools having well-defined cutting edges.

Micro to Nano level removal of material in AFM process allows production of better

surface finish, closer tolerances, and more intricate surface features.

Three modes of metal deformation so far have been identified in any abrasive

machining process which are as follows [20]:

1. Elastic deformation associated with rubbing;

2. Plastic deformation or ploughing where majority of the material is displaced

without

being removed, as shown in Figure 2.1(b);

3. Micro-cutting where removal of material takes in the form of miniature chips.

8

The occurrence of any particular mode of deformation strongly depends on the

magnitude of cutting forces acting on the individual grain, and the resulting depth of

indentation in the workpiece. In AFM, the mode of metal deformation may change

as the grain passes through the workpiece surface. Many researchers [21-25]

proposed the mode of grain-workpiece interaction and developed various theories.

Basuray[21+ and L’vov*22] proposed a model to estimate the undeformed chip

thickness at the onset of chip formation. Morochkin [24] proposed three different

regimes of grain-workpiece interactions (i.e., chip regime, plastic regime, elastic

regime). Brecker et al [25] proposed a model for the minimum depth of indentation

and minimum load on a grain required for chip formation. These theories can be

applied to AFM process considering that in AFM grain-workpiece interaction is taking

place in any one or combination of the possible modes (i.e. chipping, rubbing, and

ploughing). Bowden et al [26] have proposed a simple analysis of axial force

composed of shearing and ploughing forces. They also analysed radial force on the

single abrasive grain and the real area of contact between abrasive grain, and

workpiece surface for the case of metal to metal sliding. Gorana et al [27] conducted

Scratching experiments to simulate and acquire the knowledge about the action of

abrasive grains during the AFM process. Such Scratching experiments have been

carried out in the past by Groenou et al [28] for the processes other than AFM.

The abrasion in AFM referred to is defined as the removal of solid material from the

surface by the unidirectional sliding action of discrete particles of another material

(abrasive medium). The basic mechanism of abrasion has been the subject of many

investigators [29,30].

Khrushchov and Bavichov [29] identified two processes taking place when abrasive

grains made contact with the wearing surface:

(1) The formation of plastically impressed grooves which do not involve material

removal; (2) The separation of material particles in the form of micro-chips.

Whether chip formation takes place, or rubbing, or both depends upon the shape of

indenting particle. In particular, spherical indentors have been found to show a

change over from rubbing to at least partial chip formation when the indentation

strain (defined as the depth of indentation divided by the diameter of the indentor)

exceeds a certain value [30]. The machining action during AFM compares to a

grinding operation as medium uniformly removes material from the workpiece

surface. Shaw [31] proposed a theory of microchip formation for fine finish grinding

assuming spherical abrasive grain. Chen and Rowe [32] applied this theory for the

analysis and simulation of a grinding process. In the abrasive flow machining process,

the normal force applied to such a spherical grain will cause it to penetrate the

workpiece surface.

9

2.1 PRINCIPLE OF MATERIAL REMOVAL MECHANISM

Abrasive flow machining is complex because of the little-understood behaviour of

the non-Newtonian medium and the complicated and random nature of the

mechanical action of material removal.

Fig.2.2 Principle of material removal mechanism in two way AFM process

Commonly used AFM is Two-way AFM in which two vertically opposed cylinders

extrude medium back and forth through passages formed by the workpiece and

tooling as shown in Fig.2.2. AFM is used to deburr, radius and polish difficult to reach

surfaces by extruding an abrasive laden polymer medium with very special

rheological properties. It is widely used finishing process to finish complicated

shapes and profiles. The polymer abrasive medium which is used in this process,

possesses easy flowability, better self-deformability and fine abrading capability.

Layer thickness of the material removed is of the order of about 1 to 10 μm. Best

surface finish that has been achieved is 50 nm and tolerances are +/- 0.5 μm. In this

process tooling plays very important role in finishing of material, however hardly any

literature is available on this aspect of the process. In AFM, deburring, radiusing and

polishing are performed simultaneously in a single operation in various areas

including normally inaccessible areas. It can produce true round radii even on

complex edges. AFM reduces surface roughness by 75 to 90 percent on cast and

machined surfaces. It can process dozens of holes or multiple passage parts

simultaneously with uniform results. Also air cooling holes on a turbine disk and

hundreds of holes in a combustion liner can be deburred and radiused in a single

operation. AFM maintains flexibility and jobs which require hours of highly skilled

hand polishing can be processed in a few minutes; AFM produces uniform,

repeatable and predictable results on an impressive range of finishing operations.

10

Important feature which differentiates AFM from other finishing processes is that it

is possible to control and select the intensity and location of abrasion through fixture

design, medium selection and process parameters. It has applications in many areas

such as aerospace, dies and moulds, and automotive industries.

2.2 CLASSIFICATION OF AFM MACHINE

As mentioned earlier, AFM machines are classified into three categories: one way

AFM, two way AFM and orbital AFM. A brief discussion of the same is given below.

One way AFM process: One way AFM process [33] apparatus is provided with a

hydraulically actuated reciprocating piston and an extrusion medium chamber

adapted to receive and extrude medium uni-directionally across the internal surfaces

of a workpiece having internal passages formed therein, as shown in Fig.2.3. Fixture

directs the flow of the medium from the extrusion medium chamber into the

internal passages of the workpiece, while a medium collector collects the medium as

it extrudes out from the internal passages. The extrusion medium chamber is

provided with an access port to periodically receive medium from the collector into

extrusion medium chamber.

Fig.2.3 Unidirectional AFM process

The hydraulically actuated piston intermittently withdraws from its extruding

position to open the extrusion medium chamber access port to collect the medium

in the extrusion medium chamber. When the extrusion medium chamber is charged

with the working medium, the operation is resumed.

Two-way AFM process: Two way AFM machine [34] has two hydraulic cylinders and

two medium cylinders. The medium is extruded, hydraulically or mechanically, from

the filled chamber to the empty chamber via the restricted passageway through or

past the workpiece surface to be abraded, as illustrated in Fig.2.4. Typically, the

medium is extruded back and forth between the chambers for the desired fixed

number of cycles.

11

Fig.2.4 Two-way flow process illustration

Counter bores, recessed areas and even blind cavities can be finished by using

restrictors or mandrels to direct the medium flow along the surfaces to be finished.

Orbital AFM process: In orbital AFM[35], the workpiece is precisely oscillated in two

or three dimensions within a slow flowing ‘pad’ of compliant elastic/plastic AFM

medium, as shown in Fig.2.5. In Orbital AFM, surface and edge finishing are achieved

by rapid, low-amplitude, oscillations of the workpiece relative to a self-forming

elastic plastic abrasive polishing tool. The tool is a pad or layer of abrasive-laden

elastic plastic medium (similar to that used in two way abrasive flow finishing), but

typically higher in viscosity and more in elastic.

Fig.2.5 Orbital AFM (a) before start of finishing, (b) while finishing

12

Orbital AFM concept is to provide translational motion to the workpiece. When

workpiece with complex geometry translates, it compressively displaces and

tangentially slides across the compressed elastic plastic self-formed pad (layer of

viscoelastic abrasive medium) which is positioned on the surface of a displacer which

is roughly a mirror image of the workpiece, plus or minus a gap accommodating the

layer of medium and a clearance. A small orbital oscillation (0.5 to 5 mm) circular

eccentric planar oscillation is applied to the workpiece so that, at any point in its

oscillation, a portion of its surface bumps into the medium pad, elastically

compresses (5 to 20%) and slides across the medium as the workpiece moves along

its orbital oscillation path. As the circular eccentric oscillation continues, different

portions of the work piece slide across the medium.

Ultimately, the full circular oscillation engages each portion of the surface. To assure

uniformity, the highly elastic abrasive medium must be somewhat plastic in order to

be self-forming and to be continually presenting fresh medium to the polishing gap.

For finishing applications, AFM medium allows the use of a simple arrangement for

feeding and evacuating the abrasive medium pad to achieve uniform results. Regions

of the medium pad that overly fill the gap generally get pushed aside and are shaped

by the oscillation of the workpiece itself. Regions of medium in the gap that are

worked excessively become warmer, due to deformation heating, and consequently

become less elastic and more plastic and are squeezed out of the work gap. Orbital

AFM’s small (0.5 to 5 mm) oscillation amplitude allows finishing highly complex

geometries, since all areas except internal features that are even smaller than the

oscillation amplitude are equally worked in the process. The controlled and

cushioned, but still repeated, bumping of the workpiece against the self-shaped tool

imparts beneficial residual compressive stresses to the workpiece surfaces. The

tangential translation of the workpiece across the elastically compressed and

cushioned abrasive particles provides remarkable improvements in surface

roughness. Orbital AFM can be applied to many different workpieces from many

different industries from precision ground aerospace components to cast aluminum

wheels. Coining dies used to make proof coins can be polished from a 0.5 μm before

surface to an amazing 0.01 μm after finish after only seven minutes of Orbital AFM

processing. Orbital AFM is used to produce extremely fine finishes on the complex

geometry of prosthetic devices while maintaining critical dimensional tolerances.

Beverage container blow moulds are finished using the Orbital AFM process

dramatically reducing polishing costs while, at the same time, improving consistency,

increasing production rates, and reducing the need for skilled labour.

Rhoades [36] noted that Orbital AFM is one of the modifications as opposed to the

primitive AFM process where the relative motion between workpiece and displacer

is provided to obtain a positive displacement of the viscous abrasive medium to

finish the complex blind cavities. This movement can be gyratory, orbital,

13

reciprocatory or any combination of these with or without the combination of rotary

motion.

Liang Fang [37] studied abrasive particle movement pattern as an important factor in

estimating the wear rate of materials, especially, as it is closely related to the

burring, buffing and polishing efficiency of the abrasive flow machining (AFM)

process. There are generally two kinds of particle movement patterns in the AFM

process, i.e. sliding–rubbing and rolling. In mechanism, AFM grain–workpiece

interaction is taking place in any one or a combination of the possible modes:

elastic/plastic deformation by sliding–rubbing grain movement; elastic/plastic

deformation by rolling grain movement; chip formation (micro-cutting) by rubbing

grain movement; ridges formation by rubbing and rolling grain movement; and low-

cycle fatigue wear. Therefore, the machining efficiency of a machine part is

predominantly dependent upon the particle movement patterns. Fang [37]

investigated normal load, particle size and hardness of machine parts to understand

the involved parameters of particle movement patterns and proposed a computer

statistic prediction of particle movement patterns. It has been found that there are

two cases. In case of large-sized particles, the ratio of rolling particles is increased

with increasing normal load. For small-sized particles, the ratio of grooving particles

is increased with increasing normal load and vice versa. When normal load is light,

the particle size cannot usually give an effect on movement patterns. That influence

will be predominant under heavy normal load. Most of the particles will tend to

groove when the particle size is below a certain value. Hardness of the material and

their hardness difference for tribological pairs are other important monitors in

predicting particle movement patterns. In this research, increasing hardness of

materials results in more rolling particles, which results in much less cutting

particles.

