AKG Int Journal Tech Vol 6 No 1

AKGECINTERNATIONAL JOURNAL INTERNATIONAL JOURNAL

OF TECHNOLOGYOF TECHNOLOGYJanuary-June 2015 Vol. 6, No. 1

AKGEC

Ajay Kumar Garg Engineering CollegePublished by:

Ajay Kumar Garg Engineering College27th km Stone, Delhi-Hapur Bypass Road, PO Adhyatmic Nagar, Ghaziabad 201009 (U.P.) India

Design of a Digital Clock Using Very High Speed IC Hardware 1Description LanguageUsman Sammani Sani and Ibrahim Haruna Shanono

Application of Queuing Model: With Special Reference to Construction 4and Business Bank Adama Branch EthiopiaDr. Kamlesh Kumar Shukla and Dr. Girish Kumar Painoli

Hybrid Method For Automatically Filling of the Chemical Liquid 10into Bottles Using PLC & SCADAJagat Dhiman and Er. Dileep Kumar

Friction Stir Welding: Tool Material and Geometry 16A. Chandrashekar, B.S. Ajay Kumar and H.N. Reddappa

Determination of Natural Frequency of a Turning Specimen 21Subjected to Random ExcitationM.Z. Hussain, Dr. A.A. Khan and Dr. M. Suhaib

Circuit Architecture for Photon Counting Pixel Detector 26with Threshold CorrectionDr. Amit K. Jain

Public Sector Comparator: A Useful Decision Making Tool in Infrastructure 31Akhil Damodaran

Variation of the Capacitance of Supercapacitors with Current and Frequency 36Usman Sammani Sani and Ibrahim Haruna Shanono

An Efficient Carry Select AdderA Review 39Rishabh Rai and Rajni Parashar

FPGA Implementation of Digital Modulators 46Pronnati and Dr. K.K. Tripathi

Design and Implementation of Inset feed Square Patch Micro strip Antenna 50Array for WLAN Application Using Dielectric SubstratePriya Upadhyay, Dr. Ranjit Singh and Arundhati Tiwari

A Novel Method of Extracting Mark-Hauwink-Sakurada 55Parameters from Viscosity DataDr. Aniruddh Singh and Mohammad Asad

ISSN 0975-9514

Prof. Kripa ShankerFormer Vice ChancellorGautam Buddha Technical UniversityLucknow

Prof. H.M. GuptaDepartment of Electrical EngineeringIndian Institute of TechnologyNew Delhi

Prof. Suneet TuliCentre of Applied Research in ElectronicsIndian Institute of Technology,New Delhi

Prof. M. L. KothariDepartment of Electrical EngineeringIndian Institute of TechnologyNew Delhi

Prof. Bhim SinghDepartment of Electrical EngineeringIndian Institute of TechnologyNew Delhi

Prof. S.P. SinghDepartment of Mechanical EngineeringIndian Institute of TechnologyNew Delhi

Prof. Vinod KumarDepartment of Electrical EngineeringIndian Institute of TechnologyRoorkee

EDITORIAL ADVISORY BOARD

Prof. Durg Singh ChauhanVice ChancellorUttrakhand Technical UniversityDehradun (Uttrakhand)

Shri S. N. GuptaSecretary GeneralNGN Forum (India)New Delhi

Shri V.K. GuptaDirectorNational Power Training Institute New Delhi

Prof. Surinder PrasadDepartment of Electrical Engineering, Indian Institute of TechnologyNew Delhi

Prof. N.N. KishoreDepartment of Mechanical EngineeringIndian Institute of TechnologyKanpur

Prof. N.S. VyasDepartment of Mechanical EngineeringIndian Institute of TechnologyKanpur

Patron-in-ChiefDr. R. K. AgarwalDirectorAjay Kumar Garg Engineering CollegeGhaziabad

Editorial Team

Editor-in-Chief: Dr. Ranjit SinghEditor: Dushyant S. Chauhan

January - June 2015 Vol. 6, No. 1 ISSN 0975-9514

AKGEC International Journal of Technology

Ajay Kumar Garg Engineering College

CONTENTS

Patron-in-ChiefDr. R.K. Agarwal

Editor-in-Chief Dr. Ranjit Singh

(ISO 9001 : 2008 Certified and Accredited by NBA)27th Km Stone, Delhi-Hapur Bypass Road, PO Adhyatmic Nagar, Ghaziabad 201009 (U.P.)Phone : 0120 - 2762841-51 Fax: 0120 - 2761844 Website: www.akgec.org, E-mail: [email protected]

Address forcorrespondence

Design of a Digital Clock Using Very High Speed IC Hardware 1Description LanguageUsman Sammani Sani and Ibrahim Haruna Shanono

Application of Queuing Model: With Special Reference to 4Construction and Business Bank Adama Branch EthiopiaDr. Kamlesh Kumar Shukla and Dr. Girish Kumar Painoli

Hybrid Method For Automatically Filling of the Chemical Liquid 10into Bottles Using PLC & SCADAJagat Dhiman and Er. Dileep Kumar

Friction Stir Welding: Tool Material and Geometry 16A. Chandrashekar, B.S. Ajaykumar and H.N. Reddappa

Determination of Natural Frequency of a Turning Specimen 21Subjected to Random ExcitationM.Z. Hussain, Dr. A.A. Khan and Dr. M. Suhaib

Circuit Architecture for Photon Counting Pixel Detector 26with Threshold CorrectionDr. Amit K. Jain

Public Sector Comparator: A Useful Decision Making 31Tool in InfrastructureAkhil Damodaran

Variation of the Capacitance of Supercapacitors with 36Current and FrequencyUsman Sammani Sani and Ibrahim Haruna Shanono

An Efficient Carry Select AdderA Review 39Rishabh Rai and Rajni Parashar

FPGA Implementation of Digital Modulators 46Pronnati and Dr. K.K. Tripathi

Design and Implementation of Inset feed Square Patch Micro 50strip Antenna Array for WLAN Application Using Dielectric SubstratePriya Upadhyay, Dr. Ranjit Singh and Arundhati Tiwari

A Novel Method of Extracting Mark-Hauwink-Sakurada 55Parameters from Viscosity DataDr. Aniruddh Singh and Mohammad Asad

Copyright AKGEC International Journal of Technology

No portion of the material published in the AKGEC International Journal of Technology should be reproduced in any form without the written permission of the Editor.

DisclaimerThe views expressed by the authors do not necessarily represent those of the Editor or Publisher, or the management of the Ajay Kumar Garg Engineering College. Though every care has been taken to avoid errors, this journal is being published on the condition and understanding that all the information provided herein is merely for reference and must not be taken as having authority of or binding in any way on the authors, editor and publisher who do not owe any responsibility for any damage or loss to any person, for the result of any action taken on the basis of this work. The publisher shall be obliged if mistakes are brought to their notice.

ii

To introduce undergraduate and postgraduate courses for all

engineering branches and award of Ph.D degree. To be one of

the best engineering colleges in the country

and to be a deemed university.

Our MissionWe strive to provide and maintain academic environment and systems, enabling

maximum learning to produce competent professionals. We also aim at

achieving this through transparent academic and administrative policies

in the college. We intend to provide conducive atmosphere

for research, development and consultancy services to

our faculty at national and international level.

Our Vision

1DESIGN OF A DIGITAL CLOCK

Design of a Digital Clock Using Very High Speed ICHardware Description Language

Abstract -- Very High Speed IC Hardware Description Language(VHDL) is one of the modern languages used in designing digitalcircuits. It can be used in programming Field Programmablegate arrays (FPGAs) and Application Specific Integrated Circuits(ASICs).This paper presents the design of a digital clock usingVery High Speed Hardware Description Language in a XilinxISE 10.1 environment. The designed clock has the functionalitiestime, alarm, stopwatch and date. The clock format can be changedfrom 24 hours to 12 hours and vice versa. After the design, testingwas done on a Spartan-3- FPGA and all units were found to beperforming the desired functions.

Keywords: VHDL, Programmable Logic Devices, FPGA

I. INTRODUCTIONTHERE are different digital logic integrated circuits availablein the market. These can be used in designing circuits by circuitdesigners. But as systems become complex, there may be somefunctions that cannot be performed by these readily availableICs. Hence the use of VHDL for circuit design comes in handy[1]. VHDL is a high level programming language which ispowerful in programming Programmable Logic devices suchas field programmable gate arrays, generic array logic (GAL)and Programmable array logic [2],[3]. For the programmablelogic devices to be programmed the codes must pass throughstages such as synthesis, timing simulations, place and routeand bit file generation [4]. This work will be based on designon an FPGA. The advantage of designing digital circuits usedfor instrumentation and control using it is that the circuits canbe modified even after reaching the market due to thereprogrammability of FPGAs thereby enhancing rapidprototyping [5]. There are many manufacturers of FPGA suchas Xilinx; Inc, Altera Corporation, Perfect Parts Corporation,Achronix Semiconductor Corporation, Atmel Corporation [6],[7] etc. FPGAs contain programmable logic elements calledLogic elements LEs and a hierarchy of reconfigurableinterconnects that allow Les to be connected physically [5].In this work, a Spartan-3 FPGA development board is used.The designed circuit has the following features:

The Clock: This is designed to display time in the format hr:min: sec. System clock is scaled down to 1Hz to trigger the

Usman Sammani Sani1 and Ibrahim Haruna Shanono2

Department of Electrical Engineering,Bayero University, Kano, P.M.B. 3011, Nigeria

[email protected], [email protected]

clock and other components. The seconds count after eachclock pulse till it reaches 59 and on the next system clock pulseit resets back to 0 and then continues. At the time it resets to 0,it triggers the minute. Each time the seconds reset to 0, theminute is being triggered and when it reaches 59, it triggers thehour, the next time it is triggered by the seconds.

