CHAPTER 1

Surface roughness, often shortened to roughness, is a measure of the texture of a surface. It is quantified by the vertical deviations of a real surface from its ideal form. If these deviations are large, the surface is rough; if they are small the surface is smooth. Roughness is typically considered to be the high frequency, short wavelength component of a measured

1 INTRODUCTION 1.1 Roughness

Roughness consists of surface irregularities which result from the various machining process. These irregularities combine to form surface texture. Roughness height is the height of the irregularities with respect to a reference line. Roughness width is the distance parallel to the nominal surface between successive peaks.

surface. Roughness plays an important role in determining how a real object will interact with its environment. Rough surfaces usually wear more quickly and have higher friction coefficients than smooth surfaces. Roughness is often a good predictor of the performance of a mechanical component, since irregularities in the surface may form nucleation sites for cracks or corrosion. On the other hand, roughness may promote adhesion. Although roughness is often undesirable, it is difficult and expensive to control in

1.2 Surface roughness

manufacturing. Decreasing the roughness of a surface will usually increase exponentially its manufacturing costs. This often results in a trade-off between the manufacturing cost of a component and its performance in application. A roughness value can either be calculated on a profile or on a surface. The profile roughness parameter (Ra, Rq,...) are more common. The area roughness parameters (Sa, Sq,...) give more significant values.

1.2.1 Profile roughness parameters

Each of the roughness parameters is calculated using a formula for describing the surface. There are many different roughness parameters in use, but is by far the most common. Other common parameters include , , and . Some parameters are used only in certain industries or within certain countries. Since these parameters reduce all of the information in a profile to a single number,

Fig 1.1 roughness and waviness profiles.

great care must be taken in applying and interpreting them. Small changes in how the raw profile data is filtered, how the mean line is calculated, and the physics of the measurement can greatly affect the calculated parameter.

By convention every 2D roughness parameter is a capital R followed by additional characters in the subscript. The subscript identifies the formula that was used, and the R means that the formula was applied to a 2D roughness profile. Different capital letters imply that the formula was applied to a different profile. For example, Ra is the arithmetic average of the roughness profile, Pa is the arithmetic average of the unfiltered raw profile, and Sa is the arithmetic average of the 3D roughness.
Parameter Description Formula

Table 1.1 Surface Roughness description and formula

Ra, Raa, Ryni 1.2.2 Amplitude parameters.

Amplitude parameters characterize the surface based on the vertical deviations of the roughness profile from the mean line. Many of them are closely related to the parameters found in statistics for characterizing population samples. For example, Ra is the arithmetic average of the absolute values and Rt is the range of the collected roughness data points.

arithmetic average of absolute values

[1]

Rq, RRMS

root mean squared

[1]

Rv
The average roughness, Ra, is expressed in units of height. In the Imperial (English) system, 1 Ra is typically expressed in "millionths" of an inch. This is also referred to as "microinches" or sometimes just as "micro" (however the latter is just slang). The amplitude parameters are by far the most common surface roughness parameters found in the United States on mechanical engineering drawings and in technical literature. Part of the reason for their popularity is that they are straightforward to calculate using a computer.

maximum valley depth

Rp

maximum peak height

Rt

Maximum Height of the Profile

Rsk

skewness

Rku

Kurtosis

1.3 Surface roughness in Manufacturing. 1.2.3 Slope, spacing, and counting parameters.
Slope parameters describe characteristics of the slope of the roughness profile. Spacing and counting parameters describe how often the profile crosses certain thresholds. These parameters are often used to describe repetitive roughness profiles, such as those produced by turning on a lathe. Just as different manufacturing processes produce parts at various tolerances, they are Table 1.2 slope of surface profile. also capable of different roughnesses. Generally these two characteristics are linked: manufacturing processes that are dimensionally precise create surfaces with low roughness. Many factors contribute to the surface finish in manufacturing. In forming processes, such as molding or metal forming, surface finish of the die determines the surface finish of the work piece. In machining the interaction of the cutting edges and the microstructure of the material being cut both contribute to the final surface finish. In general, the cost of manufacturing a surface increases as the surface finish improves.

Parameter Description

Formula

In other words, if a process can manufacture parts to a narrow dimensional tolerance, the parts will not be very rough.

the RMS Rdq, R of

q

slope profile the

Due to the abstractness of surface finish parameters, engineers usually use a tool that has a variety of surface roughnesss created using different manufacturing methods.

the

within

sampling length

the absolute Rda, R

a

average slope profile the

of

the

within

sampling length

where delta i is i calculated according ASME B46.1 to

Table 1.3 surface finesh in manufacturing

1.3 SURFACE ROUGHNESS TESTER

Model SJ-201PR is comprised of two components: detachable detector/drive and display unit which can house detector drive. Unit provides functions to communicate Statistical Process Control (SPC) data with external devices via an RS-232C interface. Product is compatible with ISO, DIN, ANSI and JIS standards and supports optional printer for retrieving hard copies of profiles and parameters. When it comes to surface roughness testing, at the process is where to measure. "The Mitutoyo SJ-201PR has the added feature of gliding softly to the measuring surface, with optional nosepiece or user specified fixture the SJ-201PR proves to be almost indestructible on the shop floor," said a company spokesperson. "Virtual elimination of detector repairs saves hundreds of dollars and headaches in the first year alone." To operate the SJ-201PR, the display unit is placed on the surface to be measured; pressing the start/stop key initiates detector traverse. The detector, which is positioned 3mm above the surface, glides to the part and commences the measurement. The resulting parameter values are then displayed on a large LCD screen. The SJ-201PR comprises two components: a fully detachable detector/drive (about the size of a pocket knife), and a display unit (about the size of a cordless phone) that can house the detector drive. The SJ-201PR is able to measure 19 surface roughness parameters: Ra, Ry, Rz, Rq, S, Sm, Pc, R3z, mr, Rt, Rp, Rk, Rpk, Rvk, Mr1, Mr2, A1, A2, Vo. It is compatible with ISO, DIN, ANSI and JIS standards. An optional printer can be added for a hard copy of both profile and parameters. SurfPak software is also available. The SJ-201PR provides functions to communicate Statistical Process Control (SPC) data with external devices including PCs and printers via an RS-232C interface.

Fig 1.1 CNC Machine.

Fig 1. Photographic view of the surface roughness tester.

