FBE- Makine Mühendisliği Lisansüstü Programı - Doktora

Bu koleksiyon için kalıcı URI

Gözat

Son Başvurular

Şimdi gösteriliyor 1 - 5 / 165
  • Öge
    Modelling formation of fuel-in-oil due to post injections in diesel applications
    (Lisansüstü Eğitim Enstitüsü, 2021) Gönül, Murat ; Kutlar, Osman A ; 709916 ; Makine Mühendisliği
    Post injections are frequently used for different purposes in diesel engines. However, post injections with specific configurations can cause oil dilution due to fuel percolating through cylinder walls. The fuel addition to the oil can cause deterioration in engine oil properties, which may result in engine failure. Therefore, the fuel concentration in oil needs to be monitored, and configurations of post injections are required to be optimized against oil dilution. Usage of post-injection can be grouped under three different purposes; regeneration of Diesel Particulate Filter (DPF), desulfation of Selective Catalytic Reaction (SCR) and heat management strategies. As a consequence of the increasing necessity of post-injection in three explained conditions, there will be an enormous increase in usage of post-injection strategies with new legislation, and oil dilution caused by post injections will become an emerging problem for all truck manufacturers and a more complex problem to contain for light and medium-duty vehicle applications. There are two main mechanisms which are controlling the rate of oil dilution due to post-injection. The first mechanism is that fuel going into the oil pan through cylinder walls due to fuel contact with the wall, and the second mechanism is that fuel vapour in the oil pan might be recovered if it reaches enough vapour pressure. The first mechanism is affected by engine design parameters such as fuel injectors, ring pack and bowl design and calibration strategies such as post-injection timing and quantity and other calibration parameters. The second mechanism is mainly controlled by system design parameters such as oil operating temperature and oil volume. There is also the influence of engine speed on oil recovery. The oil dilution rate can be defined as a balance between the first and second mechanisms. In this study, the effect of calibration parameters on the first mechanism is focused, so post-injection strategies were activated for all times during steady and transient tests. The impact of the second mechanism on oil dilution measurements is normalized based on test data. A recovery test without post injections is conducted to analyze the effect of the second mechanism. Oil properties such as viscosity with increasing fuel in oil concentration can deteriorate. As a result of viscosity change, oil dilution has specific effects on engine oil pumps and other engine components lubricated with oil. Fuel in oil concentration limit is lower in heavy-duty vehicles than other applications since in-service conformity is exceptionally high in heavy-duty vehicles. Three locations are sensitive against excessive oil dilution: crank bearing, cylinder liner and turbine-compressor shaft bearing. Besides these critical regions, excessive fuel in oil can lead to the fuel-oil mixture reaching intake via blow-by. In such a case, the engine may face over speed risk. The engine can handle the deterioration in oil properties to a certain level. There is a trade-off between oil dilution rate, maximum oil deterioration that can engine bear and oil interval demand. Several different measurement techniques can be used to analyze fuel concentration in engine oil. These measurement methods can be divided into three groups. The first group of measurement methods includes comparing total hydrocarbon input and output at the system level. The second group aims to estimate the concentration of fuel in the oil sample by comparing the physical condition of the sample with the fresh sample. And finally, the third group can be sorted as measurement methods aiming to measure fuel concentration in oil samples directly. Gas Chromatography and Flame Ionized Detector (GC&FID) was used to measure fuel concentration in samples in this study. This method is under the third group of the measurement categories. The third measurement group offers the best practice in terms of accuracy among all three above mentioned options since the third group aims to measure fuel concentration directly. However, GC&FID requires 4-6 hours in total for testing and measurement. So online optimization based on the most accurate measurement method is not possible. Therefore, the modelling studies can compensate for the disadvantage of the third group of measurement methods. A series of different tests are conducted. During these tests, 208 oil samples were analyzed in GC&FID. Firstly, steady-state testing in 56 different operating conditions is completed with 168 oil samples. As an intermediate step, the recovery of Fuel in oil is characterized in the second test with 10 oil samples. In the final testing activity, 10 set of transient cycles in 5 different traces are completed with 30 additional oil samples. By these tests, the characterization of oil dilution mechanisms is aimed to be defined. Possible input sets are determined based on a literature search and test data. The necessary input parameters are calculated in the Matlab Simulink environment. The correlation investigation is done with Python libraries. Finally, four modelling methods are applied, and the results are compared. The best input set and modelling approach indicated based on model results over transient test data.
