This thesis is for the investigation of the design and analysis processes of filament wound composite tubes under combined loading. The problem is studied by using a computational tool based on the Finite Element Method (FEM). Filament wound tubes are modeled as multi layered orthotropic tubes. Several analyses are performed on layered orthotropic tubes by using FEM. Results of the FEM are examined in order to investigate characteristics of filament wound tubes under different combined loading conditions. Winding angle, level of orthotropy and various ratios of the loading conditions were the main concerns of the study. The results of the FEM analysis are discussed for each loading condition. Both pure loading and combined loading analysis results were consistent with the ones mentioned in literature, such as optimum winding angles, optimum loading ratios and optimum level of orthotropy. Modeling parameters, assumptions and source of errors are also discussed. Finally, the required data is obtained for the design of filament wound composite tubes under combined loading.

Friction is one of the important parameters in metal forming processes since it affects metal flow in the die, forming load, strain distribution, tool and die life, surface quality of the product etc. The range of coefficient of friction in different metal forming applications is not well known and the factors affecting variation are ambiguous. Commercially available FEA packages input the coefficient of friction as constant among the whole process which is not a realistic approach. In this study, utility of user-subroutines is integrated into MSC SuperForm v.2004 and MSC Marc v.2003 FEA packages, to apply a variable coefficient of friction depending on the contact interface conditions. Instead of using comparatively simple friction models such as Coulomb, Shear (constant) models, friction models proposed by Wanheim-Bay and Levanov were used to simulate some cold forging operations. The FEA results are compared with the experimental results available in literature for cylinder upsetting. Results show that, large variation on the coefficient of friction is possible depending on the friction model used, the part geometry and the ratio of contact normal pressure to equivalent yield stress. For the ratio of contact normal pressure to equivalent yield stress values above 4, coefficient of friction values are approximately same for both friction models.

The aim of this thesis is to develop a methodology for obtaining an optimum configuration for the aeroballistic range models. In the design of aeroballistic range models, there are mainly three similarity requirements to be matched between the model and the actual munition: external geometry, location of the centre of gravity and the ratio of axial mass moment of inertia to the transverse mass moment of inertia. Furthermore, it is required to have a model with least possible weight, so that the required test velocities can be obtained with minimum chamber pressure and by use of minimum propellant while withstanding the enormous launch accelerations. This defines an optimization problem: to find the optimum model internal configuration and select materials to be used in the model such that the centre of gravity location and the inertia ratio are matched as closely as possible while the model withstands the launch forces and has minimum mass. To solve this problem a design methodology is devised and an optimization code is developed based on this methodology. Length, radius and end location of an optimum cylinder which has to be drilled out from the model are selected as the design variables for the optimization problem. Built&ndash / in functions from the Optimization Toolbox of Matlab&reg / are used in the optimization routine, and also a graphical user interface is designed for easy access to the design variables. The developed code is a very useful tool for the designer, although the results are not meant to be directly applied to the final product, they form the starting points for the detailed design.

In this study, experimental investigation of solid propellant combustion instability using an end burning T-Burner setup is performed. For this purpose, a T-Burner setup is designed, analyzed, constructed and tested with all its sub components. T-Burner setup constructed is mainly composed of a base part, a control panel and the T-Burner itself. Combustion chamber, pressure stabilization mechanism, pressurization system, measurement instruments and data acquisition systems form the T-Burner. Pressure stabilization mechanism is utilized in two different alternatives, first of which is by the use of nitrogen gas and a small surge tank with a cavitating venturi. This is a brand new approach for this kind of system. The second alternative is the use of a choked nozzle for pressure stabilization. Resonance frequencies of the system with the two different pressure stabilization mechanisms are experimentally evaluated. Helmholtz frequency of the T-burner constructed is calculated and no Helmholtz instability is observed in the system. Constructed T-Burner setup is operated for a specific solid propellant. System worked successfully and pressure data are obtained. Pressure data revealed oscillatory behaviour. Decay and growth rates of pressure oscillations are used for the calculation of pressure response of the propellant tested. By the use of this T-Burner comparison of the behavior of different propellants can be performed. It can be used as a test device for measuring quantitatively the response of a burning propellant to unsteady motions.

The only aeroballistic laboratory of Turkey is the Flight Mechanics Laboratory, FML of T&Uuml / BITAK - SAGE. In FML, flight profiles of projectiles are reconstructed using their tear marks on paper sheets, called yaw cards. Tear marks are created on yaw cards as projectiles pass through them. These yaw cards are tightly stretched to metal frames which are positioned normal to the direction of projectile flight path. The use of yaw cards for flight profile reconstruction is a low cost and reliable solution. However, the yaw card method requires a heavy workload for the analysis of tear marks. Yaw cards collected from the frames are fed through an optical scanner and converted to digital images. These digital images are then processed by operators to calculate the projectile&rsquo / s flight position and angles. To automate this manual process, an algorithm is developed by using histogram based segmentation techniques, custom search algorithms, and Radon transform. This algorithm identifies and locates the projectile marks and finds angle of attack, angle of side slip and roll angle at each frame station by conducting the necessary transformations. Using this automated algorithm, a considerable amount of improvement is accomplished in terms of both decreasing the analysis time and increasing the accuracy of flight profile reconstruction.

Indentation is a practical and easy method, therefore, is a preferred method of material characterization. Main aim of this thesis study is to determine anisotropic properties of metals by indentation tests. The basic property of the indenter used in the finite element analyses and experiments is that it is specific to this process. Thesis includes studies on optimization of the indenter geometry, analyses of effects of friction coefficient, multiple indentations, tilting of the indenter and clamping of the specimen on force-displacements curves during indentation by finite element analyses. This study also includes finite element analyses of compression tests where these experiments have been necessary to prove anisotropic behavior of the specimen material. In addition to compression, tension tests are done to have a reference for indentation tests. On the other hand, the upper bound method which is an analytical solution is applied on the assumption of plane strain indentation.