Five switches stop, settime, sethr, setmin and inctare used for resetting time. To achieve this, the clock has to bestopped by first setting the stop switch to logic 1 followedby the settime switch. Then either sethr or setmin canbe set high to reset the hour and minute respectively. If sethr isset high, it means that the hour will be changed. The next thingto do is to press the inct button (increment button). Usercan continue incrementing it up to the value of 23 after whichit resets to 0 and the incrementing process can be continued.

The same principle is used to reset minutes using setmin.But in this case it reaches a maximum of 59 before resetting.Time format can be changed by setting the format switch toeither logic 1 or 0. When set to logic 1, the digital clock willbe in 24 hours mode while when set to 0, it will be in 12 hoursmode, which can reach a maximum of 11 hours.

Option switches are used for selecting the variable either time,date, alarm and stopwatch to be displayed by the sevensegments display unit of the Spartan-3 FPGA. The options arefour in number with each enabling one of the mentionedvariables. Only one of them has to be high at a time so as toenable the variable assigned to be displayed.

The minutes are displayed by two seven segment display LEDsof the Spartan-3 FPGA [8] and hour by the other two. Secondsare displayed by six LEDs which presents it in a binary formwith off state represented by a 0 and on state by a 1.

The Stopwatch: This is activated by setting a strtstop slideswitch to logic level 1 and stopped by setting it to logic 0.The stop watch can be run for 60 minutes only. When it reaches59 it resets to 0 seconds and zero minutes on the next clockcycle. Before activating the stopwatch, the correct option

2AKGEC INTERNATIONAL JOURNAL OF TECHNOLOGY, Vol. 6, No. 1

button has to be activated for it to be displayed on the sevensegments display unit. The first two seven segment LEDS givesthe seconds count and others give the minutes count.

The Alarm: The alarm has a switch setalm. When the buttonis set, user can set and adjust the alarm setting using sethr,setmin and inct. At the time when the alarm set timebecomes equal to the clock time, an output pulse is sent out.This is dependant on whether a 24 hours or 12 hours format isselected. User adjusts it by taking into consideration whichtime format is used at a particular moment.

The Date: This has a value of 0 to 31 which increments at theend of each 24 hours. It can be adjusted using an Adateswitch provided. When the Adate switch is set high, usercan use the Inct button to increment it. Before adjusting, thestop button has to be set high and the option buttonrepresenting date display most also be high. The maximum itcan reach is 31 which then resets to 0.

II. METHODOLOGYThe codes for the functionalities mentioned earlier were writtenin eight different files and were assigned names listed below:1. Tb_toplevel.vhd2. Toplevel.vhd3. Clockdivider.vhd4. Clock.vhd5. Alarm.vhd6. Stop.vhd7. Date.vhd8. Scan4digit.vhd

Tb_tplevel.vhd is a test bench file for the entire digitalclock design.

Toplevel.vhd is the top level of the design and it includessix components.

Clock Divider.vhd is a file that provides the clock pulseneeded by all components in this work. The Spartan-3FPGA is set to operate at a frequency of 50MHz. Thisfrequency cannot be used for setting the timings in thedigital clock, and so there is the need to reduce it to 1Hz.To do this, the board oscillator frequency is divided by50 x 106 to obtain a frequency of 1Hz. This is done bysetting a variable count and then increments its valueup to 50000000, covering 50 x 106 cycles. At the end ofthe 50 million cycles, it sends a pulse to the remainingcomponent parts. The variable count resets to 0 and theprocess continues.

Clock.vhd contains the codes describing thefunctionalities of the clock.

Alarm.vhd describes the components of the alarm. Date.vhd describes the function of the date. Stop.vhd is for the stopwatch. Scan4digit.vhd acts as a decoder that receives an input

signal from the various components and determine the Spartan-3 seven segment display on which to display

each digit.

The top level consists of 7 segment display decoders fordisplaying the signals from all components in the system on aseven segment display. It receives a six bit input data andconverts it to a form that represents the binary number on acommon anode seven segment display. The seven segmentdisplay of each of the variables time, alarm, stop watch anddate were generated separately.

Codes for interchanging between one display mode and theother were included in the top level.A new project was created using Xilinx ISE 8.2i with namemydigitalclock. The eight source files mentioned earlier wereadded. Relationship between the files was established throughportmaps and signals in the top level. Each file was saved andsynthesis was carried out. Some minor errors were debuggedand the synthesis stage was passed. The register transfer logicwas generated, pins were assigned using [8] as reference andfinally the bit file was generated. The bit file was programmedonto the Spartan-3 FPGA and testing was performed whichproved that the design is well done.

III. RESULTS

Figure 1. Register Transfer Level of the digital clock.

3DESIGN OF A DIGITAL CLOCK

The table below shows how the resources in the Spartan- 3FPGA were utilized:

TABLE 1: SPARTAN-3 FPGA RESOURCES UTILIZATION

Logic Utilization Used Available UtilizationTotal Number Slice Registers 124 3,840 3%Number Used as Flip flops 98Number Used as Latches 26Number of 4 Input LUTS 444 3,840 11%Logic DistributionNumber of Occupied Slices 259 1,920 13%Number of Slices Containing 259 259 100%Only Related LogicNumber of Slices Containing 0 259 0%Unrelated LogicTotal Number of 4 Input LUTS 485 3,840 12%Number Used as Logic 444Number Used as route-thru 41Number of Bonded IOBs 35 173 20%Number of BUFGMUXs 3 8 37%

IV. CONCLUSIONThe Digital Clock was designed by first creating VHDL codesand synthesizing them using Xilinx ISE 8.2i software. The codespassed synthesis and a bit file was generated. The bit file wasprogrammed onto a Spartan-3 development kit and tested forfunctionality. The circuit performed the desired functions.Design statistics also showed that the FPGA resources werehighly utilized and therefore the design is economical. Thisproves how FPGAs are desirable when dealing with complexsystems.

V. REFERENCES[1]. M. Balch, Complete Digital Design; A Comprehensive Guide

to Digital Electronics and Computer Systems Architecture, McGraw Hill, 2003, pp 221-222.

[2]. B. Holdsworth and C. Woods, Digital Logic Design (fourthedition), Newnes, 2002, pp 295-324.

[3]. J.F. Wakerly, Digital Design Principles and Practices, PrenticeHall, 2005, pp 15-16.

[4]. Enoch O.H., Digital Logic and Microprocessor Design withVHDL, Lasiera University, 2006, pp 23-26.

[5]. http://www.altera.com/products/fpga, accessed 15th July, 2014.[6]. http://www.globalspec.com/local/3127/CA, accessed 15th July,

2014.[7]. http://www.xilinx.com/index.htm, accessed 15th July, 2014.[8]. Nexys Reference Manual.

Usman Sammani Sani graduated fromBayero University, Kano in 2008, where heobtained a bachelor degree of electricalengineering. He then furthered his studies, inwhich he obtained an MSc in ElectronicCommunications and Computer Engineeringfrom The University of Nottingham MalaysiaCampus in 2011.Usman is presently a lecturer in theDepartment of Electrical Engineering, BayeroUniversity Kano. His research interests include

digital communications, digital circuits design and testing of fabricatedelectronic components.

Ibrahim Haruna Shanono received hisB.Eng and MSc. degree from Bayero UniversityKano and Nottingham University in 2008 and2012 respectively. He is currently working withthe Department of Electrical Engineering,Bayero University Kano, Nigeria. His researchinterests are in the areas of Renewable Energy,Power Electronics and Automatic Controlsystems.


Application of Queuing Model: With Special Reference toConstruction and Business Bank Adama Branch Ethiopia

Abstract -- In this paper, queuing model is applied to the datacollected on construction bank, Adama branch, Ethiopia. Theendless customers waiting for service delivery in constructionand business bank is a phenomenon that bothers both themanagement of banking institutions and the customers alike.Thus, the need to optimize total operating cost by determiningthe optimal balance between the cost of making customers towait for service and the cost of providing additional service, somestudies have claimed that service improvement can be achievedby increasing the number of servers, but to what extent this canbe done to minimize overall cost? This study examines thevalidity of multi-server queuing models for achieving reductionin waiting time and minimization of cost and characteristics ofcustomers. After adding one/two servers, further characteristicsof customer were analyzed with help of multiple server model.The average waiting time per customer in a system as well as inthe queue were found about 32 minutes and about 29 minutesrespectively. It reveals a preference for a multi-server systemthat determines its usage and suggestions for improvement inservice delivery are highlighted.

Keywords: Queuing Model, Multi-server System, Service Delivery

I. INTRODUCTIONQUEUING theory had its beginning in the research work of aDanish engineer named A. K. Erlang. In 1909 Erlangexperimented with fluctuating demand in telephone traffic.Eight years later, he published a report addressing the delaysin automatic dialing equipment. At the end of World War II,Erlangs early work was extended to more general problemsand to business applications of waiting lines.

The study of waiting lines, called queuing theory, is one of theoldest and most widely used quantitative analysis techniques.Waiting lines are an everyday occurrence, affecting peopleshopping for groceries buying gasoline, making a bank deposit,or waiting on the telephone for the first available airlinereservationists to answer. Queues, another term for waitinglines, may also take the form of machines waiting to be repaired,trucks in line to be unloaded, or airplanes lined up on a runwaywaiting for permission to take off and restaurant study takenin Indonesia (Dharmawirya and Adi, 2011; Sharma et al., 2013).

Dr. Kamlesh Kumar Shukla1 and Dr. Girish Kumar Painoli21Department of Management, School of Business and Economics,

Adama Science and Technology University, Adama, Ethiopia2Department of Accounting and Finance, School of Business and Economics,

Adama Science and Technology University, Adama, [email protected] , [email protected]

The three basic components of a queuing process are arrivals,service rate, and the actual waiting line.

