CHAPTER 1 INTRODUCTION SURFACE ROUGHNESS Surface roughness is a component of surface texture

CHAPTER 1
INTRODUCTION
SURFACE ROUGHNESS
Surface roughness is a component of surface texture. It is defined as the shorter frequency of real surfaces relative to the troughs. Surface roughness is quantified by the deviations in the direction of the normal vector of a real surface from its ideal form. Surface roughness can be measured to ensure surface quality. The nature of surface is defined by surface characteristics such as lay, waviness and roughness. Lay is the direction of the predominant surface pattern. Roughness is defined as closely spaced irregularities. Waviness is more widely spaced irregularities. Surface roughness can be measured by using following parameters such as arithmetical mean roughness(Ra), maximum height(Ry), ten-point mean roughness(Rz), mean spacing of profile irregularities(Sm), mean spacing of local peaks of the profile(S) and profile bearing length ratio(tp).

Figure 1.1. Surface Characteristics
SURFACE ROUGHNESS MEASUREMENT METHODS
Surface roughness has been measured by various methods,
Contact type measurement.

Non-contact type measurement.

In contact type, surface roughness can be measured by,
Profilometer.

Atomic Force Microscopes.

CMM.

In Non-contact type, surface roughness measured by,
Machine vision method.

Optical method.

Capacitive method.

Ultrasonic method.

Fiber Optic method.

Prediction method.

1.3CONTACT TYPE MEASUREMENT
These methods enable to determine a numerical value of the surface finish of any surface. Nearly all instruments used are stylus probe type of instruments. These operate on electrical principles. Further, these electrical instruments can be of two kinds. In first type, they operate on the carrier-modulating principle. The movements of the stylus exploring the surface are caused to modulate a high frequency carrier current. The second type includes those operating on voltage-generating principle. In these the movements of the stylus are caused to generate a voltage signal. In both these types the output has to be amplified and the amplified output is used to operate a recording or indicating instrument. The carrier modulated frequency type of instruments have the advantage that the signal fed to the recorder depends only upon the position of the stylus. While in the voltage generating type, when the oscillatory movement of the stylus stops, the output falls to zero no matter where the stylus may be. Profilometer is a measuring instrument used to measure a surface profile. It can measure small surface variations in vertical stylus displacement as a function of position. A typical Profilometer can measure small vertical features ranging in height from 10 nanometers to 1 millimeter (Marc Detour et al. 2006). The height position of the diamond stylus generates an analog signal which is converted into a digital signal, stored, analyzed, and displayed (Detour, 2006).

In atomic force microscope contact mode, the tip is “dragged” across the surface of the sample and the contours of the surface are measured either using the deflection of the cantilever directly or, more commonly, using the feedback signal required to keep the cantilever at a constant position. Because the measurement of a static signal is prone to noise and drift, low stiffness cantilevers are used to achieve a large enough deflection signal while keeping the interaction force low.

Another contact type measuring method is coordinate measuring machine (CMM) is a device for measuring the physical geometrical characteristics of an object. This machine may be manually controlled by an operator or it may be computer controlled. Measurements are defined by a probe attached to the third moving axis of this machine. Probes may be mechanical, optical, laser, or white light, among others. A machine which takes readings in six degrees of freedom and displays these readings in mathematical form is known as a CMM.

1.4NON-CONTACT TYPE MEASUREMENT
A non-contact type measuring instruments uses sensors and vision system. It involves machine vision method, optical method, capacitive method, ultrasonic method, fiber optic method and prediction method.

1.4.1 Machine Vision
Machine vision method is a method in which microcomputer-based vision system to analyse the pattern of scattered light from the surface to derive a roughness parameter. In this method, stylus parameters (Ra and Rsm) are compared with vision based parameters (Ga, R1, R2 and contrast etc.,). Machine Vision method provides a fast and accurate means for measuring surface roughness. Its repeatability and versatility compares favourably with other methods. Surface roughness resulted to be high accuracy in this method.( G Dilli Babu et.al,2010).

1.4.2 Optical Method
An optical method is the use of a laser source in which the reflected laser light intensity from the surface may represent the surface roughness of the illuminated area. In this method, the measurement of the surface roughness of the stainless-steel samples using a He-Ne laser beam. In this method, a relation can be developed between the reflected laser beam intensity and the surface roughness of the metals. In the measurement, a Gaussian curve parameter of a Gaussian function approximating the peak of the reflected intensity is measured with a fast response photodetector.( Z.Yilbas et.al,1998).

