Search

Taşıt tasarımında son yıllarda artan müşteri memnuniyeti ile birlikte üretimin minimum maliyette ve maksimum kalitede gerçekleştirilmesi beklentisinin karşılanabilmesi açısından özellikle yeni tasarımların oluşturulmasında ve ortaya çıkan problemlerin çözümünde optimizasyon yöntemlerinin kullanılma gerekliliği oldukça artmıştır. Tasarımı gerçekleştiren veya tasarım sürecinde oluşan problemleri çözmeye çalışan araştırmacı karşılaştığı optimizasyon problemlerinin çözümünde güvenilir ve etkin bir şekilde kullanabileceği optimizasyon algoritmasına gereksinim duymaktadır. Bu tez çalışmasında kullanışlı, global, gerçek optimum değere yakınsamada güvenilir olan ve aynı zamanda optimizasyon sürecinin başlangıcından bitişine kadar çok az hesaplama zamanı tüketen yeni bir algoritma geliştirmek ve bu algoritmanın taşıt tasarım optimizasyon problemlerinin çözümünde etkin bir şekilde kullanılması amaçlanmıştır.Bu çalışmada evrimsel algoritmalar arasında yer alan global, güvenilir, kullanışlı ve birçok test probleminden başarılı bir şekilde geçen diferansiyel gelişim algoritması ele alınmış ve popülasyon içindeki en iyi vektörleri fark vektörleri olarak kullanan yeni bir mutasyon stratejisi geliştirilerek yeni geliştirilen algoritmada (DEBVs) kullanılmıştır. Geliştirilen DEBVs algoritması ile literatürde mevcut bulunan ve yeni geliştirilen birçok algoritmanın test edilmesinde sıklıkla kullanılan kısıt içeren test problemleri çözdürülmüş ve diğer yöntemlere göre daha düşük sayıda hesaplama ile daha iyi sonuçlar elde edilmiştir. Böylece geliştirilen yeni algoritmanın doğruluğu mevcut sonuçlardan daha iyi sonuçlara daha hızlı bir şekilde ulaşma imkânı ile kanıtlanmıştır. Ayrıca önerilen yöntem ile literatürde mevcut tek ve çok-amaçlı çeşitli mühendislik problemleri çözdürülmüş ve mevcut sonuçlardan daha iyi sonuçlar elde edilerek geliştirilen yöntemin doğruluğu pekiştirilmiştir.Taşıt tasarımında üç farklı problem (sürüş konfor optimizasyonu, optimum sac kalıp tasarımı ve optimum motor askı takoz tasarımı) ele alınmış ve geliştirilen DEBVs algoritması ile çözülmüştür. Geliştirilen DEBVs algoritması ile diğer evrimsel algoritmalara ve klasik optimizasyon yöntemlerine göre çok daha güvenilir, global ve etkin sonuçlar elde edilmiştir. Geliştirilen yöntemin taşıt tasarım sürecinde tasarımcıya yardımcı bir araç olarak kullanılması zaman, maliyet, kalite, hız ve kolaylık açılarından büyük avantajlar sağlaması beklenmektedir. In recent years, expectation of carrying out the production in minimum cost and maximum quality has expanded with the raising customer satisfaction. In order to meet this expectation, the necessity of using optimization methods has increased especially in making brand new designs and solving problems appeared. The researcher who generates the design or tries to solve the problems which appear during the design process needs a reliable and efficient optimization algorithm when solving the optimization problems appeared. In this thesis study, it is aimed to develop a new algorithm that is user-friendly, global, reliable in converging the real optimum value, and also consuming as short time as possible from the beginning to the end of the optimization process.In this study differential evolution which was among the best evolutionary algorithms with having global, robust and useful properties and which was tested successfully from many test problems was handled and a new mutation strategy which uses the best vectors in the population as differential vectors was developed and used in the new developed algorithm (DEBVs). Constraint test problems which were frequently used as a test tool for many new algorithms were solved with the developed DEBVs algorithm and better results with less function evaluation numbers were handled when comparing the results of other algorithms. In this way the accuracy of the developed algorithm was demonstrated with reaching better solutions faster. Also various single and multi-objective engineering problems available in the literature were solved by proposed algorithm and the accuracy of the algorithm was reinforced with obtaining better results.In the vehicle design stage three different problem (vehicle ride comfort optimization, optimum sheet metal die design and optimum engine mount design) were handled and solved with developed DEBVs algorithm. More robust, fast and effective results were handled according to other classical optimization methods and evolutionary algorithms. The developed algorithm can be very helpful as an assistant tool for engineers during vehicle design and manufacturing process in terms of time, cost, quality and convenience. Yüksek Öğretim Kurumu (YÖK) TOFAŞTürk Otomobil Fabrikası A.Ş. Bayrak Plastik