As the word turns to a global village characterized by intenseand ever increasing demand, operation bank managerscontinue to experience wrenching changes, which they mustkeep up for their survival. Bank customers have also becomeincreasingly demanding. Today, they require high quality, lowprice and immediate service delivery and tomorrow, they wantadditional components of value from their chosen banker. Sinceservice delivery in banks is personal, customers are eitherserved immediately or join a queue (waiting line) if the servingsystem is busy.

Waiting line is what one experience everywhere in daily life i.e.while shopping, checking into hotels, at hospitals and clinicsetc. In situations where facilities are limited and cannot satisfythe demand made upon them, bottlenecks occur which manifestas queue but customers are not interested in waiting in queues.When customers wait in queue, there is the threat that excessivewaiting time will lead to the loss of some customers tocompetitors. But allowing them to serve themselves so easilyis a key factor in both keeping and attracting customers(Michael, 2001).

Statement of the problem: One of the goals of queuing analysisis finding the best level of service for an organization. WhenConstruction and business bank does have control, itsobjective is usually to find a high spirits, medium between twoextremes. On the one hand, a bank can retain a large number ofcustomers and provide many service facilities. This, however,can become expensive. The other extreme is to have theminimum possible number of teller windows open. This keepsthe service cost down but may result in customerdissatisfaction. When the average length of the queueincreases then the poor services will result in the loss ofcustomers and goodwill.

When services improve in speed, then the time spent in waitingwill decrease. This waiting cost may reflect loss of productivity

5APPLICATION OF QUEUING MODEL

of workers while their tools or machines are awaiting repairs ormay simply be an estimate of the cost of customers lost becauseof poor service and long queues.

This assessment is going to address the following questions:-1. Do the customers satisfied with service rate provided by

the bank?2. What are the factors that cause the customers to leave

the bank?3. Does the organization have enough servers to serve the

expected customers?4. How the waiting time will be reduced if there are

alterations in the server?

Significance of the Study: The significance of the study is tofind the best level of service and provide informationconcerning queue analysis for Construction and business bankfor better customer services at minimum cost.

Objectives :The general objective of this study is to know thecharacteristic of the customers in Construction and businessbank at Adama branch. The main objectives of the study are To determine the waiting time of customers is likely to

experience in a system To determine how the waiting time is affected If there is

increase in the number of servers To offer necessary suggestion if any to the bank based

on the analysis of the study

II. DATA AND METHODOLOGYMethod of data collection and Sample size: The study wasconducted at Construction and business bank, Adama branch,Ethiopia where the information about the characteristics ofcustomers was collected. It was carried out on the basis ofdata collection during the period of one week, (i.e. 6 workingdays from December 24, 2013 to January 6, 2014) throughobservation, interview and questionnaire methods and theVariables were analyzed by using the Queuing Models.

The variables measured include arrival rate ( ) and servicerate (). They were analyzed for simultaneous efficiency incustomer satisfaction and cost minimization through the useof multi-channel queuing models. These are compared for anumber of queue performances such as; the average time spentby each customer in the queue as well as in the system, averagenumber of customers in the queue as well as in the system andthe probability of the system being idle.

Primary data in respect of customer arrival rate and service ratewere obtained through observations while customer attitudesurvey was carried out through a questionnaire distributedamong 50 customers by using simple random sampling method.Secondary data were collected from the bank, related booksand other relevant, and published research journal.

Methods of Data Analysis: To make the data suitable for furtheranalysis, classification and editing was made. The raw datawere organized into groups. The data which were collectedfrom the primary and secondary sources were analyzed byusing statistical tools and techniques, such as tables,percentages and graphs etc. A single and multiple servers wereapplied in the simplest form of queuing system.

Definitions of queuing system variables: = the arrival rate (average number of arrivals per time

period) = the service rate (average number served per time

period) = mean arrival rate; = mean service rate And that < (customers are served at a faster rate than

they arrive), we can state the following formulas for theoperating characteristics of a single-server model.(Bernard W. Taylor III 2006, 9th edition)

Customers must be served faster than they arrive, or aninfinitely large queue will build up.

Lq = average queue length (average number ofcustomers in queue)

L = average system length (average number of customersin system, including those being served)

Wq = average waiting time in queue (average time acustomer spends in queue)

W = average time in system (average time a customerspends in queue plus service)

Lq =Wq (Littles Law)=2/(-) L =W (Littles Law)=/- L = Lq +/ W = Wq +1/=L/ U= / I=1-u=1-/P0= (1-/)Pn= (/) n.p0== [(/)n](1-/)The parameters of the multiple-server model are as follows: =the arrival rate (average number of arrivals per time

period) =the service rate (average number of customers served

per time period) per server (channel) C=the number of servers C=the mean effective service rate for the system, which

must exceed the arrival rate C>= : the total number of servers must be able to serve

customers faster than they arrive The probability that there are no customers in the system

(all servers are idle)

III. RESULTS AND DISCUSSIONData Analysis And Interpretation: According to Table 1, 56%of the respondents were served in more than 52 minutes, 10%


of the respondents were served within 21-36 minutes, 10% ofthe respondents were served within 5-20 minutes respectively.This indicates that most of the customer served for more than52 minutes. It is due large number of customers and traditionalmethods adopted by the bank in serving the customers.

TABLE 1 DISTRIBUTION OF THE NUMBER OFRESPONDENTS WITH RESPECT TO THEIR TIME

SPENT IN THE BANK

Time spent No. of respondents Percentage

5-20 minutes 5 10%

21-36 minutes 5 10%

37-52 minutes 12 24%

TABLE 2 -- DISTRIBUTION OF RESPONDENTS AND THEIROPINION ABOUT SATISFACTION WITH TIME OF SERVICE

Opinion No of respondents Percentage

Yes 16 32%

No 34 68%

Total 50 100%

Table 2 shows that 68% of the respondents were not satisfiedwith the time of service and 32% were satisfied with time ofservice. It is observed that the most of the customers were notsatisfied with the service provided by the bank. It may be dueto lack of ethical and moral behavior among the bank employeestowards their customers.

TABLE 3 -- DISTRIBUTION OF RESPONDENTS AND THEIRATTITUDE TOWARDS LONG QUEUE

Opinion No of respondents Percentage

Yes 15 30%

No 35 70%

Total 50 100%

Table 3 reveals that, 30% of the respondents were return backafter finding the long queue, 70% of the respondents wereserved after their arrival at the bank as usual. Therefore it canbe observed that bank loses many customers because of thelong queue.

Quantitative analysis: Operating characteristics computationfor seven servers

TABLE 4 NUMBER OF CUSTOMERS SERVED WITHIN AWEEK FROM DECEMBER 24, 2013 TO JANUARY 6, 2014

Month No of customersserved per a day

December 24, 2013 1618

December 28, 2013 1306

January 1, 2014 2103

January 3, 2014 1295

January 5, 2014 1388

January 6, 2014 1134

Total 8844

Computation of mean service rate and arrival rate

Average numbers of customers served= Total numbers of customers served in a week

Total numbers of days in a week

=8844/6 = 1474Mean service rate ( ) (mean service rate per hour)

= average numbers of customers servedWorking hours per day

1474/9.5 hr. = 155/ hr

(service rate per server)= average numbers of customers served

Total No servers=7

Customers served per server within a day= 1474/7 =210.57 or 211

Mean arrival rate ()TABLE 5 -- DISTRIBUTION OF ARRIVALS OF CUSTOMER

AT THE BANK FOR A WEEK

Selected day Arrival per hour

December 24 157

December 28 153

January 1 167

January 3 144

January 5 156

January 6 141

Total 918


The average arrival rate

(mean arrival rate per hour) 153/hr 155/hr 155/7hr = 22.14 customers served in each window/hr. or mean service rate for each server perC= 7Po= the probability of No customers in the service department

L= the average number of customers in the queuing systemis:-

W= the average time a customer spends in the queue system(waiting and being served) is:-W=L/=82.32/153 =0.53804hrs (32.2824 minutes)Lq= the average number of customers in the queue is:-

= 75.41 customers on average waiting to be served on the line

Wq= the average time a customer spends in the queue, waitingto be served is:-

PW = the probability that a customer arriving the system mustwait for the service (i.e. the probability that all the severs arebusy

It was observed that, customers were frustrated by the relativelylong waiting time of 32.28 minutes and the 0.9859 probabilityof waiting.

Operating characteristics of customers when one/twoserver(s) are added

Operating characteristics computation for eight server/windows:When the servers increase from 7 to 8, the service rate alsoincreases from 155 to 177 per hour

(mean service rate per hour) = 1474 + 211 = 1685 = 177/ hr. 9.5 9 .5 =177/8= 22.14 customers served in each window /hr. or meanservice rate for each server per hour=22.14 Customers served per hour in each window=153 Average customers arrive at the service station in anhour C= 8 Number of servers

The probability that there is No customer in the servicedepartment is 0.00064


L= 42.9 customers, on average in the service department

The average time that the customer spends in the queue systemis 0.2804 hr. or 17 minute.

Lq= 35.99 customers on average waiting to be served on theline

Wq= 0.24 hrs. or 14.11minute on average a customer spends inthe queue to get the service

PW= 0.6026 so the probability of all the servers are busy is0.6026

Operating characteristics computation for nine servers/windowsC=9 severs (service rate per hour) =1474+422 =1896 =197 /hr 9.5 9.5 =22.14 Customers served in each window /hr. =153 CustomersPo= 0.0008854L= 8.16 customers on average in the service departmentW=0.053hrs or 3.29 minutes. The average time customersspends in the queue systemLq=1.25 customers on average in the queue. Waiting to beserved on the lineWq=0.00816hr or 0.49 minutes will the customers to be servedon a linePW= 0.38(38%) so the probability that all severs are busy is0.38 or 38%.

