i journal of - ijsetijset.in/wp-content/uploads/2014/07/volume-2-issue-5-june-2014.pdf ·...

JUNE2014

PRINTVERSION

INTERNATIONALJOURNALOF

SCIENCE,ENGINEERINGAND

TECHNOLOGY

(IJSET)

ISSN:2348‐4098

PrintVersion,Volume02,Issue05

June2014Edition

IJSET

www.ijset.in

GENERALINFORMATION:InternationalJournalofScience,EngineeringandTechnologyPublishBi‐MonthlyJournalunderISSN:2348‐4098COPYRIGHTCopyright©2014IJSET.INAll the respective authors are the sole owner and responsible of published researchand research papers are published after full consent of respective author or co‐author(s).For any discussion on research subject or research matter, the reader shoulddirectlycontacttoundersignedauthors.AllRightsReserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem, or transmitted, in any form or by any means, electronic, mechanical,photocopying,recording,scanningorotherwise,exceptasdescribedbelow,withoutthepermissioninwritingofthePublisher.Copying of articles is not permitted except for personal and internal use, to theextent permitted bynational copyright law, or under the terms of a license issuedby the N ational Reproduction RightsOrganization.All thepublishedresearchcanbereferencedbyreaders/scholars/researchers intheirfurtherresearchwithpropercitationgiventooriginalauthors.DISCLAIMERStatementsandopinionsexpressedinthepublishedpapersarethoseoftheindividualcontributors and not the statements and opinion of IJSET. We assumes noresponsibilityorliabilityforanydamageorinjurytopersonsorpropertyarisingoutoftheuseofanymaterials,instructions,methodsorideascontainedherein.Weexpresslydisclaim any impliedwarrantiesofmerchantability orfitnessforaparticularpurpose.Ifexpertassistanceisrequired,theservicesofacompetentprofessionalpersonshouldbesought.CONTACTINFORMATIONEditor‐in‐Chief@mail:[email protected]:http://www.ijset.in

EDITORIALBOARD

Editor‐in‐Chief

Dr.KAVITASHARMA

DeputyEditor‐inChief

Prof.GUJARANANTKUMARJOTIRAM

AMGOI Vathar, Shivaji University

Kolhapur,Maharashtra,India

EditorialBoardMembers

Dr.MdEnamulHoque,

Associate Professor, University of

NottinghamMalaysiaCampus

Dr.S.KishoreReddy,

Professor,AdamaScience&Technology

University,Adama,Ethiopia

Prof.(Dr.)ShamamaAhmed,

Director, School of Engineering &

Technology, Noida International

University

Dr.DeborahOlorode,

UniversityofLagos,Nigeria

Prof(Dr.)SabrinaLuhibatto

Prof.JayaChatterjee

Dr.PatrizateTrovalusci

Prof.(Dr.)SyeedAbdel‐HamidEl‐Sayed

Ham

Prof.(Dr.)BurkhanTurkes

Dr.M.MasoodJamalKhan

Prof.(Dr.)MaittiJaffar

Prof.(Dr.)MdOmilAhmed

Dr.MahedAhmadi

Prof.(Dr.)ZaininulArifin

Dr.SalvotoreGalline

Prof.(Dr.)ArunK.Gupta

Dr.EmanSalah

Dr.XiyanJang

Dr.AlMaloom

Dr.AnilGupta,AssociateProfessor

Dr.PriyankaGupta,AssociateProfessor

Prof.MarieGeorge

Dr.PaoloDominoParshi

Dr.FransiscoABianchi

Dr.LingNyuyen

Dr.AshishKr.Sharma,

Principal&Professor

Dr.Inem,EciyiLoffi

Dr.DanielePellegrim

Dr.UmarUl‐Azam

Dr.ZathaeusUniOMOBADEGUN

Dr.MehdiHasan

Dr.MdMehdi,Emanul

Prof.(Dr.)RustamM.Mamool

Dr.EmanulHabibHoque

Dr.ChinChingSarn

Mr.SaurabhShukla,Scientist,DEFENCE

RESEARCH & DEVELOPMENT

ORGANISATION(DRDO)

Mr. Parvin Kumar, Scientist, DEFENCE


ORGANISATION(DRDO)

Dr. Sanjay Kumar, Scientist, DEFENCE


ORGANISATION(DRDO)

Dr. Raghvendra Kr. Mishra, Assistant

Professor

Mr. Narendra Singh Rathore, Director,

GlobalNutrition&Healthcare

Dr.SergieA.BukrayVinchi

DrNiroshinNirmal

Dr.LiChiKeonge

Prof.DeborahGranam

Dr.OmotoshoElice

Dr.AbrahamGoodnick

Dr.MdAbdulMandsour

Dr.Md.YasserSudElswin

Dr.AhmedAttaSobhi

Dr.MarieAngelinaLarrain

Dr.DawoodKabani

Prof.(Dr.)MohamedAudi

Prof.MohsinJamal,AssistantProfessor

Mr.RamBhool,AssistantProfessor

Mr.AshishKumar,AssistantProfessor

Dr.AyotundeOlalande

Dr.RajendraMuthoo

Dr.EnrichBattisa

Prof.(Dr.)BomieJ.Bakman

Dr.Kin‐HyungLi

Dr.YehKarKhengha

Dr.S.Balasubramanian

Dr.LaugCoradaz

Dr.DravidA.Pie

Dr.RobertManualMaitti

TABLEOFCONTENTS

S.No. MANUSCRIPTTITLEANDAUTHOR PageNo.

1. GOINGDRIVERLESSWITHSENSORS

GEETINDERKAUR, SOURABH JOSHI, JASPREET KAUR, SAMREET

KAUR

1‐8

2. PREDICTION OF CASTING DEFECTS THROUGH

ARTIFICIALNEURALNETWORK

GANESHG.PATIL,Dr.K.H.INAMDAR

9‐25

3. VERTICAL DISTRIBUTION AND ABUNDANCE OF SOIL

ACARINA IN A NATURAL FOREST AND JHUM LAND

ECOSYSTEMOFMOKOKCHUNG,NAGALAND

KRUOLALIETSURHO,BENDANGAO

26‐46

4. BANDWIDTHENHANCEMENTOFHIGHGAINANTENNA

USING CIRCULAR ARRAY OF SQUARE PARASITIC

PATCHES

BHAGYASHRIB.KALE,J.K.SINGH

47‐54

5. PERFORMANCE EVALUATION OF UNIVERSAL

DEHAZINGWITHDIRECTEDFILTERMETHOD

DINESHKUMARPATEL,AMITKUMARRAJPUT

55‐64

6. CYCLETIMEREDUCTIONOFGRINDINGPROCESSUSING

SIXSIGMAMETHODOLOGY

ALOKB.PATIL,Dr.KEDARH.INAMDAR

65‐79

7. PRODUCTIVITY IMPROVEMENT OF AUTOMOTIVE

INDUSTRYUSINGLEANMANUFACTURING

SWAPNILT.FIRAKE,Dr.KEDARH.INAMDAR

80‐101

8. AUTONOMOUSUNDERWATERROBOTUSINGFPGA

AKANKSHAGUPTA,PINKYGUPTA,KOUSHIKCHAKRABORTY

102‐108

9. EFFECT OF COMPETING CATIONS (Cu, Zn, Mn, Pb)

ADSORBEDBYNATURALZEOLITE

AFRODITAZENDELSKA,MIRJANAGOLOMEOVA

109‐118

GOINGDRIVERLESSWITHSENSORS

GEETINDERKAUR1,SOURABHJOSHI2,JASPREETKAUR3,SAMREETKAUR4

1,2,3,4ResearchScholar,DepartmentofComputerScienceandEngineering,CTInstituteofTechnology&Research,Jalandhar,India

E‐mail:[email protected],[email protected],[email protected],[email protected]

ABSTRACT

Thispaperexplorestheimpactthathasbeenworkingtowardsthegoalofvehiclesthat

can shoulder the entire burden of driving. Google driverless cars are designed to

operate safely and autonomously without requiring human intervention. They won’t

have a steering wheel, accelerator or a brake pedal because they don’t need them,

softwareandsensorsdoallthework.Ittakesyouwhereyouwanttogoatthepushofa

button.ThisTechnologysteptowardsimprovingroadsafetyandtransformingmobility

formillionsofpeople.

INDEX TERMS: Artificial Intelligence, Hardware Sensors, Google Maps, and Google

DriverlessCar.

1. INTRODUCTION

Itwasn’tthatlongagowhenroadmaps

may become extremely valuable as

Antiques. A couple of months ago a

Google CEO Larry Page drives in a car

around to pick up a friend of his. This

car has one special feature; there is no

driver at all. The car drove Larry’s

friendtwentymilestoGooglewithouta

driver. We will dream this about

decades.Alreadywehaveseenahostof

advancements to make safer drive like

Lane assists, parking assists or even

collision prevention assistance. With

more advance technologies that finds

greater emergence, future roadways

and become a mesh network along

autonomous vehicles. They share

information with each other and large

network speed, breaking and other

variables and move in a coordinated

formation. Here we are talking about

Google driverless car. A world with

increasingly connected climate, cars

take over, where humans are out of

equation.

VOLUME 2 ISSUE 5 JUNE 2014 EDITION

ISSN: 2348-4098

INTERNATIONAL JOURNAL OF SCIENCE, ENGINEERING AND TECHNOLOGY-www.ijset.in Page 1

2. AUTONOMOUSVEHIVCLE

An Autonomous vehicle (sometimes

referred as automated car or self‐

driving car) is a robotic vehicle that is

designed to fulfilling the transportation

capabilities without a human operator.

Qualifying to it as fully autonomous,

vehiclemustbeabletonavigatewithout

human input to the destination that is

predetermined over unadapted roads

andiscapabletosensetheenvironment.

Audi, BMW, Google, Ford are some of

the companies developing and testing

these vehicles. Technologies making a

system fully autonomous are Anti‐Lock

Brakes (ABS), Electronic Stability

Control (ESC), Cruise control, Lane

Departure Warning System, Self

Parking, Sensors, and Automated

GuidedVehicleSystems.

3. GOOGLE DRIVERLESS CAR

EXPLAINED

Only with occasional human

intervention, Google’s fleet of robotic

Toyota Cruises has logged more than

190,000 miles (approx. about 300,000

Km), driving in busy highways, in city

trafficandmountainousroads.Inanear

future their driverless car technology

could change the transportation.

Director of The Stanford Artificial

Intelligence Laboratory, Sebastian

Thrun guides the project of Google

DriverlessCar’swithelucidations:

Steering can be done by itself,

whilelookingoutforobstacles.

For corrections of speed limit, it

canacceleratebyitself.

OnanytrafficconditionitcanGO

orSTOPbyitself.

Figure1:GoogleDriverlessCar

4. UNDERTHEBONET

Itintegratesthreeconstituents:

GoogleMaps

HardwareSensors

ArtificialIntelligence

4.1GOOGLEMAPS

A self‐ driving computerized car has

unveiledbyGoogle;whichhasnowheel

for steering, brake or accelerator, just

hasbuttonstostart,stop,pulloveranda

computer screen to show the route.

Through GPS and Google maps to


ISSN: 2348-4098


navigate.AGooglemapprovidesthecar

with information of road and interacts

withGPStoactlikeadatabase.

4.2HARDWARESENSORS

Real time and dynamic Environmental

conditions (properties) attained by the

car. To need real time results, sensors

areattemptedtocreatefullyobservable

environment. These hardware sensors

are LIDAR, VEDIO CAMERA, POSITION

ESTIMATOR, DISTANCE SENSOR,

AERIALandCOMPUTER.

4.2.1LIDAR

(Light Detection And Ranging also

LADAR) is an optical remote sensing

technologywhichisusedtomeasurethe

distance of target with illumination to

light in the formof pulsed laser. It is a

laserrange finderalsoknownas“heart

of system”, mounted on the top of the

spoiler. A detailed #‐D map of the

environmentisgeneratedbythedevice

VELODYNE 64‐ beam Laser (for

autonomous ground vehicles and

marine vessels, a sensor named HDL‐

64E LIDAR is designed for obstacle

detection and navigation. Its scanning

distance is of 60 meters (~ 197 feet).

For 3D mobile data collection and

mapping application this sensor

becomes ideal for most demanding

perceptions due to its durability, very

highdata rates and360degree field of

view. One piece design patented the

HDL‐64E’suses64mounted lasers that

arefixedandeachof it ismountedtoa

specificverticalanglemechanicallywith

theentirespinningunit,tomeasurethe

environment surroundings. Reliability,

field of viewandpoint clouddensity is

dramatically increased by using this

approach.)

High resolution maps of the world are

combinedbythecarlasermeasurement

to produce different types of data

models that allows it to drive itself,

avoidingobstaclesandrespectingtraffic

laws. A LIDAR instrument consists of a

Laser, Scanner and a specialized GPS

receiver,principally.

Figure2:HDL‐64ELidar

HOWISLIDARDATACOLLECTED?

A beam of light is reflected by the

surface when it encounter with the


ISSN: 2348-4098


Laser that ispointedat the targetarea.

To measure the range, this reflected

light is recorded by a sensor. An

orientation data that is generated from

integrated GPS and Inertial

Measurement Unit System scan angles

andcalibrationwithposition.Theresult

obtainedisadense,and“pointcloud”(A

detail rich group of elevation points

consists of 3D spatial coordinates i.e.

Latitude,LongitudeandHeight).

4.2.2VIDEOCAMERA

A sensor that is positioned near to the

rear‐view mirror that detects the

upcoming traffic light. It performs the

same function as the mildly interested

human motorist performs. It reads the

read signs and keeps an eye out for

cyclists, other motorists and for

pedestrians.

4.2.3POSITIONESTIMATOR

An ultrasonic sensor also known as(

Wheel Encoder) mounted on the rear

wheels of vehicle, determines the

location and keep track of its

movements.Byusingthisinformationit

automatically update the position of

vehicleonGoogleMap.

4.2.4DISTANCESENSOR(RADAR)

Other sensors which include: four

radars,mountedonbothfrontandrear

bumpers are also carried by this

autonomousvehicle that allows the car

to “see” far enough to detect nearly or

upcoming cars or obstacles and deal

withfasttrafficonfreeways.

4.2.5AERIAL

A highly accurate positioning data is

demanded by a self – navigating car.

Readings from the car’s onboard

instruments (i.e. Altimeters,

Tachometers and Gyroscopes) are

combined with information received

fromGPSsatellitestomakesurethecar

knowsexactlywhereitis.

4.2.6COMPUTER

Car’s central computer holds all the

information that is fed from various

sensors so toanalyze thedata, steering

and acceleration and brakes are

adjusted accordingly. Not only traffic

laws,butalsotheunspokenassumption

of road users is needed to understand

bythecomputer.

4.3ARTIFICIALINTELLIGENCE

Artificial Intelligence provides the

autonomous car with real time

decisions. Data obtained from the

HardwareSensorsandGoogleMapsare


ISSN: 2348-4098


sent to A.I for determining the

acceleration i.e. how fast it is; when to

slowdown/stopandtosteerthewheel.

The main goal of A.I is to drive the

passenger safely and legally to his

destination.

5. WORKINGOFGOOGLECAR

Destination is set by “The Driver”

and software of car calculates a

routeandstartsonitsway.

LIDAR, a rotating, roof mounted

sensor monitors and scannes a

range of 60‐ meters around the

surroundings of car and creates

rudimentary detailed 3‐D map of

immediatearea.

An ultrasonic sensor mounted on

left rear wheel monitors

movementstodetectpositionofthe

carrelativeto3‐Dmap.

DISTANCE SENSORS mounted on

front and rear bumpers calculate

distancestoobstacles.

All the sensors are connected to

Artificial intelligence software in

the car and has input from Google

VIDEOCAMERASandstreetview.

ArtificialIntelligencestimulatesthe

real time decisions and human

perceptions o control actions such

asacceleration,steeringandbrakes.

The surface installed in the car

consults with Google Maps for

advance notification of things like

landmarks,trafficsignalsandlights.

To take control of the vehicle by

human is also allowed by override

function.

Figure3:HowitWorks

6. AN END TO TRAFFIC JAMS

FOREVER

Autonomous cars will be able to “talk”

to each other and navigate safely by

knowing where they are, by using

RADAR, CAMERAS, GPS, SENSORS and

WirelessTechnologyinrelationtoother

vehiclesandbymeanswithconnectivity

theycancommunicatewithobstaclelike

traffic signals. As a result traffic flow

becomes smoother; an end to traffic

jams and greater safety would be


ISSN: 2348-4098


achievedbyilluminatingthefrustration

and dangerous driving that’s often

triggeredbysittinginheavycongestion

for ages. When it comes to

sustainability, the self‐driving car also

holdsgreatpromiseby figuringout the

most–direct, least traffic jammedroute

by drivingwithout quickly accelerating

orbreakingtoohard,allwhich leadsto

savingonfuelconsumption.

Figure4:GoingDriverlessonroad

7. TRIALSANDTRIBULATIONS

Weseldomthinkabout ,whatneeds to

be happen behind the scenes to bring

thispotentiallylife‐changingtechnology

tothemarket,whileit’seasytogetlost

into it. Ahead of the Law is the major

problem to this technology, as

Lawmakers have a huge impact on

innovation. In the USmost federal and

stateautomobileLawsassumeahuman

operator. Before the technology can be

commercialized these need to be

repealed. To legalize the operation of

autonomous cars on the roads, Nevada

became the first state in 2012. An

attempttogainstatesupportforsimilar

changes in Law, Lobbyists fromGoogle

havebeentravellingaroundotherstates

and targeting Insurance companies as

well.The technologyalsoposesserious

puzzle to Insurance in terms of

RegulatoryissuesandLiability.

8. CONCLUSION

This paper explained about the Google

Driverlesscar revolutionwhichaimsat

the development of autonomous

vehiclesforeasytransportationwithout

a driver. For the economy, society and

individual business this autonomous

technology has brought many broad

implications.Carsthatdrivethemselves

willimprovereadsafety,fuelefficiency,

increase productivity and accessibility;

the driverless car technology helps to

minimize loss of control by improving

vehicle’s stabilityas thesearedesigned

tominimizeaccidentsbyaddressingone

ofthemaincausesofcollisions:Driving

error, distraction and drowsiness. But

stillthesecarshavealotofhurdlestogo

through before they became everyday

technology.


ISSN: 2348-4098


REFERENCES

[1]www.theneweconomy.com/insight/g

oogledriverless‐cars

[2] Jaemin Byun, Ki‐InNa, Myungchan

Noh,JooChanSohnandSunghoonKim;

ESTRO: Design and Development of

Intelligent Autonomous Vehicle for

ShuttleServiceintheESTRI.

[3] J. Markoff. (2010, October) Google

carsdrive themselves, in traffic.Online.

TheNewYorkTimes.

[4]KPMG(2012),Self‐DrivingCars:The

NextRevolution,KMPGandTheCenter

for Automotive Research; at

www.Kpmg.com/ca/en/Isuues‐And

Insight/Articles

Publications/Documents/Self‐Driving‐

Cars‐next‐revolution.pdf.

[5] Stephen E. Reutebuch, Hans‐Erik

Andersen, and Robert J.McGaughey;

Light Detection and Ranging (LIDAR):

AnEmergingtoolforMultipleResource

Inventory.

[6] bgr.com/2013/01/27/google‐

driverless‐car‐anaysis‐306756/

[7]

Spectrum.ieee.org/automation/robotics

/artificial‐intelligence/how‐google‐self‐

driving‐car‐works.

[8]LuisAraujo,KatyMasonandMartin

Spring;Self‐DrivingCars:acasestudyin

makingnewmarkets.

[9] Q. Zhang and R. Pless, “Extrinsic

CalibrationofaCameraandLaserRange

Finder”, in Proc. IEEE/RST Int.conf.

Intelligent Robots and Systems, Sendai,

Japan,2004.

[10].www.dezeen.com/2014/05/28/pu

blic‐test‐drive‐first‐driverless‐cars‐by‐

google/

[11]. Todd Litman, Victoria Transport

Policy Institute; Autonomous Vehicle

ImplementationPredictions.


ISSN: 2348-4098


BIOGRAPHIES

Research Scholar ,

Department of Computer

ScienceandEngineering,CT

Institute of Technology &

Research , Jalandhar,India ,

[email protected]

Research Scholar ,




Research , Jalandhar,India

[email protected]

Research Scholar ,




Research , Jalandhar, India ,

[email protected]

Research Scholar ,




Research , Jalandhar, India ,

[email protected]


ISSN: 2348-4098


PREDICTIONOFCASTINGDEFECTSTHROUGHARTIFICIALNEURALNETWORK

GANESHG.PATIL,DR.K.H.INAMDAR

1M.TechStudent,DepartmentofMechanicalEngineering,WalchandCollegeofEngineeringSangli,Maharastra,India,

Email:[email protected]

2Professor,DepartmentofMechanicalEngineering,WalchandCollegeofEngineeringSangli,Maharastra,India,


ABSTRACT

India is the second largest casting component producer in theworld after China. So,

foundriesrepresentimportantsectorofthemanufacturingindustry.Castingprocessis

the most widely used process in manufacturing industries especially in automobile

products. Systematic analysis and identification of sources of product defects are

essential forsuccessfulmanufacturing.Foundryindustrysuffersfromthepoorquality

andproductivityduethelargenumberofprocessparameter.Sincethequalityofcasting

partsismostlyinfluencedbyprocesscondition,howtodeterminetheoptimumprocess

condition becomes the key to improving part quality. The industry generally tries to

eliminatethedefectsbytrialanderrorwhichisanexpensiveanderror‐proneprocess.

Butittakestoomuchtimeandmanpower.Nowadaycommercialtechniquesareused

to simulate casting process. Simulation software only gives validation of results.

Improvementincastingqualityistheprocessoffindingtherootcauseofoccurrenceof

defectssuchassanddrop,blowhole,leakageandbadmouldintherejectionofcasting

andtakingnecessarystepstoreducethedefectsandhencerejectionofcasting.Inthis

dissertation work, for improvement in casting quality, the artificial neural network

technique is use for the optimize the sand andmoulding related parameters such as

greencompressionstrength (GCS),permeability,moisturepercent,metal composition

andmetaltemperature.Theneuralnetworkwastrainedwithparametersasinputsand

the presence/absence of defects as outputs. Artificial neural network is used for the

optimizationcastingparametersbyusingMATLABsoftware.Theresults indicate that


ISSN: 2348-4098


selected process parameter significantly affect the casting defect and causes the

rejection.

INDEX TERMS: Artificial neural network (ANN), Casting defects, Optimization and

Castingprocess.

1. INTRODUCTION

Metal casting has been a primary

manufacturing process for several

centuriesduringBCandissoeventoday

in the 21st century. Today, its

applications include automotive

components, spacecraft components

and many industrial and domestic

components. The principle of

manufacturing a casting involves

creating a hollow shape of themetallic

component to be made inside a sand

mouldandthenpouringtheliquidmetal

directly into the sand‐shaped mould.

Casting is a very versatile process

capable of being used in mass

production items in very large shaped

pieces,withintricatedesignsandhaving

properties unobtainable by any other

methods. The major activities involved

in making a casting are moulding,

melting, pouring, solidification, fettling,

cleaning, inspection and elimination of

defectivecastings[1].

Foundry industry suffers from poor

qualityandproductivityduetothelarge

number of the process parameters.

Global buyers demand defect‐free

castings and strict delivery schedule,

which foundries are finding it very

difficult to meet. Casting defects result

inincreasedunitcostandlowermorale

ofshopfloorpersonnel.

Casting process is also known as

process of uncertainty. Even in a

completely controlled process, defects

in casting are found out which

challenges explanation about the cause

ofcastingdefects.Thecomplexityofthe

processisduetotheinvolvementofthe

various disciplines of science and

engineering with casting. The cause of

defectsisoftenacombinationofseveral

factors rather than a single one. It is

important to correctly identify the

defect symptomsprior to assigning the

cause to the problem. False remedies

not only fail to solve theproblem, they

canconfusetheissuesandmakeitmore

difficulttocurethedefect[2].

The metal casting is one of the basic

manufacturing processes. The purpose

of process development is to improve

the performance characteristics of the


ISSN: 2348-4098


process related to customer needs and

expectations. The process development

can be achieved through

experimentation and the aim is to

reduce and control variation of a

process. Subsequently, decisions must

be made concerning which parameter

affects the performance of the process.

By properly adjusting the factors, the

variations of the process are reduced

therebythelossescanbeminimized[3].

Casting defect analysis can be carried

outbyusingtechniqueslikecause‐effect

diagrams,designofexperiments,casting

simulation, if‐then rules (expert

systems)andartificialneuralnetworks.

Among all these different techniques

mentioned above, artificial neural

networktechniquesareproposedtouse

for analysis of casting defects and

improvethecastingquality.

Figure1:Stepsinvolvedincastingprocess

The first step inmaking a casting is to

makeahallowcavityinsidesandmould

suchthattheshapeofthehallowcavity

insidethesandmouldwouldbesimilar

tothatofthecomponentwhichisgoing

to be manufactured. This process is

knownas‘moulding’.Thesecondstepis

‘melting’, which involves melting the

solid chargemetal insidea furnaceand

making the liquidmetal free fromslags

andanydissolvedgases.The thirdstep

is ‘pouring’,which involvespouring the

moltenmetalinsidethesandmouldand

allowing the liquid metal to solidify

insidethemould,thusmakingthemetal

to take the shape of the mould cavity.

The fourth step is the ‘fettling’process,

in which the sand mould is broken

(after solidification of the casting) and

the solidified casting is taken out. The

casting is also cleaned with water,

pressurized air, etc. The fifth step is

‘inspection’ that includes identification

of defective castings through different

techniquesandensuringqualitycontrol.

The sixth step is ‘elimination/dispatch’,

which includes recycling of defective

castings for re‐melting and passing on

thesoundcastingsforshipping.

Outoftheseveralstagesinvolvedinthe

castingprocess, ‘moulding’andmelting’

processesconstitutethemostimportant

Moulding Melting

Pouring

Fettling

Inspection

Dispatch


ISSN: 2348-4098


stages, as the parameters of these two

processesmostlydetermine thequality

ofthecasting.Themouldingparameters

like moisture percent, permeability,

green compression strength (GCS),

green shear strength (GSS) affect the

qualityofthefinishedcasting.Similarly,

melting parameters like liquid metal

temperature, carbon percent in liquid

metal (C), manganese percent in liquid

metal (Mn), silicon percent in liquid

metal(Si),sulfurpercentinliquidmetal

(S), phosphoruspercent in liquidmetal

(P) and chromium percent in liquid

metal(Cr)alsodeterminethequalityof

thefinishedcasting.Impropercontrolof

moulding and melting parameters

results in defective castings, which

substantiallyreducestheproductivityof

afoundryindustry[1].

Hence, in thepresentstudy,anattempt

hasbeenmadetopreventthedefectsin

castings by predicting them just before

the‘pouring’stageusingartificialneural

networks. The moulding and melting

parameters of castings ware collected

from a Suyesh Iron & Steels Pvt. Ltd.

foundry and the same were fed as

inputs to a back‐propagation neural

network. The natures of the castings

(‘sound’ or ‘defective’) were fed as

outputstothenetwork.Aftersuccessful

training, the network was able to

predictthechancesofvariousdefectsin

thecastingsthatwereabouttobemade.

In case the network has predicted the

chance of a particular casting defect,

then the possible causes for the

particular defect are to be investigated

and necessary measures have to be

taken so as to prevent the defect that

waspredictedbytheneuralnetwork.

2. RECENTTHEORIES

Recent theories give detail information

aboutpresentpracticesusedindifferent

foundries and results of advanced

researchesallovertheworld.Itisoneof

theimportantstepstobefollowedwhile

carrying out dissertation work.

