Yüke bağlı imalat kontrolü
Load-oriented manufacturing control
- Tez No: 39877
- Danışmanlar: PROF.DR. MURAT DİNÇMEN
- Tez Türü: Yüksek Lisans
- Konular: Endüstri ve Endüstri Mühendisliği, Industrial and Industrial Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1994
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 109
Özet
ÖZET Yüke bağlı sipariş serbest bırakma yaklaşımı, atölye imalatında planlama yöntemi olarak, sipariş serbest bırakmanın otomatikleştirmesine yaramaktadır. özellikle, atölye imala tında, ürünlerin çeşitli sıraları yüzünden, genelde çok uzun akış süreleri ve böylece büyük stoklar oluşmaktadır. Akış süresinin %85'i geçiş süresi (taşıma ve bekleme süreleri) denilen süreden oluşmaktadır. Bu yüksek oran, termin planla masında büyük güvensizliklere neden olmaktadır ve ayrıca erken yüklemeye ve böylece işletme araçları ömünde yığılmalara ve uzun akış sürelerine sebebiyet vermektedir. Bu yüzden acil siparişler, imalat sisteminden geçirilmektedirler. Ancak, bu da diğer siparişlerin akış sürelerini uzatmaktadır. Bu iliş ki, imalat kontrolünde hata çemberi veya akış süresi sendromu diye anılmaktadır. Bu nedenle, sipariş serbest bırakmada, siparişlerin termine uygun olarak tamamlanabilmesi için yeterli kapasitenin hazır tutulmasına dikkat edilmelidir. Burada yüke bağlı sipariş serbest bırakma devreye girmektedir. Her plan periyotunda, siparişler ilk önce aciliyetlerine göre ayrılıp sıralanmaktadırlar. Bundan sonra siparişler için, emniyet sürelerinin hesabıyla birlikte akış terminlemesi yapılmakta dır. Bütün iş sistemlerinde, verilen bir yük sınırı (örn., periyot kapasitesinin %200-300) aşılmadıkça yükleme devam ede cektir. Aksi halde, sipariş bir sonraki periyoda kaydırılmak tadır. Tecrübeler gösterdi ki yüke bağlı sipariş serbest bırakma belirli durumlarda akış sürelerini azaltmaktadır. Bu yöntemin dezavantajı ise, siparişlerin teker teker hedef li kontrolüne olanak vermemesidir. Ayrıca, yük sınırının önceden belirlenmesi zorunluluğu ayrı bir problemdir, çünkü düşük seçilmiş bir yük sınırı sistemden kötü yararlanmasına neden olmaktadır. Benzetim yardımıyla sınırının belirlenmesi önerilmektedir. Yüke bağlı sipariş serbest bırakmayla ilişkili olarak, basit ve görsel temsil imkanı huni modelidir. Burada, atölye ler huni, bunların kapasitesi huni ağızı ve iş sistemi önünde bekleyen siparişler de dolum yüksekliği olarak temsil edilmek tedir.
Özet (Çeviri)
SUMMARY LOAD-ORIENTED MANUFACTURING CONTROL Load-oriented manufacturing control is a new solution for jop shops producing in small lots. Based upon a statistical view of kop. shop manufacturing, load-oriented manufacturing control contrasts with conventional deterministic short term planning methods. The nucleus of load-oriented manufacturing control is the technique of load-oriented order release. This simple and consistent technique mas developed during research at the Institut für Fabrikanlagen of the Universitat Hannover, Germany. Since then load-oriented order release was completed by further techniques and has been successfully implemented in at least 20 factories. Several software companies are offering load-oriented manufacturing control systems. The latest state of the art can be found in Wiendahl (1987). Increasing product variety as well as shortening life cycles and shrinking lot sizes, demand a high flexibility from today's production plant. At the same time the customer's pressure on delivery times is steadly growing. However, long lead times and high inventories become more and more harmful to keeping up with the tightening competition. Especially for jop shop manufacturing, the still predominant and also most flexible production tybe, reducing lead times and work-in-pro cess rank among the most urgent objectives of manufacturing control. On the other hand, the objective of high capacity utili zation, wich is often overvalued, gradually loses weight against low inventories, short lead times and high due date performance. Short and determinable lead times facilitate on- time production as production planning becomes safer and more reliable. Satisfying the customer with on-time delivery finally guarantees a high level of employment and a successful business. Conventional manufacturing control systems are usually based on expensive computerized short-term planning methods, such as finite loading. These systems perform iterative procedures of daily load leveling and operations sequencing by means of network planning techniques and pretend to deliver an optimal solution. Due to the great number of activities this task can hardy be solved as even large computers, in general, VIneed many hours for this procedure. Beyond that, these short- term planning techniques must fail principally because they make every operation a critical one and thus do not leave any room for the natural and inevitable incidents of manufacturing reality (such as breakdowns, shortages, absenteeism, quality problems, data inaccuraties, etc.}- Therefore, the planning results very soon become obsolete and have to be continually revised. The intended high, but haryly reliable, precision and high cost of deterministic systems just do nor meet the reality of jop shop manufacturing. Contrary to that, the technique of load - oriented manufacturing control apply a statistical view of jop shop manufacturing. The idea is to limit and balance work-in- process inventory on a level as low as possible in order to accomplish a high workcentre utilization as well as a rapid and in-time flow of orders through the factory. Evaluating the feedback records from the jop shop input, output, inventory and average lead times of workcentres are monitored and compared to their planned values in a dynamic way. In addition the flow of work orders to and from the factory is controlled by measuring downstream and upstream inventory including planned