Multi Objective Optimization of Cutting Parameters in Turning

Operation Using Taguchi Method

Thesis submitted in partial fulfillment of the requirements for the Degree of

Bachelor of Technology (B. Tech.)

In

Mechanical Engineering

By

RAJA SUTAR

Roll No. 110ME0336

Under the Supervision of

Prof. K.P.MAITY

DEPARTMENT OF MECHANICAL ENGINEERING

NATIONAL INSTITUTE OF TECHNOLOGY, ROURKELA

May, 2014

1
 NATIONAL INSTITUTE OF TECHNOLOGY

ROURKELA-769008,

INDIA

Certificate of Approval
This is to certify that the thesis entitled Multi Objective Optimization of Cutting Parameters

in Turning Operation Using Taguchi Method submitted by Raja Sutar has been carried out

under my supervision in partial fulfillment of the requirements for the Degree of Bachelor of

Technology in Mechanical Engineering at National Institute of Technology, Rourkela, and this

work has not been submitted elsewhere before for any other academic degree/diploma.

Dr. K .P.MAITY

H.O.D

Date: Department of Mechanical Engineering

National Institute of Technology, Rourkela

2
 Acknowledgement

I wish to express my profound gratitude and indebtedness to Prof. K. P. MAITY, Head of

the department, National Institution of Technology, Rourkela for introducing the recent topic and

for their inspiring guidance and valuable suggestion throughout this project work.

I would also like to thank Mr. Swastik Pradhan, Ph. D. Scholar and Mr. Dilip Kumar

Bagal, M. Tech. research scholar of Production Engineering specialization for his consistent

assistance and help in carrying out experiments. Last but not the least, my sincere thanks goes to

all my friends who have extended all types of help for accomplishing this undertaking.

Date: RAJA SUTAR (110ME0336)

Department of Mechanical Engineering

National Institute of Technology, Rourkela

3
 ABSTRACT

This paper refers to parametric optimization of turning process applying Taguchi method

in order to improve quality of manufacturing goods, and engineering development of designs for

studying variation.SS-304 is used as work-piece for carrying out experiment to optimize material

removal rate and surface roughness.

There are three machining parameters i.e. spindle speed, feed rate and depth of cut.

Different experiments are done by varying one parameter and keeping other two fixed so that

optimized value of each parameter can be obtained. In this project dry turning of S. S. 304

graded steel as a work piece and carbide insert tool (SNMG120408MS, SNMG432MS) is

performed. The range of cutting parameters are cutting speed (40, 66 and 92 m/min), feed rate

(0.05, 0.1 and 0.15 mm/rev), depth of cut (0.25, 0.5 and 0.75 mm).

Taguchi orthogonal array is designed with three levels of turning parameters with the help

of software Minitab version 16. Taguchi method stresses the importance of studying the

response variation using the signal to noise (S/N) ratio, resulting in minimization of quality

characteristic variation due to uncontrolled parameters. It is predicted that Taguchi method is a

good method for optimization of various machining parameters as it reduces number of

experiments. The results indicate the optimum values of the input factors and the results are

conformed by a confirmatory test.

Keywords: - Depth of Cut; Feed; Speed; Spindle Speed; Taguchi Orthogonal Array;

4
 CONTENTS

Chapter no Description Page no.

Acknowledgement 3
Abstract 4
List of Tables 6
List of Figures 7
Chapter-1
1.1 Introduction and literature review 9
1.2 Objective of the work 25
Chapter-2
2.1 Cutting Tool Specification 28
2.2 Composition of work piece 28
2.3 Talysurf 29
2.4 Procedure followed 31
Chapter-3
3.1 Taguchi method 33
Chapter-4
4.1 Experimental observation & Analysis 36
Chapter-5
5.1 Conclusions 54

5
 List of tables

Table no Description Page no

2.1 Specification of cutting tool 28
2.2 composition of SS-304 29
2.3 mechanical properties of 29
SS-304
2.4 Taguchi design of experiment 37
4.1 observation table 39
4.2 Estimated Model Coefficients 40
for SN ratios
4.3 Analysis of Variance for SN 41
ratios
4.4 Estimated Model Coefficients 42
for Means
4.5 Analysis of Variance for 43
Means
4.6 Response Table for Signal to 44
Noise Ratios
(Smaller-is-better)
4.7 Response Table for Means 45
4.12 Response Table for Signal to 46
Noise Ratios
(Smaller-is-better)
4.13 Response Table for Means 47

6
 List of Figures

Figure no Description Page no

1.1 Nomenclature of a single point 10
cutting tool
1.2 machining process and the 11
principal cutting-tool elements
1.3 Cutting angles 11
2.1 carbide insert 28
2.2 Talysurf 29
2.3 Stylus based instruments 29
2.4 Measurement of Ra 41
2.5 Workpiece 42
4.1 Main Effects Plot for Means 43
4.2 Main Effects Plot for SN 43
ratios
4.3 Normplot of Residuals for SN 44
ratios
4.4 Residuals vs Fits for SN ratios 45
4.5 Residual Histogram for SN 45
ratios
4.6 Residuals vs Order for SN 46
ratios
4.7 Normplot of Residuals for 46
Means
4.8 Residuals vs Fits for Means 47
4.9 Residual Histogram for Means 47
4.10 Residuals vs Order for Means 48
4.11 Main Effects Plot for Means 48
4.12 Main Effects Plot for SN 49
ratios
4.13 Normplot of Residuals for SN 49
ratios
4.14 Residuals vs Fits for SN ratios 50
4.15 Residual Histogram for SN 50
ratios
4.16 Residuals vs Order for SN 51
ratios
4.17 Normplot of Residuals for 51
Means
4.18 Residuals vs Fits for Means 52

7
CHAPTER 1

8
 CHAPTER 1:

INTRODUCTION

1. Introduction

Turning is one of the major machining processes which includes metal cutting as removal

of metal chips in order to get finished product of desired shape, size and surface roughness. The

engineers have to face challenge in order to get optimal parameters for preferred output using

available sources.

Usually selection of machining parameters is very much difficult for desired product.

Actually it depends upon experience of the engineers and the table given by machine-tool

designer. So the importance of optimization arises in order to satisfy economy and quality of

machined part.

The Taguchi’s method tells about reduction in variation in order to improve quality by

method of offline or online quality control. The offline quality control helps in improving

quality of processes, where online quality control helps in maintaining conformance to the

original or intended design. The main and fundamental part of Taguchi’s design is to ensure that

the product perform well even in noise; it helps in making the product long lasting. Taguchi

method is applied in a very short period of time without lots of efforts. That is why Taguchi’s

method is adopted in various industries in order to improve the process quality in manufacturing

sectors.

Surface roughness and cutting force are two very important parameters in machining

process. Cutting force is necessary for calculation of power machining. Cutting forces influences

dimensional accuracy, deformation of work-piece and chip formation.

9
 Components of certain surface roughness are always required in industries as per

customer requirement. This can be achieved by optimization process which we are going to

discuss about.