TABLE 6 -- OPERATING CHARACTERISTICS FORALTERNATIVES SERVERS

No. of servers L Lq W Wq PW

7 82.32 75.41 32.28m 29.52m 0.9859

8 42.9 35.99 17m 14.4m 0.6026

9 8.16 1.25 3.18m 0.4896m 0.38

Above table indicates that using of 9-server in system is betterthan both 8 servers and seven servers system in allcircumstance. For instance a 9 servers system has 1.25

customers waiting in the queue for service while an 8-serverssystem and seven servers system has 35.99 and 75.41 customersrespectively on the queue. In an 8 and 7 servers system, acustomer spends 17 minutes and 32.28 minutes in the systemrespectively and customer spends 14.4minutes and29.52minutes on the queue respectively. However, in 9 serverssystem customer waits in the system for about 3.18 and waitsin the queue for 0.4896 minutes.

The probability of being busy for seven -servers system isvery high (98.59%) than either of 8 or 9 servers system. Meannumbers of arrivals was 153 customers per hour. The arrivalprocess follow queue discipline and it was first-come-first-served and queue population was infinite. From the abovestudy it was observed that the adoption of 9 servers is betterthan both 7 and 8 servers in all circumstances.

IV. SUGGESTIONSSuggestions: Based on the analysis of the study, the followingsuggestions are given for improvement of efficiency and qualityof service to customers of the bank. Addition of two service channels will reduce loss of

customers. Providing TV in the waiting hall, comfortable seats, and

toilet facilities for increasing the satisfaction of thecustomers.

The bank should educate their front line employee in theapplication of queuing models for efficient solving ofoperational problems.

Bank should enrich employees job by making them multi-skilled, through continuous training so as to enable themto eliminate unnecessary counter-check handoffs.

The queue characteristics should be viewed from thecustomers point of view i.e. whether the waiting time isreasonable and acceptable or not.

Reengineering the banking operations through ITsolutions e.g. ATM, and online banking etc., to harmonizequeuing model.

V. CONCLUSIONBased up on the above discussion it can be concluded that,for rendering better services to the bank customers modernmodels should be adopted. The average waiting time percustomer in a system as well as in the queue were found about32 minutes and about 29 minutes respectively, however 2servers added in the bank, the average waiting time percustomer in the system and queue were found about3.18minutes and 0.4896 minutes respectively. Customersatisfaction is the significant factor for any industry and moreto service industry to which all banks belong. Therefore theapplication of queuing analysis has proved that the adoptionof it in the day to day operational activities of the bank willsatisfy the customers. Satisfied customers will result in desiredgrowth for the bank and for economy of the country.


VI. REFERENCES[1]. W. Bernard Taylor III Introduction To Management Science

9th, Virginia Polytechnic Institute and State University, 2006,Prentice Hall.

[2]. www.http//safe-associate/2000[3]. T.L. Saaty, Elements of Queuing Theory 1961, New York:

McGraw-Hill Safe Associates (2002)[4]. R.B. Cooper, Introduction to Queuing Theory, 2le, Elsevier,

North Holland, 1980[5]. A.M. Lee, Applied Queuing Theory, St Marints press, New

York, 1966.[6]. Mathias Dharmawirya, and Erwin Adi, Case Study for

Restaurant Queuing Model IPEDR Vol.6 (2011) (2011)IACSIT Press, Bali, Indonesia

[7]. Ajay Kumar Sharma, Queuing theory approach with queuingmodel: a study, International Journal of Engineering ScienceInvention ISSN (Online): 2319 6734, ISSN (Print): 2319 6726 Volume 2 Issue 2 February. 2013 PP.01-11.

Dr. Kamlesh Kumar Shukla is presentlyworking as Assistant Professor in ASTU,Ethiopia since September 2013. Obtained PhDin Statistics from Banaras Hindu University,Varansi, India and Masters degree in Statistics(Gold medallist) in 1997.Selected as Secretary, Governing Council ofForum for Interdisciplinary Mathematics forthe term 2015 2017. Possesses over 12 yearsof teaching experience.Authored a book for BCA students, namely,

Elements of Statistics published by Thakur Publishers, Lucknow. Workedon six projects with organisations such as IIPS, WHO, DST, NewDelhi. Published over 12 papers and presented six papers at conferences.

Dr. Girish Kumar Painoli received theM.Com degree, from Osmania University,Hyderabad, in 1995, M.Phil degree from SGBAmaravathi University in 1999 and Ph.D.degree in management Science, from SwamiRamanand Teerth Marathwada UniversityNanded in 2012. He also qualified UGC-NET.Worked as Professor in Department ofAccounting and Finance at Adama Science andTechonology University, Adama, Ethiopia.Currently, he is a Faculty in Accounting in

Department of Business Studies at Shinas College of Technology, AlAQR, Shinas, Sultanat of Oman. His teaching and research areas includeAccounting and Finance authored/co-authored approximately twentyfive research papers and attended nearly 25 conferences.

10

AKGEC INTERNATIONAL JOURNAL OF TECHNOLOGY, Vol. 6, No. 1

Hybrid Method for Automatically Filling of the ChemicalLiquid into Bottles Using PLC & SCADA

Abstract -- In todays fast-moving, highly competitive industrialworld, a company must be flexible, cost effective and efficient ifit wishes to survive. In the process and manufacturing industries,this has resulted in a great demand for industrial control systems/automation in order to streamline operations in terms of speed,reliability and product output. Automation plays an increasinglyimportant role in the world economy and in daily experience. Aprototype of commercial Hybrid method of automatically fillingthe bottles using PLC&SCADA and show its visualization onSCADA screen, controlled using programmable logic controller(PLC) is proposed and the whole process is monitored usingsupervisory control and data acquisition. This system providesthe provision of mixing any number of liquids in any proportion.Its remote control and monitoring makes the system easilyaccessible and warns the operator in the event of any fault. Oneof the important applications of automation is in the soft drinkand other beverage industries, where a particular liquid has tobe filled continuously. The objective of this paper is to design,develop and testing of the Real time implementation of PLC,SCADA system for ratio control based bottle filling plant. Thiswork will provide low operational cost, low power consumption,accuracy and flexibility to the system and at the same time it willprovide accurate volume of liquid in bottle by saving operationaltime.

Keywords: PLC, SCADA, Sensors, Automation,VFD, Conveyor

I. INTRODUCTIONTHIS research is to design and develop the Hybrid methodof automatically filling the chemical liquid into bottles usingPLC&SCADA and show its visualization on SCADA screen.

Jagat Dhiman and Er. Dileep KumarEternal University, Baru Sahib Road, Baru Sahib, Himachal Pradesh 173001 India

[email protected]

We can operate & control automatically filling of bottles sittingfar away from the plant (for example 500km distance from plant)and we can change all the parameters of the process usingSCADA technology because SCADA system is used assupervisor to monitor the process. The purpose of this toresearch is to apply filling 2 type of chemical liquid into bottlesrandomly by using PLC as a controller. This is a batch operationwhere a set amount of inputs to be process is received as agroup, and an operation produces the finish product. In manyautomation processes it is necessary to achieve a desireddemand in some specified time. If the production rate is 35bottles per minute and the demand increases to 65 bottles perminute, the operating speed needs to be increased, whereas ifthe demand drops abruptly the production rate needs to bedecreased. Thus the research deals with overcoming the problemsof speed control in order to have improved operational parameters.

II. PLC AS SYSTEM CONTROLLERProgrammable logic controller or programmable controller is adigital computer used for automation of industrial process,such as control of machinery on factory assembly lines. Unlikegeneral-purpose computers, the PLC is designed for multipleinputs and output unlike general purpose computers, the PLCis designed for multiple inputs and output. Arrangements,extended temperature ranges, immunity to electrical noise, andresistance to vibration and impact. Programs to controlledmachine operations are typically stored in the battery backedor non volatile memory.

Figure 1: Basic PLC operation process.

11

AUTOMATICALLY FILLING OF THE CHEMICAL LIQUID

Unlike a personal computer though the PLC is designed tosurvive in a rugged industrial atmosphere and to be very flexiblein how it interfaces with inputs and outputs to the real world.a programmable logic controller is simply used in manyindustries such as oil refineries, manufacturing lines, conveyorsystems and so on. Wherever there is a need in control devicesthe PLC provides a flexible way to soft wire the componentstogether.

III. FUNDAMENTAL PRINCIPLESSCADA refers to the combination of telemetry and dataacquisition. SCADA encompasses the collecting of theinformation, transferring it back to the central site, carryingout any necessary analysis and control and then displayingthat information on a number of operator screens or displays.The required control actions are then conveyed back to theprocess. The PLC or Programmable Logic Controller is still oneof the most widely used control systems in industry. As needsgrew to monitor and control more devices in the plant, thePLCs were distributed and the systems became more intelligentand smaller in size. PLCs and DCS or (Distributed ControlSystems) are used as

The advantages of the PLC / DCS SCADA system are: The computer can record and store a very large amount

of data. The data can be displayed in any way the user requires. Thousands of sensors over a wide area can be connected

to the system. The operator can incorporate real data simulations into

the system. Many types of data can be collected from the RTUs. The data can be viewed from anywhere, not just on site.

IV. BLOCK DIAGRAMThis deals with the key components used in settling up thecomplete plant and thus explains the use and working of eachcomponent. The block diagram of the experimental set up isillustrated. The following configurations can be obtained.

The digital computer is used as an interface between PLC andSCADA. The PLC is a micro processor based system controllerused to sense, activate and control industrial equipments andthus incorporate a number of input output/modules whichallows electrical system to be interfaced. SCADA is a centralizedsystem used to supervise a complete plant and basicallyconsists of data accessing features and controlling processesremotely. The communication protocol used is Ethernet. TheVariable Frequency Drive connected to the PLC receives ACpower and converts it to an adjustable frequency adjustablevoltage output for controlling the motor operation. The analogmodule converts analog input signals to digital output signalswhich can be manipulated by the processor.

