<p>Multi-axis milling simulation of flexible parts is a highly interdisciplinary topic. It includes theories and methods in milling mechanics, structural mechanics and geometric simulation. In order to determine the acceptable cutting conditions, process planners faced with the flexible tools and thin walled structures either use very conservative cutting conditions, or in the case of high volume productions, trial and error methods. In either case, a final air cut is usually required to bring the machined surface within the tolerance. Although practicala, the above approach results in low productivity. Development of an analysis software capable of predicting the cutting forces deflections and surface errors in multi-axis milling of the thin wall structures is therefore highly desirable. By eliminating the need for trial and error methods a dynamic multi-axis process simulator can assist process planners to optimize the cutting condition and tool path for the least geometrical errors. This thesis describes such a simulator, implemented as an integrated CAD/CAM architecture consisting of both geometric and physical simulators. A static cutting force model is introduced which accounts for the rotation of the a and b axes in multi-axis side milling operations. The proposed formulation is based on a vectorial approach which reduced to a conventional formulation if a three-axis tool path is used. A dynamic multi-axis force model is also developed which is capable of modeling the tool/workpiece structural interaction. The dynamic response of the workpiece to the instantaneous cutting forces is modeled using a special finite element code. In conjunction with the finite element code, a solid based automatic mesh generation algorithm is also developed. The automatic mesh generator frequently updates the finite element mesh as the workpiece geometry changes. The dynamic parameters of the tool, experimentally found at its tip, are extrapolated for the rest of the cutter elements along the tool axis. The tool elemental parameters are then used to model the dynamic response of the cutter. Using the effective deflection of the tool/workpiece system, profile of the machined surface is simulated and its geometric deviation from the design surface is computed. This is highly desirable for side milling of ruled surfaces, where the value added is very high. For the experimental verification of the developed models, a twisted ruled surface was machined using a four-axis milling operation. The thickness of the twisted blade was reduced to 2.0 mm for the case of the dynamic cutting tests. The profile left on the machined surface was compared with the simulation result. In all comparisons a good agreement is seen between the experimental data and simulation results, which verifies the validity of the developed models and techniques.</p> / Doctor of Philosophy (PhD)

Robotic Fixtureless Assembly (RFA) refers to the performance of assembly tasks by robots without the aid of jigs and fixtures. Two control problems encountered in RFA of sheet metal parts assembly are addressed in this thesis. They are the control of vibration when handling the sheet metal parts and the control of the contact state between the parts during assembly. Both 2-D parts and 3-D parts will be considered when the two problems are addressed. A unique sensor, termed a Strain Gauge equipped Finger (SGF), was first developed which can measure the part's vibration during handling and the contact state during assembly. For the vibration control problem, a novel Learning Extremum Control (LEC) algorithm is proposed. Using SGF's mounted on the robot gripper for vibration feedback, the orientation of the part in 3-D space relative to its path is controlled to reduce vibration. Furthermore, as an alternate approach, a feedforward Input Command Shaping Method (ICSM) is employed to reduce the vibration. Experimental results confirmed the effectiveness of both control algorithms. The LEC and the ICSM reduced the vibration amplitude by up to 71% and 78%, respectively. In order to solve the contact state control problem, the contact state measurement problem in 2-D parts assembly is first considered. Measurement methods are developed based on the use of a Force/Moment Sensor (FMS) alone, SGFs alone and the sensor fusion of he FMS and SGFs information. The methods are verified in 160 robotic assembly trials. The success rate for the FMS method increased from 28% to 88% when the sheet thickness was increased from 0.32 mm to 1.88 mm. Under the same conditions, the success rate for the SGF method decreased from 100% to 55%. The sensor fusion method surpassed the best individual performances off the other methods by achieving a 100% success rate for the complete set of sheet thicknesses and contact states tested. After the contact state measurement problem is considered, the contact state control problem is examined. The SGFs a Force/Moment Sensor (FMS) are used to provide feedback about the contact condition between two sheet metal parts. An Integral Contact Controller (ICC) is used to correct angular error between the parts to ensure full contact along the joints for subsequent welding. Experimental results confirmed the effectiveness of the ICC algorithms. For 2-D parts, using the fusion of the FMS and SGF sensory information, the ICC reduced the Yaw joint angular error from 0.5˚ to 0.025˚ within 1.7 seconds. For 3-D parts, the ICC reduced the angular error in both the Pitch and Yaw angles from 0.5˚ to 0.05˚ within 2.4 and 3.6 seconds using the FMS alone and SGFs alone, respectively. Finally, an investigation is performed to determine the effectiveness of applying the Input Command Shaping Method in the contact state control experiments. This method's limitations are discussed. / Doctor of Philosophy (PhD)