1.1. Single Point Cutting Tool

Fig. 1 Nomenclature of a single point cutting tool(refer 40)

Single point cutting tools have one principal cutting edge which is mainly used for

cutting. These tools are used for turning, boring, planning etc. used in machines like lathe, boring

and shaping machines. Single point cutting tools contain following parts: - shank (this is the

main body of the tool), flank (which is adjacent below the cutting edge), face (the surface upon

which chip slides), nose radius (it is the point where cutting edge intersects with side cutting

edge). The schematic diagram of single point cutting tool is shown in Fig. 1.

10
 Fig. 2 Diagram of the machining process(refer 37)

The angle between the side relief face of the tool and machining plane is called side relief

angle α. The relief angle depends upon rate of feed parameter, if feed increases, then relief angle

increases in order to avoid friction between relief surface and cutting edge.

The angle between the top and side relief surface of the tool is known as lip angle β. The

angle between the plane perpendicular to the cutting plane and the top surface of the tool is

known as side rake angle γ. The mechanism of turning process is shown in Fig. 2.

Fig. 3 Cutting angles(refer 37)

Larger rake angles facilitate easier formation of chip in machining, but it decreases

cutting force(less power consumption). For hard material, tool with small rake angle is always

11
used. Finally, it can be observed that rake angle depends upon physical and mechanical

properties of work-piece material. The cutting angles which are used for machining shown in

Fig. 3 with projected point of view.

Turning is the removal of metal from the outer diameter of a rotating cylindrical work

piece. Turning is used to reduce the diameter of the work piece, usually to a specified dimension,

and to produce a smooth finish on the metal can be defined as the machining of an external

surface:

• With the work piece rotating

• With a single-point cutting tool, and

• With the cutting tool feeding parallel to the axis of the work piece and at a

• Distance that will remove the outer surface of the work.

1.2. Cutting factors in turning

The primary factors in any basic turning operation are speed, feed, and depth of cut.

Other factors such as kind of material and type of tool have a large influence, of course, but these

three are the ones the operator can change by adjusting the controls.

1.2.1. Speed:

Speed refers to the spindle and the work piece. When it is stated in rpm, it tells their

rotating speed. But the important feature for a particular turning operation is the surface speed

that is the speed at which the work piece material is moving past the cutting tool. It is simply the

product of the rotating speed times the circumference of the work piece before the cut is started.

It is expressed in meter per minute (m/min), and it refers only to the work piece. Every different

diameter on a work piece will have a different cutting speed, even though the rotating speed

remains the same.

12
v = 3.14DN/1000 (m/min)

Here, v is the cutting speed in turning; D is the initial diameter of the work piece in mm.

1.2.2. Feed:

Feed refers to the cutting tool and it is the rate at which the tool advances along its cutting

path. In most of power-fed lathes, the feed rate is directly related to the spindle speed and is

expressed in mm (of tool advance) per revolution (of the spindle), or mm/rev.

1.2.3. Depth of Cut:

Depth of cut is the thickness of the layer being removed (in a single pass) from the work

piece or the distance from the uncut surface of the work to the cut surface, expressed in mm. It is

important to note that the diameter of the work piece is reduced by two times the depth of cut

because this layer is being removed from both sides of the work piece.

The Taguchi’s method tells about reduction in variation in order to improve quality by

method of offline or online quality control. The offline quality control helps in improving

quality of processes, where online quality control helps in maintaining conformance to the

original or intended design. The main and fundamental part of Taguchi’s design is to ensure that

the product perform well even in noise; it helps in making the product long lasting. Taguchi

method is applied in a very short period of time without lots of efforts. That is why Taguchi’s

method is adopted in various industries in order to improve the process quality in manufacturing

sectors.

Taguchi methods provide a cost effective, efficient and systematic way to optimize

designs for performance, quality, and cost. This method has been used successfully in designing

reliable, high-quality products at low cost in such areas as automotive, aerospace, and consumer.

13
 Cutting forces, surface roughness and tool wear are among the most important technical

parameters in machining process. Cutting forces are necessary for evaluation of machine tool

components and the tool body. Cutting forces influences the deformation of the work piece

machined, its dimensional accuracy, machine stability and chip formation.

1.3. Tool Geometry:

For cutting tools, geometry shown in fig 1.3 depends on the properties of the tool

material and the work material. For single point tools, the most important angles are the rake

angles and the end and side relief angle.

(Tool geometry shown in fig 1.3)(refer. 40)

1.3.1. Flank:

A flat surface of a single-point tool that is adjacent to the face of the tool. During turning,

the side flank faces the direction that the tool is fed into the work piece. At the end flank passes

over the newly machined surface.

1.3.2. Face:

The flat surface of a single point tool through which, the work piece rotates during

turning operation is called the face of tool. On a typical turning setup, the face of the tool is

positioned upwards.

14
1.3.3. Back rake angle:

If viewed from the side facing the end of the work piece, it is the angle formed by the

face of the tool, and a line parallel to the base. A positive back rake angle tilts the tool face back,

and a negative angle tilts it forward and up.

1.3.4. Side rake angle:

If viewed behind the tool down the length of the tool holder, it is the angle formed by the

face of the tool and the centerline of the work piece. A positive side rake angle tilts the tool face

down toward the floor, and a negative angle tilts the face up and toward the work piece.

1.3.5. Side cutting edge angle:

If viewed from above looking down on the cutting tool, it is the angle formed by the side

flank of the tool and a line perpendicular to the work piece centerline. A positive side cutting

edge angle moves the side flank into the cut, and a negative angle moves the side flank out of the

cut.

1.3.6. End cutting edge angle:

If viewed from above looking down on the cutting tool, the angle formed by the end flank

of the tool and a line parallel to the work piece centerline is called end cutting edge angle.

Increasing the end cutting edge angle tilts the far end of the cutting edge away from the work

piece.

1.3.7. Side relief angle:

If viewed behind the tool down the length of the tool holder, the angle formed by the side

flank of the tool and a vertical line down to the floor is called side relief angle. Increasing the

side relief angle tilts the side flank away from the work piece.

1.3.8. End relief angle:

15
 If viewed from the side facing the end of the work piece, the angle formed by the end

flank of the tool and a vertical line down to the floor is named as end relief angle. Increasing the

end relief angle tilts the end flank away from the work piece.

1.3.9. Nose radius:

It is the rounded tip on the cutting edge of a single point tool. A sharp point of the cutting

tool is created by a zero degree nose radius.

1.3.10. Lead angle:

Lead angle is the common name for the side cutting edge angle. If a tool holder is built

with dimensions that shift the angle of an insert, the lead angle takes this change into

consideration.

Here, the experts of machining gave their opinions regarding obtained results. A

numerous studies have been carried out regarding turning operation and the applied optimization

techniques.

Zhou et al. (1995) [1] investigated on tool life criteria in raw turning. A new tool-life

criterion depending on a pattern-recognition technique was proposed and neural network and

wavelet techniques were used to realize the new criterion. The experimental results showed

that this criterion was applicable to tool condition monitoring in a wide range of cutting

conditions.