Extrude Hone Corporation, USA, originally developed the AFM process in 1966[5].

Since then, a few empirical studies [38-41] have been carried out and also research

work regarding process mechanisms, modelling of surface generation and process

monitoring of AFM was conducted.

The MR in AFM can be considered as a separation of target material when a hard

material (abrasives) slides over the former. There are three classical theories which

describe the MR process when a metal is worked against a fine abrasive viz. flow

mechanism, mechanical abrasion, and molecular removal, Jain and Jain [42]

hypothesized that in AFM, the MR occurs as a result of some processes like

microplowing, microcutting, microcracking, etc. A theoretical model on this basis has

been presented considering spherical shape of abrasive particle and certain other

restrictions like constant load on the abrasive grains, same penetration depth for

every grain, etc. Kato et al [43] suggested that Scanning Electron Microscopy (SEM) is

14

one of the most effective methods for investigating the wear behavior of materials.

The information collected from SEM photographs was correlated with load and

friction force acting during wear. It was reported that abrasion of metals occurs by

three modes, i.e., cutting, wedge forming, and ploughing. Loveless et al [13]

compared SEM photomicrographs of surfaces machined by AFM and Grinding and

reported that there are many similarities in the mechanism of MR by both the

processes. Singh et al [44] reported the evidence of metal smearing in AFM for

ductile materials.

Di Ilio et al [45] suggested the application of AFM in order to meet the

manufacturing demands of finishing new generation composites viz. Al alloy/SiC

MMC produced through almost in their final shape, to lower values of tolerances.

Jain [18] reported that micro-/nano-machining (abbreviated as MNM) processes are

classified mainly in two classes: traditional and advanced. Majority of the traditional

MNM processes are embedded abrasive or fixed geometry cutting tool type

processes. Conversely, majority of the advanced MNM processes are loose flowing

abrasive based processes in which abrasive orientation and its geometry at the time

of interaction with the workpiece is not fixed. There are some MNM processes which

do not come under the abrasive based MNM category, for example, laser beam

machining, electron beam machining, ion beam machining, and proton beam

machining. Jain [18] also proposed a generalized mechanism of material removal for

these MNM processes which include: Abrasive Flow Finishing (AFF), Magnetic

Abrasive Finishing (MAF), Magnetorheological Finishing, Magnetorheological

Abrasive Flow Finishing, Elastic Emission Machining (EEM) and Magnetic Float

Polishing. EEM results in surface finish of the order of sub-nanometer level by using

the nanometer size abrasive particles with the precisely controlled forces. Except

two (AFF and EEM), all other processes mentioned above use a medium whose

properties can be controlled externally with the help of magnetic field. This permits

to control the forces acting on an abrasive particle hence the amount of material

removed is also controlled. This class of processes is capable to produce surface

roughness value of 8 nm or lower. Using better force control and still finer abrasive

particles, some of these processes may result in the sub-nanometer surface

roughness value on the finished part. Understanding the mechanism of material

removal and rotation of the abrasives in these processes will help in rationalization

of some of the experimental observations which otherwise seem to be contradicting

with the established theories.

2.3 MAJOR AREAS OF EXPERIMENTAL RESEARCH IN ABRASIVE FLOW FINISHING

With the perspective of developing better insight of the subject and to explore the

current status of the charismatic technique, which has opened up new vistas for

finishing difficult to machine materials with complicated shapes which would have

15

been otherwise impossible. These processes are emerging as major technological

infrastructure for precision, meso, micro, and nano scale engineering. This review

provides an insight into the fundamental and applied research in the area and

creates a better understanding of abrasive flow finishing process, with the objective

of helping in the design stage of our study.

2.3.1 PROCESS PARAMETERS AND THEIR INFLUENCE ON QUALITY

CHARACTERISTICS

Experimental investigations have been carried out by various researchers to

investigate the effects of process parameters like extrusion pressure, number of

cycles, viscosity, abrasive concentration and grain size on the output responses

namely, surface finish and material removal during AFM. The controllable input

parameters are shown in Fig.2.6.

The literature on AFM [46 - 48] indicates that several research efforts have been

made in the past to understand the underlying process mechanism. Przyklenk [49]

presented results of experimental work regarding the effect of AFM parameters on

material removal, surface roughness and edge radius of different work specimens.

Loveless et al[50] reported the details of the surfaces finish-machined by AFM with

the help of scanning electron microscopy. Jain and Adsul [51] also presented the

SEM photographs of AFMed surfaces obtained under different process conditions.

Machining parameters of AFM and the rheological properties of the abrasive

medium are two key factors that will affect the efficiency in the polished process.

The surface precision can be controlled by changing the AFM parameters (such as

number of cycles, concentration of the abrasive, abrasive mesh size and medium

flow speed) when the complex hole is polished [15, 16, 51]. Moreover, the medium

viscosity and extrusion pressure will significantly affect the material removal and the

surface roughness of AFM [16, 52]. The rheological properties of the abrasive media

have also been studied by some researchers [53, 54], the experiments showed that

not only the temperature could seriously influence the viscosity of the medium but

also a small increase in the temperature would drastically reduce the medium

viscosity in AFM. The results also present that the medium viscosity increases with

the abrasive concentration but decreases with the abrasive size. Silicone rubber (a

kind of polymer gel) with high viscosity and low flow rate is a good abrasive medium

that can easily polish the WEDM surface to a smooth finish [55]. Furthermore, a new

finishing method by applying a magnetic field around the workpiece was proposed

to enhance the material removal rate and the surface roughness in AFM [56, 57].

16

Fig.2.6 Classification of major AFM research areas[58]

17

Rhoades [1, 2, 59] experimentally investigated the basic principle of AFM process

and identified its control parameters. He observed that when the medium is

suddenly forced through restrictive passage then its viscosity temporarily rises.

Significant material removal is observed only when medium is thickened. The

amount of abrasion during AFM depends on design of tooling, extrusion pressure,

medium viscosity and medium flow volume. All these parameters ultimately change

the number of particles interacting with the workpiece and the force acting on

individual abrasive grain. A higher volume of medium flow increases number of

interacting abrasive grains with the workpiece, hence more abrasion takes place.

Number of cycles depend on the velocity of medium, during a given time period.

Flow pattern of medium depends on its slug (medium exiting the workpiece) flow

speed, medium rheology and passage size (cross-sectional area). AFM can be used in

industrial applications such as precision deburring, edge contouring, surface finish,

removal of thermal recast layers, etc. [60]. Williams and Rajurkar [52] used the full

factorial experimental design to study the effect of medium viscosity and extrusion

pressure on metal removal and surface roughness. Medium’s viscosity effect is more

significant on material removal as compared to extrusion pressure. It is also reported

that major change in the surface finish is observed after finishing for a few cycle

only.

Jain and Adsul [51] reported that initial surface roughness and hardness of the

workpiece affects material removal during AFM process. Material removal and

reduction in surface roughness value are reported higher for the case of softer

workpiece material as compared to harder material. Material removal and reduction

in surface roughness increase when percentage concentration of abrasive in the

medium increases. They also concluded that among all the process parameters

studied, the dominating one is the abrasive concentration followed by abrasive mesh

size, and number of cycles. It was also reported that with higher abrasive mesh size,

both MR and improvement in ∆Ra value decrease.

Loveless [50] reported that the type of machining operation used to prepare the

specimen prior to AFM is important and affects the improvement achieved during

finishing. As compared to the turned and milled surfaces, WEDM’d surfaces are

found to be more suitable for AFM. The amounts of material removal from the

WEDM’d and milled surfaces are significantly different from that of turning and

grinding, because these machining processes produce different micro surface

contours. Davies and Fletcher [61] reported a relationship between the number of

cycles, temperature and pressure drop across the die for the given type of polymer

and abrasive concentration. Increase in temperature results in decrease in medium

viscosity and increase in volumetric flow rate. With increase in processing time,

medium temperature increases that causes a change in medium viscosity. They

18

concluded that rise in temperature is due to a combination of internal shearing of

the medium and finishing action of the abrasive grit.

Williams and Rajurkar [46, 62] showed that media viscosity and extrusion pressure

significantly determine both surface roughness and the material removal rate. The

authors indicated that the major improvement insurface finish takes place within the

first few cycles. Their later work proposed methods to estimate thenumber of

dynamic active grains involved in cutting and the amount of abrasive grain wear per

stroke. Williams et al [63] presented an experimental and qualitative analysis of the

distribution of metal removal in finishing applications and reported metal removal

and surface roughness characteristics per cycle for a single hole part and found that

the most pronounced change in the bore diameter and surface roughness occurred

on the first cycle.

Uhlmann et al [64] investigated the fundamental principles of AFM on advanced

ceramic materials such as a correlation between flow processes, surface formation

and edge rounding. Furthermore, an insight into a process model is given which was

developed using modern simulation techniques. The overall objective of this

approach was to anticipate work results like surface quality and edge rounding on

any user-defined geometry. The material removal process behaviour of ceramic

materials is mainly ductile, hence the smooth surface is extending. As a result of the

machining process, typical washed-out surface textures occur. Grain boundaries as

well as edges of micro cracks have been smoothed out. By removing the surface

layer by layer, existing damages beneath the surface have been uncovered with

AFM. Only when a failure of the grain boundary is approached, grains are breaking

off. The investigation shows that for advanced ceramics a ductile material removal

mechanism is achieved by using abrasive diamond grains with average grain size

under 44,5 micron, whereas a brittle material removal mechanism could be achieved

with grain size upon 185 micron. It has been proved that in most cases a weight ratio

of 1:2 between carrier fluid to abrasive grain is attained to success. Rising

temperature leads to descending viscosity of the grinding medium, hence the

abrasive removal rate is sinking. By raising the processing pressure and reduction of

the flow cross-section, the fluid velocity and the removal rate are increasing.

Rajeshwar et al [65] conducted the simulation tests concerning the flow of a non-

Newtonian fluid around an edge, which were confirmed experimentally, and

reported a conclusion that a linear relationship exists between the pressure in the

working chamber and velocity of the abrasive medium flow.

Dabrowski et al [66] observed material removal comes as a result of micro-cutting

process affected by abrasive grains being component of the abrasive paste. The

intensity of this process is conditioned by pressure in the abrasive media. Pressure

has to be adjusted to the machined part design. Fragile parts can be machined while

19

applying insignificant pressure on the internal or external surfaces and robust parts

allow for high pressure flow up to 10 Mpa.

Abrasive Flow Machining (AFM) process was applied in industry as early as in the

seventies [67, 68]. The AFM filled technological gap in production of parts of

complex shape, having important and not easily accessible surfaces or edges. It

became available due to specific machining system and non-conventional abrasive

medium, a paste with viscous plastic polymeric carrier as a basic component. The

machining cycle has to be planned in accordance with the technological

requirements and it can include many abrasive one-way or two-way flow runs.