  • Öge
    Aort kapağı kan akımının katı-sıvı etkileşim yöntemiyle sayısal incelenmesi
    (Lisansüstü Eğitim Enstitüsü, 2022) Amindari, Armin ; Kırkköprü, Kadir ; 712131 ; Makine Mühendisliği
    Günümüzde, insan ölümlerinin başlıca sebebi kalp hastalıkları olup bu hastalıkların çoğu ise kapak kusurları ile ilgilidir. Bu kusurların tedavisinde uygulanacak tedavinin kararlaştırılmasında, kusurun karakterize edilmesi ve hastalığın ciddiyetinin belirlenmesi hayati önem taşımaktadır. Günümüzde bu teşhisler, tıbbi görüntüleme temellidir ve teşhisi yapan hekimin tecrübesine bağlı olarak özel biçimde yapılmaktadır. Bununla birlikte, ileri seviyede hastalıklarda aort kapaklarının mekanik ve prostetik kapaklarla değişimi sağlanmaktadır. Yapay kapakların tasarımı ve kullanılan malzemenin, gerçek aort kapak fonksiyonunu yerine getirecek şekilde belirlenmesi için, aort kapak dinamiği ve malzemesinin mekanik özelliklerinin doğru bir şekilde anlaşılması gerekmektedir. Bu tez, aort kapağının elastisitesi ve geçen kan akışının sayısal ortamda modellenmesi ile ilgilidir. Çalışma kapsamında, aort kapağında gerçekleşen kan akışı ve kapak elastodinamiğinin sayısal olarak modellenmesi için bir metodoloji oluşturulması hedeflenmiştir. Bu metodoloji ile kapak kusurlarının kan akışı üzerinde oluşturduğu etkiler doğru ve hassas şekilde modellenebilecektir. Aort kapak yapısı, kan akışının oluşturduğu dinamik basınçlar altında deforme olarak kalpten bütün vücuda doğru tek yönlü kan akışını sağlamaktadır. Bu sebepten kan akışı ve aort kapak yapısı güçlü bir etkileşim içindedir. Kan akışı ve kapak elastodinamiğini gerçeğe yakın ve doğru bir şekilde modellemek için her iki çözüm alanı olan kan akış alanı ve aort kapağı katı alanının eş zamanlı ve birlikte çözülmesi gerekmektedir. Bu tez çalışması kapsamında, ileri bir mühendislik tekniği olan katı-sıvı etkileşim metotları kullanılarak, kan akışı ve aort yapısal alanlarının eş zamanlı çözülmesi için bir modelleme metodolojisi oluşturulmuştur. İlk aşamada iki boyutlu ideal aort kapak modelleri oluşturularak aort kapak üzerinde meydana gelen kireçlenme probleminin kapak dinamiği üzerindeki etkisi incelenmiştir. Bir sonraki aşamada ise farklı hastalık grupları belirlenerek gerçek ve hastaya özel üç boyutlu katı-sıvı etkileşim modelleri oluşturulmuştur. Sayısal sonuçlar gerçek ölçüm sonuçlarıyla karşılaştırılmış ve metodolojinin hassasiyeti değerlendirilmiştir. Tezin son bölümünde ise aort kapak malzemesinin doğrusal olmayan mekanik özelliklerinin kan akışı ve kapak dinamiği üzerindeki etkisi incelenmiştir. Bu aşamada, üç farklı malzeme modeli kullanılarak farklı katı-sıvı etkileşim kontrol modelleri oluşturulmuştur. Bu şekilde, kapak malzemesinin doğrusal olamayan özelliklerinin kapak stabilizasyonu ve ayrıca kireçlenme riski üzerindeki ektisi detaylı ve sayısal olarak incelenebilmiştir. Yapılan hesaplamalara göre kolajen fiberlerinin kapak kapanış evresindeki titreşimini düşürüp, kapağın kararlılığını artırdığı gözlemlenmiştir. Ayrıca, kolajen fiberlerinin çevresel yöndeki konumlanması, aort kapağın başlangıç pozisyonuna daha hızlı ve rahat şekilde dönmesini sağladığı ve bu şekilde kapak boyunca tek yönlü akışın sağlanmasında önemli rol üstlendiği gözlemlenmiştir. Öte yandan, malzemenin doğrusal olmayan elastisite özelliği sebebiyle kireçlenme problemini ciddi derecede azaldığı ve bu sebepten stenoz riskinin azaldığı sonucuna varılmıştır.