Lin et al. (2001) [2] adopted an abdicative network to construct a prediction

model for surface roughness and cutting force. Once the process parameters: cutting speed,

feed rate 44 and depth of cut were given; the surface roughness and cutting force could be

predicted by this network. Regression analysis was also adopted as second prediction model

for surface roughness and cutting force. Comparison was made on the results of both models

16
indicating that adductive network was found more accurate than that by regression analysis.

Feng and Wang (2002) [3] investigated for the prediction of surface roughness in

finish turning operation by developing an empirical model through considering working

parameters: work piece hardness (material), feed, cutting tool point angle, depth of cut,

spindle speed, and cutting time. Data mining techniques, nonlinear regression analysis with

logarithmic data transformation were employed for developing the empirical model to predict

the surface roughness.

Suresh et al. (2002) [4] focused on machining mild steel by TiN-coated tungsten

carbide (CNMG) cutting tools for developing a surface roughness prediction model by using

Response Surface Methodology (RSM). Genetic Algorithms (GA) used to optimize the

objective function and compared with RSM results. It was observed that GA program

provided minimum and maximum values of surface roughness and their respective optimal

machining conditions.

Lee and Chen (2003) [5] highlighted on artificial neural networks (OSRR-ANN)

using a sensing technique to monitor the effect of vibration produced by the motions of the

cutting tool and work piece during the cutting process developed an on-line surface

recognition system. The authors employed tri-axial accelerometer for determining the

direction of vibration that significantly affected surface roughness then analyzed by using a

statistical method and compared prediction accuracy of both the ANN and SMR.

Choudhury and Bartarya (2003) [6] focused on design of experiments and the neural

network for prediction of tool wear. The input parameters were cutting speed, feed and depth

of cut; flank wear, surface finish and cutting zone temperature were selected as outputs.

Empirical relation between different responses and input variables and also through neural

17
network (NN) program helped in predictions for all the three response variables and compared

which method was best for the prediction.

Chien and Tsai (2003) [7] developed a model for the prediction of tool flank wear

followed by an optimization model for the determination of optimal cutting conditions in

machining 17-4PH stainless steel. The back-propagation neural network (BPN) was used to

construct the predictive model. The genetic algorithm (GA) was used for model optimization.

Kirby et al. (2004) [8] developed the prediction model for surface roughness in turning

operation. The regression model was developed by a single cutting parameter and vibrations

along three axes were chosen for in-process surface roughness prediction system. By using

multiple regression and Analysis of Variance (ANOVA) a strong linear relationship among

the parameters (feed rate and vibration measured in three axes) and the response (surface

roughness) was found. The authors demonstrated that spindle speed and depth of cut might

not necessarily have to be fixed for an effective surface roughness prediction model.

Őzel and Karpat (2005) [9] studied for prediction of surface roughness and tool flank

wear by utilizing the neural network model in comparison with regression model. The data set

from measured surface roughness and tool flank wear were employed to train the neural

network models. Predictive neural network models were found to be capable of better

predictions for surface roughness and tool flank wear within the range in between they were

trained.

Luo et al. (2005) [10] carried out theoretical and experimental studies to investigate

the intrinsic relationship between tool flank wear and operational conditions in metal cutting

processes using carbide cutting inserts. The authors developed the model to predict tool flank

wear land width which combined cutting mechanics simulation and an empirical model. The

18
study revealed that cutting speed had more dramatic effect on tool life than feed rate.

Kohli and Dixit (2005) [11] proposed a neural-network-based methodology with the

acceleration of the radial vibration of the tool holder as feedback. For the surface roughness

prediction in turning process the back-propagation algorithm was used for training the

network model. The methodology was validated for dry and wet turning of steel using high

speed steel and carbide tool and observed that the proposed methodology was able to make

accurate prediction of surface roughness by utilizing small sized training and testing datasets.

Pal and Chakraborty (2005) [12] studied on development of a back propagation neural

network model for prediction of surface roughness in turning operation and used mild steel

work-pieces with high speed steel as the cutting tool for performing a large number of

experiments. The authors used speed, feed, depth of cut and the cutting forces as inputs to the

neural network model for prediction of the surface roughness. The work resulted that

predicted surface roughness was very close to the experimental value.

Sing and Kumar (2006) [13] studied on optimization of feed force through setting of

optimal value of process parameters namely speed, feed and depth of cut in turning of EN24

steel with TiC coated tungsten carbide inserts. The authors used Taguchi’s parameter design

approach and concluded that the effect of depth of cut and feed in variation of feed force were

affected more as compare to speed.

Ahmed (2006) [14] developed the methodology required for obtaining optimal process

parameters for prediction of surface roughness in Al turning. For development of empirical

model nonlinear regression analysis with logarithmic data transformation was applied. The

developed model showed small errors and satisfactory results. The study concluded that low

feed rate was good to produce reduced surface roughness and also the high speed could

19
produce high surface quality within the experimental domain.

Abburi and Dixit (2006) [15] developed a knowledge-based system for the prediction of

surface roughness in turning process. Fuzzy set theory and neural networks were utilized for

this purpose. The authors developed rule for predicting the surface roughness for given

process variables as well as for the prediction of process variables for a given surface

roughness.

Zhong et al. (2006) [16] predicted the surface roughness of turned surfaces using

networks with seven inputs namely tool insert grade, work piece material, tool nose radius,

rake angle, depth of cut, spindle rate, and feed rate.

Kumanan et al. (2006) [17] proposed the methodology for prediction of machining

forces using multi-layered perceptron trained by genetic algorithm (GA). The data obtained

from experimental results of a turning process were explored to train the proposed artificial

neural networks (ANNs) with three inputs to get machining forces as output. The optimal

ANN weights were obtained using GA search. This function-replacing hybrid made of GA

and ANN was found computationally efficient as well as accurate to predict the machining

forces for the input machining conditions.

Mahmoud and Abdelkarim (2006) [18] studied on turning operation using High-Speed

Steel (HSS) cutting tool with 450 approach angle. This tool showed that it could perform

cutting operation at higher speed and longer tool life than traditional tool with 90 degree

approach angle. The study finally determined optimal cutting speed for high production rate

and minimum cost, tool like, production time and operation costs.

Doniavi et al. (2007) [19] used response surface methodology (RSM) in order to

develop empirical model for the prediction of surface roughness by deciding the optimum

20
cutting condition in turning. The authors showed that the feed rate influenced surface

roughness remarkably. With increase in feed rate surface roughness was found to be

increased. With increase in cutting speed the surface roughness decreased. The analysis of

variance was applied which showed that the influence of feed and speed were more in surface

roughness than depth of cut.

Kassab and Khoshnaw (2007) [20] examined the correlation between surface roughness

and cutting tool vibration for turning operation. The process parameters were cutting speed,

depth of cut, feed rate and tool overhanging. The experiments were carried out on lathe using

dry turning (no cutting fluid) operation of medium carbon steel with different level of

aforesaid process parameters. Dry turning was helpful for good correlation between surface

roughness and cutting tool vibration because of clean environment. The authors developed

good correlation between the cutting tool vibration and surface roughness for controlling the

surface finish of the work pieces during mass production. The study concluded that the

surface roughness of work piece was observed to be affected more by cutting tool

acceleration; acceleration increased with overhang of cutting tool. Surface roughness was

found to be increased with increase in feed rate.