and released orders. Using these control information inventories, lead times and due dates are planned applying the technique of load-oriented order release and performing backword scheduling with true lead times. The crucial point is that the system keeps actual lead times on a planned and predetermined level in a self -regulating way. The system also points out the bottlenecks and initiates capacity adjustments and due date changes if necessary. Numerous case studies very often revealed an alarming discrepancy between planned and actual lead times. However, reliable and determinable lead times are essential for job shop manufacturing. In general the parts produced in the job shop are scheduled for the assembly of products and components on given due dates. Reliable due data planning is only possible with correct lead times. Furthermore effective planning and allocation of the required raw materials and capacities depend on correct lead times, too. Therefore wrong and uncontrollable lead times are the main reason for the failure of many production planning and control systems. The philosophy of load-oriented manufacturing control can best be explained using the funnel model of a job shop. The model is divided into the two sections: 'order stock', contai ning the planned orders, and 'job shop', containing the released orders (work-in-process). The job shop itself consists of numerous workcentres. The resulting manufacturing flow between the workcentres then becomes very complex and changes continuously according to the VIImix of orders. At each of the workcentres there are generally several orders competing at the same time to be processed. Each workcentre acts as a funnel in which the funnel outlet represents the workcentre capacity, the contents of the funnel represents the current order queue and channels the flow of orders to and from the workcentre. In order to be consistent with usual feedback systems the completions of orders processed at a workcentre are defined as output and completions of orders processed at the preceeding upstream workcentres are defined as input. The essential events in the flow of orders through the whole system: 1. Entry (order defination) 2. Release (start of manufacturing flow) 3. Input (arrival at a work center) 4. Output (departure from a workcentre) 5. Exit (end of manufacturing flow) These events depend on various decisions by the control and production personnel requiring effective manufacturing control techniques. For this purpose load-oriented manufacturing control has a planning system for three planning levels: 1. Order entry and mid-term capacity planning. 2. Order release and short-term capacity planning. 3. Operation sequencing. In addition load-oriented manufacturing control has a control system with graphs and reports to illustrate the manufacturing flow and to monitor inventories, lead times, due date deviations and capacity utilization.. Applying the philosophy of load-oriented manufacturing control, means regulating the flow of orders in such a way that all of the workcentres keep running whereas the orders proceed as fast as possible and are delivered on-time: thus the job shop turns into a flow process. The objective is to limit and balance work-in-process (WIP) step by step on a level as low as possible. Limiting and balancing work-in-process requires insight in the basic relationships at a workcentre. The important variables are: input, output, inventory (WIP), and lead time. The arriving orders (input) queue up contributing to inventory and after a certain time of waiting they are processed and completed (output). Depending on their priorities the single orders are preferred or deferred or even processed one after the other, resulting in quite different actual lead times. However, the average lead time of a VIIIworkcentre equals the ratio of average inventory and average output. Extensive overload aggravates the competition among the queueing orders and operation sequencing becomes more and more erratic. Actual priorities will be very different then from the 'natural' sequence according to the FIFO-rule and can be divided in four groups : * express orders * normal orders * deferred orders * neglegted orders The further the actual priorities differ from the FIFO-rule, the larger will be the resulting variety of individual lead times. Transforming the five rules mentioned above into applicable techniques for job shops requires a quantitative model of the relationships described above. The appropriate quantitative model is a so called *work-guantity-versus-time- chart' (Bechte 1982) or 'throughput diagram' (Wiendahl 1987) Charts like this represent a special form of input- output-control and can be generated if feedback records are continuously available for all orders processed in the job shop. Throughput diagrams are universally valid for all kinds of production processes. Applying them to job shops is the new approach of load-oriented manufacturing control. The basic form of a throughput diagram consists of an input and output curve describing the arrivals and completions of orders processed at the workcentre. The units to measure are usually hours for work quantity and days for time. The vertical distance between input and output indicates the momentary queue length, i.e. the inventory (WIP). The resulting ratio of average inventory and average output is defined as average time supply (ATS) describing the time the workcentre could be kept busy by average inventory at the current output rate without replenishment. The horizontal distance between input and output, also denoted as displacement time, indicates the time the just completed order would have stayed at the workcentre if the dispatching rule was FIFO. The resulting average displacement time (ADT) represents the average time distance between input and