  • Öge
    High precision motion control of mechatronic systems in presence of general uncertainties: Application with a heavy-duty parallel robot
    (Lisansüstü Eğitim Enstitüsü, 2022) Sancak, Kamil Vedat ; Bayraktaroğlu, Zeki Yağız ; 723041 ; Makine Mühendisliği
    This thesis study is focused on precise control of heavy duty 6 DoF 5.500 kg payload Stewart_Gough platform which is designed and manifactured at Altınay company. Experimental work aiming to identify frictions in linear actuators of Stewart_Gough platform in order to improve tracking accuracy performances brought out significant amount of nonlinear friction. The parallel mechanism consists of six linear actuators with identical kinematic chains, coupling the moving upper and the fixed lower platforms through universal joints. The linear module actuaters of SGP contain a ball-screw mechanism mounted to frameless AC servo motors which are equipped with failsafe electromagnetic brakes and high resolution absolute encoders. Thes linear modules are subjected to gravitational forces that increase the effect and importance of elastic deformations in ball screw mechanisms. An inverse dynamical model derived using the Newton-Euler approach for similar Stewart_Gough platforms includes the dynamics of the linear modules. Since the rotor inertia of frameless AC servo motors can not be neglected due to their magnitudes linear module dynamics is included to the dynamic model of SGP. The mathematical model determines the driving forces acting on the legs according to the dynamic formulation. The LuGre and GMS (Generalized Maxwell Slip) dynamic friction models proposed in related literature are used to calculate friction forces of thelinear model actuators. Using experimental data and dynamical model of GSP, friction model parameters are identified. The system performance is experimentally evaluated as well by comparising of two different dynamic friction models. The identified models offer the possibility of further advanced studies of the platform structure in order to evaluate dynamic behaviors and designing controllers for high tracking accuracy Model-based control and non-model-based control methods are applied with the assistance of Luenberger-like observer and extended state observer (ESO). The Nonlinear PD and nonlinear computed torque control methods enabled to have precise position tracking of 6 DoF SGP.
  • Öge
    Modelling and analyses of damped multi-layered structures
    (Lisansüstü Eğitim Enstitüsü, 2021) Özer, Mehmet Sait ; Şanlıtürk, Kenan Yüce ; 662857 ; Makine Mühendisliği
    Mechanical vibrations may cause several undesired effects on mechanical systems such as noise, durability problems, functional disorders, even fatal damages on the large structures. Therefore, academicians and R&D engineers have been putting great efforts for the investigation of structural vibrations and controlling (or limiting) vibration amplitudes. One of the widely used methods for reducing the vibrations is covering the mechanical system without modifying the original function of the structure. This can be done with layer damping treatments which are known as efficient, lightweight and low-cost. Besides, the mechanical structures can be directly built using the sandwich-type of panels containing damping layers. Although commonly used in automotive industry, aeronautics and aerospace applications, naval and offshore structures, the mechanical behaviour of these layered structures are not well-understood and eventually cannot be accurately modelled. Despite the existence of some finite element (FE) approaches suggested for modelling the multi-layered structures, the majority of these methods are based on undamped applications. Considering damping, in addition to the frequency-dependent material properties, increases the complexity of the problems. The suggested methods in commercial programmes are unwieldy, even for simple cases, the problem size may increase dramatically for composites having a few layers. Alternatively, theories using higher-order complex functions for modelling the deformation scheme along the composite thickness is also acknowledged. However, these approaches bring complexity for practical usage in finite elements. Therefore, there is a need for accurate, practical and cost-effective modelling approaches. Furthermore, a high precision procedure for identification of dynamic properties of commonly used and promising materials in multi-layered structures is also a requirement. This thesis aims to develop modelling methods and identify the elastic and damping properties of the multi-layered structures considered for damping treatments. New analytical, numerical and experimental studies have been performed in order to achieve these objectives. In the scope of this thesis, a