mf.unze.bamf.unze.ba/wp-content/uploads/2016/04/Masinstvo-num4-1998.pdf · grisani uticaj teorija, strategija, i alata modeliranja na ... igrati u procesu konstruisanja, mora se prethodno

Ma{instvo 4(2), 169 - 182, (1998) N.Rep~i}: JEDAN PRIKAZ OKRU@ENJA...

- 116699 -

JJEEDDAANN PPRRIIKKAAZZ OOKKRRUU@@EENNJJAA PPRRII PPRROOCCEESSUU KKOONNSSTTRRUUIISSAANNJJAA

Prof. dr. Nedàd Rep~i}, Ma{inski fakultet u Sarajevu, Univerzitet u Sarajevu,71000 Sarajevo, Vilsonovo {etali{te 9, Bosna i Hercegovina,

REZIME

Ovaj rad je jedan prikaz dizajn okruènja, koji poku{ava adresirati problem integrativnog alata u procesukonstruisanja. Porast kompleksnosti proizvodnog procesa i potrebe za vi{om efikasnosti, ve}om fleksibil-nosti, boljim proizvodnim kvalitetom, i niòm cijenom promjenili su prirodu industrijske prakse. Me|utim,primjena ra~unara je dozvolila provo|enje naprednijih tehnika modeliranja. Ra~unari su {iroko kori{teni uinènjerskim aplikacijama, ali njihova primjena je ipak ograni~ena na jednostavnije algoritamske prora~une.U stvari, mnogi inènjerski problemi nisu prikladni da se jednostavno algoritamski rje{avaju. Jedna klju~nakonkurentska prednost za moderna proizvodna preduze}a je fleksibilnost u brzom reagovanju na trì{nepritiske i potrebe kupca. Primarno ograni~enje nije trud ili kapital, nego informacija. Tradicionalna pri-mjena ra~unara u proizvodnim rje{enjima je da pomaù upravljati informacijama pridruènih pojedina~nimzadacima preko CAD-a, toka procesa, procesa kontrole i dijagnoze. Na nesre}u takvi sistemi nisu dobrointegrisani i ~ine malo olak{anje pri podjeli informacija i kordinaciju pri odlu~ivanju. Kao odgovor istraìva~ipoku{avaju razviti integrativne radne okvire koji u~inkovito povezuju zajedno konstruktore, softverske si-steme, ma{ine i aktivnosti preko jednog velikog, geografski raspr{ene proizvodne kompanije. Do sada,istraìvanje je bilo fokusirano na softverske alate (softverski programi raspore|eni preko cjelokupne mreèra~unara), me|utim organizaciona struktura alata ~ini se da je zna~ajnija od toga.

Klju~ne rije~i: Proces konstruisanja, CAD, PDM

AA DDEESSCCRRIIPPTTIIOONN OOFF AA DDEESSIIGGNN EENNVVIIRROONNMMEENNTT

Nedàd Rep~i}, Ph.D., BSc. Mech.Eng., Faculty of Mechanical Engineering, Univer-sity of Sarajevo, 71000 Sarajevo, Vilsonovo {etali{te 9,

SUMMARY

This paper is a description of a design environment, which attempts to address the problem of toolintegration in the design process. The growing complexity of manufacturing processes and the needfor higher efficiency, greater flexibility, better product quality, and lower cost have changed the natureof industrial practice. Meanwhile, the application of computers has allowed the implementation of moreadvanced modeling techniques. Computers have been widely used in engineering applications, buttheir use has been limited almost exclusively to purely algorithmic solutions. In fact, many engi-neering problems are not amenable to purely algorithmic computation. A key competitive advantagefor modern manufacturing enterprises is the flexibility to react swiftly to market pressures and cus-tomer needs. The primary limitation is not labor or capital, but information. Traditional applicationsof computers in manufacturing provide point solutions that help manage information associated withindividual tasks such as CAD, scheduling, process control and diagnosis.Unfortunately, such systemsare not well integrated and do little to facilitate the sharing of information and coordination of deci-sions. In response, researchers are attempting to develop integration frameworks that seamlessly tietogether people, software systems, machines, and activities across a large, geographically dispersed,manufacturing enterprise.Until now, research in the field has been focused on the software tools( soft-ware programs distributed over the whole network of computers); however, the organizational struc-tures of the tools seem to be significant as well.

Key words: Design process, CAD, PDM

PREGLEDNI RAD

SUBJECT REVIEW

1. UVOD

Postoji danas jedan ponovljeni interes u konstruisa-nju kao jedan dio inènjerske edukacije i kao jedanaspekt inènjerstva. Ovo je registrovano u dva prav-ca, sa poku{ajima pove}ati sadràj inènjerskogmodeliranja inènjerskih tokova i povezuju}i inènjer-stvo sa industrijskim dizajnom. Sve vi{e prakti~nadizajn aktivnost je komplementirana pomo}u lekcijakoje obuhvataju tehnike modernih dizajn analiza. Oneuklju~uju konstruiranje obzirom na zamor, FE analize,modeliranje obzirom na pouzdanost, dizajn premakvaliteti, optimizacija modeliranja i modeliranje zamontaù i proizvodnju. Mada je glavna namjera ovograda izvje{taj o pristupu preuzimanja inènjerskogmodeliranja, ipak je prvo potrebno dati pregledglavnih elemenata CIM-a. Postoji jedan pristup CIM-u koji se ~ini konzistentnijim tj. umjesto jednostavnogzamjenjivanja sa akronimom CAD/CAM ( Dizajniranjepodràno ra~unarima / Proizvodnja podràna ra~unari-ma ) sa jednim modernijim CAD/CAM-om CIM-a, pri~emu se dodaju pojmovi koji uve}avaju integrativneaplikacije. êsto, ova perspektiva je pra}ena pret-postavkom da automatizacija kreiranja i rasporednumeri~ke kontrole podataka postavljenih za obraduna alatnim ma{inama je osnovni cilj CIM-a. To nijeza sugerirati da efektivne CAD/CAM aplikacije usmje-rene prema NC generisanju nisu nu`ne. Umjesto ovihpokazatelja potreba za jednim mnogo {irim pogledomprema CIM-u, koji moè uklju~iti ove aplikacije, alikoji obuhvata i druge funkcije proizvodnih preduze}asu dobro do{le.. Drugi pogled CIM-a prepoznaje daizdanci automatizacije su kreirana u mnogim pre-duze}ima i da oni mogu biti kategorizirani u ~etirigrupe. Ove grupe su izlistane ispod sa tipi~nimizdancima automatizacije na|enih u ovim grupama.

CAD/CAMGrupna tehnologija, Inènjering podràn ra~unarima (CAE),Konstruisanje podràno ra~unarima ( CAD ),Proizvodnja podràna ra~unarima ( CAM ).Planiranje proizvodnje i kontrolaKontrola zaliha, Optere~enost prodajnih objekata,Planiranje kapaciteta, Prvenstvo u redoslijedu,Prodaja.Automatizacija preduze}aProizvodnja podràna ra~unarima, Robotika,NC/DNC/CNC, Fleksibilni proizvodni sistemi,Sistemi automatskog rukovanja materijalima,Automatizirana test oprema, Kontroleri procesa.Op{te poslovno upravljanjeOp}enito i formiranje cijene, Marketing, Ulaz poru|be,Pomo} pri odluci, Radni pregled, Platne liste.


- 117700 -

1. INTRODUCTION

There is a renewed interest in design as a part ofengineering education and as an aspect of engi-neering. This is noticeable in two ways, with attemptsto increase the engineering design content of engi-neering courses and by combining engineering withindustrial design. All the more practical design workis complemented by lectures covering moderndesign analysis techniques. These include designagainst fatigue, FE analysis, design for reliability,design for quality, design optimisation and design forassembly and manufacture. Although the main pur-pose of this paper is to report the approach takento engineering design, it is first necessary to givean overview of the main elements of the CIM.There is a view of CIM which seems to consist ofsimply replacing the acronym CAD/CAM (ComputerAided Design/Computer Aided Manufacturing) withmore fashionable one of CIM and adding the notionof increased application integration. Often, this per-spective is accompanied by the assumption thatautomating the creation and distribution of numeri-cal control data sets for operating machine tools isthe fundamental goal of CIM. This is not to suggestthat effective CAD/CAM applications directed towardNC generation are not necessary. Instead this indi-cates the need for a much broader view of CIM,that may include these applications, but whichembraces the other functions within the manufactur-ing enterprise as well. Another view of CIM recognizes that islands ofautomation have been created in many enterprisesand that these can be categorized into four groups.These groups are listed below along with typicalislands of automation found in these groups.

CAD/CAMGroup technology ,Computer-aided engineering (CAE),Computer-aided design ( CAD ), Computer-aidedmanufacturing ( CAM ).Manufacturing planning and control Inventory control ,Shop loading, Capacity planning,Master scheduling,Purchasing. Factory automation Computer-aided manufacturing,Robotics,NC/DNC/CNC, Flexible manufacturing systems, Automated materials handling systems, Automat-

edtest equipment, Process controllers. General business managementGeneral and cost accounting, Marketing, Order entry, Decision support, Labor collection, Payroll.

2. ENGINEERING DESIGN

Engineering Design methodology explains the integrat-ed influence of theories, strategies, and design toolson the final results quality. Modern technology ischanging faster then education and the marketdemands for engineering designers are changing asfast as technology changes. Historically, the engi-neering process developed through trial and errorapprenticeship, followed by a shift to scientific orient-ed engineering education. This paper deals with theproblem of applying knowledge engineering techniquesin the process of designing mechanisms. I want tostart with design, because it is a highly-creative area,and as such, one in which computers, because theyare capable only of routine tasks, would seem to beable to play no part. In fact, computers are used wide-ly in design work, which leads people to think thatthey are a lot brighter than they really are. To seewhat role computers can play in design work, we firsthave to examine the nature of the design processitself. Design is the application of technology and aes-thetics to the solution of a specific practical problem.In some areas, the technology dominates, while in oth-ers the aesthetic aspect is seen by many people asprimary, although the technological underpinning mostdefinitely remains. There are three distinct types ofoperation in the design process, and they are:

The formulation of ideas for possible solutions to the problem,The analysis of these ideas to see if theywill work,The selection of one solution as the "best",on whatever criteria one wants to use.

The first step is the creative one: how creativedepends on the novelty and the difficulty of the prob-lem. It involves formulating some possible solutions tothe problem. There is no immediate requirement thatall these solutions be workable, or even sensible.Theessential element in this process is that of synthesis.What the designer is doing is bringing together:

Technical knowledge, in the form of basic principles, often from a wide range of fields,Past experience, with similar projects and aesthetic sense.

They combine these into a mix which fits this par-ticular situation, and so come up with one or moresolutions that might work. The strategies used to dothis are many and varied, and are learned by exam-ple from other professionals, and by experience. They involve things like recasting a problem into dif-

2. IN@ENJERSKI DIZAJN

Metodologija inènjerskog dizajna obja{njava inte-grisani uticaj teorija, strategija, i alata modeliranja nakona~ne rezultate kvaliteta. Moderne tehnologijemjenjaju se brè od edukacije i trì{nih zahtjeva, akostruktori-projektanti mjenjaju se isto tako brzo kaoi tehnolo{ke promjene. Istorijski, inènjerski procesrazvijen je kroz poku{aj i gre{ke zanatstva, slijede}ipomake ka nau~no orjentisanoj inènjerskoj edukaci-ji. Ovaj rad bavi se sa problemom primjene tehnikainènjerskog znanja u procesu dizajniranja ma{inskihkonstrukcija. Ako se èli startati sa dizajnom,a budu}ije ono visoko kreativno podru~je i kao takvo u kojemra~unari su sposobni samo za rutinske zadatke, ~inise da bi trebalo biti sposoban uraditi to bez njihovogu~e{}a. U stvari, ra~unari su upotrebljeni {iroko u po-stupku modeliranja, koji vodi konstruktora ka misli dasu oni mnogo bistriji i efikasniji od onog {to sustvarno. Da bi stvarno vidjeli koju ulogu ra~unari moguigrati u procesu konstruisanja, mora se prethodno ispi-tati priroda samog procesa modeliranja. Modeliranje jeprimjena tehnologije i estetike ka rje{avanju specifi~nihprakti~nih problema. U nekim podru~jima, tehnologijadominira, dok u drugim estetski aspekt je vi|en odmnogih ljudi kao primaran, mada tehnolo{ka kompo-nenta je kona~no presudna. Postoje tri distinktivna tipaoperacija u procesu konstruisanja, i oni su:

Formulacija ideje za mogu}a rje{enja problema,Analize ovih ideja da se vidi kako }e onefunkcionisati,Izbor jednog rje{enja kao "najboljeg", ma koji kriterij se èli upotrebiti.

Prvi korak je kreativni korak: kako kreirati zavisnood novina i te{ko~a problema. On obuhvata for-mulisanje nekih mogu}ih rje{enja problema. Nesmije biti zahtjeva koja ova rje{enja ne bi moglauraditi ili ~ak osjetiti. Bitan elemenat u ovom pro-cesu je ova sinteza. Kakav konstruktor je zavisiod dva elementa:

Tehni~kog znanja, u formi osnovnih principa, ~esto iz jednog {ireg podru~ja nau~ne oblasti,Prethodno iskustvo, sa sli~nim projektima i estetskim osje}ajem.

Oni kombinuju ovo u jednu mje{avinu kojapode{ava ovu pojedina~nu situaciju, i tako se dolazido jednog ili vi{e rje{enja koja bi mogla raditi.Strategije upotrebljene da ovo u~ine su mnoge irazli~ite, a izu~avaju se pomo}u primjera od drugihprofesionalaca, i pomo}u iskustva. Oni obuhvatajustvari poput rikastinga jednog problema i razli~itimoblicima done{enih na osnovu sli~nosti sa pozna-


- 117711 -

tim problemima, razbijanjem problema u manje, pri-lagodljivije komade, dudling, i mnoge druge. Pos-toji jedan broj pojava o ovom {to ~ini da proceskonstruisanja je vrlo te{ko emulirati na ra~unaru:

înjenica je da nema dizajna koji slijedi ta~noneke procese, budu}i su razli~iti zahtjevi,Varijabilnost izvora koji bi mogli biti prisni zaop{te principe, i ~injenica da precizna primjena ovih moè da razlikuje i estetske zahtjeve.

Sve ovo ~ini to da je nemogu}e ispisati kruti setpravila zahtjevanih od ra~unara prije nego se to nepredstavi. Me|utim, ovo nije zaustavilo upotrebura~unara u asistiranju pri ovom radu. Posti}i ovaj ciljkoji je zahtjevan, mora se planirati i konstrukcijaproizvodnih sistema, kao i kvalifikovani dizajn zadatakcentralne va`nosti koji mora biti izvr{en u vezi samodeliranjem proizvoda koji }e biti proizvedeni. Ovajdizajn rad zahtjeva uvje`ban, dobro edukovan i isku-san proizvodni sistem dizajnera, opremljen sa efikas-nim alatkama. To je njihova odgovornost uvidjeti toda novi proizvodni sistemi ta~no traè funkcionalnezahtjeve koji su definisani proizvodima koji treba dabudu izra|eni. Nadalje, to je njihov zadatak dauklju~e ljudske faktore u modeliranje i time uzmuu~e{}e i psiholo{ke i fiziolo{ke potrebe radnika iproizvodnih planera. U dodavanju ovih elemenata zaovaj dizajn rad, mora sve ovo biti konkurentnokratko.

2.1 Primjena metode kona~nih elemenata

Jedna razlo`na primjena metode kona~nih elemena-ta zahtjeva pojednostavljenje procesa modeliranja,respektivno spojiti to sa CAD procesom, FEM pred-procesorima razvijenih za ovu namjenu. Rje{enjekoje je obi~no prezentovano sadrì dva koraka. Uprvom koraku jedan takozvani osnovni model seupotrebljava, koji zapravo predstavlja prethodni ste-pen modela kona~nog elementa. Osnovni modelmoè biti izveden prvenstveno iz oblika karakteristikemodela CAD sistema. On se transformi{e u stvarniFEM - model pomo}u jedne automatske generacijemreè. Jedan model kona~nog elementa moè bitizami{ljen kao jedan sistem krutih opruga.Kad se primjeni jedno optere}enje na strukturi, svi ele-menti deformi{u su dok se sile ne izbalansiraju. Zasvaki element u modelu, jedna~ine mogu biti pisaneu skladu pomjeranja i sila na ~vorovima. Element kojije uzet za primjer ovdje je jedan 2D kvadrat koji ima~etiri ~vora. Svaki ~vor ima dva stepena slobode kojisu pridruèni njima ( pomjeranje u X i Y pravcima),tako da element ima ukupno osam stepeni slobode.


- 117722 -

ferent forms to bring out similarities with known prob-lems, breaking the problem into smaller, more tractablepieces, doodling, and many others. There are a num-ber of features about this which makes the designprocess very difficult to emulate on the computer:

The fact that no design follows exactly thesame processes, because of the differing requirements,The variety of sources which can be tappedfor general principles, and the fact that the precise application of these may differ and theaesthetic requirement.

All of these make it impossible to write down the rigidset of rules required be a computer before it will per-form. However, this does not stop us using a com-puter to assist us with the hackwork.To achieve thisgoal it is required that we regard planning and con-struction of manufacturing production systems as aqualified design task of central importance which mustbe carried out in conjuction with the design of prod-ucts to be manufactured. This design work requiresskilled, well educated and experienced manufacturingsystem designers equipped with efficient tools. It istheir responsibility to see to it that the new manu-facturing production systems accurately meet the func-tional requirements as defined by the products to beproduced. Furthermore it is their task to includehuman factors in the design and thereby anticipatethe psychological and physiological needs of theworkers and production planners. In addition the leadtime for this design work must be competitively short.

2.1 Application of the Method of Finite Elements

A reasonable application of the Method of Finite Ele-ments demands to simplify the modelling process,respectively to connect it with the CAD process, FEMpreprocessors have been developed to this aim. Thesolution presented consists of two steps. In the firststep a so called basic model is used, which is aprevious stage of the finite-element-model. The basicmodel can be derived advantageously from the formfeature model of a CAD-system. It is transformed tothe real FEM-model by means of an automatic meshgenerator. A finite-element model can be thought ofas a system of solid springs. When a load is appliedto the structure, all elements deform until all forcesbalance. For each element in the model, equationscan be written relating displacements and forces atthe nodes. The element shown here, for example, isa 2D quadrilateral having four nodes. Each node hastwo degrees of freedom associated with it (displace-ments in X and Y directions), so that the element has

Tamo mora tako|e biti jedna ~vorna sila za svaki~vorni stepen, tako postoji tako|e osam ~vornih silaza elemenat. Ova pomjeranja i sile su identifikovanepomo}u jednog brojnog koordinatnog sistema za ulazkompjuterskog programa.Kad je jedna struktura modelirana, pojedina~ne garni-ture matri~nih jedna~ina su automatski generisane zasvaki elemenat. Elementi u modelu dijele zajedni~ke~vorove tako pojedina~nih garnitura matri~nih jedna~inai mogu biti kombinovane u jedan globalni set matri~nihjedna~ina. Ovaj globalni set je u relaciji svih ~vornihstepeni slobode ka ~vornim silama, i ~vorni stepenislobode rije{eni su simultano iz globalne matrice. Kadasu poznata pomjeranja za sve ~vorove, onda je tako|erpoznato stanje deformacije svakog elementa. A kadaje poznata deformacija svakog elementa onda su po-znati i naponi i pomjeranja svakog elementa. Za obi~nestati~ke analize metod kona~nog elementa je dvoste-peni proces. ^vorna pomjeranja su prvo istovremenoizra~unata iz elementa krutosti i ~vornih sila, iunutra{njih i vanjskih. Slijede}e, izra~unavaju se naponi,op}enito kod svakog element-centroida.Budu}i su pomjeranja izra~unata samo za kona~anbroj ta~aka u strukturi, metoda kona~nih elemenataje ipak jedna numeri~ka aproksimacija prili~no uda-ljena od ta~nog rje{enja. Parametarske karakteristikepostaju sve vi{e zajedni~ke u FEA paketima. Klju~napovoljnost parametarskih karakteristika je da onedopu{taju korisnicima da vide efekte dizajn promje-na veoma brzo. Sa adekvatnim planiranjem. Korisni-ci mogu definisati jedan FE model cjelokupno uizrazima varijabli ili parametara. âk mre`ne karak-teristike mogu biti definisane kao parametri.

2.2 Uop{teno-ssvrha kretanja-aanalizasoftvera pokrenutih na PC-uu

Prikaz je dat sa namjerom da ukaè da postoje i diza-jneri koji èle isprobati nekoliko ideja na brz na~in.Drugi paketi, modelirani za zahtjevnije zadatke, suraspoloìvi na UNIX radnim stanicama. One mogu nesamo analizirati optere}enje na zglobovima i pinovima,nego tako|er optimiziraju konstrukcije pomo}uodre|ivanja najbolje duìne za moment krakova, i takodalje. Paketi dizajnirani za upotrebu na PC-u tipi~nosu mehani~ki dinami~ki simulatori. Korisnik defini{eobjekte sa specificiranim masama, tada posmatra tokao program koji pokazuje {ta se doga|a pod datimsetom uslova. Geometrije mogu biti ili u samom pro-gramu ili importovane preko dxf fajlova. Modeli imajufii~ka svojstva koja uklju~uju masu, stati~ko i kineti~kotrenje, elasticitet,moment inercije, i elektrostati~ki naboj.Raspoloìva ograni~enja u PC baziranim programimauklju~uju zglobove, prigu{iva~e, uàd, neelasti}nepoluge, opruge, ispravlja~e, i motore spojenih masa.Korisnici ograni~avaju mase sa pinovima, prorezima,


- 117733 -

a total of eight degrees of freedom. There must alsobe a nodal force for each nodal degree, so there arealso eight nodal forces for the element. These dis-placements and forces are identified by a coordinatenumbering system for entry the computer program.When a structure is modeled, individual sets ofmatrix equations are automatically generated for eachelement. The elements in the model share commonnodes so individual sets of matrix equations can becombined into a global set of matrix equations. Thisglobal set relates all the nodal degrees of freedomto the nodal forces, and the nodal degrees of free-dom are solved simultaneously from the globalmatrix. When displacements for all nodes are known,the state of deformation of each element is known.And, when deformation of each element is know, thestress and strain within the element are also known. For simple static analysis, the finite-element methodis a two-step process. Nodal displacements arefirst simultaneously calculated from the elementstiffness and the nodal forces, both internal andexternal. Next, stresses are calculated, generally atthe each element's centroid. Because displace-ments are calculated for only a finite number ofpoints in the structure, the finite-element method isa numerical approximation rather than an exactsolution.Parametric features are becoming more common inFEA packages. The key benefit of parametric fea-tures is that they let users see the effects of designchanges quickly. With adequate planning, users candefine an FE model entirely in terms of variables orparameters. Even mesh characteristics can bedefined as parameters.