Simulation research using non-Newtonian fluid [69] showed that for forcing paste

through the hole of 29 mm in diameter there is linear relationship between applied

pressure and speed of flow [69].Uhlmann and Szulczynski [70] showed that there is

non-linear relationship between number of flow cycles, pressure and productivity . It

has been experimentally confirmed. Such high pressure of forcing viscous elastic

medium through holes and slots of various cross-sections considerably changes

condition of flow due to different friction forces. This results in local elevation of

paste temperature and consequently viscosity and productivity of the paste is

lowered. Hence, additional expenses have to be allocated to cool down the

machined parts. Because of the above, the pastes of low viscosity could not be

applied in conventional AFM because of their limited abrasive machinability. This

disadvantage was eliminated when new types of paste were applied. Their viscosity

was low but their behaviour was similar to that of electrolytes applied in

electrochemical machining.

Abrasive flow finishing (AFF) is a kind of advanced finishing process that can perform

operations such as polishing, deburring, radiusing, and removing recast layers that

are produced by electric discharge machining (EDM) [1, 2, 50, 59, 60, 71]. With small-

bore diameter of workpiece geometry, more grains come in contact with the

workpiece surface; hence, material removal rate (MRR) increases [49]. MRR is high in

the first few cycles due to higher initial coarseness of workpiece surface, and

thereafter, it starts slightly decreasing in every cycle [51]. Percentage of abrasives in

the medium grain size and viscosity of base medium are important parameters that

influence stock removal and medium velocity [54]. Depth of penetration of abrasive

particle depends on extrusion pressure, abrasive medium viscosity, and grain size.

Due to the combined effect of radial force and axial force, the material is removed in

the form of microchip [27, 72]. To enhance the performance of AFF, magnetic field

has been applied to the AFM process [44, 56]. Researchers rotated the medium by

using different shaped rods to improve finishing ability, to increase number of

dynamic active grains and to achieve better circularity of a finished cylindrical part

[73, 74]. Some researchers planted spiral-fluted screw in the medium flowing path to

improve surface quality [75]. Some researchers have presented mathematical

20

models for computer simulation of AFF process while finishing cylindrical workpieces

[65]. Models have been proposed to predict radial stresses at the workpiece surface

and material removal rate during abrasive flow finishing [42]. AFF process modelling

has been done using neural networks and genetic algorithms. The predicted

machined surface quality and MRR results were found to agree well with the results

obtained using neural network model [76]. Modeling of the abrasive flow finishing

using a cascade-correlation neural network approach was studied and proved that it

is a better method compared to the back propagation technique [77].

If processing oil content in the medium increases, then its viscosity decreases and

abrasive particles get loosely bounded with the base medium (polymer). If

comparatively, a low viscosity medium flows over the workpiece surface, some of

the abrasive grains may slide instead of indenting in it [78]. Therefore, as the oil

content increases, the percentage improvement in surface finish decreases in both

the cases.

It is important to know cutting force components and active grain density during

abrasive flow machining (AFM) as this information could be used to evaluate the

mechanism involved in AFM. The results show that cutting force components and

active grain density govern the surface roughness produced during AFM process. In

this paper, an attempt has been made to study the influence of these two

parameters, namely cutting force and active grain density, on the surface roughness.

This study will help in developing a more realistic theoretical model.

Several theories [79-81] have been put forward to explain the mechanism of

abrasion by abrasive particles. Solid particle erosion proposed by Finnie [79] can be

considered as the basic mechanism of material removal in AFM with some

modifications. In abrasive jet machining the energy of the striking abrasive particle is

imparted by the high speed of the medium stream but in AFM the required energy

to the abrasive particles is provided by high pressure acting on the viscoelastic

carrier medium. The medium dilates and the abrasive particles come under a high

level of strain due to the pressure acting in the restriction. The momentum that

abrasive particles acquire due to these conditions can be considered to be

responsible for microploughing and microchipping of the surface in contact with the

abrasive. Microploughing causes plastic deformation on the surface of the metal.

Initially no material removal takes place. However, the surface atoms become more

vulnerable to removal by subsequent abrasive grains. More abrasive particles attack

the surface repeatedly, which causes the detachment of merial often referred to as

‘cutting wear’.

The simultaneous increase in MR and ∆Ra indicates the unique behaviour of AFM

when compared with other machining processes[56] and these results support the

findings reported by Williams and Rajurkar [82]. One possible reason could be that,

21

in AFM, the material removal takes place first from hills or peaks of the surface

profile. More material removal produces a smoother surface. In other words, the

more material removal the smaller is the height of hills on the surface, and hence

the lesser is the roughness of the surface. This holds good until all of the high hills

are removed and quite a smooth surface is produced.

Gorana et.al [72] developed a suitable two-component disc dynamometer for

measuring axial and radial force components during AFM. The influence of three

controllable variables (extrusion pressure, abrasive concentration and grain size) on

the response (material removal, reduction in surface roughness (Ra value), cutting

forces and active grain density) are studied. Gorana et al reported from preliminary

experiments a high percentage reduction in surface roughness was obtained for 80

to 220 Grain sizes, beyond 180 Grain size, the behaviour of percentage reduction in

surface roughness became erratic. They further reported a linear relationship

between percentage reduction in surface roughness and extrusion pressure, with

increase in percentage reduction in surface roughness with increase in extrusion

pressure, considering that both the axial and radial forces increase with increase in

extrusion pressure. Furthermore, the main effects of extrusion pressure, grain size

and abrasive concentration on average percentage reduction in surface roughness

were all positive. These findings were collaborated by Jain and Jain[83] who have

also reported that reduction in surface roughness increases with increase in

extrusion pressure and abrasive concentration, but had also observed that reduction

in surface roughness was higher with increase in average grain size (Mesh Number).

Williams and Rajurkar [84] also reported that extrusion pressure and grain size main

effects are significant.

Hsinn-Jyh Tzeng [85] conducted an investigative methodology based on the Taguchi

experimental method for the micro slits of biomedicine was developed to determine

the parameters of AFM, including abrasive particle size, concentration, extrusion

pressure and machining time. The parameters that influenced the machining quality

of the micro slits were also analyzed. Furthermore, in the shape precision of the

micro slit fabricated by wire-EDM and subsequently fine-finished by AFM was also

elucidated using a scanning electron microscope (SEM).

Several researchers have studied the applications of polishing surfaces by abrasive

flow machining. Loveless et al [50] pointed out the results of an investigation of the

effects of AFM on surfaces produced by turning, milling, grinding, and wire-EDM. In

particular, all of the wire-EDM surfaces were improved greatly using AFM. Haan et al

[48] studied the effect of AFM on surface finishing. Their work indicated that AFM

exhibited a wide variation of MRR as various levels of the machining parameter were

set to various levels. Petri et al [86] examined the machining characteristics of AFM.

They attempted to develop a process modeling system for AFM to predict surface

22

finishing and dimensional modification. Jain et al [83] presents the improvement of

the surface resulting from machining and the removal of the material based on the

parameters: machining time, extrusion pressure, concentration, abrasive grains, and

decreasing rate. Adsul et al [51] also showed that the surface roughness

improvement and the material removal rate have a direct rational relation by using

AFM method. Jain et al [87-89] established the model and selected the optimum

parameters by using a neural network process on the AFM. Moreover, the specific

energy and tangential force according to machining parameters of AFM were also

drawn. Jain et al [54] elucidated that, in AFM, abrasive medium can remove the

material and promote the surface smoothness respectively, with variety of viscosity,

concentration, abrasive grain size and working temperature. Singh et al [56,90]

determined the effect of magneto abrasive flow machining on the surface roughness

and the removal of the material. The parameters of material removal were

optimized and the surface roughness was reduced by applying the magneto abrasive

flow machining process in the Taguchi experimental design and ANOVA. Jain et al

[91] investigated the effect of variation in the viscosity on the removal of material

and the surface roughness by simulating the AFM with finite element. Gorana et al.

[72] studied the effect of extrusion pressure, concentration and abrasive particle size

on the material removal, the surface roughness, the cutting forces and the active

grain density using AFM. The above studies address the machining of the surfaces of

holes with large radii, rather than micro meter orifices.

AFM is an effective method to deburr, polish and remove the recast layers by wire

electrical discharge machining (WEDM) [13, 14, 92]. The surface precision can be

controlled by changing the AFM parameters (such as number of cycles,

concentration of the abrasive, abrasive meshes size and medium flow speed) when

the complex hole is polished [51,92]. The material removal and surface roughness of

AFM are significantly affected by the medium viscosity and extrusion pressure [52,

92].

Sehijpal Singh et al.[44] reported the results of an experimental study conducted

with the objective to understand the mechanism of material removal (MR) and the

wear behavior of some materials when processed by AFM and magnetically assisted

abrasive flow machining. Scanning electron microscopy (SEM) has been used to gain

insight into the underlying wear pattern on the surfaces of different materials. The

results suggest that the magnetic field has a strong effect on the MR in AFM.

Furthermore, the nature of work material plays an important role in controlling the

MR on the surface.

Sehijpal Singh et al.[93] studied the mechanism of material removal (MR) in Abrasive

Flow Machining (AFM) process . Representative components of pure Aluminum and

Brass were processed by AFM under similar process conditions. The processed

23

surfaces were analyzed with the help of Scanning Electron Microscopy (SEM). SEM

photographs reveal noticeable difference between abrasion patterns produced on

the processed surfaces of both the materials. A mechanism of MR has been

proposed by examining the nature of interaction between the flowing abrasive

medium and target work surfaces of selected materials.

Fang et al.[94] investigated the work efficiency, considered as most concerned

target, in abrasive flow machining (AFM) and reported many influence factors, such

as, temperature, media viscosity, abrasive hardness, particles sharpness and density,

workpiece hardness, pressure, piston moving speed, etc. The influence of

temperature on work efficiency is most critical. It has been shown from AFM tests

that media viscosity decreases continuously with increasing temperature. Media

temperature increases with increasing cycles, which means media viscosity

decreases with cycles increasing. AFM tests shows that increasing cycles extensively

decrease materials removal and surface roughness decreasing efficiency. When

media with different viscosity is used, media with high viscosity has more effective

material removal efficiency. The high viscosity media to surface roughness

improvement is also better than the low viscosity media at the initial several cycle

numbers. With further increasing cycles the roughness improvement difference

among different media with different viscosity is reduced. It is found from Mooney

viscosity–temperature relation of media that temperature rising directly results in

the decrease of media viscosity. When work cycles are increased the media

temperature is quickly increased. The media viscosity is also decreased dramatically.

In order to understand the mechanism of decrease of material removal efficiency

with temperature, computational fluid dynamics (CFD) approach was applied to

predict the abrasive particles movement tendency and a two-dimensional model was

constructed for AFM process. The simulation results showed that the temperature

rising of media results in increasing the rolling tendency of abrasive particles which

causes work efficiency deteriorated.

Many researchers, Shaw [95] and Lal [96], have evaluated force ratio in grinding to

correlate it with the percentage change in surface roughness value of the machined

workpiece surface. It has given a satisfactory correlation. Hence, it was envisaged

that a similar correlation may give some useful insight into the AFM process.

Therefore, in AFM the force ratio is estimated as the ratio of radial force to axial

force.