comprehensive review for analytical and finite element modelling of the damped multi-layered structures is performed. A family of multi-layered solid elements (in 8-, 12-, and 20-noded configuration) is developed, implemented in in-house FE code and evaluated. Although the 12-node solid element is not widely used in practice as it does not satisfy the geometric isotropy, it has been deliberately involved in this research in order to have higher interpolation function along the thickness direction and decrease the computation time. The proposed solid elements are assessed using some numerical studies of multi-layered structures. Moreover, a new equivalent shell finite element (FE) for modelling damped multi‐layered structures is presented in this thesis. The method used for developing the new FE for such structures is based on the idea that the strain energy of the equivalent single‐layer FE must be equal to the sum of the strain energies of individual layers. The so‐called energy coefficients are defined for this purpose for the extensional, bending and shear deformations of the composite structure. These coefficients are then determined and used as correction multipliers during stacking the elemental matrices of individual layers. Several assumptions for strain and stress distributions are examined. Among those, the ones based on second‐order strain or stress distribution assumption through the composite thickness, are investigated for deriving the shear energy coefficients. The damping capability of the FE developed here arises from using complex Young's modulus to define the material properties of individual layers. The resulting equivalent single‐layer shell element with four nodes has six degrees‐of‐freedom per node. The accuracy, advantages and limitations of the composite FE developed in this work are investigated using experimental as well as theoretical results. In the light of the finding of these investigations, further enhancement in the formulation is made by also utilising a new shear correction factor for the individual layers in the equivalent shell element. Conclusive results for free‐ and constrained-layer structures confirm that the enhanced equivalent shell FE developed in this thesis can be used effectively for the prediction of the modal properties of damped multi‐layered structures. The importance of the shear deformation in the damped multi-layered structures are emphasised, and it is proved that the accurate modelling approach for such structures can only be provided by determining the appropriate shear stress behaviour through the composite thickness. An alternative composite shell FE formulation using existing shear correction factor formulations in the literature is performed. Accordingly, several shear correction formulations are inspected, and their applicability in the proposed FE is investigated. Using a single shear correction factor for the whole composite structure and employing individual shear correction factors for individual layers in the proposed formulation of the composite shell FE are assessed for frequency and loss factor predictions of multi-layered structures for practical applications. The resulting composite FE is validated using the results of the state-of-art 3D solid FEs. Beam, as well as, plate type of structures are analysed, and the results reveal that the proposed composite FE can predict the modal parameters of multi-layered structures with great accuracy. Oberst Beam Method is a commonly used experimental procedure for identification of the frequency-dependent properties of materials. The use of a non-contact electromagnetic excitation system is highly recommended in the literature using the Oberst Beam Method. However, it is not possible to test a specimen made of non-magnetic material using the Oberst beam test rig, comprising of an electromagnetic exciter, unless the specimen is modified using some magnetic particles or small discs made of a ferromagnetic material. Although doing so makes it possible to perform the test, this results in an undesirable modification to the test specimen, leading to unquantified levels of errors in the estimated material properties. An approach for eliminating the adverse effects of such mass modification to the test specimen, which also allows subsequent removal of the electromagnetic stiffening effects produced by the electromagnetic exciter, is introduced in the thesis. The proposed method is validated using both FE simulations and experimental data. Results confirm that the proposed method for the removal of the adverse effects of mass modification, combined with the subsequent removal of the electromagnetic stiffening effects, is very effective, making it possible to determine the material properties of non-magnetic materials with high accuracy.