1.4 Surface roughness in CNC machining process.

Nowadays, due to the increasing demand of higher precision components for its functional aspect, surface roughness of a machined part plays an important role in the modern manufacturing process. Turning is a machining operation, which is carried out on lathe. The quality of the surface plays a very important role in the performance of turning as a good quality turned surface significantly improves fatigue strength, corrosion resistance, or creep life. Surface roughness also affects several functional attributes of parts, such as, contact causing surface friction, wearing, light reflection, heat transmission, ability of distributing and holding a lubricant, load bearing capacity, coating, or resisting fatigue. To achieve the desired surface finish, a good predictive model is required for stable machining. Generally, these models have a complex relationship between surface roughness and operational parameters, work materials, and chip breaker types

1.5 Programmable logic Controller:A PLC matches the NC to the machine. PLCs were basically as replacement for hard wired relay control panels. They were basically introduced as replacement for hard wired relay panels. They developed to be re-programmed without hardware changes when requirements were altered and thus are re-usable. PLCs are now available with increased functions, more memory and larger input/output capabilities. In the CPU; all the decisions are made relative to controlling a machine or a process. The CPU receives input data, performs logical decisions based upon stored programs and drives the output s. connection to a computer for hierarchical control are done via the CPU

10

Tool number (T). Tool offset function (D). Miscellaneous functions (M,H, etc. ) End of block (EOB / *)

1.6 PROCEDURE OF PART PROGRAMMING.

Study the relevant component drawing thoroughly. Identify the type of material to be machined. Determine the specifications and functions of machine to be used. Decide the dimension and mode i.e. Metric or inch. Decide the co-ordinate system i.e. Absolute or incremental Identify the plane of cutting. Determine the cutting parameters for the job / tool combination. Decide the feed rate programming. Check the tooling required. Establish the sequence of machining operations. Identify whether use of special features like subroutines, mirror imaging etc. is required or not. Decide the mode of storing the part program once it is completed.

1.8 CNC CUTTING PARAMETERS

Different parameters in CNC Machining to obtain surface finish. 1.8.1 Depth of cut,

The Thickness of material removed by one pass of cutting tool. Cutting speed and feed rate come together with depth of cut to determine the material removal rate, which is the volume of work piece material that can be removed per unit time.

1.8.2

Cutting speed,

The rate at which the cutting edge of the tool moves past the work piece surface at the

1.7 STRUCTURE OF A PART PROGRAMMING

Part program defines a sequence of NC machining operations. The information contained in the program can be dimensional or non-dimensional like speed, feed, auxiliary functions, etc... The basic unit of a part program input to the control is called a block. Each block contains adequate information for the machine to perform a movement and or functions. Block in turn is made up of words and each word consists of a number of characters. All blocks are terminated by the block end character. The maximum block length for each CNC is fixed. A block may contain any or all the following: Optional block skip (/). Sequence or block number (N). Preparatory functions (G). Dimensional information (X, Y,Z etc. ) Decimal point (.). Feed rate (F). Spindle speed (S).
11

point of contact. Cutting speed (also called surface speed or simply speed) may be defined as the rate (or speed) that the material moves past the cutting edge of the tool, irrespective of the machining operation used. A cutting speed for mild steel, of 100 ft/min (or approx. 30 meters/min) is the same whether it is the speed of the (stationary) cutter passing over the (moving) workpiece, such as in a turning operation, or the speed of the (rotating) cutter moving past a (stationary) workpiece, such as in a milling operation. What will affect the value of this surface speed for mild steel are the cutting conditions: For a given material there will be an optimum cutting speed for a certain set of machining conditions, and from this speed the spindle speed (RPM) can be calculated. Factors affecting the calculation of cutting speed are: The material being machined (steel, brass, tool steel, plastic, wood) (see table below) The material the cutter is made from (Carbon steel, high speed

steel (HSS), carbide, ceramics)

12

The economical life of the cutter (the cost to regrind or purchase new, compared to the quantity of parts produced) Cutting speeds for various materials using a plain high speed steel cutter Meters (MPM) 1518 3038 1824 20-37 21-40 & per min Surface feet per min (SFM) 5060 100125 6080 65120 70130

Cutting speeds are calculated on the assumption that optimum cutting conditions exist, these include: Metal removal rate (finishing cuts that remove a small amount of material may be run at increased speeds)

Material type

Steel (tough) Mild steel

Full and constant flow of cutting fluid (adequate cooling and chip flushing) Rigidity of the machine and tooling setup (reduction in vibration or chatter) Continuity of cut (as compared to an interrupted cut , such as machining square section material in a lathe) Carbon steels (C1008-C1095) Cast iron (medium) Alloy steels (13209262)

Condition of material (mill scale, hard spots due to white cast iron forming in castings) Free cutting steels (B1111-B1113 C1108-C1213) 35-69 115225

The cutting speed is given as a set of constants that are available from the material manufacturer or supplier, the most common materials are available in reference books, or charts but will always be subject to adjustment depending on the cutting conditions. The following table gives the cutting speeds for a selection of common materials under one set of conditions. The conditions are a tool life of 1 hour, dry cutting (no coolant) and at medium feeds so they may appear to be incorrect depending on circumstances. These cutting speeds may change if, for instance, adequate coolant is available or an improved grade of HSS is used (such as one that includes cobalt).
Aluminium 75105 250350 Leaded steel (Leadloy 12L14) 91 300

Stainless steels (300 & 400 series) Bronzes

23-40 2445

75130 80150

Brass

90-210

300-700 (Max. spindle speed)

.
13 14

1.8.3

Feed rate ,

Formula to determine feed rate This formula can be used to figure out the feed rate that the cutter travels into or around the work. This would apply to cutters on a milling machine, drill press and a number of other machine tools. This is not to be used on the lathe for turning operations, as the feed rate on a lathe is given as feed per revolution.

The rate that the cutting tool travels along the surface of the work piece. Feed rate is the velocity at which the cutter is fed, that is, advanced against the workpiece. It is expressed in units of distance per revolution for turning and boring (typically inches per revolution [ipr] or millimetres per revolution). It can be expressed thus for milling also, but it is often expressed in units of distance per time for milling (typically inches per minute [ipm] or millimetres per minute), with considerations of how many teeth (or flutes) the cutter has then determining what that means for each tooth. Feed rate is dependent on the: Type of tool (a small drill or a large drill, high speed or carbide, a boxtool or recess, a thin form tool or wide form tool, a slide knurl or a turret straddle knurl). Surface finish desired. Power available at the spindle (to prevent stalling of the cutter or workpiece). Rigidity of the machine and tooling setup (ability to withstand vibration or chatter). Strength of the workpiece (high feed rates will collapse thin wall tubing) Where:

FR = the calculated feed rate in inches per minute or mm per minute. RPM = is the calculated speed for the cutter. T = Number of teeth on the cutter. CL = The chip load or feed per tooth. This is the size of chip that each tooth of the cutter takes.