2.2 General-ppurpose motion-aanalysis packages run on PCs.

They are aimed at designers who want to try outa few ideas quickly. Other packages, designed formore demanding tasks, are available on Unix work-stations. These might not only analyze loads onpivots and pins, but also optimize designs bydetermining the best length for moment arms, andso forth. Packages designed for use on PCs typi-cally are mechanical dynamics simulators.The userdefines objects with specified masses, then watch-es as the program shows what happens undergiven sets of conditions. Geometries can either bedevised in the program itself or imported via .dxffiles. Models have physical properties that includemass, static and kinetic friction, elasticity, momentof inertia, and electrostatic charge. Available con-straints in PC-based programs include pulleys,joints, dampers, ropes,inflexible rods, springs, actu-

ators, and motors to join masses. Users constrainmasses with pin, slot, keyed slot, and rigid joints,and specify forces acting on the model. Motionsimulators calculate equations of motion throughnumerical integration. Users define the time stepsin the simulation to determine accuracy. Large timesteps are acceptable for slowly moving objects, butmay throw off accuracy. These simulators provideseveral different numerical methods that trade offaccuracy versus speed of calculation. PC-basedmotion simulators work strictly with 2D rigid bod-ies at this point. For situations where deflectionsplay a part in the simulation, and where mecha-nisms move in 3D, workstation simulators arerequired.SolidWorks is the leader of a new class of low-priced, feature-based solid modeling programs forWindows that is challenging the dominance ofPro/Engineer. Yet learning to use this new softwareisn't as easy as it looks. Designing Parts with Solid-Works helps new users of SolidWorks mastering fea-ture-based modeling using innovative approaches.Some of the unique benefits of this book include:

Maximum use of graphic examples withminimal reading. A project-oriented approach that has users creating part models while learning. The most efficient ways to designing parts ineach situation. Techniques for capturing design intent, enablingworkers to reuse models reliably. Tips and tricks designers need to know to get their work done with time to spare.

The essential guide for CAD managers on the wayup. It will help you get more out of your CAD sys-tems after you've bought them. You'll learn how toplan productivity increases, set goals and measureyour progress, and overcome fears of subordinates,superiors, and project managers who distrust CAD. Itshows you how to simplify administrative tasks,upgrade training programs,improve communicationswith other departments, and apply CAD technology totasks that yield the highest return. Searching for CADsoftware that fits your needs? It gives you featurelists, functional capabilities, and computing require-ments for CAD and CAM programs. Disciplinesinclude mechanical design and drafting, architecturaldesign, mapping, three-dimensional modeling, struc-tural analysis, NC tooling, and animation.

zatvorenim prorezima, i krutim zglobovima, i specifici-ranim silama koje djeluju na modelu. Kretanje simu-latora ra~una jedna~ine kretanja kroz numeri~ku inte-graciju. Korisnici defini{u vrijeme koraka u simulacijida odrede ta~nost. Velika vremena koraka su prih-vatljiva za sporija kretanja objekata, ali mogu znatnosmanjiti ta~nost. Ovi simulatori snabdjeveni su sanekoliko razli~itih numeri~kih metoda koji se dis-tribuiraju prema ta~nosti verzije brzine ra~unanja. PC-bazirani simulatori kretanja rade strogo sa 2D krutimtijelima kod neke ta~ke. Za situacije gdje ugibi igra-ju neku ulogu u simulaciji, i gdje mehanizmi se kre}uu 3D radnim stanicama simulatori su zahtjevniji.SolidWorks je predvodnik jedne nove klase relativnoniske cijene, feature-based solid programa za mode-liranje za Windows, koji je posljedica izazova domi-nacije Pro/Engineer-a. Tek izu~avanje upotrebe ovognovog softvera nije uop{te lagano kako to izgleda.Modeliranje dijelova sa SolidWorks-om pomaè novimkorisnicima SolidWorks-a mastering feature-basedmodeliranje koriste~i inovativne pristupe. Neke odunikatnih povoljnosti ovog softvera odnose se na:

Maksimalnu upotrebu grafi~kih primjera sa minimalnim ~itanjem,Jedan projektno-orjentisani pristup kojikorisnicima kreira dijelove modela dok ~itaju,Najefikasniji na~ini dizajniranja dijelova u svakojsituaciji,Tehnike za prihvatanje dizajn namjera, osposob-ljavaju}i radnike da pouzdano ponovo upotrebljavaju modele,Tips and tricks dizajneri treba da znaju da dobijaju i to, {to njihov rad ~ini efikasnijim.

Nuàn je vodi~ za CAD menadère na ovom putu.To }e pomo}i da se dobiju informacije za CAD sis-teme, nakon njihove instalacije i konfiguracije. Tu seizu~ava kako planirati uve~anje produktivnosti, posta-vljanje ciljeva, i mjerenje napretka, i prevazilaènjestrahova subordinatora, superiora, i projekt-menadèrakoji nevjeruju CAD-u. Ujedno ovo pokazuje kakopojednostaviti administrativne zadatke, prilagoditi pro-grame za obuku, pobolj{ati komunikaciju sa drugimodjeljenjima, i primjeniti CAD tehnologiju ka zadacimakoji su najkorisniji. Uvijek treba traìti softvere samoza pode{avanje vlastitih potreba. Ovo daje karakter-isti~nu listu funkcionalnih sposobnosti, kao i kom-pjuterskih zahtjeva za CAD i CAM programima. Dis-cipline uklju~uju ma{insko konstruiranje i crte`,arhitektonsko crtanje, mapiranje, trodimenzionalnomodeliranje, strukturalne analize, NC alate, i animaci-je.


- 117744 -

2.3 Korisnik i posluìlac priroda

Progresi u {tampa~ima, radnim stanicama i mediji-ma za arhiviranje i odlaganje podataka pomaùinènjerima da budu produktivniji. Hijerarhija upotre-bljena u CAD/CAM mreàma danas tipi~no je hijer-arhija jedne korisnik/posluìlac prirode. Svaka radnastanica, ima njenu vlastitu centralnu procesorjedinicu ( CPU ) i lokalni disk za odlaganje podata-ka. Mre`ni rad moè tako|e imati jedan fajl serverkoji sadràva CPU i tvrdi disk za odlaganje podata-ka, kao i magnetske trake za arhiviranje.Jedan klju~ni element client/server sistema je ovajfile-server softver snabdjeven fajlom rezimea kakorisniku. Korisnik radnih stanica ne mora znati gdjefajlovi su fizi~ki smje{teni. Ovo pona{anje jepovoljno i {to se zove jedan zami{ljeni fajl sistem,koji dopu{ta slanje fajl sistem zahtjeva ka jednojlokalnoj radnoj stanici fajl sistema ili ka udaljenomserveru putem mreè. Inènjerske radne staniceodnedavno imaju i multimedijalne sposobnosti. Sili-con Graphics Inc. je ve} duì period predvodnik ugrafi~kom procesiranju. SGI Indy serija radnih stani-ca, bazirana na jednom 64-bit Risc procesoru, imaunutra{nji audio capture i opcionalnu kameru zavideo telekonferenciju. Druge karakteristike uklju~ujudisplej rezoluciju ka 1.280x1.034, glavnu memorijukapaciteta ka 256MB, i povoljnost dodavanja plo~akoje ubrzavaju grafi~ku manipulaciju 3D modela.Sli~no, Sun SPARCstation radne stanice imajudovoljno procesor snage da upravljaju video telekon-ferencijama i dekompresiranje MPEG video bezupotrebe unutra{njeg dodavanja plo~a.20-HS11upotrebljava jedan hiper SPARC procesor pokrenutkod 266 MHz i moè upravljati i sa 64MB glavnememorije.Unutra{nje dodavanje kartica ubrzava kliping i rotaci-ju 3D slika. Neke napredne mreè startaju evolvi-ranjem drugog sloja hijerarhije iznad koje je fajlserver. Ideja je spojiti fajl server preko jednog{irokog podru~ja mre`nog rada ka masivnom arhivi-ranju olak{ane sadràvanjem opti~kih diskova juke-boxes, i magnetnih traka ili biblioteke opti~kih disko-va. Ovaj dodatni sloj je osnovno jedan near-line sis-tem arhiviranja koji drì informaciju raspoloìvu sapristupnim vremenom reda sekunda ili duè. Zaprimjer, mogu}e je slanje samo 2D crteà jednogdijela da bude u servisnom birou kreiran kao solidmodel, a na osnovu printa. Za relativno jednostavnedijelove, ovo }e vjerovatno uzeti vremena isto kaoda se kreira solid model sa jednim CAD programom.Ali situacija se dramati~no mjenja kada dijelovipostaju vi{e i vi{e kompleksni. Neki CAD programiba{ zbog ovoga rade brè jedan od drugog.Feature-based ili modeleri varijacione geometrije kao{to je Pro/Engineer mogu obi~no generisati modele


- 117755 -

2.3 Client / server nature

Advances in printers, workstations, and storagemedia help make engineers more productive. Thehierarchy used in CAD/CAM networks today typicallyis one of a client/server nature. Each workstation,orclient, has its own central processing unit (CPU) andlocal disk storage. The network may also have a fileserver that contains CPU and hard disk storage, aswell as magnetic tape storage. A key trait ofclient/server systems is that the file-server softwareprovides file abstraction to the client. The worksta-tion user doesn't need to know where the files phys-ically reside. This behavior is courtesy of what iscalled a virtual-file system which allows sending filesystem requests to a local workstation file system orto a remote server across the network. Engineering workstations have in recent years sport-ed multimedia capabilities. Silicon Graphics Inc. haslong been considered the leader in graphics pro-cessing. The SGI Indy series workstation, based ona 64-bit Risc processor, has internal audio captureand optional camera for video teleconferencing. Otherfeatures include display resolution to 1,280 × 1,034,main memory capacity to 256M bytes, and the avail-ability of add-in boards that accelerate graphic manip-ulations of 3D models. Similarly, Sun SPARCstationworkstations have sufficient processing power to han-dle video teleconferencing and decompressing MPEGvideo without the use of add-in boards. The 20-HS11 uses a hyperSPARC processor runningat 64 MHz and can handle up to 64M bytes of mainmemory. Add-in cards accelerate the clipping androtation of 3D images. Some advanced networks arestarting to evolve another layer of hierarchy abovethat of the file server. The idea is to connect thefile server over a wide area network to massive stor-age facilities consisting of optical-disk jukeboxes,and mag tape or optical-disk libraries. This addi-tional layer is basically a near-line storage systemthat keeps information available with access timeson the order of seconds or longer. For example,sending only 2D drawings of a part to be fabricat-ed forces the service bureau to create the solidmodel from the prints. For relatively simple parts, itwill probably take about as long to create a solidmodel with one CAD program as it would withanother. But the situation changes dramatically asparts become more and more complex. Some CADprograms are just faster to work with than others. Feature-based or variational geometry modelers suchas Pro/Engineer can usually generate models muchmore quickly than modelers based on Boolean oper-ators. The difference in modeling time becomes morepronounced in complicated models that incorporatefeatures such as sculpted surfaces, numerous bends

mnogo brè od modelera baziranih na Bulovimoperacijama. Razlika u vremenu modeliranja postajevi{e izraèna u komplikovanijim modelima koji inkor-poriraju karakteristike takve kao skulptorske povr{ine,brojne prelome, savijanja,radijuse itd.Izraz virtualnarealnost dolazi iz zna~enja sposobnosti " {etatiokolo" 3D matemati~kog modela i pogleda na njihkao da oni fizi~ki egzistiraju u prostoru.Neki CAD modeli sada mogu biti pogledani u virtu-alnoj stvarnosti kroz upotrebu specijalnih pogleda{ljemova ili ~a{a, i softvera koji formate video slikakreira utiskom dubine. Utisak dubine rezultira izvi|enja dva perspektivna pogleda vi|enih pomo}una{eg lijvog i desnog oka.Imaju dva generalnana~ina kreiranja ovih pogleda. Pristup upliva u virtu-alnu stvarnost {ljemova je da treba upotrebiti dvamala ekrana, jedan za svako oko, i poslati svakipogled respektivnom ekranu. Jedan drugi metod,kojije vi{e tipi~an upotrebljava se u CAD aplikacijama,multipleks dvije slike na sekvencijalnim video poljimajednog obi~nog monitora. Posmatra~ koristi jeftinijiset ~a{a koje sadràvaju jedan infrerirani prijemnik ipolarizirane lijeve i desne le}e.Jedan infreriranioda{ilja~ sinhronizovan ka promjenljivom video polji-ma {alje signale ka oku koji uzrokuje svaka le}a za-tvaranjem i otvaranjem odgovaraju}e vrijeme, tako dasvako oko vidi korespondiraju}i lijevi ili desni pogled.Konstruisati trodimenzionalne displeje iz jednog mod-ela u jednu CAD bazu podataka, softveru mora prvobiti data jedna ta~ka pogleda posmatra~a. Sa infor-macijom o ta~ki pogleda, kompjuter moè ra~unatikoje povr{ine CAD modela trebaju biti vidljive iz ovepojedina~ne perspektive. Tipi~an pristup sortiranjuvidljivih povr{ina starta sa onim najzatvorenijim kaposmatra~u i radi obrnuto. Najzatvoreniji objektiblokiraju pogled nekih objekata udaljenijih, i takonadalje. Postoje tako|er dvije kliping operacijepridruène sa pogledom na ekranu. Prva operacijaupotrebljava rastojanje izme|u posmatra~a i modela,ra~unaju}i koji dio modela treba biti vidljiv naekranu. Jo{ mnogo toga izvan posmatranog okviraje ignorirano. Druga operacija odlu~uje koji predmetine bi trebali biti vidljivi, budu}i oni su kompletnoskriveni putem objekata bliìh posmatra~u.Jedanputkada kompjuter prora~una koje povr{ine su vidljive,on odre|uje boju svake vidljive povr{ine, koje zavisei od osvjetljenja i od originalne boje povr{ine.Napredni CAD programi mogu ~esto dodijeliti jednuraznolikost simuliranih svjetlosnih izvora takva kaospot svjetla, obojena svjetla, usmjereni svjetlosniizvori kao poput svjetla koja dolaze kroz jedan pro-zor, i tome sli~no.Jednom CAD program odre|ujeosnovnu koloraciju, i on prora~unava sjenu za svakipiksel na displeju. Napredni programi optimizirajurealnost i mogu uposliti sofistificirane metode zagenerisanje ove vrste {ejdinga. Prora~uni involvirani,


- 117766 -

and radii, and so forth. The term virtual reality hascome to mean the ability to "walk around" 3D math-ematical models and view them as though they phys-ically existed in space. Some CAD models now canbe viewed in virtual reality through use of specialviewing helmets or glasses, and software that formatsvideo images to create the impression of depth.The impression of depth results from seeing two per-spective views of a three-dimensional object that cor-respond to the same views seen by our left and righteyes. There are two general ways to create theseviews. The approach employed in virtual reality hel-mets is to use two small video screens, one for eacheye, and send each view to the respective screen. Asecond method, which is more typically used in CADapplications, multiplexes the two images onto thesequential video fields of an ordinary monitor. Theviewer wears an inexpensive set of glasses that con-tain an infrared receiver and polarized left and rightlenses. An infrared transmitter synchronized to thealternating video fields sends signals to the eyewearwhich cause each lens to shutter on and off at theappropriate time so that each eye sees the corre-sponding left or right view. To construct three dimen-sional displays from a model in a CAD database, soft-ware must first be given a viewpoint of the observer.With information about the viewpoint, the computercan calculate which surfaces of the CAD model wouldbe visible from that particular perspective. The typical approach to sorting visible surfacesstarts with those closest to the viewer and worksback. The closest objects block the view of someobjects farther away, and so forth. There are alsotwo clipping operations associated with the view onthe screen. The first operation uses the distancebetween the viewer and the model to calculatewhich part of the model would be visible on thescreen.Everything else outside the viewing box isignored. The second operation decides whichobjects would not be visible because they are com-pletely hidden by objects closer to the viewer.Once the computer has calculated which surfacesare visible, it determines the color of each visiblesurface, which depends both on lighting and on theoriginal surface color. Advanced CAD programs canoften assign a variety of simulated light sourcessuch as spot lights, colored lights, directional lightsources as like light coming through a window, andso forth. Once the CAD program has determinedbasic coloration, it calculates a specific shade foreach pixel in the display. Advanced programs optimized for realism mayemploy sophisticated methods for generating this sortof shading. The calculations involved, however, arelengthy and work best on high-end graphic work-stations. High-end software may also add refine-

me|utim su duì i rade najbolje na krajnje visokografi~kim radnim stanicama. High-end softver moètako|e dodati rafinate takve kao antialiasing smanjitijaggedness dijagonalnih linija, transparentnost ve}produciranih efekata, takvi kao magla ili zamazana~a{a, i tekstura mapinga umjesto ra~unanja kolor vri-jednosti. Ovo daje izgled povr{ina kao {are drveta.

3. UPRAVLJANJE PODACIMA PROIZVODA

U mnogim firmama, porast obima dizajna baziranog nara~unarima i analize podataka prijete preoptere~enjusistema za upravljanje ovim podacima. Upravljanjepodacima proizvoda ( PDM ) softver je dizajniran takoda osigura one inènjere koji su upotrebljavali praveverzije dizajniranja proizvoda podacima, spre~avaju}ineautorizovane osobe od pregledavanja ili promjenepodataka, i osiguravanja onih promjena koje su slo-bodne samo poslije odgovaraju}eg pregleda i dozv-ole. Najsofistificiraniji sistemi mogu trasirati koje verzi-je podataka su upotrebljene u izradi mnogih proizvo-da. Ne kao specifi~ni CAD ili CAE softver, PDM sis-temi dodiruju organizaciju cjelokupnog dizajn procesa.Vi ne moète kupiti jedno PDM mjesto da vidite kakoto dobro radi budu}i jedan sistem koji upravlja malimiznosima podataka moè biti prekinut pod punimoptere~enjem va{e kompanije podacima modeliranja iprocedurama. Pronalaza~, da li brzo prototipisanje jekompatibilno sa va{om teku~om praksom i {ta va{afirma mora u~initi da bi to uspje{no upotrebljavala. U~iti o novim aplikacijama i produktivnosti-gradnjatehnika iz posebnih razgovora sa korisnicima eksper-tima. Za primjer, prona}i kako kreirati brzo pokretni iproizvodni alat koriste}i brzo prototipisanje i kakopobolj{ati ta~nost i izlaz postoje}ih sistema. Vi }etetako|e izu~iti koliko to ko{ta i vrijeme osiguranja kojemoète o~ekivati od ovih novih metoda. Feature-basedmodeliranje usmjereno je ka konstrukciji geometrijakao jednoj kombinaciji karakteristi~nih oblika. Konstruktor specificira karakteristike u inènjerskimizrazima takve kao otvore, proreze ili {upljine radijenego kao geometrijske izraze takve kao krugovi ilikutije. Features-i mogu tako|e pohraniti negrafi~keinformacije isto tako dobro kao i geafi~ke. Ova infor-macija moè biti upotrebljena u aktivnostima takvekao {to su drafting, NC, analiza kona~nih elemena-ta i kinematska analiza. Nadalje, feature-based paketi~esto snimaju geometrijsku konstrukciju i modifikaci-ju sekvenci upotrebljenih i gradnji modela. Nekeverzije softvera za solid modeliranje koriste primitiveili grani~nu reprezentaciju kao gradnja blokova zamodelirane dijelove. Sa primitivima, elementarni obli-ci su kombinovani u jednoj gradnji-blok modernihkreiranih novih oblika. Bulove logi~ke komande, takve kao unija,razlika i


- 117777 -

ments such as antialiasing to reduce the jaggednessof diagonal lines,transparency for producing effectssuch as fog or tinted glass, and texture mappinginstead of calculating a color value. This providesthe appearance of surfaces such as wood grains.