Indentation tests are widely used with simultaneous measurements of indentation depth and force especially for determining material properties. In this study / numerical and experimental investigation of the force-indentation measurements is presented. For indentation tests on anisotropic metals, a novel indenter which is not self similar is used with three transducers to measure the displacements. It is seen that in order to have high repeatability and accuracy at the tests, workpiece and indenter parameters have crucial importance. These parameters in the indentations are analyzed by finite element methods. Ideal dimensions of the workpiece are determined. It is shown that plane strain conditions can only be achieved by embedded indentations. Effect of surface quality and clamping on repeatability are investigated. It is shown that surface treatments have significant effects on the results. Also it is seen that clamping increases the repeatability drastically. Moreover, indentation tests are conducted to verify the results of numerical simulations. Effect of anisotropy on the force-displacement curves is clearly observed.

Fatigue failure of metallic structures operating under dynamic loading is a common occurrence in engineering applications. It is difficult to estimate the response of complicated systems analytically, due to structure&amp / #8217 / s dynamic characteristics and varying loadings. Therefore, experimental, numerical or a combination of both methods are used for fatigue evaluations. Fatigue failure can occur on systems and platforms as well as components to be mounted on the platform. In this thesis, a helicopter&amp / #8217 / s Missile Warning Sensor - Cowling assembly is analyzed. Analytical, numerical and experimental approaches are used wherever necessary to perform stress and fatigue analyses. Operational flight tests are used for obtaining the loading history at the analyzed location by using sensors. Operational vibration profiles are created by synthesizing the data (LMS Mission Synthesis). Numerical fatigue analysis of the assembly is done for determining the natural modes and the critical locations on the assembly by using a finite element model (MSC Fatigue). In addition, numerical multiaxial PSD analysis is performed for relating the experimental results (Ansys). Residual stresses due to riveting are determined (MSC Marc) and included in experimental analysis as mean stresses. Bolt analysis is performed analytically (Hexagon) for keeping the v assembly stresses in safe levels while mounting the experimental prototype to the test fixture. Fatigue tests for determining the accelerated life parameters are done by an electromagnetic shaker and stress data is collected. Afterwards, fatigue test is performed for determining whether the assembly satisfies the required operational life. Resonance test is performed at the frequency in which the critical location is at resonance, since there was no failure observed after fatigue testing. A failure is obtained during resonance test. At the end of the study, an analytical equation is brought up which relates accelerated life test durations with equivalent alternating stresses. Therefore, optimization of the accelerated life test duration can be done, especially in military applications, by avoiding the maximum stress level to reach or exceed the yield limit.

iv Sheet metal forming technologies are challenged especially by the improvements in the automotive industry in the last decades. To fulfill the customer expectations, safety requirements and market competitions, new production technologies have been implemented. This study focuses on the assessment of conventional and new sheet metal forming technologies by performing a systematic analysis. A geometry spectrum consisting of six different circular, elliptic, quad cross-sections are selected for the assessment of conventional deep drawing, hydro-mechanical deep drawing and high-pressure sheet metal forming. Within each cross-section, three different equivalent drawing ratios are used as a variant. More than 200 numerical experiments are performed to predict the forming limits of three competing processes. St14 stainless steel is used as the material throughout the assessment study. The deformation behavior is described by an elasto-plastic material model and all numerical simulations are carried out by using dynamic-explicit commercial The process validation is done by interpreting the strain results of numerical experiment. Therefore, the reliability of predictions in the assessment study highly depends on the quality of simulations. The precision of numerical experiments are verified by comparing to NUMISHEET benchmarks, analytical formulation, and experiments to increase the assets of the assessment study. The analyses revealed that depending on the workpiece geometry and dimensional properties certain processes are more preferable for obtaining satisfactory products. The process limits for each process are established based on the analyzed crosssections of the spectrum. This data is expected to be useful for predicting the formability limits and for selecting the appropriate production process according to a given workpiece geometry.Dynamic-explicit FEM, Deep drawing, Hydroforming, Forming limits, Process evaluation

The globalization of the markets and the worldwide competition forces the SMEs to implement new technologies and organize themselves using new concepts in order to maintain their competitivity. This type of temporary alliance is called as Virtual Enterprise (VE). SMEs seem to be appropriate units for building this type of temporary alliances when their properties (such as flexibility, adaptability and agility) are taken into account. This study is concerned with the simulation modeling of shop floor activities for SMEs in VEs. Analyzing the SMEs with their current and new work load over the existing one by using simulation tool may help the VE management unit to see the most appropriate SMEs for the projects. Because of mentioned advantages, this thesis will test whether the simulation tool will or will not be used in the selection of the VE partner. The simulation methodology for modeling shop floor activities of SMEs was developed by using ARENA simulation tool in this thesis. A hydraulic cylinder company was selected for pilot application. Manufacturing of twelve basic hydraulic cylinders was studied in the developed model. Four different queue rules were applied to the developed model to optimize the system efficiency. By analyzing the output statistical results of ARENA which were obtained with the usage of the input variables of SME (such as resource capacities, process times, setup and remove times of parts, variables wrt to workers, etc.) best manufacturing policy for pilot SME was able to be found. To see the response of the system under different circumstances, grinding and hardening operations for a drive shaft manufacturing were assigned to the company. This new job was applied on two models (Base Model and Optimized System Model) and the statistical results of each were examined.