Literature review not only gives the

historyofaparticularproblembutalso

providesresultsofrecentresearcheson

the same. At present, other than the

artificialneuralnetworks(ANN),casting

defect analysis is carried out using

techniques like historical data analysis,

cause‐effect diagrams, if‐then rules

(expertsystems),simulationanddesign

ofexperiment.

An artificial neural network is

computational model of the human

brain, where information processing is

distributed over some interconnected


ISSN: 2348-4098


processingelements, callednodes (also

called neurons). They are structured in

some layers. These layers are called as

input, output and hidden layers and

theyhavebeenoperatedparalleltoeach

other. The outputs of the node in one

layer are transmitted tonodesof other

layer through connections. While

transmitting outputs from one layer to

anotherviasomeconnections,theymay

be amplified (if necessary) through

weight factors. The net input to each

node(otherthaninputnode)isnetsum

of the weighted output of the nodes

feedingthatnode[1].

M. Perzyk et al. [4] studied that

defectsin castings often appear

unexpectedly and it is difficult to

identify their source as they can be

brought about by alarge number of

randomly changing production

parameters.ANNwasusedfordetection

of the causes of gas porosity defects in

steel castings. The applied procedure

includedsystematicstoringoftwotypes

of information: about the process

parameters, materials used and even

employees involved in the production

(as the network inputs) and about the

appearance of a given defect (as the

network output). The trained network

was able to detect interdependences

amongvariousfactorsinfluencingwater

vapourpressure in themouldand thus

to indicate the most probable cause of

porosity.

Karunakar and Datta [1] have applied

back propagation neural networks for

analysis and prediction of casting

defects. In this paper they had applied

theneuralnetworktothemetalcasting

forpredictionofcastingdefectssuchas

coldshut,sanddrop,slaginclusionsand

microstructurerelateddefects.

Jiang Zheng et al. [5] this study

represent systematic approach to high‐

pressure die casting is a versatile

processforproducingengineeredmetal

parts. There are many attributes

involved which contribute to the

complexityoftheprocess.Itisessential

for the engineers to optimize the

process parameters and improve the

surface quality. However, the process

parameters are interdependent and in

conflict in a complicated way and

optimization of the combination of

processes are time‐consuming. In this

work, an evaluation system for the

surface defect of casting has been

established to quantify surface defects,

and artificial neural network was

introducedtogeneralizethecorrelation

betweensurfacedefectsanddie‐casting


ISSN: 2348-4098


parameters, such asmold temperature,

pouring temperature, and injection

velocity. It was found that the trained

network has great forecast ability.

Furthermore, the trained neural

networkwas employed as an objective

functiontooptimizetheprocesses.

3. INTRODUCTIONTOANN

Neural networks, which are simplified

modelsofthebiologicalneuronsystem,

is a massively parallel distributed

processing system made up of highly

interconnected neural computing

elements that have the ability to learn

and thereby acquire knowledge and

make it available for use. The neural

network could predict cracks, misruns

and air‐locks accurately in most of the

cases. The neural network could also

predict other defects successfully [6].

ANNsarewidelyacceptedastechnology

offering an alternativeway to simulate

complexand ill‐definedproblems.They

havebeenusedindiverseapplicationin

control, robotics, pattern recognition,

forecasting, power system,

manufacturing, optimization, signal

processing etc. they are particularly

useful in system modeling. A neural

network is computational structure

consisting of number of highly

interconnected processing unit called

neuron.Theneuronsumweightedinput

andapplieslinearornonlinearfunction

to the resulting sum to determine to

output and the neuron are arranged in

layer and are combined though

excessiveconnectivity[7].

Identification and control are the two

fundamentaltasksofsolvingaproblem.

The identification and control of

nonlinear systems are still challenging

tasks. Recently, considerable effort has

been invested in the use of artificial

neural networks (ANN) for nonlinear

controlandidentification.Bothpractical

andtheoreticalresultsestablishtheuse

of neural control as one of the most

promising areas of neural network

applications. Neural networks have the

ability to learn from their environment

andadapttoitinaninteractivemanner

similar to their biological counterparts.

A very important feature of these

networks is their adaptive nature,

where ‘learning by example’ replaces

‘programming’ in solving theproblems.

This featuremakes such computational

models very appealing in application

domains where one has little or an

incomplete understanding of the

problem to be solved, but where

training data are readily available.

Neuro‐computingcanplayanimportant


ISSN: 2348-4098


role in solving certain problems in

science and engineering that would

otherwisebedifficulttosolve,problems

such as pattern recognition,

optimization, event classification,

control and identification of nonlinear

systems, and statistical analysis, etc. In

addition toneuralnetwork’susefulness

in solving complexnonlinearproblems,

theyareattractive inviewof theirhigh

execution speed and their relatively

modest computer hardware

requirements[1].

4. PROBLEMDESCRIPTION

ThedataiscollectedfromSuyeshiron&

steel Pvt. Ltd. Foundry for give the

training to artificial neural network.

Thisfoundryismakingcomponentslike

sideframe,bolster,yoke,etc.Themajor

defectsoccurredinthisfoundryarehot

cracks,misrun,scab,blowholes,airlock

andleakage.Hence,anattempthasbeen

madetopredictthesefivedefectsusing

a back‐propagation neural network

before the pouring stage, which is the

thirdstepofthecastingprocess.Before

dealing with the problem, study of the

defectswhich is occurred in foundry is

given below. Hot cracks are formed

becauseofcastingishotandmayoccur

during cooling in the mould, during

knocking‐outhotorduringcoolingafter

hot knock‐out. Hot cracking can also

occur in the event of uneven cooling

conditions.Misrungenerallytakesplace

when the pouring temperature and the

pouringspeeddrasticallydecrease.This

defect arises especiallywhen themetal

intheladleisabouttobeexhaustedand

hence the small amount of metal at

comparatively lower temperature

present inside the ladle flows into the

mould with considerably low speed.

Pouring temperature, however, exert

themajorinfluenceonmisrun[8,9].

Scabsoccurasaresultoftheformation

ofashellofdriedsandonthehotmould

surface.Ascabisformedwhenaportion

of the mould face lifts due to thermal

expansion and liquid metal flows

underneath in a thin layer. Scabbing

generallytakesplaceif(a)pouringtime

is comparatively high, (b) excessive

moisture or volatile matter content

presentinthemould,(c)sandgrainsize

distribution being non‐uniform, (d)

mould having low green strength, (e)

comparatively low active clay content

and/orhighdeadclaycontentand (f)

lowpermeabilityofmould[10].

Blowholesoriginatefromaspontaneous

evaporation of the water present in a

thin surface layer of the mould. Iron

oxideimplantedonthemouldwalloften


ISSN: 2348-4098


results in blowholes. Carbonmonoxide

gas produced by the reaction of fluid

slag from furnace or ladle with carbon

in the iron may also result in the

formation of blowholes. Air lock may

generallyariseasaresultofentrapment

oftheairwithinthemouldbytheliquid

metalduetoafaultygatingsystem.

Theshopfloorinthisfoundryrecorded

fourmouldingproperties,namelygreen

compression strength, green shear

strength, mould permeability and

moisture percent in mould for each

component. The shop floor also

recorded seven melting parameters

namely carbon percent in charge,

manganese percent in charge, silicon

percent in charge, sulfur percent in

charge, phosphorus percent in charge,

chromiumpercentinchargeandmolten

metaltemperatureforeachcomponent.

Thedata is regarding thenatureof the

casting. i.e. either soundordefective. If

the nature of casting is defective, were

alsorecorded.Theabovemouldingand

melting parameters for different

components that were collected from

thesaidfoundry.Incasethecastingwas

defective, the nature of defect is note

downintheremarkscolumn.

5. IMPLEMENTATIONOFANN

Implementtheartificialneuralnetwork

for prevent the casting defects such as

crack, misrun, blowholes, scab and

airlock. Prevent the casting defects by

predictingthemjustbeforethepouring

stage using artificial neural network.

Artificial neural network are several

different type of algorithm but in this

paper back‐propagation algorithm is

used. In back‐propagation algorithm

initially weight are calculated,

consequently output are calculated

randomly. However, these calculated

outputs compare with the

actual/desired output by the neural

networkanderroristransmittedtothe

initial layer, which result in correction

of the weights. The training iteration

processmay be terminated either by a

convergence limitor simplyby limiting

thetotalnumberofiterations.Thesteps

of the ANN calculation during training

usingback‐propagationalgorithmareas

follows.

Step 1: The network synaptic weights

areinitializedtosmallrandomvalues.

Step 2: From the set of training

input/output pairs, an input pattern is

presented and thenetwork response is

calculated.


ISSN: 2348-4098


Step3:Thedesirednetworkresponseis

comparedwith theactualoutputof the

network, and all the local errors to be

computed.

Step 4: The weights preceding each

output node are updated according to

thefollowingupdateformula:

Δwij(t)=ηδioi+αΔwij(t‐1)(1)

Where,

η‐Learningrate

δ–Localerrorgradient

α‐Momentumcoefficient

oi–Outputofithinput

wijrepresentstheweightconnectingthe

ith neuron of the input vector and the

jth neuron of the output vector. The

localerrorgradientcalculationdepends

on whether the unit into which the

weightsfeedisintheoutputlayerorthe

hiddenlayers.Localgradientsinoutput

layersaretheproductofthederivatives

of thenetwork’serror functionand the

units’ activation functions. Local

gradients in hidden layers are the

weighted sum of the unit’s outgoing

weights and the local gradients of the

unitstowhichtheseweightsconnect.

Step5:The cycle (step 2 to step 4) is

repeated until the calculated outputs

haveconvergedsufficientlyclose to the

desiredoutputsoraniterationlimithas

beenreached.

A processing element accepts one or

more signals, which may be produced

by other processing elements. The

various signals are individually

amplified, or weighted, and then

summedtogetherwithintheprocessing

element.Theresultingsumisappliedto

a specific transfer function, and the

function value becomes the output of

the processing element. Transfer

function used in the back‐propagation

networkisknownas‘sigmoidfunction’,

whichisshownbelow.

F(s)=1/(1+e‐s) (2)

Where, s is thesumof thenode inputs.

Clearlythenodeoutputwillbeconfined

totherange0<f(s)<1.

Pre‐processingof input signalsprior to

input to the neural network is carried

outasfollows.Allinputandoutputdata

arescaledsothattheyareconfinedtoa

subinterval of [0...1]. A practical region

for the data is chosen to be [0.1 ....0.9].

In this case each input or output

parameterXisnormalizedasXnbefore

being applied to the neural network,

according to the following equation,

shownbelow


ISSN: 2348-4098


(3)

Where, Xmax and Xmin are the

maximum and minimum values,

respectively, of the data parameter X.

The network starts calculating its

output values by passing the weighted

inputstothenodesinthefirstlayer.The

resultingnodeoutputsof that layerare

passedon,throughanewsetofweights,

to thesecond layer,andsoonuntil the

nodes of the output layer compute the

finaloutputs.

5.1TESTINGOFANN

The four mould properties and the

sevenmeltingparametersmentionedin

the previous sections were fed to the

neuralnetworkasinputs.Thenaturesof

the castings (sound or defective) were

fed as outputs. The ‘presence’ and

‘absence’ofeachdefectwasinputtedto

neural network by the indication of ‘1’

and ‘0’, respectively. The data were N

scaled between 0.1 and 0.9 as per Eq.

(3), before feeding to the network. A

back‐propagation neural network was

constructed for the present task. The

network used one hidden layer and 23

hiddenneurons.Theinputsandoutputs

that were used to train the neural

networkareshownbelow.

INPUTPARAMETERS:

Greencompressionstrength(N/m2)

Greenshearstrength(N/m2)

Permeability

Moisturecontent

Carbonpercentincharge

Manganesepercentincharge

Siliconpercentincharge

Sulfurpercentincharge

Phosphoruspercentincharge

Chromiumpercentincharge

MoltenmetaltemperatureinCelsius

OUTPUTPARAMETERS:

Presence/Absenceofdefect

Eighty‐four samples of data were

collected from the shop floor and the

same have been used to carry out the

present investigation. The first 65

samples were used for training the

networkand the remaining19 samples

werereservedexclusivelyfortestingthe

accuracyof the trainednetwork. In the

neural network program, momentum

ratewassetas0.7and learningrateas

0.5. The error goal was set as 0.01.

Other network architectures were also

tried with two and three hidden

neurons, respectively, but the onewith

single hidden layer gave comparatively

betterresultswithanoptimumtraining

time. The program was run using


ISSN: 2348-4098


MATLAB neural networks toolbox and

thenetworkconvergedtotheerrorgoal

of0.01after4000iterations.Errorgoals

smaller than 0.01 were also tried but

they resulted inmemorizing of data by

the neural network rather than

capturing the generality of inputs and

outputs. After training was over, the

networkwastestedforitsaccuracy.The

inputparametersofthe19samplesthat

were not used in training the neural

network were fed to the trained

networkand thenetworkwasasked to

predictthepossibleoutputs(presence/

absenceofeachdefect).Apart,fromthe

above 19 samples, the inputs of the 65

samples(usedfortrainingthenetwork)

were also fed to the trained neural

networkand thenetworkwasasked to

predictthepossibleoutputs(presence/

absence of each defect). The trained

neural network predicted the presence

of each defect by a decimal close to ‘1’

anditsabsencebyadecimalcloseto‘0’.

5.2DISCUSSIONS

Thepredictions of thedefectsmadeby

the back‐propagation neural network

are satisfactory in most of the cases.

However, the neural network did not

predict ‘1’or‘0’preciselytopredictthe

presence or absence of a defect, but a

decimal value closer to them. In the

present analysis, if the decimal value

was higher than 0.5 for the occurrence

of a defect, then the prediction was

treatedasgoodandifthedecimalvalue

was lowerthan0.5, thentheprediction

was treated as poor. In Tables 1 an

accurateandgoodpredictionofadefect

ishighlightedasbolddecimal,whereas

an inaccurate and poor prediction is

highlightedasunderlinedbolddecimal.

Among all the predictions, predictions

of crack and misrun seem to be most

accurate and good and the predictions

are correct in all the cases. Predictions

of scab were accurate in most of the

cases except in sample number 55 in

which thepredictiondecimalhappened

tobelittlelowerthan0.5,resultingina

weak prediction. Furthermore, other

decimalvaluesofthepredictionforthis

defect were not very close to ‘1’, as in

the case of predictions of crack and

misrun.However,theywerestillhigher

than0.5,whichmakesthepredictionof

the network acceptable. Predictions of

air‐lock were accurate in most of the

cases except in sample number 23, in

which the network predicted the

occurrence of air‐lock when the actual

castingwasasoundone.


ISSN: 2348-4098


Coming to the blowhole, the network’s

predictionwas not very good as in the

case of the other defects. In sample

number 36, the network predicted the

occurrence of blowhole by a decimal

little lower than0.5. In samplenumber

69alsotheoccurrenceofblowholewas

predictedbyadecimal little lowerthan

0.5. In sample number59, the network

predictedtheoccurrenceofmisrunand

blowhole when the casting actually

possessedonlymisrun.Thismaybedue

tothefactthattheneuralnetworkcould

not learn the generality of this defect,

duetotheincorrectentryofdatainthe

shopfloor.

It should not be overlooked that the

soundcastingswerepredictedassound

in all the cases, which would further

make the network’s prediction more

accurate and acceptable. Though the

presentworkonmodellingmakesuseof

rather limited data obtained from a

singlefoundryandthisexercisehasnot

been tried with data from another

industry,itisstillfeltthattheworkhas

fairly conclusively demonstrated

usefulness and capability of ANN

modelling. However, the present work

predicts and thereby prevents the

casting defects just before the castings

are about to be made, by the use of

artificialneuralnetworks.

6. CONCLUSIONS

The ANN is the effective technique in

shop floor for the prediction of casting

defectsinfoundry,itwarnsthefoundry‐

man whenever a defective casting is

about to be manufactured. Thus, ANN

minimizes the defective castings and

increases productivity. The neural

network has to be trained using shop

floordata.Thenumberofhiddenlayers

and hidden neurons should be fixed

optimally and this task requires time

andskill.However,thisisaonetimejob

that is to be done in the beginning,

duringthetrainingofthenetwork.After

the training and testing tasks are

completed successfully, the weights of

the trainednetworkare tobestored in

thecomputer.Atthetimeofacastingis

going to be manufactured, the data

relatedtomouldpropertiesandmolten

metalaretobefedtothetrainedneural

networkandtheneuralnetworkwould

predict the nature of the casting

(whether sound or defective). If the

predicted nature of the casting is

‘sound’, then the remaining steps of

casting Process like pouring, fettling,

etc., are tobe carriedout.On theother

hand, if the neural network has


ISSN: 2348-4098


predicted some defect, then the take

some action on mould properties and

molten metal and again reset the

parameter and check by trained neural

network.

TABLE‐1:PREDICTEDDEFECTBYNEURALNETWORK

Sr.

No

Nameof

casting

Predicteddefects Actualdefects

Crack Misrun Scab AirlockBlowhol

e

Crack

Misrun

Scab

Airlock

Blowhole

1 SideFrame ‐0.0616 0.0035 ‐0.0160 ‐0.1245 0.0153 0 0 0 0 0

2 SideFrame 0.0250 ‐0.0057 ‐0.0069 ‐0.0403 ‐0.1249 0 0 0 0 0

3 SideFrame ‐0.0280 ‐0.0015 0.0021 ‐0.1198 0.0191 0 0 0 0 0

4 SideFrame ‐0.0374 ‐0.0073 ‐0.0024 ‐0.0029 0.0766 0 0 0 0 0

5 SideFrame 0.0000 ‐0.0163 0.0036 ‐0.0211 ‐0.0051 0 0 0 0 0

6 SideFrame ‐0.0062 ‐0.0144 ‐0.0009 0.5767 ‐0.0968 0 0 0 1 0

7 SideFrame ‐0.0168 1.0264 0.0663 ‐0.0020 ‐0.0432 0 1 0 0 0

8 SideFrame ‐0.0696 0.0375 ‐0.0008 0.0512 0.1545 0 0 0 0 0

9 SideFrame ‐0.0132 0.0939 0.0017 0.8675 0.0075 0 0 0 1 0

10 SideFrame ‐0.0507 0.9949 ‐0.0105 ‐0.1250 0.0026 0 1 0 0 0

11 SideFrame 0.0422 ‐0.0059 ‐0.0057 ‐0.0792 ‐0.0898 0 0 0 0 0

12 SideFrame ‐0.0276 ‐0.0034 ‐0.0343 ‐0.0150 ‐0.0020 0 0 0 0 0

13 Bolster ‐0.0829 0.0193 ‐0.0041 ‐0.0714 0.5724 0 0 0 0 1

14 Bolster ‐0.0882 ‐0.0074 0.7539 ‐0.0586 ‐0.0413 0 0 1 0 0

15 Bolster 0.9276 0.0064 ‐0.0006 ‐0.1249 0.0024 1 0 0 0 0

16 Bolster 0.0384 ‐0.0018 0.0001 0.0053 ‐0.0480 0 0 0 0 0

17 Bolster 0.0371 0.0033 ‐0.0211 ‐0.0266 0.0083 0 0 0 0 0

18 Bolster ‐0.0171 ‐0.0056 0.0030 ‐0.0236 0.0187 0 0 0 0 0


ISSN: 2348-4098


19 Bolster ‐0.0064 ‐0.0103 0.0054 ‐0.0544 ‐0.1224 0 0 0 0 0

20 Bolster ‐0.0369 ‐0.0003 ‐0.0059 ‐0.0261 ‐0.0564 0 0 0 0 0

21 Bolster ‐0.0003 0.0057 ‐0.0062 ‐0.1250 0.0490 0 0 0 0 0

22 Bolster ‐0.0111 ‐0.0124 0.0059 0.1245 0.1084 0 0 0 0 0

23 Bolster 0.0103 0.0029 ‐0.0085 .05062 ‐0.1249 0 0 0 0 0

24 NTTBody 0.0330 ‐0.0118 0.7463 0.0218 ‐0.1204 0 0 1 0 0

25 NTTBody 0.9923 ‐0.0028 0.0595 0.0217 ‐0.1191 1 0 0 0 0

26 NTTBody 0.0127 0.0062 ‐0.0098 0.0159 0.0068 0 0 0 0 0

27 NTTBody ‐0.0312 ‐0.0004 ‐0.0006 0.0962 0.0182 0 0 0 0 0

28 NTTBody ‐0.1179 0.0565 0.0016 ‐0.0895 0.0213 0 0 0 0 0

29 NTTBody ‐0.0245 ‐0.0126 ‐0.0060 ‐0.1052 0.0028 0 0 0 0 0

30 NTTBody 0.0018 ‐0.0009 0.0123 0.0327 ‐0.0435 0 0 0 0 0

31 NTTBody 0.0094 0.0082 ‐0.0039 ‐0.0185 0.5028 0 0 0 0 1

32 NTTBody ‐0.0779 0.0117 1.0027 ‐0.0319 0.2725 0 0 1 0 0

33 NTTBody ‐0.0805 ‐0.0104 ‐0.0030 0.1223 ‐0.0998 0 0 0 0 0

34 NTTBody 0.0150 0.0017 ‐0.0037 ‐0.0182 ‐0.0587 0 0 0 0 0

35 NTTBody 0.0123 ‐0.0001 ‐0.0035 ‐0.0396 0.0712 0 0 0 0 0

36 NTTBody 0.0346 0.0209 0.0192 0.0265 0.4434 0 0 0 0 1

37 Yoke 0.1139 ‐0.0145 0.8529 ‐0.0504 0.0042 0 0 1 0 0

38 Yoke 0.9826 ‐0.0031 ‐0.0029 0.8954 0.0836 1 0 0 1 0

39 Yoke ‐0.0346 0.8967 0.0216 ‐0.0058 0.7698 0 1 0 0 1

40 Yoke 0.0756 0.0014 0.7513 ‐0.1247 0.0372 0 0 1 0 0

41 Yoke 0.0734 ‐0.0095 ‐0.0052 ‐0.0170 0.0225 0 0 0 0 0

42 Yoke ‐0.0304 ‐0.0025 0.0031 ‐0.1250 ‐0.0097 0 0 0 0 0


ISSN: 2348-4098


43 Yoke ‐0.0619 0.0147 ‐0.0044 ‐0.0186 ‐0.0877 0 0 0 0 0

44 Yoke 0.0032 ‐0.0105 0.0013 ‐0.1243 ‐0.0294 0 0 0 0 0

45 Yoke 0.0241 ‐0.0055 ‐0.0055 ‐0.0241 ‐0.0696 0 0 0 0 0

46 Yoke 0.0379 ‐0.0064 ‐0.0073 0.0209 0.9634 0 0 0 0 1

47 Yoke 0.0394 ‐0.0022 0.0236 0.0251 0.0226 0 0 0 0 0

48 Yoke 1.0411 ‐0.0517 ‐0.0021 0.0024 0.0117 1 0 0 0 0

49 SideFrame 0.1054 ‐0.0093 0.2009 ‐0.0226 0.3435 0 0 0 0 0

50 SideFrame 0.0834 0.9177 0.0673 ‐0.1250 0.0471 0 1 0 0 0

51 SideFrame 0.0105 0.1026 ‐0.0064 ‐0.2181 ‐0.1235 0 0 0 0 0

52 SideFrame 0.9054 ‐0.0040 0.0173 ‐0.0240 0.2365 1 0 0 0 0

53 SideFrame ‐0.0942 ‐0.0086 0.0597 0.1800 0.8029 0 0 0 0 1

54 SideFrame ‐0.1234 ‐0.2091 0.0394 0.0376 0.0570 0 0 0 0 0

55 SideFrame 0.0685 ‐0.0027 0.4015 ‐0.2233 0.0299 0 0 1 0 0

56 SideFrame ‐0.0165 0.2035 0.0039 ‐0.1193 ‐0.0310 0 0 0 0 0

57 SideFrame 0.0670 ‐0.0026 ‐0.0120 0.0182 0.7841 0 0 0 0 1

58 SideFrame 0.0113 ‐0.0044 0.2346 0.0451 ‐0.1240 0 0 0 0 0

59 SideFrame 0.0954 0.8037 0.0089 ‐0.1250 0.5143 0 1 0 0 0

60 SideFrame ‐0.0053 ‐0.0070 ‐0.0033 0.5904 0.3066 0 0 0 1 0

61 SideFrame ‐0.0200 0.0074 ‐0.0002 ‐0.0136 ‐0.1109 0 0 0 0 0

62 SideFrame 0.2543 1.0014 0.0075 ‐0.1122 ‐0.0116 0 1 0 0 0

63 SideFrame 0.0035 ‐0.0027 ‐0.0033 ‐0.1174 ‐0.0404 0 0 0 0 0

64 Bolster ‐0.0626 ‐0.0016 ‐0.0041 ‐0.0376 0.0407 0 0 0 0 0

65 Bolster ‐0.0037 ‐0.0024 ‐0.0483 ‐0.0025 ‐0.0340 0 0 0 0 0

66 Bolster ‐0.0415 ‐0.0025 0.0393 ‐0.0669 ‐0.0460 0 0 0 0 0


ISSN: 2348-4098


67 Bolster ‐0.0377 0.0002 ‐0.0047 ‐0.1129 0.0086 0 0 0 0 0

68 Bolster 0.0015 0.0399 ‐0.0049 ‐0.0211 0.0213 0 0 0 0 0

69 Bolster ‐0.1007 ‐0.0036 ‐0.0108 ‐0.0420 0.4638 0 0 0 0 1

70 Bolster ‐0.0115 ‐0.0032 0.6494 ‐0.1250 ‐0.0093 0 0 1 0 0

71 SideFrame 1.1834 0.0161 0.0007 ‐0.0445 ‐0.1093 1 0 0 0 0

72 SideFrame ‐0.0326 0.0100 0.0013 ‐0.0923 0.1293 0 0 0 0 0

73 SideFrame 0.0400 0.0564 0.8341 0.0214 ‐0.1237 0 0 1 0 0

74 SideFrame 0.9602 0.0090 0.0011 ‐0.0101 ‐0.1029 1 0 0 0 0

75 Bolster ‐0.0658 ‐0.0032 0.0044 ‐0.0231 ‐0.0246 0 0 0 0 0

76 Bolster 0.0457 ‐0.0053 ‐0.0046 ‐0.1241 0.0155 0 0 0 0 0

77 Bolster ‐0.0340 ‐0.0062 0.0238 0.0005 ‐0.0804 0 0 0 0 0

78 Bolster 0.0279 ‐0.0098 0.0182 0.8427 0.0464 0 0 0 1 0

79 Bolster ‐0.1236 1.0017 ‐0.0031 0.0110 0.0047 0 1 0 0 0

80 Bolster 0.0631 0.0004 0.0046 ‐0.1247 ‐0.0986 0 0 0 0 0

81 Bolster 0.0331 ‐0.0018 0.0056 ‐0.0131 0.7394 0 0 0 0 1

82 SideFrame ‐0.0358 ‐0.0070 1.1079 0.0163 0.8531 0 0 1 0 1

83 SideFrame 1.1248 0.0154 0.8619 0.0282 ‐0.0060 1 0 1 0 0

84 SideFrame 0.0018 ‐0.0520 ‐0.0042 ‐0.0172 0.0013 0 0 0 0 0

REFERENCES:

[1].D.BennyKarunakarandG.L.Datta,

“Preventionofdefects in castingsusing

back propagation neural networks,”

International Journal of Advance

Manufacturing Technology, 2008, vol.

39,pp.1111‐1124.

[2].S.Guharaja,A.NoorulHaqandK.M.

Karuppannan, “Parameter optimization

ofCO2castingprocessbyusingTaguchi

method,” International Journal of

Advanced Manufacturing Technology,

2009,vol.3,pp41‐50.


ISSN: 2348-4098


[3].RichardHeine,CarlLoperandPhilip

Rosenthal, “Principlesofmetalcasting”,

Tata MaGraw Hill Publications (1984),

NewDelhi.