Fang et al.[97] reported abrasive particle movement pattern as an important factor

in estimating the wear rate of materials, especially, as it is closely related to the

burring, buffing and polishing efficiency of the abrasive flow machining (AFM)

process. There are generally two kinds of particle movement patterns in the AFM

process, i.e. sliding–rubbing and rolling. In mechanism, AFM particle–workpiece

24

interaction is taking place in any one or a combination of the possible modes:

elastic/plastic deformation by grooving particle movement; elastic/plastic

deformation by rolling particle movement; chip formation (micro-cutting) by

grooving particle movement, ridge formation by grooving and rolling particle

movement, and low-cycle fatigue wear. Grooving particle movement pattern has a

greater contribution to wear mass loss of workpiece than rolling mode. Considering

the machining efficiency of a machine part is predominantly dependent upon its

wear mass loss speed, it can be concluded that particle movement patterns are key

parameters to machining efficiency in AFM. Fang et al.[97] investigated ellipsoidal

particles to understand particle movement patterns. An analytical model of

ellipsoidal geometry to determine particle movement patterns in AFM is proposed

with given particle ellipticity, normal load, particle size and material hardness. From

the analytical model and particle movement pattern criterion proposed by the

present authors, a statistic prediction of particle movement patterns is completed by

computer programmed by C++ language. It is found that a seat position of ellipsoid is

an easy grooving position for a particle and a large ellipticity value predominantly

increases grooving particle numbers. Smaller workpiece hardness, larger particle

radius and higher normal load promote grooving of the particles. Sharper particles

are much more easy to groove; moreover, grooving pattern will be predominant if

particle ellipticity is below 0.8. Increasing workpiece hardness tends to decrease

grooving regime while other parameters are fixed in AFM process. In three-body

abrasion, hard material paired with soft material will result in more rolling particles.

Abrasive contour and material hardness in many variables are two predominant

parameters to give distinct influence on particle movement pattern.

Dong et al [98] based his study on the equations of motion and rheological theory of

abrasive flow in the slit, analysed the effect between wall slip and grinding and

developed the grinding flow model of abrasive flow. Wall pressure, wall shear stress,

wall slip and the grinding force equation of abrasive flow were derived using the

model. The results show that wall slip is one of the necessary conditions in abrasive

flow machining, and show that main factors influencing wall slip are viscosity

coefficient , grinding coefficient of abrasive flow, and the first normal stress

difference. In the meantime, the results also show that grinding force in slit wall is

proportional to the entrance pressure of slit. That is to say that the grinding force of

per unit length in flow direction increasing dramatically with decreasing slit height.

Mali and Manna[99] reviewed the current status of AFM and observed that Extrusion

pressure, flow volume, grit size, number of cycles, media, and workpiece

configuration are the principal machining parameters that control the surface finish

characteristics. Recently there has been a trend to create hybrid processes by

merging the AFF process with other non-conventional processes.

25

2.3.2 PROCESS MODELING AND OPTIMIZATION

To advance the understanding of complex process it is often necessary to construct a

simple model, experimental or mathematical, in which the variables can be changed

in an orderly fashion and their effects can be analysed. The model, in general,

provides the information, which gives an insight into the nature of phenomenon

occurring in the real life situation. In the case of grinding, there have been many

attempts in the past to model the surface generated using statistical approaches

[100].

Yoshikawa and Sata[101] simulated the grinding process by the Monte Carlo

method. Law et al [102] developed a grinding model in terms of grains distribution

on the abrasive wheel and kinematic grinding conditions. Hamed et al.[103] applied

an approach to generate random surfaces by computer simulation which involves

the use of a series of ‘unit events’ to produce a surface profile. The work confirmed

that the engineering surfaces could be described statistically by a profile having

ordinate heights with a Gaussian distribution together with a form of exponential

autocorrelation function. Abrahamson et al.[104] showed the effect of initial surface

finish on the wear of sliding surfaces. Jain and Jain [105] analysed and simulated the

profile of the finished surface by the interaction of abrasive grains with the

workpiece in abrasive flow finishing process. Pandey et al.[106] proposed the

equation to estimate the centre line average value of a parabolic profile. Spurr[107]

also gave the equation that estimates the C.L.A. surface roughness value of metals

after they have been slid against various grades of abrasive paper and confirmed it

experimentally. Dowson and Whomes [108] reported that the inclusion of roughness

in the analysis inevitably introduces difficulty in mathematical representation of the

surface topography. The surface cannot be described exactly by a simple equation,

nor can it be considered entirely random in nature. Any model of the surface will

therefore be an approximation. Sakamoto and Tsukizoe[109], the asperities

distributed over a machined surface could generally be assumed to be of conical

protuberances. Tsuwa [110] reported a simple concept of probability to find out

effective spacing between the cutting edges in the grinding wheel. The model of

active grains is proposed from the work of Tanaka and Ikawa [111] with a few

modifications. The approaches discussed above [104-111], can be applied to model

the surface roughness obtained in the abrasive flow machining process with certain

modifications, assuming that there exists similarity in distribution of abrasive

particles in grinding wheel and abrasive particles in medium used as AFM, depending

upon the AFM conditions. The difference is that in grinding wheel the medium is a

rigid body while in AFM it is a flexible viscoelastic putty. It is considered that the

diameter of all the grains is the same. Let the diameter of a representative grain be

d₆ = 28/M¹˙¹e, where Me is the mesh number [112]. The shape of an abrasive grain is

approximated as a sphere, and not composed of acute cutting edges [20].

26

Process modelling and optimization are very important issues in manufacturing

engineering. Manufacturing processes are usually too complicated to warrant

appropriate analytical models and most of the time, analytical models are developed

based on many assumptions which contradict reality. More importantly, it is

sometimes difficult to adjust the parameters of the models according to the actual

situation of the machining process [113]. Because of the complexity of the machining

process, optimization as well as optimal control are difficult to perform.

Chryssolouris and Guillot [114] modelled the machining processes by a multiple

regression method and neural network, and concluded that neural networks are

superior to conventional multiple regression method considering that neural

networks can map the input/output relationships and possess massive parallel

computing capability. Rangwala and Dornfeld[115] presented a scheme that used a

multi-layered perceptron neural network to model the turning process and an

augmented Lagrange multiplier (ALM) method to optimize the material removal

rate(MRR). A simple neural network model for grinding mode identification and

surface quality prediction in grinding of silicon nitride is also established by Zouaghi

and Ichida [116]. Sathyanarayan et al.[117] used a neural network model to study

creep feed grinding of super alloys, but the optimization was done analytically using

an off-line multi-objective programming technique. The central motivation

underlying the development of artificial neural systems is to provide a new type of

computer architecture in which knowledge is acquired and stored over time through

the use of adaptive learning algorithms [118]. Neural net models are specified by the

net topology, node characteristics, and training or learning rules. Back-propagation

neural network [119] is usually referred to as feed forwarded, multi-layered network

with a number of hidden layers trained with a gradient descent technique. This

algorithm is based on the error correction learning rule. Genetic algorithms [120] are

computerized search and optimization algorithms based on the evolutionary process

of biological organisms in nature. During the course of evolution, natural populations

evolve according to the principles of natural selection and ‘survival of the fittest’.

Williams and Rajurkar [63] developed a stochastic model of AFM generated surfaces

by using Data Dependent Systems (DDS) methodology. They have estimated the

ratio of surface roughness peak to valley height (Rz) to centerline average surface

roughness value (Ra) by DDS methodology and found to be between 1.4 and 2.2 for

the AFM process. They have established that AFM finished surface profiles possess

two distinct wavelengths, a large wavelength that corresponds to the main path of

abrasive while the small wavelength is associated with the cutting edges. Good

agreement is found between the primary frequency ranges obtained in DDS

modeling and those derived from spectral analysis function. It is stated that these

frequency bands are related to different material removal modes in AFM;

consequently, the mechanism of material removal in AFM is considered to consist of

27

ploughing responsible for creation of characteristic flow lines and micro-cutting.

They also proposed an expression for estimating the abrasive grain wear and the

number of active grains (Cd). The estimated value of Cd is used as a cutting life

criterion for abrasives. For small number of cycles its value should remain fairly

stable but with more and more processing the abrasive particles may fracture

thereby increasing the Cd value. The downturn of Cd value indicates that the

medium has absorbed too much work piece material and need replacement. Jain et

al. [42] also carried out simulation of finished surface profile and material removed

considering the interaction of abrasive grains with workpiece. Rajeshwar et al. [121]

proposed a mathematical simulation model to determine the characteristics of the

medium flow during finishing and its experimental verification was carried out. This

model was developed using constitutive equations of Maxwell model considering the

medium characteristics as non-newtonian flow. They reported that a linear

relationship exists between shear stress acting on the surface and the layer thickness

of material removed. A finite element approach was developed by Jain et al. [42] for

prediction of the stresses developed during finishing of a cylindrical passage by AFM

process (axi-symmetric flow). In their study it is assumed that medium exhibits linear

viscous flow property and medium properties are independent of temperature and

are constant with regards to time and space. They also presented a theoretical

model which is based upon the consideration that abrasion process in AFM i.e.,

combination of micro-ploughing and micro-cutting by assuming that all the abrasive

particles are spherical in shape having a single cutting edge with same size. It is also

assumed in this model that the load acting on each particle is constant and every

grain achieves the same penetration depth depending upon applied load. Gorana et

al. [27] developed a theoretical model of forces acting on a single abrasive grain for

studying the finishing mechanism of AFM process. Comparison of theoretical model

results with that of experimental data of force and active abrasive grains density

obtained during AFM process was done. Fletcher et al. [122] studied the relationship

between medium rheological properties and the AFM process. Shear rate of the

polymer increases when it passes through the restriction (or reduced crosssectional

area). Capillary rheometer is used to find the relationship between wall shear stress

and shear rate for medium viscosity of polyborosiloxane medium. They concluded

that coefficient of viscosity decreases but shear stress increases as shear rate

increases. Variation of wall shear stress with time is also studied. They also

concluded that greater finishing action could be achieved as a result of longer piston

stroke durations, due to higher wall shear stress generated. Petri et al. [123] adopted

neural network modelling technique for developing a comprehensive model for

AFM. They presented three neural network models (Polishing applications, surface

removal applications with a circular flow path and surface removal application with a

non-circular flow path). Petri et al. [86] reported a set of neural network models that

predict the surface finish and dimensional change. These neural network models are

28

then paired with a heuristic search algorithm to select sets of machine setup

parameters for the AFM process. Lam and Smith [124,125] applied Cascade-

Correlation neural network modeling to finishing of automotive engine air intake

manifold. They used it to predict the instant at which the finishing process should

terminate to meet the airflow specifications. Jain and Jain [87] proposed a

generalized back propagation neural network model and a second network which

parallelizes the augmented Lagrange multiplier (ALM) algorithm. The model

determines optimal finishing parameters by minimizing a performance index subject

to appropriate operating constraints. Sarah et al. [77] presented a neural network

model as an off-line controller for AFM of automotive engine manifold to predict

when the AFM process should be stopped to achieve the required airflow rate

through manifold body.