1.4.3 Capacitive Method
Capacitive method is a method in which it uses capacitive sensor for the measurement of surface roughness. In this method, it characterize dielectric films and surface roughness on conducting substrates with a capacitance sensor. The dielectric constant and the thickness of dielectric films can be measured using a corrugated electrode and a flat electrode. In this method, the statistical parameters of a rough surface with gaussian statistics can be obtained from two capacitance measurements. (N. C. Bruce et.al,2003).
1.4.4 Ultrasonic Method
Ultrasonic method is a Doppler based method which uses Doppler effect for surface roughness measurements. In this method. an ultrasonic transmitter emits sound pulse that travel across to under test surface and then the reflected wave is separated into many weak sound pluses, which are received by the receiver. The Doppler effect affects the frequency of the received wave with respect to the surface roughness at the reflecting point. Roughness can be analyzed by the relation between the Doppler shift and the roughing slope. Implementing the Doppler effect instead of the ultrasonic transit-time to describe the roughing parameters will result in a lower dependency of the sensor on the sound speed in air which provides a more precise measurement.( J. Rezanejad Gatabi et.al,2009).

1.4.5 Fiber Optic Method
Fiber optic method is a method which uses multi-wavelength fiber sensor for measuring surface roughness. In this method, the laser used as light source with 650 nm, 1310 nm and 1550 nm for measuring specimens with different surface roughness. The advantages of using multi-wavelength fiber sensor is that it reduces unsystematic errors and increase accuracy in surface roughness. The surface roughness has a linear relationship with light scattering intensity ratio. The accuracy of surface roughness can be measured is high by using multi-wavelength fiber sensor when compared single wavelength fiber sensor.( Zhu Nan-nan et.al,2016).

1.4.6 Prediction Method
In this method, surface roughness was predicted by different cutting parameters such as speed, feed and depth of cut. Artificial Neural Network(ANN) and Multiple regression method is used to predict the surface roughness in this method. Artificial neural networks (ANNs) emulating the biological connections between neurons are known as soft computing techniques. Multiple regression is a statistical technique that allows us to determine the correlation between a continuous dependent variable and two or more continuous or discrete independent variables. ANN model estimates the surface roughness with high accuracy compared to the multiple regression model.( Ilhan Asilturk et.al,2011).

1.5OUTLINE OF THE PROJECT
Many researchers working on above mentioned methods. Industries are very keen in online measurements to improve the productivity and to reduce the unproductive time. Measurements were considered as the unproductive process. To satisfy the need of industries we have two options which we may suggest for online measurement of surface roughness. One is non-contact method and the other one in surface roughness prediction.
In this project, it is planned to predict surface roughness by using various parameters such as feed rate, cutting speed, depth of cut, spindle vibration and additionally a capacitive sensor which is going to be placed nearer to the machining part. The parameters considered in this prediction type are measured using different types of sensors. The sensors such as accelerometer, capacitance sensor in combination with feed, speed and depth of cut has to be given as input to the neural network to predict the roughness in grinding process.

CHAPTER 2
LITERATURE REVIEW
2.1NON-CONTACT TYPE TECHNIQUES
G Dilli Babu et al. (2010) discussed on evaluation of surface roughness using
machine vision. In this paper, surface roughness of milled surfaces has to be determined using machine vision system. Design of Experiments (DoE) technique used for checking the effectiveness of the machine vision based results for a wide range of surface roughness were generated by CNC milling centre. In this work, stylus based parameters Ra and Rsm are compared with vision based parameters (Ga, R1, R2 and contrast,etc,.). In this experiments, a 100mW florescent light to illuminate the rough surface and image was captured by Nikon coolpix 4500 digital camera. An image processing toolbox in Matlab software was used for analysis of the captured image. Vision parameters such as Ga, R1, R2 and contrast,etc,. were computed from the stored image using Matlab program. Then the vision parameters were compared with the average surface roughness (Ra), and Mean line peak spacing (Rsm). Surface roughness resulted to be reasonable accuracy in this method.