Al-Ahmari (2007) [21] developed empirical models for tool life, surface roughness and

cutting force for turning operation. The process parameters used in the study were speed, feed,

depth of cut and nose radius to develop the machinability model. The methods used 48 for

developing aforesaid models were Response Surface Methodology (RSM) and neural

networks (NN).

Thamizhmanii et al. (2007) [22]applied Taguchi method for finding out the optimal

value of surface roughness under optimum cutting condition in turning SCM 440 alloy steel.

21
The experiment was designed by using Taguchi method and experiments were conducted and

results thereof were analyzed with the help of ANOVA (Analysis of Variance) method. The

causes of poor surface finish as detected were machine tool vibrations, tool chattering whose

effects were ignored for analyses. The authors concluded that the results obtained by this

method would be useful to other researches for similar type of study on tool vibrations,

cutting forces etc. The work concluded that deppth of cut was the only significant factor

which contributed to the surface roughness.

Natarajan et al. (2007) [23] presented the on-line tool wear monitoring technique in

turning operation. Spindle speed, feed, depth of cut, cutting force, spindle-motor power and

temperature were selected as the input parameters for the monitoring technique. For finding

out the extent of tool wear; two methods of Hidden Markov Model (HMM) such as the Bar-

graph Method and the Multiple Modeling Methods were used. A decision fusion centre

algorithm (DFCA) was used for increasing the reliability of this output which combined the

outputs of the individual methods to make a global decision about the wear status of the tool.

Finally, all the proposed methods were combined in a DFCA to determine the wear status of

the tool during the turning operations.

Ozel et al. (2007) [24] carried out finish turning of AISI D2 steels (60 HRC) using

ceramic wiper (multi-radii) design inserts for surface finish and tool flank wear investigation.

For prediction of surface roughness and tool flank wear multiple linear regression models and

neural network models were developed. Neural network based predictions of surface

roughness and tool flank wear were carried out, compared with a non-training experimental

data and the results thereof showed that the proposed neural network models were efficient to

predict tool wear and surface roughness patterns for a range of cutting conditions. The study

22
concluded that best tool life was obtained in lowest feed rate and lowest cutting speed

combination.

Wang and Lan (2008) [25] used Orthogonal Array of Taguchi method coupled with

grey relational analysis considering four parameters viz. speed, cutting depth, feed rate, tool

nose run off etc. for optimizing three responses: surface roughness, tool wear and material

removal rate in precision turning on an ECOCA-3807 CNC Lathe. The MINITAB software

was explored to analyze the mean effect of Signal-to-Noise (S/N) ratio to achieve the multi-

objective features. This study not only proposed an optimization approaches using Orthogonal

Array and grey relational analysis but also contributed a satisfactory technique for improving

the multiple machining performances in precision CNC turning with profound insight.

Srikanth and Kamala (2008) [26] evaluated optimal values of cutting parameters by

using a Real Coded Genetic Algorithm (RCGA) and explained various issues of RCGA and

its advantages over the existing approach of Binary Coded Genetic Algorithm (BCGA). They

concluded that RCGA was reliable and accurate for solving the cutting parameter

optimization and construct optimization problem with multiple decision variables. These

decision variables were cutting speed, feed, depth of cut and nose radius. The authors

highlighted that the faster solution can be obtain with RCGA with relatively high rate of

success, with selected machining conditions thereby providing overall improvement of the

product quality by reduction in production cost, reduction in production time, flexibility in

machining parameter selection.

Sahoo et al. (2008) [27] studied for optimization of machining parameters

combinations emphasizing on fractal characteristics of surface profile generated in CNC

turning operation. The authors used L27 Taguchi Orthogonal Array design with machining

23
parameters: speed, feed and depth of cut on three different work piece materials viz.

aluminum, mild steel and brass. It was concluded that feed rate was more significant

influencing surface finish in all three materials. It was observed that in case of mild steel and

aluminum feed showed some influences while in case of brass depth of cut was noticed to

impose some influences on surface finish. The factorial interaction was responsible for

controlling the fractal dimensions of surface profile produced in CNC turning.

Reddy et al. (2008) [28] adopted multiple regression model and artificial neural

network to deal with surface roughness prediction model for machining of aluminium alloys

by CNC turning. For judging the efficiency and ability of the model in surface roughness

prediction the authors used the percentage deviation and average percentage deviation. The

study of experimental results showed that the artificial neural network was efficient as

compared to multiple regression models for the prediction of surface roughness.

Wannas (2008) [29] carried out experiments for hard turning of graphitic cast iron

for the prediction of status of tool wear by using radial basis function neural network

(RBFNN) model. The RBFNN had three inputs: speed, feed and depth of cut and one output:

state variable node. The error was less as obtained from neural network model than the

regression model.

Lan et al. (2008) [30] considered four cutting parameters: speed, feed, depth of

cut, and nose runoff varied in three levels for predicting the surface roughness of CNC turned

product.

Thamma (2008) [31] constructed the regression model to find out the optimal

combination of process parameters in turning operation for Aluminium 6061 work pieces. The

study highlighted that cutting speed, feed rate, and nose radius had a major impact on surface

24
roughness. Smoother surfaces could be produced when machined with a higher cutting speed,

smaller feed rate, and smaller nose radius.

Fnides et al. (2008) [32] studied on machining of slide-lathing grade X38CrMoV5-1

steel treated at 50 HRC by a mixed ceramic tool (insert CC650) to reveal the influences of

cutting parameters: feed rate, cutting speed, depth of cut and flank wear on cutting forces as

well as on surface roughness. The authors found that tangential cutting force was very

sensitive to the variation of cutting depth. It was observed that surface roughness was very

sensitive to the variation of feed rate and that flank wear had a great influence on the

evolution of cutting force components and on the criteria of surface roughness.

Biswas et al. (2008) [33] studied that on-line flank wear directly influenced the

power consumption, quality of the surface finish, tool life, productivity etc. The authors

developed a Neuro-Fuzzy model for prediction of the tool wear. From the orthogonal

machining of aluminium with high-speed steel tool for various rake angles, feed and velocity

the experimental data were obtained and input along with other machining parameters ratio

between cutting force and tangential forces was collected. These were used to predict the tool

wear. The final parameters of the model were obtained by tuning the crude values obtained

from mountain clustering method by using back-propagation learning algorithm and finally

the present Neuro-Fuzzy system which predicted the flank.

Fu and Hope (2008) [34] established an intelligent tool condition monitoring

system by applying a unique fuzzy neural hybrid pattern recognition system. The study

concluded that armed with the advanced pattern recognition methodology, the established

intelligent tool condition monitoring system had the advantages of being suitable for different

machining conditions, robust to noise and tolerant to faults.

25
 Wang et al. (2008) [35] studied on Hybrid Neural Network-based modeling

approach integrated with an analytical tool wear model and an artificial neural network that

was used to predict CBN tool flank wear in turning of hardened 52100 bearing steel.

Experimental results showed that the proposed Hybrid Neural Network excelled the analytical

tool wear model approach and the general neural network-based modeling aproach.