1.8.4 Other CNC Cutting Parameters

1. Tool nose radius, 2. Tool overhang, 3. Approach angle, 4. Work piece length, 5. Work piece diameter, 6. Work piece material.

Characteristics of the material being cut, chip flow depends on material type and feed rate. The ideal chip shape is small and breaks free early, carrying heat away from the tool and work. Threads per inch (TPI) for taps, die heads and threading tools. When deciding what feed rate to use for a certain cutting operation, the calculation is fairly straightforward for single-point cutting tools, because all of the cutting work is done at one point (done by "one tooth", as it were). With a milling machine or jointer, where multitipped/multi-fluted cutting tools are involved, then the desirable feed rate becomes dependent on the number of teeth on the cutter, as well as the desired amount of material per tooth to cut (expressed as chip load). The greater the number of cutting edges, the higher the feed rate permissible: for a cutting edge to work efficiently it must remove sufficient material to cut rather than rub; it also must do its fair share of work. The ratio of the spindle speed and the feed rate controls how aggressive the cut is, and the nature of the swarf formed.

15

16

CHAPTER 2 LITERATURE SURVEY

To study the Analysis of Surface Roughness on Different Materials in CNC Turning Operation. This literature study guides my project to undergo in same procedure and principles. Ahilana.C, et al (2012),Decision-making process in manufacturing environment is increasingly difficult due to the rapid changes in design and demand of quality products. To make decision making process (selection of machining parameters) online, effective and efficient artificial intelligent tools like neural networks are being attempted. This paper proposes the development of neural network models for prediction of machining parameters in CNC turning process. Experiments are designed based on Taguchis Design of Experiments (DoE) and conducted with cutting speed, feed rate, depth of cut and nose radius as the process parameters and surface roughness and power consumption as objectives. Developed models are validated and reported. Signal-to-noise (S/N) ratios of responses are calculated to identify the influences of process parameters using analysis of variance (ANOVA) analysis. Results obtained in this work are intended for use by numerical control or manually operated machines. AmanAggarwal, et al (2008), this paper optimizes multiple characteristics (tool life, cutting force, surface roughness and power consumption) in CNC turning of AISI P-20 tool steel using liquid nitrogen as a coolant. Four controllable factors of the turning process viz. cutting speed, feed, depth of cut and nose radius, were studied. Face centred central composite design was used for experimentation. Response surface methodology was used for modelling the responses. Desirability function was used for single and multiple response optimization. Models developed were adequate in explaining the effect of independent parameters on responses.3D plots for overall desirability function revealed the desirability range when responses are given equal weight age. As clear from the plots low level of cutting speed, feed and depth of cut are desirable for getting high value of desirability. Likewise high value of nose radius is also desirable forgetting high value of desirability. Confirmation experiments were done as output in Table and the % variation between actual experimental data and predicted data was in between 6.25 and 2.6% which validates the results drawn from desirability plot.
17

AmanAggarwal, et al(2008). An experimental investigation into the effects of cutting speed, feed rate, depth of cut, nose radius and cutting environment in CNC turning of AISIP20 tool steel.response surface methodology (RSM) and Taguchis technique, have been used to accomplish the objective of the experimental study. L27 orthogonal array and face centred central composite design have been used for conducting the experiments. The effects of feed rate and nose radius were found to be insignificant compared to other factors. Time required for conducting experiments using RSM technique was almost twice as that needed through Taguchi technique. It is attributed to the fact that 180 (302, 302, 302) were performed using face centred central composite design for three (dry, wet and cryogenic) cutting environments whereas only 81 (273) experiments were performed using L27 orthogonal array. Also ramp function graphs tell the exact level of parameters for desired level of response. Thus RSM can better predict the effect of parameters on response and is a better tool for optimization. Anil Gupta, et al(2011), This paper presents the application of Taguchi method with logical fuzzy reasoning for multiple output optimization of high speed CNC turning of AISI P-20 tool steel using TiN coated tungsten carbide coatings. The machining parameters (cutting speed, feed rate, depth of cut, nose radius and cutting environment) are optimized with considerations of the multiple performance measures (surface roughness, tool life, cutting force and power consumption).Taguchis concepts of orthogonal arrays, signal to noise (S/N) ratio, ANOVA have been fuzzified to optimize the high speed CNC turning process parameters through a single comprehensive output measure (COM).The result analysis shows that cutting speed of 160 m/min, nose radius of 0.8 mm, feed of 0.1 mm/rev, depth of cut of 0.2 mm and the cryogenic environment are the most favourable cutting parameters for high speed CNC turning of AISI P-20 tool steel. In the multi-response problem, all the four responses tool life, power consumption, cutting force and surface roughness were simultaneously considered. It can be concluded that middle level of cutting speed(160 m/min) and nose radius (0.8 mm) and lower level of feed (0.1 mm/rev) and depth of cut (0.2 mm) yield the optimal result. Both single response and multi-response optimization analysis proved that cryogenic machining environment E3 is favourable in increasing tool life and reducing surface roughness, cutting force and power consumption compared to wet and dry machining.