ÖZET Makine tasarım optimizasyon problemleri, genellikle doğrusal olmayan yapıda, kısıtların yer aldığı ve çok sayıda tasarım değişkeninden oluşan çok amaçlı modeller olduğu için çoğu zaman analitik yöntemlerle çözülememektedir. Bu tip problemlerin çözümünde sayısal çözüm yöntemleri kullanılmakta ve gerekirse model üzerinde dönüşümler uygulanmaktadır. Problemin kısıt içermeyen bir yapıya veya doğrusal modele dönüştürülmesi sıkça kullanılan yaklaşımlardır. Son yıllarda, tasarım problemlerinin optimizasyonuna yönelik çalışmalar önem kazanmıştır. Birçok araştırmacı değişik yöntemler geliştirerek daha etkin optimumu bulmaya yönelik yaklaşımlar geliştirmeye çalışmaktadır. Bu çalışmada, literatürde mevcut makine tasarım test problemleri ele alınmıştır. Geliştirilen genetik algoritma tabanlı yaklaşım ile elde edilen sonuçlar literatürde yer alan sonuçlarla karşılaştırılmıştır. Ayrıca ele alınan makine tasarım problemleri için uygun genetik parametreler araştırılmıştır. Bu çalışmada, karmaşık matematiksel modeller içeren çok amaçlı problemlerin klasik yöntemlerle çözümünde karşılaşılan zorlukları gidermek için şekil optimizasyonuyla entegre çalışan genetik tabanlı çoklu arama yaklaşımının geliştirilmesi ve tasarım problemlerine uygulanması işlemi gerçekleştirilmeye çalışılmıştır. İlk kısımlarda genetik algoritmalar üzerinde durulmuş ve genetik algoritma kavramı açıklanmaya çalışılmıştır. Daha sonra şekil optimizasyonu konusu üzerinde durulmuş ve klasik şekil optimizasyon işleminin nasıl gerçekleştirildiği açıklanmıştır. ANSYS programında şekil optimizasyon işleminin nasıl gerçekleştiği açıklandıktan sonra şekil optimizasyonu ile entegre çalışan genetik algoritma tabanlı bir yaklaşım üzerinde çalışılmıştır. Son kısımda örnek makine tasarım problemleri ele alınarak bu problemlerin genetik algoritmalarla çözümü sonucu elde edilen sonuçlar, performans kriterleri doğrultusunda literatürde yapılan diğer çalışmalarla karşılaştırılmıştır. Performans kriterleri bağlamında elde edilen kazanımlarınİİ literatürde mevcut sonuçlarla karşılaştırmaları yaklaşımın geçerliliği konusunda fikir verecektir. Ayrıca son kısımda, şekil optimizasyonuyla entegre çalışan genetik algoritma tabanlı yaklaşımın taşıt elemanlarının optimum tasarımında etkin olarak kullanılabileceğini araştırmak amacıyla ticari araçların sürgülü kapılarının belli bir eksende kızak üzerinde kayarak açılıp kapanması sağlayan taşıt sürgü kolunun şekil optimizasyonu çalışması yapılmıştır. Tasarım modeline genetik algoritma tabanlı şekil optimizasyonu uygulanarak taşıt sürgü kolunun optimum boyutlarına ulaşılmıştır. İİİ ABSTRACT Machine design optimization problems usually can not be solved with analytical methods because of having multi objective functions, large numbers of design variables and constraint functions. In the solution process of these problems numerical solution methods are used and if necessary transformations on the model are applied. Transforming of the problem to the unconstraint form or to the linear model approaches are frequently used approaches. In recent years, design optimization problem studies have been gained great importance. Many researchers have been worked to develop different approaches which have more effective optimum solutions. In this study, particularly product design test problems available in literature have been taken into account. The solutions obtained from the developed genetic algorithm based approach have been compared with the solutions available in literature. Besides, suitable genetic parameters have been investigated for the product design problem which examined in this study. In this study, the processes of developing the shape optimization approach which works with multi-search based genetic algorithm and applying to the product design problem have been carried out for eliminating the difficulties which are come across to the solutions of multi-objective problems in the complex mathematical models. In the first stage of this work, genetic algorithms have been discussed and have been tried to explain them. Then classic shape optimization process has been tried to explain. After the shape optimization technique works with ANSYS programme was explained, it has been studied on the approach that works with the genetic algorithm based shape optimization. In the last stages of this work, the solutions of the sample product design problems solved by other researchers in literature have been compared with performance criterions. In the view of performance criterions, the obtained advantages can be given an idea to the researchers when compared with the solutions available in the literature.IV In addition, to show the ability of using the shape optimization approach which works with genetic algorithms effectively in the optimum design of vehicle components, the study of shape optimization of the vehicle harrow arm which provides doors to open and close along a certain axis on the sledge has been done. Optimum dimensions of the vehicle harrow arm has been obtained, applying genetic algorithm based shape optimization to the design model. 95

