Search

The economic application of carbon fiber reinforced plastics in large-scale aerospace structures demands cost-efficient production technologies. In recent years, much progress was achieved in automation engineering, like automated fiber placement (AFP) and automated tape laying (ATL) technologies. In the design process, new methods have been established to incorporate boundary conditions of the production process. Henceforth the optimization process is not only focused on weight-reduction, but on an improved cost/weight ratio. Most research on this topic has been done in the field of conceptual design, as the highest percentage of the later arising manufacturing cost is defined by decisions made in the early design phase. But there is also a potential for reducing production cost in the detailed design phase. In the Composite Design department of the DLR Institute of Composite Structures and Adaptive Systems the detailed design of a wing cover skin section was optimized for the AFP-process. Ply shapes and ramp geometries have been modified to reduce the number of courses needed for ply-layup, and thus to reduce production time. Uncomplete courses with less than all of the available tows, as well as repeated stops and acceleration of the fiber-placement head due to unnecessary tow-cutting processes have been avoided. With these approaches the total layup time was reduced by 3,4% whereas the on-surface time of the fibre placement heads decreased by even 5% compared to the reference design, while structural weight remained constant. The optimization strategies, originally developed for the AFP-process, are also applicable to the ATL-process. The optimized design was analyzed in 3 sections and compared to the reference design for 408 combinations of longitudinal, transversal and shear loads, showing only minor differences in strength and stability.

Die Halbleiterfertigung ist ein äußerst komplexer Wertschöpfungsprozess, der aufgrund eines schnell wechselnden Marktes und der immensen Investitionskosten unter extremen Kostendruck steht. Um konkurrenzfähig zu bleiben, müssen aktuell jährlich ca. 15 % der Fertigungskosten durch eine erhöhte Anlagen- und Fabrikproduktivität eingespart wer-den. Im Rahmen der Arbeit wurden zum ersten Mal grundlegende Untersuchungen zur Modellierung und Simulation sowie der Optimierung von Halbleiterfertigungsprozessen auf Geräte- bzw. Prozesssequenzebene durchgeführt. Um die Anlagen- und Fabrikproduktivität zu erhöhen und damit implizit Fertigungskosten zu reduzieren, wurden bestehende und bereits eingesetzte logistische Strategien auf die Anlagenebene transferiert, neuartige und innovative Optimierungsstrategien auf Geräte- oder Prozesssequenzebene entwickelt und mit Hilfe der diskreten Simulation untersucht. Schwerpunkte waren die Batchbildung bei flexiblen Fertigungen, Mehrkammerprozessanlagen und eine innovative vorausschauende Stichprobenplanung. In flexiblen Fertigungen kommt es häufig zu logistischen Problemen bei Batchprozessen, da es aufgrund von fehlenden, geeigneten Losen teilweise zu großen Wartezeiten und somit eventuell zu Überschreitungen der Kundenliefertermine kommt. Es wurde im Rahmen der Arbeit ein Konzept entwickelt, welches kleinere Batchgrößen und Füllerscheiben nutzt. Hiermit konnten kürzere Durchlaufzeiten durch das Gesamtsystem (80 – 96 %) realisiert werden, ohne dabei den Durchsatz negativ zu beeinflussen. Der Trend in der Halbleiterfertigung geht aber weiter in Richtung Einzelscheibenprozesse. Als Folge dessen sind Mehrkammerprozessanlagen zu unverzichtbaren Werkzeugen bei der Herstellung von Halbleiterbauelementen geworden. Bei den Untersuchungen an Mehrkammerprozessanlagen wurde die Integration von Messtechnik, die optimalen Dimensionierung und generell Möglichkeiten zur Steigerung der Produktivität untersucht. Das bei linearen Fertigungslinien bekannte und breit eingesetzte Konzept der Taktzeitanpassung wurde erstmals zur Anwendung bei Mehrkammerprozessanlagen adaptiert, dabei konnten beim Durchsatz Steigerungen von 15 - 23 % erreicht werden. Auf Prozesssequenzebene wurden Untersuchungen zur Stichprobenplanung durchgeführt. Die entwickelte vorausschauende Stichprobenplanung zeigte ein sehr großes Potenzial zur Kostenreduktion durch Einsparungen von Messungen bei Defektdichtemessketten und zur Senkung des Risikos von nicht kotrollierten Anlagen. Bei den untersuchten Anwendungsfällen konnte die Anzahl der benötigten Messungen um 6 – 50 % gegenüber der bisher eingesetzten Stichprobenstrategie reduziert werden, gleichzeitig wurde dabei das Risiko von längeren Fertigungszeiträumen ohne Kontrollmessungen um bis zu 85 % verringert. Generell konnte gezeigt werden, dass die diskrete Ablaufsimulation ein hervorragend geeignetes Werkzeug zur Untersuchung von Halbleiterfertigungsanlagen, –prozessen und deren Optimierung darstellt. Es können logistische Optimierungsstrategien jeglicher Art damit analysiert und bewertet werden. Semiconductor manufacturing is a rather complex value-added process, which is facing an extreme cost pressure due to fast changing markets combined with very high capital investments. Currently a manufacturing cost reduction of 15 % per year has to be achieved by an increased equipment and fab productivity in order to compete on the global market. Within this work fundamental investigations on modeling and simulation as well as optimization of semiconductor manufacturing processes at equipment and process sequence level have been performed for the first time. In order to increase equipment and fab productivity and to reduce manufacturing costs thereby, known and established logistical strategies were transferred to the equipment level and new and innovative optimization strategies were elaborated and investigated by the application of the discrete event simulation. This work focuses on batching strategies for flexible manufacturing, on cluster tools and on an innovative predictive sampling strategy. For flexible manufacturing, batching strategies became problematic. Due to a lack of compatible lots for the completion of full batches, long waiting times occur and customer delivery dates were exceeded from time to time. A concept using smaller batch sizes and filler wafers to speed up the batching process was developed. With this concept shorter cycle times (80 – 96 %) could be realized without a negative effect on the throughput of the system. Nevertheless, semiconductor manufactures tend to prefer single wafer processes. As a consequence of this, cluster tools became essential for semiconductor manufacturing nowadays. Within this work investigations on the integration of metrology, on optimal hardware configurations and on general optimization possibilities of cluster tools were performed. The established concept of identical process times for subsequent process steps in a linear process flow was adapted to be applicable for cluster tools for the first time. Thereby, a throughput increase of 15 – 23 % could be achieved. On process sequence level investigations regarding sampling strategies were performed. The developed predictive sampling strategy offered a great potential to reduce manufacturing costs by reducing the amount of required metrology operations for defectivity control chains. Furthermore, the risk of uncontrolled manufacturing equipment could be also decreased. For the investigated applications the amount of required metrology operations could be reduced by 6 – 50 % compared to the sampling strategies, which are currently used in semiconductor manufacturing. At the same time the risk of uncontrolled process operations could be reduced by up to 85 %. In general, it could be demonstrated, that the discrete event simulation is an excellent option to analyze the performance of semiconductor manufacturing equipment and processes and to investigate optimization possibilities. All kind of optimization strategies could be evaluated.