The output of the VFD is given to the 3-phase induction motorwhich in turn with the help of a pulley mechanism is used tovary the speed of the conveyor belt. An inductive sensor is anelectronic proximity sensors used to detect metallic objectswithout touching them. The solenoid valve is a normally closeddirect acting valve used to pour the liquid in the bottle wheneverit gets a signal from the proximity sensor

V. PLC AND RELATED SOFTWARESThe PLC used is Micro Logix 1200 as it has 10 inputs and 6outputs and has an interface for Ethernet. The Micro Logix1400 system offers higher I/O count, faster high-speed counter/PTO, and enhanced network capabilities The programmingsoftware used is RSLOGIX 500 and the communication softwareused is RS LINX 500.

Features of MicroLogix 1200:Ethernet port provides Web server capability, email capabilityand protocol support Built-in LCD with backlight lets youview controller and I/O status Built-in LCD provides simpleinterface for messages, bit/integer monitoring and manipulationExpands application capabilities through support for as manyas seven 1762 Micro Logix Expansion I/O modules with 256discrete I/O . As many as six embedded 100 kHz high-speedcounters (only on controllers with DC inputs) Two serial ports

Figure 2. Block Diagram.

12


with DF1, DH-485, Modbus RTU, DNP3 and ASCII protocolsupport.

Proximity Sensors: Proximity Sensors are available in two typesnamely;1) Inductive sensors2) Capacitive Sensors.

Inductive Sensors are cheaper and allow detection of metalobjects whereas capacitive sensors are costly and allowdetection of metal, plastic and glass objects as well.

VI. VARIABLE FREQUENCY DRIVEWhen an induction motor starts, it will draw very high inrushcurrent due to the absence of the back EMF at start. Thisresults in higher power loss in the transmission line and also inthe rotor, which will eventually heat up and may fail due toinsulation failure. The high inrush current may cause thevoltage to dip in the supply line, which may affect theperformance of other utility equipment connected on the samesupply line.

Adding a variable frequency drive (VFD) to a motor-drivensystem can offer potential energy savings in a system in whichthe loads vary with time. VFDs belong to a group of equipmentcalled adjustable speed drives or variable speed drives.

that have a safety factor. This often leads to energy inefficiencyin systems that operate for extended periods at reduced load.The ability to adjust motor speed enables closer matching ofmotor output to load and often results in energy savings

The components of the drive system are broken into four majorcategories: source power, rectifier, dc bus, and inverter. Othercomponents exits such as resolver and encoder feedbackdevices, tachometers, sensors, relays and help supplementthe system.

First, the source power must be converted from alternatingcurrent to direct current. This conversion is accomplished bymeans of a rectifier; a diode is used for more intelligentrectification. The power source that was 460volts ac, 60 Hertznow converted to 650 volts dc. This AC to DC conversion isnecessary before the power can be changed back to AC at avariable frequency.

The diode bridge converter that converts AC-to-DC issometimes just referred to as a converter. The converter thatconverts the dc back to ac is also a converter, but to distinguishit from the diode converter, it is usually referred to as aninverter. It has become common in the industry to refer toany DC-to-AC converter as an inverter.

(Variable speed drives can be electrical or mechanical, whereasVFDs are electrical.) The operating speed of a motor connectedto a VFD is varied by changing the frequency of the motorsupply voltage. This allows continuous process speed control.Motor-driven systems are often designed to handle peak loads

Figure 3. Block diagram of VFD ac to dc converter.

VII. METHODOLOGYThis department of the plant works on distribution of any kindof chemical liquid into different tank to a main two bufferstorage tank. This distribution takes place automatically usingthe Programmable Logic Controller (PLC).

13


It commonly applied application of PLC where five differentchemical liquids are mixed in required proportion to form abatch .Rate of the flow is already fixed. We only control thetime of the flow. Level of the liquids in the tank is sensed bythe level sensor switches. The ratio of five different liquidswill decided as per the required mixed liquid that we needed inthe bottle. There is a stirrer motor is also fitted to mix these twoliquids of definite amount in the main tank.

Bottles are kept in position in a carton over conveyor belt;they are sensed to detect their presence. Capacitive sensorsare used for sensing the bottles. Depending on the output ofthe sensor the corresponding valve switch on and filling

Figure 4. Screen shot of SCADA software of manufacturing department of the chemical liquid.

Figure 5. Screen shot of the SCADA software of the automatically filling of the chemical liquid into the bottles.

operation takes place. If the particular bottle is not presentthen the valve in that position is switched off, thereby avoidingwastage of the liquid. The filling process is done based ontiming. Depending on the preset value of the timer the valve isswitched on for that particular period of time and the filling isdone.

The motor continues to run even when the bottle moves awayfrom the first sensors range, i.e. the output of the motor islatched as explained in the ladder logic section of PLC. Whensensor 2 senses the bottle, it also gives a high output to the

14


PLC. The PLC instructs the inverter to stop the motor. Thehigh output bit of sensor 2 is also given to the timer for thesolenoid valve. The timer used is TON. It counts for a predefinedvalue of time (18 sec). It gives two outputs, Enable output anddone output. The Enable output remains high while the timeris counting and the output goes high after the timer has finishedcounting. The Enable output of TON is given to the solenoidvalve, and so the solenoid valve is open for the predefinedvalue of time (18 sec). The Done output bit is used to turn ONthe motor again in the running . And this all the process arerepeat again and half the bottles fill again to in front of thesecond chemical tank and the bottles full filled and the doneoutput bit is used to turn on the induction motor again. Allthis are described in this ladder programming of the PLC.

VIII. PLC LADDER PROGRAM

IX. CONCLUSIONThis paper presents a automated liquid filling to bottles ofusing PLC and SCADA. A total control is made in a filling isachieved. The present system will provides a great deal ofapplications in the field of automation, especially in massproduction industries where there are large number ofcomponents to be processed and handled in a short period oftime and theres need for increased production. Theprogramming to this system developed is flexible, quickly andeasily. This will increase the total production output; thisincrease in production can yield significant financial benefitsand savings. This concept can be used in beverage and foodindustries, milk industries, medicine industries, mineral water,chemical product industries and manufacturing industries. The

Figure 6. PLC ladder program for filling of the chemical liquid into the bottles.

15


present work is motivated to develop an online scheme tomonitor and control a hybrid method of automatically filling ofthe chemical liquid into the bottles using PLC and SCADA.

X. REFERENCES[1]. Sager P. Jain, Dr.Sanjay l.Haridas ,energy efficient atomized

bottling plant using plc and SCADA with speed variableconveyor assembly (Iosr Journal Of Electronic AndCommunication Engineering, e-ISSN:2278-2834, p- ISSN8735.

[2]. Mrs. Jignesha, Air design and Development of PLC andSCADA Based Control Panel For Continuous Monitoring Of3-Phase Induction Motor.

[3]. K.Gowri Shankar Control of Boiler Operation Using PLC-SCADA International Multiconference of Engineers andComputers Scientists 2008 Volume 2 imecs 2008, 19-21 March,2008, Hong Kong.

[4]. N. Shaukat, PLC Based Automatic Liquid Filling Process,Multitopic Conference, 2002, IEEE publication.

[5]. T. Kalasiselvi and R. Praveena PLC Based Automatic BottleFilling and Capping System with user Defined VolumeSelection, International Journal of Emerging Technology AndAdvanced Engineering, Volume 2, Issue 8, August 2012).

[6]. Arvind N. Nakiya, Mahesh A. Makwana An Overview of aContinuous Monitoring and Control System for 3-PhaseInduction Motor Based on Programming Logic Controller andSCADA Technology. (IJEET volume 4 , issue 4, July August(2013).

[7]. V. Mathavi & Dhivya Static Application Panels (SAP)Controlled by PLC(International Journal of Applied

Information Systems (IJAIS)-ISSN: 2249-0868 Foundationof Computer Science FCS, New York, USA Volume 5No 6,April 2013.

[8]. Hemant Ahuja and Arika Singh Automatic Filling ManagementSystem For Industries, International Journal of EmergingTechnology and Advanced Engineering,Volume-4, special issue,February 2014.

[9]. Mahesh Nandaniya( Automatic Canal Gate Control of 3-Induction Motor with PLC and VFD, Powered by Solar Systemand Monitoring International Journal of Emerging Trends inElectrical and Electronics (IJETEE) , Vol. 1, Issue 1, March2013.

[10]. Sujith John Mathew and B.Hemalatha, Fault Identificationand Protection of Induction Motor using PLC and SCADAinternational journal of advanced research in electrical, electronicand instrumention engineering vol3, issue 4, April 2014.

[11]. Pampashree and Md. Fakhruddin Ansari Design andImplementation of Scada Based Induction Motor Control.Journal of Engineering Research and Applicationswww.ijera.com ISSN : 2248-9622, Vol. 4, Issue 3 Version 5,March 2014, pp.05-1.

[12]. Ahmed Ullah Abu Saeed & Md. Al-Mamun IndustrialApplication of PLCs in Bangladesh, International Journal ofScientific & Engineering Research, Volume 3, Issue 6, June-2012, ISSN 2229-5518.

[13]. V. Vikash and H.J. Amarendra, PLC Based Sensor OperatedObstacle Detection Carriage Vehicle, International Journal ofApplied Engineering Research ISSN 0973-4562 Volume 9,Number 7 (2014) pp. 847-851.

16


Friction Stir Welding: Tool Materialand Geometry

Abstract -- The Friction stir welding is a dynamically developingversion of pressure welding processes by which High-quality weldscan be created. The mixing the material flow conditionsspecifically affect the quality of the weld, so the Tool geometry isvery important. Tool design and selection of process variablesare critical issues in the usage of FSW process. The Developmentof cost effective and durable tools, which lead to structurallysound welds, is still awaited. Material selection and designintensely affect the performance of the tools. Here we reviewedseveral important aspects of FSW tools such as tool materialselection & its importance, geometry and load bearing abilityand process economics for applications in this article.