[4]. M. Perzyk and A. Kochanski,

“Detection of Causes of Casting Defects

AssistedbyArtificialNeuralNetworks,”

Institute of Materials Processing,

Warsaw University of Technology,

Warsaw, Poland, 2003, vol. 217, pp.

1279‐1284.

[5]. Jiang Zheng, Qudong Wang, Peng

Zhao and CongboWu, “Optimization of

High‐Pressure Die‐Casting Process

Parameters using Artificial Neural

Network,” International Journal of

Advance Manufacturing Technology,

2009,vol.44,pp.667‐674.

[6].S.RajasekaranandG.Vijayalakshmi

Pai, “NeuralNetworks, FuzzyLogic and

Genetic Algorithm,” Prentice ‐ Hall of

India,EasternEconomyEdition,2008.

[7]. Lakshmanan Singaram, “Improving

Quality of Sand Casting using Taguchi

Method and ANN Analysis,”

International Journal on Design and

Manufacturing Technologies, January

2010,vol.4,no.1,pp.1‐5.

[8]. Greenhill JM, “The prevention of

cracking in iron castings during

manufacture,”TheBritishFoundry‐man,

1969,vol.62(10),pp.378–391.

[9]. Goodrich GM, “Investigating cast

iron defects: Four foundries

experiences,” Mod Cast, 1999, vol.

89(12),pp.36–39.

[10]. Boenisch D and Patterson W,

“Discussion of the scabbing tendencies

ofgreensand,”AFSTrans,1966,vol.74,

pp.470–484.

BIOGRAPHIES:

Ganesh G. Patil is currentlystudent of M. Tech(Mechanical – ProductionEngineering). He iscompletedhis graduation inMechanical Engineering in2012 from Dr. BAMUAurangabad, Maharashtra,India.

Dr. K. H. Inamdar is working inDepartment of Mechanical Engineering,WalchandCollegeofEngineering, Sangli.Hehaspublishedmorethan80technicalpapersinvariousnational/internationalconferencesaswell as journals.Hisareaof interest is in quality control and heacquiredpatentrelatedtoit.


ISSN: 2348-4098


VERTICALDISTRIBUTIONANDABUNDANCEOFSOILACARINAINANATURALFORESTANDJHUMLAND

ECOSYSTEMOFMOKOKCHUNG,NAGALAND

1KRUOLALIETSURHO,2BENDANGAO

1AssistantProfessor,DepartmentofZoology,FazlAliCollege,Mokokchung,Nagaland,

India,Email:[email protected]

2AssistantProfessor,DepartmentofZoology,NagalandUniversity,Lumami,Zunheboto,

Nagaland,India,Email:[email protected]

ABSTRACT

SeasonaldistributionanddifferencesofsoilAcarinaabundanceinanaturalforestand

jhumlandecosystemofMokokchungdistrictofNagalandwasassessedduringJanuary

2009 to December 2011. Itwas observed that the total annual population density of

Acarinaandtheirverticaldistributionpatternin3(three)differentdepthsi.e.,0‐10cm,

10‐20 cm, 20‐30 cm of the soil layers showed higher population density in forest

ecosystemas compared to jhum land ecosystem, and thismaybebecauseof the rich

vegetation,physico‐chemical factorsandabsenceofhumaninterferenceinthenatural

forest.Incaseofthejhumlandecosystem,thelowerpopulationdensityofAcarinamay

be due to slash and burn, sparse vegetation and anthropogenic practices. Increase in

depth showed a significant decrease in the population density in both the sites ‐ the

highest density being recorded during the rainy season, and the lowest during the

winter.Theeffectofphysicalparametersrevealedsignificantcorrelationwith thesoil

Acarina,but therewasnoappreciablerelationshipwithotherchemical factorsexcept

forsoilpotassium.Thus,thedistributionofsoilmicroarthropodsisaffectedbyvarious

propertiesinanagroecosystem,aswellashabitatquality.

KEYWORDS:Acarina,microarthropods,verticaldistribution,jhumland.


ISSN: 2348-4098


1. INTRODUCTION

Acarina are minute, free‐living

microarthropods found abundantly in

the soil and litter. They are the

dominant group amongst the soil‐litter

sub system and play an important role

in nurturing or maintaining the

sustainabilityofanecosystem.Theyare

one of the most important organisms,

because of the fact that they play an

essential role in soil fertility via

decomposition of organic matter, soil

mineralization, maintenance of soil

physical structure, nutrient cycling,

energy flow and enhancing primary

productivity(BadejoandStaalen,1993;

Hofer et al., 2001; Yang and Chen,

2009). But the direct contributions of

soil microarthropods are often subtle

(Seastedt1984;Huntetal.1987;Hunter

et al. 2003). Their high population

densitymaybeattributedtoavailability

of nutrient, dense vegetation and litter,

canopy covering and optimum physic‐

chemical factors.Environmental factors

suchas temperature,soilmoisture,and

pH also commonly affect their biology

(vanGestel and vanDiepen1997; Choi

et al. 2002; Cassagne et al. 2003;Ke et

al. 2004), and are thus likely to have

bothdirectandindirect impactsonsoil

systems(Rethetal.2005).

Soil Acarina are divided into four sub

orders viz. Cryptostigmata,

Mesostigmata, Prostigmata and

Astigmata. Despite their small size,

whichrangesbetween0.2to9mm,they

are important component in the sense

that they are associated with highly

organic, decomposing mater (Christian

andBellinger,1980).Anydisturbancein

the microclimatic conditions through

changes in climate, physico‐ chemical

properties of soil, type of vegetative

cover,typeanddepthoflitteretc.would

be accompanied by negative effects on

their reproduction and survival and

adverselyaffect theirpopulation(Price,

1973; Seastedt, 1984; Badejo and

Staalen,1993;WardleandGiller,1996).

It is also to be stressed that

microarthropods often respond to

environmental factors in a nonlinear

manner, even fluctuating over seasons.

Itisthereforedifficulttoextrapolatethe

net effect of fluctuating environmental

controls. Therefore a detailed

assessmentofsoilmicroarthropodsand

their response to changing

environmentalconditionsisessential.

In Nagaland, the drastic reduction of

vegetative cover due to deforestation

poses serious environmental concerns.

A case in point is the finding by Duolo


ISSN: 2348-4098


and Kakati (2009) who recorded a

higher population in natural forest as

compared to a degraded forest, and

therefore, it is imperative to study the

effectsofurbanization/deforestationon

soil microarthropod populations. Thus

thepresentstudywastakenuptostudy

the effects of deforestation, as also the

effects of climatic and edaphic factors

on their abundance, distribution and

diversity in a natural forest and jhum

landecosystemsinMokokchungdistrict

ofNagaland.

2. MATERIALSANDMETHODS

2.1STUDYSITES

The present study was carried out in

twoadjacentareasofnaturalforestand

jhumlandecosystemsinMopongchuket

village and Chuchuyimpang village

under Mokokchung district, Nagaland

which lies at 26°11'36’’ North latitude

andinbetween94°17'44’’ to94°45’42’’

(E)longitude.Theforestsitecomprised

of rich vegetation which had not been

disturbed for more than twenty years

while the jhum land had almost no

vegetation due to frequent human

activitiesandinterference.

The natural forest comprised of rich

vegetation with a distinct vertical

stratification. The canopy layer has an

average height of 20 metres or more,

comprising of Albizia procera, Schima

wallichii, Alnus nepalensis, Castinopsis

indica, Lithocarpus elegans, Michellia

champaca andPersia villosa. Emergent

trees that overshoot the canopy layers

were not present. The smaller trees

mostly belong to the families of

Lauraceae, Euphobiaceae, Araliaceae,

Ficaseae and Rubiaceae. The average

heightofthesemembersisfoundtobe5

to15mts.The ground flora is rich and

epiphytes,climbersandlianaswerealso

found to be growing abundantly. The

jhumland,ontheotherhandwasnotas

wellstratifiedasthenaturalforest.The

treespeciespresentarethespeciesthat

wereleftuncutwhileclearingtheforest

and the stumps that survived the jhum

cultivation. Quercus serrata, Erythrina

striata,Albiziaprocera, Schimawalichii

were the dominant species present in

thejhumareas.

2.2CLIMATE

The climate of the area is monsoonal,

withwarmmoistsummersandcooldry

winters. Themeteorological databased

onthreeyears(2009‐2011)asshownin

tabular as well as graphical forms

(tables1‐3andfigures1‐3)revealsthat

JunetoOctoberconstituteswetmonths

andNovember toMay the drymonths.


ISSN: 2348-4098


The dry period can be further divided

into summer (March to May) and cool

dry season (November to February).

Thusthereisdistinctsummer(Marchto

May), rainy (June to October) and

winter (November to February)

seasons. March constitutes the

transitionalmonthbetweenwinter and

summer whereas October is the

transitional month between rainy and

winterseason.

The maximum and minimum air

temperature was 21.4°C (August) and

6.3°C (January) respectively in 2010

(Average:Max=21.4°C Min=8.1°C).

The maximum and minimum relative

humidity was 85% (August) in 2009

and 35.55 (December) in 2011

respectively (Average: Max = 83.3%

Min = 54.5%). The maximum and

minimum total rainfall was 972.5 cms

(July)in2011andtheminimumwas3.7

cms (March) in 2009 (Average: Max =

572.5cmMin=11.3cm,totalaverage

rainfall=1859.93cms).

2.3SAMPLINGANDEXTRACTION

In both the forest and jhum land

ecosystems, the sampling collection

sites were divided according to the

elevation because of the terrain viz.

upper elevation site, middle elevation

site and lower elevation site. In each

elevation site, three different plots

havingasizeof10mx10m,eachat25‐

30 m apart were selected from where

soil sampleswere taken randomly. Soil

samples were taken at one month

intervals in the middle week of each

month during the study period. All the

collectionsweremade in themornings

between 10:00 and 11:00 AM. The soil

sampleswerecollectedwiththehelpof

iron cylindrical core with sampler size

of 3.925 cm, which are 10cm in depth

and 5cm in diameter. Three replicates

were collected from each area or

collection site.The samples were

immediately bound in polytene bags,

labelled and brought to the laboratory

foranalysis.Ineachstudysiteatotalof

1944soilsampleswerecollectedduring

the whole study period. The soil

sampleswere thanpackedandbrought

to the laboratory within an average of

one hour after the field collection. The

sampleswerethendividedintosections

and placed in a Tullgren funnel as

describedbyCrossleyandBlair(1991).

The soil microarthropods were

extractedintocollectingvialscontaining

70% alcohol. After the extraction, the

vialsandthecontentswere transferred

into a petridish and vialswerewashed

several times with 70% alcohol. The


ISSN: 2348-4098


extracted soil microarthropods were

preserved in70%alcohol towhich few

drops of glycerine were added to

prevent desiccation. Identification and

counting was done under a binocular

microscope,anddensitycalculated.

2.4SOILANALYSIS

Physico‐chemicalfactorsofthesoillike

temperature, moisture, pH, organic

carbon, total nitrogen, available

phosphorus, and potassium were

analyzedduringeachsamplingperiodin

order to study the impact of these

factorsinthepopulationchangesofsoil

microarthropods. The methodologies

utilized for each are as follows: Soil

temperature (soil thermometer), Soil

moisture content (gravimetric method

according to Misra, 1968 andWilde et

al., 1985), Soil pH (portable glass

electrode pH meter (according to

Jackson, 1958), Soil organic carbon

(oxidation calorimetric method i.e.,

modified Walkey and Black method

according to Anderson and Ingram,

1993),Soiltotalnitrogen(aciddigestion

Kjeldahl procedures according to

Anderson and Ingram, 1993),

Phosphorus (ammonium molybdate

stannous chloridemethod according to

Sparlinget al., 1985),Potassium (flame

photometer according to Steward,

1971).

3. RESULTANDDISCUSSION

The total annual population density of

AcarinaInforestecosystem,was428.42

x 102m‐2 amounting to 43.38% of the

total soil microarthropod population,

while in jhum landecosystem, the total

annual population density of Acarina

was 264.70 x 102 m‐2 amounting to

30.97%tothetotalsoilmicroarthropod

population. The population density of

Acarinashowedadecreasingtrendwith

increase in soil depth in both the

ecosystems (Table 1). The values

recordedindicatethatthepercentageis

higher in the forest ecosystem. This

difference in values between the two

areas or ecosystemsmay be attributed

todifferences inthe localmicroclimatic

conditions, vegetative and litter cover

(Stanton, 1979). This has also been

corroborated by Hazra (1991),

Chitrawati (2002), Doulo and Kakati

(2009) etc. Maximum population

density was observed during rainy

seasonfollowedbysummerandwinter

respectively in both the sites (table 2).

Thistrendhasbeenobservedbyvarious

workers(ChakrabortiandBhattacharya,

1996;YadavaandSingh,1998;Narulaet

al. 1998; Reddy and Venkataiah, 1990;


ISSN: 2348-4098


Gope et al., 2007; Duolo and Kakati,

2009 etc) and may be attributed to

optimumrainfall, soilmoistureandsoil

organic matter that promotes the

growth and activity of the organism

during rainy season, desiccation of soil

surfaceduetohightemperatureduring

summer, and lower rainfall and post

monsooneffect(organismsbecomeless

active)duringwinters.

The monthly variation of total

population density of Acarina in forest

ecosystemandjhumlandecosystemwas

foundtobethehighest inthemonthof

August (68.21 x 102 m‐2) and (53.18 x

102 m‐2) respectively (Fig. 1). The

monthlypopulationdensitiesinthetwo

ecosystemsatdifferentsoildepthsagain

show similarities and differences

according to depth in relation to the

month (Figures 2 and 3). For instance,

themaximumvaluesatallsoildepthsin

both the ecosystems is seen in the

month of August, but the minimum

values differ by being either in the

month of December or January in both

the ecosystems (Figures 2 and 3). The

seasonalverticaldistributionofAcarina

decreaseswithincreasingdepthinboth

the forest and jhum land ecosystems

(Table 4). In both the ecosystems, the

highest readings were in the 0‐10 cm

depth during rainy season, while the

lowest readings were in the 20‐30 cm

depth during winter. The seasonal

variations observed in both the sites

may be due to a cumulative effect of

different factors rather than a single

factor (Petersen, 1980), although

Acarina in the upper soil layers are

primarily influenced by moisture

content and temperature (Strong,

1967).We surmise that the abundance

in the upper layer may be due to

constant deposition of decaying

materials. Moreover, increased

temperature due to solar radiation in

theupperlayersmayindirectlyalterthe

soil micro arthropod communities by

causingashiftofabundanceverticallyin

abundance, and composition of soil

organisms upon which they prey

(Kardoletal.,2011).

Physico‐chemicalparametersofthesoil

showed significant positive correlation

to Acarina population, specifically in

relation to soil moisture, soil

temperature, rainfall and humidity

(table 3). Corpuz‐Raros (1980), Santos

andWhitford(1983)etc.Haveobserved

that moisture and organic matter are

more important than other physico‐

chemicalpropertiesof thesoil in terms

of microarthropod abundance and


ISSN: 2348-4098


diversity, but the relationship between

theotherfactorsandmicroarthropodan

density does not appear to be that

simple.All thesoildepths (0‐10cm,10‐

20 cm, and 20‐30 cm) in both the

ecosystems showed a positive

correlation to soil moisture. The

positive relationship between Acarina

andsoilmoisturecontentsoestablished

across a range of ecosystems as

reported by Lindberg et al. (2002),

Badejo and Akinwole (2006), Chikoski

et al., (2006), Classen et al., (2006)

showsthatAcarinamightbeadaptedto

strong seasonal fluctuations in soil

moisturecontent.Rainfallandhumidity

also showed a positive correlation in

both the ecosystems, as was the case

with soil temperature, except for a

negative correlation (r = ‐0.4635, p <

0.05,atthe10‐20cmdepthoftheforest

ecosystem in the case of soil

temperature, and this might be due to

thefactthatforestsoilisinfluencedand

defined by disturbances such as tree

falls,roottip‐ups,subterraneanlogsleft

untouchedbypassingfiresandaffected

by moisture (Moldenke and

Lattin,1990b). It has also been opined

that thebiochemical signatureofa tree

isimprintedonthelocalsoilecosystem,

evenlongafterthetreeblowsdown,cut

orisburneddownsothatitcontinuesto

influence the soil around it long

afterwards.Inthisregardwewouldlike

topointoutthatdivergentfindingshave

been reported with respect to soil

temperature i.e., Mukharji and Singh

(1970), Sanyal (1982), Hattar et al.

(1998),Reddy(1984),Chitrapati(2002)

etc.havereportedapositivecorrelation,

whileDuoloandKakati(2009)reported

anegativecorrelationinanaturalforest

inNagaland.

Table4showsthatnitrogenandorganic

carbon, has a positive and significant

correlation in relation to distributional

patterns at all soil depths in both the

ecosystems. In the case of soil pH a

negativecorrelation(r=‐0.52,p<0.05)

wasobserved inonlyoneplace i.e., the

0‐10cmdepthofthejhumland,butthe

remaining sampling depths in both the

ecosystems showed a positive

correlation. This negative correlation

may be due to agricultural

intensificationwhichdisturbstheupper

soil layers (alteration of soil pH), thus

disturbingsoil faunaniches(Moreiraet

al., 2006). While the carbon: nitrogen

ration is of considerable importance in

controlling bacterial population, it is

integratedwithseveralotherfactorsall

of which ultimately determine

populationlevel.


ISSN: 2348-4098


Seastedt(1984)observedthatsoilfauna

enhance nitrogen mineralization

markedly by up to 25%. Olsen (1933),

Caldwell and DeLong (1950), Bocock

(1964)etchavedocumentedincreasein

nitrogenindecomposingleaves.Suchan

increase might be due to the

accumulation and retention of nitrogen

in microorganisms (Witkamp, 1963).

Although the effect of soil fauna on

decomposition rates has been reported

bySeastedtandCrossley(1983)‐being

higher in forest ecosystems (Seastedt,

1984), it is not well demonstrated in

agro‐ecosystems (Cromacket al.1975).

In jhumland ecosystems, the effect of

fauna on decomposition rates appears

tobeoflessersignificance.Theeffectof

soil fauna on nutrient cycling in agro‐

ecosystem may be of particular

importance in reducing fertilization

schedules by increasing the use

efficiencyoffertilizerinput.

In the case of phosphorus and

potassium, a negative correlation was

observedinallsamplingdepthsofboth

the areas except for a single positive

correlation in the case of potassium in

the 20‐30 cm depth of the forest

ecosystem.Thismightbeduetovarious

factors, for instance, microbial

immobilization of potassiumwhich can

shift the equilibrium between available

and bound potassium (Bear, 1964).

Potassium may be derived from

weathering of primary and secondary

potassium‐bearing minerals as well as

derived from atmospheric sources

(Black, 1957; Likens et al 1967). The

potassium taken up by the treeswhich

accounts for about 55‐65% is returned

tothelitter/soilinleaffall(Black,1957;

Lutz andChandler, 1946). Potassium is

subsequently mobilized via microbial

decomposition and a shift in exchange

complex equilibrium as a result of

potassium loss. Moreover it is

susceptible to leaching from living and

deadorganicmatterandthesoil(Tukey,

1970; Lutz and Chandler, 1946; Black,

1957;Bear,1964),andmicroorganisms

alsoproduceacidswhichareimportant

in thereleaseof insolublepotassium in

soilminerals.

Phosphorusmaybeorganicorinorganic

(Black, 1957), and may also occur in

othersecondaryformsascompoundsof

calcium, magnesium, iron and

alluminium (Lutz and Chandler, 1946).

Soil litter is the primary reservoir of

available phosphorus (Lutz and

Chandler, 1946), and anymanipulation

thatalters the rateofmineralizationor

nutrient status of the litterwill in turn


ISSN: 2348-4098


affect its amount as well as its

availability. Litter decomposition

releases a large amount of phosphorus

which is then immobilized by chemical

precipitationandmicrobialfixation,and

thus, is not lost through leaching

(Bengston,1970).

According to House et al. (1989), the

ability of forests to maintain large

amounts of nutrients in circulation

appears to explain the relatively high

productivity of forests on soils of high

nutrient status. This also proves that

nutrient cycling is interdependentwith

all components of the ecosystem and

that decomposition is adjusted to

nutrient uptake, and vice versa (Likens

et al. 1970). Moreover, in contrast to

agriculturalsoils,forestsoilsrestrictthe

loss of nutrient elements via leaching

(Overrein, 1969). The influence of

vegetation in the separation of soil

communities is ambiguous i.e., there

appears to be no relationship between

vegetation type and microarthropod

communitystructureongrasslandsoils

(Curry,1978).

Thepresentstudyshowssimilaritiesas

well as differenceswith the findings of

earlierworks,butthismaybeattributed

to local micro‐climatic factors, as also

opined by Wallwork (1970). Under

natural conditions, the subtropical type

climate would tend to favour the

developmentofsubsurfacefauna(Price,

1973). However, the additional effects

of cultivation, fallowing, irrigation and

other habitat disturbances associated

with agriculture are difficult to assess.

Incaseofjhumlandecosystem,thesoil

fauna are no doubt disturbed or

modified considerably by agriculture,

and therefore, their effect on

decomposition rates appears to be less

significant. The data suggests that such

disturbancesmayhaveagreaterimpact

on population densities in the surface

layersthanonthoseindeepersoil.

Earlier studies have found higher

nitrogen in crop residues, as well as

lower lignincontentwhencomparedto

forest ecosystems. In this respect,

Cromacketal.(1975)observedthatthe

effect of soil fauna on nutrients

dynamicsandcalciumdynamicsremain

undemonstrated in agro‐ecosystem,

although they contain number of

oribatid mites (Acarina) which are

important in the calcium dynamics of

forest ecosystem. Therefore, the effect

ofsoilfaunaonnutrientcyclinginjhum

land ecosystems may be of particular

importance in the sense of organic


ISSN: 2348-4098


farming practices, by rescheduling and

minimizinguseoffertilizers.

REFERENCES

Anderson, J.H.andIngram, J.S. (1993).

Tropical soil biology and fertility. A

handbookofmethod.UK.C.A.B.Int.

Badejo, M. A., Akinwole, P. O. (2006).

Microenvironmental preferences of

oribatid mite species on the floor of a

tropicalrainforest.Exp.Appl.Acarol.40,

145‐156.

M.A.Badejo,M.A.andVanStraalen,N.M.

(1993). Seasonal abundance of

Springtails in two contrasting

Environments.Biotropica.25:222‐228.

Bear,F.E.(1964).Chemistryofthesoil.

Reinhold Publishing Co., New York, pp

515.

Bengston, G. W. (1970). Forest soil

improvement through chemical

amendments.J.Forest68:343‐347.

Black, C. A. (1957). Soil‐plant

relationships. John‐Wiley & Sons Inc.,

NewYork,pp332.

Bocock, K. L. (1964). Changes in the

amountofdrymatter,nitrogen,carbon,

and energy in decomposing woodland

leaf litter in relation to the activities of

soilfauna.J.Ecol.52:273‐284.

Caldwell,B.B.andDeLong,W.A.(1950).

Studiesofthecompositionofdeciduous

forest tree leaves before and after

partialdecomposition.Sci.Agr.30:456‐

466.

CassagneN,GersC,GauquelinT(2003)

Relationships between Collembola, soil

chemistry and humus types in forest

stands (France). Biol. Fertil. Soils

37:355–361

Chakraborti, P and Bhattacharya, T.

(1996). Fluctuations of soil

microarthropods ina rubberplantation

and an adjacent wasteland. J. Soil Biol.

Ecol.16(1):54‐59.

Chikoski,J.M.,Ferguson,S.H.,Meyer,L.

(2006).Effectsofwateradditiononsoil

arthropods and soil characteristics in a

precipitation‐limited environment. Acta

Oecol.Int.J.Ecol.30,203–211.

Chitrapati,C.(2002).Ecologicalstudyof

soilmicroarthropodsinthesub‐tropical

forest ecosystem at Khonghampat,

Manipur. Ph. D. thesis. Manipur

University.

Choi, W. I., Ryoo, M. I., Kim, J. (2002)

Biology of Paronychiurus kimi

(Collembola: Onychiuridae) under the

influencesoftemperature,humidityand

nutrition.Pedobiologia.46:548–557.


ISSN: 2348-4098


Christiansen,KandBellinger,P.(1980).

ThecollembolaofNorthAmerica:North

of Rio Grande. Grinnel College, Frinnel

IA.

Classen, A. T., DeMarco, J., Hart, S. C.,

Whitham,T. G., Cobb,N. S., Koch, G.W.

(2006). Impacts of herbivorous insects

ondecomposercommunitiesduringthe

early stages of primary succession in a

semi‐aridwoodland.SoilBiol.Biochem.

38,972‐982.

Corpuz‐Raros (1980). Phillipine

Entomol.4(4):193,figs32–36.

Cromack, K., Jr andMonk, C. D. (1975)

Litter production, decomposition, and

nutrient cycling in a mixed hardwood

watershedandwhitepinewatershed.In

“Mineral Cycling in Southeastern

Ecosystems” (eds.Howell, F. G., Gentry,

J. B. and. Smith,M.H.) pp.609‐624.US

Energy Research and Development

Admin. Symposium Series, CONF‐

740613,Washington,DC.

Crossley, D. A., Blair, J. M. (1991). A

high‐efficiency, low‐technology

Tullgren‐type extractor for soil

microarthropods. Agric. Ecosyst.

Environ.34,187–192.

Curry, J. P. (1978). Relationships

between microarthropod communities

andsoilandvegetationaltypes.Sc.Proc.

R.Dublin.Soc.Ser.A.6:131‐141.

Doulo, V. and Kakati, L. N. (2009).

Vertical distribution and seasonal

variation of soil microarthropods in

natural and degraded forest ecosystem

at Lumami, Nagaland. Ph.D. thesis,

submittedtoNagalandUniversity.

Gope, R., Ray, D. C. and Hazra, A. K.

(2007). Seasonal distribution and

community structure of edaphic

Collembollans in home garden and

secondary successional soil in sub‐

tropical humid climate of Barak Valley

(NE,India).TheEkologia,7(1‐2):63‐70.

Hattar, S.J.S., Alfred, J.R.B. and Darlong,

V.T. (1998). Animal diversity in some

managed and protected forests of

North‐East India with particular

reference to soil fauna. Agro Botanica,

Bikaner. (Eds. Kotwal, P. C. and

Banerjee,S.)

Hazra, A. K. (1991). Effect of

deforestation on the soil macro‐

microarthropod fauna of West Bengal,

India. In Advances inmanagement and

conservation of soil fauna. Bangalore,

pp. 399‐411. (Eds. Veersh, G. K.,

Rajagopal,D.andViraktamath,C.A.)

Hofer H, Hanagarth W, Garcia M,

MartiusC,Franklin,RombkeJ,andBeck


ISSN: 2348-4098


L (2001). Structureand functionof soil

fauna communities in Amazonian

anthropogenic and natural ecosystems.

EuropeanjournalofsoilBiology37.229

–235.

House,G. J., Stinner,B.R. andCrossley,

Jr. D. A. (1984). Nitrogen cycling in

conventional and no‐tillage agro‐

ecosystems: analysis of pathways and

processes. Journal of Applied Ecology.

21:991‐1012.

House, G.J., Del Rosario Alzugaray, M.,

1989. Influence of cover cropping and

no‐tillage practices on community

composition of soil arthropods in a

NorthCarolinaagroecosystem.Environ.

Entomol.18,302‐307.

Hunt, H.W., Coleman, D. C., Ingham, E.

R.,Ingham,R.E.,Elliot,E.T.,Moore,J.C.,

Rose, S. L., Reid, C. P. P., Morley, C. R.

(1987). The detrital food web in a

shortgrassprairie.BiolFertilSoils.3:57‐

68.