3. PRODUCT DATA MANAGEMENT

In many firms, the growing volume of computer-based design and analysis data threatens to over-whelm systems for managing that data. Productdata management (PDM) software is designed toassure that engineers are using the right version ofproduct design data, prevent unauthorized peoplefrom viewing or changing data, and assure thatchanges are released only after appropriate reviewand approval. The most sophisticated systems cantrack which versions of data are used to manufac-ture each lot of products. Unlike specific CAD orCAE software, PDM systems affect an organization'sentire design process. You can't buy one PDM seatto see how well it works because a system thatmanages small amounts of data might break underthe full load of your company's design data andprocedures.Discover whether rapid prototyping is compatiblewith your current practices and what your firm mustdo to use it successfully. Learn about new applica-tions and productivity-building techniques from exclu-sive interviews with expert users. For example, findout how to create short-run and production toolingusing rapid prototyping and how to improve accu-racy and output of existing systems. You'll also learnwhat cost and time savings you can expect fromthese new methods.Feature-based modeling refers to the construction ofgeometries as a combination of form features. Thedesigner specifies features in engineering termssuch as holes, slots, or bosses rather than geo-metric terms such as circles or boxes. Features canalso store nongraphic information as well. This infor-mation can be used in activities such as drafting,NC, finite-element analysis, and kinematic analysis.Furthermore, feature-based packages frequentlyrecord the geometric construction and modificationsequences used in building the model. Some ver-sions of solid-modeling software use primitives orboundary representations as the building blocks formodeled parts. With primitives, elementary shapesare combined in a building-block fashion to createa new shape. Boolean logic commands, such as union, difference,and intersection, aid in forming new shapes. Withboundary definitions, 2D surfaces are swept through

presjek pomaù u formiranju novih oblika. Sagrani~nim definicijama, 2D povr{ine su provu~enekroz prostor trasiraju}i volumene. Mnogi sistemi nudenekoliko tipova svipova pomaù}i kreiranju jedneraznolikosti oblika. Parametarski metodi zavise odsekvenci operacija upotrebljenih u konstruisanjunekog ma{inskog dijela. Softver odràva jedan istori-jat promjena u specificiranim parametrima. Ta~ka spa-janja ove istorije drì trag operacija koje zavise jednaod drugih, tako da kad god se traì promjena jednespecifi~ne dimenzije, sistem moè prilagoditi sveoperacije referentne prema ovoj dimenziji.Za primjer, jedan krug predstavljaju}i jedan otvor zaklin moè biti konstruisan tako da je to koncentri~noka kru`nom prorezu. Ako prorez se kre}e tako ~inii krug za klin. Parametri su obi~no pokazani i izraz-ima dimenzija ili natpisa i sluè kao mehanizmipomo}u kojih se mjenja geometrija. Konstruktormoè promjeniti parametre ru~no mjenjaju}i jednudimenziju, ili pomo}u referencije njih ka jednoj var-ijabli u jedna~ini koja je rje{ena ili pomo}u samogmodeling programa ili pomo}u vanjskih programatakvih kao {to su spreadsheets.Parametarsko modeliranje je najefikasnije u radu sadizajnima gdje promjene su vjerovatne ka sadràjudimenzionalnih promjena, radije nego velikim razlika-ma geometrija. Ovaj ~lanak predstavlja jedan kon-cept i strukturu sistema za modeliranje. U skladu sadizajn iskustvom u inènjerstvu, sistem osposobljavaskicu dizajna i analizu napona formiraju}i jednu bazuza detaljisanje. Jedan komplet softver program jerazvijen za upotrebu sa CNC PUNCH/LASER PLAS-MA ma{inama. Sistem potpomaè cjelokupni ciklusCNC operacije i uklju~uje drafting, spre~avagu`vanje materijala, automatsko/ interaktivno gener-isanje G-koda, grafi~ku simulaciju CNC programa,gnje`|enje, downloading i uploading. Softver potpo-maè jedno {iroko podru~je ma{ina uklju~uju}iAMADA, TRUMPF, WHITNEY, BISTRONIC itd.U dodatku, Nesting softverski paket snabdjeven jejednim kompletnim rje{enjem za interaktivno iautomatsko gnje`|enje razli~itih `lijebova. Windows-osnovna aplikacija je kompatibilna sa jednim {irokimpodru~jem CAD/CAM sistema ( DFX,IGES,CADL).Pomak promjena u inènjerstvu i nau~nim ra~unari-ma, CAD/CAM i CAE sistemima i brzo prototipisan-je predstavljaju jednu dilemu. Da li uglavnom koris-titi alate koji ve} su na raspolaganju, ili se upustitiu domen novih tehnologija koje su revolucionarnena trì{tu koje je izrazito konkurentsko. Izvje{taji naproizvodne pogre{ke omogu}uju da se izbjegnu sis-temi sa gre{kama u ~ipovima i mati~nim plo~ama.Tehni~ki prora~uni zavise od poslovnih aplikacija.Inènjeri, nau~nici, ekonomisti i ostalo tehni~koosoblje daju brojne intenzivne zadatke. Konstruktoriproizvoda, postrojenja, gra|evina i infrastrukture traè


- 117788 -

space to trace out volumes. Most systems offer sev-eral types of sweeps to help create a variety ofshapes. Parametric methods depend on thesequence of operations used to construct thedesign. The software maintains a history of changesin specific parameters. The point of capturing thishistory is to keep track of operations that dependon each other, so that, whenever it is told tochange a specific dimension, the system can updateall operations referenced to that dimension. Forexample, a circle representing a bolt hole may beconstructed so it is always concentric to a circularslot. If the slot moves, so does the bolt circle. Para-meters are usually displayed in terms of dimensionsor labels and serve as the mechanism by whichgeometry changes.The designer can change para-meters manually by changing a dimension, or by ref-erencing them to a variable in an equation that issolved either by the modeling program itself or byexternal programs such as spreadsheets. Parametricmodeling is most efficient in working with designswhere changes are likely to consist of dimensionalchanges rather than grossly different geometries. This paper presents a concept and the structure ofthe system for the design. According to designexpirience in engineering, system enables layoutdesign and stress analysis forming a base for detail-ing.A complete software program has been devel-oped for use with CNC PUNCH/LASER PLASMAmachines. The system supports the entire cycle of CNC oper-ation and includes drafting, unfolding,automatic/interactive G-code generation, the graphicsimulation of CNC programs, nesting, downloadingand uploading. The software supports a wide range of machinesincluding AMADA, TRUMPF, WHITNEY, BISTRONICand more. In addition, the Nesting software pack-age provides a complete solution for interactiveand automatic nesting of different true-shape pat-terns. The Windows-based application is compati-ble with a wide range of CAD/CAM systems (DXF,IGES, CADL). The pace of change in engineeringand scientific computers, CAD/CAM and CAE, andrapid prototyping presents you with a dilemma.You have to make the most of the tools youalready have, while keeping up with new tech-nologies that are revolutionizing the market inwhich you compete.Reports on manufacturing defects, so you can avoidsystems with faulty chips and boards. Technicalcomputing differs from business applications. Engi-neers, scientists, economists, and other technicalprofessionals perform numerically intensive tasks.Designers of products, plants, buildings, and enter-tainment media need fast graphics performance to

brze grafi~ke performanse da bi radili efikasno.Korisnici tehni~kih prora~una ne trebaju da se nekeperiode posve}uju aplikacijama kao {to suprora~uni, knjigovodstvo, bankarstvo, osiguranjeprava u procesu i obnova. Oni trebaju i to zasluùjujedan newsletter posve~en unikatnim zahtjevimatehni~kih profesionalaca. Trasiranje ta~nih i detaljnihinformacija o mena|mentu podacima proizvoda (PDM ) kao sistema je vrlo te{ko. Razli~iti magaziniizvje{tavaju o raznim temama u procesu konstru-isanja, pa ~ak njihove korice su ~esto koloritne dabi zainteresovale dizajnere. To je tako|e skupo jerPDM sistem je takav sistem koji mora odgovaratispecificiranim aplikacijama.Zato je potrebno izna~i iprepoznati na osnovu iskustva stvarni PDM sistemkoji radi pouzdano, a izbje}i onaj koji ne odgo-vara.Ovaj rad pomaè u ovakvoj odluci , ali i kadase mjenja pejza` PDM sistema. On {tedi vrijemebudu}i su prezentovane informacije koje su potreb-ne da bi konstruktori i projektanti donijeli praveodluke. Ovo je odgovor i na dilerove zahtjeve, a bal-ansira i propagandne istupe.Na taj na~in ima se prava informacija, a da se negubi previ{e vremena oko toga. Informacije dobiveneiz ovog ~lanka tako|er obezbje|uju izbjegavanjevendora i tehnologija koje ne mogu ili ne}e zado-voljiti trenutne ali i budu}e zahtjeve dizajnera.Tako|er obuhvataju i ko{tanje i proizvodne trendoveu mnogo detalja, daju}i spoznaje pri pregovaranju onabavci nekog softvera. Naravno svaki dizajner bibio prezadovoljan kada bi postojao jedn tim ekspe-rata da analizira sve promjene, da osigura uvijek radsa najboljim raspoloìvim tehnologijama, i da nikadne investira u novi proizvod prije nego otkloni svebagove u njemu. Mnogi ovakvi radovi daju}i ~itaoci-ma prave informacije u poslovnim i dizajnerskimtrendovima predvidili su i kolaps nekoliko biv{ihCAD/CAM industrijskih predvodnika, prije nego {tosu i do{li u probleme. U 1988 godini neki CAD~asopisi sa entuzijazmom opisali su Pro/ Engineer,koji je tada bio jedan novi solid modeling softver,iz male firme zvane Parametric Technology. Danas,PTC je vode}a industrija u prodaji softvera za solidmodeliranje, i ostvaruje ogromne profite za razliku odmnogo drugih koje jedva preìvljavaju.Vrativ{i se nazad u 1983 godinu mnogi dizajneri suse sve vi{e okretali ka personalnim ra~unarima ipredvi|ali da }e i CAD i CAE sistemi biti popularniba{ na ovim platformama. U 1991 godini neki dizaj-neri su objasnili za{to }e i veliki gigant poput IBM-a do}i u te{ko~e, jer nije ispo{tovao strategijukupaca u hardverskom i softverskom smislu, a kojesu bile poznate 1990 godine.CAD radovi pravilnogovore kako inovirati CAD/CAM tehnologije u ciljupodizanja produktivnosti. Oni izvje{tavaju kojeaplikacije vra~aju najve}u korist, i {to je jo{ bitnije


- 117799 -

work efficiently. Technical computing users don'tneed to wade through periodicals devoted to appli-cations such as accounting, banking, insuranceclaims processing, and retailing. They do need anddeserve a newsletter devoted to the uniquedemands of technical professionals.Tracking downaccurate and detailed information about product datamanagement (PDM) systems is hard. Trade maga-zines barely report on the topic, and even whenthey do, their coverage is often colored by the inter-ests of advertisers. It's also costly to benchmarkPDM systems side-by-side because they must becustomized to your specific application. You need tofind out from the experience of actual PDM userswhich PDM systems work reliably and which youshould avoid. This paper helps you keep up with the ever-chang-ing landscape of PDM. It saves you time becauseit presents the information you need to make deci-sions and explains the details that other publica-tions gloss over. It casts a critical eye on vendorclaims, balancing corporate propaganda with verifi-able case histories. You stay informed withouthours of reading.The information you get frompaper e will help you avoid vendors and tech-nologies that can't or won't meet your current andfuture needs.Also covers pricing and producttrends in detail, giving you an edge in purchasingnegotiations. Wouldn't it be great if you could relyon a team of experts to analyze these changes, tohelp ensure that you're always working with thebest technology available, and that you never investin new products before the bugs have beenworked out? Of course, yes. Keeping readers alert to business trends that mightjeopardize their investments, these papers and jour-nals predicted early the collapse of several formerCAD/CAM industry leaders before their troublesbecame widely known. In 1988, CAD"s journals enthu-siastically described Pro/Engineer, what was then anew solid-modeling product from a small firm calledParametric Technology. Today, PTC leads the industryin sales of solid-based design software, while whatwere once billion-dollar competitors struggle for sur-vival. Way back in 1983, many designers foretold thedecline of turnkey CAD and predicted personal com-puters would become popular platforms for CAD andCAE. In 1991, some designers explained why evengiants like IBM are no longer safe buys, and we out-lined hardware and software purchasing strategiessuitable for the 1990s. The CAD papers regularly tellshow innovative CAD/CAM owners are using the tech-nology to boost productivity and save money. Itreports which applications yield the highest return,and -- just as importantly -- which should be avoid-ed with today's systems. They also write about soft-

koje bi trebalo izbje}i za dana{nje sisteme. Onitako|er pi{u kao i ovaj ~lanak o softveru koji upravl-ja i kontroli{e CAD/CAM podatke u svim elektron-sko inènjerskim uredima.Tako|er u mnoge poslove dizajniranja investirana suogromna sredstva u cilju brzog prototipisanja sistemasamo da se otkrije da li ne{to moè funkcionisati iline, ili je to ve} poznato. Rad pomaè upravo da seizbjegnu takve gre{ke i da se prepoznaju oneCAD/CAM tehologije i PDM sistemi koji daju ~vrstinuu odluci i daju va`ne informacije u ovoj oblasti,kako bi se predvidile sve tehnolo{ke prednosti nekogtek potencijalnog sistema za kori{tenje. Najva`nije,rad daje jasnu sliku koje su postoje}e industrije soft-vera najefikasnije za dana{nje okruènje, kako bi semoglo planirati {ta ~initi u budu}nosti. Novi razvoji umemoriji, mikroprocesorima, multiprocesorima, plote-rima i input /output ure|ajima obezbje|uju najboljeperformanse i pouzdanost uz minimalne tro{kove.Osim toga, prilagoditi najnovije operativne sisteme,uklju~uju}i tu i Windows 95 i Windovs NT i razvojekoji jo{ uvijek mogu dati budu}nost UNIX-u.Ovaj rad je i pronalaza~ kad ne{to u~initi pa i plati-ti premiju za UNIX ili Microsoft Windows NT, a kadato ne ~initi.Isto tako tutoriali na mreàma, grafi~ki dis-pleji i multiprocesing sistemi omogu}uju da se proc-jeni koja od ovih tehnologija radi najbolje za odre|enuaplikaciju. Grafovi budu}ih kompjuterskih vrijednostitako|er usmjeravaju za izbor najoptimalnijeg sistema iizbjegavanje onih sistema koji su preskupi. Trì{te dijelii finansijske podatke za vode}e tvorce kompjutera,tako da je i to jedan od pokazatelja koje tehnologijenestaju odnosno odumiru. Sistem PDM obuhvata dakle{iroko podru~je tematskih cjelina a sve sa ciljem dadizajneri budu dobro informisani u svako vrijeme, a tose prvenstveno odnosi na:

Nau~iti kako staviti pod kontrolu tro{kove,Istraìti kako druge firme su koristile PDM sistem i da li efikasno,Monitoring finansijskog stanja PDM softver kompanija, da bi se izbjegli mogu}i problemi,Prou~iti koji PDM snabdjeva~i su u vrhu novihtehnologija, a koje su u padu,Razumjeti koji standardi oblikovanja se primjenjuju u PDM sistemu razvoja proizvoda,Odrediti koji PDM softver je najbolji za trenutnepotrebe neke kompanije,Izu~iti kako su drugi upravljali nekim funkcijama PDM zahtjeva,Pripreme za PDM softver pomo}u smjernicaprocesa i maping radni tok,Izbje}i PDM provo|enje gre{aka koje buduotkrivene,Prona}i kako male,srednje i velike kompanije


- 118800 -

ware to help you manage and control CAD/CAM datain an all-electronic engineering office. Too many businesses have invested hundreds ofthousands of dollars in a rapid prototyping systemonly to discover that it can't do what they believedor that it is already obsolete.The paper helps youavoid such mistakes by printing in-depth, objectiveappraisals of the strengths and weaknesses of eachtechnology and keeps you aware of important newresearch and development activities and interpretsthis research so you can foresee technologicaladvances. Most important, paper gives you the big picture ofwhere this exciting new industry is heading, so youcan plan for the future. New developments in mem-ory, microprocessors, multiprocessing, plotters, andinput/output devices that boost performance andreliability while lowering prices. Updates on the lat-est operating systems, including the maturing ofWindows 95 and Windows NT, and developmentsthat may yet give Unix a future. Discover when it makes sense to pay premiums forUnix or Microsoft's Windows NT and when you don'tneed too. Tutorials on networking, graphics displays,and multiprocessing systems, so you can evaluatewhich of these technologies works best for yourapplications. Graphs of future computer values soyou can dispose of existing systems at fair priceswhile avoiding paying too much for used systems.Market shares and financial data on leading com-puter makers so that you don't risk getting strand-ed with a technology that's dying. Product DataManagement covers a wide range of topics andyou'll be armed and ready to get the best dealevery time and it will help you:

Learn how to put the brakes on out-of-controlspending. Find out how other firms have used PDM systems effectively. Monitor the financial health of PDM software companies, so you don't get stuck with a financially troubled supplier. Learn which PDM suppliers are on the cutting

edge of technology and which are falling behind. Understand how standards are shaping PDM product developments. Determine which PDM software products are best for your company's needs and budget. Learn how others have managed cross-functional PDM requirements. Prepare for PDM by streamlining processes and mapping workflow. Avoid PDM implementation mistakes that can wreck your career. Discover how small, medium, and large companies budget for PDM.

ulaù u PDM sistem,Izu~iti kako objasniti PDM koncepte jednostavnomenadèrima i ostalim korisnicima,Za{tititi investicije pomo}u osiguranja da PDMsistem uvijek bude pouzdan,Upore|ivati vlastiti PDM napredak sa drugim kompanijama.

Implementacija i kori{tenje PDM sistema je rizi~no ikompleksno, ali PDM pomaè da se minimiziraju ovirizici daju}i niz informacija. Ako PDM pomogne dase izbjegnu npr. gre{ke na nekoliko crteà, u{tededaleko nadma{uju uloèna sredstva u PDM sistem.

4. ZAKLJUÂK

Najefikasniji op{ti alati za ovaj proces konstruisanjasu kompjuterski sistemi. Specifi~niji kompjuterski sis-temi su potrebni onda kada inkorporiraju i modeli-ranje alata baziranog na proizvodnim naukama. Ovamodeliranja alata moraju nas snabdjeti sa dovoljnimmogu}nostima za ta~nost i korektno kreiranje mod-ela na{e industrijske realnosti i sa njihovim proizvodi-ma, trì{nim uslovima, konstrukcijama, alatnimma{inama,upravljanje sistemima i ergonomskim zaht-jevima. Ovi modeli i kompjuterski podràni sistemibazirani na njima moraju biti razumljivi za inènjere.Uspje{an dizajn proizvoda zahtjeva realizaciju neko-liko razli~itih koraka. Za primjer, jedan dizajn procesmoè involvirati korake takve koji su spregnuti sapotrebama kupca i osiguranjem toga oni su podesnoformulisani kao jedan dizajn problem, zatim dekom-pozicija velikog problema u zadatke veli~ine sa koji-ma je mogu}e upravljati,,zatim dodjela zahtjevnimzadacima aktivnih sudionika na svakom zadatku iosiguranje da rezultati proizvedeni pomo}u zahtjevnihzadataka su integrativni i prihvatljivi.Svaki korak je poduzet od dizajn predstavnika, soft-verskih alata, ili kombinacija i jednog i drugog. Inte-gracioni alati i ljudstvo i dodjeljena uloga svakom odnjih su kriti~ne faze za uspje{no kompletiranje proce-sa konstruisanja. Postoje}i konstruktorski sistemi teèupotrebi razli~itih struktura za ljudstvo i za softverskealate ( CAD alati ). U cilju humanosti , npr arhitektedopu{taju i predplaniranje i spontanu kolaboracijuizme|u alatki u formi grupa, komiteta i timova. Kaorezultat, fleksibilnost je relativno visoka.U suprotnom,kreacija kompleksnih organizacija softverskih alata jejedan teàk zadatak. Relativni nedostatak fleksibilnihsoftverskih mjesta alata ne{to teè nagla{ava ljudskuulogu u procesu konstruisanja. Me|utim, uve~avaju}ikompleksnost dizajna i potrebe za kra}im dizajn cik-lusom zahtjeva od dizajnera, da se pitaju kakvi mogubiti oni iza njihovih sposobnosti .Premda, softverskialati i dizajneri trebaju biti u kooperaciji i zajedni~ki


- 118811 -

Learn how to explain PDM concepts simply to managers, peers, and subordinates. Protect your investment by ensuring that PDM systems are used wisely. Compare your PDM progress against other companies.

Implementing and using PDM systems is risky andcomplex and the PDM helps you minimize thoserisks by giving you information . If the PDM helpsyou avoid a six-figure mistake, it will have paid foritself hundreds of times over.

4. CONCLUSION

The most efficient general tools for this design workare computer systems. More specifically, computersystems are needed which incorporate modellingtools based on manufacturing science. These mod-elling tools must provide us with sufficient possibil-ities for accurately and correctly creating models ofour industrial realities with their products, marketconditions, buildings, machine tools, handling sys-tems and working human beings. These models andthe computer support systems based on them mustbe understandable for engineers. The successfuldesign of a product requires the realization of sev-eral distinct steps. For example, a design processmay involve steps such as capturing the customers"needs and ensuring they are properly formulated asa design problem; decomposing the overall probleminto tasks of manageable size; assigning a taskforce of active agents to each task; and ensuringthat the results produced by the task forces areintegrated into an acceptable, overall solution. Each step is undertaken by human agents, softwaretools, or a combination of both. The integration oftools and humans and the assignment of the roleto each one of them are critical phases for the suc-cessful completion of the design process. Existingdesign systems tend to use different structures forhumans and for software tools ( CAD tools ).For humans, architectures allow for both preplannedand spontaneous collaborations among the tools inthe form of groups, committees and teams. As aresult, flexibility is relatively high. In contrast, the cre-ation of complex organizations of software tools isa difficult task. The relative lack of flexible software tools places aheavy emphasis on the human"s role in design sys-tems.However, the increasing complexity of designsand the need for shorter design cycles demandfrom humans what may be beyond their capabilities.Thus, software tools and humans need to be ableto co-exist and cooperate in a design environment.

egzistirati u jednom dizajn okruènju. Jedno mogu~erje{enje problema integracije alata i dizajnera obuh-vata tri linije rada: prvo, razvoj fleksibilnosti, dvodi-menzionalne strukture u kojima alati mogu brzo ilagano biti integrisani; drugo, postizanje alata sposob-nih da u~estvuju u ovim strukturama; i tre}e, razvojkontrolnih shema za operisa-nje rezultuju}im organi-zovanjima.Glavna namjera ovog rada je pobolj{ati efikasnostprocesa konstruisanja i kvaliteta rezultuju}ih dizajna. Zaproizvodnu industriju odgovoriti ~ak uve~anim zahtjevi-ma, za fleksibilnost, kvalitet, brzinu i ekonomiju, kon-struktori moraju znati kako primjeniti moderneprora~une i automatizaciju tehnologije. Dizajn i plani-ranje proizvodnih sistema je jedna aktivnost ogromnogzna~aja za industrijske konkurentnosti. Zdrave analizepomaù modeliranje i proces planiranja i to je klju~postizanja efikasnih operacija takvih sistema. Strogaoptimizacija metodologija i ta~no modeliranje tehnikepropusta, u mnogim slu~ajevima, adresiraju ogromnukompleksnost proizvodnog sistema. Integracija dizajna iproizvodnje je va`na ne samo za ve}u efikasnost, negotako|e pobolj{ava kvalitet i u dizajnu i u proizvodnji.Moderni proizvodni sistemi su visoko tehnolo{ki alatikoji nas snabdjevaju sa sredstvima produciranja fizi~kihproizvoda sa funkcijama koje ta~no predstavljaju indi-vidualne potrebe razli~itih korisnika. Ovi sistemi,podesno kori{teni, daju nam prakti~ne, produktivnemogu}nosti postizanja korisnikove prilago|eneproizvodnje direktno prema naru|bi. Kupca Jedan per-spektivni dizajn ovih sistema daje nam mogu}nostadaptiranja sistemskih performansi prilago|enih zahtje-vima radnika i proizvodnih planera. U ovom slu~ajuljudski faktor postaje jedan vaàn sastojak u uspje{nojoperaciji modernih proizvodnih sistema.