18

Daniel Kirby.E, et al (2007), this paper discusses the development of a surface roughness prediction system for a turning operation, using a fuzzy nets modelling technique. The goal is to develop and train a fuzzy-nets-based surface roughness prediction (FN-SRP) system that will predict the surface roughness of a turned work piece using accelerometer measurements of turning parameters and vibration data. The FN-SRP system has been developed using a computer numerical control (CNC) slant-bed lathe with a carbide cutting tool. The system was trained using feed rate, spindle speed, and tangential vibration data collected during experimental runs. A series of validation runs indicate that this system has a mean accuracy of 95%.Providing machine tools with the capability to monitor quality characteristics such as surface roughness is an essential component for the ability to create reliable unmanned machining cells. This study explored the use of vibrations measured at the cutting tool in a turning operation to predict surface roughness. A viable FN prediction system was developed, using feed rate, spindle speed, and vibrations in the Y-axis, which yielded excellent results. Test data was used to validate the model, which yielded an average error rate of 95%. Future research into varying the FN techniques, applying these techniques to different experimental setups, and exploring ways to increase prediction accuracy Exploration of the effect of various conditions such as different materials, material variations, coolant, internal and external vibration sources, and tool wear Development of a true in-process system, to determine the capability to predict surface roughness deviations under various conditions. Development of an adaptive control system which utilizes this predictive capability. DurmusKarayel (2009). In this study, a neural network approach is presented for the prediction and control of surface roughness in a computer numerically controlled (CNC) lathe. Experiments have been performed on the CNC lathe to obtain the data used for the training and testing of a neural network. The parameters used in the experiment were reduced to three cutting parameters which consisted of depth of cutting, cutting speed, and feed rate. Each of the other parameters such as tool nose radius, tool overhang, approach angle, workpiece length, work piece diameter and workpiece material was taken as constant. The number of iterations was 8000 and no smoothing factor was used. Ra, Rz and Rmax were modelled and were evaluated individually. One hidden layer was used for all models while the numbers of neurons in the hidden layer of the Ra model were five and the numbers of
19

neurons in the hidden layers of the Rz and Rmax models were ten. A new surface roughness value was determined by sending the cutting parameters to the observer (ANN block).The feed rate is a dominant parameter and the surface roughness increases rapidly with the increase in feed rate. If the developed control algorithm is used, cutting parameters corresponding to any surface roughness can be obtained to produce the desired surface roughness. If the research is repeated with different parameters for other machine tools, it can be generalized and so can be applied to other machining types. IlhanAsiltrk and HarunAkkus (2011), this study focuses on optimizing turning parameters based on the Taguchi method to minimize surface roughness (Ra and Rz). Experiments have been conducted using the L9 orthogonal array in a CNC turning machine. Dry turning tests are carried out on hardened AISI 4140 (51 HRC) with coated carbide cutting tools. Each experiment is repeated three times and each test uses a new cutting insert to ensure accurate readings of the surface roughness. The statistical methods of signal to noise ratio (SNR) and the analysis of variance (ANOVA) are applied to investigate effects of cutting speed, feed rate and depth of cut on surface roughness. As a result, nine experiments were conducted instead of the full factorial 27 experiments. Ra and Rz S/N ratios were found as a result of experiments conducted according to the L9 orthogonal array. Joseph Davidson. M, et al(2008), Design of experiments has been used to study the effects of the main flow-forming parameters such as the speed of the mandrel, the longitudinal feed, and the amount of coolant used on the surface roughness of flow-formed AA6061 tube. A mathematical prediction model of the surface roughness has been developed in terms of the above parameters. The effect of these parameters on the surface roughness has been investigated using response surface methodology (RSM). Response surface contours were constructed for determining the optimum forming conditions for a required surface roughness. The surface roughness was found to increase with increase in the longitudinal feed and it decreased with decrease in the amount of the coolant used. The verification experiment carried out to check the validity of the developed model predicted surface roughness within 6% error. RSM has been used to determine the surface roughness attained by the flow-formed tubes for various input parameters namely the feed, speed and the coolant. A RSM model can successfully relate the above process parameters with the response, surface roughness. The verifying experiment has shown that the predicted value agrees with the experimental evidence.
20

model is further interfaced with the GA to optimize the cutting conditions for desired surface roughness. The GA reduces the surface roughness value in the mold cavity from 0.412m to Muammer Nalbant, et al(2009), In this study the machining of AISI1030 steel (i.e. orthogonal cutting) uncoated, PVD- and CVD- coated cemented carbide insert with different feed rates of 0.25, 0.30, 0.35, 0.40 and 0.45mm/rev with the cutting speeds of 100, 200 and 300m/min by keeping depth of cuts constant (i.e.2mm), without using cooling liquids has been accomplished. The surface roughness effects of coating method, coating material, cutting speed and feed rate on the work piece have been investigated. Among the cutting tools with 200mm/min cutting speed and 0.25mm/rev feed rate the TiN coated with PVD method has provided 2.16 mm. While the uncoated cutting tool with the cutting speed of 100m/min and 0.25mm/rev feed rate has yielded the surface roughness value of 2.45 mm. The training and test data of the ANNs have been prepared using experimental patterns for the surface roughness. Decreasing the friction of coefficient and thermal conductivity of the cutting tool decrease the average surface roughness of the machining material. There is a positive linear relationship between the average surface roughness of uncoated and coated cemented carbide cutting tools and the feed rate used at cutting operation. The bestaveragesurfaceroughnessvaluesat200m/min cutting speed with the feed rate of 0.25mm/rev are the following: TiN - coated cutting tool with CVD method 2.16 mm, TiAlN- coated cutting tool with PVD method2.3 mm, AlTiN coated cutting tool with PVD method 2.46 mm, In the case of uncoated cutting tool, at the cutting speed of100m/min and feed rate of 0.25mm/rev the result is2.45 mm. Oktem. H, et al(2005), This paper focuses on the development of an effective methodology to determine the optimum cutting conditions leading to minimum surface roughness in milling of mold surfaces by coupling response surface methodology (RSM) with a developed genetic algorithm (GA).RSM is utilized to create an efficient analytical model for surface roughness in terms of cutting parameters: feed, cutting speed, axial depth of cut, radial depth of cut and machining tolerance. For this purpose, a number of machining experiments based on statistical three-level full factorial design of experiments method are carried out in order to collect surface roughness values. An effective fourth order response surface (RS) model is developed utilizing experimental measurements in the mold cavity. RS
21

0.375mcorresponding to about 10% improvement. Optimum cutting condition produced from GA is verified with the experimental measurement. In this study, a fourth order RS model for predicting surface roughness values in milling mold surfaces made of Aluminium (7075-T6) material was developed. Surface roughness of the mold surfaces, which was 0.412 _m before optimization, was reduced to 0.375 _m after optimization. GA improved the surface roughness by about 10%.Difference was found to be less than 1.4%. This indicates that the optimization methodology proposed in this study by coupling the developed RS model and the developed GA is effective and can be utilized in other machining problems such as tool life, dimensional errors, etc.as well. Ramesh S, et al (2012), the effect of cutting parameters on the surface roughness in turning of titanium alloy has been investigated using response surface methodology. The experimental studies were conducted under varying cutting speeds, feed and depths of cut. The chip formation and SEM analysis are discussed to enhance the supportive surface quality achieved in turning. The work material used for the present investigation is commercial aerospace titanium alloy (gr5) and the tool used is RCMT 10T300 MT TT3500 round insert. The equation developed using response surface methodology is used for predicting the surface roughness in machining of titanium alloy. Using the Design of Experiments concept, the experiments was designed using Taguchis orthogonal array principle titanium alloy was machined using a RCMT 10T300 MT TT3500 round insert in dry machining condition.Taguchi ANOVA analysis was performed. The most influencing parameter was identified as the feed. The order of importance was feed, followed by depth of cut and cutting speed.The surface damage owing to the interaction tool/work observed through SEM analysis.