Keywords: Friction Stir Welding; Tool Material; Tool Geometry;Rotational Speed.

I. INTRODUCTIONFRICTION stir welding is a solid state joining process using arotating tool moving along the joint interface, generating heatand resulting in a re-circulating plasticized material flow nearthe tool surface. This plasticized material is subjected toextrusion by the tool probe rotational and linear movementsleading to the formation of stir zone. This stir zone formation isaffected by the material flow behavior under the action ofrotating tool. It was developed in England by The WeldingInstitute (TWI) in 1991 [1]. The friction stirring tool consists ofa pin, or probe, and a shoulder as shown in Fig.1. Contact ofthe pin with the workpiece creates frictional and deformationalheating and softens the workpiece material; contacting theshoulder to the workpiece increases the workpiece heating,expands the zone of softened material, and constrains thedeformed material. Naturally, there are important effects to thetool during welding: abrasive wear, high temperature anddynamic effects. Therefore, the good tool materials have thefollowing properties: good wear resistance, high temperaturestrength, temper resistance, and good toughness.

So there are two important aspects of friction stir welding tooldesign: tool material and geometry [2]. Most importantChallenges of Friction Stir Welding are application of hightemperature materials, Tool material selection, Development ofTool Materials, Tool design and Complex geometries anddissimilar materials.

A. Chandrashekar, B. S. Ajay Kumar and H. N. ReddappaDepartment of Mechanical Engineering, Bangalore Institute of Technology, Bangalore, India.

[email protected]

Figure 1. Schematic drawing of friction stir welding.

II. INFLUENCE OF TOOL MATERIALAND GEOMETRY ON WELD QUALITY

The tool of FSW is composed of two parts: a tool body and aprobe. The tool technology is the heart of friction stir weldingprocess. The tool shape determines the heating, plastic flowand forging pattern of the plastic weld metal. The tool shapedetermines the weld size, welding speed and tool strength.The tool material determines the rate of friction heating, toolstrength and working temperature, the latter ultimatelydetermines which materials can be friction stir welded [3]. Twodifferent tool pin geometries (square and hexagonal) and threedifferent process variables, i.e. rotational speeds and weldingspeeds were selected for the experimental investigation ofAA6101-T6 alloy. It was observed that square pin profile gavebetter weld quality than the other profile. Besides, the electricalconductivity of the material was maintained up to 95% of thebase metal after welding. Arora et al [4] proposed and tested acriterion for the design of a tool shoulder diameter (consideredthree Shoulder diameters 15, 18, & 21mm) based on the principleof maximum utilization of supplied torque for traction.

The optimum tool shoulder diameter computed from thisprinciple using a numerical heat transfer and material flow modelresulted in best weld metal strength in independent tests andpeak temperatures that are well within the commonlyencountered range. The optimum shoulder diameter of 18 mmat 1200 rpm has resulted in superior tensile properties inindependent tests. Elangovan and Balasubramanian [5] havealso reported that the tool with an 18 mm shoulder diameter

17

FRICTION STIR WELDING

provided the best weld joint strength at a rotational speed of1200 rpm, as shown in Table 1.

TABLE I- THE MECHANICAL PROPERTIES OF WELDSMADE USING A CYLINDRICAL PIN PROFILE [4]

TABLE II - WELDING TEMPERATURE RANGEOF VARIOUS ALLOYS [5]

TABLE III - TOOL MATERIALS USED INFSW FOR SOFT ALLOYS [5]

Materials such as aluminium or magnesium alloys, andaluminium matrix composites (AMCs) are commonly weldedusing steel tools. Steel tools have also been used for the joiningof dissimilar materials in both lap and butt configurations. Toolwear during welding of metal matrix composites is greater whencompared with welding of soft alloys due to the presence ofhard, abrasive phases in the composites. Total wear was foundto increase with rotational speed and decrease at lower traversespeed, which suggests that process parameters can be adjustedto increase tool life [6]. Lakshman Rao et al [7] highlight therole of tool geometry in their investigation, because toolgeometry plays a major role in FSW. Proper selection of a tool

material and shape of the pin reduces number of trials andtooling cost. In addition this study also highlights the weareffect due to friction between sliding surfaces. The effect ofFriction Stir Welding process parameters on the mechanicalproperties of the AA 2014-T6 alloy joints produced by frictionstir welding have been discussed by Vagh and Pandya [8].Effects of tool design, tool rotation speed & tool travels speedon mechanical properties have been analysed using Taguchiorthogonal array design of experiments technique. There arethree different tool rotation speeds (1000, 1400 & 2000 rpm)and three different tool traverse speeds (14, 20, 28 mm/min).For each combination of tool rotation speeds and tool traversespeeds three different types of tool pin profiles (threadedcylindrical pin, Stepped pin and Threaded cone pin) have beenused. The study indicates that Tool design is the main processparameter that has the highest statistical influence onmechanical properties.

Figure 2. Different FSW tool geometries used in the experiment [8].

Prasanna et al [10] studied the effect of four different tool pinprofiles on mechanical properties of AA 6061 aluminum alloy.Four different profiles have been used to fabricate the buttjoints by keeping constant process parameters of tool rotationalspeed 1200rpm, welding speed 14mm/min and an axial force7kN. Different heat treatment methods like annealing,normalizing and quenching have been applied on the jointsand evaluation of the mechanical properties like tensilestrength, percentage of elongation, hardness andmicrostructure in the friction stirring formation zone areevaluated. Of the four tool profiles, the maximum tensile strengthand % of elongation of 210 MPa and 20.9 respectively wasobserved on Hexagonal pin profile tool with annealing process.The tensile strength and percent of elongation of the hexagonaltool profile with annealing process has reached about 90 %and 80 % respectively of the parent metal. Lee et al [11] weldedAlMg alloy with low carbon steel in lap joint configurationusing tool steel as tool material without its excessive wear byplacing the softer AlMg alloy on top of the steel plate andavoiding direct contact of the tool with the steel plate. Tungstenbased alloys have also been used for the welding of both lowand high melting point alloys [12]. For example, Edwards and

18


Ramulu [13] used a WLa alloy tool to study FSW of Ti6Al4V alloy. Tools made of a tungsten alloy Densimet (compositionnot reported) were used by Yadava et al [14] to weld AA 6111-T4 aluminium alloy. Table 4 shows properties of tool materials.

to the larger contact area. With the increase in shoulder diameterthe total torque increases continuously even when the stickingtorque decreases for large shoulder diameters.

Figure 3. Total torque required during FSW of AA6061 as a functionof the tool shoulder diameter for rotational speeds

of 900, 1200 & 1500 rpm [3].

TABLE 4 - PROPERTIES OF COMMON TOOL MATERIALS [11]

Tool geometryTool geometry affects the heat generation rate, traverse force,torque and the thermo-mechanical environment experiencedby the tool. The flow of plasticised material in the workpiece isaffected by the tool geometry as well as the linear and rotationalmotion of the tool. Important factors are shoulder diameter,shoulder surface angle, pin geometry including its shape andsize, and the nature of tool surfaces [12]. It was also observedfrom the previous data that the friction stir weld tool geometryhas a significant effect on the weldment reinforcement, microhardness, and weld strength.

Shoulder diameterIn order to determine the optimum tool geometry, the twocomponents of the torque are plotted in Fig. 4 for variousshoulder diameters. As the shoulder diameter increases, thesticking torque, increases, reaches a maximum and thendecreases [4]. This behavior, which shows that two main factorsaffect the value of the sticking torque. First, the strength of thematerial, shear stress decreases with increasing temperaturedue to an increase in the shoulder diameter. Second, the area

over which the torque is applied increases with shoulderdiameter. As a result, the product of these two componentsshows the trend indicated in the Fig.3. The sliding torque,increases continuously with increasing shoulder diameter due

19

FRICTION STIR WELDING

Pin (probe) geometryFriction stirring pins produce deformational and frictionalheating to the joint surfaces. The pin is designed to disruptthe faying, or contacting surfaces of the work piece, shearmaterial in front of the tool, and move material behind the tool.In addition, the depth of deformation and tool travel speed aregoverned by the pin design [3].

Figure 4. The computed values of sticking, sliding and total torque for various shoulder diameters at 1200rpm [3].

Tool costWhile the energy cost for the FSW of aluminium alloys issignificantly lower than that for the fusion welding processes[25] the process is not cost effective for the FSW of hardalloys. Tools made of pcBN are often used for the welding ofhard materials. However, pcBN is expensive due to hightemperatures and pressures required in its manufacture [12].Santella et al [22] did an approximate cost benefit analysis forFSW with a pcBN tool versus resistance spot welding (RSW)of DP 780 steel. The equipment and utility costs for FSSWwere assumed to be 90 and 30% respectively of the costs inRSW; however, they did not report the dollar amounts of thesecosts. They further assumed that a typical RSW tool tip lasts5000 welds and costs $0.65 per tip [12]. Considering the costsinvolved with equipment, utility and the tool, they estimatedthat in order for the FSSW to be cost competitive with respectto RSW, each FSSW tool, costing ~$100, needs to make 26 000spot welds. Since the cost of each pcBN tool was significantlygreater than $100 and typical tool life was between 500 and1000 welds, they suggested lowering tool costs as an importantneed. Feng et al [24] produced over 100 friction stir spot weldson dual phase steel (ultimate tensile strength 600 MPa) andmartensitic steel (ultimate tensile strength 1310 MPa) withoutnoticeable degradation of the pcBN tool. The costs of Si3N4and TiB2 tools were less than 25% of the cost of pcBN tools[22]. Tools of WRe or WLa alloys are relatively lessexpensive than that of pcBN tool but suffer considerably more

wear compared with super abrasives due to their relativelylower high temperature strength and hardness [12]. Mohantyet al [9] investigated the effects of different friction stir weldingtool geometries on mechanical strength and the microstructureproperties of aluminum alloy welds. Three distinct toolgeometries with different types of shoulder and tool probeprofiles were used in the investigation according to the designmatrix.