Hunter, M. D., Adl. S, Pringle, C. M.,

Coleman,D.C.(2003).Relativeeffectsof

macroinvertebrates and habitat on the

chemistry of litter during

decomposition. Pedobiologia. 47:101–

115.

Jackson, M. L. (1958). Soil chemical

analysis. Prentice Hall Inc, New Jersey,

USA.

Kardol,P.,Cregger,M.A.,Campany,C.E.,

Classen, A. T. (2010b). Soil ecosystem

functioningunderclimatechange:plant

speciesandcommunityeffects.Ecology.

91:767–781.

Kardol P, Reynolds WN, Norby RJ and

Classen AT (2011). Climate change

effects on soil microarthropod

abundance and community structure.

AppliedSoilEcology47(1)37‐44.

Ke X, Yang YM, YinWY, Xue LZ (2004)

Effects of low pH environment on the

collembolan Onychiurus yaodai.

Pedobiologia48:545–550

Likens,G.E.,Bormann,F.H.,Johnson,N.

M.,andPierce,R.S.(1967).Thecalcium,

magnesium, potassium, and sodium

budgets of a small forested ecosystem.

Ecology48:772‐785.

LikensG.E.,Bormann,F.H.,Johnson,N.

M.,Fisher,D.W.,andPierce,R.S.(1970).

Effects of forest cutting and herbicide

treatment on nutrient budgets in the

Hubard Brook Watershed Ecosystem.

Ecol.Monogr.40:23‐47.

Lindberg, N., Bengtsson, J., Persson, T.,

(2002). Effects of experimental


ISSN: 2348-4098


irrigation and drought on the

composition and diversity of soil fauna

in a coniferous stand. J. Appl. Ecol. 39,

924–936.

Lutz, H. J. and Chandler, R. E. (1946).

Forest soils. John‐Wiley & Sons Inc.,

NewYork,pp514.

Misra, R. (1968). Ecology work book.

OxfordandIBHPub.Co.Calcutta,India.

Moldenke, A.R., & Lattin, J.D. (1990b).

Dispersal characteristics of old‐growth

soilarthropods:Thepotentialforlossof

diversity andbiological function.

NorthwestEnvironmental Journal,6(2),

408‐409.

MorieraFMS, Iquera JOandBrussardL

(2006). Soil organisms in Tropical

ecosystems: a key role forBrazil in the

Global quest for the Conservation and

sustainable use of Biodiversity in

Amazonian and other Brazilian

Ecosystems. CAB International,

Wallingfork,UK.Pp:1‐12.

Mukharji, S. P. and Singh, J. (1970).

Seasonalvariationsinthedensitiesofa

soil arthropod population in the

densitiesofasoilarthropodpopulation

in a rose garden at Varanasi (India).

Pedobiologia.105:442‐446.

Narula, A., Vats, L. K. and Handa, S.

(1998). Collembolas and mites of

deciduousforeststand.IndianJournalof

forestry.21(2):147‐149.

Olsen, C. (1933). Studies of nitrogen

fixation. 1. Nitrogen fixation in dead

leaves of forest beds. Compt. Rend. des

Trav.derLab.Carlsberg19(9):1‐36.

Overrein,L.N.(1969).Lysimeterstudies

ontracernitrogeninforestsoil.SoilSci.

107:149‐159.

Petersen, H. (1980). In Soil Biology as

relatedtolandusepractices(ed.Dindal,

D.L.).Proc.VIIIInt.SoilSoilZool.Colloq.

pp.806‐833.

Price, D. W. (1973). Abundance and

verticaldistributionofmicroarthropods

inthesurfacelayersofaCaliforniapine

forestsoil.Hilgardia.42:121‐174.

R.V. Reddy, R.V. (1984). Seasonal

fluctuation of different edaphic micro

arthropods population densities in

relation to soil moisture and

temperature in a Pine, Pinus kesiya

Royle plantation ecosystem.

InternationalJournalofBiometeorology

1:55‐59.

Reddy, V.M & Venkataiah, B. (1990).

Seasonal abundance of soil surface

arthropods in reation to some


ISSN: 2348-4098


meteorologicalandedaphicvariablesof

thegrasslandandtreeplantedareasina

tropical semi – arid savanna.

International Journalofbiometeorology

34(1)49–59.

RethS,ReichsteinM,FalgeE(2005)The

effects of soil water content, soil

temperature,soilpH‐valueandtheroot

mass on soil CO2 efflux—a modified

model.PlantSoil268:21–33

Santos, P. F. &Whitford, W. G. (1983).

Seasonalandspatialvariationinthesoil

microarthropod fauna of the white

sands national monument. South

westernNaturalist.28:417–421.

Sanyal,A.K. (1982). Soil oribatidmites

and their relation with soil factors in

WestBengal.J.SoilBiolEcl.2(1):8‐17.

Seastedt, T. R. (1984). The role of

microarthropods in decomposition and

mineralization processes. Ann. Rev.

Entom.29:25‐46.

Seastedt, T. R. and Crossley, Jr. D. A.

(1983). Naphthalene and artificial

throughfall effects on forest floor

nutrient dynamics: a field microcosm

study. Soil Biology & Biochemistry.

15:159‐165.

Sparling, G. P., Milne, J. D. G. and

Vincent, K. W. (1985). Effect of soil

moisture regime on the microbial

contribution to Olsen phosphorus

values. New Zealand Journal of

AgriculturalResearch.38:79‐84.

Stanton,N.L.(1979).Patternsofspecies

diversityintemperateandtropicallitter

mites.Ecology60:295‐304.

Steward,E.A.(1971).Chemicalanalysis

of ecological materials. Blackwell

ScientificPublication,Oxford.

Strong, J. (1967). Ecology of terrestrial

arthropodsatPalmersatation,Antarctic

Peninsula. Antarctic Res. Ser. 10: 357‐

371.

Tukey, H. B. Jr. (1970). The leaching of

substances from plants. Ann. Rev. of

PlantPhysiol.21:305‐324.

VanGestel,C.A.M.,VanDiepen,A.M.F.

(1997). The influence of soil moisture

content on the bioavailability and

toxicity of Cadmium for Folsomia

candida Willem (Collembola:

Isotomidae). Ecotoxicol Environ Saf.

36:123–132.

Wallwork, J. A. (1970). Ecology of soil

animals. Mc graw‐Hill publishing

companylimited.283.

Wardle,D.A.&Giller,K.E. (1996).The

quest for a contemporary ecological

dimension to soil biology. Discussion.


ISSN: 2348-4098


SoilBiologyandBiochemistry.28:1549‐

1554.

Wilde, S. A., Corey, R. B., Iyer, J. G. and

Viogt, G. K. (1985). Soil and plant

analysisfortreeculture.OxfordandIBH

Pub.Co.NewDelhi.

Witkamp, M. (1963). Microbial

population of leaf litter in relation to

environmental conditions and

decomposition.Ecology44:370‐377.

Yadava, P. S. and Singh, E. J. (1988).

SomeaspectsofecologyofOakforestin

Shiroy Hills, Manipur (North Eastern

India). International J. Ecol. & Environ.

Sci.14:111‐116 .

Yang X and Chen J (2009). Plant litter

quality influences the contribution of

soil fauna to litter decomposition in

humid tropical forests, southwestern

China. Soil Biology and Chemistry 41:

910‐918.


ISSN: 2348-4098


TABLESANDFIGURES

Table1.SeasonalvariationofAcarina(Numbers±S.E)x102m‐2

Area Season Soillayers Total

0‐10cm 10‐20cm 20‐30cm

Natural

forest

Winter 42.61±0.90 20.43±0.26 14.21±0.71 77.25±0.53

Summer 65.02±0.52 27.54±0.83 23.64±0.43 116.20±0.42

Rainy 97.76±0.28 82.71±0.21 50.90±0.85 231.37±0.14

Annual 205.39±0.33 130.68±0.27 88.75±0.69 424.82±0.68

Jhum

land

Winter 33.45±0.30 21.21±0.12 6.73±0.17 61.39±0.48

Summer 42.91±0.15 29.26±0.11 8.56±0.83 80.73±0.97

Rainy 66.07±0.49 41.71±0.24 14.80±0.21 122.58±0.33

Annual 142.43±0.81 92.18±0.52 30.09±0.58 264.70±0.98


ISSN: 2348-4098


Table2:TotalnumbersandpercentageofAcarina

Area Soillayer(cm) Numbers±S.E. A B

Natural

Forest

0‐10 205.39±0.33 48.34 53.43

10‐20 130.68±0.27 30.76 42.84

20‐30 88.75±0.69 20.89 33.89

Total 424.82±0.68 100.00 43.38

Jhum

Land

0‐10 142.43±0.81 53.80 43.44

10‐20 92.18±0.52 34.82 32.86

20‐30 30.09±0.58 11.36 16.62

Total 264.70±0.98 100.00 30.97

(A=Percentagecontributionamong thesoil layers i.e.0‐10cm,10‐20cmand20‐30cm

andrepresentthenumberofmicroarthropodsinthelayerwithrespecttototalofallthe

layers in that sampled area. B= Percentage contribution to the total soil

microarthropodsineachlayerrespectively).


ISSN: 2348-4098


Table3.CorelationshipsbetweenAcarinaandphysicalfactors

Factors Soil layers

(cm)

ForestEcosystem JhumlandEcosystem

r2 r p r2 r p

Soil

moisture

(%)

0‐10

10‐20

20‐30

58.09

65.60

52.94

0.76

0.80

0.72

p<0.05

p<0.05

p<0.05

72.51

66.98

40.88

0.85

0.81

0.63

p<0.05

p<0.05

p<0.05

Soil

temp(0C)

0‐10

10‐20

20‐30

40.24

21.49

30.74

0.63

‐0.46

0.54

p<0.05

p<0.05

p<0.05

29.27

43.23

47.30

0.54

0.65

0.68

p<0.05

p<0.05

p<0.05

Rainfall

(cm)

0‐30

90.21

0.94

p<0.05

92.27

0.96

p<0.05

Humidity

(%)

0‐30

72.55

0.81

p<0.05

78.48

0.88

p<0.05


ISSN: 2348-4098


Table4.CorrelationshipbetweenAcarinaandchemicalfactors

Factors Soillayers

(cm)

ForestEcosystem JhumlandEcosystem

r2 r p r2 r p

Soil

pH

0‐10

10‐20

20‐30

43.28

54.47

43.40

0.65

0.73

0.65

p<0.05

p<0.05

p<0.05

27.73

54.47

45.21

‐0.52

0.73

0.67

p<0.05

p<0.05

p<0.05

Soiltotal

nitrogen

(%)

0‐10

10‐20

20‐30

60.04

49.82

50.45

0.77

0.70

0.71

p<0.05

p<0.05

p<0.05

40.19

36.98

45.39

0.63

0.60

0.67

p<0.05

p<0.05

p<0.05

Soil

potassium

(%)

0‐10

10‐20

20‐30

27.36

29.82

40.43

‐0.52

‐0.54

0.63

p<0.05

p<0.05

p<0.05

26.53

7.88

12.21

‐0.51

‐0.28

‐0.34

p<0.05

p<0.05

p<0.05

Soil

available

Phosphorus

(%)

0‐10

10‐20

20‐30

23.65

10.58

12.56

‐0.32

‐0.48

‐0.35

p<0.05

p<0.05

p<0.05

10.59

0.36

2.73

‐0.30

‐0.06

‐0.69

p<0.05

p<0.05

p<0.05

Soilorganic

carbon(%)

0‐10

10‐20

20‐30

42.36

68.82

60.39

0.65

0.82

0.77

p<0.05

p<0.05

p<0.05

30.20

63.69

65.91

0.54

0.79

0.81

p<0.05

p<0.05

p<0.05


ISSN: 2348-4098


0

10

20

30

40

50

60

70

80

Jan

Feb

Mar

Apr

May Jun

Jul

Aug

Sep

Oct

Nov

Dec

Forest ecosystem

Jhumland ecosystem

Figure1:MonthlyvariationoftotalAcarinapopulationdensityinforestandjhumland

ecosystem(Numbersx102m‐2)

0

10

20

30

40

50

60

70

80

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0‐10 cm

10‐20 cm

20‐30 cm

Figure2:MonthlyvariationoftotalAcarinapopulationdensityindifferentsoillayers

offorestecosystem(Numbersx102m‐2)


ISSN: 2348-4098


0

10

20

30

40

50

60

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0‐10 cm

10‐20 cm

20‐30 cm

Figure3:MonthlyvariationoftotalAcarinapopulationdensityindifferentsoillayers

ofjhumlandecosystem(Numbersx102m‐2)


ISSN: 2348-4098


BANDWIDTHENHANCEMENTOFHIGHGAINANTENNAUSINGCIRCULARARRAYOFSQUAREPARASITICPATCHES

1BHAGYASHRIB.KALE,2J.K.SINGH

1M.E.Student,Dept.ofE&TC,VACOE,Ahmednagar,Maharashtra,PuneUniversity,India

2Asst.Prof.,Dept.ofE&TC,VACOE,Ahmednagar,Maharashtra,PuneUniversity,India

Email:[email protected],[email protected]

ABSTRACT

This paper presents the design of Microstrip Antenna (MSA) with circular array ofsquareparasiticpatches(CASPPs)onasuperstratelayerforbandwidthenhancement.The antenna structure consists of a MSA, which feeds circular array of 36 parasiticpatches (PPs) printed below a FR4 superstrate and positioned at about 0.5λ0 heightfrom the ground plane. The antenna structure provides peak gain of 17.15 dBi withimpedancebandwidthof950MHz(16.6%)whichcovers5.25‐5.875GHzISMfrequencybandand5.9‐6.2GHzup‐linkC‐bandforsatellitecommunication.High‐gainandbroad‐band performance is obtained by resonatingMSA and PPs at different frequencies in5.25–6.2GHz band. Results obtained verify that the proposed antenna structure isattractive solution for several wireless communication systems, such as satellitesystems,basestationcellularsystems,andpoint‐to‐pointlinks.

KEYWORDS: High gain wideband antenna, directive antenna, multilayer, stackedantenna,ISM,Fabry‐PerotCavity.

1. INTRODUCTION

MSAisoneofthemostusableantennasat frequencies greater than 1GHz. MSAhas several advantages like lowprofile,lowcost, easy to fabricate, easy to feedetc. Beside all these advantages MSAsuffers from disadvantages like lowgain, lowbandwidth, lowefficiencyetc.[1].Gainenhancement techniquesbasedonFabry‐Perot cavity (FPC) where apartiallyreflectingsurface(PRS)formed

byadielectriclayeroraperiodicscreenat approximately 0.5λ above a groundplane is used. The reflection coefficientof PRS and radiation characteristic offeed antenna affects the gain of PRSantenna[2‐5]. Highgainantennaswithartificialmagnetic conductors basedonFPC model have been proposed [6].High gain antennas using a frequencyselective surface, electromagnetic bandgapresonator[7‐8].High gain antennas using PPs on asuperstrate have been reported. Such


ISSN: 2348-4098


antennas offer high efficiency, low sidelobe level and avoid feed network butsuffer from narrow bandwidth [9‐10].Theseantennasexhibithighgainbutthebandwidth performance is poor. Thetechniques for improving the gain andbandwidth by arranging parasiticelements above the feeding MSA areinvestigated [11‐14]. A high gain andwidebandFPCantennawithCASPPsonasuperstratelayerhavebeenproposed[15].

Inthispaper,antennastructureforhighgainandwidebandwidthapplicationsisinvestigatedanddesignedusingCASPPsat about 0.5λ0 height. The antennastructureconsistofaMSA,which feedscircular arrayof 36 squarePPsprintedbelowaFR4superstrateandpositionedat about 0.5λ0 from the ground plane.The antenna structure is designed tooperate over 5.25‐6.2 GHz band,whichcovers 5.25‐5.875 GHz ISM band and5.9‐6.2 GHz up‐link C‐band for satellitecommunication. Here, the feed‐linenetwork is completely avoided soantennastructure iseasy todesignandfabricate. By resonating the MSA, PPsandFPCatdifferentnearbyfrequencies,gain aswell as bandwidth is improved.The different element of a structureresonating at different close byfrequencies results in gain andbandwidth improvement. The antennadesign and optimization is carried outusing commercial method‐of‐momentbasedIE3Dsoftware[16].Thefollowingsections deal with the antennageometry, design theory, simulationresults. Radiation pattern andimpedance variation of antennastructureisalsodescribed.

2. ANTENNADESIGNMETHODOLOGYANDGEOMETRY

In this section, antenna designmethodology and antenna geometry isdescribed.Thesideviewof thecirculararray of 36 square PPs belowsuperstrate layer antenna structure isshowninFig‐1.TheFeedPatch(FP)isametallicMSA of 0.5mm thickness. It isplaced at a height h = 3mm from thegroundplane.ThePPsarefabricatedatbottomsideofFR4superstrate layeratabouths=0.5λ0height,whereλ0 is thefreespacewavelengthcorrespondingtocentral frequency 5.7 GHz. Relativepermittivity and loss tangent of FR4superstrateis4.4and0.02respectively.Air is used between FP and groundplane, superstrate layer and FP as adielectric medium to achieve higherefficiency.A50Ωcoaxialprobe isusedto feedtheFP.Theantenna isdesignedtooperateover5.25‐6.2GHz frequencyband.GeometryofCASPPs(topview)isshowninFig‐2.

Figure1:Geometryofantennastructure(sideview)


ISSN: 2348-4098


Figure2:Geometryoftheantennastructure(topview)

The antenna structure can beconsidered as a cavity resonator withFSS or superstrate. The antennastructure is an extension of a halfwavelength FPC consisting of a groundplane and a partially reflecting surfacewhich results in multiple reflectionsbetween superstrate and groundplane.A broadside directive radiation patternresults if the distance between theground plane and superstrate is suchthatitcausesthewavesemanatingfromsuperstrate to be in phase in normaldirection. If reflection coefficient of thesuperstrate is ρejψ and f(α) is thenormalized field pattern of feedantenna,thennormalizedelectricfieldEandpowerSatanangleαtothenormalisderivedin[2]

)(cos221

21

fE

(1)

)(2

cos221

21

fS

(2)

Here, is thephasedifferencebetweenwaves emanating from superstrate.Boresight gain ( = 0˚) and bandwidthare functionof reflection coefficient [2‐3]

1/1G

(3)

5.0/)1)(2/(/ ro LffBW

(4)

Resonant distance Lr between groundplaneandsuperstrateisgivenby

22)5.0

360( 0

NLr

(5)

Here0 isexpressedindegreeandN=0,

1,2,3etc.

When a MSA feeds CASPPs on asuperstrate layer, high gain broadsideradiationcanbeachievedifthePPsarefed in phase and current induced atpatches are in phase. Since thePPs arepositioned at different location and atdifferent distance from FP, therefore,feedtoeachelementinvolvesamplitudetapering and phase delay. Beside theamplitude tapering due to distance,there is additional amplitude taperingduetotheradiationpatternofMSA.Theamplitude tapering results in decreasein gain but it improves bandwidth andside lobe level. There is little phasedelay in feed to different PPs, whichleads to bandwidth improvement.Hence, high gain wide band arrayantennawithalowSLLcanbeachieved.


ISSN: 2348-4098


Gain and bandwidth of such structuredepends on the reflection coefficient ofsuperstrate. The gain increases but theband width decreases with reflectioncoefficient of superstrate. Therefore,PPs on a dielectric layer are fabricatedtoenhancethereflectioncoefficientandthe gain. As thewaves emanating fromthe superstrate must be cophasal, thegain of the antenna depends on thespacing between patches and theirdimensions. Identical size of PPs leadsto different close resonant frequenciesresultingintowidebandwidth.

3. ANALYSIS ON INFINITEGROUNDPLANE

MSA using a metallic patch of 0.5 mmthickness at a height h =2mm fromthe infinite ground plane is designedand then a superstrate layer of FR4 aths 0.5λ0 height is placed and thestructure is optimized.MSA provides again of 8 dBi, which increases to 10.5dBiwhenFR4superstrateof1.59mmisplaced above MSA. Placing superstrateabove MSA results into increase incapacitiveimpedance.Tocompensateit,h is increased to 3 mm and hs isoptimized to 30.9 mm. As a result,impedance bandwidth is improved andVSWRlessthan2isobtainedover5.15‐5.875GHz.

InnerCASPPsconsistingof6PPsofsize16 mm × 16 mm is placed abovesuperstrate and structure is optimizedwhich provides VSWR less than 2 over5.15‐5.875 GHz frequency band and12.5 dBi gain. Then another circulararrayconsisting12squarePPsisplaced

below the superstrate layer and thestructure is optimized. The structureprovidesgainof15dBiwithimpedancebandwidthof13.8%,whichcovers5.25‐5.875 GHz ISM bands. Square PPs is of16mm×16mmeach.DistancebetweenPPs is optimized to obtain desiredbandwidthperformance.

OuterCASPPsof18elementsofsize15mm × 15 mm is placed below thesuperstrate layer in addition to twoinnerarraysandstructureisoptimized.ThesizeofPPsinouterarrayisslightlyless than that of inner arrays whichcompensates the amplitude taperingand these PPs resonate at higherfrequency than inner arrays resultinginto wideband performance. Thestructureprovidespeakgainof16.5dBiwith impedance bandwidth of 16.6%which covers5.25‐5.875GHz ISMbandand 5.9‐6.2 GHz up‐link C‐band forsatellitecommunication.Radialdistancebetween PPs is optimized to obtaindesired gain bandwidth performance.TheradialdistancebetweenPPsiskeptclosetoλ,whereλisthewavelengthinFR4dielectric.Theoptimumdimensionsare h = 3mm, hs = 30.9mm, whereassquarePPsininnertwoarraysareof16mm × 16 mm each and PPs in outercircular array are of 15 mm × 15 mmeach.

VSWRvsfrequencyofthefinalantennastructureoninfinitegroundisshowninFig‐3, which shows the operatingfrequency band of the designedstructure.VectorcurrentdistributionattheFPandPPsat5.5GHzand6.1GHzisshown in Fig‐4. The superstrate affectsthephaseandamplitudedistributionoffields. The phase distributions of the


ISSN: 2348-4098


fieldswbemorsuperstfocusinand taperturimprovfrequenpropospeakga

Figure3

withasupere uniformtrate. Theng or phasthus incrre area,vement [4ncy plot issed antennainof16.5

:VSWRvs.fr

erstratearem than onee superstrse smoothreases thresulting

4‐5]. Thes shown inna structudBioninfin

requencyoni

eobservedwithout thrate hashening effehe effectivg in gae gain vn fig‐5. Thure providniteground

infinitegroun

tohea

ectveainvsheesd.

nd

F

4

Trep9an

Figure4:C

Figure5:Gain

4. ANTENFINITE

he antennedesignedlane of siz50MHz(16nd maximu

Currentdistri

nvs.frequencplane

NNA REAEGROUND

na structuron squar

ze 4λ0×4λ06.6%)impeum gain o

ibutioninCA

cyoninfinitee

ALIZATIOD

re with 36re finite0. Structureedancebanof 17.15 dB

ASPPs

eground

ON ON

6 PPs isgrounde offersndwidthBi. Gain


ISSN: 2348-4098


variatioshownwith finconstruradiateparticuThis anefficienefficienFig‐7.Vin Fig‐groundCross pSLL ismore thighgapatternbroadsusabilitapplica

Figure

Figur

onofstructin Fig‐6. Gnite grounuctive ined andular dimenntenna strncymorethncy more tVSWRvsfr‐8. Radiatid at 5.5 GHpolarizatioless thanthan 25,whainwidebann is directiidedirectity of the aations.

e6:Gainvs.fr

re7:Efficiencgro

tureonfinGain incrend. Thismanterferencereflectedsions of firucture offhan85%athan 90%requencypion patterHz is shown is less t‐18 dB wihich is appndantennave and symonwhich cantenna fo

requencyonplane

cyvs.frequenoundplane

itegroundases slightay be duee betweewaves

nite grounfers antennandradiatioas shownplotisshowrn on finiwn in Fig‐than ‐20 dth F/B ratpreciable fas.Radiatiommetricalconfirms thor wideban

finiteground

ncyonfinite

istlytoenatnd.naoninwnite‐9.dB,tioforoninhend

d

F

5

AwpTlosupobfrpbinca

Figure8:VSW

Figure9:Rad

5. CONCL

Ahighgainwith circulaatches is ihe antennow costuperstrate.attern banbserved wrequencybeak gain oandwidthndicate thaapableofg

WRvsfrequeplane

diationpattergroundp

USION

andwidebar array ofinvestigatena structur

easily. Impedanndwidth chwhich covband.Thesof 17.15dBof 16.6

at the progenerating

encyonfinitee

nat5.5GHzlane

bandFPCaf square ped and prere is desigavailable

nce and raharacteristver 5.25–structurepBiwith imp%. Theposed antefficientd

ground

onfinite

antennaparasiticesented.gned one FR4adiationtics are–6.2GHzprovidespedanceresultsenna isdirective


ISSN: 2348-4098


radiation patterns in the desiredfrequencyband.Thestructurehasaflat,conformalprofile,andcanbeembeddedintothehostvehicle.

REFERENCES

[1]. G. Kumar and K P. Ray, BroadbandMicrostrip Antennas, Norwood, MAArtechhouse,2003.

[2]. G. V. Trentini, “Partially reflectingsheet arrays,” IRE Trans. AntennasPropagat.vol.4,pp.666–671,1956.

[3]. A.P.FeresidisandJ.C.Vardaxoglou,“High gain planar antenna usingoptimized partially reflective surfaces,”IEE Proc. Microw. Antennas Propagat148,pp.345–350,2001.

[4]. R. Gardelli, Matteo Albani, andFilippo Capolino, “Array thinning byusing antennas in a Fabry–PerotCavityfor gain enhancement,” IEEE Trans.Antennas Propagat AP‐ 54, pp. 1979–1990,2006.

[5]. A.R.DjordjevićandAlenkaG.Zajić,“Optimization of resonant cavityantenna,” in Proc. of EuropeanConference on Antennas andPropagation,2006.

[6]. S. H. Wang, A. P. Feresidis, G.Goussetis, J. C. Vardaxoglou, “Low‐profile resonant cavity antenna withartificial magnetic conductor groundplane,” Electron. Lett., vol. 40, 405‐406,2004.

[7]. E.A.Parker, “Thegentleman’s guidetofrequencyselectivesurfaces,”inProc

17th Q.M.W. Antenna Symposium,London,1991.

[8]. Y. J. Lee, J. Yeo,R.Mittra, andW. S.Park, “Design of a high directivityelectromagnetic band gap resonatorusing a frequency selective surfacesuperstrate,” Microwave Opt. Technol.Lett.,vol.43,pp.462‐467,2004.

[9]. R. K. Gupta and J. Mukherjee,“Efficient high gain with low SLLantenna structures using circular arrayof square parasitic patches on asuperstrate layer,”, Microwave andOptical Technology Letters, Vol. 52, pp.2812‐2817,December2010.

[10]. R. K. Gupta and J. Mukherjee,“Effectofsuperstratematerialonahighgain antenna using array of parasiticpatches,”Microw.Opt.Technol.Lett.52,pp.82–88,2010.

[11]. Zhi‐Chen Ge, Wen‐Xun Zhang,Zhen‐GuoLiu,Ying‐YingGu,“Broadbandand High gain printed antennasconstructed from Fabry‐ Perotresonator structure using EBG or FSScover,” Microw. Opt. Technol. Lett., 48,pp.1272–1274,2006.

[12]. H. Legay and L. Shafai, “A newstacked microstrip antenna with largebandwidthandhighgain,”inProc.IEEEAP‐SInt.Symp.,pp.948–951,1993.

[13]. EgashiraS.,NishiyamaE.,“Stackedmicrostrip antenna with widebandwidth and high gain,” IEEE Trans.Antennas Propag, 44, pp. 1533‐1534,1996.