Bu çalışma, 07-08 Haziran 2010 tarihleri arasında Bursa[Türkiye]’da düzenlenen 5. Automotive Technology Conference (OTEKON)’da bildiri olarak sunulmuştur. A greater need to enhance comfort characteristics during vehicle design process has recently forced the manufacturers to develop simulation-based approaches. In this study, a simulation-based model of a full-car suspension system is proposed to predict the ride comfort. A simulation model was created for calculating ride comfort effectively. This simulation uses seat-back, seat-surface, and feet acceleration values collected from four different road vehicles which were run on six different roads. Parameters which effect ride comfort were also investigated. Using these parameters, a simulation-based model of a full-car suspension system including engine and seat is created for predicting the ride comfort. The correlation between the results of physical tests and the simulation is very promising. It was found that the effect of an engine has a substantial influence on the ride comfort. To find the optimum values of each parameter, an optimization process was executed properly and added in the model. Using this model, the best ride comfort values were computed without the need of physical prototypes. The developed algorithm can be very helpful as an assistant tool for engineers during vehicle design and manufacturing process. TOFAŞ-PLATFORM R&D A.Ş

This paper presents a framework for re-designing a failed automotive component subjected to cyclic loading. In the automotive industry, some parts fail before completing their design life. These failed components have to be re-designed using modern optimisation tools. In this paper, a failed clutch fork was completely re-designed using topology and the shape optimisation approach. Stress-life fatigue analysis was conducted to correlate the crack location between the failed component and the simulation model. A new design proposal was determined with the topology optimisation approach, and then design optimisation by response surface methodology was effectively used to improve the new clutch fork design. The mass reduction obtained was 24%. A maximum stress reduction of 9% was achieved, and the rigidity was improved up to 37% in comparison to the original clutch fork.

The engine mount systems have a significant effect with respect to reducing the vehicle development time and cost regarding NVH characteristics and shortcomings not solved yet. There is still a need to enhance the performance of engine mount systems for vehicle ride comfort. This research presents a simulation-based approach to design optimised engine mount components to isolate the road- and engine-induced vibrations without the need for physical prototypes. Experimental tests are performed to define hyperelastic and viscoelastic models for non-linear finite element simulations of rubber components. After the correlation between the results of simulations and the product tests, a virtual engine mount model is presented to design optimum engine mount systems with the desired static and dynamic characteristics. The stiffness curves of different shapes of designs are investigated, and the best curve is chosen as an optimum design solution using the design of experiments approach.

Although genetic algorithm and multi-objective optimization techniques are widely used to solve problems in the design and manufacturing area, further improvements are required to develop more efficient techniques regarding multi-objective optimization problems. The main goal of the present research is to further develop and strengthen the genetic algorithm based multi-objective optimization approach to generate real-world design solutions in the automotive industry. In this research, a new hybrid approach based on Taguchi's method and a genetic algorithm is presented to achieve better Pareto-optimal set solutions for multi-objective design optimization problems. In addition, fatigue damage and life are also considered to evaluate the results of the design optimization process. The validity and efficiency of the proposed approach are evaluated and illustrated with test problems taken from the literature. It is then applied to a vehicle component taken from the automotive industry.

The goal of this research is to describe how simulation and topology design can be used as a die design frame tool to introduce optimal die design structures with the desired maximum rigidity and minimum deformations using response surface methodology and differential evolution algorithms. In the simulation process, not only die deflection, but also press table deflection is taken into account in order to achieve more realistic results. The validation of the present approach is evaluated by a comparison of test and simulation results. In the experimental test, acceleration and strain data were measured from critical points of the die structure in order to obtain the maximum displacement and stress values. The optimal shape parameters for the die structure were obtained using response surface methodology and differential evolution as a fundamental optimisation technique. Significant results were obtained: The mass was reduced approximately about 24 %, and the current maximum stress decreased approximately about 72 %. By using this methodology in the design stage of die and sheet metal stamping, major improvements can be made to the vehicle development process, such as reducing the weight and the cost of die, reducing the labour costs during the tryout process and reducing the environmental damage and CO2 emissions by reducing the amount of cast iron.