The effects of each tool shoulder and probe geometry on theweld was evaluated.

Figure 5. Micro hardness profile for various tool geometry [8].

The micro hardness of weld nugget TMAZ obtained withdifferent tool profiles is shown in Fig. 5. It is observed that theweld nugget exhibits a higher micro hardness compared to thethermo-mechanically affected zone (TMAZ) and the base metal[9].

III. CONCLUDING REMARKSThe joints of different tool pin profiles like straight cylindrical,Taper cylindrical, triangular, square, trepezoidal and hexagonaltool etc., with different rotational speeds, weld speeds andaxial force were reviewed in this paper. The following importantconclusions were made: Based on the literature survey, Toolshoulder-to-pin diameter ratios play an important role in stirzone development. The diameter of the pin is equal to thethickness of the parts to be welded and its length is slightlyshorter than the thickness of the part. Tool material propertiessuch as strength, fracture toughness, hardness, thermalconductivity and thermal expansion coefficient affect the weldquality, tool wear and performance. Heat generation rate andplastic flow in the workpiece are affected by the shape and sizeof the tool shoulder and pin. Although the tool design affectsweld properties, defects and the forces on the tool. The pincross-sectional geometry and surface features such as threadsinfluence the heat generation rates, axial forces on the tool andmaterial flow. Tool wear, deformation and failure are also muchmore prominent in the tool pin compared with the tool shoulder.There is a need for concerted research efforts towards

20


development of cost effective durable tools for commercialapplication of FSW to hard engineering alloys.References

[1]. L.V. Kamble, S.N. Soman, P.K. Brahmankar, Effect of ToolDesign and Process Variables on Mechanical Properties andMicrostructure of AA6101-T6 Alloy Welded by Friction StirWelding, IOSR Journal of Mechanical and Civil Engineering(IOSR-JMCE) ISSN(e) : 2278-1684, ISSN(p) : 2320334X,Pp.30-35.

[2]. Akos meilinger and imre torok, The importance of friction stirwelding Tool, Production Processes and Systems, vol. 6, 2013,no. 1, Pp. 25-34.

[3]. H. K. Mohanty, M. M. Mahapatra, P. Kumar, P. Biswas andN. R. Mandal, Effect of Tool Shoulder and Pin Probe Profileson Friction Stirred Aluminum Welds a Comparative Study,J. Marine Sci. Appl., 11: 200-207, DOI: 10.1007/s11804-012-1123-4, 2012, Pp. 200-207.

[4]. A. Arora, A. De and T. Deb Roy, Toward optimum frictionstir welding tool shoulder diameter, Acta Materialia Inc.Elsevier, doi:10.1016/j.scriptamat.2010.08.052, 2010, Pp-9-12.

[5]. K. Elangovan and V. Balasubramanian, Mater. Des. 29, 2008,Pp.362.

[6]. S.K. Selvam and T. Parameshwaran Pillai, Analysis of HeavyAlloy Tool in Friction Stir Welding, International Journal ofChemTech Research CODEN (USA): IJCRGG ISSN: 0974-4290, Vol.5, No.3, 2013, p 1346-1358.

[7]. M. Lakshman Rao, P. Suresh Babu, T. Rammohan And Y.Seenaaiah, Study of Tool Geometry In Friction Stir WeldingApplications, AKGEC International Journal Of Technology,Vol. 3, No. 2, Pp.15-18.

[8]. A.S Vagh and S. N. Pandya, Influence Of Process ParametersOn The Mechanical Properties Of Friction Stir Welded AA2014-T6 Alloy Using Taguchi Orthogonal Array, InternationalJournal of Engineering Sciences & Emerging Technologies,ISSN: 2231 6604 Volume 2, Issue 1, 2012, Pp. 51-58.

[9]. H. K. Mohanty, M. M. Mahapatra, P. Kumar, P. Biswas andN. R. Mandal, Effect of Tool Shoulder and Pin Probe Profileson Friction Stirred Aluminum Welds a Comparative Study,J. Marine Sci. Appl., 11: 200-207, DOI: 10.1007/s11804-012-1123-4, 2012, Pp. 200-207.

[10]. P. Prasanna, Ch. Penchalayya and D. Anandamohana Rao,Effect Of Tool Pin Profiles And Heat Treatment Process InThe Friction Stir Welding Of AA 6061 Aluminium Alloy,American Journal of Engineering Research (AJER) e-ISSN:2320-0847 p-ISSN : 2320-0936 Volume-02, Issue-01, 2013,Pp.07-15.

[11]. Y. Lee, D. H. Choi, Y. M. Yeon and S. B. Jung, Dissimilarfriction stir spot welding of low carbon steel and AlMg alloyby formation of IMCs, Sci. Technol. Weld. Join., 14, (3), 2009,Pp.216220.

[12]. R. Rai, A. De, H. Bhadeshia and T. Deb Roy, Review: frictionstir welding tools, Institute of Materials, Minerals and Mining,Science and Technology of Welding and Joining, Vol.16, No.4,2011, Pp- 325-342.

[13]. P. Edwards and M. Ramulu, Effect of process conditions onsuper plastic forming behaviour in Ti6Al4V friction stirwelds, Sci. Technol. Weld. Join.,14, (7), 2009, Pp.669680.

[14]. M. K. Yadava, R. S. Mishra, Y. L. Chen, B. Carlson and G. J.Grant, Study of friction stir joining of thin aluminium sheetsin lap joint configuration, Sci. Technol. Weld. Join., 15, (1),2010, Pp. 7075.

[15]. C. Meran, V. Kovan and A. Alptekin, Friction stir welding ofAISI 304 austenitic stainless steel, Materialwiss.Werkstofftech., 38, 2007, Pp.829835.

[16]. W. Gan, Z. T. Li and S. Khurana, Tool materials selection forfriction stir welding of L80 steel, Sci. Technol. Weld. Join., 12,(7), 2007, Pp.610613.

[17]. B. K. Jasthi, W. J. Arbegast and S. M. Howard, Thermalexpansion coefficient and mechanical properties of frictionstir welded invar (Fe36%Ni), J. Mater. Eng. Perform, 18,(7), 2009, Pp. 925934.

[18]. E. A. Brandes and G. B. Brook, Smithells metals referencebook, 1992, Oxford, Butterworth Heinemann.

[19]. J. F. Shackelford and W. Alexander, CRC materials science andengineering handbook, Boca Raton, Florida, 2001, CRC Press.

[20]. A. de Pablos, M. I. Osendi and P. Miranzo, Effect ofmicrostructure on the thermal conductivity of hot-pressedsilicon nitride materials, J. Am. Ceram. Soc., 85, (1), 2002,Pp. 200 206.

[21]. J. Z. Jiang, H. Lindelov, L. Gerward, K. Stahl, J. M. Recio, P.Mori-Sanchez, S. Carlson, M. Mezouar, E. Dooryhee, A. Fitchand D. J. Frost, Compressibility and thermal expansion ofcubic silicon nitride, Phys. Rev. B, 2002, 65B, 161202.

[22]. M. Santella, Y. Hovanski, A. Frederick, G. Grant and M. Dahl,Friction stir spot welding of DP780 carbon steel, Sci. Technol.Weld. Join., 2010, 15, (4), 271278.

[23]. G. Grant, Y. Hovanski and M. Santella, Friction stir spotwelding of advanced high strength steels, Oral presentation,Proc. DOE Hydrogen Program and Vehicle TechnologiesProgram Annual Merit Review and Peer Evaluation Meeting,Arlington, VA, May 2009, DOE.

[24]. Z. Feng, M. L. Santella, S. A. David, R. J. Steel, S. M. Packer,T. Pan, M. Kuo and R. S.Bhatnagar, Friction stir spot weldingof advanced high-strength steels a feasibility study, SAEtechnical paper 2005-01-1248, SAE International, Warrendale,PA, USA, 2005.

[25]. R. Hancock, Friction welding of aluminum cuts energy costby 99%, Weld. J., 2004, 83, 40.

21

DETERMINATION OF NATURAL FREQUENCY

Determination of Natural Frequency of a TurningSpecimen Subjected to Random Excitation

Abstract -- This paper deals with determination of naturalfrequency of a turning specimen subjected to random cuttingand thrust force. In this paper, a machine tool system is modeledas a cross-coupled linear system. The natural frequency of aturning specimen subjected to random excitation can bedetermined by using Volterra series. The Volterra seriesrepresents the response of a system in a functional form, througha series of first and higher order convolution integrals, involvingexplicit operations on the input to the system. Exploring theinput identification procedure for the estimation of excitationforces from the knowledge of system parameters and response oflinear cross-coupled system having cross-coupling in stiffness aswell as damping. Modeling is done for a machine turning toolsystem which is being excited by random forces. The turningspecimen can be modeled as a two-degree-of freedom systemwith both direct as well as cross coupling effect has to beconsidered in linear stiffness and damping terms. The equationof motion has been written in a non-dimensional form.Illustration of procedure is done through numerical simulation.The assumption involved and the approximations are alsodiscussed. The procedure for identification of response andconsecutively natural frequency is illustrated through numericalsimulation using FORTRAN language.

Keywords: Turning Specimen, Machine Tool System, Random Excitation,Volterra Series, Cross Coupling Effect.

I. INTRODUCTIONTHE AIM of this paper is to develop an algorithm that can beused to determine the natural frequency by knowing theresponse and the system parameter using Volterra series. Suchanalysis have attracted considerable attention in the recentpast, partly due to growing awareness of the significance ofthe Random nature of forces produced by during operation.Indirect estimation of excitation force using model co-ordinatetransformation has been carried out by De Sanghere et.al. [1],classification of different force identification problem has beencarried out by Stevens [2]. A study for finding inverse methodfor estimation of impulsive loads has been conducted by Maet.al. [3]. A non-linear vibration problem of estimating theexternal forces for a single degree of freedom system usingconjugate gradient method has been developed by Huang [4-5].