[14]. Lee R.Q., Lee K.F., “Experimentalstudy of two layer electromagnetically


ISSN: 2348-4098


coupled rectangular patch antenna,”IEEE Trans. Antennas Propag., 38, pp.1298‐1302,1990.

[15]. A. R. Vaidya, S.K. Mishra, R. K.Gupta and J. Mukherjee, “Efficient HighGain Wideband Antenna with CircularArray of Square Parasitic Patches,” inProc.IEEEAPCAP,pp.39‐40,2012.

[16]. IE3D release 14.0, ZelandsoftwareInc.,Fremont,CA.,USA,2008.


ISSN: 2348-4098


PERFORMANCEEVALUATIONOFUNIVERSALDEHAZING

WITHDIRECTEDFILTERMETHOD

1DINESHKUMARPATEL,2AMITKUMARRAJPUT

1Student,ElectronicsandCommunicationEngineering,RITSBhopal,MadhyaPradesh,

India,Email:[email protected]

2Astt.Professor,ElectronicsandCommunicationEngineering,RITSBhopal,Madhya

Pradesh,India,Email:[email protected]

ABSTRACT

AbstractHazeisanatmosphericindividualitythatsignificantlydegradesthevisibilityof

outdoorscenes.Thisismainlyduetotheatmosphereparticlesthatabsorbandscatter

the light. We build the spread map by estimating the atmospheric light except a

continuous region which has no edge information. Themethod performs a per‐pixel

manipulation,whichisstraightforwardtoimplementandthenapplytheDirectedfilter

to improve the imagequality. The experimental results demonstrate that themethod

yields results comparative to and even better than themore complex state‐of‐the‐art

techniques,havingtheadvantageofbeingappropriateforreal‐timeapplications.

INDEXTERMS:Hazedetection,Dehazing,Directedfilter,universaldehazingandsingle

imagedehazing.

1. INTRODUCTION

Haze is an irritating factor when it

shows up in the image since it causes

poor visibility. This is the major

problem of some applications in the

field of computer vision, such as

surveillance, object recognition, etc. In

order to obtain the clear images, haze

removal is inevitable. Fog, mist and

some other particles that disgrace the

scene image are the results of

atmospheric combination and light

scattering. The radiance achieved to

cameraalong the sightline isdecreased

due to atmospheric light and it is


ISSN: 2348-4098


replaced by previously scattered light,

which is called the airlight. This

degradationwillcausetheimagetolose

contrast and color correctness.

Furthermore, the airlight which affect

the image depends on the depth of the

scene. This knowledge is commonly

used for dehazing problems. We also

adopt this clue to solve the haze

removal problem. Image haze removal

has gotten a growing interest recently.

Moreandmoremethodsareintroduced

in the past three years. Nevertheless,

dehazingisachallengingtopicsincethe

haze is dependent on the unknown

depth information.Often, the imagesof

open‐air scenes are degraded by bad

weather conditions. In such cases,

atmospheric phenomena like haze and

fogdegradesignificantlythevisibilityof

thecapturedscene.Since theaerosol is

misted by additional particles, the

reflected light is scattered and as a

result, distant objects and parts of the

scene are less visible, which is

characterized by reduced contrast and

faded colors. Restoration of images

taken in these specific conditions has

caught increasing attention in the last

years. This task is important in several

outdoor applications such as remote

sensing, intelligent vehicles, object

recognition and surveillance. In remote

sensing systems, the recordedbandsof

reflected light are processed [1], [2] in

order to restore the outputs. Multi‐

image techniques [3] solve the image

dehazingproblembyprocessingseveral

input images that have been taken in

different atmospheric conditions.

Anotheralternative[4]istoassumethat

anapproximated3Dgeometricalmodel

of the scene is given. In this paper of

Treibitz and Schechner [5] different

angles of polarized filters are used to

estimate the haze effects. A more

challenging problem is when only a

single degraded image is accessible.

Solutions for such cases have been

introducedonlyrecently[6]–[9].Inthis

paper we introduce an alternative

single‐imagebasedstrategy that isable

to accurately dehaze images using only

the original degraded information. An

extended abstract of the core idea has

beenrecentlyintroducedbytheauthors

in [10]. Our technique has some

similarities with the previous

approaches of Tan [7] and Tarel and

Hautière[9], which enhance the

visibility in such outdoor images by

manipulating their contrast. However,

in contrast to existing techniques, we

built our approach on universal

dehazingwithdirectedfilter.Wearethe

first to demonstrate the utility and


ISSN: 2348-4098


effectiveness of a fusion‐based

technique for dehazing on a single

degraded image then we made the

universal image dehazing model with

directed filter. In thiswork, our goal is

to develop a simple therefore; all the

universal dehaze processing steps are

designed in order to support these

important features. The main concept

behind universal dehaze based

techniqueisthattwoinputimagesfrom

the original input with the aim of

recoveringthevisibility foreachregion

of the scene in at least one of them.

Additionally,theuniversaldehazeimage

enhancement technique estimates for

each pixel the desirable perceptual

based qualities (called weight maps)

that control the contribution of each

input to the final result. In order to

derive the images that fulfill the

visibility assumptions (good visibility

for each region in at least one of the

inputs) required for the fusionprocess,

we analyze the optimal model for this

typeofdegradation.

2. HAZE DETECTION BY

UNIVERSAL DEHAZING

METHOD

Human eyes are more susceptible to

brightnessthancolor.Thereforeweuse

the atmospheric light estimation and

produceatransmissionmapinthe

colorchannels.Theatmosphericlightis

estimated from the most dense pixel.

Theexistingalgorithmpicksup the top

0.1% brightest pixels in the dark

channel prior. Sine an image does not

haveinformationontheedgeofthesky

orawall in thearea, themis‐estimated

valueoftheatmosphericlightresultsin

failure of the defogging (dehazing)

algorithm. Therefore we use the edge

information to represent the

neighboring pixel’s relative depth

information. With this relative depth

information we can construct the

corresponding atmospheric light to

restrain the edge halation.We produce

thetransmissionmapbyestimatingthe

atmospheric light except a continuous

region which has no edge information.

Andthetransmissionmapisgivenas,

(4)

(5)

Where 0 t restricts the transmission t

(x) to a lower bound 0 t ,whichmeans

thatasmallamountoffogarepreserved

in very dense fog regions. In the

experiment we used 0 t _ 0.1 .Color


ISSN: 2348-4098


distortion problem may occur in the

compensation process. To solve this

problem, the image restored by color

correctionusingstatisticalRGBchannel

feature extraction of image. We

calculatetheRGBchannelratiobetween

foggy and defogged images for color

correction with weighted image. The

RGBchannelratioisdefinedas,

WhereRrepresentsthedefoggedimage

anOthefoggyimage.Asaresult,wecan

obtain the color‐corrected image using

color matching of RGB channels of

restoredimage,suchas

Where J represents the color‐corrected

image,andkthenumberofpixels.

3. DIRECTED FILTER IMAGE

MODELLING FOR HAZE

EXTRACTION

The observed brightness of a capture

image in the presence of haze can be

modelled based on the atmospheric

optics[6,7,11]via

(6)

Where,I(x) istheobservedhazeimage,

J(x) is scene irradiance(the clear haze‐

free image), A is the airlight that

represents the ambient light in the

atmosphere. t(x)ϵ[0, 1] is the

transmissionofthelightreflectedbythe

object, which indicates the depth

information of the scene objects

directly.J(x)t(x)ontherighthandsideis

called direct attenuation, which

describes the scene radiance and its

decay in themedium. The second term

A(1‐t(x))is the atmospheric veil

(atmospheric scattering light), which

causes fuzzy, color shift, and distortion

inthescene.Thegoalofhazeremovalis

torecoverJ(x),Aandt(x)fromI(x).

4. IMAGEDEHAZING

Inthissection,wewilldescribeindetail.

The rough down‐sampled transmission

and the air‐light are estimated firstly,

then the transmission is smoothed and

up sampled using a directed filter, and

finallythehaze‐freeimageisrestored.

4.1EXTRACTTHETRANSMISSION

Thecoreofhazeremovalforanimageis

to estimate the airlight and

transmissionmap.Assumingtheairlight

is already known, to recover the haze


ISSN: 2348-4098


freeimage,thetransmissionmapshould

beextractedfirst.Heetal.[8]foundthat

the minimum intensity in the non‐sky

patches on haze free open‐air images

should have a very low value,which is

calleddark channel prior. Formally, for

an image J, the dark channel value of a

pixelxisdefinedas:

(x)=

Where, isacolorchannelofJ;Ω(x)isa

patch around x. By assuming the

transmissioninalocalpatchisconstant

andtakingtheminoperationtoboththe

patchandthreecolorchannels,thehaze

imaging model in (4) can be

transformedas:

= (x)

+(1‐ (x))

(7)

where, (x) is the patch transmission.

SinceA is alwayspositiveand thedark

channel value of a haze‐free image J

tends to be zero according to the dark

channelprior,wehave

→0

Then the transmission can be exacted

simplyby:

(x) =

(8)

Althoughthedarkchannelpriorisnota

good prior for the sky regions,

fortunately, both sky regions and non‐

sky regions canbewell handledby (8)

sincetheskyisinfinitelydistantandits

transmission is indeed close to zero. In

practice, the atmosphere is not

absolutely free of any particle even in

clear weather. Therefore, a constant

parameter ω(0<ω≤1) is introduced

into(8) to keep a small amountof haze

forthedistantobjects:

(x) = 1‐ ω

(9)

Theestimatedtransmissionmapsusing

(9) is practical. Themainproblemsare

some halos and block artifacts. This is

because the transmission is not always

constant in a patch. Several techniques

were proposed to refine the

transmissionmap, such as softmatting

anddirected joint bilateral filter.These

techniques were functional on the

transmissionmapsoftheoriginal foggy

images and usually several operations

shouldbeusedtoachieveagoodresult,

whichcouldbecomputationalintensive.

For image haze removal, the time

complexity is a critical difficulty that


ISSN: 2348-4098


needs to be addressed. High time

complexity of dehazing may make the

algorithmimpracticable.

4.2REFINETHETRANSMISSION

Toimprovetheefficiency,inthepresent

execution, the transmission map is

obtained form a down‐sampled

minimum channel image. Then, it is

refined and up‐sampled by using

directed filter, which can be explicitly

expressedby[11]:

( ) (10)

( ) =

(11)

Where, istheguidanceimage; and

arethemeanandvarianceof in ;|w|

is the number of pixels in . is a

regularization parameter. The refined

operationonadown‐sampledminimum

channel image leads to a low time

complexity and helps to reduce halos

and block artifacts. Joint up sampling

usingdirectedfilterisappliedtoobtain

the full transmissionmap.Thedirected

filter is reported to be a fast and non‐

approximate linear‐time algorithm,

which can perform as an edge

preserving,smoothingoperatorlikethe

bilateral filter,butdoesnotsuffer from

the gradient reversal artifacts.

Moreover, the directed filter has an

O(N) time (in the number of pixels

N)exact algorithm for both gray‐scale

andcolorimages.

4.3PERFORMANCEPARAMETERS

For a good algorithm, values of these

evaluationmetricsshouldbehigh.

Modelling the Markov pdf

parametrically involves thedatadriven

optimal estimation of the parameters

associated with the potential functions

Vc. The model parameters must be

estimatedforeachdatasetaspartofthe

image processing algorithm. In our

algorithms, the noise variance σ2 in

(10) and the parameter a in the

coefficient MRF pdf in (11) are

unknown. Thus, we need to estimate

these parameters in our algorithms.

Becauseweassumethatthenoiseinthe

fusion model is a Gaussian noise, it is

straightforward to estimate the noise

variance by the maximum likelihood

(ML)criterion.Itisgivenby

(13)

The direct ML estimation of the

parametersassociatedwiththepdfofH


ISSN: 2348-4098


isknowntobeadifficultproblem[32].

TheMLestimateofais

(14)

The potential function can be

simply computed. However, the

normalization term ZH involves a

summation over all possible

configurationsofH,which ispractically

impossibleduetothelargecomputation

time. Note that, for two source images

with size 300 *300, H has a total of

490000 possible configurations. An

alternativemethodforapproximationto

ML estimation is maximum pseudo

likelihood(MPL)estimation,whichwas

proposed by Besag[15]. The MPL

estimation method is a suboptimal

method,whichisgivenby

= .

(15)

Thedifferencesamongthefusedresults

areusuallydifficulttobemeasuredonly

basedonobservation,particularlywhen

the fused images are multiband.

Objective and quantitative analysis can

benefit to a comprehensive evaluation.

Variousimagequalityindiceshavebeen

developed for the purpose of image

fusion [12]–[13]. Some of these indices

validate the spatial resolution, while

others focus on the spectral properties

of the obtained fused result. In this

paper,weemploythreesuchindices.

4.3.1SNR

TheSNRindecibels,asshownin(19),is

a direct index to compare the fused

image to the reference one [16].For

multiband images, it can be calculated

band‐by‐band and also globally

averagedSNR

(16)

4.3.2 Universal Image Quality Index

(UIQI)

A UIQI [14] has been widely used for

imagesimilarityevaluationandwasalso

applied to validate fusion techniques

[13]. UIQI of two images (A and B) is

definedas

(17)

Thisqualityindexmodelsanydistortion

as a combination of three different

factors: loss of correlation, luminance

distortion, and contrast distortion. The

dynamic range of Q is [−1,1], and the

bestvalue1 isobtained ifA=B.When

applying this index to a multiband


ISSN: 2348-4098


image, it is applied band‐by‐band and

averagedoverallbands.[16].

4.3.3 Performance of the image

compressioncoding

Itisnecessarytodefineameasurement

that can estimate the difference

between the original image and the

decoded image. Two common used

measurements are the Mean Square

Error (MSE) and the Peak Signal to

Noise Ratio (PSNR), which are defined

in (2.3) and (2.4), respectively. f(x,y) is

thepixelvalueoftheoriginalimage,and

f’(x,y)is the pixel value of the decoded

image.Mostimagecompressionsystems

are designed tominimize theMSE and

maximizethePSNR.

(18)

(19)

5. RESULTANALYSIS

The algorithm proposed here will

remove haze from an image surface

without former knowledge of the haze

location upon that surface. The

proposed method is based on

determining the illumination profile of

the image surface. This profile is then

used to remove the haze. It is

implemented using MATLAB 7.9.0

(R2009b) on i‐5 processor with 4‐GB

RAM.Thesimulationshavebeen tested

on aerial images in figure 2; Figure

2showstheOriginal Imageof forestand

hazeRemovedImage.

Figure: 3 (a) Original Image of forest, (b)

Dehazed Image by usingMulti scale Fusion (c)

DehazedImagebyusingUniversalDehazing(d)

DehazedImageAfterDirectedfilter.

Table1Comparisonparametersforforestimage

Method Variance Mean SNR UIQI

Multiscale

Fusion

0.1237 0.4923

6.6155

4.8557

Universal

Dehazing

0.0807

0.3263 8.1238 6.4175

(a) (b)

(c) (d)


ISSN: 2348-4098


6. CONCLUSION AND FUTURE

SCOPE

Inthispaper,afastandeffectivemethod

for real‐time imageandvideodehazing

isproposed.Inthepresentedalgorithm,

the airlight and the down‐sampled

transmission can be estimated and

extracted easily. Then using a directed

filter, the transmission can be further

refined and up‐sampled. Results

demonstrate the presented method

abilities to remove the haze layer and

achieve real‐time performances. It is

believedthatmanyapplications,suchas

outdoor surveillance systems,

intelligent vehicle systems, etc, could

benefitfromtheproposedmethod.

REFERENCES

[1].P.Chavez,“Animproveddark‐object

subtraction technique for atmospheric

scattering correction of multispectral

data,”RemoteSens.Environ.,vol.24,no.

3,pp.459–479,1988.

[2].G.D.MoroandL.Halounova,“Haze

removal and data calibration for high‐

resolution satellite data,” Int. J. Remote

Sens.,pp.2187–2205,2006.

[3]. S. Narasimhan and S. Nayar,

“Contrast restoration of weather

degraded images,” IEEE Trans. Pattern

Anal. Mach. Intell,, vol. 25, no. 6, pp.

713–724,Jun.2003.

[4]. J. Kopf, B. Neubert, B. Chen, M.

Cohen, D. Cohen‐Or, O. Deussen, M.

Uyttendaele, and D. Lischinski, “Deep

photo: Model‐based photograph

enhancementandviewing,”ACMTrans.

Graph.,vol.27,no.5,p.116,2008.

[5]. T. Treibitz and Y. Y. Schechner,

“Polarization: Beneficial for visibility

enhancement?” in Proc. IEEE Conf.

Comput. Vis. Pattern Recognit., Jun.

2009,pp.525–532.

[6]. R. Fattal, “Single image dehazing,”

ACM Trans. Graph., SIGGRAPH, vol. 27,

no.3,p.72,2008.

[7].R.T.Tan, “Visibility inbadweather

fromasingleimage,”inProc.IEEEConf.

Comput. Vis. Pattern Recognit., Jun.

2008,pp.1–8.

[8]. K. He, J. Sun, and X. Tang, “Single

imagehazeremovalusingdarkchannel

prior,” in Proc. IEEE Conf. Comput. Vis.

Pattern Recognit., Jun. 2009, pp. 1956–

1963.

[9]. J.‐P. Tarel and N. Hautiere, “Fast

visibilityrestorationfromasinglecolor

or gray level image,” in Proc. IEEE Int.

Conf. Comput. Vis., Sep.–Oct. 2009, pp.

2201–2208.


ISSN: 2348-4098


[10]. C. O. Ancuti, C. Ancuti, and P.

Bekaert, “Effective single image

dehazing by fusion,” in Proc. IEEE Int.

Conf. Image Process., Sep. 2010, pp.

3541–3544.

[11]. H. B. Mitchell, Image Fusion:

Theories, Techniques and Applications.

New York, NY, USA: Springer‐Verlag,

2010.

[12]. M. Grundland, R. Vohra, G. P.

Williams, and N. A. Dodgson, “Cross

dissolvewithout cross fade: Preserving

contrast, color and salience in image

compositing,” Comput. Graph. Forum,

vol.25,no.3,pp.577–586,2006.

[13]. T. Mertens, J. Kautz, and F. V.

Reeth, “Exposure fusion: A simple and

practical alternative to high dynamic

range photography,” Comput. Graph.

Forum,vol.28,no.1,pp.161–171,2009.

[14]. L. Schaul, C. Fredembach, and S.

Süsstrunk, “Color imagedehazingusing

the near‐infrared,” in Proc. IEEE Int.

Conf. Image Process., Nov. 2009, pp.

1629–1632.

[15]. H. Koschmieder, “Theorie der

horizontal ensichtweite,” in

BeitragezurPhysik der Freien

Atmosphare. Munich, Germany:

Keim&Nemnich,1924.

[16].G.FinlaysonandE.Trezzi,“Shades

of gray and colour constancy,” in Proc.

12thColorImag.Conf.,2004,pp.37–41.

BIOGRAPHIES

DineshkumarPatel

Student,E&C,RITSBhopal

AmitKumarRajput

AssistantProfessor,E&C,

RITSBhopal


ISSN: 2348-4098


CYCLETIMEREDUCTIONOFGRINDINGPROCESSUSINGSIX

SIGMAMETHODOLOGY

1ALOKB.PATIL,2DR.KEDARH.INAMDAR

1ResearchStudent,MechanicalEngineering,WalchandCollegeofEngineering,Sangli,

Maharashtra,INDIA,Email:[email protected]

2Professor,MechanicalEngineering,WalchandCollegeofEngineering,Sangli,

Maharashtra,INDIA,Email:[email protected]

ABSTRACT

Nowadaysforthecycletimereductionintheindustriessixsigmamethodologyisvery

famous and helpful. Also it is a systematicmethodology tomove towards defect less

processes or production. It uses a detailed analysis of the process to determine the

proposesofcycletimereductionandcausesofcycletimedeviation."Define–Measure‐

Analyze–Improve–Control" (i.e. DMAIC) is the one of the approach from the various

approaches adopted while following the six sigma methodology. It is the classic Six

Sigmaproblemsolvingprocess.However,DMAICisnotexclusivetoSixSigmaandcan

beusedas the framework for improvementapplications. Itusesadetailedanalysisof

the process to determine the causes of the problem and proposes a successful

improvement. Cycle time reduction is nothing but the process improvement. Process

improvementmeansthestudytheexistingprocessandmakingtheprocesschangesto

improve cycle time of production by keeping the quality of product, reduce process

costs, accelerate productivity and etc. Most process improvement work so far has

focusedondefectreduction,butthereisanotherpointforprocessimprovementworkis

cycletimereduction.Nowaday’sindustriesarefacingthedowntimeproblemsduring

the production hours due to some technical or nontechnical issue like Cycle time

Deviation. For the cycle time reduction through DMAIC approach there are some

statistical analysis toolsavailable suchasANOVA,RegressionAnalysis,EVOP,Process

Capability Study, Pareto Analysis, etc. This paper presents the cycle time reduction

usingtheDMAICapproach,astheDMAICprovedtobethemostpreferredtechniquefor


ISSN: 2348-4098


thedefectidentificationandprocessimprovementbyuseofvariousstatisticaltools.In

thisstudythemajorproblemwasdowntimeoccurredonthefurtheroperationsinthe

period of last seven months. The 430 hours downtime was occurred in the seven

months due to this the overall efficiency of the face grinding process is get down to

42%. Initially theoverallefficiencyof facegrindingprocess iscalculatedbasedonthe

machineutilizationpercentageand themachineproductivityover theavailablehours

forproduction.ThentheParetoanalysiswasusedtodetect thecritical issuescausing

thedowntimeandfurthertheysolvedthroughtheDMAICapproach.

INDEX TERMS: Six Sigma, DMAIC, Cycle Time Reduction, Downtime, Grinding

Allowance,CycleTimeDeviation

1. INTRODUCTION

In real several manufacturing areas at

present, real challenges are arising for

the cycle time improvements of the

manufacturing process or operation,

also the challenges in quality

improvements of the products,

efficiencyimprovementofthemachines,

machineutilizationimprovement,etc.to

do such improvement Six Sigma

methodology isveryhelpful, andoutof

all the six sigma's approaches the

DMAIC approach (Define–Measure–

Analyze–Improve–Control) is very

helpfulforsuchsituation.

Six Sigma is a well‐structured

methodology that focuses on reducing

the various defects occurring in the

processesaswellasintheproducts.Six

Sigma methodology was originally

developed byMotorola in 1980s and it

targetedadifficultgoalof3.4partsper

million defects. Six Sigma has been on

an incredible run over 25 years,

producing significant savings to the

bottom line of many large and small

organizations. Six Sigma was initially

introduced inmanufacturingprocesses;

today, however,marketing, purchasing,

billing, invoicing, insurance, human

resource and customer call answering

functionsarealso implementingtheSix

Sigma methodology with the aim of

continuously reducing defects

throughout the organization’s

processes[1]. Six Sigma methodology

have two main methodologies DMAIC

and DMADV. Define, Measure, Analyze,

Improve, and Control (DMAIC)


ISSN: 2348-4098


methodology was followed for process

improvement and DMADV (Define,

Measure, Analysis, Design Verify) was

followedforproductimprovement.

Processimprovementisnothingbutthe

cycle time reduction because the

process improvement means

understanding of an existing process

and introducing process changes to

improve quality of product, reduce

costs, overall efficiency of process or

accelerate productivity. Generally the

overallefficiencyofmachineorprocess

is calculated based on the machine

utilization percentage and themachine

productivityovertheavailablehoursof

production.

1.1DMAICVSDMADVAPPROACH

Despite the shared first three letters of

their names, there are some notable

differences between them. The main

differenceexistsinthewaythefinaltwo

stepsof theprocess arehandled. With

DMADV, the Design and Verify steps

deal with redesigning a process to

match customer needs, as opposed to

the Improve and Control steps that

focus on determining ways to readjust

and control the process. DMAIC

typicallydefinesabusinessprocessand

howapplicableitis;DMADVdefinesthe

needsofthecustomerastheyrelatetoa

serviceorproduct.

With regards to measurement, DMAIC

measures current performance of a

process while DMADV measures

customer specifications and needs.

Control systems are established with

DMAIC in order to keep check on the

business’ future performance, while

with DMADV, a suggested business

modelmustundergosimulationteststo

verifyefficacy.

DMAIC concentrates on making

improvements to a business process in

order to reduce or eliminate defects;

DMADV develops an appropriate

business model destined to meet the

customers’requirements.

1.2DMAICAPPROACH

DMAIC is similar in function such as

Plan‐Do‐Check‐Act and the Seven Step

methodofJuranandGrynaforproblem

solving approaches. In the theory of

organizational routines, DMAIC is a

meta‐routine: a routine for changing

established routines or for designing

new routines. DMAIC is applied in

practice as a generic problem solving

and improvement approach [2].DMAIC

should be used when a product or

processisinexistenceatacompanybut


ISSN: 2348-4098


isnotaspercustomerspecificationsor

is not performing adequately. DMADV

should be used when a product or

processisnotinexistenceandoneneed

to be developed or when the existing

product or process has been optimized

and still does not meet the level of

customer specification or six sigma

level.

1.3 ADVANTAGES OF DMAIC

APPROACH

Can realize genuine cost savings:

DMAICisaparticularlyastutemeansof

identifying waste and unnecessary

rework. A successful DMAIC

implementation can pay for itself

severaltimesoverbygreatlyincreasing

theeffectivenessofaprocess.Thecycle

ofDMAICisreusabletoobusinessescan

continually repeat the process,

identifying further enhancements and

improvementsovertime.

Structured thinking: The DMAIC

process is systematic and thorough. It

enables decisions to bemade based on

actual data and measurement. The

varioustoolsandtechniquesusedinthe

analysis phase can flush out problems

and issues that might not have been

exposed otherwise and the approach

often brings a freshway of thinking to

establishedprocesses.

Looks at the longer term: DMAIC

implementation is seldom about quick

fixes.Theapproachlendsitselftolonger

term process resolution so for

established businesses or businesses

withparticularlycomplicatedprocesses,

DMAICworks verywell. Many projects

toywith a problem, implement a quick

fix and then walk away. The control

phase of the DMAIC methodology

ensuresthatthisneverhappens.

2. GRINDINGOPERATION

The study was conducted at a leading

manufacturer of Bearings of DGBB

(Deep Grove Ball Bearing), TRB (Taper

Roller Bearing) types. Fig.1 shows the

types of the bearing rings. In firm the

turned rings as a raw material is

processed with operations like Heat

treatment, Face Grinding, OD Grinding,

Bore Grinding, Track Grinding and

Honing and then assembly. TheCritical

operations in the firm is Face grinding

as there is any amount of less

productivity is occurs itwill make big

no material downtime on the channels

on which the further processes are

carried out. Fig.2 shows the 3D bone

structure of the DDS face grinding


ISSN: 2348-4098


machine present at the firm. The DDS

face grinding machine means Double

Disk face grinding machine, this DDS

machine contains two vertical spindles

with two grinding wheels placed as

shown inFig.2.Thereare twopressure

plates placed one at entry side and

another at exit side, also there are two

guiderailspresenttoguidetheringflow

fromentrytoexitside.Thetwogrinding

wheels are rotated opposite to each

other.

(a) (b)

Figure1:(a)DeepGroveBallBearing,(b)Taper

RollerBearing

Forfacegrindingoperationofinnerand

outer rings of both Deep Grove Ball

BearingsandTaperRollerBearings,the

+0 to ‐50µmtoleranceon thewidthof

the rings is allowable. To achieve this

tolerance the grinding allowance is

provided on the width i.e. excess

material is provide on the face side of

the bearing rings, it is of +150 to+250

µm for each type of bearing rings. This

excessmaterialisattheturnstageafter

theheattreatmentgrowththisgrinding

allowance goes to +200 to +300 µm.

Thisexcessmaterialisremovedthrough

face grinding operation. This face

grindingoperationisdoneinnumberof

passeswith one final finishing pass. As

per the machine capability and to get

reliable quality from the process the

machinecanremove~250µmatonce.