LITERATURA - REFERENCES

[1] Obermann, K.: CAD/CAM, Handbuch, VerlagComputergrafik GmbH, Muenchen, 1995.

[2] Williams, D:J.: Manufacturing Systems, Chapmanand Hall, London, 1992.

[3] Lee, M.H.: Intelligent Robotics, Chapman andHall, London, 1991.

[4] Knight, P.T., Kim, H.S.: A knowledge system forintegrated product design, Journal of Intelligent Man-ufacturing, London,1991.

[5] Rep~i}, N.: CAD metode, Ma{inski fakultet Sara-jevo, Sarajevo, 1996.

[6] Hasegawa, Y.: Evaluation and Economic Consid-erations, Handbook of Industrial Robotics,,John Wiley & Sons, New York, 1988.


- 118822 -

A possible solution to the problem of integrating toolsand humans involves three lines of work; first, todevelop flexible, two dimensional structures into whichtools can quickly and easily be integrated; second,to obtain tools capable of participating in these struc-tures; and third, to develop control schemes for oper-ating the resulting organizations.The main purpose ofpaper is to improve the efficiency of the designprocess and the quality of the resulting designs.For manufacturing industries to respond to everincreasing demands, for flexibility, quality, speed andeconomy, they must know how to apply moderncomputing and automation technology.The designand planning of manufacturing systems is an activ-ity of immense importance for industrial competi-tiveness. Sound analysis to support the design andplanning process is the key to achieving efficientoperation of such systems. Strict optimization methodologies and exact model-ing techniques fail, in many cases, to address thetremendous complexity of the manufacturing sys-tem.The integration of design and manufacture isimportant not only for greater efficiency, but also toimprove quality in both design and manufacture.Modern manufacturing systems are high technolo-gy tools that provide us with the means to producephysical products with functions that accurately meetthe individual needs of the various customers. These systems, properly utilized, give us practical, pro-ductive possibilities to achieve customer adapted pro-duction directly upon customer order. A perceptivedesign of these systems gives us the possibility toadapt system performance to suit the requirements ofthe workers and production planners. In this wayhuman factors become an important ingredient in thesuccessful operation of modern manufacturing systems.

[7] Maddux, K.C., Jain, S.C.: CAE for the manufac-turing engineer, Manufacturing Simulation andProcesses, ASME,20, New York,1990.

[8] Jackson, P., Reichgelt, H.: Logic Based Knowl-edge Representation, The MIT Press, Cambridge,USA,1991.

[9] Kochan, D.: CAM developments in computerintegrated manufacturing, Springer Verlag, Berlin,1992.

[10] Cheng, Y.Q., Yang, J.Y.: An expert system forunderstanding assembly drawings, Advances in Mod-elling and Simulation, 21, 1991.

[11] Web site-http:// www. cad/cam.com / science /engineering / design /

1. INTRODUCTION

Hardness measurements represent one of the mostconvenient methods for quality assurance in the fieldof mechanical characterization of materials. Due to itssimplicity and its quickness of execution, Rockwellhardness test is the most widely employed. In lastyears particular attention has been devoted to the influ-ence of the indenter, which can be considered oneof the main source of uncertainty in Rockwell hard-ness measurements [1, 2] This influence can be relat-ed either to the shape of the indenter (geometry, sur-face roughness) and to its mechanical behavior(deformations under the action of loads, hysteresis) [3].

Ma{instvo 4(2), 183 - 192, (1998) G.Barbato,...: INFLUENCE OF THE INDENTER SHAPE...

- 118833 -

IINNFFLLUUEENNCCEE OOFF TTHHEE IINNDDEENNTTEERR SSHHAAPPEE IINN RROOCCKKWWEELLLL HHAARRDDNNEESSSS TTEESSTT*

Giulio Barbato, Maurizio Galetto, (DSPEA, Politecnico di Torino, 10129 Turin, Italy)Alessandro Germak, Fabrizio Mazzoleni, (Istituto di Metrologia "G. Colonnetti" -CNR, 10135 Turin, Italy)

SUMMARY

One of the main problems related to Rockwell hardness measurements is the influence of the inden-ter shape. In the present work the effects of the macro-geometry of the diamond indenter are ana-lyzed as influencing parameters in Rockwell C test. Both an experimental analysis is carried out, usingIMGC National Standard Machine, and a finite element simulation (FEM) is performed. Four different theoretical models are proposed to interpret experimental results: the first model isbased on Brinell definition of hardness, the second on Meyer analysis and the last two are basedon the evaluation of the deformation work through the analysis of the indentation volume. These dif-ferent corrections are compared to evaluate their performances.

Key words: indenter shape, hardness measurements,FEM,

UUTTIICCAAJJ OOBBLLIIKKAA UUTTIISSKKIIVVAA^̂AA KKOODD IISSPPIITTIIVVAANNJJAA TTVVRRDDOO]]EE PPOO RROOCCKKWWEELLLLUU*

Giulio Barbato, Maurizio Galetto, (DSPEA, Politecnico di Torino, 10129 Turin, Italy)Alessandro Germak, Fabrizio Mazzoleni, (Istituto di Metrologia "G. Colonnetti" -CNR, 10135 Turin, Italy)

REZIME

Jedan od osnovnih problema vezanih za mjerenje tvrdo}e po Rockwellu je uticaj oblika utiskiva~a. Uovom radu se analiziraju efekti makro-geometrije dijamantskog utiskiva~a kao uticajni parametri u Rock-well C postupku. Izvr{ene su i eksperimentalna analiza, kori{tenjem IMGC nacionalne standardnema{ine, i simulacija metodom kona~nih elemenata (MKE).Predloèna su ~etiri razli~ita teoretska modela za interpretaciju eksperimentalnih rezultata: prvi modelse bazira na Brinellovoj definiciji tvrdo}e, druga na Meyerovoj analizi a posljednja dva se baziraju narazvijanju deformacionog rada kroz analizu zapremine utiskivanja. Ove razli~ite korekcije su upore|eneda bi se razvile njihove performanse.

Klju~ne rije~i: oblik utiskiva~a, mjerenje tvrdo}e, MKE

1. UVOD

Mjerenja tvrdo}e predstavljaju jedan od najprikladni-jih metoda za osiguranje kvaliteta u polju mehani~kekarakterizacije materijala. Zahvaljuju}i jednostavnosti ibrzini provo|enja postupka, naj~e{}e se upotreblja-va ispitivanje tvrdo}e po Rockwellu. Posljednjih god-ina posebna pa`nja se poklanja uticaju utiskiva~a,za koji se smatra da je osnovni uzrok mjerne nesig-urnosti mjerenja tvrdo}e po Rockwellu [1, 2]. Ovajuticaj se moè povezati i sa oblikom utiskiva~a(geometrija, hrapavost povr{ine) i sa njegovimmehani~kim osobinama (deformacije usljedoptere}enja, histerezis) [3].

CCOONNFFEERREENNCCEE PPAAPPEERR

RRAADD IIZZLLOO@@EENN NNAA NNAAUU^̂NNOOMM SSKKUUPPUU

* This paper was published at HARDMEKO ’98, Beijing, China on September 21-23, 1998, * Ovaj rad je objavljen na konferenciji HARDMEKO ’98, Peking, Kina, 21-23. septembra 1998. godine

The effect of indenter shape is widely investigated andmany authors present different kind of approach tosolve this problem [4, 5]. The errors in Rockwell Chardness values caused by angle and radius errors ofthe indenter are summarized in Fig. 1.a and Fig. 1.b1,[6, 7, 8, 9]. The zone between the two continuous linescontains the errors, as resumed by Petik [1], due to aradius difference of +0,01 mm and an angle differenceof +30’. Inside of the two included zones the linescorresponding to the theoretical evaluation due toBochmann & Hild (corresponding to "Brinell method")[6, 10] and the regressions proposed by Yamashiro &Uemura [8] and Stepanow [7] are reported.Fig. 1.a and Fig. 1.b are in agreement with thepractical rule proposed by Yamamoto & Yano [11],according to which the uncertainty of hardness stan-dardizing machines due to shape errors of theindenter should be of the order of ±1 HRC. It must be noted that other experimental data byWood et al. [12], who analyzed the performance of20 indenters from 5 different producers, result intoan hardness error higher than the zone given in Fig.1.a and Fig. 1.b.

The problem has two main drawbacks: on the onehand the experimental plans of the star type thatseems to be used in the past cannot give infor-mation about second-order effects, on the secondhand factors outside of geometry can produce anexperimental noise that hide geometry effects. Forthese reasons we tried to use an experimental planbased on a factorial and a star scheme comparedwith the results obtained by the simulation with FEM.To keep small the relative effect of various influence


- 118844 -

Uticaj oblika utiskiva~a je dosta istraìvan i mnogiautori su dali razli~ite pristupe rje{avanju ovog prob-lema [4, 5]. Gre{ke u vrijednosti tvrdo}e po Rock-wellu C izazvane gre{kama ugla i radijusa utiskiva~asu prikazane na slikama 1.a i 1.b1, [6, 7, 8, 9].Zona izme|u dvije kontinuirane linije sadrì gre{ke,kao {to je zaklju~io Petik [1], usljed razlike radijusaod +0,01 mm i razlike ugla od +30'. Unutar dvijeobuhva}ene zone nalaze se linije koje odgovarajuteoretskoj interpretaciji prema Bochmannu i Hildu(prema "Brinell metodi") [6, 10] i regresiji koju supredloìli Yamashiro i Uemura [8] i Stepanow [7]. Slike 1.a i 1.b se slaù sa prakti~nim pravilom kojesu predloìli Yamamoto i Yano [11], po kojem bimjerna nesigurnost ma{ina za standardizaciju tvrdo}eusljed gre{aka oblika utiskiva~a trebala biti redaveli~ine ±1 HRC. Mora se napomenuti da drugi eksperimentalnipodaci, po Woodu i dr. [12], koji su analiziralipona{anje 20 utiskiva~a od 5 razli~itih proizvo|a~a,rezultiraju gre{kama tvrdo}e iznad zone date naslikama 1.a i 1.b.

Problem ima dvije glavne posljedice: s jedne straneeksperiment po rasporedu tipa zvijezda koje sekoristilo u pro{losti ne moè dati informacije o uti-cajima drugog reda, dok s druge strane faktori kojinisu geometrijski mogu dovesti do rasipanja rezulta-ta eksperimenta koja sakrivaju uticaje geometrije. Izovih razloga poku{ali smo koristiti eksperiment zas-novan na shemi faktorijala i zvijezde za pore|enjesa rezultatima dobijenih simulacijom pomo}u MKE.Da bi se postigle male vrijednosti razli~itih uticaja,

Bochmann & Hild

Yamashiro & Uemura

Stepanow

0,0

0,2

0,4

0,6

10 20 30 40 50 60 70

Hardness [HRC]

∆∆ H

ardn

ess [

HR

C]

Bochmann & Hild

Yamashiro & Uemura

Stepanow

0,0

0,2

0,4

0,6

0,8

1,0

10 20 30 40 50 60 70

Hardness [HRC]

∆ H

ardn

ess

[HR

C]

Fig.1.a.: Hardness errors due to a radius deviationof +0,01 mm in function of the hardness level.Slika 1.a.: Gre{ke tvrdo}e usljed devijacije radijusaod +0,01 mm u funkciji nivoa tvrdo}e

Fig.1.b.: Hardness errors due to an angle deviationof +30' in function of the hardness level.Slika 1.b.: Gre{ke tvrdo}e usljed devijacije ugla od+30' u funkciji nivoa tvrdo}e

kao {to su hrapavost povr{ine i mehani~ke osobineutiskiva~a, provedena je analiza na polju uzorakaznatno ve}em od standardnih tolerancija, od oko118° do 122°, za ugao konusa, i od 180 µm do220 µm, za radijus kugle.

2. EKSPERIMENTALNA ANALIZA

Eksperimentalna analiza je provedena kori{tenjemtri bloka za kalibraciju sa sljede}im nivoimatvrdo}e: 25 HRC, 44 HRC i 55 HRC. Kao {to jepomenuto, kori{teno je devet razli~itih dijamantnihutiskiva~a, raspore|enih po shemi faktorijal i zvi-jezda, da bi se utvrdili uticaji radijusa i uglakonusa drugog reda. Dimenzije utiskiva~a su dateu tabeli 1.

Kuglice sfernog utiskiva~a su karakterizirane pomo}umjernog sistema IMGC [13]. Izmjerene su ~etiri sek-cije svake kuglice. Prosje~ne vrijednosti ~etiri radi-jusa vezanih za pomenute sekcije su navedene utabeli 1. Relevantna pro{irena gre{ka (2σ) iznosi oko0,4 µm. Mjerenje ugla konusa je provedeno pomo}uIMGC ure|aja baziranog na opti~koj interferenciji,koji omogu}uje postizanje rezultata na koji uti~epro{irena mjerna nesigurnost (2σ) reda veli~ine0,03°. Za svaki kalibracioni blok, izvr{eno je petmjerenja tvrdo}e pomo}u IMGC ma{ine za ispitivanjesa pasivnim optere}enjem [14]. U tabeli 1 su dateprosje~ne vrijednosti za svaki blok.Utvr|eno je da je pro{irena mjerna nesigurnost (2σ)vezana za svaku od ovih prosje~nih vrijednosti redaveli~ine od 0,15 HRC.

3. SIMULACIJA MKE

Prednost ovog pristupa je u tme da on omogu}uje,u principu, izoliranje razli~itih uticaja na gre{ke umjerenju tvrdo}e. Za simulaciju je usvojen ABAQUSprogram za prora~un metodom kona~nih elemenata.Problem je modeliran tako {to su uvedeni sljede}iulazni parametri: • savr{eno kruti utiskiva~, sa zahtijevanim radijusom

kuglice i uglom konusa;• cilindri~na epruveta sa dovoljno velikim radijusom;• epruveta izra|ena od homogenog materijala;• trenje izme|u epruvete i utiskiva~a η = 0;• karakteristi~ni parametri za elasto-plasti~ne

deformacije ispitivanog materijala.

quantities, like surface roughness and mechanical prop-erties of the indenter, the analysis has been performedon a sample field much larger than the standard tol-erances, ranging from about 118° to 122°, for the coneangle, and from 180 µm to 220 µm, for the tip radius.

2. EXPERIMENTAL ANALYSYS

The experimental analysis has been performed usingthree calibration blocks having respectively the fol-lowing hardness levels: 25 HRC, 44 HRC and 55HRC. As previously reported, nine different diamondindenters, arranged following a factorial and a starscheme, have been utilized in order to find out firstand second-order effects of the radius and of thecone angle. Indenter dimensions are given in Table 1.

The spherical tips of indenters have been charac-terized using the IMGC measurement system [13].Four sections of each tip have been measured.Average values of the four radii related to the abovementioned sections are reported in Table 1. The rel-evant extended uncertainty (2σ) is of about 0,4 µm.The measurement of cone angle has been per-formed using IMGC device based on optical inter-ference, which allows obtaining result affected by anextended uncertainty (2σ) of the order of 0,03°.For each calibration block, five hardness measure-ments have been performed using the IMGC dead-weight testing machine [14]. In Table 1 average val-ues for each block are reported.The extended uncertainty (2σ) related to each ofthose average values has been estimated to be ofthe order of 0,15 HRC.

3. FEM SIMULATION

The advantage of this approach is that it allows, inprinciple, to isolate different contributes of hardnessmeasurement errors. For that simulation the finiteelement code ABAQUS has been adopted. Theproblem has been modeled introducing the follow-ing input parameters:• perfectly rigid indenter, shaped with required tip

radius and cone angle;• cylindrical specimen with sufficiently large radius;• specimen constructed with homogeneous material;• friction coefficient between specimen and

indenter η = 0;• characteristic parameters for the elasto-plastic

deformations of the tested material.


- 118855 -

Due to the geometry of the problem it is possibleto analyze it with an axialsymmetric model, whichallows reducing remarkably the complexity of thesimulation. Therefore the analysis has been per-formed on a radial section of the specimen and ofthe indenter, shown in its main dimensions in Fig.2 and, for the indentation zone, in Fig. 3 (radialsection of the indentation after the removal of themajor load).

In Table 2 the differences between simulation andexperimental results are reported. It must be speci-fied that, as found by a careful analysis, the effectintroduced with the discretization of the indentingdomain produces an extended uncertainty (2σ) ofthe order of 0,3 HRC.


- 118866 -

Geometrija problema omogu}uje analizu pomo}uaksisimetri~nog modela, koji znatno smanjujesloènost simulacije. Stoga je analiza izvr{ena naradijalnom dijelu epruvete i utiskiva~a, koji su sasvojim glavnim dimenzijama prikazani na slici 2 i, zazonu utiskivanja, na slici 3 (radijalna sekcija utiski-vanja nakon uklanjanja glavnog optere}enja).

U tabeli 2 su date razlike izme|u simulacije ieksperimentalnih rezultata. Mora se navesti da, kao{to se utvrdilo pa`ljivom analizom, diskretizacijadomene utiskivanja ima za posljedicu pro{irenu mjer-nu nesigurnost (2σ) reda veli~ine 0,3 HRC.

IndenterHardness level:

25 HRCHardness level:

44 HRCHardness level:

55 HRC

Nr.Tip radius [µm]

Cone angle [°]

H

[HRC]∆H

[HRC]H

[HRC]∆H

[HRC]H

[HRC]∆H

[HRC]1 175,5 118,62 22,68 -2,36 41,92 -2,28 52,64 -2,102 192,3 118,64 22,91 -2,13 42,27 -1,93 52,83 -1,913 227,7 118,33 23,54 -1,50 43,57 -0,63 54,69 -0,054 180,8 120,38 25,33 0,29 44,08 -0,12 54,37 -0,375 205,8 119,96 25,19 0,15 44,22 0,02 54,84 0,106 202,4 120,41 25,98 0,94 44,92 0,72 55,49 0,757 183,5 122,45 28,15 3,11 46,13 1,93 55,92 1,188 185,9 122,37 28,16 3,12 46,32 2,12 56,04 1,309 220,2 122,33 28,70 3,66 47,27 3,07 57,46 2,72

Table 1.: Hardness values (H) measured for each hardness level with the nine indenters and differences(∆H) from central values (estimated with the averages of the results coming from all corrections).Tabela 1.: Vrijednosti tvrdo}e (H) izmjerene za svaki nivo tvrdo}e sa devet utiskiva~a i razlike (∆H) od sred-njih vrijednosti (koje su procijenjene preko prosjeka rezultata iz svih korekcija)

Figure 2.: Scheme of the model utilized for thesimulation.

Slika 2.: Shema modela pripremljenog za simulaciju.

Figure. 3.: Shape of the indentation impression afterthe removal of total load

Slika 3.: Oblik modela utiskivanja nakon uklanjanjaukupnog optere}enja

4. RESULTS ANALYSIS

Some differences from experimental and simulationresults exist; they are reported in Table 2. Thosediscrepancies seem to become bigger with theincreasing of the hardness. The reason of that hasto be found in the limits of both the experimentalanalysis and the simulation. Experimental results aremainly affected by three sources of error: the inden-ter shape, which could not be so regular as in aideal case, the indenter roughness, which dependon machining differences, and the indenter hystere-sis, which is due to the production technology.On the other side, simulation data present errors com-ing from the domain discretization and from theassumptions done for the construction of the model.Particular influence is due to contact parameters at theinterface between indenter and specimen, for that rea-son friction coefficient η and surface discretization ofsample are supposed to be the main source of error.

Those differences, being justified by uncertainty eval-uation, shall be acceptable. For that reason we shallconsider the numerical model as validated for thisfirst step of the approach, within the said uncer-tainties. The purpose for the future work is toreduce those discrepancies increasing the numberof elements, at least in the region with the maximaldeformation, and to investigate on friction effects. Experimental data have been analyzed in order toobtain a correction of hardness measurements ver-sus tip radius and cone angle errors. Operating alinear correlation analysis it has not been possibleto find out significant second-order effects at hard-ness levels 25 HRC and 44 HRC. Second-ordereffects become relevant at 55 HRC level.


- 118877 -

4. ANALIZA REZULTATA

Postoje neke razlike izme|u eksperimentalnih rezul-tata i rezultata simulacije; one su navedene u tabeli2. Za ove razlike se moè re}i da rastu spove}anjem tvrdo}e. Razlog za to se moè na}i uograni~enjima i eksperimentalne analize i simulacije.Na eksperimentalne rezultate uglavnom uti~u tri vrstegre{aka: oblik utiskiva~a, koji moè odstupati od ide-alnog, hrapavost utiskiva~a, koja zavisi od razlike uma{inskoj obradi i histerezis utiskiva~a, koji se javl-ja usljed tehnologije proizvodnje.S druge strane, podaci dobijeni simulacijom sadrègre{ke usljed diskretizacije domene i pretpostavkiuvedenih kod konstrukcije modela. Poseban uticajimaju kontaktni parametri na dodiru epruvete i utiski-va~a, pa se smatra da su zbog toga koeficijent tren-ja η i diskretizacija povr{ine uzorka osnovni izvorigre{aka.