Satheesh Kumar.N, et al (2012), the effect of process parameters in turning of Carbon Alloy Steels in a CNC lathe. The parameters namely the spindle speed and feed rate are varied to study their effect on surface roughness. The five different carbon alloy steels used for turning are SAE8620, EN8, EN19, EN24 and EN47.The study reveals that the surface roughness is directly influenced by the spindle speed and feed rate. It is observed that the surface roughness increases with increased feed rate and is higher at lower speeds and vice
22

versa for all feed rates. The better surface finish may be achieved by turning carbon alloy steels at low feed rate and high spindle speeds. It should also be noted that the turning operation for all work pieces carried out sequentially. SuleymanNeseli, et al (2011), This investigation focuses on the influence of tool geometry on the surface finish obtained in turning of AISI 1040 steel. In order to find out the effect of tool geometry parameters on the surface roughness during turning, response surface methodology (RSM) was used and a prediction model was developed related to average surface roughness (Ra) using experimental data. The results indicated that the tool nose radius was the dominant factor on the surface roughness. In addition, a good agreement between the predicted and measured surface roughness was observed. Therefore, the developed model can be effectively used to predict the surface roughness on the machining of AISI 1040 steel within 95% confidence intervals ranges of parameters studied. The result of ANOVA proved that the quadratic mathematical models allow prediction of surface roughness parameter with a 96% confident interval. Tool nose radius is the most significant factor on surface roughness with 51.45% contribution in the total variability of model. The quadratic effect of tool nose radius little provides little contribution to the model. Also, approach angle and rake angle are significant factors on surface roughness with 18.24% and17.74% contribution in the total variability of model, respectively. It can be said that the interaction between all factors has no significant effect on surface roughness. Using response optimization show that the optimal combination of machining parameters are (0.4 mm, 60, 3) for tool nose radius, approach angle and rake angle, respectively. Suresh P.V.S, et al (2002), Due to the widespread use of highly automated machine tools in the industry, manufacturing requires reliable models and methods for the prediction of output performance of machining processes. The prediction of optimal machining conditions for good surface finish and dimensional accuracy plays a very important role in process planning. The present work deals with the study and development of a surface roughness prediction model for machining mild steel, using Response Surface Methodology (RSM). The experimentation was carried out with TiN-coated tungsten carbide (CNMG) cutting tools, for machining mild steel work-pieces covering a wide range of machining
23

conditions. This model gives the factor effects of the individual process parameters. An attempt has also been made to optimize the surface roughness prediction model using Genetic Algorithms (GA) to optimize the objective function. The GA program gives minimum and maximum values of surface roughness and their respective optimal machining conditions. The two-stage effort of obtaining a surface roughness model by surface response methodology, and optimization of this model by Genetic Algorithms, has resulted in a fairly useful method of obtaining process parameters in order to attain the required surface quality. This has validated the trends available in the literature, and extended the data range to the present operating conditions, apart from improving the accuracy and modelling by involving the most recent modelling method. VikasUpadhyay, et al (2011), in this work, an attempt has been made to use vibration signals for in-process prediction of surface roughness during turning of Ti6Al4V alloy. The investigation was carried out in two stages. In the first stage, only acceleration amplitude of tool vibrations in axial, radial and tangential directions were used to develop multiple regression models for prediction of surface roughness. The first and second order regression models thus developed were not found accurate enough (maximum percentage error close to 24%). In the second stage, initially a correlation analysis was performed to determine the degree of association of cutting speed, feed rate, and depth of cut and the acceleration amplitude of vibrations in axial, radial, and tangential directions with surface roughness. Subsequently, based on this analysis, feed rate and depth of cut were included as input parameters aside from the acceleration amplitude of vibrations in radial and tangential directions to develop a refined first order multiple regression model for surface roughness prediction. This model provided good prediction accuracy (maximum percentage error 7.45%) of surface roughness. Finally, an artificial neural network model was developed as it can be readily integrated into a computer integrated manufacturing environment. Pearson correlation coefficient was used to determine the correlation between surface roughness and cutting parameters and acceleration amplitude of vibrations. Pearson correlation coefficient for feed rate was maximum followed by acceleration amplitude of vibration in radial direction, depth of cut and acceleration amplitude of vibration in tangential direction. As this model was found accurate enough, neural network model was developed using the same combination of input parameters. To check the adequacy of developed models, the models were validated with the data not used in development of models.

24

CHAPTER 3
3.1 Material Preparation Aluminum alloys may be strengthened by addition of copper, silicon and Tin.

3.2 EXPERIMENTAL DESIGN AND RESULTS

Experimental design involves variation of three factors (cutting speed, feed rate and depth of cut) at three levels as mentioned Table 1.4 Machining parameters and levels