Figure2:BoneStructureofDDSfacegrinding

machine(DrawninCATIA‐V5)

Ifthetargetsizeisachievedin3passes,

then out of that 3 passes the first two

are rough passes in which 100µm

material of width is removed and

OuterringFace

Inner ringFace

TopGrindingWheel

BottomGrindingWheel

RingsIn

RingsOut


ISSN: 2348-4098


remaining 40 to 50µms material is

removedinlastonei.e.infinishingpass.

3. BACKGROUND

The grinding process under

consideration is a special purpose

process whichwas specially developed

forformingthefacesofthebearingring.

The name of this special process is

knownasfacegrindingoperationonthe

bearing inner and outer rings. The

operations performed in the firm on

turned (i.e. raw ring) inner and outer

rings are heat treatment, face grinding,

OD grinding, Bore Grinding, Track

grinding and honing to track of the

innerandouter ringsof thebearing. In

firstoperationi.e.theHeattreatmentof

theturnedringisdoneandthentheface

and OD grinding is done by separate

machinesandforthefurtheroperations

rings goes on to the channels. For face

grindingDDS(DoubleDisk)and forOD

the CL‐46 (Center‐less Grinding)

machines are available. DDS grinding

machine has a two co‐axial vertical

spindles with horizontal ring through

feeding as shown in Fig.2. For such

specific continuous feeding of the

bearingring for facegrindingthere isa

specialfeedingunitisinstalled.

Inthefirmthereisat leastoneproduct

changeover is happened in a shift as

firmhasbatch typeproduction isdone.

Duringthestudyofthedowntimeofall

the processes, found that the DDS face

grinding machine has created the no

material down time on the next

processes i.e. on the channels because

DDSmachineitselfhadsomedowntime

problem.Thiswasaseriousproblemto

mate the delivery date. The Pareto

analysis is done regarding the Hours

lost in recent sevenmonths and itwas

foundthattheproductchangeovertime

andcycle timedeviationhasbottleneck

issues.

Thereforetheobjectiveofthestudywas

to minimize the product changeover

timeandreducethecycletimedeviation

without affecting the quality of the

product To solve these issues the Six

Sigma technique was selected. In this

paper out of the two bottleneck issues

thecycletimedeviationisfocused.The

cycletimedeviationremovedwithhelp

oftheDMAICapproach.

4. CYCLETIMEREDUCTION

As the study aimed at cycle time

reduction of the existing face grinding

process,DMAICapproach is considered

[3], it consists of five phases that are


ISSN: 2348-4098


namely: Define, Measure, Analyze,

ImproveandControl.

AnySixSigmaprojectstartswithDefine

phase and is defined based on the

customer requirement and company

strategyandmission[4].Measurephase

helps the project team member to

collect thedata related to problemand

begin the search for various causes of

theproblem.

Table‐1:RationalreconstructionoftheDMAIC

procedure

Define: Problem selection and benefit

analysis

D1.Identifyandmaprelevantprocesses.

D2.Identifystakeholders.

D3.Determine and prioritize customer

needsandrequirements.

D4.Makeabusinesscasefortheproject.

Measure: Translation of the problem

into a measurable form, and

measurement of the current situation;

refineddefinitionofobjectives

M1.SelectoneormoreCTQs.

M2. Determine operational definitions for

CTQsandrequirements.

M3. Validate measurement systems of the

CTQs.

M4.Assessthecurrentprocesscapability.

M5.Defineobjectives.

Analyze: Identification of influence

factors and causes that determine the

CTQs'behavior

A1.Identifypotentialinfluencefactors.

A2.Selectthevitalfewinfluencefactors.

Improve:Designand implementationof

adjustments to the process to improve

theperformanceoftheCTQs

I1. Quantify relationships between Xs and

CTQs.

I2.Designactions tomodify theprocessor

settings of influence factors in such a way

thattheCTQsareoptimized.

I3. Conduct pilot test of improvement

actions

Control: Empirical verification of the

project's results and adjustment of the

processmanagementandcontrolsystem

in order that improvements are

sustainable

C1.Determinethenewprocesscapability.

C2.Implementcontrolplans.

InAnalyzephase,thecollecteddataare

analyzed, causes found are analyzed

usingvariousdataanalysistoolsandthe

data is validated for Improvement


ISSN: 2348-4098


phase. Improvement phase helps in

finding solutions and implementing

them so that the problems can be

eliminated.InControlphase,thegainof

the project is sustained. The

performance of the process after

improvementismeasuredroutinelyand

accordingly adjustments are made in

operations. If the Control phase is not

implemented, itmay revert the project

toitspreviousstates[5].Table‐1.Shows

the flow diagram of the DMAIC

approachwithitsfivemainphases.

In the study presented, the six sigma's

DMAICapproachisappliedtodiagnosis

the probable bottleneck issues of face

grinding operation downtime in

machine performance and successfully

reduced one of the issue. In proposed

studyonlythesecondissueofcycletime

deviation is considered for the

improvement. The following sections

explainthemethodologyappliedforthe

purpose[6].

4.1DEFINEPHASE

Define is the first phase of the DMAIC

methodologyofSixSigma.Thepurpose

is to define the project team’s

understanding of the problem to be

addressed and the output is stated in

the project charter. In the charter, the

team normally indicates the objectives

of theproject,expectedtimeline,scope,

andmembersof the team.Also created

during thisphase is a suppliers, inputs,

process, outputs, customers (SIPOC)

diagram that identifies the process

being examined, the inputs to and

outputsoftheprocess,andtherelevant

suppliers and customers to ensure that

teammembersacquireabird’s‐eyeview

oftheproject.Anotherimportantaspect

of the define phase is the gathering of

voice of the customer data. The Six

Sigmaprojectteamisfocusedonfinding

out directly from customers what they

wantandhowwell thecurrentprocess

meetstheirneeds.

Problem Statement: Selected firm is

theleadingbearingmanufacturerinthe

country and is known for its quality

bearings. But currently due to some

internal production efficiency loss

companyfacingthedowntimeproblem.

FaceandODGrindingdepartmentisone

of the low efficient department in the

firm.The firmworks for24by7hours

withthreeshiftsfirstandsecondeachof

8 hours and third shift of 7.3 hours, so

the total working hours for seven

months are ~5000 Hrs. For the recent

sevenmonthsJan2013toJuly2013due

to low efficiency at Face and OD


ISSN: 2348-4098


grinding department creates the

downtimeof430Hrsoutof4939Hrsat

the channels on which further

operations are carried out. Downtime

wasnearly~9%.

Efficiencylosslargelydependsuponthe

performance of the process. Hence,

processimprovementshavetobedone.

By doing this we can reduce the

downtimeoftheotherchannels.

Key Objectives: The Main two key

objectives to solve the cycle time

deviation problem of the DDS Cell face

grindingmachineareasfollows:

Grindingallowancereduction.

ReductioninGrindingPass

4.2MEASUREPHASE

The measure phase establishes

techniques for collecting data on the

current performance of the process

identifiedinthedefinephase.Themain

objective is to collect data pertinent to

thescopeoftheproject.Leaderscollect

reliable baseline data to compare

against future results. Teams create a

detailedmapofallinterrelatedbusiness

processes toelucidateareasofpossible

performance enhancement [8,9]. This

phase is used to determine sources of

variationandservesasabenchmarkto

validate improvements. A detailed

processmapisalsocreatedinthisphase

together with indications of possible

variationsexistingwithintheprocess.

In the proposed study, first find the

bottleneck machine of the Face‐OD

grinding department for that studied

the recent seven months efficiencies.

See the Table‐2. This gives the month

wiseefficienciesofalltheFacegrinding

machinesofthedepartment.

From the Table. I the efficiency of face

grinding machine DDS Cell is nearly

20% less than the other face grinding

machines can conclude thatDDS cell is

one of the major bottleneck from Face

and OD department. So DDS cell is the

first Bottleneck for the downtime

problemfromFaceandODdepartment

inthefirm.


ISSN: 2348-4098


Table‐2:Monthwiseefficienciesoftheface

grindingmachines

M/

C

No.

MonthEfficiency(in%)

Ja

n

Fe

b

M

ar

Ap

r

M

ay

Ju

nJul

DD

S

Cell

41.

4

%

42.

5

%

46

%

44.

1

%

42.

7

%

44.

2

%

41.

2

%

DD

S

544

66.

2

%

61.

7

%

64.

2

%

60.

3

%

62.

0

%

59.

6

%

69.

9

%

Gar

dne

r

101

6

58.

4

%

55.

8

%

60.

1

%

63.

5

%

60.

8

%

65.

2

%

63.

5

%

Gar

dne

r

160

1

66.

5

%

58.

4

%

64.

8

%

59.

8

%

55.

2

%

62.

5

%

63.

8

%

4.3ANALYSISPHASE

The purpose of the analyze phase is to

allow the project team to target

improvement opportunities by taking a

closerlookatthedatatodeterminethe

rootcausesoftheprocessproblemsand

inefficiencies. This involves discovering

why defects are generated by further

probing into the key variables

(identified in the previous measure

phase) that are most likely to cause

processvariation.Statisticalanalysisisa

key component of this phase and used

to demonstrate and confirm these

relationships.

TheAnalyzephasedeploysanumberof

tools for collecting team input and

conducting objective experiments to

identify or confirm top causes. The

mostcommonlyusedoftheseare‐

ParetoChart

FishboneDiagram

5‐Why

HypothesisTesting

RegressionAnalysis

TimeSeriesPlots

Multi‐VariAnalysis

Histograms

ScatterDiagrams

TreeDiagrams

PFMEA

Pareto analysis of bottleneck machine

DDS cell form Jan‐July 2013 is carried


ISSN: 2348-4098


out to find the hours lost as there is a

430Hrs downtime in Face‐OD grinding

department. Chart‐1. shows the pareto

graphformachineDDSCell.Thepareto

graphdrawn forhours loston theDDS

cellfacegrindingprocessisdrawnusing

the"MINITAB16"software.

FromParetoanalysisitisobservedthat

20% of the activities causing the 80%

effect formaking theDDS cellmachine

bottleneck. The three main hours lost

reasonsareasfollows:

New Type Setting (Product

ChangeoverTime)

CycleTimeDeviation

DressingOperation

Hence to improve the performance of

the DDS cell it is required to work on

thesethreecauses.

Housr lost 7.0922.2 596.5 432.3 123.9 100.3 50.7 37.4 27.3Percent 0.340.1 26.0 18.8 5.4 4.4 2.2 1.6 1.2Cum % 100.040.1 66.1 84.9 90.3 94.7 96.9 98.5 99.7

ActivityOthe

r

No M

ateria

l

Wheel

Chan

ge

Quali

ty Ad

justm

ent

Mainten

ece

No O

perat

or

Dres

sing

Cycle

Time D

eviat

ion

New Ty

pe S

etting

2500

2000

1500

1000

500

0

100

80

60

40

20

0

Hou

sr lo

st

Perc

ent

Pareto Chart of Activity for Jan to July 2013

Chart1:ParetoChartfortheDowntimeofDDS

facegrindingmachine

4.4IMPROVEPHASE

The main objective at the end of this

stage is to complete a test run of a

change that is to be widely

implemented. Teams and stakeholders

devisemethods to address the process

deficiencies uncovered during the data

analysis process. Groups finalize and

testachangethatisaimedatmitigating

the ineffective process. Improvements

are ongoing and include feedback

analysisandstakeholderparticipation.

In the proposed study, the cycle time

reduction issue is take for the

improvement. TheCycle timedeviation

means the number of hours required

more than that of the standard hours

required to produce the same quantity

of rings. The cycle time deviation is

given in terms of hours lost, from the

Paretoanalysisitisseenthatnearabout

26%oftotaldowntimeisoccurreddue

to the cycle time deviation on the DDS

facegrindingmachine.

There are the two ways to reduce the

cycletimedeviationofthefacegrinding

process. One is to optimize the input

parametersofthemachinetogetproper

production output rate. The input

parameters of the DDS face grinding

machineare ring feeding rate (m/min),


ISSN: 2348-4098


topandbottomgrindingwheelvelocity

(rpm) and top‐bottom grinding wheel

compensation (µm). With help of the

Taguchi's design of experiments

method, the numbers of runs are

performed on process or machine and

by analyzing the result one can fix the

input parameter to prevent the cycle

timedeviationof theDDS facegrinding

machine.

Second way to prevent cycle time

deviation is to reduce the facegrinding

allowancepresenton thebearing inner

and outer rings. This reduction can be

doneonthebasisoftheheattreatments

growthof thebearing ring in itswidth.

In the proposed case study this second

way is chosen to reduce the cycle time

deviation. As in the firm the face

grindingisdoneinthe2‐3passes,hence

byreducingthegrindingallowance,one

can reduce the number of passes

required to manufacture finish ring

indirectly theproduction time required

per pass is get eliminated and the

standardproductionrateisachieved.

Table‐3:Dimensionalchangesofthe61902

bearingtypeinnerandouterringsafterthe

grindingallowancereduction

Type

61902Status

Face GAGA

reductio

nFace

(mm)

Turn

Face

Size

(mm)

Finish

Face

size

(mm)

Min

(mm)

Max

(mm)

OR

Actual 7.300 7.000 0.200 0.300 ‐

Propos

ed7.180 7.000 0.130 0.230 0.120

IR

Actual 7.300 7.000 0.200 0.300 ‐

Propos

ed7.180 7.000 0.130 0.230 0.120

TheinnerandouterringofDeepGrove

Ball Bearing (DGBB) type61902 is one

of the bearing type had the cycle time

deviation in face grinding operation.

Cycle timedeviationmeans, toproduce

finish rings of this type requires more

time than that of the standard

production time. Initially there were

three passes to be done to meet the

tolerance of 0 to ‐100µm on width of

size 7 mm. The old width grinding

allowances on inner ringwas +300 µm

and on the outré ring +300 µm. This

grindingallowanceistheadditionofthe

turned ring grinding allowance pulse

heattreatmentgrowth.


ISSN: 2348-4098


Asmachinecarried the threepasses, in

first pass material removed was

~150µm, in second pass was ~100µm

and in last finish pass was ~50µm

materialwas removed.By reducing the

face grinding allowance the number of

passes gets reduced. Table.3 shows the

dimensional changes of the 61902

bearingtypeinnerandouterringsafter

thegrindingallowancereduction.

4.5CONTROLPHASE

The objective of the last stage of the

methodology is to developmetrics that

willhelpleadersmonitoranddocument

continued success. Six Sigma strategies

areadaptiveandon‐going.Adjustments

can bemade and new changesmay be

implemented as a result of the

completion of this first cycle of the

process. At the end of the cycle

additional processes are addressed or

theinitialprojectisthencomplete.

After completing the Improve phase,

factorsaffectingthecycletimedeviation

of the face grinding process on the

bearing inner and outer rings were

proposed. The actions proposed were

implemented in the manufacturing

process. The results of these

improvements were monitored in

Control phase. A control plan was

prepared which is the major action of

this phase. This control plan consisted

ofalltheactionsthatwereproposedfor

reducingthecycletimedeviationofthe

DDS face grindingmachine. It included

training and certifying the operators,

employees, maintenance plan

preparation, regular inspection, and

preparation of control charts. And thus

from Fig.4 it can be observed that the

goal set of reducing or preventing the

downtime bottleneck issues were

achieved.

5. CONCLUSION

Industries have to deal with a host of

problems related to productivity and

quality control. Substandard

productivity hampers the internal

customerdemandoftheproductswhich

directly affects the company targets.

Organizationshavetosufferhugelosses

which are not easy to cope up with.

Thus there is a need to improve the

processsimultaneouslykeepinginmind

the quality and the productivity of the

product. Six Sigma can be effectively

applied and the existing business

processes can be improved and made

error free, downtime free. Six Sigma

provides statistical proof to each and

every action, thus helping making

decisions more efficient. It can work


ISSN: 2348-4098


evenwithlessnumberofreadingsinthe

database. Thus Six Sigma is completely

an industry oriented methodology of

qualityandproductivityimprovement.

In the presented study the downtime

was much higher, i.e. ~9% of total

workinghoursoffirmforsevenmonths.

The firm had to sustain the downtime

costand thewastageof theman‐hours.

Establishing the relationship between

the issues for the downtime and the

effectof these issues isachallenge ina

complex system like the one discussed

above. The decision of using Six Sigma

methodologyprovedtobefacile.Pareto

graph was implemented to find all the

key issues that are causing the

downtime. Thus there was significant

improvement in the productivity and

lossesthefirmincurred.

Table5.3:Numberofpassesreduceddueto

grindingallowancereduction

Bearing

Type

GrindingAllowance No.ofPasses

Before After Before After

OR‐61902200 to 300

µm

180 to

210µm3 2

IR‐61902200 to 300

µm

180 to

210µm3 2

Table.4showsthecyclereductionofthe

61902 bearing type i.e. from now

onwardswhenthistypeiscomeforthe

production it requires only 2/3rd time

of the old standard production time,

meansthe1/3rdcycletimereductionis

achieved.

ABBREVIATIONS

DMAIC: Define, Measure, Analyze,

Improve,Control

DMADV: Define, Measure, Analyze,

Design,andVerify

DGBB:DeepGroveBallBearing

TRB: TaperRollerBearing

DDS: DoubleDisks

GA: GrindingAllowance

REFERENCES

[1] E. V. Gijo, Johny Scaria and Jiju

Antony, “Application of Six Sigma

Methodology to Reduce Defects of a

Grinding Process,” Published online in

Wiley Online Library,

(wileyonlinelibrary.com) DOI:

10.1002/qre.1212.

[2] De Mast, J., and Lokkerbol, “An

analysisoftheSixSigmaDMAICmethod

from the perspective of problem

solving,” International Journal of


ISSN: 2348-4098


Production Economics 139(2), 2012,

p.p.604–614.

[3] U. D. Kumar, D. Nowicki, J. E.

Ram´ırez‐M´arquez, and D. Verma, “On

the optimal selection of process

alternatives in a six sigma

implementation,” International Journal

ofProductionEconomics,vol.111,no.2,

pp.456–467,2008.

[4] M. Swink and B. W. Jacobs, “Six

sigmaadoption:operatingperformance

Impacts and contextual drivers of

success,” Journal of Operations

Management,vol.30,no.6,pp.437–453,

2012.

[5] E. V. Gijo, J. Scaria, and J. Antony,

“Application of six sigma methodology

toreducedefectsofagrindingprocess,”

Quality and Reliability Engineering

International, vol. 27, no. 8, pp. 1221–

1234,2011.

[6] D.Starbird,“Businessexcellence:Six

Sigma as a management system,” in

Proceedings of the Annual Quality

Congress, pp. 47– 55, Milwaukee,Wis,

USA,May2002.

[7] What is Six Sigma,

http://www.isixsigma.com/new‐to‐

sixsigma/getting‐started/what‐six‐

sigma/.

[8] G. W. Frings and L. Grant, “Who

moved my sigma ...effective

implementation of the six sigma

methodology to hospitals,” Quality and

Reliability Engineering International,

vol.21,no.3,pp.311–328,2005.

[9] R. McAdam and A. Evans,

“Challenges to six sigma in a high

technology mass‐manufacturing

environments,” Total Quality

Management and Business Excellence,

vol.15,no.5‐6,pp.699–706,2004.

BIOGRAPHIES

Alok B. Patil is currently

studentofsecondyearM.Tech

(Mechanical Engg.

specialization Production

Engineering). He is completed

his graduation in Mechanical

Engineeringin2011fromPune

University,Maharashtra,India.

Dr. Kedar H. Inamdar is

working in Department of

Mechanical Engineering,

Walchand College of

Engineering, Sangli. He has

published more than 80

technical papers in various

national / international

conferencesaswellasjournals.

Hisareaofinterestisinquality

controlandheacquiredpatent

relatedtoit.


ISSN: 2348-4098


PRODUCTIVITYIMPROVEMENTOFAUTOMOTIVEINDUSTRY

USINGLEANMANUFACTURING

1SWAPNILT.FIRAKE,2DR.KEDARH.INAMDAR

1Student,DepartmentofMechanicalEngineering,WCE,Sangli,Maharashtra,INDIA,


2Professor,DepartmentofMechanicalEngineering,WCE,Sangli,Maharashtra,INDIA,


ABSTRACT

Leanmanufacturingisdefinedasasystematicapproachtoidentifyingandeliminating

wastethroughcontinuousimprovement,flowingtheproductatthepullofthecustomer

inpursuitofperfection.Theidentificationandmeasurementofbestpractices, inLean

Production implementation, followed by the evaluation of its usage level, in the

organizations,aretheadequatewaythroughtheeliminationorminimizationofwaste.

However, the lack of a coordinated and structured roadmap, in the Lean Production

implementation, may result in poor and disappointing results. In that sense, it is

important to identify thestepsrequiredtoassess thestagesofcompanies towardthe

LeanProductionsystem.

The automotive industry under study includes assembly, testing and pre‐dispatch

inspection department. Kaizen improvements and 5S are the two lean tools that are

takenintoconsiderationforimprovements.Thedataiscollectedforthetimestudyand

analyzedwith the leanmetrics. Line balancing of production line is done in order to

removetheunnecessarystepsandthusshortentheleadtime.Theleanmanufacturing

reducestheleadtimeandalsoincreasesthequalityoftheproduct.

INDEXTERMS:Productivity,LeanManufacturing,Linebalancing,Kaizen,5S.


ISSN: 2348-4098


1. INTRODUCTION TO LEAN

MANUFACTURING

Lean manufacturing is one of the

initiatives that many major

manufacturingplants inAsia,especially

inMalaysiahavebeentryingtoadoptin

order to remain competitive in an

increasingly competitive globalmarket.

The focus of the approach is on cost

reductionthrougheliminatingnonvalue

added activities via applying a

management philosophywhich focused

on identifying and eliminating waste

from each step in theproduction chain

respective of energy, time, motion and

resources alike throughout a product’s

value stream, known as lean. Since the

birth of Toyota Production System,

manyofthetoolsandtechniquesoflean

manufacturing (e.g., just‐in‐time (JIT),

cellularmanufacturing, totalproductive

maintenance,single‐minuteexchangeof

dies, production smoothing) have been

extensively used. This activity is more

towards to Toyota Production System

(TPS),asystematicapproachtoidentify

and eliminate waste activities through

continuous improvement. All these

effort is objectively to keep cost down

andstayaheadintherace.

Industrial organizations have

increasingly sought to optimize the

resources needed for the manufacture

of itsproducts fromthecompetition, in

order to maintain their profit margins.

Thesearchforbalanceofresourcesand

balanceddistributionoftasksinvarious

types of industrial environments is

calledbalancing.Whenadjustmentsare

madeandadequacyofanassemblyline

thatisalreadyinoperation,thisprocess

is called rebalancing. Productivity of a

manufacturingsystemcanbedefinedas

the amount of work that can be

accomplished per unit time using the

availableresources.

Lean manufacturing has emerged

relatively recently as an approach that

integratesdifferenttoolstofocusonthe

elimination of waste and produce

products that meet customer

expectations. It helps in reduction of

resourcesandpresentsbenefitssuchas:

reduced delivery time, reduced

inventory, bettermanagement and less

rework[1].

1.1LINEBALANCING

Line balancing (LB) is usually

undertaken to minimize imbalance

between machines or personnel while

meetingarequiredoutputfromtheline.

Line balancing is a tool to improve the

throughputofaworkcellor linewhich


ISSN: 2348-4098


at the same time reducing manpower

and cost needed. It is often used to

develop product based layout. LB job

descriptionistoassigntaskstoaseries

of connected workstations where the

number of workstations and the total

amountofidletimeareminimizedfora

givenoutput level. The line is balanced

if the amountofworkassigned to each

workstationisidentical.

Line balancing is commonly technique

tosolveproblemsoccurredinassembly

line. Line balancing is a technique to

minimize imbalance between workers

and workloads in order to achieve

required run rate. This can be done by

equalizing the amount of work in each

station and assign the smallest number

of workers in the particular

workstation.

Generally,LBtechniqueisusedbymany

companies to improve theproductivity,

decreases the man power, decreases

idletimeandbufferoreventoproduce

more than two products at the same

time.LBtechniqueisusedtoachievethe

minimization of the number of

workstations, theminimizationof cycle

time, the maximization of workload

smoothness and the maximization of

workrelatedness.

Linebalancingiscommonlytechnique

to solve problems occurred in

assemblyline.Basically,linebalancing

triestominimizeimbalancebetween

workersandworkloadinordertoget

higher efficiency. There are some

methods to solve line balancing

problem; Heigesson Birnie Method,

Moodie‐Young Method, Immediate

Update First‐Fit Heuristic, and Rank‐

and‐AssignHeuristic.

1.2‘5S’METHODOLOGY

It is one of the simplest tools of lean

manufacturing.5Sisasystemtohave

less waste, optimize quality and

productivity through maintaining an

orderly workplace and using visual

signs to achieve operational results.

The practice of 5S comes from first

letter of five Japanese words and

translates as: sort, set in order, shine,

standardizeandsustain.

i) Sort:isthefirst“S”andrefersto

sorting tools, equipments on the work

place, relocate or remove all

components that is unnecessary or not

usedoften.

ii) Set inorder:means“aplace for

everything and everything in itsplace”.

Itaimstoorganizetheworkplace.


ISSN: 2348-4098


iii) Shine: refers to clean the work

area. It involves improving the

appearance of the work area and

housekeepingefforts.Everythingshould

stayclean.

iv) Standardize: everyone in the

organizationmustbeinvolvedinthe

5S effort. 5S should be implemented

withthesamewaytoeverywhere.

v) Sustain:referstomakingsure5S

implementation is followed by the

personnel. 5S is a culture and it has to

beingrainedintotheorganization[2].

1. CRITERIAINLINE

BALANCING

Therearesomecriteriawhichshouldbe

considered in a line balancing process.

These are takt time, cycle time,

downtime and minimum number of

workstationswhichcanbeexplainedas

below:

A. TAKTTIME

Takt time is pre‐requisite procedure in

doing line balancing task. Takt time is

the pace of production that aligns

production with customer demand. It

showshowfasttheneedtomanufacture

product in order to fill the customer

orders. Producing faster than takt time

results in over‐production which is a

typeofwastewhereasproducingslower

than takt time results in bottlenecks

where the customerordersmaynotbe

filled in time. The takt time is

determinedbyusingEq.1.

dayperdemandCustomer

daypertimeAvailableTaktTime

(1)

B. CYCLETIME

Cycle time shows how often the

productionlinecanproducetheproduct

withcurrentresourcesandstaffing.Itis

an accurate indicator to represent of

how the line is currently set up to run.

Cycletimeistheexpectedaveragetotal

production time per unit produced. On

anassembly lineor in awork cellwith

multiple operators, each operator will

have his own time associated with

completingtheworkheisdoing.

Takt time and cycle time are definitely

not the same. Takt time represents the

maximum time allowed to meet the

customerdemandwhereascycletimeis

the actual time necessary for an

operator to perform an activity or

completeonecycleofhisprocess.Both

takttimeandcycletimearedetermined

by customer demand. Using Eq.2, we

can calculate the cycle time for one

enginecompleteassembly.


ISSN: 2348-4098


requiredoduction

timeproductionActualCycleTime

Pr (2)

C. DOWNTIME

Downtime can be defined as that time

that is non value added. It is often

related with the 7 wastes that are:

defects, overproduction, waiting,

transportation, unnecessary inventory,

unnecessary motion and inappropriate

processing.

D. MINIMUM NUMBER OF

WORKSTATIONS

Aworkstationisaphysicalareawherea

workerwithtools,aworkerwithoneor

more machines, or an unattended

machine performs particular sets of

worktogether.Numberofworkstations

working is the amount of work to be

done at a work center expressed in

numberofworkstations.