1.2. Objective of Work

The objective of work is to observe the cutting parameters in turning and to calculate the

optimum value of the parameters in order to optimize the surface roughness and tool wear using

Taguchi method. The statistical analysis is to be performed for better machining operation which

can be used for quality control of machining parts. This will help to concerned R and D

researchers or industrial experts.

26
CHAPTER 2

27
 CHAPTER 2:

BRIEF DESCRIPTION OF APPARATUS USED

2. Cutting tool specification

The specification of cutting tool used is SNMG 120408MS and the dimensions are as

follows.

Table 1: Specification of cutting tool (mm)

Cutting Edge Inscribed Circle or Hole Corner Side

Thickness
Length Height Diameter Radius Clearance
12.7 12.7 4.76 5.16 0.8 0°

2.1. Composition and application of work piece

S.S. 304 is a most widely used austenitic steel popularly known as 18/8 stainless steel.

The Fig. 4 shows the experimental set up of turning operation with work piece of S. S. 304

graded steel.

Fig. 4 Experimental set up for machining

28
2.2. Physical data

Table 2: S. S. 304 steel physical data

Density (lb/cu.in.) 0.285

Specific Gravity 7.9
Specific Heat (Btu/lb/Deg F) 0.12
Electrical Resistivity (microhm-cm) 432
Melting Point (Deg F) 2650
Modulus of Elasticity Tension 28

2.3. Composition

Table 3: Chemical composition of S. S. 304 steel

C Mn Si Cr Ni P S
0.08 2.0 1.0 18-20 8-10.05 0.045 0.03

2.4. Material notes

Essentially non-magnetic, becomes slightly magnetic when cold worked, it has excellent

corrosion and forming characteristics and highly ductile. This has corrosion resistance mostly

with oxidizing acids and salt spray.

2.5. Applications

This steel is widely used in petrochemical industries, diary, household and

pharmaceutical purposes and cryogenic vessels, and as heat exchanger in air conditioning

refrigeration factories.

2.6. Talysurf measurements

Talysurf (shown in Fig. 5) is a device used for measurement of surface roughness which

known as Portable profilometer (Taylor Hobson Surtronic 3+). This is a stylus based instrument

29
which is based on the principle of a probe running across the surface to measure the variation of

height as a function of distance.

Fig. 5 Talysurf

In this instrument, a transducer is used which converts the vertical distance into electrical

signal. The signal is processed later in order to get the surface roughness value. Some error may

arise due to some factors like stylus speed, stylus load, size and shape of the stylus and lateral

deflection.

2.7. Centre line average roughness

It refers to arithmetic average or center line average (CLA). It is denoted as Ra. This is

calculated as the average distance of the surface from the mean line looking at all the points of

the profile.

30
 (reference 37)Fig. 6 Measurement of Ra

The principle regarding measuring surface roughness is described in Fig. 6. Based on this

method of the profilomer of Talysurf the surface roughness is measured and recorded as per each

run.

2.8. Procedure followed for experimentation

First of all, the work-piece (S.S. 304) is mounted on the head stock of lathe. The other end

of material is center-bored using center drill and then fixed with the tail stock respectively. Then,

according to the design of experiment (Table 4), different levels of parameters are set to get 9

numbers of run.

The work piece was given initial roughness pass. Then the surface roughness (R a) value is

calculated for each run using Talysurf. And tool wear at each respective cutting edge of the tool

is also calculated using Tool maker microscope. Statistical analysis of obtained data carried out

using Taguchi optimization technique.

31
CHAPTER 3

32
 Chapter 3:

Methodology

3.1. Taguchi Method

Taguchi method is one of the powerful tools for optimization technique. This is based on

the “Orthogonal Array” experiments which give much optimum setting of control parameters

with reduced variance. So after optimization the best result is obtained from the Taguchi method.

“Taguchi Orthogonal Array” gives a well-balanced (minimum) set of parameter. The signal-to-

noise ratios which are log function of desired output serves as objective function for method of

optimization which helps in data analysis and finding the optimal result. The optimization

problems which involve selection of best levels of parameters in order to get a desired output are

called as “static problem”.

There are three signal-to-noise ratios of common interest for optimization of static

problems:-

1) Smaller-The-Better

n = -10 Log10 [mean of sum of squares of measured data] (1)

This is usually the chosen S/N ratio for all undesirable characteristics for which the ideal

value is zero. But when the ideal value is zero, then the difference between measured data and

ideal value is expected to be as small as possible. The generic form of S/N ratio becomes:-

n = -10 Log10 [mean of sum of squares of {measured - ideal}] (2)

2) Larger-The-Better

n = -10 Log10 [mean of sum squares of reciprocal of measured data] (3)

33
 By taking the reciprocals of measured data and taking the value of S/N ratio as in

smaller-the-better case, we can convert it to smaller-the-better case.

3) Nominal-The-Best

n = 10 Log10 (square of mean/ variance) (4)

This case arises when a specified value is most desired, meaning that neither a smaller

nor a larger value is desirable.

34
CHAPTER 4

35
 Chapter 4:

Experimental Observation & Analysis

According to Taguchi’s orthogonal array theory L9 orthogonal array is adopted for the

whole experimentation for turning operation of S. S. 304 graded steel. In L 9 orthogonal array, 9

experimental runs are conducted and the corresponding out puts is evaluated by Taguchi

optimization technique. Here, Tool wears and means of surface roughness are measured by

above said instruments and these values are taken out put responses in Taguchi optimization

method. Table 4 shows the standard structure of L9 orthogonal array which levels of each

parameters are taken as 1, 2 and 3 respectively.

Table 4: Taguchi orthogonal array

Sl. no. Cutting speed Feed Depth of cut

1 1 1 1
2 1 2 2
3 1 3 3
4 2 1 3
5 2 2 1
6 2 3 2
7 3 1 2
8 3 2 3
9 3 3 1

Here, the process variables are cutting speed, feed rate and depth of cut. These are the

input parameters for the Taguchi optimization. So, nine experiments are carried out as per this

orthogonal array and corresponding output data are recorded serially. The surface roughness was

measured thrice times at different parts of the surface of work piece and then calculate the mean

of those value. The full structure of experimentation is tabulated in Table 5 as per L9 orthogonal

array.

36
 Table 5: Observation table

Feed Doc Tool Wear S. R. S. R. S. R. Average

Run Speed (Vc)
(f) (d) (micron) (1) (2) (3) S. R.
1 40 0.15 0.75 0.810 1.84 1.64 1.76 1.746
2 40 0.1 0.5 1.051 1.36 1.16 1.24 1.253
3 40 0.05 0.25 1.008 1.42 1.44 1.52 1.46
4 66 0.15 0.25 1.127 1.44 1.28 1.3 1.34
5 66 0.1 0.75 0.800 1.98 1.74 1.86 1.86
6 66 0.05 0.5 0.659 1.66 1.58 1.66 1.63
7 92 0.15 0.5 0.928 1.88 1.74 1.84 1.82
8 92 0.1 0.25 1.302 1.52 1.40 1.38 1.43
9 92 0.05 0.75 1.381 3.56 3.28 3.42 3.42

37
CHAPTER 5

38
 Chapter 5:

Results & Analysis

These above data are analyzed by a power full statistical tool named Minitab software of

latest version 16. First of all the input parameters are defined in the software as per their

corresponding value and then give the responses data to optimize. Here, the main objective of the

problem is to minimize the value of tool wear and surface roughness. So, the criterion of

Smaller-The-Better is adopted for the optimization. The analysis of S/N ratio and Means are

carried out by the software. Then, ANOVA analysis of each parameter is done after simulation

of optimization. And lastly the influences of residuals on parameters are carried out by plotting

graphs.