output and describes average operation lead time under FIFO. The basic relationship between average lead time and average inventory-output ratio leads to the principle of load limiting. By means of this principle input, output, inventory IX(WIP) and lead time at the workcentres are planned using a variable parameter called 'load limit'. Considering a fixed planning period with a given capacity or planned output, the level of planned lead time requires a certain level of planned inventory and planned input. The planned relation of input and output is defined by the load limit, expressed as a percentage of capacity. Strictly speaking, the load limit is computed as follows: Planning period+Planned lead time Load limit = *100 Planning period The procedure of load limiting begins with measuring the initial load at the beginning of each planning period. Then the permitted incoming load is established according to the load limit so that the planned inventory turns out at the end of the planning period. So far it was assumed that the capacity or planned output is fixed and load limiting decides on what volume of load may flow towards the workcentre within the next planning period. However, if the capacity allows adjustments over the short- term load limiting is used to establish the required capacity for an urgent volume of load flowing towards the workcentres as follows: Planning period+Planned lead time Load limit = *100 Planning period Initial load+Incoming (urgent) load Required capacity *100 Initial load+Incoming (urgent) load Required capacity = *100 Load limit Thus load limiting either rejects load if capacity is fixed or plans capacity if it is flexible. In both cases the desired planned lead time stays the same. The planned lead time is not a rule to be followed by every single operation but more a standard that should be met on an average. At the beginning of each planning period, in general one week, the procedure of load-oriented order release decides on which of the planned orders to be released in the following planning period. The following prerequisites are necessary: 1. Order and operation due dates are scheduled. 2. Required materials are available. 3. Required tools and devices are available.4. Initial loads due to work-in-process are known. 5. Capacities of next planning period are known. The two steps of load-oriented order release are: Step 1: Establish urgent orders Step 2: Release workable orders The first step is necessary to meet the due dates. As the planned orders were scheduled backwards with reliable operation lead times the resulting scheduled release dates, which are still preliminary, are realistic and therefore are used as priorities. By means of the ''time limit", which is a variable parameter and longer than the planning period, the number of urgent orders is narrowed down to avoid early order release. Selecting the time limit provides manufacturing control with the option to consider a larger or smaller number of future orders if underload or overload is expected as the case may be. The second step is necessary to guarantee the planned levels of work-in-process and lead times. Beginning with the highest priority order the urgent orders are released as long as the load limit of all workcentres involved are not exceeded. In case of release, the workcentre loads are increased accordingly. Initial and incoming load are calcula ted by means of load conversion. Time phasing and time bucketing are not necessary. Orders that are rejected despite a high priority have a better chance to be released in the next planning period. This preliminary rejection is the only practicable compromise in a bottleneck situation as more urgent orders have already loaded the workcentres involved up to the limit. If manufac turing control has the option of changing the capacities over the short-term, load-oriented order release sets the required flags for appropriate decisions. Having adjusted the capaci ties the second step of the procedure has to be repeated. The procedure can also be used daily or whenever suitable with an overlapping planning period that is long enough to see the bottlenecks as early as possible and to give manufacturing control and production personnel enough time for capacity adjustments. In this case only a smaller volume of the workable orders should be released accordingly. XI
Benzer Tezler
- Hibrit kompozitlerde farklı metal bağlantılarının yapıştırma özelliklerinin belirlenmesi
Determination of adhesive bonding properties of different metals in hybrid composites
HANDE OSMANOĞLU
Yüksek Lisans
Türkçe
2018
Makine Mühendisliğiİstanbul Teknik ÜniversitesiMakine Mühendisliği Ana Bilim Dalı
PROF. DR. MUSTAFA BAKKAL
- Turbojet motoru kompresör çarkının yapısal analizi
Structural analysis of turbojet engine compressor wheel
TALHA ENSAR BAŞ
Yüksek Lisans
Türkçe
2021
Mühendislik Bilimleriİstanbul Teknik ÜniversitesiUçak ve Uzay Mühendisliği Ana Bilim Dalı
PROF. DR. ZAHİT MECİTOĞLU
- Yalın trapezoidal bükümlü çelik saçların kullanma ve taşımı sınır durumları üzerine deneysel bir inceleme
An Experimental research on serviceability and ultimate limit states of simple trapezoidal corrugated steel sheets
HALET ALMILA BÜYÜKTAŞKIN
- Atölye tipi üretim için modüler çizelgeleme ve kontrol paketi
Modular scheduling and control package for job shop production
EMİN GÜNDOĞAR
Doktora
Türkçe
1991
Endüstri ve Endüstri Mühendisliğiİstanbul Teknik ÜniversitesiPROF.DR. MURAT DİNÇMEN
- Eksenel statik çekme yüküne maruz betonarme tekil kazık davranışı ve orijinal bir kazık deney ve imalat yaklaşımı
A novel construction and testing approach for cast in-situ piles subjected to tension/uplift load
ORHAN ESAT İNANIR
Doktora
Türkçe
2024
İnşaat Mühendisliğiİstanbul Teknik Üniversitesiİnşaat Mühendisliği Ana Bilim Dalı
PROF. DR. AYKUT ŞENOL
PROF. DR. MEHMET MUHİT BERİLGEN