Minimumnumberofworkstation is the

least number of workstations that can

providetherequiredproduction.Actual

number of workstation is the total

numberofworkstationsrequiredonthe

entireproductionline,calculatedasthe

next integer value of the number of

workstationsworking[3].

2. THEORIES RELATED TO

LEANMANUFACTURING

Literature is studied for lean

manufacturing. Literature review gives

detail information about present

practices in lean manufacturing and

results of advanced researches all over

the world. Literature review not only

givesthehistoryofaparticularproblem

but also provides results of recent

researchesonthesame.

3.1ABRIEFHISTORYOFLEAN

Mention ‘lean’ andmost ‘lean thinkers’

willknowthatthis isareferencetothe

leanproductionapproachpioneeredby

Toyota but also the subject of The

MachinethatChangedtheWorld,abook

which first highlighted Japanese

production methods as compared to

traditional Western mass production

systems,italsohighlightedthesuperior

performanceof the former.The follow‐

on book, Lean Thinking: Banish Waste

andCreateWealthinyourOrganization

is equally a key step in the history of

lean as it summarizes the lean

principles which ‘guide action’. It also

coinedthephrase‘LeanProduction’.

3.2 RECENT THEORIES AND

PRACTICES

The recent researches are studied for

the lean manufacturing concept, line


ISSN: 2348-4098


balancing approach and 5S

implementation.

2.2.1 Leanmanufacturing

Jostein Pettersen suggested that the

Lean principles are applicable to any

industry. If this is correct, then the

Japanese should logically have

distributed the knowledge of these

principles throughout all domestic

Japanese industry. This does not seem

to be the case. The only ‘true’ Lean

producers in Japan are confined to the

automobile industry, represented by,

e.g.Toyota,HondaandMazda,whereas

other areas of industry are performing

at the same level as or worse than

western competitors. This was pointed

outthattheprinciplesconstitutingLean

Productionhavenotreceivedanywide‐

spread attention outside the auto‐

industry. He argues that the possibility

to become ‘Lean’ through JIT in

particular is highly dependent upon

businessconditionsthatarenotalways

met, thus limiting the ‘universality’ of

theconcept[4].

3.2.2Linebalancing

R. B. Breginski et al. explained the line

balancing methods for flexible

manufacturing processes. It includes

Heuristic Method of line balancing

whichnormallyusedforbalancingof

number of group activities to be

performed during operation and

explained the problem solution which

gives optimum output for such

problems. This method is very easy to

understand as well as implement into

actual analysis is of problem. No extra

expense is required to analyze the

problemaswellasfindingsolution[5].

Hudli Mohd. Rameez et al. explained

main purpose of implementing lean

manufacturing is to increase

productivity, reduce lead time and

cost and improve quality thus

providing the up most value to

customers. Lean Manufacturing is an

operational strategy oriented toward

achieving the shortest possible cycle

time by eliminating waste. Lean

manufacturing techniquesarebasedon

the application of five principles to

guide management’s action toward

success. Lean production method is an

effective way to improve management,

enhance the international

competitiveness of manufacturing

enterprises[6].

3.2.3 ‘5S’ methodology: P. M. Rojasra

and M. N. Qureshi demonstrate the

implementation of 5S as lean

manufacturing technique in small scale


ISSN: 2348-4098


industry. Leanmanufacturing is one of

theoptions toreducenon‐value‐added‐

activity or waste and improve

operational efficiency of the

organization. The efficient

implementationof5Stechniqueleads

to subsequent improvement in

productivity of the manufacturing

plant.The5Simprovesenvironmental

performanceandthusrelateprimarily

in reduction of wastes in

manufacturing. It promotesneatness in

storage of raw material and finished

products.The5Simplementationleads

to the improvement of the case

company organization in many ways

[7].

3. CASESTUDY‐LEAN

IMPLEMENTATION

In order to increase the productivity,

theautomotiveindustrydecidedtotake

initiative of lean implementation. The

Greaves Cotton Limited is one of the

leadingindustriestomanufacturesingle

cylinder diesel engines. The case study

in this paper is regarding the

manufacturing of 3‐wheeler TML

engine. The different departments

include assembly department, testing

departmentandpre‐dispatchinspection

department. The data is collected and

analyzedforalldepartments.

The following assumptions are used to

definetheproblem:

a) The assembly line is never

starved,

b) Set‐up times are not taken into

consideration.Because ina real system

the setup process is usually

accomplishedattheendoftheworking

time,

c) No maintenance process is

performedduringtheworkingperiod,

d) Transportation of raw materials

is performed by workers who aren’t

usedforassemblyoperations.

The method used for improvements is

showninfollowingFig‐1.

Figure1:ProcessImprovementFlowDiagram


ISSN: 2348-4098


4.1 COMPANY AND PROCESS

BACKGROUND

The Industry currently is one of the

leadingindustriestomanufacturesingle

cylinder diesel engines. The industry is

heading towards becoming the world

class leader. They have current peak

capacity of assembly of 225

engines/shift, testing of 154

engines/shift and PDI of 200 engines.

The industry is seeking to increase its

capacity so that it can satisfy the

increased demand of the existing

customer in the future and also seeks

the other customers to bring towards

them.

In automotive industries, themain aim

whileincreasingproductivityisfoundto

be the increase in number of output

units that are manufactured. The lean

implementation also takes care of

quality of the product that is

manufactured.Thelinebalancingofthe

assemblylineistheinitiatetowardsthis

fulfillment.

Some of the tools used are kaizen

improvements, pokayoke, motion

reduction,transportationreduction,line

balancing,5S implementation, takttime

etc. The steps include process review

and data collection, data analysis,

observations and data collection after

the improvement, results and

discussions.

4.2ENGINEFLOWDIAGRAM

Thisdiagramshowswhichsequencethe

product, engine, flows from one

departmenttoanotherdepartment.The

engine is assembled in the assembly

department on the conveyor. The

assembled engine is tested in engine

testingdepartment.Thetorquesettings

are done in this testing and also the

engine is checked for any leak while

actual running of the engine. The OK

engine from testing department then

transferred to pre‐dispatch inspection

(PDI) department. In this department,

all the accessories,markings are added

to engine and also the tappet settings

aredone.

Figure2:EngineFlowDiagram

EngineAssemblyDepartment

This is the first and the main area of

concern towards the lean

implementation. There are 30‐Online

workstationsand9‐Offlineworkstations

ontheconveyor.Theconveyoravailable


ISSN: 2348-4098


is power and free conveyor which is

already installed. From the data

analysis, the target in the productivity

improvement is set for assembly of

engines,whichis16%improvement.

ProcessreviewanddatacollectionPrior

to Lean Production implementation, a

process review on TMLwas conducted

toinvestigatetheexistingmethodof its

actual assembly processes through

direct observation. Hardcopy

information on actual manufacturing

activities is based on their Operations

Manuals and the Standard Operating

Procedure (SOP). The Cycle Time or

Processing Timemeasured is observed

by the video taken from the ongoing

process of the assembly in order to

establish the baseline for data analysis.

Further to that, line observation was

conducted tomonitor and to grasp full

understanding on the current practice

at the assembly line as well as to

identifytypesofwastesintheprocess.

The engine is assembled on conveyor

which is already available. The

conveyor used is power and free

conveyor. The following data is

collected as before improvement data.

The issues that are found out from the

presentassemblylineareasfollows:

a) The activities contain value‐

added as well as non‐value‐added

activities. Non‐value‐added activities

aretakingtimethataddsnovaluetothe

final product. So the time required is

more and again there ismotion loss. It

leadstolowproductionrate.

b) Line is not balanced and one

station is taking too much time to

complete the set of activities that are

subjected to be done on that

workstation only. That leads to

bottlenecks and the next station is idle

fortheremainingtime.

c) Thereareidlestationspresentin

thelinethatarenotaddinganyvalueto

the product, thus leads to take more

timetoassembletheengineandleadsto

lowproductionrate.

d) The activities are taking too

much time than the actual time they

havetotake,sincetheycontainvarious

non‐valueaddedmotions.Thusthetime

taken for completion of thework for a

particularstationismore.

Table‐1 shows the detailed operations

intheassemblylinewiththeprocessing

times at respective stations. The

observations before the improvements

for assembly line related to total

workstations, total capacity of the line


ISSN: 2348-4098


andtheoutputpermanaremadewhich

areasbelow:

a) Numberofworkstations:40

(Online‐30,Offline‐10)

b) Capacity (Engines/Shift) = 225

Engines

c) Output/Man: Total Team

Associate(TA)=35

Hence, Output/TA = 225/35 = 6.42

Engines/TA

As it can be seen from the above data,

thereare30onlineworkstations.Outof

these, five stations are without Team

Associate. Out of these, on the two

stations there is In‐ProcessVerification

(IPV) setup is installed as a quality

check point and three stations are

idle/man‐lessstations.

a) DATAANALYSIS

For thedata analysis, total shift time is

taken 8Hour and 30minutes, i.e. 510

minutes. Excluding the unproductive

time like lunch and other time, the

available productive time is 450

minutes.


ISSN: 2348-4098


Table‐1:G600WIIIOperationsDetails

WSNo. Operationstobedone Timeinseconds

01 Loading of crank case, History card scanning and locate at identified

location74.4

02 Crankshaftfitment,starterplate,retainerplate 94.203 IPV‐01(AxialPlayMeasurement) 52

04 Fitmentofcamshaft,rollertappet&PRTsupport,FWEcover 123.6

05 Fitmentofcylinderheadstud&PTOcoverfitment 126

06 Barrelfitment,Piston&con.rodfitment 88.8

07 Fitmentofstrainer,studonadaptorforfilterfitment 122

08 Fitmentofoilpanwithloctite&pressureswitchonadaptor 68

09 Oilpantightening,LOFadaptorfitment 80

10 Positivelubricationpipefitment 114

11 Studfitmentonstarterplate 86

12 FWEtighteninginsequence,alternatorbracketstand,enginefeettrolley 85

13 IPV‐II(Torquetoturn) ‐

14 Flywheel&crankshaftpulleyfitment 81

15 Fitmentoffwaterpump&waterfeedpumppulley 81

16 FitmentofFIP,checkBDCwithdialgauge 126

17 Bumpingclearance&TDCmarking 75

18 PRT,Pushrod,Cylinderheadfitment 68

19 Cylinderhead tightening,DCNRtorquing,Tappetsetting 85

20 Waterfeedpumppulley,inletandexhauststudfitment 120

21 Thermostatcoverwithsealantandstopbracketfitment 79

22 Fuelfeedpumpandricovalvefitment 79

23 Rockerlubricationpipe,oilreturnpipefitment 73

24 FeedpumptoFIPpipeandoverflowpipefitment 87

25 Rockercover,Intakemanifoldfitment 78

26 Idlestation ‐

27 Engineairleaktest,highpressurepipefitment 121

28 Idlestation ‐

29 Idlestation ‐

30 Breather cap,nameplate fitment, oil dispensing, register entry, barcode

scanninganddeclareengineformovingtowardstestbed

84


ISSN: 2348-4098


FromEq.2, the cycle time is calculated

andwhichisfoundtobe2minutes.This

means that there is assembly of one

engine for every 2minutes that is 120

seconds. The time taken for pallet

movement from one station to another

station is found to be 13 seconds. This

shows that the available time for

working on the each station is 107

seconds.

Basedon this, the target is set for16%

improvement which gives 262 engines

per shift andwith the takt time of 103

seconds. Excluding the prior available

palletmovementtimebetweenstations,

the time available for completing the

work at each station is found to be 90

seconds,i.e.,theassemblylineshouldbe

balancedfor90seconds.Thisisthetakt

timeforassemblydepartment.

b) IMPROVEMENTS INASSEMBLY

LINE

The improvements done are done by

kaizen improvements and 5S

implementation on the assembly line.

Theseimprovementsreducethemotion

losses, waiting losses, transportation

losses, etc. Apart from these

improvements,pokayokeimprovements

arealsodoneasqualityimprovements.

KAIZENIMPROVEMENTS:

Thesearedoneforreducingfatigueand

motion losses. These improvements

help to reduce the time for completing

thetask.Thenon‐value‐addedactivities

ormotionsareeliminated/reducedwith

the help of proper kaizen

improvements.

Rebalancing is done by the proper

shifting or distribution of activities at

variousworkstationssuchthatidentical

time is required at all workstations to

complete the activities distributed on

them.

These are continuous improvements.

The basic idea of improvement is got

form actual walking on the assembly

line. The operator on the station

provides the idea of the improvement.

This helps to reduce the fatigue to the

worker.

Someofthekaizenimprovementsareas

showninfollowingTables.


ISSN: 2348-4098


Table‐2:KaizenImprovements(1)

KaizenObjective:Toreducedtimeandfatigue Idea: To provide location arrangement forgasketandTDCplate

Problem/Present status: More time and fatigueduringtakingofmaterial

Countermeasure: Gasket and TDC plate binprovidedinrightsideofoperator

Before After

Description: Material taking on back side (Timerequired=11sec)

Description: Materialtakingonrightside(Timerequired=7sec)

AsshowninTable‐2,thelocationforthe

bin ischanged.Thisreducesthefatigue

totheworker.Thelocationofthebinis

also placed at the good height and

distance.


KaizenObjective:Toreducedtimeandfatigue Idea: ToprovidearrangementforFIFOrack

Problem/Present status: More time and fatigueduringtakingofmaterial

Countermeasure: Small bin attached over FIFOrackatcomfortablelevel

Before After

Description: Material bin is not at comfortablelevel(Timereqd=09sec)

Description: Small bin provided at comfortablelevel(Timereqd=3sec)

As shown from Table‐3, it is seen that

the bin before was not at the

comfortablelevel.Thecomfortablelevel

ofbinreducesfatiguetotheworker.


ISSN: 2348-4098



KaizenObjective:Toreducedtimeandfatigue Idea: ToprovidearrangementforFIFOrack

Problem/Presentstatus:oretimeandfatigueduringtakingofsolenoidbracket

Countermeasure: Slant tray provided in thesideofoperator

Before After

Description:Solenoidbrackettakingfrombackside(Timereqd=8sec)

Description: Solenoidbrackettakingfromsideofoperator(Timereqd=4sec)

Table‐4 shows that before kaizen

improvement, the solenoid bracket has

to be taken from the back side of the

worker. After, the tray is provided by

thesideoftheoperator.


KaizenObjective:Toreducedtimeandfatigue Idea: Toprovidearrangementforbackplate&

airshroud

Problem/Present status: More time & fatigue

duringBOMissue

Countermeasure: Hangertypetrolleyprovided

forbackplate&airshroud

Before After

Description:BeforeBOMissueforbackplate,it

isinbox(Timereqd=09sec)

Description: Trolley for Air shroud & back

plate(Timereqd=6sec)


ISSN: 2348-4098


Table‐5 shows before kaizen

improvements, the back plate was

provided in box on the assembly line

whichcauses fatigueduringBOMissue.

The trolley is provided for air shroud

Andbackplate,reducesfatigueandtime

duringBOMissue.

5SIMPROVEMENTS

These are workplace related

improvements. These cause the best

utilizationoftheworkplace.

Table‐6:5'S'Improvements(1)

5S Objective: To improved 5's' and increaseworkingspace

Idea: Toprovidelocationarrangementfors/aofEGRvalve

Problem/Present status: 5's' notmaintained anddifficultyforworking

Countermeasure: Hanger provided for s/a ofEGRvalve

Before After

Description:S/aofEGRvalveonworkingtable Description: S\aofEGRvalveonhanger

Table‐7:5'S'Improvements(2)

5SObjective:Toimproved5's' Idea: Toprovidelocationarrangementforoilreturnpipefitmentgauge

Problem/Presentstatus:5's'notmaintained Countermeasure: location change and separatestandprovidedforgauge

Before After

Description: Gauge location andmaterial binlocationissamenotseparate

Description: separate location provided for gaugelocation


ISSN: 2348-4098


AsshowninTable‐6andTable‐7, these

are 5'S' improvements. The right

locationissetforallthematerials.After

usage, same is placed to its respective

location. These are out of some of the

5'S' improvements. Table‐8 shows the

time before (P) line balancing and the

timerequiredafter(Q)linebalancingat

allworkstations.Actiontakentoreduce

the timing at each workstation is also

includedinthetable.All thetimestudy

isdoneinseconds.

Table‐8:TimeBeforeandAftertheImprovementsandtheActionsTaken

WSNo. P Q ActionsTaken

01 74.4 85 Activityrebalanced for90sec02 94.2 88 Activityrebalancedfor90sec03 52 52 IPV‐1(EndFloat)Man‐less04 123.6 82 Layoutchanged&rebalancingdone

05 126 81.4 Motionlossreduced

06 88.8 81.6 Motionlossreduced

07 121.2 69 Motionlossreduced

08 67.2 82 Activityrebalancedfor90sec

09 79.2 67.8 Activityrebalancedfor90sec



12 85.2 85.8 Activityrebalancedfor90sec13 62 62 IPV‐2TorqueToTurn(Man‐lessstation)14 81.6 79.2 Motionlossreduced15 81 84 Activityrebalancedfor90sec

16 126 72 Motionlossreduced

17 81 66.6 Activityrebalancedfor90sec&motionlossreduction



20 120.6 69 Motionlossreduced

21 78.6 90 Activityrebalancedfor90sec&motionlossreduced22 78.6 78 Activityrebalancedfor90sec



25 0 82 Idle station used to utilize conveyor & Activityrebalancedfor90sec


27 121.2 73.2 Motionlossreduced

28 0 0 Idlestation29 0 0 Idlestation

30 84 72.6 Motionlossreduced


ISSN: 2348-4098


Thus,thetotaltasktimerequiredbefore

line balancing is 2426.4 seconds and

that required after line balancing is

2180 seconds. Fig‐3 below shows the

graph for above data for OnlineWS‐01

to WS‐30. WS‐28 and WS‐29 are idle

stations. The activities at all

workstationsarebalancedfor90secby

usingmotionlosses,changedlayoutand

rebalancingofactivities.

Figure3:TimestudyandTaktTime

As shown in above Figure 3 , the cycle

time isrebalanced for90secondsatall

workstations,whichistakttime.

c) DATA AFTER LEAN

IMPLEMENTATION

i) Productioncapacity:

Before Improvement = 225

Engines/Shift

AfterImprovement=262Engines/Shift

PercentageImprovement=(262‐225)/

225×100

=16%improvement.

ii) Production lead time: Time from

startofphysicalproductionoffirstsub‐

module/part to production finished

(readyfordelivery).

FromTable‐7, theproduction lead time

before was 2426.4 seconds and that

after line balancing is 2180 seconds.

Thusproductionleadtimeisalsofound

tobereduced.


ISSN: 2348-4098


iii) Product yield per employee or

Output/Man = 262/34 = 7.71 engines,

which was 6.67 engines before

improvement.

iv) It determines optimize use of

labor. It measures effectiveness of

manufacturingprocessandproductivity

ofemployee.Thus, in thiscasestudy, it

isfoundtobeincreased.

v) Additional 925 (= 37×25) engines

canbemadeinoneshiftbasisonly.

vi) With225engines/shiftwecanran

single shift up to maximum 5625

engines/month but with 262

engines/shift, we can achieve 6550

engines per month with the same

manpower.

d) Results Observed after Lean

Implementation

i) ProductivityImprovement

Increase in the number of engine

assembly leads to increase in the

productivity. Here, the number of

enginesassemblyisincreasedfrom225

to 262 engines per shift. Percentage

improvement observed is 16%

improvement. This shows that the lead

time is also reduced since there is

increase in the number of engine

assembliesinthesameamountoftime.

ii) LineEfficiency

Eq. 3 below shows the formula for

calculating the efficiency of the

assemblyline.FromTable‐7,addingthe

data for before line balancing, the

equationgivesthelineefficiencybefore

improvement[7].

TimeCycleestLnsWorkstatioofNumber

TimeTaskEfficiencyLine

arg

................(3)

LineEfficiency=2426.4/(30×126)=

64.19%

Now, adding the data for after line

balancing,Eq.3becomes,

Line Efficiency = 2180 / (30 × 90) =

80.74%

Thus, it can be seen that there is

improvement in line efficiency from

64.19%to80.74%.

I. G600WIII Engine Testing

Department

In this department, every assembled

engineistestedforitsperformanceand

thevarioussettingsarealsodonewhile

testing of the engine. Some of these

settingsincludesmaximumRPMsetting,

low RPM, idle RPM settings, rico valve

setting, etc. Testing of engine includes

loading of engine on the testbed then

running the engine and setting the


ISSN: 2348-4098


differentenginespecificationsand then

again unloading of the engine. Testing

departmentconsistsoftotal11testbeds

onwhichengineistested.

Following roadmap is prepared for the

improvementinthetestingdepartment.

i) VideotobetakenfromLoading+

Connection+Removal of connection to

unloading.

ii) Conduct time study (Loading +

Running+Unloading).

iii) Identify wastages/Improvement

opportunities.

iv) Implementkaizen.

v) Checkresults.

a) Data Collection before and

afterImprovement

The data included the activities to be

performed alongwith the time require

for those activities before and after

improvement.Theimprovementactions

taken are also included in the Table‐8.

The improvements done are mainly

kaizen improvements. The data

collection is done for the three steps,

loading of engine, running in cycle of

engineandunloadingoftheengine.

Table‐8:EngineTestingDepartmentSummary

Terms Before After Improvement

Loading(sec) 419 153 266

Runningcycle(sec) 1166 982 184

Unloading(sec) 344 139 205

Totaltime(sec) 1929 1274 655

Cycletime(min) 32.14 21.23 10.55

Totalhrworking (min) 450 450

Output per engineer(engine/engineer)

14 21.19 7

Testing capacity(engines/day)

447 670 223

b) DataAnalysisforObservations

BeforeandAftertheImprovements

Manpower required for 447

engines/day=2.8shifts/day.

Before required = 447/14= 32 TA

(approx.)

After required = 447/21.19 =21

TA(approx.)


ISSN: 2348-4098


Benefit it can be seen that, after the

improvements, required 11 Team

associates(TA)lessthanprevious.

Investmentfortest‐bedimprovement:

Per test cell investment (Rs. in lacs) =

Rs.0.5lacs.

Total test bed (11nos.) investment =

11×0.5=Rs.5.5lacs.

The required investment is regained

backwiththereductioninTA.

II. G600WIIIPDIDEPARTMENT

Thisisthelastdepartmentunderstudy.

This includes the per‐dispatch

inspectionofengine.Italsoincludesthe

addition of OK tags, markings, final

tappet setting, and applying anti‐rust.

The data collection, data analysis,

observations and data collection after

the improvements are the major steps

includedinthisstudy.

a) Data Collection and

Improvements

There are 10 Workstations, Operation‐

10 to Operation‐100. The following

Table‐9 shows time taken at various

workstations before and after

rebalancing and actions taken to

rebalancingthePDIline.

Table‐9:EnginePDIDepartmentSummary

WSNo.TimeRequired(sec)

ActionstakenBefore After

OP‐10 88 110 Linebalancedfor110sec

OP‐20 85 0 Retorquingeliminated

OP‐30 128 100 Motioneliminated


OP‐50 100 100 ‐


OP‐70 135 106 PTOretorquingeliminated


OP‐90 70 102 Linebalancedfor110sec

OP‐100 105 109 Paintmarkingeliminatedforoneplace


ISSN: 2348-4098


Figure4:TimestudyandTaktTime

Figure 4 shows the graph for the time

study for various workstations in PDI

department. It can be seen that, all the

workstations are balanced for the takt

time of 110 seconds. The operation‐20

is eliminated after the improvements

[8].

b) Observations after the

Improvements

i) Capacity improved to 270

engines/shiftfrom200engines/shift.

Percentage improvement = (270‐

200)/200×100

=35%improvement.

ii) Output per man is improved

from20nos.enginesto27nos.engine.

iii) Additional 1750 nos. of engines

per month can be made in one shift

basisonly.

iv) With 200 nos. engines/shift, we

can run a single shift upto max. 5000

engines/month, but with 270 nos. of

engines/shift,wecanachieve6750nos.

of engines/month with optimum

manpower.

5. CONCLUSIONS

Thisisconcludedthattheassemblyline

balancingisoneofthemajorsteptobe

taken into consideration while

increasing productivity of automotive

industries. Line balancing is done with

taking in account the takt time, cycle

time and downtime and thus reduces

theproductionleadtimewithincreased

number of output engines. Continuous

improvement is the step to reduce

unnecessary downtime losses. The

productivity of engine assembly line is

thus found to be increased. The testing


ISSN: 2348-4098


department and PDI department also

have some non‐value‐added activities.

Thosearealsoreducedoreliminatedby

the kaizen improvements and 5'S'

changes and the operation are

rebalanced taking in account the takt

time. The productivity of both testing

andPDIdepartmentsisalsofoundtobe

increased. Thus lean manufacturing

concept when deployed increases the

productivity. The primary lean tools

used are kaizen improvements and the

5S implementation. By using line

balancing and Lean techniques,

practitioners can better calculate the

timeandeffortneededtocompletetheir

products or services, and also utilize

theirresourcestothefullesttoproduce

theoutputdemandedbythecustomer.

6. REFERENCES

[1] Abhishek Dwivedi, "An analysis

and development of software for

assembly line balancing problem of

manufacturing industry, " VSRD‐MAP,

vol.2(2),pp.74‐87,2012.

[2] DuarteF.Gomes,ManuelPereira

Lopes and Carlos Vaz de Carvalho, "

SeriousGames for LeanManufacturing:

The 5S Game, " IEEE, vol. 8, no. 4, pp.

191‐196,2013.

[3] DanielKitaw,AmareMatebuand

Solomon Tadesse , "Assembly line

balancingusingsimulationtechniquein

agarmentmanufacturingfirm,"Journal

ofEEA,vol.27,pp.69‐80,2010.

[4] JosteinPettersen, “DefiningLean

Production: Some Conceptual and

Practical Issues,” The TQM Journal,

2009,vol.21,no.2,pp.127‐142.

[5] R.B.Breginski,M.G.CletoandJ.

L.SaasJunior,"AssemblyLineBalancing

using Eight Heuristics, " 22nd

Internatinal Conference on Production

Research.

[6] HudliMohd.RameezandDr.K.

H. Inamdar, “Areas of Lean

Manufacturing for Productivity

Improvement in a Manufacturing

Unit,” World Academy of Science,

Engineering and Technology, 2010, pp.

584‐587.

[7] P.M.RojasraandM.N.Qureshi,"

Performance Improvement through 5S

in Small Scale Industry: A case study, "

International Journal of Modern

EngineeringResearch,vol.3,issue3,pp.

1654‐1660,2013.

[8] Scholl. A, Balancing and

sequencing of assembly lines, 2nd ed.,

Physica,Heidelberg,pp.62‐63.


ISSN: 2348-4098


AUTONOMOUSUNDERWATERROBOTUSINGFPGA

1AKANKSHAGUPTA,2PINKYGUPTA,3KOUSHIKCHAKRABORTY

1M.TechScholar,DepartmentofECE,JayotiVidyapeethWomen’sUniversity,

Jaipur‐INDIA,Email:[email protected]

2,3AssistantProfessor,DepartmentofElectronics&CommunicationEngineering,

JayotiVidyapeethWomen’sUniversity,Jaipur‐INDIA

ABSTRACT

New perspectives have been opened by underwater exploration and whatever the

environmentis,theaimofroboticsistodeveloptoolsthatcanbeusedtofacilitatethe

work in real world domains. The robots under water are thus known as Remotely

OperatedVehicles(ROV)orAutonomousUnderwaterVehicles(AUV).Theyreducerisk

of offshore exploration and due to their smaller size; they can performmissions that

othercraftscannot.Suchtypesofrobotsaredesignedforaresearchprototypeplatform.

Therobotsareequippedwithcamerastomakecomputervisionsystem.Avisionsystem

thusresultsinmotionestimationandlocalizationofanunderwaterrobot.Thiscamera

providesrichinformationaboutrobotnavigationinsidethewaterbody.