Petri et al.[126] developed a predictive process modeling system for the AFM

process that relates all of the critical parameters using strictly empirical techniques,

namely neural networks [60]. Their system addresses process settings for AFM for a

variety of products and material types. They focussed their attention on one

particular product type (i.e., engine manifolds) but with the demand of more precise

control to meet stringent specifications.

A prototype neural network based process monitor and controller for abrasive flow

machining of engine manifolds was developed for a consortium including an AFM

manufacturer and a U.S. automotive manufacturer. The first objective of this

research was to improve the functional performance of U.S. automotive engines,

hence generate the economic benefits of reduction in fuel consumption. The second

objective was to enable predictive process control of the AFM process, with an

understanding of the relationship between the AFM media to the specified air flow

rate of the engine manifolds. The development of the process model is an attempt

to capture the behavior of both the independent and interaction effects of these

variables in order to accurately predict the flow of the orifice fluid (viz., air) through

the manifold.

Theoretical models and numerical methods were developed to predict the polishing

behavior of the abrasive medium during AFM [27,42,76,83,87,88,91,105]. The

material removal rate and surface roughness were estimated using the finite

element method [42, 91]. Stochastic simulation was used to determine the active

grain density on the medium surface as well [88]. This method could easily extend to

simulate the surface generation in AFM. Furthermore, the material deformation

produced by the abrasive was developed to predict the force models of AFM [27].

The scratching experiments were used to study the material removal mechanism in

the abrasive process. However, it is not easy to get uniform roughness in complex

surface during AFM, because the shear forces acting on the complex surface will not

29

be the same if the flow path is not regular[16]. In addition, there were no researches

to develop a method to create the uniform surface roughness during AFM. So a non-

Newtonian flow model was used to simulate the motion of the abrasive medium in

AFM. In this simulation, excellent passageway could be designed when the uniform

shear forces were found on the machined surface. These kinds of passageway could

be applied to identify the uniform roughness of the complex hole in AFM.

Jain and Jain [91] developed a finite element model proposed for analysing the flow

of a viscoelastic medium in the AFM process. The results of the finite element

analysis have been used to estimate the material removal and surface finish in AFM

process. The theoretical results obtained by the finite element simulation have been

compared with experimental results. The effects of the process parameters on the

process performance have been discussed. They concluded that the extrusion

pressure and normal stresses increase linearly with increase in piston velocity. The

design of the tooling is important to achieve the appropriate normal pressure, so

that the desired surface finish in AFM can be achieved. At higher reduction ratios,

the rate of increase in extrusion pressure is higher. The normal stresses on the

workpiece surface increase with increase in reduction ratio. Material removal

increases with increase in extrusion pressure and percentage concentration of

abrasives in the medium. The change in surface roughness value increases with

increase in extrusion pressure, percentage concentration of abrasives and reduction

ratio for a specified number of cycles.

Jain [42,83] used finite elements simulation approach to model AFM process used to

evaluate the stresses and forces by assuming the sliding action of active abrasive

particles for modelling material removal and surface finish. However, according to

Fang et al[94], the abrasive particles in media are not only sliding, but also rolling

while loaded. The rolling particles will decrease material removal efficiency [127-

129]. Gates[130] and Wirojanupatump[131], also noticed the influence of abrasive

particle rolling to material removal efficiency. Therefore, the estimation for abrasive

particles movement pattern should be stressed to AFM.

Gorana et al.[132] developed an analytical model to simulate and predict the surface

roughness for different machining conditions in abrasive flow machining (AFM). The

kinematic analysis is used to model the interaction between grain and workpiece.

Fundamental AFM parameters, such as the grain size, grain concentration, active

grain density, grain spacing, forces on the grain, initial topography, and initial surface

finish (Ravalue) of the workpiece are used to describe the grain-workpiece

interaction. The AFM process is studied under a systematic variation of grain size,

grain concentration and extrusion pressure with initial surface finish of the

workpiece. Simulation results show that the proposed model gives results that are

consistent with experimental results. Gorana et al.[132] concluded that active grain

30

density during the AFM process increases with an increase in extrusion pressure and

percent abrasive concentration in the medium. This results in an increase in percent

reduction in Ra value.

The numerical simulation of viscous flow had been the topic of many researchers

[133-135]. Early numerical formulation of viscous materials involved the finite

element technique based on the Galerkin models, and the finite difference

technique based on stream function – vorticity formulations. Zienkiewicz et al. [133]

have presented the flow formulation approach in forming and extrusion,

investigating two techniques, namely the pressure – velocity formulation with

Lagrangian constraints and the penalty – function approach. Dixit and Dixit [134]

carried out deformation analysis of the steady state wire-drawing process by the

finite element model (FEM) using the mixed pressure – velocity formulation. The

initial developments of finite difference and finite element methods used for solving

non-Newtonian flow have been reviewed by Crochet et al.[135]. Wilson et al.[136]

reported that, if the settling velocity of discrete solid particles transported by a

carrier liquid is very small, then the slurry can be considered as a single phase. If the

mixture can be transported with a uniform solid concentration across the pipe, then

during such transport the mixture can be viewed as homogeneous.

Das et al.[137] explored a new precision finishing process called magneto-rheological

abrasive flow finishing (MRAFF), which is basically a combination of abrasive flow

machining (AFM) and magneto-rheological finishing (MRF), developed for nano-

finishing of parts even with complicated geometry for a wide range of industrial

applications. They dealt with the theoretical investigations into the mechanism of

MRAFF process to study the effects of various process parameters and an attempt

has been made to analyse the medium flow through the fixture by finite difference

method by assuming the medium as Bingham plastic to evaluate the stresses

developed during the process. A capillary viscometer has been designed and

fabricated to study the effect of magnetic field on the rheological properties of the

medium. Microstructure of the mixture of ferromagnetic and abrasive particles in

magneto-rheological polishing fluid (MRPF) has been proposed, and normal force on

the abrasive particles is calculated from the applied magnetic field. A model for the

prediction of material removal and surface roughness has also been presented.

Theoretical results compare well with the experimental data available in the

literature.

Analytical models that explain a highly non-linear relationship with interactions

among process variables are difficult to obtain. Moreover, there are no analytical

models that capture the dynamics of the entire abrasive flow machining process.

Artificial intelligence techniques, such as neural networks and expert systems, have

been increasingly used to successfully model process behavior in areas where

31

analytical models are unavailable. The use of neural networks is motivated because

of their accommodation of non-linearities, interactions, and multiple variables.

Neural networks are also tolerant of noisy data and can operate very quickly in

software, and in real time in hardware. Unlike statistical models which generally

require assumptions about the parametric nature of the factors (which may or may

not be true), neural networks do not require apriori assumption of the functional

form of the model. Recent works in using neural networks for modelling

manufacturing processes include [138-141].

2.3.3 MONITORING OF AFM PROCESS

For online monitoring of material removal and surface roughness in AFM process,

Williams [142] applied acoustic emission technique (elastic stress waves generated

by the rapid release of strain energy within a material due to a rearrangement of its

internal structure is called "acoustic emission"). In a full factorial experiment, the

effect of extrusion pressure, medium viscosity, abrasive grit size, number of cycles,

and work piece material was investigated on material removal, root mean square of

acoustic emission signal (AERMS), and surface roughness improvement. From the

above parameters only grit size showed insignificant effect on material removal.

They studied acoustic emission signals for grinding to analyse the mechanism

involved in AFM and found that the acoustic emission signal is highly dependent on

the characteristics of the initial surface roughness of the workpiece. The AERMS of

the signal is sensitive to extrusion pressure and other AFM process parameters,

which affect material removal. They used Data Dependent Systems technique to

analyze the acoustic emission signal acquired during AFM processing. In DDS

analysis, they found that main root of frequency of signal was around 160 kHz and

secondary root with less power had a frequency around 80-90 kHz. They reported

the fact that the higher frequency component is associated with the ploughing

mechanism which would agree with the results obtained during grinding. They also

found that aluminum workpiece give stronger signals than steel workpiece because

of higher material removal from aluminum workpiece as compared to steel under

identical finishing conditions. This suggests a strong correlation between material

removal rate and acoustic signal in AFM.

Dabrowski et al [143] attempted theoretical analysis of the media flow in the AFM

process. The mathematical representation of the media flow in the AFM process

involves the equations of continuity and momentum, and the constitutive equations.

The following assumptions are made, to simplify the analysis:

1. The medium used in AFM is composed of a semisolid carrier mixed with abrasives

that exhibits a linear viscous flow property.

2. The medium is isotropic and homogeneous. The medium properties are

independent of fluid temperature and constant in time and space.

32

3. Since a cylindrical workpiece is considered, flow of the medium is taken as axial-

symmetric.

4. The flow of the medium is steady.

Such assumptions enable the use of the finite element method (FEM) for simulation

of rigid plastic flow through a slot in a part. It is observed that the present FEM

analysis gives a linear relationship between the piston pressure and the velocity of

the piston. In addition, to achieve the same flow velocity of the medium, the tool

angle should be low at low pressure. Thus, for a given piston velocity, the highest

pressure is required at a ¼ 90 tool angle, which is a commonly used configuration.

Hence, the tool should be designed using lower angles rather than the 90 angle

configuration. Normal stress also varies linearly with piston velocity and its level is

higher for a higher angle value of the tool. At a lower radial pressure, the expected

penetration depth is lower. They compared the experimental and theoretical

relationships between the piston velocity and material removal. It is found that

material removal increases with an increase of piston velocity. The simulation results

for material removal agree well with the experimental data. However, at higher

velocities, the number of abrasive grains taking part in actual cutting may be lower,

owing to the rolling of grains at a higher velocity. The effect of this phenomenon may

be a drop in pressure in the narrow channels through which the medium is flowing

[57]. Hence, actual material removal will be less than theoretical removal values.

Williams [144] emphasized that a deterministic and stochastic analysis techniques

are needed to understand the process, due to the complex and random nature of

AFM, and to find a quantitative measure of the metal removal mechanism which can

be integrated with an on-line control strategy. They dealt with the following specific

topics related to the AFM process: metal removal and surface roughness

characteristics per cycle for a single hole part, machining of multiple hole

components, prediction of metal removal and flow rate in AFM using an acoustic

emission monitoring system, and stochastic modeling and analysis of AFM surface

roughness profiles and acoustic emission signal data. In studying the machining

characteristics per cycle, the largest change in bore diameter and surface roughness

always occurred from the raw workpiece to the first AFM cycle. A better starting

surface roughness was associated with a better final roughness value. A velocity

distribution of the abrasive-laden media was proposed for a multiple hole

component. Experimental results showed that the center hole had, on an average

30% more metal removal as compared to the outer holes. The effect of media

temperature on flow rate also helped to clarify the results. An AFM monitoring

system using acoustic emission (AE) technology was designed, built and tested.