Experience gleaned in the field of product development shows that the early development phase is the phase during which product developers are most apt to be plagued by an acute sense of insecurity. Many of the product’s characteristics have yet to be determined. The results of the work that has been carried out are at this stage not accurate enough, incomplete and sometimes even flawed. The product developer is nonetheless obliged to make conceptual decisions on the basis of these criteria and to define the product’s central characteristics. These decisions not only determine fundamental product characteristics but also have a direct affect on the product development time and the product development costs. If the product that is to be developed has a high degree of novelty, then there is only a limited number of reference points available which can be used to perform a cost calculation using established procedures. This paper presents a method which allows products with a high degree of novelty to be assessed at a very early stage of their development with regard to their manufacturing costs. This method precedes the actual cost calculation. It closes the gap to the calculation process which requires established functions and solution principles in order for the costs to be assessed. To achieve this end, this paper deals with the development of a methodical and systematic approach to cost prediction using Bayes networks. The purpose of a Bayes network is to predict and diagnose factors or events, which because of existing information are either not obviously apparent or too complex to determine. Bayes networks are a special form of theoretical probability models. Following a comprehensive introduction into the fundamental principles of Bayes networks, this paper goes on to present guidelines for the generation and implementation of own networks. Tools that are needed for each of the necessary steps that are connected with generating and working with a network are also presented. This method is not only excellently suited for classic elements of engineering, but also for other fields and sectors like mechatronics or electrical engineering. In addition, universal network structures are developed which can be used for almost all product domains thereby making it easier to set up an own network. A detailled example is used to illustrate the approach to the solution of this problem.

High-quality manufacturing processes are fundamental to be successful on global markets. Especially the German industry relays on its reputation for high-quality products with top performance in combination with high reliability. To ensure and further improve the manufacturing quality, this paper evaluates the potential to adapt the Failure Process Matrix (FPM) to assembly lines with small quantities. The FPM was originally developed to reduce production failure and improve assembly quality of high-volume engine assembly lines in an automobile context. To adapt the methodology to small-quantity assembly lines, the first step was the analysis and evaluation of typical variations. Based on the findings, requirements for the adaptation of the methodology and the supporting software tool were developed. The adaptations include changes to the methodology especially in regards to the gathering of manufacturing data on the shop floor. Also, a new spreadsheet based software tool has been developed to fit the processes of small-quantity assembly lines. Furthermore, new key performance indicators to visualize und understand assembly quality throughout the whole assembly line, were included in the software tool. The adapted methodology with the matching software tool were successfully utilized to optimize the assembly quality of two small-quantity assembly lines for complex industrial large diesel engines used for various fields of application. The case study proved the potential of the FPM for a holistic understanding and improvement of assembly quality of small-quantity assembly lines with complex products.