KEYWORDS:AUV,UnderwaterVehicle,RemotelyOperatedVehicle,UnderwaterRobot

IntelligentSystem(URIS).

1. INTRODUCTION

In the fast growing world, real time

controllingofanyroboticapplication is

very important. The idea behind

developing the system is to make a

singlecontrolsystemtocontrolmultiple

robotic applications simultaneously

because individual controller for every

application is very difficult to

synchronize and development of logic

for every different application is also

verytimeconsuming.Sothesystemisto

be developed such that it can control

any robotic application from remote

area simultaneously in real time.

Communicationinterfacesarealsotobe

developed using which user can easily


ISSN: 2348-4098


control any application remotely. A

highly versatile architecture deals with

high‐levelreal‐timeprocessingroutines.

The hardware has been designed to

workco‐operativelywithahost,leaving

thehostfreetodealwiththefinalsteps

concerning scene understanding and

interpretationtasks.

2. MOTIVATIONBEHINDTHE

WORK

In this paper, the whole framework is

about underwater vehicle [1] but its

heart relies in the cameramounted on

it. The camera captures all rich

information underwater including the

localization and motion estimation.

Thesefeaturesincludethedetectionand

quality of images. One of the main

objectives of thiswork is to obtain the

rate performance for the execution of

tasks performed by the vehicle. This

procedure involves both hardware and

software. The work investigates the

possibility of accelerating parts by

means of hardware implementation.

This framework is used to give a new

derivation of VLSI and FPGA

architectures[2]whichhasbeenproved

to work better in underwater imaging.

The software part including FPGA can

achieve a much higher performance

while maintaining a high level of

flexibility.

3. OPERATIONALOVERVIEW

Foragivenapplication,weimplementit

using custom hardware or software

design.Buttofulfillalltherequirements

ofhostthebestchoiceistocombinethe

advantages of both hardware and

software.

3.1HARDWARE

Customizedtotheproblem

Relativelyfast

3.2SOFTWARE

Flexible,taskscanbemodifiedby

changing the instruction stream in

rewritablememory

Generalpurposecomputing

4. SYSTEMCOMPONENTS

4.1 UNDERWATER PLATFORM

(HARDWARE)

Underwater Robot Intelligent System

(URIS)hasbeendesignedinasmallsize,

andsousedasaresearchprototype[3].

It was build with the aim to perform

missions either in a controlled

environment suchaswater tanks,or in

naturalenvironmentslikewaterbodies.


ISSN: 2348-4098


Themainelectronicdevicesrequiredfor

a successful underwater robot are as

follows:

4.1.1IRSensor

The sensors will instruct the robot

about direction, clashing, or any other

sudden change. The IR sensor is

considered the best, otherwise

QRD1114 optical sensor will also be

suitablefortheproject.

4.1.2LM324comparatorIC

TheLM324comparatorICcomparesthe

inputsandgives thedigitaloutput.The

digital output from the comparator is

the signal used to control the motor

driver.

4.1.3 L293DMotorDriverIC

L293D IC controls the circuit by

controlling the motors [4] which are

used to drive the back wheels of the

robot independently. It can supply

600mA continuous and 1.2A peak

currents.MoreoveritconsistsoftwoH‐

bridgeswhichcontrols theswitchingof

thedevice.

4.2SYSTEMDESIGNING(SOFTWARE)

Testing and validation of the design is

describedinVHDL/Verilog[5]andthen

synthesized for the FPGA device. The

whole system is implemented using

Xilinx tool corresponding to every step

of design. The application is targeted

forFPGAdevicesduetovariousreasons,

sinceitassistsateverydesignstage.The

VHDLandVeriloglanguagesarechosen

for hardware design. The Xilinx

simulation tool is used for design

verification together with Spartan III

(seriesXC3S400)[6].Thegivenboardis

chosen due to its various features (like

480Mbpsdatatransmissionspeed,16X2

LCD module interface, 60 General

Purpose I/O's, 128 Kbit EEPROM etc.)

[7].

5. HARDWAREINTERFACING

The inputs are taken from the sensors

array and LM324 comparator and the

datafromthesedeviceswillbeinbinary

form. That data is feed in FPGA board

which controls the motor driver IC

L293D.Thesemotorsaredrivenbytwo

wheels. The whole hardware system

willdetectobstaclesandcontrolsspeed.

It also drives, controls and gives the

directiontotherobot.

The interfacing of whole system is

designed as follows [8]:


ISSN: 2348-4098


Figure1:HardwareinterfacingofrequiredcomponentswithFPGABoard

6. ADVANTAGESAND

APPLICATIONS

Thetheoryofpaper iscurrentlyamong

the most intensively studied and

promisingareasinVLSIfieldwhichwill

certainly play a primary role in future.

An FPGA hardware implementation is

proposed,theimplementationshoweda

significant speedup and reduction in

power consumption compared to the

traditional PC based software

implementation.Theprojectislowcost,

safe and convenient and also easy to

manipulate due to fast transmission

speed like characteristic of XC3S400‐

series FPGA kit. A hardware

implementation of navigation approach

of an underwater autonomous mobile

robot presents a wide application in

edge detection, 3D reconstruction and

localizationandobjectrecognition.

7. RESULT

The FPGA designs are implemented in

Xilinx ISE software, which provides a

variety of performance analyses,

including resource utilization, speed,

and power consumption. The Xilinx

performance report is based on

simulationsofthehardwaredesign.

The given figures shows Prototype

board built with FPGA XC3S400 with

simulation result of seven segment

display, stepper motor control and IR

sensorrespectively:


ISSN: 2348-4098


Figure2:Simulationresultofseven‐segmentdisplayVHDLcode

Figure3:SimulationresultofsteppermotorVHDLcode

Figure4:SimulationresultofIRSensorVHDLcode


ISSN: 2348-4098


RTLSchematicofMobileRobothasbeengivenbelow:

Figure5:RTLSchematicofUnderwaterRobot

Figure6:SimulationresultofUnderwaterRobotVHDLcode

8. CONCLUSIONANDFUTURE

SCOPE

In this paper, FPGA based real time

control system is developed for

wirelesslycontrolanytypeofmotorsin

robotic applications. The use of the

underwater mobile robots for the

purpose like object finding, capturing

special or required moments. It allows

the integration of several important

areas of knowledge and a low cost

solution. The main objective of this

workwastoproposeagenericplatform

for a robotic mobile system inside a

water body. Another objective was to

present practical solutions for


ISSN: 2348-4098


terrestrial problems, such as

maintenance,supervisionandtransport

of materials. Autonomous underwater

robots can be used to deliver parts

underwater, being complementary

platformsinasecuritysystemandthey

also can be used in hazardous areas

where humans cannot stay for a long

periodoftime.

The proposed framework remains

simpleanduser friendly;additionally it

provides enough flexibility for the

specific application. The approach can

be extended to more demanding

applications by adding more modules,

or other peripheral interfaces. This

workwastheprototypeofabigsystem.

Byappliedlogic,total16motorscanbe

controlled but by taking advantage of

FPGA, more number of logics can be

added into the developed control

system.Thefutureresearchareacanbe

integrating more numbers of different

modules by finding different aspects of

underwatervehicles.

ACKNOWLEDGEMENT

We would like to thanks to Dr. S. Lal,

Dean, Faculty of Engineering and

TechnologyandalltheFacultymembers

of ECE department, Jayoti Vidyapeeth

Women’s University, Jaipur for their

kindsupportandencouragement.

REFERENCES

[1] Recognising and locating objects

with local sensors ,Jan De Geeterl, H.

VanBrusse1,J.DeSchutter,M.DecrCton

[2] Prabhas Chongstitvatana. “A FPGA‐

based Behavioral Control System for a

Mobile Robot”. IEEE Asia‐Pacific

Conference on Circuits and Systems,

Chiangmai,Thailand,1998.

[3] S. Commuri, V. Tadigotla, L. Sliger ”

EfficientControllerimplementationsfor

Robot Control ” Circuits, Systems,

Electronics,Control&SignalProcessing,

Dallas,USA,November1‐3,2006

[4] Application Note : Motor Control

usingFPSLIC™/FPGA

[5] Samir Palnitkar, "Verilog HDL, A

guide to digital design and synthesis",

SunSoftpress1996

[6] Xilinx, "Spartan‐3 FPGA Starter Kit

Board User Guide", UG230 (v1.2)

January20,2011

[7]www.datasheetcatalog.com

[8] Volnie A.Pedroni. “Circuit Design

with VHDL” MIT Press, Cambridge,

Massachusetts,London,England.


ISSN: 2348-4098


EFFECTOFCOMPETINGCATIONS(Cu,Zn,Mn,Pb)

ADSORBEDBYNATURALZEOLITE

1AFRODITAZENDELSKA,2MIRJANAGOLOMEOVA

1,2FacultyofNaturalandTechnicalSciences,GoceDelcevUniversity,Stip,Macedonia


ABSTRACT

The aim of this work was to investigate the influence of the presence of competing

cations on the individual adsorption of Cu2+, Pb2+, Zn2+ and Mn2+ from a solution

containingamixtureof all thesemetal ions,bynatural zeolite. In thiswork is shown

compares the adsorption of each heavy metal ion from both single‐ and multi‐

component solutions. The amount adsorbed from multi‐component solutions was

affected significantly, except for Pb2+where the difference between single andmulti‐

component solution is minimal, almost insignificant. It was also determine the

selectivityofnaturalzeolite, fortherespectiveheavymetal ions.Theselectivityseries

obtained for singlecomponent solutionwas:Pb2+>Cu2+>Zn2+>Mn2+, and formulti‐

componentsolutionwasPb2+>Cu2+>Mn2+>Zn2+.

INDEX TERMS: copper, zinc, manganese, lead, zeolite, competing cation, selectivity

series.

1. INTRODUCTION

Zeolite is a natural porous mineral in

whichthepartialsubstitutionofSi4+by

Al3+ results in an excess of negative

charge. This is compensated by alkali

andalkalineearthcations(Na+,K+,Ca2+

or Mg2+). Zeolites have been used as

adsorbents, molecular sieves,

membranes, ion‐exchangers and

catalysts, mainly because zeolite

exchangeable ions are relatively

innocuous. Thus, zeolites are

particularly suitable for removing

undesirableheavymetal ions (e.g. lead,

nickel, zinc, manganese, cadmium,

copper, chromium and/or cobalt),

radionuclides as well as ammoniacal

nitrogen (ammonia and ammonium)

from municipal wastewater, industrial

wastewater,acidminedrainage,mining

operations, fertilizers, battery


ISSN: 2348-4098


manufacture, dyestuff, chemical

pharmaceutical, electronic device

manufacturesandmanyothers[1].

Most of heavy metals are highly toxic

and are non‐biodegradable; therefore

theymustberemovedfromthepolluted

streams in order to meet increasingly

stringent environmental quality

standards.

Industrial wastewater and acid mine

drainage typically contain many

differentmetalionsasamixture.These

ions have the potential to affect the

effectivenessofanadsorbentintreating

the wastewater and that is based on

their competition for exchange sites on

and in the adsorbent. Therefore, it is

important to investigate the impact of

competing cations on the removal of

eachpollutantfromsolution.

Theaimofthisworkwastoinvestigate

the influence of the presence of

competing cations on the individual

adsorptionof Cu2+, Pb2+, Zn2+ andMn2+

fromasolutioncontainingamixtureof

allfourmetalions,bynaturalzeolite.In

this work is shown compares the

adsorptionofeachheavymetalionfrom

both single‐ and multi‐component

solutions. Also, according to the

maximum adsorption capacity (qe)was

determine the selectivity of natural

zeolite, for the respective heavy metal

ions. There are a large number of

selectivity series assigned to zeolites

thatcontainclinoptilolite(Table1).

Table‐1:Examplesofexperimentallyderived

selectivityseriesofnaturalzeolitefordifferent

heavymetalsfromliterature

Blanchard et

al.,1984[3]

Pb2+ > NH4+ > Ba2+ > Cu2+≈

Zn2+>Cd2+≈Sr2+>Co2+

Zamzowetal.,

1990[4]

Pb2+ > Cd2+ > Cs2+ > Cu2+ >

Co2+ > Cr3+ > Zn2+ > Ni2+ >

Hg2+

Moreno et al.,

2001[5]

Fe3+≈Al3+>Cu2+>Pb2+>Cd2+

>Zn2+ >Mn2+ >Ca2+≈Sr2+

>Mg2+

Inglezakis et

al.,2002[6]

Pb2+>Cr3+>Fe3+>Cu2+

Alvarez‐Ayuso

etal.,2003[7]

Cu2+ > Cr3+ > Zn2+ > Cd2+ >

Ni2+

Erdem et al.,

2004[8]

Co2+>Cu2+>Zn2+>Mn2+

B.Calvoetal.,

2009[9]

Pb2+ >Cu2+>Zn2+

Sprynskyy,

2009[10]

Cd2+ > Pb2+ > Cr3+ > Cu2+ >

Ni2+

Motsi, 2010

[11]

Fe3+>Zn2+>Cu2+>Mn2+

SabryM. S. et

al.,2012[12]

Pb2+ >Cu2+>Zn2+>Cd2+ >

Ni2+


ISSN: 2348-4098


The selectivity of zeolite to adsorb

various cations is the result of the

complex combined effect of follow

parameters: 1.Parameters related to

work conditions: the static or dynamic

nature of the regime of adsorption,

solid:liquid ratio,working temperature,

initial concentration and pH of contact

solutions, stirring intensity of the

heterogenous system as well as the

nature of the cation and accompanying

anion; 2.Parameters related to the

characteristics of zeolite: the average

diameterofparticles,mineralogicaland

chemical composition, initial activation,

internal structure of macropores and

microporesand3.Parametersrelatedto

the characteristics of adsorbed ions:

hydrated radiusof the ion, tendency to

form hydrocomplexes in solutions,

hydration energy and ionicmobility, as

wellasotherfactors[2].

2. MATERIALSANDMETHODS

2.1ADSORBENT

The natural zeolite‐ clinoptilolite was

usedintherecentstudyasanadsorbent

for adsorptionofheavymetals, suchas

Cu, Zn, Mn and Pb. The particle size

rangeof thenaturalzeoliteusedinthis

studywas0.8to2.5mm.

The chemical compositions of natural

zeolitearepresentedinTable2.

Table2:Chemicalcompositionofzeolite

samples

Typicalchemicalcompositionin%wt

SiO2 69.68 CaO 2.01

Al2O3 11.40 Na2O 0.62

TiO2 0.15 K2O 2.90

Fe2O3 0.93 H2O 13.24

MgO 0.87 P2O5 0.02

MnO 0.08 ratioSi/Al 4.0‐5.2

Cation exchange

percation

K+41meq/100g

Na+16.10meq/100g

Ca2+67.14meq/100g

Mg2+3.88meq/100g

Total cation

exchangecapacity 1.8‐2.2meq/g

X‐Ray Difractometer 6100 from

Snimadzu was used to investigate the

mineralogical structure of natural

zeolitesamples.Thistechniqueisbased

onobserving the scattering intensityof

an X – Ray beam hitting a sample as a

functionofincidentandscatteredangle,

polarization,andwavelengthorenergy.

The diffraction data obtained are

compared to the database maintained


ISSN: 2348-4098


by the International Centre for

DiffractionData,inordertoidentifythe

material in the solid samples. The

results of XRD (Fig. 1) shown that the

natural zeolite contained clinoptilolite

inthemajority.

Figure1:X–Raydiffractionofnaturalzeolite

The surface morphology of natural

zeolite was studied using a scanning

electronmicroscope, VEGA3 LMU. This

particularmicroscopeisalsofittedwith

anInca250EDSsystem.EDS,standsfor

EnergyDispersiveSpectroscopy,itisan

analytical technique used for the

elementalanalysisofasamplebasedon

the emission of characteristic X – Rays

bythesamplewhensubjectedtoahigh

energy beam of charged particles such

aselectronsorprotons.Micrographsof

natural zeolite samples obtained from

SEM analysis are given in Fig. 2. The

micrographs clearly show a number of

macro‐pores in the zeolite structure.

Themicrographsalsoshowwelldefined

crystalsofclinoptilolite.

Figure2:Micrographsofnaturalzeolite

samplesobtainedfromSEManalysis

An electron beam was directed onto

differentpartsofthesamplesinorderto

getamoreaccurateanalysis(Fig.3)and

the elemental composition of natural

zeolite (clinoptilolite) are presented in

Table3.

Figure3:EDSanalysisshowingthescanning

methodfornaturalzeolite


ISSN: 2348-4098


Table‐3:EDSanalysisshowingtheelementalcompositionfornaturalzeolite

Element Spect1 Spect2 Spect3 Average Standarddeviation

O 58.46 55.4 58.83 57.56 1.882

Na 0.27 0.15 0.3 0.24 0.079

Mg 0.72 0.66 0.77 0.72 0.055

Al 5.28 5.52 5.03 5.28 0.245

Si 29.55 31.36 29.47 30.13 1.068

K 2.73 2.96 2.44 2.71 0.26

Ca 1.9 2.42 1.66 1.99 0.388

Fe 1.1 1.53 1.5 1.38 0.24

Total 100 100 100 100

ResultsofEDSanalysisshowedthatthe

predominant exchangeable cations in

natural zeolite (clinoptilolite) structure

wereK+andCa2+.

2.2ADSORBATE

The heavy metals, Cu, Zn, Mn and Pb

were used as adsorbate in the recent

investigations. Synthetic single and

multi‐component solutions of these

metals were prepared by dissolving a

weighed mass of the analytical grade

salt CuSO4.5H2O, ZnSO4.7H2O,

MnSO4.H2O and Pb(NO3)2,

appropriately,in1000mldistilledwater.

2.3EXPERIMENTALPROCEDURE

Adsorption of heavy metals ions on

zeolite was performed with synthetic

single and multi‐component ion

solutions of Cu2+, Zn2+ Mn2+ and Pb2+

ions with initial concentration of 25

mg/l. Initial pH value 3.5 of prepared

solutions was adjusted by adding 2%

sulfuric acid and controlled by 210

Microprocessor pH Meter. The

experimentswereperformedinabatch

mode in a series of beakers equipped

with magnetic stirrers by contacting a

mass of zeolite (5g) with a volume of

solution, 400ml. Zeolite sample and

aqueous phase were suspended by

magnetic stirrer at 400 rpm. The

agitation time was varied up to 360


ISSN: 2348-4098


minutes. At the end of the

predetermined time, the suspension

was filtered and the filtrate was

analyzed. The final pH value was also

measured. All experiments were

performed at room temperature on

20±1oC. The initial and remaining

concentrations of metal ions were

determined by Liberty 110, ICP

Emission Spectrometer, Varian.

Inductively coupled plasma atomic

emission spectroscopy(ICP‐AES) is an

analytical technique used for the

detection of trace metals. It is a type

ofemission spectroscopy that uses

theinductively coupled plasmato

produce excited atoms and ions that

emitelectromagnetic radiationat

wavelengths characteristic of a

particularelement.The intensity of this

emission is indicative of the

concentrationoftheelementwithinthe

sample.

The adsorption capacitywas calculated

byusingthefollowingexpression:

,(mg/g) (1)

where: isthemassofadsorbedmetal

ionsperunitmassofadsorbent(mg/g),

and are the initial and finalmetal

ion concentrations (mg/l), respectively,

Visthevolumeoftheaqueousphase(l)

andmisthemassofadsorbentused(g).

Degree of adsorption, in percentage, is

calculatedas:

(2)

3. RESULTSANDDISCUSSION

Experiments were carried out to

investigatetheinfluenceofthepresence

of competing cations on the individual

adsorption of Cu2+, Zn2+,Mn2+ and Pb2+

fromasolutioncontainingamixtureof

all4metalions,bynaturalzeolite.

On Chart 1 is made comparison of the



solutions.


ISSN: 2348-4098


Chart‐1:Comparisonoftheadsorptioncapacity

ofnaturalzeoliteforCu,ZnMnandPbfrom

singleandmulti–componentsolutions

The amount adsorbed from multi‐

component solutions was affected

significantly, except for Pb2+where the

difference between single and multi‐

component solution is minimal, almost

insignificant. The results show that

amount adsorbed Cu2+ from multi‐

component solution was decreased

approximately 10%, and 25‐50% for

Zn2+andMn2+compared to their single

componentsolutions.

Moreover, the total amount of heavy

metal ions adsorbed (all four cations)

per unit mass of natural zeolite

increasedofmulti‐componentsolutions

compared to the amount of solute

adsorbed from single component

solutions.

According to the obtained results was

determinetheselectivityofusedzeolite.

This was done by comparing the

maximum adsorption capacity (qe) of

natural zeolite for the respectiveheavy

metal ion. The selectivity series

obtained in single component solution

was: Pb2+ > Cu2+ > Zn2+ > Mn2+, but in

multi‐component solution was Pb2+ >

Cu2+>Mn2+>Zn2+.

Thedifferenceinadsorptioncapacityof

the natural zeolite for the heavymetal

ionsmaybeduetoanumberof factors

whichincludehydrationradii,hydration

enthalpies and solubility of the cations.

The hydration radii of the cations are:

rHZn2+=4.30Å,rHCu2+=4.19Å,rHPb2+=

4.01ÅandrHMn2+=4.38Å[13][14].The

smallest cations should ideally be

adsorbedfasterandinlargerquantities

comparedtothelargercations,sincethe

smaller cations can pass through the

micropores and channels of the zeolite

structure with ease [8]. Furthermore,

adsorption should be described using

hydrationenthalpy,whichistheenergy

that permits the detachment of water

moleculesfromcationsandthusreflects

theeasewithwhichthecationinteracts


ISSN: 2348-4098


withtheadsorbent.Therefore,themore

a cation is hydrated the stronger its

hydration enthalpy and the less it can

interact with the adsorbent [11].

Because of its high Si:Al ratio,

clinoptilolitehasalowstructuralcharge

density.Therefore,divalentcationswith

low hydration energies are sorbed

preferably compared to cations with

high hydration energies [15]. The

hydration energies of the cations are: ‐

2010, ‐1955, ‐1760 and ‐1481 kJmol‐1

for Cu2+, Zn2+, Mn2+ and Pb2+

respectively [13] [14].According to the

hydration radii the order of adsorption

shouldbePb2+>Cu2+>Zn2+>Mn2+,and

according to the hydration enthalpies

the order should be

Pb2+>Mn2+>Zn2+>Cu2+.

Accordingtothehydrationenergiesand

hydration radii, the zeolite will prefer

Pb over Cu, Mn and Zn in multi‐

component solutions.Therefore, it is to

beexpectedthathighPbconcentrations

willlimittheuptakeofCu,MnandZn.

The above series according to the

hydration radii is same with the

experimentally obtained series for

single component solution, which is

Pb2+ > Cu2+ > Zn2+ > Mn2+. But the

experimentally obtained series for

multi‐component solution is different

from above series according to the

hydrationradiiandenthalpies.

4. CONCLUSIONS

The investigation for influence of the

presence of competing cations on the

individualadsorptionofCu2+,Zn2+,Mn2+

and Pb2+ from a solution containing a

mixtureofallthismetalions,bynatural

zeolite is done by comparing the



solutions. From this is concluded that

the amount adsorbed from multi‐

component solutions was affected

significantly, except for Pb2+where the

difference between single and multi‐

component solution is minimal, almost

insignificant.TheamountadsorbedCu2+

from multi‐component solution was

decreased approximately 10%, and 25‐

50% for Zn2+ and Mn2+ compared to

theirsinglecomponentsolutions.

The own unique selectivity series on

investigatedzeoliteinsinglecomponent

solutionwas:Pb2+>Cu2+>Zn2+>Mn2+,

but in multi‐component solution was

Pb2+ >Cu2+ >Mn2+ > Zn2+.According to

the hydration energies and hydration

radii, thezeolitewillpreferPboverCu,

Mn and Zn in multi‐component

solutions.Therefore,itistobeexpected


ISSN: 2348-4098


that high Pb concentrations will limit

theuptakeofCu,MnandZn.

REFERENCES

[1] J.R.SilvioRobertoTaffarel,“Onthe

removal of Mn2+ ions by

adsorption onto natural and

activated Chilean zeolites,”

Minerals Engineering 22 , p. 336–

343,2009.

[2] L. Mihaly‐Cozmuta, A. Mihaly‐

Cozmuta, A. Peter, C. Nicula, H.

Tutu, Dan Silipas, Emil Indrea,

“Adsorptionofheavymetalcations

by Na‐clinoptilolite: Equilibrium

and selectivity studies,” Journal of

Environmental Management 137,

pp.69‐80,2014.

[3] G. Blanchard, M. Maunaye, G.

Martin, “Removal of heavy metals

from waters by means of natural

zeolite,” Water Res. 18, pp. 1501‐

1507,1984.

[4] M.J. Zamzow, B.R. Eichbaum, K.R.

Sandgren,D.E.Shanks,“Removalof

heavymetalandothercationsfrom

wastewaterusingzeolites,”Sep.Sci.

Technol.25,pp.1555‐1569,1990.

[5] Moreno, N., Querol, X., Ayora, C.,

“Utilization of zeolites synthesised

from coal fly ash for the

purification of acid mine waters,”

Environmental Science and

Technology, 35, pp. 3526‐3534,

2001.

[6] Inglezakis, V.J., Loizidou, M.D.,

Grigoropoulou, H.P., “Equilibrium

andkineticionexchangestudiesof

Pb2+, Cr3+, Fe3+ and Cu2+ on

natural clinoptilolite,” Water

Research,36,pp.2784‐2792,2002.

[7] Alvarez‐Ayuso, E., Garcia‐Sanchez,

A.,Querol,X.,“Purificationofmetal

electroplating waste waters using

zeolites,” Water Research, 37, pp.

4855‐4862,2003.

[8] E. Erdem, N. Karapinar, R. Donat,

“The removal of heavy metal

cationsbynaturalzeolites,”Journal

of Colloid and Interface Science,

Volume 280, Issue 2, p. 309–314,

2004.

[9] B.Calvo,L.Canoira,F.Morante,J.M.

Martínez‐Bedia, C. Vinagre, J.E.

García‐González, J. Elsen, R.

Alcantara, “Continuous elimination

ofPb2+,Cu2+,Zn2+,H+andNH4+

from acidic waters by ionic

exchange on natural zeolites,”

Journal of Hazardous Materials,


ISSN: 2348-4098


Volume 166, Issues 2–3, p. 619–

627,2009.

[10]M. Sprynskyy, “Solid–liquid–solid

extraction of heavymetals (Cr, Cu,

Cd,NiandPb) inaqueoussystems

of zeolite–sewage sludge,” Journal

of Hazardous Materials, Volume

161, Issues 2–3, p. 1377–1383,

2009.

[11]T.Motsi,Remediationofacidmine

drainage using natural zeolite,

Doctotal thesis, United Kingdom:

School of Chemical Engineering,

The University of Birmingham,

2010.

[12]SabryM. Shaheen,Aly S.Derbalah,

Farahat S. Moghanm, “Removal of

Heavy Metals from Aqueous

Solution by Zeolite in Competitive

Sorption System,” International

Journal of Environmental Science

andDevelopment,Vol.3,No.4,pp.

362‐367,2012.

[13]E. R. Nightingale,

“Phenomenological theory of ion

solvation. Effective radii of

hydrated ions.,” J. Phys. Chem., 63

(9),p.1381–1387,1959.

[14]I. Mobasherpour, E. Salahi, M.

Pazouki, “Comparative of the

removalofPb2+,Cd2+andNi2+by

nano crystallite hydroxyapatite

from aqueous solutions:

Adsorption isotherm study,”

ArabianJournalofChemistry ,том

5,бр.4,p.439–446,2012.

[15]C. Colella, “Ion exchange equilibria

in zeolite minerals,” Mineralium

Deposita,31,pp.554‐562,1991.


ISSN: 2348-4098