  • Öge
    Constitutive failure modelling and analysis of steel wire rope structures subjected to impact loading
    (Lisansüstü Eğitim Enstitüsü, 2021) Candaş, Adem ; İmrak, Cevat Erdem ; 685723 ; Makine Mühendisliği
    Dynamic fracture is an important research topic in the science of fracture mechanics. The crack initiation and propagation is a problem that has received considerable attention because of its technical consequences. In case of impact loading and related failure mechanism both in macro- and micro-level should be carefully investigated. An impact load may adversely affect the system's operation, especially in cases where brittle structural elements are subjected to this load. Besides, brittle materials have advantages such as hardness and wear resistance, their deficiencies in terms of toughness and brittleness significantly restrain their usage in practice. This is the main reason that the problem of crack propagation at both macro- and micro levels is a problem of frequent discussion in the recent literature. The dynamic fracture behaviour of brittle materials that contain micro-level cracks were examined when material subjected to impact loading. The investigation on the effect of micro-cracks on the crack propagation was carried out in the first step. The macro-crack initiates from notch tips in the Kalthoff– Winkler experiment, a classical impact problem. A micro-crack cluster was designed to decelerate this crack propagation. To define pre-defined micro-cracks in three-dimensional space, a two-dimensional micro-crack plane definition was proposed in the bond-based Peridynamics (PD). PD is a non-local form of classical continuum theory. Randomly distributed micro-cracks with different number densities in a constant area and number in expending area models were examined to monitor the toughening of the material. The velocities of macro-crack propagation and the time required for completing fractures were considered in several pre-defined micro-cracks cases. It has been observed that toughening mechanism only initiates by exceeding a certain number of micro-cracks; therefore, it can be said that there is a positive correlation between the density of pre-defined micro-cracks and macro-crack propagation rate and, also, toughening mechanism. The classical impact problem was explained in details and then, wire rope structures that one of the most important elements in material handling were examined. The complexity of material handling area needs to manage many different machine and equipment. Therefore, accidents can inevitably occur in these areas. However, in general, there are further factors that affect the failure of ropes in an accident. Wire ropes are designed for static axial loading owing to its nature of structure. In that manner, an impact load can result in an undetermined mechanical response of the rope. Moreover, corrosion, insufficient lubrication, porosities in the working area, and wear can decrease the strength of wire ropes. The lifetime prediction of a rope system is a very complicated task because of the complex structure of ropes and different loading conditions. However, to determine the reliability of material handling require more specific information about each element. With the help of proposed methods and findings in Kalthoff-Winkler problem, a theoretical scheme of analysing cable systems and wire ropes subjected to impact load with peridynamics was handled. Numerical studies were carried out, and the simulation parameters were discussed. It can be estimated that the failure of a wire in a strand does not affect its neighbours, because crack propagation in a wire cross-section ends at the outer surface of that wire. The resulting stress concentration that will cause crack propagation in adjacent wires is not observed. However, of course, there is some local transition of the load should be taken into account because of inter-contact states between wires. With regard to this, the work presented in this study can be extended to examine the inter-contact interaction between wires. As a consequence, the effect of micro-cracks on a macro-crack propagation was investigated in Kalthoff-Winkler problem. The one most obvious finding to emerge from this study is that the less than a certain number of randomly located micro-cracks around the crack tip has no positive effect on fracture toughening mechanism. Nevertheless, adding more amounts of pre-defined micro-cracks in the same region can decrease crack propagation velocity and significantly increase the toughness. The second major finding was that there needs a certain number of micro-cracks for occurring of toughening effect. This study has found that an insufficient number of micro-cracks cannot decelerate the propagation of cracks. A certain number of micro-cracks should be placed in the body in order to obtain the toughening effect. In general, therefore, it seems that the density of micro-cracks in a constant area and the number of micro-cracks in expending areas are significant parameters on toughening mechanism in a brittle material subjected to impact load. The findings of this investigation complement those of earlier studies. These findings support the PD's competence as an alternative to classical continuum mechanics for modelling of fracture and thus, designing of strengthen geometries. Although the study has successfully demonstrated that crack propagation and fracture characteristics, it has certain limitations in terms of properties of micro-cracks. Dynamic crack propagation and failure in a wire cross section were studied. The programs used in the study were evaluated. Developed scripts that can be useful for further researchers were provided. Compared with average velocities in m-convergence tests, velocities in δ-convergence tests differentiate much more. It can be inferred that the minimum value of m (as an indicator of material points within a horizon) should be 3 for the models with given parameters and dimensions. The average velocities of m = 3,4, and 5 models are very close to each other. The data for δ = 0.00450 test should be considered an outlier because the crack did not propagate in contrast to other models. This result indicated that the horizon value, δ = 0.00450 is not applicable for the model with given parameters. With the understanding of wave progression and mode transition relation, the model δ = 0.0015 can be considered as a better parameter choice for the given model. The Mode I crack opening transition in the reference model indicates a routing of the crack in horizontal direction. These findings are thought of as a basis for the simulation of fracture mechanisms in wire ropes with PD.