Ove razlike, koje su utvr|ene odre|ivanjem mjernenesigurnosti, }e biti prihvatljive. Iz tog razloga sma-tra}emo da se numeri~ki model moè koristiti uovom prvom koraku, sa navedenom mjernom nesi-gurno{}u. Cilj budu}eg rada je smanjenje ovih rasi-panja pove}anjem broja elemenata, barem u regionusa maksimalnom deformacijom, te ispitivanje uticajatrenja.Eksperimentalni rezultati su analizirani da bi se dobi-la korekcija mjerenja tvrdo}e u odnosu na gre{keradijusa kuglice i ugla konusa. Vr{e}i linearnu kore-acionu analizu nije bilo mogu}e prona}i va`ne uti-caje drugog reda na tvrdo}ama 25 HRC i 44 HRC.Uticaji drugog reda postaju relevantni na nivou od55 HRC.

Indenters Difference between simulation and experimental data [HRC]Tip radius[µm]

Cone angle[°]

Hardness level: 25 HRC



175,5 118,62 -0,18 -0,60 -0,21192,3 118,64 -0,16 -0,21 0,53227,7 118,33 -0,19 0,26 0,47180,8 120,38 -0,05 -0,42 -0,56205,8 119,96 0,01 0,05 0,02202,4 120,41 -0,07 -0,42 -0,62183,5 122,45 0,18 -0,37 -0,57185,9 122,37 0,15 -0,57 -0,63220,2 122,33 0,26 -0,22 -0,43

Table 2. Differences between simulation and experimental data.Tabela 2. Razlike izme|u simulacije i eksperimentalnih rezultata.

Relacije koje su se dobile za tri razli~ita nivoatvrdo}e su navedene u tabeli 3, gdje ∆H ozna~avagre{ku tvrdo}e izmjerenu na ta~kama Rockwell Cskale, ∆α gre{ku ugla u stepenima, i ∆r gre{ku radi-jusa u mikrometrima. Rezultati koji su dobijeni ovimrelacijama su na granicama regiona prikazanog naslikama 1.a i 1.b. Ista analiza je izvedena na rezultatima simulacije. Uovom slu~aju postoje relevantni uticaji drugog redana sva tri nivoa tvrdo}e, a dobijene relacije su nave-dene u tabeli 4.

Niske vrijednosti standardne devijacije pokazuju daje prevladan veliki broj slu~ajnih uticaja.Primijeti}emo, ipak, da su rezultati dobijeni relacija-ma u tabeli 4 samo djelimi~no unutar granica saslika 1.a i 1.b. To je verovatno zato {to gre{kadiskretizacije uzorka ima svoj najve}i uticaj kod naj-manje dubine utiskivanja. a to se de{ava za vrijemeapliciranja optere}enja na tvrde materijale.

5. PREDLO@ENE KOREKCIJE

Usljed ~injenice da korekcije koje dolaze iz linearnekorelacije eksperimentalnih rezultata ne daju fizi~kuinterpretaciju fenomena oji se de{avaju, korisno jeuvesti razli~it pristup interpretaciji devijacija tvrdo}ekoje se javljaju usljed gre{aka geometrije utiskiva~a.Iz tog razloga predloèna su ~etiri razli~ita modela,zasnovana na interpretaciji procesa utiskivanja.

The relations obtained for the three different levelsof hardness are reported in Table 3, where ∆Hdenotes hardness error measured on Rockwell Cscale points, ∆α angle error in degrees, and ∆rradius error in micrometers. The results obtainedwith those relations are on the boundaries of theregions reported in Fig. 1.a and Fig. 1.b. The same analysis has been performed on simula-tion results. In this case there are relevant second-order effects at all the three hardness levels, therelations obtained are reported in Table 4.

The low values of standard deviation indicate that agreat part of random effects have been overcome.We shall notice, however, that the results obtained withrelations in Table 4 are only partially inside of the lim-its of Fig. 1.a and Fig. 1.b. This is probably due tothe fact that sample discretization error has its maxi-mal effect with smaller indentation impression, and thishappens during load application in hard materials.

5. PROPOSED CORRECTIONS

Due to the fact that corrections coming from linercorrelation of experimental data do not give a phys-ical interpretation of the involved phenomena, it isuseful to introduce a different approach to interprethardness deviations induced by indenter geometryerrors. For that reason four different models, basedon indenting process interpretation, are proposed.


- 118888 -

Table 3. Relations between tip radius, cone angle and hardness errors (experiment)Tabela 3. Relacije izme|u radijusa kuglice, ugla konusa i gre{aka tvrdo}e (eksperiment).

Hardness level[HRC] Formula

Standarddeviation

[HRC]

Correlationcoefficient

25 r023,038,1H ∆⋅+α∆⋅=∆ 0,13 0,997

43 r035,005,1H ∆⋅+α∆⋅=∆ 0,16 0,993

55 22 r0006,012,0r045,096,0H ∆⋅+α∆⋅−∆⋅+α∆⋅=∆ 0,19 0,990

Table 4. Relations between tip radius, cone angle and hardness errors (simulation)Tabela 4. Relacije izme|u radijusa kuglice, ugla konusa i gre{aka tvrdo}e (simulacija).

Hardness level[HRC]

FormulaStandarddeviation

[HRC]

Correlationcoefficient

25 2032,0r024,051,1H α∆⋅−∆⋅+α∆⋅=∆ 0.09 0,999

4328,0r0044,0r00035,0

058,0r047,007,1H2

2

−∆⋅α∆⋅−∆⋅+

+α∆⋅−∆⋅+α∆⋅=∆0,03 0,999

55 31,0r0019,0021,0r052,0594,0H

2 −∆⋅α∆⋅−α∆⋅−∆⋅+

+α∆⋅=∆0,03 0,999

The first correction take rise from Brinell definition ofhardness, according to which, the ratio betweenapplied load and indentation surface should be con-stant for a given material. The second one is basedMeyer analysis, and consists on imposing the invari-ance of the ratio between applied load and the pro-jection of the indented surface. The last two correc-tions are based on the evaluation of the deformationwork through the analysis of the indentation volume,in that way they are directly correlated to mechanicalprocess during the indentation. It must be remarkedthat a significant source of error for these correctionsis the approximation of the indentation shape with theindenter shape, thus neglecting the effects due to theelastic recovery of the tested material. For those corrections the four formulas in Table 5have been utilized.

Referring to Table 5, F is the imposed load, S isthe contact surface, A is the projection of contactsurface, V is the volume of the indenting impres-sion, h is the depth of the impression and kB, kM,kv1, and kv2 are constants depending on the spec-imen material. The values of the constants are cal-culated from the measured hardness values, takinginto account the measured geometry of the inden-ter. Thereafter it is possible to calculate the inden-tation depth under preload and total load expectedfor an indenter of ideal shape and, after that, toobtain the corrected values of hardness.Comparisons between those four corrections for thethree level of hardness, directly calculated on ourexperimental data, are reported in Table 6.a andTable 6.b, where ∆H is the difference between themeasurements for each indenter and the central val-ues for each hardness level, which is representedby the average of the results coming from all cor-rections. In general, mainly with the "Brinell" and"Meyer" criteria, one obtains a reduction of theexperimental errors to less than 30%. It must be noted that the angle deviation of theused indenters is larger than the radius deviation inrespect to the standard specifications. In those con-ditions residuals are bigger when corrections areapplied on angle in comparison to radius errors.


- 118899 -

Prva korekcija je izvedena iz Brinellove definicijetvrdo}e, prema kojoj odnos izme|u narinutogoptere}enja i povr{ine utiskivanja mora biti kon-stantan za dati materijal. Druga se zasniva na Mey-ervoj analizi, i sastoji se od utvr|ivanja invarijanteodnosa izme|u narinutog optere}enja i projekcijepovr{ine utiskivanja. Posljednje dvije korekcije sezasnivaju na odre|ivanju deformacionog rada krozanalizu zapremine utiskivanja, tako da su oni udirektnoj korelaciji sa mehani~kim procesima kojise de{avaju za vrijeme utiskivanja. Mora se ista}ida je znatan izvor gre{aka za ove korekcijeaproksimacija oblika utiskivanja sa oblikom utiski-va~a, ~ime se zanemaruju uticaji elasti~nogpona{anja ispitivanog materijala.Za ove korekcije izvedene su ~etiri formule u tabeli 5.

Prema tabeli 5, F je narinuto optere}enje, S je kon-taktna povr{ina, A je projekcija kontaktne povr{ine,V je zapremina utiskivanja, h je dubina utiskivanja,a kB, kM, kv1, i kv2 su konstante koje zavise odmaterijala epruvete. Veli~ine ovih konstanti sera~unaju iz izmjerenih vrijednosti tvrdo}e, uzimaju}iu obzir izmjerenu geometriju utiskiva~a. Nakon togaje mogu}e izra~unati dubinu utiskivanja o~etnomsilom i ukupnim optere}enjem koje se o~ekuje zautiskiva~ idealnog oblika i, nakon toga, dobitikorigovane vrijednosti tvrdo}e. Pore|enja te ~etiri korekcije za tri nivoa tvrdo}e,direktno izra~unata nad na{im eksperimentalnimrezultatima, data su u tabelama 6.a i 6.b, gdje je∆H razlika izme|u izmjerenih vrijednosti za svakiutiskiva~ i srednje vrijednosti za svaki nivo tvrdo}e,koja se predstavlja putem prosjeka rezultata dobi-jenih iz svih korekcija. Uop{te, koriste}i uglavnom"Brinell" i "Meyer" kriterije, postiè se smanjenjeeksperimentalnih gre{aka do ispod 30%. Mora se napomenuti da je devijacija ugla kori{tenihutiskiva~a ve}a od devijacije radijusa prema stan-dardnim specifikacijama. Pod tim uslovima rezidualisu ve}i kad se korekcije primijene na gre{ke uglanego kad se primijene na gre{ke radijusa.

Table 5. Correction formulas for hardness errors inferred by angle and radius errorsTabela 5. Korekcione formule za gre{ke tvrdo}e izvedene iz gre{aka ugla i radijusa.

“Brinell” correction “Meyer” correction Volume correction (V1) Volume correction (V2)

FS

kB=FA

kM= ( )F

VkV

3 2 1=FVh

kV

= 2

6. CONCLUSION

The analysis performed using a factorial and a starexperimental plan shows that second order and inter-action effects are significant at least for high hard-ness level. Nevertheless, we shall observe that, evenapplying regressions completed with the secondorder terms, the correction is not complete. There-fore experimental results can be considered asaffected by other effects connected with micro-geom-etry (roughness) or mechanical deformations of theindenter under load. For evaluating these effects, anattempt was made to simulate Rockwell hardnessmeasurement with a numerical model based onFinite elements method. The developed simulationpermits to obtain a good agreement between exper-imental and theoretical data in the range of 0,3 HRC,


- 119900 -

6. ZAKLJUÂK

Analiza koja je provedena kori{tenjem eksperimen-ta po rasporedu tipa faktorijal i tipa zvijezdapokazuje da su uticaji drugog reda i uticaji inter-akcije znatni, barem za visoke nivoe tvrdo}e.Uprkos tome, primijeti}emo da, ~ak i kad sekoriste regresije izvr{ene pod uslovima drugogreda, korekcija nije kompletna. Zbog toga se zarezultate eksperimenta moè re}i da na njih uti~udrugi uticaji vezani za mikro-geometriju (hrapavost)ili mehani~ke deformacije utiskiva~a podoptere}enjem. Za procjenu tih uticaja, poku{alo sesimulirati izmjerenu Rockwelovu tvrdo}u pomo}unumeri~kog metoda baziranog na metodi kona~nihelemenata. Izvedena simulacija omogu}uje dobroslaganje izme|u eksperimentalnih i teoretskih rezul-

Table 6.a Comparison between corrections at different hardness levelsTabela 6.a Pore|enje korekcija pri razli~itim nivoima tvrdo}e.

Table 6.b Comparison between corrections at different hardness levelsTabela 6.b Pore|enje korekcija pri razli~itim nivoima tvrdo}e.

∆Η Difference from central value [HRC]Indenter “Brinell” correction “Meyer” correction

Nr. Hardness level Hardness level25 44 55 Average 25 44 55 Average

1 -0,37 -0,23 -0,08 -0,23 -0,09 0,13 0,26 0,102 -0,32 -0,49 -0,69 -0,50 0,00 -0,23 -0,47 -0,233 -0,13 -0,53 -0,55 -0,41 0,10 -0,45 -0,54 -0,304 -0,02 0,23 0,27 0,16 -0,12 0,27 0,38 0,185 0,11 -0,18 -0,16 -0,08 0,10 -0,21 -0,19 -0,106 0,40 0,29 0,39 0,36 0,30 0,21 0,32 0,287 0,15 0,40 0,35 0,30 -0,43 0,10 0,15 -0,068 0,25 0,60 0,44 0,43 -0,28 0,32 0,23 0,099 0,48 0,42 0,39 0,43 -0,06 -0,04 0,00 -0,03

Average 0,06 0,06 0,04 -0,05 0,01 0,02

∆Η Difference from central value [HRC]Indenter ( )V3 2

correction Vh correction

Nr. Hardness level Hardness level25 44 55 Average 25 44 55 Average

1 -1,38 -1,18 -0,84 -1,14 -0,42 -0,15 0,37 -0,072 -0,80 -0,90 -1,04 -0,91 -0,11 -0,31 -0,46 -0,293 0,45 -0,05 -0,47 -0,02 0,57 -0,50 -1,52 -0,494 -0,53 -0,32 -0,13 -0,32 -0,39 0,05 0,51 0,065 0,28 -0,01 -0,06 0,07 0,19 -0,17 -0,30 -0,096 0,55 0,42 0,49 0,49 0,34 0,23 0,31 0,297 0,13 0,23 0,31 0,22 -0,68 -0,11 0,36 -0,148 0,30 0,51 0,45 0,42 -0,51 0,14 0,45 0,039 1,51 1,30 1,09 1,30 0,22 0,09 -0,20 0,04

Average 0,06 0,00 -0,02 -0,09 -0,08 -0,05

and, being the aim of the work only the analysis ofvariations, this was certainly encouraging. In fact weshall remark that the level of significance of theregression coefficients, together with correlation coef-ficients, are usually higher for the simulation resultsas compared with the relevant experimental results,and the relevant standard deviations are much small-er. This seems to indicate a smaller noise effect, butwe shall note that the freedom of FEM results fromdisturbing effects, like roughness and hysteresis ofindenters, is masked by numerical and discretizationeffects. Therefore it has not been possible to isolatethe errors of indenters not produced by the geo-metrical effects. Much work shall be devoted to animprovement of the simulation, and this could berealized increasing the number of elements in thezones of maximal deformation and evaluating contri-butions coming from the friction effect.

A second way attempted for isolating non-geometri-cal effects was to use a fit based on physical inter-pretations of the involved phenomena. A check wasdone using the Brinell definition of hardness, assuggested by Bochmann & Hild [6], the Meyer def-inition and an evaluation based on the indentationvolume (V1), as previously attempted [9] and finallya new model based again on the indentation vol-ume (V2). The results show that performances of allthis corrections are independent from the hardnesslevel, but "Brinell" correction and V1 correction showa systematic residual effect of the geometry, while"Meyer" correction and V2 correction (that is a mod-ified "Meyer" correction) seems to clean very wellthe results from geometrical effects, as the residu-als appears to be random.Future work is needed to evaluate possible effectsoutside the geometrical ones, therefore specificexperiments, to evidence effects of mechanical defor-mation, and a better FEM model, also with the intro-duction of the effect of friction, are in preparation.

7. REFERENCES - LITERATURA

[1] Petik F., Factors influencing hardness measure-ment, OIML P 11, Paris, 1983.

[2] Song J. F., Low S., Pitchure D., Germak A., Des-ogus S., Polzin T., Yang H. Q., Ishida H., BarbatoG., Establishing a world-wide unified Rockwell hard-ness scale with metrological traceability, Metrologia,Nr. 34, 1997, 331-342.[5] Polzin T., Schwenk D., Measurement of Rockwellindenters by laser interferometry, VDI Berichte, Nr.1194, 1995, 265-273.


- 119911 -

tata u rasponu od 0,3 HRC, i, ako je cilj radasamo analiza varijacija, ovo je sigurno ohrabruju}e.U stvari, nazna~i}emo da su nivoi signifikantnostikoeficijenata regresije, zajedno sa koeficijentimakorelacije, obi~no vi{i kod rezultata simulacije negokod relevantnih eksperimentalnih rezultata, i rele-vantne standardne devijacije su znatno manje. Topokazuje da je rasipanje rezultata manje, ali }emonapomenuti da je neosjetljivost MKE rezultata naporeme}ajne faktore, kao {to su hrapavost i his-terezis utiskiva~a, sakrivena uticajima diskretizacije inumeri~ke analize. Zbog toga nije bilo mogu}eizolirati gre{ke utiskiva~a koje nisu proizveligeometrijski uticaji. Dosta rada }e se posvetitiunapre|enju simulacije, i to bi se moglo ostvaritipove}anjem broja elemenata u zonama najve}ihdeformacija i uzimanjem u obzir uticaja trenja.

Drugi na~in kojim smo poku{ali izolovati nege-ometrijske uticaje bilo je kori{tenje sklopa zasno-vanog na fizi~kim interpretacijama fenomena koji sejavljaju. Provjera je izvr{ena pomo}u Brinellovedefinicije tvrdo}e, kao {to su predloìli Bochmann iHild [6], pomo}u Meyerove definicije i prora~unazasnovanog na zapremini utiskivanja (V1), kao uprethodnom poku{aju [9] i kona~no novi model zas-novan ponovo na zapremini utiskivanja (V2). Rezul-tati pokazuju da performanse svih ovih korekcija nezavise od nivoa tvrdo}e, nego "Brinell" korekcija iV1 korekcija pokazuju sistematski rezidualni efekatgeometrije, dok izgleda da "Meyer" korekcija i V2korekcija ({to je modificirana "Meyer" korekcija)veoma dobro ~iste rezultate od geometrijskih utica-ja, a reziduali se pojavljuju kao slu~ajni. Ubudu}e jepotrebno jo{ raditi na utvr|ivanju mogu}ih uticajakoji nisu geometrijske prirode, {to zna~i da su upripremi specifi~ni eksperimenti, za dokazivanje uti-caja mehani~ke deformacije, kao i bolji MKE model,tako|e s uvo|enjem uticaja trenja.

[3] Yamamoto H., Yamamoto T., Minagawa T., Effectsof the hardness test conditions (importance of rigid-ity of indenter), XIV IMEKO World Congress, Tam-pere, Finland, 1997, vol. III, 252-257

[4] Song J. F., Smith J. H., Vorburger T. V., RudderF. F., Stylus technique for direct verification of Rock-well diamond indenters, VDI Berichte, Nr. 1194,1995.


- 119922 -

[6] Bochmann G., Hild K., The influence of the devi-ation of indenter geometry on the hardness Rock-well C , Zeitschrift für Instrumentenkunde 68, S. 155,1960, (in German).

[7] Stepanow S. S., The influence of the geometri-cal parameters of conical indenters on the Rockwellhardness value, Trudi VNIIM 37(97), Moskow-Leningrad, 1959, 106-111, (in Russian).

[8] Yamashiro S., Uemura Y., Effects of test forcesand indenter geometry errors on the Rockwell mea-surements, VDI Berichte, Nr. 41, 1961, 109-121, (inGerman).

[9] Barbato G., Desogus S., The meaning of thegeometry of Rockwell indenters, Technical ReportIMGC n. R128, 6, 1978.

[10] Hild K., Influence of geometrical tolerances ofindenters on Rockwell C hardness measurements,Zeitschrift für Instrumentenkunde 66, 1958, S. 202,(in German).

[11] Yamamoto K., Yano H., Standardization of Rock-well C scale done by NRLM Tokyo, Bulletin NRLM,series Nr. 13, October 1966, 1-9.

[12] Wood J. G., Cotter J., Nash P. J., Hardness test-ing: a survey of Rockwell C scale diamond inden-ters, NPL Report MOM40, 1980.

[13] Barbato G., Desogus S., Measurement of theSpherical Tip of Rockwell Indenters, Journal of Test-ing And Evaluation, Vol 16, N.4, 1988, 369-374.

[14] Barbato G., Desogus S., Levi R., Design Stud-ies and Characteristics Description of the StandardDeadweight Hardness Tester of Istituto di Metrologia"G Colonnetti" (IMGC), VDI Berichte 212, 1978, 97-103.

1. UVOD

Alati za upravljanje kvalitetom, o kojima je rije~ u ovomradu, su veoma va`no sredstvo za postizanje, pra}enjei unapre|ivanje kvaliteta proizvoda i procesa u smisluotkrivanja gre{aka i otklanjanja njihovih uzroka.ISO 9004-4 navodi 11 alata i tehnika za pobolj{anjekvaliteta. U ovom slu~aju obradjuje se sedam alatza djelotvorno prepoznavanje i analizu problema kaotemelj korekcijskih mjera.

Ma{instvo 4(2), 193 - 204, (1998) E.Seferovi},...:ALATI ZA UPRAVLJANJE...