Relationships were proposed between the acoustic emission level and the AFM

process parameters. It was shown that metal removal in AFM could be fairly

accurately predicted knowing the RMS of the AE signal and the levels of the

33

machining parameters. A high correlation was found between AFM flow rate and AE

RMS over certain ranges. Data Dependent Systems (DDS) analysis of the acoustic

emission signals revealed distinct frequency bands during AFM that were linked to

the process mechanisms. Good agreement was found between DDS frequency

decomposition and the results of a spectral analysis option on the new data

acquisition system.

Smith and Slaughter [145] discussed the preliminary development of a neural

network-based process monitor and off-line controller for abrasive flow machining

of automotive engine intake manifolds. The process is only observable indirectly, yet

the time at which machining achieves the specified air flow rate must be estimated

accurately. A neural network model is used to estimate when the process has

achieved air flow specification so that machining can be terminated. This model uses

surrogate process parameters as inputs because of the inaccessibility of the product

parameter of interest, air flow rate through the manifold during processing.

[146] developed a method for controlling the abrasive machining, especially abrasive

flow machining of a part which includes measuring any acoustic emission signal

caused by the abrasive machining of such part; thereafter comparing a high

frequency, e.g. greater than 100 kHz, component of said signal with a low frequency,

e.g. less than 100 kHz, component of said signal to generate a control ratio. When

the control ratio reaches a predetermined ratio, or predetermined change in the

ratio over time, a control action of the machining process is effected, such as

termination of the flow of abrasive.

2.3.4 MEDIA COMPOSITION AND ITS EFFECT

AFM utilizes a medium loaded with abrasive grain as the cutting tool. The base is a

friable, highly resilient organic polymer such as polyborosiloxane blended with

special lubricating diluents [147]. The main abrasives commonly used in AFM are

aluminum oxide, silicon carbide, boron carbide and polycrystalline diamond. The

composition of the medium includes special viscoelastic polymers that change in

viscosity with the different shear stresses seen while being forced through the

workpiece.

Abrasive media typically consist of a polyborosiloxane, a silicone rubber (commonly

known as “silly putty”) and conventional abrasives such as diamond, silicon carbide

and aluminium oxide. Small quantities of plasticisers and viscosifiers are included to

modify the rheology of the abrasive media [148].

Fletcher et al [122] studied the thermal and fluid flow properties of polymers used in

AFM. They showed that the rheology of the media contributes significantly to the

success of the AFM process. The velocity of the extruded media is dependent upon

34

the principal parameters of viscosity, pressure, passage size, geometry and length

[149]. Sankar et al.[150] noted that the flow of viscoelastic medium under high

pressure through a fixture and workpiece having varying cross-sections changes

medium properties. This is due to the variance in shear rate to which it is subjected.

Agrawal et.al. [151] developed a viscometer set-up based on the principle of visco-

elasticity. The creep compliance and the bulk modulus have been determined and

the viscosity of the abrasive media has been subsequently calculated. Measurements

have been conducted for obtaining viscosity along with an assessment of specimen

length and initial load, the influence of reduced data points and the repeatability of

the experiments. Besides this, experiments have been conducted at varying

concentration and temperature of the abrasive media. Experiments show that the

viscosity of the media increases with the percentage concentration of abrasives and

decreases with temperature. Viscosities at different concentration of the abrasives

were compared with the values obtained from a capillary viscometer and the

comparison was found to be good.

Kar et al.[152] studied the media used for AFM, developed from the various

viscoelastic carriers and has been contrasted through experimental investigations.

The viscoelastic media are selected on the basis of existing media through the

studies of thermogravimetric analysis and arecharacterized by mechanical, as well as

rheological, properties with the help of a universal testing machine and a rheometer.

The performance of the medium is evaluated through the finishing criteria on a two-

way AFM setup. The investigation reveals that the styrene butadiene rubber (SBR)

medium gives a good improvement in surface finish. The surface improvement

through SBR media is 88%. It is also found that the strain, temperature, shear

rate, time of applied constant stress, cyclic loading, etc. have animpact on the

mechanical and rheological properties of the newly developed medium, which are

ultimately governed by the performance of the medium in the target applications.

Wang & Weng [55] developed Lower-cost and effective abrasive media to enhance

the roughness of the WEDM surface. It has been concluded that Vinyl-silicone

polymer (or silicone rubber) has good deformation and low flow effect; it can flow

through the complicated holes easily. Moreover, the silicone rubber will not stick on

the workpiece surface after machining. In this study, abrasive particles and silicone

rubber are mixed uniformly to form the flexible media. Then a chain hole, cut by

WEDM, is polished by these media in AFM. The abrasive medium with high viscosity

has excellent deformation. It can easily polish the WEDM surface to a smooth finish.

Furthermore, high abrasive concentration also has fine polishing result. They

reported that the surface roughness decreases from 1.8 to 0.28 μm Ra after five

machining cycles (with abrasive concentration 60% (wt.%)). In this case, the

roughness improvement rate (RIR) reaches 84%. On the other hand, small abrasive

35

sizes can produce smoother surfaces than large abrasive sizes. But the polishing time

might be several times longer than that by large abrasive sizes. However, the

polishing efficiency is good when new abrasive media are used in AFM.

Wang et al.[153]noted that Abrasive media are key elements which dominate the

polished results in AFM. But it is hard to develop the machining model of these

abrasive gels because of its complicated mechanism. They used a non-Newtonian

flow to set up the abrasive mechanism of the abrasive media in AFM. Power law is a

main equation of the non-Newtonian flow to describe the motion of the abrasive

media. Viscosities vs. Shear rates of different abrasive gels are used to establish the

power law in CFD-ACE+ software first. And the working parameters of AFM were

applied as input to study the properties of the abrasive gels in AFM. Finally, the

relationships between the simulations and the experiments were found. And the

abrasive mechanism of the abrasive gels was set up in AFM. The simulated results

show that the abrasive gel with high viscosity can entirely deform in the complex

hole than the abrasive gel with low viscosity. And the abrasive gel with high viscosity

generates a larger shear force than the abrasive gel with low viscosity in the same

area. Moreover, the strain rate is seriously changed when the abrasive gel cross over

the narrow cross-section of the complex hole. It also means that abrasive gel will

produce large finish force in that area. And these results indeed consistent with the

experiments in AFM.

Mark [154] & Ghose [155] suggested mixing of processing oil in the medium for

improved surface finish. They noted that processing oil is lower molecular weight

material and ‘polymer’ is higher molecular weight material. So, when plasticizer is

mixed with polymer, it tries to soften the polymer by either breaking the long

polymer chains into small ones or by entering between the polymer chains. When

the plasticizer enters between the polymer chains, the free space between two

polymer chains increases so that inter-molecular forces between the polymer

molecular chains reduce. Thus, the polymer becomes softer when processing oil

mixes with it. As the processing oil content increases, normal inter-polymer

molecular forces gradually decrease and the medium starts attaining better flexibility

or self- deformability.

Wan et al. [156] realized the need for an abrasive flow machining process that could

function at relatively low pressures using equipment commonly available in the

typical machine shop. To this end, low viscous abrasive media was formulated from

readily available materials, and a pneumatic shop-air driven abrasive flow machining

system was designed and fabricated. They present experimental results on the

polishing performance of the media on small diameter holes EDM wire cut in

aluminium. Analysis of the results revealed the critical role played by the normal

stress differences in determining whether or not material removal and hence

36

polishing will occur. Thus, while the media must be sufficiently viscous to suspend

the abrasives, all three viscometric must be considered in characterising the efficacy

of the media used in the experiments.

2.3.5 RECENT DEVELOPMENTS IN AFM PROCESSES

The basic principle behind AFM process is to use a large number of random cutting

edges with indefinite orientation and geometry for effective removal of material

with chip sizes far smaller than those obtained during machining with tools having

defined edges. Because of extremely thin chips produced in abrasive machining, it

allows better surface finish, close tolerances, and generation of more intricate

surface features. One of the limitations of AFM process is the low productivity. The

time to achieve the required surface finish is longer in AFM process as compared to

other finishing processes. Researchers have tried to overcome this difficulty by using

hybrid approach [157–160] and have reported improvement in process efficiency of

AFM when centrifugal force was applied on the abrasive media while it abrades the

workpiece.

Walia et al. [161] recently explored centrifugal force assisted abrasive flow

machining (CFAAFM) process as a hybrid machining process with the aim towards

the performance improvement of AFM process by applying centrifugal force on the

abrasive laden media with a rotating centrifugal force generating (CFG) rod

introduced in the work piece passage. For optimization of process parameters, an

approach based on a Utility theory and Taguchi quality loss function (TQLF) has been

applied to CFAAFM for simultaneous optimization of more than one response

characteristics. Three potential response parameters i.e., material removal, %

improvement of surface finish and scatter of surface roughness over the fine finished

surface of a sleeve type work piece of brass are examined. Utility values based upon

these response parameters have been analysed for optimization by using Taguchi

approach.

The enhancement of the process efficiency of different non-traditional machining

processes has been explored by researchers [162-164]. In the case of AFM, Singh et

al.[165] have shown the improvement in the process efficiency when the magnetic

field was applied to the workpiece.

Many researchers have been working to overcome the limitations, such as low

finishing rate, and incapability to correct the form geometry, and at the same time to

improve the finishing rate, surface integrity and compressive residual stresses

produced on the workpiece surface. Singh and Shan [56] applied magnetic field

around the workpiece in AFM and developed a set-up for a composite process

termed magneto abrasive flow machining (MAFM), and the effect of key parameters

on the performance of the process has been studied. It was observed that magnetic

37

field significantly affect the material removal and change in surface roughness. With

the application of magnetic field, less number of cycles are required for the higher

material removal. Higher material removal and higher change in surface roughness

are observed (in case of brass as workpiece material) with the low flow rates of

medium and high magnetic flux density. Experimental results indicate significantly

improved performance of MAFM over AFM [56].

Fig.2.7 Sectional front view of tooling in the finishing region in DBG-AFF process and

top view of medium splitting through twin slotted fixture plate[166]

Fig.2.8 Abrasive direction due to active abrasive grains in DBG-AFF (a) macro-view,

and (b) micro-view [18]

Ravi Sankar et al. [166] tried to improve finishing rate, material removal and surface

texture by placing drill bit in the medium flow path called Drill Bit Guided Abrasive

Flow Finishing (DBG-AFF), as illustrated in Fig.2.7. The inner part of medium slug

flows along the helical flute which creates random motion among the abrasives in

inner region of the medium, as shown in Fig.2.8. This causes reshuffling of abrasive

particles at outer region. Hence, comparatively more number of new and fresh

abrasive grains interact with the workpiece surface. From the experimental results, it

38

is concluded that the abrasive traverse path is longer than the AFM abrasive traverse

path in each cycle. It results in higher finishing rate in DBG-AFF as compared to AFM.