3.1.1 Silicon (Si) Symbol Some of aluminum based bearing alloys contain silicon. Silicon has very high hardness and its inclusions distributed over the aluminum matrix serve as abrasive particles polishing the mating journal surface. F 3.1.2 Tin (Sn) A Aluminum based bearing alloys commonly contain tin as a soft component. Tin is distributed in aluminum matrix as a separate phase in form of a reticular (network) structure along the edges of aluminum grains. Tin imparts to the material anti-friction properties (compatibility, conformability, embedability). 3.1.3 Copper (Cu) The aluminum-copper alloys typically contain 1.2% copper, with smaller additions of other elements. The copper provides substantial increases in strength and facilitates precipitation hardening. The introduction of copper to aluminum can also reduce ductility and corrosion resistance. 3.1.4 Material to be selected. By convention every 2D roughness parameter is a capital R followed by additional o Phosphor-bronze(Cu (85%) , Sn(10-12%), P(0.001%), Ni(1%) with boron carbide (8 micron size) o Aluminum with (Sn(22%), Cu(1.2%)) o Aluminum with (Si(36%), Mg(1.5%) characters in the subscript. The subscript identifies the formula that was used, and the R means that the formula was applied to a 2D roughness profile. Different capital letters imply that the formula was applied to a different profile. For example, Ra is the arithmetic average of the roughness profile, Pa is the arithmetic average of the unfiltered raw profile, and Sa is the arithmetic average of the 3D roughness. Since these parameters reduce all of the information in a profile to a single number, great care must be taken in applying and interpreting them. Small changes in how the raw profile data is filtered, how the mean line is calculated, and the physics of the measurement can greatly affect the calculated parameter. Depth of cut mm 1 1.5 2 Feed rate mm/rev 0.15 0.2 0.25 V Cutting speed rpm 1500 2000 2500 Control factor Unit Level 1 Level 2 Level 3

3.2.1 Surface Roughness Measurement.

Each of the roughness parameters is calculated using a formula for describing the surface. There are many different roughness parameters in use, but is by far the most common. Other common parameters include , , and . Some parameters are used only in certain industries or within certain countries.

25

26

Table 1.5 Experimental results of training patterns and actual surface roughness values Sample number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Cutting speed (v) (rpm) 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 Feed rate (f) (mm/rev) 0.15 0.15 0.15 0.2 0.2 0.2 0.25 0.25 0.25 0.15 0.15 0.15 0.2 0.2 0.2 0.25 0.25 0.25 0.15 0.15 0.15 0.2 0.2 0.2 0.25 0.25 0.25 Depth of cut (d) (mm) 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 Actual roughness (Ra) (m) Artificial neural networks (ANNs) are information processing systems, and since their 3.47 3.87 4.41 4.94 4.72 5.46 5.61 5.72 5.82 1.68 1.97 2.17 2.47 2.97 3.44 3.57 4.73 4.81 0.52 0.68 0.88 1.68 1.79 1.91 2.33 2.36 2.69
[16]

3.3 ARTIFICIAL NEURAL NETWORKS

inception, they have been used in several areas of engineering applications. In experimental studies, some of the operating conditions of a system can be investigated. For this type of experimental work, experts and special equipment are needed. It also requires too much time and high cost.

3.3.1 Development of neural network model In the present work, a feed-forward back-propagation training algorithm is employed for predicting the surface roughness in CNC turning process. Training begins with all weights set to random numbers. For each data record, the predicted value is compared to the desired (actual) value and the weights are adjusted to move the prediction closer to the desired value. Many cycles are made through the entire set of training data with the weights being continually adjusted to produce more accurate predictions.

3.3.2 Execution of experiments The experimental set-up and the tests were performed on a CNC turning center. The training data set was developed through experiments based on L27 Taguchi orthogonal array . The orthogonal array were assigned to cutting speed (v), feed rate (f), and depth of cut

(d), respectively, and accordingly 27 experiments were performed under different combinations of the factor levels. The aluminium and bronze composite metal specimens with dimensions each of diameter 30 mm and length of 150 mm was clamped onto to the turret of the machine table. Surface roughness measurement was done in the off-line with the usage of TIME TR100 surface roughness tester. The radius of the stylus point is 10.02.5 micron and the traversed length is 6 mm. The experimental set-up consists of a CNC machine, battery unit for backup purpose, power supply, and the whole set-up is connected to the computer interface. A computer numeric control (CNC) program was written to perform the turning process. According to the acceptable ranges of cutting speed and feed rate when cutting brass with a carbide insert with a tool holder PCLNR120408 and nose radius of 0.8, and then an NC program was written to execute the cutting operations. Three levels of each

27

28

factor were selected as shown in Table 1. Including test runs, there were totally 6 specimens machined in this experiment. All the specimens in this experiment were machined without coolant. At high cutting speeds, there will be no BUE. Furthermore, in CNC, sculptured tool holders are used, and hence, the inserts are replaced in its position, depending on the wear rate of the tool inserts. In addition, the following assumptions were made (1) Cutting tools used are identical in property. (2) The hardness of each work piece is same throughout the length of the work piece. (3) Surface roughness values are not affected by abnormal factors. (4) Surface roughness values measured within the measuring area are sufficient to represent (5) The roughness of entire work piece. (6) The effect of approach angle is not considered. (7) Vibration is negligible. (8) Tool nose radius is constant. Where n is the learning step, is the learning rate and is the momentum constant. In Eq. 4, pj is the error term, which is given as follows: For output layer:

pk =(dkp - okp )(1 - okp ), k = 1,.K

(3)

After turning all the specimens, the surface roughness (Ra) was measured by using the surface roughness tester and shown in Table 2. The measurement was done separately and the measured Ra values are utilized for the purpose of training the developed neural networks. The parameters which influence the surface roughness are taken into consideration. A combined set of test values are also obtained with the values of surface roughness, which can be compared after implied with the neural network procedure. 3.3.3 Architecture of the proposed artificial neural network Figure 3 shows the developed architecture of artificial neural network and comprises one input layer with 3 neurons, one output layer with 1 neuron and one hidden layer with 9 neurons in the layers, respectively. The links with synaptic weights are connected between neurons and the back-propagation training algorithm is based on weight updates so as to minimize the sum of squared error for K number of output neurons, given as Where J is the number of neurons in the hidden layer. The training process is initialized by assigning small random weight values to all the links. The inputoutput patterns are presented one by one and updating the weights each time. The mean square error (MSE) at the end of each epoch due to all patterns is computed as For hidden layer: Fig. 3 ANN architecture with a single hidden layer.

pj = opj(1 - opj )

, j = 1,.J

(4)

E=

(1) MSE =
Where dk,p=desired output for the pth pattern. The weights of the links are updated as

(5)

Where NP=number of training patterns. The training process will be terminated when the specified goal of MSE or maximum number of epochs is achieved. The activation function
30

w ji(n+1)= w ji(n) + pj opi + wji(n)

