Kanban esaslı bir üretim hattında benzetim çalışması
Simulation study of a kanban based production line
- Tez No: 39854
- Danışmanlar: PROF.DR. GÖNÜL YENERSOY
- Tez Türü: Yüksek Lisans
- Konular: Endüstri ve Endüstri Mühendisliği, Industrial and Industrial Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1993
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 160
Özet
ÖZET Günümüzün rekabet koşullarında modern üretim yöneti mi yaklaşımları dünyada olduğu gibi Türkiye'de de büyük önem kazanmaktadır. Günümüzde firmaların rekabet koşul larına dayanabilmeleri için, sahip oldukları kaynakları en yüksek verimlilikle ve israf etmeden kullanmaları, ay rıca maliyetleri en aza indirgemeleri gerekmektedir. Dünya' da bir çok endüstriyel sektörde kullanılan tam zamanında üretim sisteminin felsefesi, gereken miktarda, gereken zamanda, gereken yerde ve istenilen kalitede üre tim yapmaktır. îşgörenlere de verdiği sorumluluk ve yet ki ile verimlilik artışı ve kalite kontrolü gibi diğer a- lanlarda da kullanılmaktadır. Ideal olarak bir tam zamanında üretim sisteminde hiç ara stok bulunmaz ve bu duruma sadece üretim aşamalarının operasyon sürelerinin sabit ve eşit olduğu hallerde ula- şılabilinir. Ancak bu duruma, gerçek üretim ortamlarında ulaşılması imkansızdır. Bu nedenle üretim sisteminde bu lunacak kanban sayıları ve kanban taşıyıcı kapasiteleri nin belirlenmesi önem kazanmıştır. Bu çalışmada, kanban esaslı çekme sistemi uygulanan bir üretim hattında SIMAN benzetim dili kullanılarak bir benzetim model leme çalışması yapılmış ve elde edilen so nuçların maliyet bazında değerlendirilmesi için bir mali yet modeli kurulmuştur. - xii
Özet (Çeviri)
SUMMARY Simulation Study of a Kanban Based Production Line In the competitive world markets of the 90 's, manufacturing companies are learning that survival is predicated upon a commitment towards continual process and product improvement. In today's rapidly changing mar kel; place, the company that can deliver a high quality, low cost product to the marketplace will win the most market share. On the production side, this translates into a need for management to continually find ways to improve manufacturing conformance quality, while simultaneously seeking to reduce excessive“waste”in the system. Just-in-time (JIT) encompasses a philosophy which can assist companies in this regard. The just-in-time philosophy has gained significant acceptance in the world in the last decade. Japanese just-in-time production systems are so different from the production systems currently in use in North America and Europe. The Japanese“pull system”which is in the JIT technique is designed to minimise in-process inventory and its fluctuations. A“Pull”manufacturing strategy is a necessary ingredient for providing good and efficient control over all system inventory levels. The pull system simplifies Inventory controls, prevents amplified transmission of demand fluctuations from stage to stage and raises the level of shop control through decentralization. Another benefit of the pull system is an earlier awareness of quality problems, thus reducing reworking and improving product quality. Ideally the just-in-time production system holds no buffer inventory since the production is just in time. This is achieved only when the stage's operation times are constant and equal but realistically the variability of operator performance and the unequal distribution of task times precludes such an ideal situations. xm -In a pull system, the succeding stage demends andwithdraws in-process units from the preceding stage onlyat the rate the succeding stage consumes the items. Theideal pull system with one unit of inventory at eachstage is not achievable in a real manufacturing systems,where a variation in processing times, imbalance ofworkloads between stages, demand fluctuations and machinebreakdowns are inevitable. Consequently, a substantialamount of work-in-process (WIP) must be maintained ateach stage of the pull system in order to reduce theeffects of these variations and machine breakdowns.In a JIT system, inventory level between stages iscontrolled by the number of Kanbans allocated. A Kanbancard is used as a communication tool between twoproduction stages. A Kanban is sent from a workstationto the preceeding workstation to initiate production of aunit or a specified number of unit. Since the goal whichone unit of inventory must be at each production stage isnot achiavable in real manufacturing enviroments, thedetermination of the number of Kanbans is an issue ofconsiderable interest for researchers.There are many research about Kanban-based pullsystems. Efforts to model kanban-based pull systems havefollowed three main ways:(i) Simulation models, where digital simulation isused to explore the effects of different systemparameters and configurations,(ii) Deterministic models, where a mathematical modelof system behaviour is developed that assumes onlydeterministic relations,(iii) Stochastic models, where demand and processing arecharacterized by stochastic processes.A simulation model seeks to“duplicate”the behaviorof the system under investigation by studying the inter-actions among its components. Simulation must be treatedas a statistical experiment. Unlike the mathematicalmodels the output of the model represents a long-runsteady state behavior, the results obtained from runninga simulation model are observations that are subject to- xiv -experimental error.A simulation experiment differs from the regularlaboratory experiment in that it can be conducted totallyon the computer systems. By expressing the interactionsamong the components of the system as mathematical rela-tionships. We are unable to gather the necessary infor-mation in very much the same way as if we were observingthe real-life system.The nature