- 119933 -

AALLAATTII ZZAA UUPPRRAAVVLLJJAANNJJEE KKVVAALLIITTEETTOOMM UU PPRRIIMMJJEENNIIna primjeru vanjske mehani~ke obrade (struganje)

~eli~nog cilindri~nog dijela na kotu F 104,7±0,04 mm

Dr. Edhem Seferovi}, profesor, Ma{inski fakultet, Univerzitet u Sarajevu, Vilsonovo{etali{te br.9, 71000 SarajevoDipl.ing. Dervi{ êngi}, saradnik, Ma{inski fakultet, Univerzitet u Sarajevu, Vil-sonovo {etali{te br.9, 71000 Sarajevo, "Pretis-NNIS", d.o.o. Vogo{}a

REZIME

Alati za upravljanje kvalitetom posebno su va`no sredstvo za pra}enje i postizanje stabilnosti procesa teosiguranja projektovanog kvaliteta proizvoda. To se u pravilu postiè zapisom podataka o kvalitetu kao idokumentovanom sprije~avanju gre{aka i rje{avanjem problema.U radu se govori o sedam alata za upravljanje kvalitetom kroz primjer mehani~ke obrade struganjem gdjesu ovi alati uspje{no primjenjeni.

Klju~ne rije~i: histogram, poreto dijagram, korelacijski dijagram, kontrolna karta.

QQUUAALLIITTYY CCOONNTTRROOLL TTOOOOLLSS IINN AAPPPPLLIICCAATTIIOONNthrough example of external mechanical turning of

steel ccyylliinnddeerr to diameter 104,7±0,04 mm

Edhem Seferovi}, Ph.D., Faculty of Mechanical Engineering, University of Sarajevo,71000 Sarajevo, Vilsonovo {etali{te 9, Bosnia and Herzegovina,Dervi{ êngi}, BsC., Mech. Engineering, Faculty of Mechanical Engineering, 71000Sarajevo, Vilsonovo {etali{te 9, "Pretis-NNIS", d.o.o. Vogo{}a,

SUMMARY

Quality control tools have especially important role like devices for tracing and achieving stability ofprocess and assurance of designed quality of products. It is achieved with booking appropriate infor-mation about quality and also with documented errors prevention and solving the problem.Seven tools for quality control through successful example of application for mechanical turning isanalysed in this paper.

Key words: histogram, poreto diagram, and correlation diagram, control map

1. INTRODUCTION

Quality control tools, which are elaborated in thispaper, are essential to obtain, follow and improveproduct and process quality in terms of error check-ing and removing causes of errors.ISO 9004-4 gives 11 tools and techniques for qual-ity improvement. In this particular case, seven toolsfor efficient recognition and problem analysis as thebasis of correction measures are elaborated.

STRU^NI RAD

PROFESSIONAL PAPER

2. SEDAM ALATA ZA UPRAVLJANJEKVALITETOM

u primjeni na mehani~koj obradi stru-ganjem ~eli~nog cilindri~nog dijela namjeru F 104,7±0,04 mm

2.1. Obrazac za prikupljanje podataka - "Strichlista"

Obrazac za prikupljanje podataka omogu}ava si-s-temati~no prikupljanje podataka i olak{ava pozna-vanje zakonitosti, odnosno nagomilavanja, zbog dobi-janja jasne slike stvarnosti. On ~ini osnovu za daljugrafi~ku obradu i bitno olak{ava analizu.Ovom jednostavnom metodom utvrdjuju se kategori-je gre{aka nekog prozvoda i pojavljivanje tih gre{akazapisuje u obliku "Strichliste" (zapis pomo}u crtica).U na{em primjeru, struganje ~eli~nog cilindri~nogdijela, izdvojeno je 106 elemenata ~ije su mjere dateu spisku mjera istim redoslijedom kako su elemen-ti obra|ivani (na univerzalnom strugu "ADA" - Potis-je).

Grupisanjem dobijenih rezultata mjerenja moè seformirati slijede}i obrazac za prikupljanje podataka -"Strichlista" koji pruà mnogo vi{e informacija uodnosu na spisak mjera.Iz obrazca za prikupljanje podataka moè se uo~itiodre|ena stabilnost procesa (oblik Gausove krive),te da je ve}ina elemenata ura|ena na mjeru od104,68 - 104,72 mm {to je sredina tolerantnog polja.

Ma{instvo 4(2), 193 - 204, (1998) E.Seferovi},...: ALATI ZA UPRAVLJANJE...

- 119944 -

2. SEVEN QUALITY CONTROL TOOLS applied to mechanical turning of steelcylinder to diameter 104,7±0,04 mm

2.1. Data acquisition form - "Strichlist"

Data acquisition form enables systematic data acqui-sition and alleviates recognition of laws, i.e. accu-mulation, to obtain clear image about reality. It rep-resents basis for further graphical manipulation andsignificantly alleviates analysis.This simple method is used to establish error cate-gories of a product and to register occurrence ofthese errors in the form of "Strichlist" (bulleted list).In our example, machining of a steel cylinder, 106elements are separated, whose dimensions aregiven in a dimension list at the same order as theywere machined (at the universal turning machine"ADA"-Potisje).

Following data acquisition form - "Strichlist" can becreated by grouping obtained measuring results.Such a form gives much more information thandimension list.

Data acquisition form shows certain process stabili-ty (shape of Gaussian curve), and a fact that mostelements are machined to dimension between104,68 and 104,72 mm, that is middle of tolerancefield.

SEDAM ALATA ZA UPRAVLJANJEKVALITETOM

1. Obrazac za prikupljanje podataka 2. Dijagram toka 3. Dijagram uzroka i posljedica (Ishikawa dijagram) 4. Kontrolna karta 5. Histogram 6. Pareto dijagram 7. Korelacijski dijagram

SEVEN QUALITY CONTROL TOOLS

1. Data acquisition form

2. Flow chart

3. Causes-Consequencues chart(Ishikawa chart)

4. Control charts

5. Histogram

6. Pareto chart

7. Correlation chart

2.2. Flow chart

Flow chart is a tool whose purpose is to show com-plicated flows with different competencies in suchway that their structure and logic are clear andtransparent. Participants can easily recognise theirparticular tasks from the flow chart. Descriptive for-mulation of complete flow would be much harder tounderstand and to perceive.

Figure 1 shows flow chart for our example ofmechanical turning.

2.2. Dijagram toka

Dijagram toka je alat ~ija je svrha da komplikovanetokove sa razli~itim nadle`nostima i zadacima tokaprikaè, da njihova struktura i logika bude jasna itransparentna. Iz dijagrama toka u~esnici lahko pre-poznaju svoje razli~ite zadatke. Opisno formulisanjecjelokupnog toka bilo bi mnogo teè shvatiti i sagle-dati.

Na slici br. 1. prikazan je dijagram toka za na{ pri-mjer mehani~ke obrade elemenata.


- 119955 -

104,70 104,71 104,70 104,71 104,72 104,69104,72 104,74 104,73 104,72 104,73 104,74104,75 104,66 104,65 104,67 104,69 104,69104,68 104,69 104,70 104,71 104,70 104,71104,72 104,71 104,72 104,70 104,70 104,71104,70 104,71 104,72 104,73 104,72 104,73104,74 104,76 104,67 104,66 104,64 104,66104,67 104,68 104,68 104,69 104,69 104,68104,69 104,69 104,70 104,69 104,70 104,71104,69 104,70 104,70 104,71 104,71 104,72104,71 104,72 104,70 104,70 104,73 104,72104,71 104,71 104,72 104,73 104,73 104,74104,75 104,67 104,65 104,66 104,67 104,68104,68 104,67 104,68 104,69 104,69 104,70104,69 104,70 104,68 104,69 104,68 104,68104,69 104,71 104,70 104,72 104,70 104,71104,71 104,73 104,71 104,73

Mjera(φ 104,7±0,04)

Broj elementata Σ

104,64 I 1104,65 II 2104,66 IIIII 5104,67 IIIII III 8104,68 IIIII IIIII I 11104,69 IIIII IIIII IIIIII 16104,70 IIIII IIIII IIIII III 18104,71 IIIII IIIII IIIII II 17104,72 IIIII IIIII II 12104,73 IIIII IIII 9104,74 IIII 4104,75 II 2104,76 I 1

Ukupno 106

Spisak mjeraDimension list

Obrazac za prikupljanje podatakaData acquisition form

2.3. Dijagram uzroka i posljedica(Ishikowa dijagram)

Dijagram uzroka i posljedica (koji se tako|e nazivaIshikawa dijagram ili dijagram "riblja kost") je grani~kiprikaz, koji prikazuje uzroke u zbijenom oblikulogi~no i stepenasto. Ova metoda se pripisuje KaoruIshikawi, koji ju je po~etkom 50-tih godina uveo ujapansku industriju ~elika.Prema njegovom iskustvu odre|eni u~inak rijetkopo~iva na jednom jedinom uzroku, pogotovo ne naonome, koji se ~ini da leì na dlanu. Naprotiv,mogu}e uzroke naj~e{}e treba traìti u slijede}a~etiri polja:

• ~ovjek, • stroj,• metoda,• materijal.

Pri prakti~noj primjeni metode slobodno se defini{uvlastita polja, koja odgovaraju ispitivanom u~inku.tako se ~esto kao druga polja mogu}ih uzro~nikajavljaju Mitwelt (okolina) i Management (upravljanje).


- 119966 -

2.3. Causes-CConsequences chart(Ishikawa chart)

Causes-Consequences chart (also known asIshikawa chart or "Fish-bone" chart) is boundaryreview, which shows causes in compressed waylogically and stepped. Kaoro Ishikawa, who intro-duced it in 1950's into the Japanese steel produc-tion, develops this method.According to his experience, certain impact isseldom caused by one single cause, especiallyon the obvious one. Contrary, possible causesshould be searched for in the following fourfields:

• man• machine• method• material

In practical use of this method, it is permitted todefine own fields that are appropriate for selectedperformance. Mitwelt (environment) and Managementare often found as other fields of possible causes.

START

DOSTAVAPRIPREMAKA

Kontrolapripremaka

OBRADASTRUGANJEM

KontrolaDORADA

[KART

STOP

ZA[TITA IPAKOVANJE

STOP

DA

DA

NE

[KART

STOP

NE

Slika 1. Dijagram tokaFigure 1. Flow chart

Da bi se analizom odredili uzroci za odre|enuposljedicu, korisno je ispitati sve mogu}e uzroke zajedan identifikovani problem prema dijagramu uzro-ka i posljedica, ispitati njihovu prisutnost i u~inak,te pomo}u principa islju~ivosti (teìnskom analizom)izbaciti uzroke koji ne dolaze u obzir.Pri formiranju dijagrama uzroka i posljedica prvo seunese ispitivani u~inak (u na{em slu~aju "dorada i{kart") desno na kraju vodoravne crte. Zatim segranaju skupine glavnih uzroka (~ovjek, alat, ma{inai ostalo) u obliku riblje kosti. Pri tome se primje-njuju kreativne tehnike, kao npr. "branstorming", uzu~e{}e {to vi{e stru~njaka kojim je nastali problemblizak. Na taj na~in se defini{u mogu}i pojedina~niuzroci unutar glavnih uzroka (npr. motivacija,stru~nost, bolest i umor unutar uzroka ~ovjek) iupisuju se na horizontalnim strelicama (slika 2.).

Me|usobnim upore|ivanjem va`nosti uzroka, odno-sno, teìnskom analizom (slika 3.) dolazimo do ste-pena prioritetnosti uzroka kako pojedina~nih tako iglavnih.Na dijagramu se prioriteti u odnosu na uklanjanjegre{aka obiljeàvaju odgovaraju}im oznakama kao naslici 2.Analiziraju}i dijagram uzroka i posljedica za na{primjer moè se zaklju~iti:


- 119977 -

To obtain causes for some consequence by analy-sis, it is useful to examine all possible causes forsome identified problem according to the cause-con-sequence chart, to examine their presence andeffect, and to eliminate inappropriate cause by theexclusiveness principle (weight analysis).

During the creation of cause-consequence chart,examined effect should be entered first (in our case"upgrade and waste") to the right, at the end of hor-izontal line. Then, groups of main causes (man, tool,machine et al.) are branched out in the shape ofa fish-bone. Creative techniques should be usedduring that process, for example "brainstorming",with participation of as much experts for that par-ticular problem as possible. In such way, possiblesingle causes are defined and entered at the hori-zontal arrows (fig. 2.).By comparison of the importance of causes, i.e.weight analysis (Fig. 3.) priority level of causes isobtained, single ones, as well as main ones.

Priorities, compared with removal of errors, aremarked with appropriate signs at the chart, asshown in Fig. 2.By analysing cause-consequence chart for our exam-ple, following can be concluded:

Slika 2. Dijagram uzroka i posljedica (Ishikowa dijagram)Fig. 2. Causes-Consequences chart (Ishikawa chart)

To eliminate waste and upgrade in analysedprocess, it is necessary to give priority to the solv-ing of the following causes:- cutting tool,- worker motivation,- worker skillness.Cause-consequence chart can be universally used,especially for:- Improvement of process because of optimisationproductivity, expenses etc.- Analysis of errors, complaints and other imperfections.

Graphical representation documents complexity ofproblem being researched. Primary influence causesbecome clear. According to that chart, different pos-sible causes can be examined, confirmed or reject-ed, with main goal: to concentrate measures to thecentre of cause.

Da bi se eliminisali {kart i dorada u analiziranomprocesu potrebno je prioritetno rje{avati slijede}euzroke:

- rezni alat,- motivacija radnika,- stru~nost radnika.

Dijagram uzroka i posljedica je univerzalno prim-jenljiv naro~ito za:• Pobolj{anje procesa sa svrhom optimiziranja, pro-duktivnosti, tro{kova itd.• Analizu gre{aka, reklamacija i drugih nedostataka.

Grafi~kim prikazivanjem dokumentuje se komplek-snost istraìva~kog problema. Pos-taju jasni uticajniprimarni uzroci. Na osnovu tog dijagrama mogu seispitati, potvrditi ili odbaciti razli~iti mogu}i uzroci, sasvrhom koncentrisanja mjera na teì{te uzroka.


- 119988 -

ôvjek Motivacija Stru~nost Bolest Umor ΣMotivacija 2 2 2 6Stru~nost 0 2 2 4Bolest 0 0 1 1Umor 0 0 1 1

Alat Rezni alat Stezni alat Hla|enje Podmazivanje ΣRezni alat 2 2 2 6Stezni alat 0 1 1 2Hla|enje 0 1 1 2

Podmazivanje 0 1 1 2

Ma{inaBroj

obrtajaSuport

Elektro-motor

Centri~nost Σ

Br. obrtaja 1 1 1 3Suport 1 1 1 3El.motor 1 1 1 3

Centri~nost 1 1 1 3

Ostalo Materijal TehnologijaUslovirada

Temperatura Σ

Materijal 2 2 2 6Tehnologija 0 2 2 4Uslovi rada 0 0 1 1Temperatura 0 0 1 1

ôvjek Alat Ma{ina Ostalo Σôvjek 1 2 2 5Alat 1 2 2 5

Ma{ina 0 0 1 1Ostalo 0 0 1 1

Slika 3. Teìnska analiza uzorakaFigure 3. Weight analysis of causes

2.4. Kontrolne karte

Kontrolna karta, kao ~etvrti alat u ovom pregledualata za upravljanje kvalitetom, omogu}ava prikazi-vanje pada toka o kvalitetu u jednom dijagramu sadefinisanim grani~nim vrijednostima, sa svrhompravovremenog prepoznavanja i pode{avanja odstu-panja od planiranog toka procesa.Cilj svakog preduze}a je ekonomi~na proizvodnjakoja odgovara zahtjevima za kvalitet proizvoda. Iz tograzloga proizvodni procesi moraju biti tako koncipi-rani, pra}eni i upravljani da prakti~no sve proizve-dene jedinice ispunjavaju zahtjeve postavljene uspecifikacijama. Kontrolne karte su alat koji je veomaefikasan za pra}enje i upravljanje procesima.To je obrazac za grafi~ko prikazivanje vrijednosti, kojese dobivaju ispitivanjem neprekidnog niza uzoraka ikoji se, nakon njihovog upisivanja, upore|uju sa kon-trolnim granicama i ako je to potrebno, sa granica-ma upozorenja, sa svrhom upravljanja kvalitetom.Poznato je vi{e modela kontrolni karata koje seprimjenjuju u zavisnosti od karaktera procesa koji seprati.U na{em primjeru mehani~ke obrade, obzirom da seprati samo jedan parametar - vanjski pre~nik ele-menata, najprikladnija je X - karta ili karta mjera -kako se jo{ naziva.Smatraju}i izuzetu koli~inu, od 106 elemenata, je-d-nim uzorkom ura|ena je odgovaraju}a X - karta satolerancijskim granicama i granicama upozorenja,slika 4.


- 119999 -

2.4. Control charts

Control chart, as the fourth tool in this survey ofquality managing tools, enables demonstration ofquality flow decession in a chart with definedboundary values, to recognise and to adjust retreatof planned process flow at the right time.

Every company's aim is economical production,which is in accordance with product qualityrequests. Because of that, production processesshould be designed, followed and managed to prac-tically all produced units fit into requests set inspecification. Control charts are tools, which is veryeffective to follow and manage the process.It is a form for graphical representation of valuesobtained by examination of continuous set of spec-imens, which are, after entering the data, comparedwith control boundaries and, if necessary, with warn-ing boundaries, to perform quality control.There are several models of control charts, whichare being used depending on followed processcharacter.In our example of mechanical turning, since onlyone parameter is being followed, external diameterof the cylinder, the most appropriate chart is X-chart,sometimes called "dimension-chart".Excepted quantity of 106 elements is assumed asone specimen, appropriate X-chart with toleranceboundaries and warning boundaries is created, fig.4.

Slika 4. X-karta (karta mjera)Figure 4. X-chart (dimension chart)

Analiziraju}i ura|enu kontrolnu kartu moè se jasnouo~iti da je potrebno nakon svakih 20 - 25 ura|enihelemenata izvr{iti regulisanje procesa kako bi sesprije~ila dorada elemenata. Regulisanje procesa uovom slu~aju se vr{i korekcijom reznog alata iliparametara operacije struganja. Tako|e se moèuo~iti nepravovremeno reagovanje na pribliàvanjemjera elemenata granicama upozorenja {to je izaz-valo {kart - 3 elementa i doradu - 3 elementa.

2.5. Histogram

Histogram je klasi~ni grafi~ki prikaz relativnih u~estalostivrijednosti karakteristika (izmjerenih vrijednosti) nekogprocesa, sa svrhom razja{njenja {irine rasipanja iteì{ta raspodjele (poloàj, oblik). Na taj na~in se lahkomoè uo~iti zakonitost i trend procesa.

Izgled histograma za primjer koji je predmet ovograda prikazan je na slici 5.Ve} na prvi pogled moè se zaklju~iti da histogramima pravilan oblik - Gausova kriva {to ukazuje naodre|enu stabilnost procesa. Simetri~no je ra-spore|en oko sredine tolerantnog polja {to dokazu-je centriranost procesa. Ono {to je negativno uovom slu~aju je {irina rasipanja, {to uzrokuje pojavudorade i {karta jer odre|en broj elemenata izlazi izpropisanih tolerancijskih granica, F 104,7±0,04 mm.


- 220000 -

By analysing finished control chart, it is clear thatafter every 20 to 25 produced elements, processshould be regulated to prevent upgrade of elements.Process regulation in this case is being performedby correction of cutting tool or machining operationparameters. Also, it is noticeable that slow reactionto element's dimensions approaching to warningboundaries caused waste of 3 elements andupgrade of 3 elements.

2.5. Histogram

Histogram is classical graphical exposition of rela-tive frequencies of characteristic's values (measuredvalues) of some process, whose aim is to explainthe dissipation width and median (place, shape). Inthat way, it is easy to notice existence of principlesand process trend.The shape of histogram for the example, which issubject of this paper, is shown on the figure 5.At the first sight, one can conclude that histogramhas regular shape - Gaussian curve, and that showscertain stability of process. It is symmetrically dis-tributed around the middle of tolerance field, whichproves centreness of process. What is negative inthis case is width of dissipation, which causeswaste and necessity of upgrade, because some ele-ments are out of allowed tolerance boundaries,104,7±0,04 mm.

Slika 5. HistogramFigure 5. Histogram

2.6. Pareto dijagram (ABC analiza)

Pareto analiza je slikovito prikazivanje informacijazbog pronalaènja iz velikog broja uticajnih faktora,koji su sa odredjenog stanovi{ta, na primjer sastanovi{ta tro{kova, od najve}e va`nosti. Ova ana-liza pokazuje, da vrlo ~esto mali broj uzroka proizvo-di najve}i dio u~inka.S tim u vezi poznato je pravilo 80/20, koje kaè da20% uzroka ~ini oko 80% sveukupnog u~inka /posljedica. Ova se iskustva mogu upotrijebiti zaodre|ivanje prioriteta korekcionih i preventivnihmjera.Za izradu Pareto dijagrama prvo se prikupe svizapaèni fenomeni (npr. vrste gre{aka) istraìvanogproblema. Kao slijede}e mora se utvrditi, koji }e kri-terij (u~nika) odlu~ivati s obzirom na istraìvani pro-blem o reagiranju pojedinih kategorija.

Kao slijede}i koraci:• za svaku kategoriju se odre|uje udio u sveukup-nom u~inku i izra~unava postotni udio;• utvr|uje se redoslijed kategorija prema pripada-ju}im u~incima;• za svaku se kategoriju u skladu sa tim redoslije-dom sabiraju u~inci;• taj zbir u~inaka se grafi~ki prikazuje redoslijedomkategorija.

Za na{ primjer analizirano je koliki su u~inci poje-dinih uzroka u doradi elemenata izraèno u vre-menskim jedinicama. Rezultati te analize dati su utabeli na slici 6.