5.1. Taguchi Analysis: Tool Wear versus Speed, Feed, Doc

The following terms cannot be estimated, and were removed.

Speed*Feed

Speed*Doc

Feed*Doc

The interaction terms are not be analyzed by Taguchi.

5.2. Linear Model Analysis: S/N ratios versus Speed, Feed, Doc

Table 6: Estimated Model Coefficients for S/N ratios

Term Coef. S.E. Coef. T P

Constant 0.15600 0.7928 0.197 0.862
Speed 40 0.28701 1.1212 0.256 0.822
Speed 66 1.35134 1.1212 1.205 0.351
feed 0.05 0.09374 1.1212 0.084 0.941
feed 0.10 -0.41802 1.1212 -0.373 0.745
doc 0.25 -1.28930 1.1212 -1.150 0.369
doc 0.50 1.12376 1.1212 1.002 0.422
S = 2.378 R-Sq = 67.5% R-Sq (adj) = 0.0%

39
 Table 7: Analysis of Variance for S/N ratios

%
Source D.O.F. Seq. S. S. Adj. S. S. Adj. M. S. F P
contribution
Speed 2 13.7781 13.7781 6.8890 1.22 0.0451 39.57
Feed 2 0.8661 0.8661 0.4330 0.08 0.0929 2.49
Doc 2 8.8576 8.8576 4.4288 0.78 0.0561 25.44
Residual
2 11.3140 11.3140 5.6570 32.49
Error
Total 8 34.8157 100

5.3. Linear Model Analysis: Means versus Speed, Feed, Doc

Table 8: Estimated Model Coefficients for Means

Term Coef SE Coef T P

Constant 1.00733 0.08956 11.247 0.008
Speed 40 -0.05100 0.12666 -0.403 0.726
Speed 66 -0.14533 0.12666 -1.147 0.370
feed 0.05 0.00867 0.12666 0.068 0.952
feed 0.10 0.04367 0.12666 0.345 0.763
doc 0.25 0.13833 0.12666 1.092 0.389
doc 0.50 -0.12800 0.12666 -1.011 0.419
S = 0.2687 R-Sq = 68.1% R-Sq (adj) = 0.0%

Table 9: Analysis of Variance for Means

Source DF Seq SS Adj SS Adj MS F P

Speed 2 0.18681 0.18681 0.093404 1.29 0.436
Feed 2 0.01416 0.01416 0.007081 0.10 0.911
Doc 2 0.10688 0.10688 0.053440 0.74 0.575
Residual Error 2 0.14439 0.14439 0.072194
Total 8 0.45224

Table 10: Response Table for Signal to Noise Ratios

Smaller-is-better

Level Speed Feed Doc

1 0.4430 0.2497 -1.1333
2 1.5073 -0.2620 1.2798
3 -1.4824 0.4803 0.3215
Delta 2.9897 0.7423 2.4131
Rank 1 3 2

40
 Table 11: Response Table for Means

Level Speed Feed Doc

1 0.9563 1.0160 1.1457
2 0.8620 1.0510 0.8793
3 1.2037 0.9550 0.9970
Delta 0.3417 0.0960 0.2663
Rank 1 3 2

Main Effects Plot for Means

Data Means
Speed feed
1.2

1.1

1.0
Mean of Means

0.9

40 66 92 0.05 0.10 0.15

doc
1.2

1.1

1.0

0.9

0.25 0.50 0.75

Fig. 7 Main Effects Plot for Means

Main Effects Plot for SN ratios

Data Means
Speed feed

0
Mean of SN ratios

-1

40 66 92 0.05 0.10 0.15

doc

-1

0.25 0.50 0.75

Signal-to-noise: Smaller is better

Fig. 8 Main Effects Plot for S/N ratios

41
 Normal Probability Plot
(response is SN ratios)
99

95
90

80
70

Percent 60
50
40
30
20

10

1
-3 -2 -1 0 1 2 3
Standardized Residual

Fig. 9 Normal plot of Residuals for S/N ratios

Versus Fits
(response is SN ratios)
1.0

0.5
Standardized Residual

0.0

-0.5

-1.0

-1.5
-4 -3 -2 -1 0 1 2 3
Fitted Value

Fig. 10 Residuals vs Fits for S/N ratios

42
 Histogram
(response is SN ratios)

3.0

2.5

2.0

Frequency
1.5

1.0

0.5

0.0
-1.5 -1.0 -0.5 0.0 0.5 1.0
Standardized Residual

Fig. 11 Residual Histogram for S/N ratios

Versus Order
(response is SN ratios)
1.0

0.5
Standardized Residual

0.0

-0.5

-1.0

-1.5
1 2 3 4 5 6 7 8 9
Observation Order

Fig. 12 Residuals vs Order for S/N ratios

43
 Normal Probability Plot
(response is Means)
99

95
90

80
70
Percent 60
50
40
30
20

10

1
-3 -2 -1 0 1 2 3
Standardized Residual

Fig. 13 Normal plot of Residuals for Means

Versus Fits
(response is Means)
1.5

1.0
Standardized Residual

0.5

0.0

-0.5

-1.0
0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
Fitted Value

Fig. 14 Residuals vs Fits for Means

44
 Histogram
(response is Means)

3.0

2.5

2.0
Frequency
1.5

1.0

0.5

0.0
-1.0 -0.5 0.0 0.5 1.0 1.5
Standardized Residual

Fig. 15 Residual Histogram for Means

Versus Order
(response is Means)
1.5

1.0
Standardized Residual

0.5

0.0

-0.5

-1.0
1 2 3 4 5 6 7 8 9
Observation Order

Fig. 16 Residuals vs Order for Means

5.4. Taguchi Analysis: S. R. versus Speed, Feed, Doc

The following terms cannot be estimated, and were removed.