29

(2)

for the input and the two hidden layers is chosen as tansigmoidal function. The activation function for the output layer is chosen as pure linear function. The network is then simulated for the input values and a graph is plotted between the output and target (neural network output) values. The network created is trained for the input and output values. The stopping criterion for training was number of epochs and is given as 590. The network is again simulated for the input values and the target values of the experiments conducted. The input values for the test readings are then given and the network is trained. The target value is then obtained and compared with actual output. The predicted Ra values are compared with the actual Ra values and the predicted Ra values obtained from the present study show minimal in variations. The parameters taken could be confidently used for the above method for predicting the Ra values. The behaviors of the parameters are also noted. The predicted value of Ra is compared with the respective measured average values of Ra and the absolute percentage error is computed, which is given as 3 4 1 2

Table 2.1 Experimental results and predicted reading and actual roughness:

Sample number

Cutting speed (v) (rpm)

Feed rate (f) (mm/rev)

Depth of cut (d) (mm) 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5 2 1 1.5

Actual roughness (Ra) (m)

Predicted roughness (Ra) Difference % Error

2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000 1500 2500 2000

0.15 0.15 0.15 0.2 0.2 0.2 0.25 0.25 0.25 0.15 0.15 0.15 0.2 0.2 0.2 0.25 0.25

3.47 3.87 4.41 4.94 4.72 5.46 5.61 5.72 5.82 1.68 1.97 2.17 2.47 2.97 3.44 3.57 4.73

3.9869 4.4269 4.8155 5.318 5.3179 5.8918 6.1038 6.1336 6.1915 2.0335 2.3238 2.6168 2.9522 3.5105 4.0122 4.0921 5.3122

0.5169 0.5569 0.4055 0.378 0.5979 0.4318 0.4938 0.4136 0.3715 0.3535 0.6538 0.4468 0.4822 0.5405 0.5722 0.5221 0.5822

17.5036402 16.8090308 10.1261081 8.28583954 14.5047427 8.58756613 9.65169462 7.79436153 6.81839038 26.6490765 49.6733019 25.9285051 24.2579736 22.247376 19.9525769 17.1298271 14.0363566

% Absolute error =

(6)

5 6 7 8 9 10 11 12 13 14 15 16 17

Where Ra, actual is the measured value (average) and Ra, predicted is the ANN predicted value of the response for i th trial.

31

32

18 19 20 21 22 23 24 25 26 27

1500 2500 2000 1500 2500 2000 1500 2500 2000 1500

0.25 0.15 0.15 0.15 0.2 0.2 0.2 0.25 0.25 0.25

2 1 1.5 2 1 1.5 2 1 1.5 2

4.81 0.52 0.68 0.88 1.68 1.79 1.91 2.33 2.36 2.69

5.2972 0.6146 0.7942 1.0022 1.9302 2.0211 2.0603 2.4611 2.4922 2.7504

0.4872 0.1946 0.2642 0.1222 0.2502 0.2311 0.1503 0.1311 0.1322 0.0604

11.2704728 59.803319 63.5401635

CHAPTER 4 4 RESULTS AND DISCUSSION

The actual roughness values have been calculated for each set of readings and the

16.1256268 17.4989509 14.8245558 8.54122862 5.96207195 5.9341054 2.29692729

same is compared with predicted roughness values. The behaviour of roughness with various parameters is analyzed and the influence of each parameter over surface roughness is identified from the experiments carried out. The percentage deviation between actual roughness values and predicted roughness values have been obtained and shown in Table 3 and also it is calculated that the average percentage of error is 12.93%. 4.1 Comparison of graphical results Figure 6 shows comparison between actual and predicted roughness values and it is observed that there is a good agreement between experimental and predicted Ra values. The test values obtained from the neural network after training the values are found to be closer to the experimental values. The 3D plots were drawn using Minitab 5.0 software package to identify the influence of the parameters over surface roughness for the experimental values. Optimization of cutting parameters can be obtained for the usage of the same in machining area in mass production. This will reduce the inspection of the product, which is a quality check in any industry. The material is widely used in the pump industry where surface roughness plays an important role. The methodology using ANN for predicting parameters are utilized for the same actual roughness values. The test values obtained from the neural network after training the values are found to be closer to the experimental values. The 3D plots were drawn to identify the influence of the parameters over surface roughness for the experimental values. Optimization of cutting parameters can be obtained for the usage of the same in machining area in mass production. This will reduce the inspection of the product, which is a quality check in any industry. The material is widely used in the pump industry where surface roughness plays an important role. The methodology using ANN for predicting parameters are utilized for the same. Figure 9.1 shows the interaction plot between surface roughness and speed with respect to feed. If the feed rate increases gradually towards the speed, the surface roughness value decreases proportionately. The surface roughness will be improved by increasing the feed for higher speeds.

33

34

Figure 9.2 shows the interaction plot between surface roughness and f with respect to speed. If the Speed is increased with respect to the feed, the surface roughness value increases and decreases. The surface roughness will be improved for higher speeds by increasing the feed rate. Figure 10.1 shows the interaction plot between surface roughness and depth of cut with respect to speed. Depth of cut (DOC) influences more on surface roughness and if DOC is increased, the surface roughness values will also increase for variable speeds. The roughness values are decreased for minimum DOC values with respect to speed. Figure 10.2 shows the interaction plot between surface roughness and speed with respect to DOC. If the speed is increased gradually, the surface roughness value increases with respect to the DOC. The surface roughness will be improved by decreasing the speed for depth of cut values. The cutting parameters play an important role in obtaining the surface roughness of a machined part. These variables are independent and hence it is analyzed with a methodology using ANN to obtain a model which will be useful for the industries. Figure 11.1 shows the interaction plot between surface roughness and DOC with respect to feed. If the feed rate is increased gradually towards the DOC, the surface roughness value decreases .The surface roughness will be improved by increasing the feed for DOC values. The surface roughness value is higher when the feed rate is not increased for higher depth of cut values. The DOC plays an important role as a parameter in CNC turning. Hence, the parameters are analyzed for obtaining improved surface roughness characteristics. Optimization of the same may be done with the help of ANN. Figure 11.2 shows the interaction plot between surface roughness and feed with respect to DOC. If the feed rate is increased gradually towards the DOC, the surface roughness value decreases.

Fig 6 Comparison of actual and predicted roughness

35

36

Fig 9 The actual roughness Ra value obtained aluminium with silicon composite material.