Na osnovu ura|ene analize, do koje se do{lopra}enjem procesa, ura|en je Pareto dijagram slika 7.Iz Pareto analize i dijagrama se jasno vidi da triuzroka (rezni alat, motivacija i stru~nost) "proizvode"83% dorade elemenata. Zbog toga su na dijagramu


- 220011 -

2.6. Pareto chart (ABC analysis)

Pareto analysis is illustrative way to display informa-tion, so it is possible to find between large num-ber of influence factors which one of them, forexample costs, is of most importance. This analysisshows that very often, small number of causes haveinfluence in most of results. In relation with it, the rule 80/20 is known. That rulesays that 20% of causes makes around 80% of totalresults / consequences. These experiences can beused to define priority of correction and preventivemeasures.To create Pareto chart, it is necessary to collect allnoticed phenomena (for example, kinds of errors) ofexamined problem. Next step is to decide which cri-terion (of result) will be chosen for decision con-cerning researched problem about reaction of sin-gle categories.As following steps:- part of total result is defined for each category,also percental;- order of categories is defined according to theirresults;- results for every category are summarised;- that sum of results is then graphically displayedin order of categories.

In our example, it is analysed how big are resultsof single causes in upgrade of element, expressedin time units. Results of that analysis are given intable at figure 6.

According to performed analysis, which has beenderived by following the process, Pareto chart iscreated, as in fig. 7.It is clear from Pareto analysis and chart that threecauses (cutting tool, motivation and skillness) "pro-

Red.br.

Uzrok dorade Dorada (min) % Kumulativ

1. Rezni alat 226 40,7 40,72. ôvjek-motivacija 138 24,8 65,53. ôvjek-stru~nost 97 17,5 834. Pripremak 35 6,4 89,45. Tehnologija 22 3,9 93,36. Ma{ina 18 3,3 96,67. Radni uslovi 7 1,3 97,98. Hla|enje 6 1,1 99,09. Podmazivanje 4 0,7 99,710. Stezni alat 2 0,3 100,0

Σ 555 100%

Slika 6. Pareto analizaFigure 6. Pareto analysis

i izdvojena 3 podru~ja A, B i C (zbog ~ega se ovaanaliza naziva i ABC analiza) ~iji su u~inci u dora-di slijede}i:

- Podru~je A (1, 2, 3) = 83,0%- Podru~je B (4, 5, 6) = 13,6%- Podru~je C (7, 8, 9 i 10) = 3,4%

Ova analiza ukazuje na neophodnost rje{avanjaproblema dorade otklanjanjem prvenstveno uzroka upodru~ju A ~ime bi se eliminisalo 83% dorade.Svaki drugi pristup zna~io bi utro{ak vremena ienergije uz veoma male efekte (jer recimo podru~jeC sa 4 uzro~nika uti~e na "samo" 3,4% dorade) {tobi svakako demotivisalo stru~njake koji rade nasmanjenju dorade.


- 220022 -

duce" 83% of element upgrade. Because of that,there are 3 areas at the chart: A, B and C (becauseof that this analysis is also called ABC analysis)whose results in upgrade are as follows:- Area A (1, 2, 3) = 83,0 %- Area B (4, 5, 6) = 13,6 %- Area C (7, 8, 9 and 10) = 3,4 %

This analysis shows that it is necessary to solve theproblem of upgrade by removing firstly causes inthe area A, which eliminates 83 % of upgrade. Anyother approach would be waste of time and ener-gy with very small effects (because area C with 4causes has influence to "only" 3,4 % of upgrade)what would surely demotivate experts who work onupgrade reduction.

Slika 7. Pareto dijagramFigure 7. Pareto chart

2.7. Korelacijski dijagram

Korelacijski dijagram ili dijagram rasipanja je grafi~kiprikaz odnosa pripadaju}ih varijabli, zbog prepozna-vanja veze izmedju njihovih vrijednosti. Iz njega sevidi kako se sa promjenom nezavisne varijable Xmijenja zavisna varijabla Y.Za primjer posmatrane mehani~ke obrade usposta-vljena je korelacija izmedju vanjskog pre~nika i maseobradjivanog elementa.U tabeli na slici 8. dati su odnosi pre~nika eleme-nata i njihove mase.

Na osnovu te tabele ura|en je korelacijski dijagramprikazan na slici 9.

Iz dijagrama se vidi da je izmedju navedenih para-metara (pre~nik i masa) uspostavljena jaka pozitivnakorelacija.Korelacijski dijagram ima zna~ajnu prakti~nu prim-jenu u procesima u kojima je oteàno pratiti nekiparametar direktno, ve} se mora to u~initi indirekt-no preko drugog lak{e mjerljivog parametra, pri~emu se mora uspostaviti odgovaraju}a korelacijaizmedju istih.

3. ZAKLJUÂK

Primjenom sedam alata za upravljanje kvalitetom naprimjenu mehani~ke obrade struganjem sa ciljemsmanjenja dorade i {karta uz pronalaènje uzroka iotklanjanja istih moè se zaklju~iti:

- Proces obrade elemenata struganjem nije dovoljnostabilan, te ga treba nakon ura|enih 20 - 25 ele-


- 220033 -

2.7. Correlation chart

Correlation chart or dissipation chart is graphicaldisplay of relations between variables, because ofrecognition of connection between their values. Itshows that change in independent variable X resultswith change in dependent variable Y.In examined example of mechanical turning, corre-lation between external diameter and mass ofmachined element is established.Table shown in figure 8 gives relations betweendiameter and mass of element.

According to that table, correlation chart shown infigure 9 is created.

From that chart it is obvious that there is strongpositive correlation between these parameters (diam-eter and mass).Correlation chart has significant practical use inprocesses where it is hard to follow some para-meter directly, and it must be done indirectlythrough some other parameter that is better formeasuring and there must be correlation betweenthem.

3. CONCLUSION

With application of seven quality management toolsto example of mechanical turning with aim to reduceupgrade and waste, with recognition and removal ofcauses, following can be concluded:

- Process of machining is not stable enough, andit has to be regulated after 20 - 25 produced ele-

Masa (kg)14,40 14,41 14,42 14,43 14,44 14,45 14,46 14,47 14,48 14,49 14,50 14,51 14,52

104,64 X104,65 X X104,66 X X104,67 X X104,68 X X104,69 X X104,70 X X104,71 X X104,72 X X104,73 X X104,74 X X104,75 X X104,76 X

Slika 8. Tabela korelacije pre~nika / masaFigure 8. Table of correlation between diameter and mass

menata regulisati. Regulisanje se odnosi prvenstvenona korekciju reznog alata ili parametara mehani~keobrade.- Glavni faktori koji uti~u na pojavu dorade i {kartasu: rezni alat i ~ovjek (motivacija i stru~nost). Elim-inisanjem ova dva uzroka dorade i {kart bi bilismanjeni za vi{e od 80%.

LITERATURA:

[1] OSKAR - Centar za razvoj i kvalitet d.o.o.,Zagreb: "Sistemi upravljanja kvalitetom u primjeni"

[2] JP UNIS - "PRETIS", Vogo{}a: "Kontrolni izvje{tajii zapisi (o pra}enju procesa)"


- 220044 -

ments. Regulation is related mainly to correction ofcutting tool or parameters of machining.- Main factors that have influence in need forupgrade and increased waste are: cutting tool andman (motivation and skillness). By eliminating ofthese two causes, upgrade and waste would bereduced for more than 80 %.

14.39

14.40

14.41

14.42

14.43

14.44

14.45

14.46

14.47

14.48

14.49

14.50

14.51

14.52

14.53

104.62 104.64 104.66 104.68 104.70 104.72 104.74 104.76 104.78

pre~nik (mm)

masa

(kg)

Slika 9. Korelacijski dijagramFigure 9. Correlation chart

Ma{instvo 4(2), 205 - 214, (1998) B.Tihi: EVOLUCIJA STRATE[KOG MENAD@MENTA...

- 220055 -

1. UVOD

Od samih po~etaka formiranja preduze}a, generalnimenadèr je uvijek imao najve}u odgovornost zadono{enje strate{kih odluka. Stoga je iznimno intere-santno iz istorijske perspektive posmatrati kako jeproces evolucije strate{kog menad`menta uticao naizmjene traènih li~nih osobina generalnogmenadèra u preduze}u.

2. EVOLUCIJA STRATE[KOGMENAD@MENTA I OSOBINE GENERALNOG MENAD@ERA SAVREMENOG PREDUZE]A

Poznati autori Igor Ansoff i Edward Mc Donell [1]smatraju da postoje ~etiri prepoznatljive faze uevoluciji strate{kog menad`menta:

EEVVOOLLUUCCIIJJAA SSTTRRAATTEE[[KKOOGG MMEENNAADD@@MMEENNTTAA II OOSSOOBBIINNEEGGEENNEERRAALLNNOOGG MMEENNAADD@@EERRAA SSAAVVRREEMMEENNOOGG PPRREEDDUUZZEE]]AA

Prof. dr. Boris Tihi, Ekonomski fakultet u Sarajevu, Univerzitet u Sarajevu,Bosna i Hercegovina,

REZIME

Zahtjevi koji se postavljaju pred generalnog menadèra u domenu strate{kog menad`menta su semijenjali kroz istoriju saglasno promjenama u preduze}u i njegovom okruènju.U radu se daje pregled promjena karakteristika generalnog menadèra u korelaciji sa promjenamastrate{kog menad`menta. Definira se skup osnovnih osobina menadèra budu}nosti.

Klju~ne rije~i: strate{ki menad`ment, evolucija, osobine generalnog menadèra, savremeno preduze}e

EEVVOOLLUUTTIIOONN OOFF SSTTRRAATTEEGGIICC MMAANNAAGGEEMMEENNTT AANNDD CCHHAARRAACCTTEERRIISSTTIICCSS OOFF GGEENNEERRAALL MMAANNAAGGEERR

OOFF MMOODDEERRNN CCOOMMPPAANNYY

Boris Tihi, PhD., professor, Faculty of economic, University of Sarajevo, Bosniaand Herzegovina

SUMMARY

Requests for general manager in the domain of strategic management have changed through histo-ry, in accordance with changes inside the company and its environment. This paper gives the sur-vey of changes in general manager characteristics in correlation with changes in strategic manage-ment. The collection of basic characteristics of future manager is defined.

Key words: Strategic Management, Evolution, General Manager Characteristics, Modern Company

1. INTRODUCTION

Since very first companies were established, gener-al manager always had highest responsibility inbringing strategic decisions. It is extremely interest-ing to watch from historical perspective howprocess of evolution of strategic management influ-enced changes in required personal characteristicsof general manager in the company.

2. EVOLUTION OF STRATEGIC MANAGEMENT AND CHARACTERISTICSOF GENERAL MANAGER OF MODERNCOMPANY

Well-known authors Igor Ansoff and Edward McDon-nell [1] anticipate that there are four recognisablephases in strategic management evolution:

PPRREEGGLLEEDDNNII RRAADD

RREEVVIIEEWW PPAAPPEERR

1. Upravljanje pomo}u kontrole,2. Upravljanje ekstrapoliranjem pro{losti,3. Upravljanje pomo}u anticipiranja promjena i4. Upravljanje pomo}u fleksibilnog i brzog reago-

vanja.

Navedene ~etiri faze citirani autori su povezali sa nji-hovom vremenskom dimenzijom, predvidivo{}ubudu}nosti, primjenjenim tehnikama i nivoima turbu-lencija u okruènju, pa sve to prikazali modelom uTabeli 1. [2]

Vremenski raspon od 1900. godine pa sve do krajadvadesetog vijeka i smje{tanje pojedinih faza u tajokvir, dobra su orijentacija op{te tendencije ali tone zna~i da se sve vrste biznisa identi~no pona{aju.âk i danas, u vrijeme koje karakteri{u nepredvidlji-va iznena|enja, postoje biznisi gdje je nivo turbu-lencija stabilan, reaktivan ili anticipiraju}i. Problem je,me|utim, u tome {to je neophodno za svaki biznisocijeniti nivo turbulencije u okruènju i tome pri-lagoditi strate{ki manad`ment. Ne moè se, na pri-mjer, danas u kompjuterskom biznisu upravljatiputem kontrole i imati generalnog menadèra saosobinama koje su bile potrebne 1900. godine!Postoje, dodu{e, mi{ljenja i veoma uglednih autora,kao {to je na primjer Mintzberg [3], da treba reduci-rati kompleksnost u strate{kom menad`mentu, sma-njiti horizont dugoro~nog planiranja i dati prednostintuiciji i iskustvu. Ovo se bazira na konceptugrani~ne racionalnosti, koji je razvio dobitnikNobelove nagrade Herbert Simon. Njegova istraìva-nja su pokazala da pojedinci, kao i organizacje, nisuu stanju da kontroli{u probleme kada ovi pre|uodre|eni nivo kompleksnosti. U takvim slu~ajevimamenadèri ne mogu vi{e razumjeti {ta se de{ava uokruènju, niti upravljati racionalno strategijom pre-duze}a.Ipak, pojednostavljenje strate{kog menad`mentamogu}e je samo u onoj mjeri u kojoj okruènje todozvoljava. Zbog toga je jedino rje{enje zamenadère koji nisu u stanju nivo kompleksnostiupravljanja strategijom preduze}a podi}i na onaj ste-pen koji zahtijeva okruènje datog biznisa, da napustetaj biznis i potraè "mirnije vode". Neprihvatanje ovelogike je siguran put u bankrotstvo!Upravo potreba da se nivo kompleksnosti strate{kogmenad`menta uvijek prilago|ava stepenu turbulen-tnosti okruènja, ~ini izuzetno interesantnom analizuli~nih osobina generalnog menadèra u razli~itimsituacijama.Osobe koje su vodile preduze}e u 19 vijeku nisusigurno imale probleme koje imaju njihovi nasljedni-ci danas. Zbog toga nisu morali imati ni sveosobine koje se danas traè od generalnog


- 220066 -

1. Management through control2. Management through extrapolation of the past3. Management through anticipating the changes

and4. Management through flexible and fast response.

Above mentioned authors connected these fourphases with their time dimension, predictability ofthe future, applied techniques and levels of turbu-lence in environment, and at last showed that in thefollowing model: [2] (Table 1.)

Time period from 1900 until the end of twentiethcentury and putting single phases into that frame,is good orientation of general tendencies, but itdoes not mean that all kinds of business act thesame way. Even today, in the time characterised byunpredictable surprises, there are businesses wherelevel of turbulence is stabile, reactive or anticipat-ing. Problem is, therefore, in necessity to estimateturbulence level in environment for every businessand to adjust strategic management to it. For exam-ple, computer business today can not be managedby control and with general manager with samecharacteristics as in 1900.There are some opinions, even by respectableauthors like Mintzberg, [3] that complexity in strate-gic management must be reduced, horizon of long-term planning should be lowered and intuition andexperience should be put at the top. This is basedon the boundary rationality concept, developed byNobel-prize winner Herbert Simon. His researchesshowed that individuals, as well as organisations,can not control problems when they exceed certainlevel of complexity. In such cases managers can nolonger understand what is going on in their envi-ronment, neither to rationally manage company'sstrategy.Nevertheless, simplification of strategic managementis possible only as much as environment allows.Because of that, the only solution for managers whocan not raise complexity level of managing compa-ny's strategy to level required by business environ-ment, is to abandon that business and look for"pacific places". Rejection of such logic is certainway to bankruptcy!The constant need to adjust complexity level ofstrategic management to the level of turbulence ofenvironment makes analysis of personal characteris-tics of general manager in different situationsextremely interesting.Persons who lead companies in 19th century sure-ly did not have problems that their successors facenowadays. Because of that, they did not have tohave all characteristics required now for general

Tabela 1. Evolucija strate{kog menad`menta [1]

PROMJENLJIVOST 1900 1930 1950 1970 1990

POZNATOST DOGA\AJA

NEPREDVIDLJIVOST BUDU]NOSTI POZNATI MOGU SE

EKSTRAPOLIRATI POZNATI U DISKONTINUITETU NOVI U DISKONTINUITETU

NAÎNI STRATE[KOG UPRAVLJANJA

♦ PRIRU^NICI ZA PROCEDURE I SISTEME PONAVLJA SE

♦ FINANSIJSKA KONTROLA

UPRAVLJANJE KONTROLOM

♦ TEKU]E FINANSIJSKO PLANIRANJE

♦ FINANSIJSKO PLANIRANJE INVESTICIJA

♦ UPRAVLJANJE POMO]U CILJEVA

PREDVIDIVA POMO]U

EKSTRAPOLACIJE

♦ DUGORO^NO PLANIRANJE

UPRAVLJANJE EKSTRAPOLIRANJEM

PRO[LOSTI

MOGU]E JE PROCIJENTI

OPASNOSTI I [ANSE

♦ PERIODI^NO STRATE[KO PLANIRANJE ♦ UPRAVLJANJE POZICIONIRANJEM

UPRAVLJANJE POMO]U

ANTICIPIRANJE PROMJENA

♦ PLANIRANJEM NEPREDVI\ENIH STANJA

♦ UPRAVLJANJE STRATE[KIM DOGA\AJIMA

♦ UPRAVLJANJE NA OSNOVU SLABIH SIGNALA

DJELIMI^NO PROCJENLJIVE

[ANSE

♦ UPRAVLJANJE IZNENA\ENJIMA

UPRAVLJANJE POMO]U

FLEKSIBILNOG I BRZOG

REAGOVANJA

NIVOI TURBULENCIJE 1

STABILAN

2 REAKTIVAN

3 ANTICIPIRAJU]I

4 IZVI\A^KI

5 KREATIVAN

Table 1. Evolution of strategic management [1]

VARIABILITY 1900 1930 1950 1970 1990

EVENT KNOWLEDGE

UNPREDICTABILITY OF THE FUTURE

KNOWN CAN BE

EXTRAPOLATED KNOWN IN DISCONTINUITY NEW IN DISCONTINUITY

KINDS OF STRATEGIC

MANAGEMENT

♦ GUIDES FOR PROCEDURES AND SYSTEMS REPEATING

♦ FINANCIAL CONTROL

CONTROL MANAGEMENT

♦ CURRENT FINANCIAL PLANNING

♦ FINANCIAL INVESTMENT PLANNING

♦ MANAGEMENT BY GOALS

PREDICTABLE THROUGH

EXTRAPOLATION

♦ LONG-TERM PLANNING

MANAGEMENT THROUGH

EXTRAPOLATION OF THE PAST

POSSIBLE TO ESTIMATE HAZARD AND CHANCES

♦ PERIODIC STRATEGIC PLANNING ♦ MANAGEMENT BY POSITIONING

MANAGEMENT THROUGH

ANTICIPATION OF CHANGES

♦ PLANNING OF UNPREDICTABLE STATES

♦ STRATEGIC EVENT MANAGEMENT

♦ MANAGEMENT BY WEAK SIGNALS

PARTIALLY ESTIMABLE CHANCES

♦ MANAGEMENT BY SURPRISES

MANAGEMENT BY FLEXIBLE AND RAPID

RESPONSE

TURBULENCE LEVEL 1

STABLE

2 REAC-TIVE

3 ANTICIPATING

4 EXPLORATIVE

5 CREATIVE

menadèra. U ono doba je regulativna uloga dràvebila minimalna, a biznis je nezavisan i za{ti~enautoritetom jakih li~nosti, tipa John D. Rockeffeller-aili Henri Ford-a. Zatim je do{ao period poduzetnikai inovatora. Proizvodi kreirani od osoba kao {to je,na primjer, bio Thomas Edison, brzo su se razvijalii doprinijeli do tada nezamislivom ekonomskom pros-peritetu.

Drugi svjetski rat je sve izmijenio. Superiornostameri~ke industrije bazirala se na timskom radu. Toje bila era efikasnosti i tehnokrata. Menadèrske le-gende, kakva je bio Alfred Sloan iz General Motors-a, fokusirale su pa`nju na interne probleme pre-duze}a. Potro{a~i su bili samo tolerisani, vlasniciakcija patronizirani, a {ire okruènje slabouvaàvano.ûveni Peter Drucker u uvodu knjige Alfred-a Sloan-a "My Years with General Motors" [4], sumira bitnestavove ovog legendarnog menadèra svoga vremena:

♦ Menad`ment je profesija i zato menadèr mora-biti profesionalac u punom smislu te rije~i,

♦ Kao i svaka profesija, ljekarska ili advokatska,menadèr ima "klijenta"-preduze}e. On morasvoje vlastite interese podrediti klijentovim,obaveze prema klijentu,

♦ Profesionalc ne donosi odluke na bazi mi{ljenjaili preferencija, nego na bazi ~injenica,

♦ Posao profesionalnog menadèra nije da se svi|aljudima, niti da ih mijenja, nego da njihove snagepreto~i u akciju. Njihova performansa je jedino{to se vrednuje i o ~emu menadèr mora vodi-ti ra~una,

♦ Neslaganja, ~ak konflikti su potrebni. Bez toganema razumijevanja, a bez razumijevanja sedonose pogre{ne odluke,

♦ Liderstvo nije "karizma", to nisu "odnosi sajavno{}u" ili predstave. To je u~inak, konzistent-no pona{anje, povjerenje,

♦ Kona~no, mo`da najva`nije, profesionalnimenadèr je ipak sluga. Poloàj ne donosi priv-ilegije i mo}, nego name}e obaveze.

Moète se ne sloìti sa nekim od ovih stavova ismatrati ih zastarjelim ali to samo potvr|uje na{uhipotezu o me|uzavisnosti okruènja, kompleksnostimenad`ment i li~nih osobina menadèra.Prije nego {to poku{amo bitne li~ne osobine gene-ralnog menadèra definisati i povezati sa okruènjemu kojem njegovo preduze}e djeluje, korisno jedetaljnije analizirati istorijski razvoj biznisa. Odabralismo SAD kao primjer, mada bi sli~na slika za nekodrugo podru~je sa duòm tradicijom trì{neekonomije, tako|e posluìla svrsi:(Tabela 2.) [5]


- 220099 -

manager. In that time regulative role of a state wasminimal, and business was independent and pro-tected by authorities of strong person's, like JohnD. Rockfeller or Henry Ford. After that, there wereperiod of entrepreneurs and innovators. Productscreated by persons like Thomas Edison were devel-oped fast and they contributed incredible economicprosperity.