M.Z. Hussain1, Dr. A. A. Khan2 and Dr. M. Suhaib31,3 Department of Mechanical Engineering, Jamia Millia Islamia, New Delhi 110025 India

2Department of Mechanical Engineering, Aligarh Muslim University, Aligarh 202002 [email protected], [email protected], [email protected]

Chatter in turning or cutting is usually assumed to be aregenerative process (there exist also other explanations likethermoplastic changes in the material). Because of externaldisturbance, work piece starts an oscillation relative to thetool, producing a wavy surface. Therefore, the chip thicknessthat has to be cut in the next round will also vary randomly.Since random cutting force depends on chip thicknessfluctuation. In processes like cutting and turning the vibrationsthat produce the random cutting force are orthogonal to therotation. But there would be a random input force in thedirection of rotation to generate chatter. In this work the inputrandom force is assumed to have white noise type of powerspectral density. Such white noise analysis is considered to bean effective tool for gaining a maximum of information with aminimum number of assumptions about the system.

In the present, analysis a two-degree of freedom turning toolsystem has been modeled having cross-coupled linear dampingand stiffness parameters. The equations of motion have beenderived from the configuration of the system. These equationsare then non-dimensionalised. Laplace transforms areemployed to derive expressions for the first order direct andcross-Kernel transforms from the non-dimensionalisedequation of motion. These first order direct and cross-Kernelsare obtained in the frequency domain, from the knowledge ofthe system parameters, which are then used to estimate theexcitation force.

II. LITERATURE REVIEWThe forward analysis of response generation of a multi degreeof freedom system subjected to random excitation force hasbeen the major thrust area in the recent past. The completeinverse analysis of determination of the forcing function is acomplex phenomenon and requires specialized techniquesparticularly when the excitation is random in nature.

The methods that are being used nowadays either involves adetailed knowledge of the material characteristics of the teststructures (finite element model approach) Dobling & Farrar[7] or make very restrictive assumption about the excitationRandall & Swevers [8]. Recently, a novel approach waspresented for the normalization of operational mode shapes

22


on a basis of in operational model models only Parloo &Guillaume [9]. In this method it is shown that the operationalmode shapes can be normalized by means of the measuredshift in natural frequencies between the original and massloaded condition just by adding or removing, for instance one(or more) masses (with well known weights) to the teststructures. A complete modal model can be reconstructed bynormalizing the operational mode shapes. Using this model,an inverse problem can be formulated for the identification ofthe unknown forces that gives rise to the measured responses.In reference Guillaume, Verboven & Sitter [10] the forcedidentification problem was addressed and an inverse solverwas proposed and is then compared to classical approach (e.g.pseudo-inverse). For force identification on the basis of outputonly data an experimental work has been carried out by E Parlooet al. [13].

In this paper the inverse problem of input identification is basedon the Volterra approaches, and is discussed in the context ofmachine tool system. Problem involving vibration arises quitefrequently in turning process, especially those involvingrandom vibration of sliding contact element of machine toolsystem due to varying chip thickness. Modeling of turningprocess has been done by Wang and Su [17] and theydeveloped the relationship between the cutting force and chipthickness fluctuation will be treated as a hysteresis model.Wang and Su [17] have modeled the turning process as a twodegree of freedom system.

The Volterra kernels extraction is proposed for the estimationof fourteen-linearised machine tool parameter subjected torandom cutting force by Khan and Khan [12]. The proceduredeveloped gives very good engineering estimates of themachine tool parameters. It can be shown that the large samplesize is preferable for obtaining sharp peaks which gives moreaccurate results. It has been also found that the proceduredeveloped is robust to the measurement noise and can besuccessfully employed for modeling of the setup.

III. MACHINE TOOL SYSTEMProblems involving vibration arise quite frequently in machinetool application, especially those involving random vibrationsof cutting tools and tool chatter, excited by random cuttingforce. In some cases, deterministic models prove to beinadequate or at least extremely complex and the phenomenoncan be adequately described only within the framework ofstatistical models. Statistical dynamics concerned with thestudy of various random phenomena in dynamic systemsenriches the classical basic theory of oscillations and extendsthe possibilities for its applications to the description andanalysis of real response processes in dynamic systems.Inverse problems in vibration analysis require techniques withrigorous theoretical base, which provide valid routes to inputidentification. Further the estimation procedure becomes more

complex if the excitation is random in nature. The Volterra [6]series provides a basis for these requirements. The basic theoryof Volterra series involves modeling the relationship betweenthe system response and input in terms of a series of first andhigher order convolution integrals. It employs multi-dimensional kernels, which upon convolution with the appliedexcitation express the response in the form of a power series.The kernels of the system are considered as multi-dimensionalunit impulse response functions.

IV. GOVERNING EQUATION OF MOTIONA cutting tool used for plain turning operation can berepresented as shown in Fig.4.1, where the flexibility isrepresented by eight springs & damping coefficients isdiscussed earlier. The excitation forces are denoted as thethrust and main cutting forces exciting the structure.

The equations of motion in the vertical and horizontal directionrelating the displacement to the forces applied to it is given by

Figure 1. A plain turning tool system with cross-couplings.

1( )xx xy xx xymx c x c y k x k y F t+ + + + =&& & & (1)2 ( )yx yy yx yymy c x c y k x k y F t+ + + + =&& & & (2)

where m is the mass, ,xx yyc c are direct linear damping terms; are cross-coupled linear damping coefficients and are the direct linear stiffness terms, while are

the cross-coupled linear stiffness term. , representsthe excitation forces given to the system in x and y direction inthe above equationIn order to write the equation of motion in a non-dimensionalform, let us define

23

DETERMINATION OF NATURAL FREQUENCY

V. COMPUTER SIMULATIONThe input identification procedure is illustrated throughnumerical simulation of the response for the non-dimensionalcoupled equation (3) and (4). The forcing functions chosen forresponse simulation are normalized zero mean random forces,

1( )f

and

2 ( )f

. The excitation forces are simulated throughrandom number generating subroutines and are normalizedwith respect to their maximum values. The governing equationsare then numerically solved using a fourth order Runge-Kuttamethod, to obtain the responses in and directions (

x

and

y

)from equations (3) and (4). Using FFT the power spectrum ofthe response averaged over the ensemble of 2000 samples isdetermined.

Owing to the statistical nature of the problem, the procedure isillustrated for various sets of direct and coupled stiffness anddamping parameters. The case studied for a particular set ofstiffness as well as damping parameter have been designed tofind the natural frequency of the system using the algorithmdeveloped.

VI. RESULTS AND DISCUSSIONCase I:

For the above set of values of the parameters, the response iscomputationally simulated using equations (3) and (4). Thegoverning equations are then numerically solved using a fourthorder Runge-Kutta method, to obtain the responses in anddirections (

x

and

y

). Their corresponding power spectrumsaveraged over 2000 samples are shown in Fig. 5 and 6. Fromthe power spectrum of the responses the fundamentalfrequencies of the system are found to be as

1 0.094358 =

per

unit of non-dimensional frequency 2 =0.256031 and per unitof non-dimensional frequency these responses are fed as inputsto the input identification algorithm. The various direct and

cross-coupled first order kernels are obtained from theknowledge of the system parameters. Typical sample ofexcitation force used for simulation is shown in Fig. 2. Volterrakernel transforms exhibit the two fundamental frequencies ofthe system, as it is a two-degree of freedom system. Theexcitation force has been estimated using these kernels andresponse. Figure 3 shows the Power spectrum of input force.Figure 4 shows the FFT of the simulated excitation forceaveraged over the ensemble of 2000 samples and is used forboth x and y directions for the case. The correspondingestimated force can be seen in Fig 4.11 and can be comparedwith the simulated force as shown in Fig. 4. The error in theestimate that is the difference between the Fig.4 (simulatedforce) and Fig. 7 (estimated force) can be seen in Fig. 8.

Figure 2. Typical sample of normalized input force.

Figure 3. Power spectrum of input force.

Figure 4. Fast Fourier Transform of the simulated input

( )( )iF

.

24


Figure 5. Power spectrum of the response : Case 1.

Figure 6. Power spectrum of the response : Case 1.

Figure 7. Estimate of the input force F1(): Case I.

Figure 8. Estimate of the input force : Case I.

VII. CONCLUSION AND SCOPE FOR FUTURE WORKThe present work is primarily concerned with the determinationof two fundamental frequencies of cross-coupled turning toolsystems. Volterra theories have been employed for the analysisof the problem. A frequency domain approach has been adoptedin order to reduce the computation time. The coupled systemsconsidered are machine tool system. The work piece mountedon chuck with cross coupling in damping and stiffness hasbeen considered.

The input identification procedure in cross-coupled system isdeveloped in steps. This serves to illustrate general nature ofthe procedures adopted for identification of excitation force,which are random in nature. Few case studies have been carriedout with different sets of non-dimensional parameters in orderto check the accuracy of estimated forcing function. Theaccuracy of the estimates with various other non-dimensionalparameters and level of excitation can be expected to followthe same trends as discussed. The response of the system isexpressed through first order direct and cross Volterra kernels.These Volterra kernels are then processed to estimate theexcitation force. Reasonably good estimates are found fordifferent sets of both linear damping and stiffness parameters.

The excitation force are found to be estimated with a gooddegree of accuracy in given case, if the programme run forother cases also however the accuracy of the estimates arefound to vary with the values of non-dimensional parameterchosen for simulation. The present study can be used to designexperiments to choose appropriate set of excitation levels forthe expected sets of parameters. This work can be extended toinclude the kernels of higher order to increase the accuracy ofestimates. Further we are considering stiffness and dampingto be linear in order to keep the algebra simple, however anextension to the present work can be made by conside

Documents

AKG Int Journal Tech Vol 6 No 1