of simulation thus allows greaterflexibility in representing complex systems that arenormally difficult to analyze by standard mathematicalmodels. Although simulation is a flexible technique,both time consuming and costly, particularly when one istrying to optimize the simulated system.Simulation is a useful tool to assist companies inevaluating the oppurtunities that exist for implementingJIT. Existing simulation languages generally fall shortin their potential for use in simulating JIT systems. ButSIMAN was selected for this simulation study because ithas uniquely open architecture (Pegden et al. 1990).The subject of this master thesis is related to thesimulation and cost modelling of a kanban basedproduction line. As the simulation language, SIMAN hasbeen used to build the model representing a productionline which has ten assembly workstations.A SIMAN simulation is divided into three distinctactivities; system model development, experimental framedevelopment and data analysis. Within these three acti-vities, the SIMAN software consists of five individualprocessors which interact, through four data files;(i) The model processor is used to construct a blockmodel diagram model. The data file that is generated iscalled the model file.(ii) The experiment processor is used to define theexperimental frame for the system model. The data file- xv -that is generated is called the experiment file.(iii) The link processor combines the model file andthe experiment file to produce the program file.(iv) The program file is input to the run processorwhich executes the simulation runs and writes the resultson the output file.(v) The output' processor is used to analyze, formatand display the data contained in the output file.In the simulation study, some assumptions have beenused. A production day consist of 480 minutes, containercapacity is two hundred items, the production line isdedicated to only one product, the processing time ateach workstation is independent and identicallydistributed, there is unlimited and instantenous supplyof raw material available, a single unit of raw materialor unfinished product is used to make one unit offinished product, ten same number of Kanbans is used ateach stations, transportation time between stations isnegligible, jobs are scheduled on first-come-first-servedbasis, no breakdowns occur at any station and no scrapis produced.The simulation model was run for a given number ofKanbans with diffrent demand and processing timedistributions. The mean demand of 20 container full ofproducts (i.e. 2000 items per day). Mean processing timeis different at each workstation. The distribution ofexperiments which has been used in this study is shown asfollow :DemandConstantNormal (CV=0.1)Normal (CV=0.5)Number ofProcessing timeKanbansConstant1-6Normal (CV=0.1)1-6Normal (CV=0.5)1-6Exponential1 - 6Uniform (CV=0.1)1-6- xviFor these experiments, 3 demand levels, 6 process I. ngtimes and 6 Kanban levels were considered. Thus, a totalof 108 different simulation conditions are examined. Thefollowing system performance measures were considered:(i) Average utilization of all workstations (for 30days of production) ;(ii) Average work-in-process inventory (units beingprocessed and waiting to be processed for 30 days ofproduction) ;(iii) Average demand meeting rate (for 30 days ofproduction).The data about performance measures from eachsimulation run were collected. The cost model has beenbuilt to analyse the data about performance measures atbase of cost.In this model; inventory holding costs, non-meetingdemand costs, machine-run costs and idle-machine costsare considered. Total cost for each kanban number in theproduction line is sum of the value of four cost types.The Kanban number which gives minimum total cost isoptimum number of kanban in that demand and processingconditions at the kanban based production line.As future research directions, the relation betweencost types should be investigated. Furthermore, there aremany factors which affect system performance measuressuch as random machine breakdown, varying set-up times,unbalanced work stations, the product mix, non-zerotransportation time between workstations. All of themshould be investigated.- xvii -
Benzer Tezler
- Üretim sistemlerinde çekme tipi esaslı iç lojistik sistemi tasarımına yönelik bütünleşik bir metodoloji önerisi: Beyaz eşya sektöründe bir uygulama
An integrated methodology proposal for towing type based internal logistics system design: An application in the household appliances sector
TUBA TOPUZ
Yüksek Lisans
Türkçe
2021
Endüstri ve Endüstri Mühendisliğiİstanbul Teknik ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
PROF. DR. EMRE ÇEVİKCAN
- Tam zamanında üretim ortamında çizelgeleme ve bir simülasyon çalışması
Scheduling and simulation study in environment of just in time
HÜLYA ÖZKALAYCIOĞLU
Yüksek Lisans
Türkçe
1997
Endüstri ve Endüstri MühendisliğiSakarya ÜniversitesiEndüstri Mühendisliği Ana Bilim Dalı
YRD. DOÇ. DR. ORHAN TORKUL
- Yalın üretim tekniklerinin incelenmesi ve imalat sektöründe bir uygulaması
Examination of lean manufacturing techniques and an application for manufacturing sector
NERMİN KESİT
Yüksek Lisans
Türkçe
2020
Endüstri ve Endüstri Mühendisliğiİstanbul Üniversitesi-CerrahpaşaEndüstri Mühendisliği Ana Bilim Dalı
DOÇ. DR. FATİH TÜYSÜZ
- CONWIP esaslı üretim sistemlerinin tasarımı ve yönetimi
Design and management of CONWIP production control systems
ÖZGÜR KABADURMUŞ
Yüksek Lisans
Türkçe
2008
Endüstri ve Endüstri Mühendisliğiİstanbul Teknik ÜniversitesiEndüstri Mühendisliği Bölümü
PROF. DR. M. BÜLENT DURMUŞOĞLU
- KOBİ'lerde tam zamanında üretim felsefesinin uygulanabilirliğinin araştırılması
An Eximination of the applicability of just in time manufacturing philosophy in small and medium enterprises
SENCER GÜNEER