Ilhan Asilturk et al. (2011) discussed on modeling and prediction of surface roughness in turning operations using artificial neural network and multiple regression method. In this study, surface roughness was predicted by different cutting parameters such as speed, feed and depth of cut. Artificial neural networks (ANN) and multiple regression approaches are used to model the surface roughness of AISI 1040 steel. Artificial neural networks (ANNs) emulating the biological connections between neurons are known as soft computing techniques. In this method, the back-propagation training algorithms such as the scaled conjugate gradient (SCG) and Levenberg–Marquardt (LM), were used for ANNs training. Multiple regression is a statistical technique that allows us to determine the correlation between a continuous dependent variable and two or more continuous or discrete independent variables. Finally, ANN model estimates the surface roughness with high accuracy compared to the multiple regression model.
F. Forouzbakhsh et al. (2009) discussed on a new measurement method for ultrasonic surface roughness measurements. In this paper, it proposes the application of Doppler-based ultrasonic method to surface roughness measurements. In this method, an ultrasonic transmitter emits sound pulses that travel across to the under-test surface and then the reflected wave is separated into many weak sounds, which are received by the receiver. The frequency of the received wave to be shifted caused by the Doppler effect with respect to the surface roughness at the reflecting point. Implementing the Doppler
effect instead of the ultrasonic transit-time to describe the roughing parameters will result in a lower dependency of the sensor on the sound speed in air which provides a more precise measurement in various environmental situations
A. Guadarrama Santana et al. (2003) discussed on a new approach for measuring surface parameters by a capacitive sensor. In this method, it characterize dielectric films and surface roughness on conducting substrates with a capacitance sensor. The dielectric constant and the thickness of dielectric films can be measured using a corrugated electrode and a flat electrode. In this method, the statistical parameters of a rough surface with gaussian statistics can be obtained from two capacitance measurements. Parametric technique is defined as the latter types of technique for random surface characterization. Capacitance is very sensitive to displacement, shape, or dielectric constant of materials. In this method, by using perturbation theory possible to determine the thickness of dielectric film and dielectric constant.

Qianzhu Liang et al.(2017) discussed on low-cost sensor fusion technique for surface roughness discrimination with optical and piezoelectric sensors. Sensor fusion is widely utilized in robotic tactile sensing as multiple sensors or multimodal sensors can be combined to improve system performance. In this method, optical sensor incorporate with piezoelectric tactile sensor to recognize surface roughness. To discriminate surface
roughness various working mechanisms were demonstrated. In this paper, sensor fusion includes two approaches feature level fusion and decision level fusion. Piezoelectric sensor measures change in mechanical force because of piezoelectric effect. Optical sensor has the capability to measure surface roughness in a non-contact method. The highest classification accuracy of 99.88% and 98.83% can be obtained with decision-level fusion and feature-level fusion respectively.
T. f. van Niekerkl et al. (2004) discussed on multi-sensor fusion model for on-iine surface roughness prediction. This paper describes a statistical method to identify relevant sensors and the neuro-fuzzy (NF) modeling technique to predict surface roughness on-line for a machining process. A laser measuring system, which employs a linear charge coupled device sensor and a neural network to process captured light patterns scattered from the work piece surface, was developed to predict the maximum peak to-valley roughness. The on-line measurement of surface roughness in machine
turning implies assessing the conditions of a work piece just behind the cutting edge of the tool. In this method, FL model for surface roughness measurement was generated by FuzzyTech NF module. Force in the direction of the feed (Fx), force in the direction of cut (Fz),tool work piece vibration (Vy), cutting sound (Sc), spindle current (Is), cutting tool temperature (Tt) and power in the cut (Pc) were machining parameters measured. Finally, Ra with an accuracy of 86 .44%. has been measured using the FL model.

Zhu Nan-nan et al. (2016) discussed on surface roughness measurement based on fiber optic sensor. In this paper, the multi-wavelength fiber sensor for measuring surface roughness and surface scattering characteristics were investigated. The experimental results indicate that multi-wavelength fiber sensor can accurately measure surface roughness, and can effectively reduce the unsystematic error. The accuracy for measuring surface roughness by multi-wavelength fiber sensor is about twice as large as that by single wavelength fiber sensor.

Laura Aulbach et al. (2017) discussed on non-contact surface roughness measurement by implementation of a spatial light modulator. This article proposes a new non-contact method for measuring the surface roughness that is straight forward to implement and easy to extend to online monitoring processes. The key element is a liquid-crystal-based spatial light modulator, integrated in an interferometric setup. By varying the imprinted phase of the modulator, a correlation between the imprinted phase and the fringe visibility of an interferogram is measured, and the surface roughness can be derived.The experimental results are compared with values obtained by an atomic force microscope and a stylus profiler.

CHAPTER 3
METHODOLOGY
CHAPTER 4
WORK CAPTIAL EXPENSES
CHAPTER 5
CONCLUSION