Speed*Feed

Speed*Doc

Feed*Doc

45
5.5. Linear Model Analysis: S/N ratios versus Speed, Feed, Doc

(table 4.8:-Estimated Model Coefficients for SN ratios)

Term Coef. S. E. Coef. T P

Constant -4.5835 0.2885 -15.886 0.004
Speed 40 1.2212 0.4080 2.993 0.096
Speed 66 0.5248 0.4080 1.286 0.327
feed 0.05 -1.4869 0.4080 -3.644 0.068
feed 0.10 1.0982 0.4080 2.691 0.115
doc 0.25 1.6049 0.4080 3.933 0.059
doc 0.50 0.7821 0.4080 1.917 0.195
S = 0.8656 R-Sq = 97.2% R-Sq(adj) = 88.8%

(table 4.9:-Analysis of Variance for S/N ratios)

%
Source D. O. F. Seq. S. S. Adj. S. S. Adj. M. S. F P
contribution
Speed 2 14.446 14.446 7.2231 9.64 0.094 27.10
Feed 2 10.704 10.704 5.3521 7.14 0.123 20.08
Doc 2 26.656 26.656 13.3280 17.79 0.053 50.00
Residual
2 1.499 1.499 0.7493 2.81
Error
Total 8 53.305 100
5.6. Linear Model Analysis: Means versus Speed, Feed, Doc

(table 4.10:-Estimated Model Coefficients for Means)

Term Coef SE Coef T P

Constant 1.7732 0.1199 14.784 0.005
Speed 40 -0.2869 0.1696 -1.691 0.233
Speed 66 -0.1632 0.1696 -0.962 0.437
feed 0.05 0.3968 0.1696 2.339 0.144
feed 0.10 -0.2589 0.1696 -1.526 0.266
doc 0.25 -0.3632 0.1696 -2.141 0.166
doc 0.50 -0.2056 0.1696 -1.212 0.349
S = 0.3598 R-Sq = 92.4% R-Sq(adj) = 69.7%

46
(table 4.11:-Analysis of Variance for Means)

Source D. O. F. Seq. S. S. Adj. S. S. Adj. M. S. F P

Speed 2 0.9346 0.9346 0.4673 3.61 0.217
Feed 2 0.7304 0.7304 0.3652 2.82 0.262
Doc 2 1.4931 1.4931 0.7465 5.77 0.148
Residual Error 2 0.2589 0.2589 0.1295
Total 8 3.4171
(table 4.12:-Response Table for Signal to Noise Ratios)

Smaller-is-better

Level Speed Feed Doc

1 -3.362 -6.070 -2.979
2 -4.059 -3.485 -3.801
3 -6.330 -4.195 -6.971
Delta 2.967 2.585 3.992
Rank 2 3 1
(table 4.13:-Response Table for Means)

Level Speed Feed Doc

1 1.486 2.170 1.410
2 1.610 1.514 1.568
3 2.223 1.635 2.342
Delta 0.737 0.656 0.932
Rank 2 3 1

Main Effects Plot for Means

Data Means
Speed feed

2.2
2.0
1.8
Mean of Means

1.6
1.4
40 66 92 0.05 0.10 0.15
doc

2.2
2.0
1.8
1.6
1.4
0.25 0.50 0.75

Fig. 17 Main Effects Plot for Means

47
 Main Effects Plot for SN ratios
Data Means
Speed feed
-3

-4

-5

Mean of SN ratios
-6

-7
40 66 92 0.05 0.10 0.15
doc
-3

-4

-5

-6

-7
0.25 0.50 0.75
Signal-to-noise: Smaller is better

Fig. 18 Main Effects Plot for SN ratios

Normal Probability Plot

(response is SN ratios)
99

95
90

80
70
Percent

60
50
40
30
20

10

1
-3 -2 -1 0 1 2 3
Standardized Residual

Fig. 19 Normal plot of Residuals for SN ratios

48
 Versus Fits
(response is SN ratios)
1.5

1.0

Standardized Residual
0.5

0.0

-0.5

-1.0

-10 -9 -8 -7 -6 -5 -4 -3 -2 -1
Fitted Value

Fig. 20 Residuals vs Fits for SN ratios

Histogram
(response is SN ratios)

3.0

2.5

2.0
Frequency

1.5

1.0

0.5

0.0
-1.0 -0.5 0.0 0.5 1.0 1.5
Standardized Residual

Fig. 21 Residual Histogram for SN ratios

49
 Versus Order
(response is SN ratios)
1.5

1.0

Standardized Residual
0.5

0.0

-0.5

-1.0

1 2 3 4 5 6 7 8 9
Observation Order

Fig. 22 Residuals vs Order for SN ratios

Normal Probability Plot

(response is Means)
99

95
90

80
70
Percent

60
50
40
30
20

10

1
-3 -2 -1 0 1 2 3
Standardized Residual

Fig. 23 Normal plot of Residuals for Means

50
 Versus Fits
(response is Means)
1.5

1.0

Standardized Residual
0.5

0.0

-0.5

-1.0

1.0 1.5 2.0 2.5 3.0 3.5

Fitted Value

Fig. 24 Residuals vs Fits for Means

Histogram
(response is Means)

3.0

2.5

2.0
Frequency

1.5

1.0

0.5

0.0
-1.0 -0.5 0.0 0.5 1.0 1.5
Standardized Residual

Fig. 25 Residual Histogram for Means

51
 Versus Order
(response is Means)
1.5

1.0

Standardized Residual
0.5

0.0

-0.5

-1.0

1 2 3 4 5 6 7 8 9
Observation Order

Fig. 26 Residuals vs Order for Means

52
CHAPTER 6

53
 Chapter 6:

Summery

6. Conclusion

 Conclusion can be derived from the experimentation done using S. S. 304 graded steel and

carbide cutting tool.

 A set of levels of parameter is obtained in order to minimize surface roughness as well as

tool wear.

 It is found that cutting velocity affects more while calculating tool wear and where depth of

cut affects more while experimentation of surface roughness.

 A conformation test is done in order to get optimal setting, it is evidenced that A2 B3 C2 for

measuring tool wear and it is found to be 0.659 micron and A1 B2 C1 for surface roughness,

it is found to be 1.253 micron.

54
CHAPTER 7

55
 Chapter 7:

References
1. Zhou Q., Hong G. S. and Rahman M., (1995), “A New Tool Life Criterion For Tool

Condition Monitoring Using a Neural Network”, Engineering Application Artificial

Intelligence, Volume 8, Number 5, pp. 579-588.

2. . Lin W. S., Lee B. Y., Wu C. L., (2001), “Modeling the surface roughness and cutting

force for turning”, Journal of Materials Processing Technology, Volume 108, pp. 286-

293.

3. Feng C. X. (Jack) and Wang X., (2002), “Development of Empirical Models for

Surface Roughness Prediction in Finish Turning”, International Journal of Advanced

Manufacturing Technology, Volume 20, pp. 348–356

4. Suresh P. V. S., Rao P. V. and Deshmukh S. G., (2002), “A genetic algorithmic

approach for optimization of surface roughness prediction model”, International

Journal of Machine Tools and Manufacture, Volume 42, pp. 675–680.

5. Lee S. S. and Chen J. C., (2003), “Online surface roughness recognition system using

artificial neural networks system in turning operations” International Journal of

Advanced Manufacturing Technology, Volume 22, pp. 498–509.

6. Choudhury S. K. and Bartarya G., (2003), “Role of temperature and surface finish in

predicting tool wear using neural network and design of experiments”, International

Journal of Machine Tools and Manufacture, Volume 43, pp. 747–753.

56
7. Chien W.-T. and Tsai C.-S., (2003), “The investigation on the prediction of tool wear

and the determination of optimum cutting conditions in machining 17-4PH stainless

steel”, Journal of Materials Processing Technology, Volume 140, pp. 340–345.