Material removal is found to decrease with decrease in drill bit diameter. Biing-Hwa

et al. [167] placed spiral fluted screw in the medium flowing path to improve surface

quality and proposed a spiral polishing method and a device for micro-finishing

purposes. This novel finishing process has wider application than traditional

processes. This offers both automation and flexibility in final machining operations

for deburring, polishing, and removing recast layers, thereby producing compressive

residual stresses even in difficult to reach areas. Application of this method can

obtain a fine polished surface by removing tiny fragments via a micro lapping

generated by transmission of an abrasive medium through a screw rod. The effect of

the removal of the tiny fragments can be achieved due to the function of micro

lapping. The method is not dependent on the size of the work-piece's application

area in order to carry out the ultra-precise process. The application of this technique

can be extended to various products of precision ball-bearing lead screw. The

proposed method produces products with greater precision and more efficiently

than traditional processes, in terms of processing precisions and the surface quality

of products. These parameters used in achieving maximum material removal rate

(MRR) and the lowest surface roughness (SR) are abrasive particle size, abrasive

concentration, gap, revolution speed and machining time.

Walia et al. [168] rotated different shaped tiny rods at the centre of the medium

flow path and used a low viscosity medium to finish. They concluded that the better

surface finish is achieved due to centrifugal action caused by the rod on the

abrasives and this process is called centrifugal force assisted abrasive flow machining

(CFAAFM). But all these three rotating medium methods may rotate the medium at

and near the axis of the medium but the probability of rotating the medium at the

abrasive-workpiece interaction region is very low. Ravi Sankar et al [169] developed

a new set-up to rotate the workpiece so that the probability of active abrasive

particle rotation in the workpiece finishing region is high which improves both

surface finishing rate and material removal rate.

Dabrowski et al [143] studied the possibility of developing Electrochemical aided

abrasive flow machining (ECAFM) using polymeric electrolytes. The ion conductivity

of electrolytes is many times lower than the conductivity of electrolytes employed in

ordinary electrochemical machining (ECM). Additions of inorganic fillers to

electrolytes in the form of abrasives decrease conductivity even more. These

considerations explain why the inter electrode gap through which the polymeric

electrolyte is forced should be small. This in turn results in greater flow resistance of

polymeric electrolyte, which takes the form of a semi-liquid paste. Rheological

properties are also important for performance considerations. Experimental

investigations have been carried out for smoothing flat surfaces and process

39

productivity in which polymer electrolytes as gelated polymers and water-gels based

on acryloamide were used.

Dabrowski et al.[170] studied the effect of progressive removal of the material

allowance due to the doubly advantageous micro-cutting influence on de-passivation

of the machined surface and that of the anodic dissolution on the mechanical

material de-cohesion. Modification of the original AFM method includes replacing a

conventional abrasive paste with a polymeric electrolyte which plays a role as a

transport medium for abrasive grains. Application of new abrasive pastes based on a

polymeric electrolyte in a machining layout equivalent to the ECM layout is the start

of and new hybrid superficial abrasive machining technique, denoted as ECAFM

(electrochemical abrasive flow machining).

Dabrowski et al. [66] presented the results of the research on the electrochemically

assisted abrasive flow machining (ECAFM). The experimental evaluation of several

solid electrolytes with various bonds has been carried out (the polypropylene glycol

PPG with the NaI salt share and ethylene glycol PEG with KSCN salt share have been

subjected to the tests). The abrasive properties of the electrolytes have been

enhanced by adding the Al2O3 and SiC grains and its consistence has been adjusted

by the SiO2 addition. The important technological and economical advantages of the

investigated ECAFM process have been demonstrated. Dabrowski et al. [66]

observed material removal comes as a result of micro-cutting process affected by

abrasive grains being component of the abrasive paste. The intensity of this process

is conditioned by pressure in the abrasive media. Pressure has to be adjusted to the

machined part design. Fragile parts can be machined while applying insignificant

pressure on the internal or external surfaces and robust parts allow for high pressure

flow up to 10 MPa.

Liao et al.[171] pointed out that conventional AFM methods have difficulty achieving

uniform roughness of an axial distribution in circular hole polishing due to limited

unitary axial motion of abrasive media. Therefore, Liao et al.[171] developed

mechanism designs for different passageways to obtain multiple flowing paths of

abrasive medium, whose flowing behavior enhances polishing effectiveness by

increasing the abrasive surface area and radial shear forces. The motion of the

abrasive medium is studied by utilizing different mold cores, which mold shapes

include the circular, hollow and helical passageway. The optimum design of the

passageways is then verified using CFD-ACE+ software, numerical results indicate

that passageway with six helices performed better in the uniform surface roughness

than others’ do. Experimental results show that roughness deviation of six helices

passageway of approximately 0.100 m Ra is significantly better than those on a

circular passageway of around 0.1760 m Ra. Additionally, the six helices

passageway is also superior to circular passageway in reducing roughness

40

improvement rate (RIR) by roughly 87% compared with RIR 67.7% for the circular

passageway.

Walia et al.[172] identified one serious limitation of this process as its low

productivity in terms of improvement in surface roughness and proposed developing

improved fixturing as a technique for productivity enhancement in terms of surface

roughness (Ra). A rotating Centrifugal Force Generating (CFG) rod was used inside

the cylindrical work piece, which provides the centrifugal force to the abrasive

particles normal to the axis of work piece. The effect of the key parameters on the

performance of process has been studied. The results show that for a given

improvement in Ra value the processing time can be reduced. It is seen that the

significant process parameters are rpm of CFG rod, extrusion pressure and abrasive

mesh size. Walia et al.[173] applied centrifugal force on the abrasive media using a

rotating rod introduced in the workpiece passage. This variant of AFM, called

centrifugal force assisted abrasive flow machining (CFAAFM), has been developed.

This paper presents finite element modelling (FEM) of a polymer-based non-

Newtonian viscoelastic fluid used in the CFAAFM process and the same is used to

evaluate the resultant pressure, velocity, and radial stresses during a working cycle.

All FEM analysis were performed using a commercial finite element package ANSYS.

The results are compared with the data available in the literature, and close

agreement has been found between the two.

Reddy et al.[174] discussed a few of the issues arising when attempting to enhance

the capabilities of the conventional AFM process. In this process, a centrifugal force

is imparted to the abrasive particles in the medium, making the process a hybrid

one. The set-up has been suitably modified for this purpose, and the resulting

process is called centrifugal force assisted abrasive flow machining (CFAAFM). The

effect of key parameters on the performance of the process has been studied

through response surface methodology (RSM). Relationships were developed for

material removal and percentage improvement in surface finish of cast Al alloy

(2014) cylindrical components. Experimental results indicate the significantly

improved performance of CFAAFM over AFM in terms of enhanced surface finish and

material removal. It was observed that the combination of a high extrusion pressure

and a higher speed of the centrifugal force generating (CFG) rod is more favourable

to obtain a higher degree of surface finish, while the combination of a larger grain

size and a higher speed of the CFG rod causes higher material removal.

41

Fig.2.9 Forces and velocity components in R-AFF process in the finishing region (a)

Overview of various forces and velocities in workpiece, (b) Enlarged view of shearing

of roughness peaks on the workpiece surface by abrasive grains, (c) Free body

diagram of forces and velocities in R-AFF process[169]

Sankar et al.[175] focussed on abrasive flow finishing (AFF), a process that finishes

complex internal and external geometries with the help of viscoelastic abrasive

medium, while keeping in mind its low finish and material removal rates (MRR).

Researchers have often strived to improve finishing rate and MRR. As an attempt to

overcome the said limitations, this paper discusses rotational abrasive flow finishing

(R-AFF) process wherein complete tooling is externally rotated and the medium

reciprocates with the help of hydraulic actuators. Fig.2.9 shows the force and

velocity components in R-AFF process. In this study, preliminary experiments are

conducted on Al alloy and Al alloy/SiC metal matrix composites (MMCs) at different

extrusion pressures, and medium compositions are employed for finding optimum

conditions of the same for higher change in roughness (ΔRa). The same optimum

conditions are used to study the effect of workpiece rotational speed on (ΔRa),

material removal (MR), change in workpiece hardness and surface topology. It is

noted that as the workpiece rotational speed increases from 2 to 10 RPM, the

experimental helix angle decreases from 22° to 9° and the helical path length

increases from 67 to 160 mm. Based on these findings the mechanism of material

removal of matrix and reinforcement in MMC using R-AFF have been proposed, as

shown in Fig.2.10. Here the matrix material is removed by micro-cutting and three

methods of material removal mechanisms for reinforcement are also explained. The

scientific logic behind finishing mechanism of matrix and reinforcement, cross hatch

patterns, helical path directions, micro-scratch (μ-scratch) width and depth variation

with size, orientation and support that active abrasive grain obtains from

neighboring abrasives is derived from scanning electron microscopy micrographs.

Finally this study establishes that R-AFF can produce 44% better ΔRa and 81.8% more

42

MR compared to the AFF process. Accordingly, R-AFF generates micro cross hatch

pattern on the finished surface that can improve lubricant holding capabilities.

Walia et al.[176] suggested Tiny rods, each shaped differently, are rotated at the

centre of the medium flow path and a low-viscosity medium is used to finish.

Improved surface finish is achieved due to centrifugal action caused by the rod on

the abrasives with the help of centrifugal force assisted abrasive flow machining.

Fig.2.10 Mechanism of material removal during AFF of MMC (a) Reinforcement

pullouts in MMC in the agglomerated area, (b) Corresponding model of indentation

in the matrix due to reinforcement pullout in AL alloy/SiC MMC, and (c) Abrasive

embedded in the surface

Das et al.[177] developed a new precision finishing process called magneto-

rheological abrasive flow finishing (MRAFF), which is basically a combination of

abrasive flow machining (AFM) and magneto-rheological finishing (MRF), for nano-

finishing of parts even with complicated geometry for a wide range of industrial

applications. In this paper microstructure of the mixture of magnetic and abrasive

particles in magneto-rheological polishing fluid (MRPF) has been proposed, and

normal force on the abrasive particles is calculated from the applied magnetic field.

A model for the prediction of material removal and surface roughness achieved has

also been presented. And, finally theoretical results are compared with the

experimental data available in the literature, and they are found to agree well.

Jha and Jain [178] developed Magnetorheological Abrasive Flow Finishing (MRAFF)

for finishing complex internal geometries up to nanometer level using

Magnetorheological polishing (MRP) fluid which comprises of MR-fluid with fine

abrasive particles dispersed in it. Experiments were conducted on stainless steel and

silicon nitride workpieces using silicon carbide, boron carbide and diamond

abrasives. The best surface finish obtained on the stainless steel workpieces with the

present set-up configuration is 30 nm. The effect of number of finishing cycles,

extrusion pressure and magnetic flux density is studied on change in surface

roughness (Ra) value. Capillary magnetorheometer is designed and fabricated for

rheological characterisation of MR-polishing fluid. Bingham plastic, Herschel Bulkley

43

and Casson's fluid model are used to fit the obtained data points. Fig.2.11 illustrates

certain Advanced Abrasive Flow Machining Processes.