World War II changed everything. Superiority ofAmerican industry was based upon teamwork. It wasage of efficiency and technocrats. Manager legends,like Alfred Sloan from General Motors, focused theirattention to internal problems of the company. Con-sumers were only tolerated, shareowners patronisedand wider environment weakly appreciated.Famous Peter Drucker, in the preface of AlfredSloan's book "My Years with General Motors", [4]summarise important attitudes of this legendary man-ager:

♦ Management is profession; therefore manager mustbe professional with the full meaning of that word.

♦ As any other profession, lawyer or doctor, man-ager has a "client" - the company. He has tosubordinate his own interests to client's interests,his duties to the client.

♦ Professional does not make decisions according to hisopinion or preferences, but according to the facts.

♦ The professional manager's job is not to be like-able to the people, nor to change them, but toconvert their forces into the action. Their perfor-mance is the only thing being valuated andabout what manager has to take care.

♦ Disagreements, even conflicts are necessary. Withoutthem there is no understanding, and without under-standing, wrong decisions are being made.

♦ Leadership is not "charisma"; it is not "public rela-tions" or shows. It is effect, consistent behaviour,confidence.

♦ Finally, perhaps most important, professional manag-er is nevertheless a servant. His position does notcarry privileges and power, but gives new duties.

You can disagree with some of these attitudes and thinkabout them as obsolete, but it just confirms our hypoth-esis about interaction of environment, management com-plexity and personal characteristics of the manager.Before we try to define and connect with the envi-ronment, in which his company acts, important per-sonal characteristics of general manager, it is useful toanalyses more in details historical development of busi-ness. We have chosen USA as an example, but sim-ilar picture about some other area with long traditionof market economy would also be good example: [5]

Tabela 2. Istorijski razvoj biznisa u SAD

PERIOD NAZIV OSNOVNE KARAKTERISTIKE

1820 – 1900 INDUSTRIJSKA REVOLUCIJA

- Velike tehnolo{ke inovacije (parna ma{ina, ma{ine za tekstil, bessemerov postupa prerade ~elika i sl.)

- Formiranje preduze}a kao osnovnih privrednih subjekata - Pojava konkurencije krajem perioda

1900 – 1930 PERIOD MASOVNE PROIZVODNJE

- Proizvodnja standardnih proizvoda u velikim koli~inama i sa niskim cijenama - Proizvodni mentalitet menadèra i fokus na interne probleme preduze}a - Minimalna politi~ka i dru{tvena kontrola preduze}a (“[to je dobro za general motors, dobro je i za dràvu”)

1930 – 1955 PERIOD

MASOVNOG MARKETINGA

- Prelazak sa pasivnog na aktivni odnos prema potro{a~ima i njihovim potrebama (“kupac je kralj”) - Proizvo|a~i potro{nih dobara i tehnolo{ki intenzivni proizvo|a~i prvi usvajaju marketing koncept a kasnije i ostali

- Ja~a IR funkcija u preduze}u (DU PONT, BELL, GE) - Iznena|enja su bila izuzetak

1955 - DANAS

POSTINDUSTRIJSKI PERIOD (VRIJEME DISKONTINUITETA

- “Dru{tvo obilja”, potro{a~i sve vi{e zahtijevaju - Inflacija, dràvne restrikcije, nezadovoljni potro{a~i, strana konkurencija, brze inovacije u tehnologiji, izmjene stavova prema radu

- IR sam generira neplanirani razvoj - Socijalne aspiracije prelaska sa kvantiteta na kvalitet ìvota - Privredni giganti postaju opasnost za ekonomsku efikasnost (monopoli) i za demokratiju (snaga) - Od preduze}a se zahtijeva i ekonomska efikasnost i dru{tvena odgovornost - Dru{tveno ekonomski odnosi preduze}a sa okruènjem su pitanje “@ivota i smrti” - Preduze}e iz poznatog svijeta marketinga i proizvodnje ulazi u nepoznati svijet novih tehnologija, novih konkurenata, novih stavova potro{a~a, novih dimenzija dru{tvene kontrole i nove uloge preduze}a u dru{tvu

- Ubrzanje i difuzija promjena, koje se te{ko predvi|aju - Potreba brzog prilago|avanja na promjene i implementacije odgovora

Table 2. Historical development of business in USA

PERIOD NAME BASIC CHARACTERISTICS

1820 – 1900 INDUSTRIAL REVOLUTION

- Great technological innovations (steam engine, textile machinery, Bessemer procedure of steel production et al.)

- Formation of companies as basic economic subjects - Appearance of competition near the end of the period

1900 – 1930 MASS

PRODUCTION PERIOD

- Production of standard products in large quantities and with low prices - Productive mentality of managers and focus into internal company problems - Minimal political and social control of the company ("What is good for General Motors, it is good also for the country")

1930 – 1955 MASS MARKETING

PERIOD

- Advance from passive to active relationship with consumers and their needs ("Consumer is the king") - Producers of consumer goods and technologically intensive producers are first to accept marketing concept, followed by others

- IR function in the company strengthens (Du Pont, Bell, Ge) - Surprises were exceptions

1955 - TODAY

POSTINDUSTRIAL PERIOD (AGE OF DISCONTINUITY)

- "Society of wealth", consumers require more and more - Inflation, state restrictions, dissatisfied consumers, foreign competition, rapid innovation in technology, changes in attitudes to the work

- IR generates unplanned development by itself - Social aspirations of transfer from quantity to the quality of life - Economic giants become hazard for economic efficiency (monopolies) and for democracy (power) - Company is required to have both economic efficiency and social responsibility - Social and economic relations between the company and environment are question of "life and death

- Company moves from known world of marketing and production to unknown world of new technologies, new competition, new consumer attitudes, new dimensions of social control and new role of the company in the society

- Acceleration and diffusion of hardly predictable changes - The need for rapid adjusting to changes and response implementation

Iz prezentiranog nije te{ko vidjeti i da li~nosti kojesu vodile bizis u razli~itim etapama njegovog razvo-ja, nisu trebale imati sve iste osobine.Sigurno je da je "vrijeme diskontinuiteta" zahtijevavi{e specifi~nih osobina od generalnog menadèra,nego "staro dobro vrijeme" u kojem nije bilo naglihpromjena ili su se bar neke mogle predvidjeti.Autor Mitchell Ford [6] u tom smislu daje jednuinteresantnu klasifikaciju. Neke od osobina gene-ralanih menadèra, po mi{ljenju ovog autora, bile suoduvijek neophodne, kao {to su i danas. Drugagrupa su osobine koje zahtijeva novo okruènje ikoje ranije nisu bile neophodne.

U prvu grupu, koju moèmo nazvati konvencionalnimosobinama generalnog menadèra, spadaju njegovesposobnosti da bude dobar:

1. donosilac odluka,2. planer,3. ekonomista i 4. strateg

Svaka od navedenih osobina zahtijeva odre|enoobja{njenje. Biti dobar donosilac odluka (1), zna~i u su{tinisposobnost da se, nakon saslu{anih mi{ljenja sviheksperata i pro~itanih izvje{taja o datom problemu,kona~no kaè: "Idemo u akciju" ili "Ne"! To jepovezano sa poduzetni~kim kvalitetama osobe,sklonosti ka riziku i sl.Ako generalni menadèr odigra dobro svoju uloguplanera (2), korektne i aktuelne informacije su priku-pljene i date svima kojima su potrebne u preduze}u.To stvara klimu sistematskog pripremanja strate{kihodluka, pove}ava njihov kvalitet i olak{ava prvuulogu iste osobe.Kada se kaè da generalni menadèr mora biti iekonomista (3), misli se prije svega na njegovuodgovornost za racionalno kori{tenje ograni~enihresursa preduze}a i njihovu alokaciju. To nikako nezna~i da prvi ~ovjek u preduze}u mora biti po obra-zovanju ekonomista. On samo treba da po{tuje bitneekonomske zakonitosti pri dono{enju svojih odluka.Kori{tenje opreme, materijala, ljudi, novca, vremenai ostalih resursa mora pro}i test ekonomi~nosti usvim prilikama i zato je ova uloga generalnogmenadèra vidljiva i bitna.Kao glavni strateg (4), prvi ~ovjek preduze}a je injegov futurista. On u tom smislu treba da imasposobnost identifikovanja promjena i kreiranja me-hanizama za upravljanje njima.Navedene ~etiri uloge generalnog menadèra bile sui ranije zna~ajne ali danas dolaze do izraàja jo{~etiri, koje su vi{e povezane sa interpersonalnimodnosima unutar i van preduze}a;


- 221111 -

It is easy to see from the survey presented alsothat persons, who lead business in different periodsof its development, did not need to have samecharacteristics.It is sure that "age of discontinuity" requests morespecific characteristics from general manager than"good old times" when there were no sudden changesor these changes could at least be predicted.Author Mitchell Ford [6] in that sense gives one inter-esting classification. Some characteristics of generalmanager, according to this author, were necessarysince the beginning, just like they are today. Secondgroup consists of characteristics requested by newenvironment, which were not necessary earlier.Into the first group, which can be called conven-tional characteristics of general manager, belong hiscapabilities to be good:

1. Decision maker2. Planner3. Economist and4. Strategist.

Each one of these characteristics needs furtherexplanation.To be good decision maker (1), means in generalto be capable to, after listening all expert's opinionsand reading all reports about given problem, finallysay: "Let's do the action" or "No"! It is connectedwith entrepreneurial qualities of a person, riskiness,etc.If general manager plays well his role of planner(2), correct and actual information are collected anddistributed to all who need them in the company.That creates climate of systematic preparation ofstrategic decisions, increase their quality and allevi-ates first role of the same person.When we say that general manager has to be goodeconomist (3), we mean firstly about his responsi-bility to rationally use limited company resourcesand their allocation. That surely does not mean thatfirst man in the company must be economist by hiseducation. He only needs to respect important eco-nomic laws while making his decisions. Use ofequipment, materials, people, money, time and otherresources must pass the economical test in all cir-cumstances and because of that, this role of gen-eral manager is visible and important.As a main strategist (4), first man of the companyis also its futurist. In that sense, he needs capabil-ity to identify changes and to create mechanisms tomanage them.These four roles of general manager were signifi-cant even earlier, but today there are four more,which are more connected to interpersonal relationsinside and outside the company:

5. savjetnik,6. arbitar,7. lider u filozofskom smislu i8. borac

Kao prvi ~ovjek preduze}a, generalni menadèr jeglavni savjetnik (5) svima koji done{ene odluke trebada provode. Smatra se da je danas ova uloga svazna~ajnija, po{to je u savremenoj decentralizovanojorganizaciji i "plitkoj formaciji", sve manje odlukakoje generalni menadèr direktno sam donosi.U~estali konflikti izme|u funkcionalnih menadèra,zaposlenih, vlasnika i ostalih koji su zainteresovaniza rezultate preduze}a, ~ine sva zna~ajnijom ulogumenadèra kao arbitra (6). Treba mnogo znanja isposobnosti da se ova uloga kvalitetno odigra udana{njem kompleksnom mikro i makro okruènjupreduze}a.U savremenom preduze}u generalni menadèr bi tre-bao biti i lider u filozofskom smislu rije~i (7), odno-sno simbol personalnosti i karaktera biznisa. To zahti-jeva od njega vi{e intelektualne {irine i spremnostida prihvati nove stavove prema kvalitet rada i ìvota.Sa druge strane, treba znati odbraniti integritet iugled biznis od svih ve}ih pritisaka iz okruènja.Kona~no, prvi ~ovjek preduze}a mora danas biti iborac (8). Pod tim se prvenstveno podrazumijevakontinuirana borba protiv "status quo" situacije upreduze}u. Treba, dakle, imati kreativnu sposob-nost da se vidi problem gdje ga drugi jo{ ne vide,da se osjeti glavobolja prije nego {to postanefizi~ka bol. Formula "ne talasaj", smatraju mnogi, jeuni{tilo vi{e preduze}a nego svi drugi uzrocizajedno!

Ve} citirani autori Ansoff i McDonell [7] tako|e supoku{ali da, predvi|aju}i karakteristike preduze}abudu}nosti, odrede i nove osobine koje }e moratiimati uspje{an generalni menadèr.Njihova polazna hipoteza je da svaka faza razvojaokruènja i preduze}a zahtijeva odre|enu osobinugeneralnog menadèra i zato je on u pro{losti bio"~ovjek trenutka".Posmatrano sa istorijskog aspekta, u prvoj polovi-ni na{eg vijeka kada su se industrije formirale,uloga poduzetnika je bila najva`nija. Nakon toga,tokom faze ranog rasta biznisa uloga administrato-ra je bila kriti~na, da bi se postigli minimalnitro{kovi i trì{no u~e{}e. Zato slijedi uloga plan-era, pa kontrola da bi se maksimizirao dotokgotovine u preduze}u.

Koncept "~ovjek trenutka" nije, me|utim, vi{e mogu}u devedesetim godinama na{eg vijeka i traì sekombinacija novih osobina, ne zanemaruju}i stare.


- 221122 -

5. Advisor6. Arbiter7. Leader in philosophical sense and8. Fighter.

As a first man in the company, general manager ismain advisor (5) to everyone who need to realise givendecisions. It is assumed that this role is more andmore important today, since in modern decentralisedorganisation and "low formation", there are less andless decisions that general manager makes on his own.Often conflicts between functional managers,employees, owners and other who are interested incompany results, increase importance of the arbitraryrole of the general manager (6). A lot of knowledgeand capabilities are required to play this role goodin today's complex micro and macro environment ofthe company.In modern company, general manager should be aleader in philosophical sense (7), i.e. symbol of per-sonality and character of the business. That requiresmore intellectual wideness and capability to acceptnew attitudes to work and life quality. On the otherhand, he needs to know how to defend integrityand reputation of the business to increasing pres-sure from the environment.Finally, first man in the company must be a fighter(8). That mainly encounters continual fight against"status quo" situation in the company. Therefore heshould have creative capability to see the problemwhere other people still does not see it, to feelheadache before it becomes physical pain.Formula "do not surge", as lots of people think,destroyed more companies than all other causestogether!Already mentioned authors Ansoff and McDonnell [7]also tried to, by predicting characteristics of thefuture company, define new characteristics that suc-cessful general manager should have.Their starting hypothesis is that every phase ofdevelopment of the company and environmentrequires certain characteristic of general manager,what made him "man of the moment" in the past.Historically reviewed, in the first half of our century,when industries started to form, the role of entre-preneur was the most important. After that, duringthe phase of early business growth, the role ofadministrator was critical, to gain minimal expensesand to take part at the market. Then role of plan-ner follows, then controller to maximise cash incometo the company.

Concept "man of the moment" is not possible anymore in nineties and combination of new charac-teristics is required, without ignoring old ones.

Nekoliko bitnih karakteristika preduze}a budu}nosti ,odredi}e i nove osobine koje se o~ekuju od gene-ralnog menadèra.a) menadèri budu}nosti }e djelovati u znatno {irem

institucionalnom okruènju. Socijalni problemi }ebiti nametnuti od strane dràve i zahtijevati ve}uinterakciju izme|u preduze}a i dru{tva. Konkuren-cija dràvnih agencija, nevladinih institucija, uni-verziteta, fondacija i sl., u alokaciji resursadru{tva }e biti sve ve}a i preduze}e ne}e vi{edominirati u tom pogledu. Sistemi vrijedno-sti i usamim preduze}ima }e se izmjeniti i vi{e usm-jeriti ka kvalitetu rada i ìvota. Globalizacijabiznisa }e uvu}i menadère u vi{e razli~itih kul-tura, nivoa ekonomske razvijenosti, sistemaupravljanja i sl. To }e sve zahtijevati da budu}igeneralni menadèr bude i dobar dràvnik, sasmislom za istoriju, politiku i kulturu i sposob-no{}u da u tom kontekstu uspje{no pregovara iuti~e na poloàj svoga preduze}a.

b) Drugi trend ka preduze}u budu}nosti je informa-tizaciona eksplozija. Ona }e do}i zbog novihtehnologija ali i {irenja globalnih ciljeva pre-duze}a. To }e zahtijevati novu organizaciju iznanja svih ali prvenstveno generalnogmenadèra. Zbog toga }e on morati postati iuspje{an kreator novih sistema, bez obzira napodr{ku specijalista.

c) Tre}i prepoznatljivi trend je porast kompleksnostipreduze}a kao bihejvioristi~kog sistema. Poredinformacionog sistema, dakle komplikuju seodnosi izme|u ljudi, {irenjem razli~itih aspiracija,vrijednosti, normi i kultura. Za razliku od nekihmi{ljenja, preduze}e budu}nosti }e biti radnointenzivno i vi{e nego ikada zavisiti od imagi-nacije, kreativnosti i inicijative ljudi. To sugeri{eda generalni menadèr bude, ne samo kreatorinformacionih sistema, nego i sistema pona{anjaljudi.

d) Sve ve}i diskontinuitet promjena u okruènju je~etvrti trend sa kojim }e se suo~iti preduze}abudu}nosti. Intenzivna konkurencija, izmjene utra`nji, tehnologijama, marketingu i sl., ponovo }eaktuelizirati potrebu za poduzetni~kim kvalitetimageneralnog menadèra. Isto kao {to je to bilo udoba Industrijske revolucije, talent, njuh, imagi-nacija i sklonost ka riziku }e bit èljene osobineprvog ~ovjeka preduze}a, ukoliko ono èli daopstane u novom, dinami~kom okruènju.


- 221133 -

Some important characteristics of future entrepreneurwill define new characteristics requested from gen-eral manager.a. Future managers will act in much wider institu-

tional environment. Social problems will beimposed by the state and they will require moreinteractions between the company and society.Competition of state agencies, non-governmentalinstitutions, universities, foundations et al., in allo-cation of social resources will increase and thecompany will not dominate in that sense. Systemsof values in companies themselves will changeand redirect to the quality of work and life. Glob-alisation of the business will introduce managersto more different cultures, levels of economicdevelopment, managing systems etc. All that willrequire general manager to be a good governor,with sense for history, politics, culture and capa-bility to successfully negotiate and influence hiscompany's position in that sense.

b. Second trend to future company is informationexplosion. It will occur because of new tech-nologies, but also because of enlarging of glob-al goals of the company. That will require neworganisation and knowledge of all, but especial-ly general manager. Because of that, he will haveto become successful creator of new systems,regardless of support of specialists.

c. Third recognised trend is growth in complexity ofthe company as behaviourist system. Exceptinformation system, human relations becomemore complicated, by distributing different aspi-rations, values, norms and cultures. Opposite tosome opinions, future company will work inten-sive and more the ever will depend on imagi-nation, creativity and human initiative. That sug-gests that general manager should be, not onlycreator of information systems, but systems ofpeople behaviour.

d. Growing discontinuity of environmental changes isfourth trend that future companies will be facedwith. Intensive competition, changes in needs,technologies, marketing etc will actualise againthe need for entrepreneur qualities of generalmanager. Just like it was during industrial revo-lution, talent, sense, imagination and riskiness willbe desired characteristics of company's firstman, if the company wants to survive in new,dynamic environment.

3. CONCLUSION

When all characteristics that general managersshould have are summarised, it is easy to concludethat there is extremely small number of individualswith such qualities. Because of that, most of themneed to become expert to be able to use experts.That is the only way to successfully manage mod-ern and future company. To know how to managechanges and how to use experts, will probably becritical mass of future general manager quality.In more and more complex strategic decision mak-ing processes, just because individual can not haveal important characteristics required, the main pointwill be team management (corporate office), com-mon authority and responsibility.Real question and real challenge to all countries intransition is surely, how to fulfil all these require-ments when environment, companies and people arechanging so drastic? Which are extra requirementsexcept these analysed?Answers to these questions will be researched longtime in theory and practice, but it is sure that edu-cation and researches should start immediately!


- 221144 -

3. ZAKLJUÂK

Kada se saberu sve osobine koje bi generalnimenadèri trebali imati, lako je do}i do zaklju~ka daje izuzetni mali broj pojedinaca sa takvim kvaliteti-ma. To je razlog {to ve}ina njih mora postati eksper-tima za kori{tenje eksperata. Jedino na taj na~inmoè se uspje{no upravljati savremenim preduze}emi preduze}em budu}nosti. Znati upravljati promjena-ma i znati koristiti eksperte, vjerovatno }e biti kriti~namasa kvaliteta generalnog menadèra budunosti.U sve kompleksnijim procesima strate{kog odlu~iva-nja, upravo zbog nemogu}nosti da pojedinac imasve bitne osobine koje se traè, teì{te }e biti natimskom upravljanju (corporate office), zajedni~komautoritetu i odgovornosti.Pravo pitanje i pravi izazov za sve zemlje u tranzi-ciji je svakako, kako zadovoljiti sve ove zahtijeve uuslovima kada se drasti~no mijenja okruènje, pre-duze}e i ljudi? Koji su dodatni zahtijevi, pored ovihkoje smo analizirali?Odgovore na ova pitanja dugo }emo traìti i u teori-ji i u praksi ali je sigurno da sa edukacijom iistraìvanjima treba po~eti odmah!

LITERATURA-RREFERENCES

[1] J. Ansoff; E. McDonell: "Implanting StrategicMnagement", Prentice hall, London, 1990, strana 12

[2] Opp.cit., stran 13.

[3] H.Minzberg: "Organization Design or Fit"? HarvardBusiness Review, Jan/Feb, 1981.

[4] A Sloan: "My Years with General Motors", Dou-bleday, New York, 1990, strana xi

[5 L.E. Boome; D.L. Kurz: "Conterporay Business",The Dryden Press, Chicago, 1990, str. 16.

[6] T.M. Ford: "The Job of the Chief Exsecutive",Management Handbook, John Wiley, New York,1981, strana 67.

[7] J. Ansoff; E. McDonell: Opp.cit., strana 291.

Documents

mf.unze.bamf.unze.ba/wp-content/uploads/2016/04/Masinstvo-num4-1998.pdf · grisani uticaj teorija, strategija, i alata modeliranja na ... igrati u procesu konstruisanja, mora se prethodno