8. Kirby E. D., Zhang Z. and Chen J. C., (2004), “Development of An Accelerometer

based surface roughness Prediction System in Turning Operation Using Multiple

Regression Techniques”, Journal of Industrial Technology, Volume 20, Number 4, pp.

1-8.

9. Özel T. and Karpat Y., (2005), “Predictive modeling of surface roughness and tool

wear in hard turning using regression and neural networks”, International Journal of

Machine Tools and Manufacture, Volume 45, pp. 467–479

10. Antony J., (2000), “Multi-response optimization in industrial experiments using

Taguchi’s quality loss function and Principal Component Analysis”, Quality and

Reliability Engineering International, Volume 16, pp.3-8

11. Kohli A. and Dixit U. S., (2005),”A neural-network-based methodology for the

prediction of surface roughness in a turning process”, International Journal of

Advanced Manufacturing Technology, Volume 25, pp.118–129.

12. Pal S. K. and Chakraborty D., (2005), “Surface roughness prediction in turning using

artificial neural network”, Neural Computing and Application, Volume14, pp. 319–

324.

13. Singh H. and Kumar P., (2006), “Optimizing Feed Force for Turned Parts through the

Taguchi Technique”, Sadhana, Volume 31, Number 6, pp. 671–681.

57
14. Ahmed S. G., (2006), “Development of a Prediction Model for Surface Roughness in

Finish Turning of Aluminium”, Sudan Engineering Society Journal, Volume 52,

Number 45, pp. 1-5.

15. Abburi N. R. and Dixit U. S., (2006), “A knowledge-based system for the prediction

of surface roughness in turning process” Robotics and Computer-Integrated

Manufacturing, Volume 22, pp. 363–372.

16. Zhong Z. W., Khoo L. P. and Han S. T., (2006), “Prediction of surface roughness of

turned surfaces using neural networks”, International Journal of Advance

Manufacturing Technology, Volume 28, pp. 688–693.

17. Kumanan S., Saheb S. K. N. and Jesuthanam C. P., (2006), “Prediction of Machining

Forces using Neural Networks Trained by a Genetic Algorithm”, Institution of

Engineers (India) Journal, Volume 87, pp. 11-15.

18. Mahmoud E. A. E. and Abdelkarim H. A., (2006), “Optimum Cutting Parameters in

Turning Operations using HSS Cutting Tool with 450 Approach Angle”, Sudan

Engineering Scoeiety Journal, Volume 53, Number 48, pp. 25-30.

19. Doniavi A., Eskanderzade M. and Tahmsebian M., (2007), “Empirical Modeling of

Surface Roughness in Turning Process of 1060 steel using Factorial Design

Methodology”, Journal of Applied Sciences, Volume 7, Number17, pp. 2509-2513.

20. Kassab S. Y. and Khoshnaw Y. K., (2007), “The Effect of Cutting Tool Vibration on

Surface Roughness of Work piece in Dry Turning Operation”, Engineering and

Technology, Volume 25, Number 7, pp. 879-889.

21. Al-Ahmari A. M. A., (2007),”Predictive machinability models for a selected hard

material in turning operations”, Journal of Materials Processing Technology, Volume

58
 190, pp. 305–311.

22. Thamizhmanii S., Saparudin S. and Hasan S., (2007), “Analysis of Surface Roughness

by Using Taguchi Method”, Achievements in Materials and Manufacturing

Engineering, Volume 20, Issue 1-2, pp. 503-505.

23. Natarajan U., Arun P., Periasamy V. M., (2007), “On-line Tool Wear Monitoring in

Turning by Hidden Markov Model (HMM)” Institution of Engineers (India) Journal

(PR), Volume 87, pp. 31-35.

24. Özel T. and Karpat Y., (2005), “Predictive modeling of surface roughness and tool

wear in hard turning using regression and neural networks”, International Journal of

Machine Tools and Manufacture, Volume 45, pp. 467–479.

25. Wang M. Y. and Lan T. S., (2008), “Parametric Optimization on Multi-Objective

Precision Turning Using Grey Relational Analysis”. Information Technology Journal,

Volume 7, pp.1072-1076.

26. Srikanth T. and Kamala V., (2008), “A Real Coded Genetic Algorithm for

Optimization of Cutting Parameters in Turning IJCSNS”, International Journal of

Computer Science and Network Security, Volume 8 Number 6, pp. 189-193.

27. Sahoo P., Barman T. K. and Routara B. C., (2008), “Taguchi based practical

dimension modeling and optimization in CNC turning”, Advance in Production

59
 Engineering and Management, Volume 3, Number 4, pp. 205-217

28. . Reddy B. S., Padmanabhan G. and Reddy K. V. K., (2008), “Surface Roughness

Prediction Techniques for CNC turning”, Asian Journal of Scientific Research,

Volume 1, Number 3, pp. 256-264.

29. Wannas A. A., (2008), “RBFNN Model for Prediction Recognition of Tool Wear in

Hard Turning”, Journal of Engineering and Applies Science, Volume 3, Number 10,

pp. 780-785.

30. . Lin W. S., Lee B. Y., Wu C. L., (2001), “Modeling the surface roughness and cutting

force for turning”, Journal of Materials Processing Technology, Volume 108, pp. 286-

293.

Thamma R., (2008), “Comparison between Multiple Regression Models to Study

Effect of Turning Parameters on the Surface Roughness”, Proceedings of the 2008

96IAJC-IJME International Conference, ISBN 978-1-60643-379-9, Paper 133, ENG

103 pp. 1-12.

31. Fnides B., Aouici H., Yallese M. A., (2008), “Cutting forces and surface roughness in

hard turning of hot work steel X38CrMoV5-1 using mixed ceramic”, Mechanika,

Volume 2, Number 70, pp. 73-78.

32. Biswas C. K., Chawla B. S., Das N. S., Srinivas E. R. K. N. K., (2008), “Tool Wear

Prediction using Neuro-Fuzzy System”, Institution of Engineers (India) Journal (PR),

Volume 89, pp. 42-46.

60
33. Fu P. and Hope A. D., (2008), “A Hybrid Pattern Recognition Architecture for Cutting

Tool Condition Monitoring” Technology and Applications, Volume 24, Number 4, pp.

548-558.

34. Feng C. X. (Jack) and Wang X., (2002), “Development of Empirical Models for

Surface Roughness Prediction in Finish Turning”, International Journal of Advanced

Manufacturing Technology, Volume 20, pp. 348–356.

35. Amitav Bhattacharyya (Chapter-Thermodynamics of Chip formation)

36. Dogra D. M., "Effect of tool geometry variation on finish turning: A Review," Journal

of Engineering Science and Technology Review, 2010

37. Danilevskii V. V., "Geometry of a Cutting Tool".

38. El-Hofy H.,

Available:"http://www.crcnetbase.com/doi/abs/10.1201/9781420043402.ch10,"

[Online].

39. Ron Amaral L. H. C., "surface roughness," December 2, 2002.

40. Basic Definitions and Cutting Tool Geometry, Single Point Cutting Tools, Winston

Churchill's message to President Roosevelt in a radio broadcast on 9 February 1941.

61