Surface Plot of Ra vs Speed, Feed

6.00
Actual roughness

5.50 5.00 4.50

Feed rate
Depth of cut

4.00 3.50 3.00 1500 2000 Cutting speed 2500

0.25 0.2 0.15

Surface Plot of Ra vs Speed, depth of cut

6 5.5
Actual roughness

5 4.5 4 3.5 3 2.5 2 1500 0.1 2000 Cutting speed 2500 2 1.5

Fig 10 The actual roughness Ra value obtained aluminium with tin composite material.

37

38

Fig 11 The actual roughness Ra value obtained for bronze composite material

Surface Plot of Ra vs Speed, Feed

Surface Plot of Ra vs Speed, Feed

6.00 5.00
Actual roughness

4.00 3.50
Actual roughness

4.00 3.00 2.00

Feed rate

3.00 2.50 2.00 1.50 1.00 0.50 0.00 1500 2000 2500 0.15 0.2 0.25
Feed rate
Depth of cut

1.00 0.00 1500 2000 Cutting speed 0.2 0.15 2500

0.25

Surface Plot of Ra vs Speed, depth of cut

5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 1500 0.1 2000 Cutting speed 2500

Cutting speed

Surface Plot of Ra vs Speed, depth of cut

3 2.5
Actual roughnes s

Actual roughnes s

2 1.5 1 0.5 1.5 0 1500 0.1 2000 2500 Cutting s peed 2

1.5

39

Depth of cut

40

CHAPTER 4 4. PROBLEM DEFINITION

Surface roughness was important role in machining process, it may varied to its material, cutting parameters, type of cutting tools used. To analysis the surface roughness on different machining parameters on CNC turning operation for the prediction of surface finish. To bring out a finished work piece with required surface quality in order to gives a perfect mating with other parts. The parameters namely the spindle speed, depth of cut and feed rate are varied to study their effect on surface roughness.

CHAPTER 5 METHODOLOGY
OBJECTIVE

STUDY OF CNC TURNING PARAMETERS

STUDY OF SURFACE ROUGHNESS ANALYSIS. Phase-1

LITERATURE SURVEY

4.1 OBJECTIVE REPORT OF PHASE -1 The objective of this project work is the prediction and analysis of surface roughness on different materials for different machining parameters during turning operations in a Computer Numerically Controlled (CNC) Lathe. The main purpose of the project to bring out a finished work piece with required surface quality in order to gives a perfect mating with other parts. TURNING OPERATION Phase-2 SURFACE ROUGHNESS MEASUREMENT/ ANALYSIS. COMPARISION & TABLE PREPARATION SELECTION OF TURNING PARAMETERS LIKE TOOL, SPEED, FEED,. Etc. . SELECTION OF MATERIALS

CONCLUSION/ REPORT PREPARATION

41

42

[8]. Joseph Davidsona.M, Balasubramanian.K, Tagore.G.R.N. Surface roughness prediction of flow-formed AA6061 alloyby design of experiments. Journal of materials processing technology 202 (2008) Page no; 4146. [9]. Raja Dhasd.J.,

CHAPTER 6 REFERENCES
[1]. Ahilana.C,SomasundaramKumanan, Sivakumaranc.N, Edwin

MuammerNalbant,

HasanGokkaya,

IhsanToktas,

Gokhan

Sur,.

The

experimental investigation of the effects of uncoated, PVD- and CVD-coated

Modeling and prediction of machining quality in CNC turning processusing intelligent hybrid decision making tools. Journal of ASOC-1559 (2012). [2]. AmanAggarwal, Hari Singh,Pradeep Kumar, Manmohan Singh,. Optimization of multiple quality characteristics for CNCturning under cryogenic cutting environment using desirability function. Journal of materials processing technology 205 ( 2008 ) Page no; 42 50. [3]. AmanAggarwal, HariSinghb, Pradeep Kumar, Manmohan Singh. Optimizing power consumption for CNC turned parts usingresponse surface methodology and Taguchis technique Acomparative analysis. Journal of materials processing technology 200 (2008) Page no; 373384. [12]. SatheeshKumar.N, Ajay Shetty, Ashay Shetty, Ananth K, Harsha Shetty. Effect of [4]. Anil Gupta, Hari Singh, AmanAggarwal,. Taguchi-fuzzy multi output optimization (MOO) in high speed CNC turningof AISI P-20 tool steel. Expert Systems with Applications 38 (2011) Page no; 68226828. [5]. Daniel Kirby.E, Joseph C. Chen,. Development of a fuzzy-nets-based surface roughness prediction system in turning operations. Computers & Industrial Engineering 53 (2007) Page no; 3042. [6]. DurmusKarayel,. Prediction and control of surface roughness in CNC lathe using artificial neural network. Journal of materials processing technology 209 (2009) Page no;31253137. [7]. IlhanAsiltrk and HarunAkkus, Determining the effect of cutting parameters on surface roughnessin hard turning using the Taguchi method. Journal of Measurement 44 (2011) Page no; 16971704. spindle speed and feed rate on surface roughness of CarbonSteels in CNC

cemented carbide inserts and cutting parameters on surface roughness in CNC turning and its prediction using artificial neural networks. Journal of Robotics and Computer-Integrated Manufacturing 25 (2009) Page no; 211223. [10]. Oktema.H, Erzurumlu.T, Kurtaran.H,. Application of response surface methodology
in the optimization of cutting conditions for surface roughness. Journal of Materials Processing Technology 170 (2005) Page no; 1116. [11]. Ramesh.S, Karunamoorthy.L, Palanikumar.K,. Measurement and analysis of surface roughness in turning ofaerospace titanium alloy (gr5). Journal of Measurement 45 (2012) Page no; 12661276.

turning.International Conference on Modeling, Optimization and Computing (ICMOC 2012).Procedia Engineering 38 ( 2012 ) Page no; 691 697. [13]. SuleymanNeseli ,SuleymanYaldz, ErolTurkes,. Optimization of tool geometry parameters for turning operations basedon the response surface methodology. Journal of Measurement 44 (2011) Page no; 580587. [14]. Suresh P.V.S, VenkateswaraRao P, Deshmukh S.G, .A genetic algorithmic approach for optimization of surface roughness prediction model. International Journal of Machine Tools & Manufacture 42 (2002) Page no; 675680. [15].VikasUpadhyay, P.K. Jain, N.K. Mehta. In-process prediction of surface roughness in turning of Ti6Al4V alloy using cutting parameters and vibration signals. International Journal of Fatigue 32 (2010) 247255.

43

44