(a) MAF

(b) MRF

(c) MRAFF

44

(d) MFP

Fig.2.11 Advanced Abrasive Flow Machining Processes (a) Magnetic Abrasive

Finishing (MAF), (b) Magnetorheological Finishing (MRF), (c) Magnetorheological

Abrasive Flow Finishing (MRAFF), and (d) Magnetic Float Polishing (MFP)

Jha and Jain [179] conducted a detailed study through statistical design of

experiments for nano-finishing of stainless steel workpieces through

Magnetorheological Abrasive Flow Finishing (MRAFF) process. Response surface

regression analysis was done to analyse the effect of number of finishing cycles,

extrusion pressure and magnetic flux density on percent change in surface

roughness. The experimental results were discussed and optimum finishing

conditions were identified through contour plot. The best surface finish obtained on

the stainless steel workpieces with the present set-up configuration is 30 nm.

Sadiq and Shunmugam [180] studied Magnetorheological Abrasive Honing (MRAH)

as a recently developed process to finish engineering surfaces. The process makes

use of a magnetically stiffened abrasive-mixed magnetorheological fluid as the

flexible tool and rotation-cum-reciprocation movements between the finishing

medium and the workpiece surface for providing finishing action. In the present

work, a finite element analysis with Mechanical/Emag module of ANSYS is performed

to understand the nature of magnetic field developed in the process and verification

is done with actual measurements. Considering the simulated magnetic field, a

model to predict final roughness value (Ra) is developed. The model, when applied

for different work materials and various process parameters, such as magnetic flux

density, process duration and workpiece rotation, yields results that are in good

agreement with experimental results.

Sadiq and Shunmugam[181] developed Magneto-Rheological Abrasive Honing

(MRAH) setup and studied the effect of magnetic field on specimens mounted in a

holder and subjected to MRAH. It is observed that the change in the surface finish on

45

magnetic specimens is insignificant, while the surface finish on non-magnetic

specimens improves appreciably. By a novel method of introducing magnetic

specimens along with non-magnetic specimens, the magnetic field near the non-

magnetic specimens is enhanced. The results obtained on the non-magnetic

specimens by this approach for different magnetic field and process duration

confirm that the finishing capability of the process is greatly improved, and yields a

maximum improvement in finish of 41.7% and 43.5%, respectively while the

previously reported results are 6.7% and 24.2%.

Sadiq and Shunmugam[182] suggested a scheme to finish external curved surfaces,

by imparting rotation while the abrasive-mixed magnetorheological fluid (or

abrasive-mixed MR fluid) is pushed up and down. Since the relative motions

resemble those present in conventional honing, the proposed method is named as

‘Magnetorheological Abrasive Honing’ (MRAH). Experiments are conducted with

aluminum and austenitic stainless steel workpieces to understand the effect of

magnetic field. Effect of initial roughness, workpiece rotation and process duration

on finishing was investigated with ground austenitic stainless steel workpieces. It is

observed that the improvement in finish is better for rougher surface and higher

rotation speed of workpiece and a reduction in roughness is consistent with process

duration.

Jones and Hull [183] reported the modifications of existing AFM by applying

ultrasonic waves in the medium for machining blind cavities. The orbital flow

machining process suggested by Gilmore [184] has been recently claimed to be

another improvement over AFM, which performs three-dimensional machining of

complex components. These processes can be classified as hybrid machining

processes (HMP) – a recent concept in the advancement of non-conventional

machining. The reason for developing a hybrid machining process is to make use of

the combined or mutually enhanced advantages and to avoid or reduce some of the

adverse effects the constituent processes produce when they are individually

applied. Rajurkar and Kozak [185] have described around 15 various processes under

this category. Shinmura and Yamaguchi [186] and more recently Kim et al.[187],

Kremen et al. [188] and Khairy [189] have reported studies on successful use of

magnetic fields for micromachining and finishing of components, particularly

circular tubes, by magnetic abrasive finishing (MAF). The process under investigation

is the combination of AFM and MAF, and is given the name magneto abrasive flow

machining (MAFM).

2.3.6 APPLICATIONS OF AFM

AFM is a non-traditional finishing process that is used to deburr, polish or radius

surfaces of critical components. It has been applied in the aerospace, automotive,

electronics, defence, surgical, die-making industries and tool manufacturing

46

industries. AFM can process many selected passages on a single workpiece or

multiple parts simultaneously. However, AFM has not been widely used because of

the lack of theoretical support for the behaviour of the process. A large range of

process parameters such as extrusion pressure, flow volume, number of cycles,

media viscosity, media rheology, abrasive size and type, grit size and grit type, part

geometry, and others (such as the highly non-Newtonian nature of the AFM “fluid”)

must be taken into consideration when developing an application.

Mali and Manna [190] appreciated the recent trend to create hybrid processes by

merging the AFF process with other non-conventional processes which has opened

up new vistas for finishing difficult to machine materials with complicated shapes

which would have been otherwise impossible. These processes are emerging as

major technological breakthrough for precision, meso, micro, and nano scale

engineering. This review provides an insight into the fundamental and applied

research in the area and creates a better understanding of this finishing process,

with the objective of helping in the selection of optimum machining parameters for

the finishing of varied workpieces in practice. Fig.2.12 shows surface finish

improvement before and after on different machine components and Fig.2.13 shows

Photomicrographs depicting complete removal of EDM recast layer.

Nowadays, a wide application range of AFM is a result of its advantageous properties

of meeting the strict requirements imposed on important parts of machines [143],

for example:

(a) improvement of the flow characteristics for compressor and turbine elements,

exhaust collectors, nozzles, and diffusers after rough milling, electrochemical

machining (ECM), electric discharge machining (EDM), or precision casting;

(b) removal of burrs, rounding, and polishing of the slot edges for improvement of

fatigue resistance of blades, discs, hubs, and shafts;

(c) controlled removal of heat damage taking place during laser machining or EDM;

(d) smoothing of spray nozzles, injector bodies, and parts of bearings;

(e) removal of carbon deposits from regenerated parts, cleaning the surface, and

improvement of their reflexivity.

47

Fig.2.12 Surface finish improvement before and after on (a) internal passages within

turbine engine diffuser (b) Medical implants (c) complex automotive engine parts

Fig.2.13 Photomicrographs showing complete removal of EDM recast layer

Williams et al. [191] explored one recent application which has gained significant

attention as the improvement of air flow and fluid flow characteristics for injector

nozzles and cast automotive engine parts such as cylinder heads, intake manifolds

and exhaust manifolds. Previous research has demonstrated the viability of acoustic

emission (AE) monitoring of AFM using traditional AE instruments and techniques.

The root mean square (RMS) voltage of the AE signal was found to be mainly

determined by the material removal and related AFM process parameters. However,

this traditional AE approach was found to have limitations when applied to air flow

applications of AFM. Williams et al.[191] reported the development of a new

approach to AE monitoring and its application to the AFM process. This approach

builds on the fundamental sources of AE in machining and uses results of modal

analysis of AE signals to monitor specific high and low frequency components of the

AE signal. Advantages of this technique over traditional methods will be

demonstrated for the AFM process.

WANG et al. [192] investigated that the shear forces in the polishing process and the

flow properties of the medium in AFM play the roles in controlling the roughness on

the entire surface. A power law model was firstly set up by utilizing the effect of

shear rates on the medium viscosities, and the coefficients of the power law would

be found by solving the algebraic equation from the relations between the shear

rates and viscosities. Then the velocities, strain rates and shear forces of the medium

acting on the surface would be obtained in the constant pressure by CFD software.

Finally, the optimal mold core put into the complex hole could be designed after

these simulations. The results show that the shear forces and strain rates change

48

sharply on the entire surface if no mold core is inserted into the complex hole,

whereas they hardly make any difference when the core shape is similar to the

complex hole. Three experimental types of mold core were used. The results

demonstrate that the similar shape of the mold core inserted into the hole could find

the uniform roughness on the surface.

Williams et al.[193] conducted a study to determine the feasibility of sealing and

finishing conformal cooling/heating channels in profiled edge laminae (PEL) rapid

tooling (RT) using abrasive flow machining (AFM).Sample PEL tools constructed of

both aluminum and steel were designed and assembled for finishing by AFM. Output

parameters of interest included the material removal, surface roughness

improvement and, most importantly, the ability to withstand a pressurized oil leak

test. The findings suggested that AFM significantly improved the finish in the

channels for aluminum and steel PEL tooling. Leak testing found that AFM also

improved the sealing of both stacks at static pressures up to 690 kPa. The steel

tooling appeared to benefit more from the AFM process. It has been postulated that

the primary cause of the sealing is the plastic deformation of workpiece material in

the ploughing mode. Although, the conformal channels studied had a simple cross-

sectional geometry and straight runs. The PEL tools were only made of two

materials. However, the investigations show great promise for large RT, including

thermoforming and composite forming molds where temperature control is a critical

issue. The ability to seal the interfaces between individual laminae expands the

potential application of AFM tremendously. AFM also has the potential to finish a

wide range of internal passages in a variety of RT.AFM has been previously used for

finishing stereolithography prototypes. This is the first known attempt to seal and

finish channels in laminated RT using AFM.

Uhlmann et al. [194] reported AFM is also well suited to process advanced ceramic

materials. Especially advanced ceramics are playing increasingly a significant role as a

substitute for metals. However the high costs for the inevitable finishing process on

ceramics prevent a more frequent use. This paper represents the technological

results of a research-project discovering the fundamental principles of AFM on

advanced ceramic materials such as a correlation between flow processes, surface

formation and edge rounding. Furthermore an insight into a process model is given,

which was developed using modern simulation techniques. The overall objective of

this approach is to anticipate work results like surface quality and edge rounding on

any user-defined geometry.

Barletta [195] reviewed the use of abrasive fluidized bed equipment in a broad range

of manufacturing processes. In particular, applications in deburring and finishing of

complex-shaped metal components, in super-finishing of dies for injection molding,

in cleaning and polishing of electronic devices, and in surface preparation of

49

tungsten carbide milling tools are reviewed. Attention is focused on the effects of

the most important process parameters, such as machining time, abrasive type and

mesh size, and flow or jet speed. The extent of material removal and the change in

surface roughness as a function of the process parameters are addressed. Selected

numerical and analytical models that are useful for automation and control purposes

are discussed. Finally, the industrial sustainability of the processes and equipment

investigated is highlighted.

The literature available on AFM indicates that certain good research initiatives have

been undertaken in the direction of development, applications and capabilities of

the process [17,196], application of acoustic emission technique for the adaptive

control of the process [142], effect of temperature, composition and other

rheological characteristics of media on performance of AFM [61], modelling of the

process with artificial neural networks (ANN), response surface methodology (RSM),

Genetic Algorithm (GA), etc. [123,197,198,199], analytical and finite element

modeling (FEM) of material removal (MR) and surface generation [91], enhancement

of efficiency and capabilities of AFM [56,73,183,199]. However, each of these studies

considered a subset of the process parameters and ignored other critical

parameters. Some research studies on AFM give conflicting opinions as regards to

the change in surface roughness and material removal rate in various materials. It

has been recognized that very few research studies have been conducted on the

optimization of the process parameters for enhanced quality characteristics.

Therefore, it is required to study the AFMed surfaces to get more insight into the

real interaction between flowing abrasive particles and the target surfaces.