<p>Linear Isoparametric elements and elements based on the Constant Strain Triangle are developed and incorporated into a computer program, for two dimensional stress analysis on mini computers. These elements are compared at element level using the Weighted Eigenvalue method and in assemblage using sensitivity analysis. Further, a method of predicting results based on the data generated during the comparisons at the element level is also presented. As an application of the computer program, analysis of a porcelain insulator is also included.</p> / Master of Engineering (ME)

<p>The purpose of this study is to examine the role that Open Architecture Control concepts have in the application of Intelligent Machining Systems. Open Architecture Control is a relatively new field whose original intent was based on the use of "Open System" computer science concepts in the development of integrated manufacturing systems. Various manifestations of Open Architecture Control systems has been published in the open literature however, in many ways the original intent of the idea has been often obfuscated by the sometimes dissimilar interests of the engineering, computer science, academic and commercial realms involved. Intelligent Machining Systems refers to application of sensing, monitoring and control technologies to machining processes with the intent of: - improving the economic performance of machining systems, - controlling the processes involved in a comprehensive manner. Consequently, such systems represent an integration of technologies which individually address only a limited part of the overall potential for economic gain. The use of "higher level" information and control functions is a common element in Intelligent Machining Systems. These may include components of Artificial Intelligence, process planning, and supervisory control technologies. The need for a high level of integration in Intelligent Machining Systems (IMS) has presented a difficulty for their practical application. Open Architecture Control (OAC) address the integration problem directly and, if the proper Open Systems issues are well considered, can enable IMS technology for commercial use. The OAC system developed in this work does address Open Systems concepts, and has been designed with consideration for the needs of IMS's. This has been achieved through the formulation and implementation of an IMS on the designed OAC. Technologies such as process sensing, monitoring and control, in addition to planning, simulation and supervisory control have been tested and verified with the OAC for an application to turning a CNC turret lathe.</p> / Doctor of Philosophy (PhD)

<p>This thesis describes the development of a software system, which integrates the optimization, finite element, expert system and computer aided design techniques to achieve a highly automatic structural design.</p> <p>New knowledge has been developed for the theory of expert systems, including an knowledge acquisition approach by linear programming and machine learning.</p> <p>Several algorithms have been developed to increase the efficiency of the finite element based optimization.</p> / Doctor of Philosophy (PhD)

<p>Increasing demands for high production rates as well as cost reduction have emphasized the potential for the industrial application of hard turning technology during the past few years. Machining instead of grinding hardened steel components reduces the machining sequence, the machining time, and the specific cutting energy. Hard turning is characterized by the generation of high temperatures the formation of saw toothed chips, and the high ratio of thrust to tangential cutting force components. Although a large volume of literature exists on hard turning, the change in machined surface physical properties represents a major challenge. Thus, a better understanding of the cutting mechanism in hard turning is still required. In particular, the chip formation process and the surface integrity of the machined surface are important issues which require further research. In this thesis, a mechanistic model for saw toothed chip formation is presented. This model is based on the concept of crack initiation of the free surface of the workpiece. The model presented explains the mechanism of chip formation. In addition, experimental investigation is conducted in order to study the chip morphology. The effect of process parameters, including edge preparation and tool wear on the chip morphology. The effect of process parameters including edge preparation and tool wear on the chip morphology, is studied using Scanning Electron Microscopy (SEM). The dynamics of chip formation are also investigated. The surface integrity of the machined parts is also investigated. This investigation focusses on residual stresses as well as surface and sub-surface deformation. A three dimensional thermo-elasto-plastic finite element model is developed to predict the machining residual stresses. The effect of flank wear is introduced during the analysis. Although residual stresses have complicated origins and are introduced by many factors, in this model only the thermal and mechanical factors are considered. The finite element analysis demonstrates the significant effect of the heat generated during cutting on the residual stresses. The machined specimens are also examined using x-ray diffraction technique to clarify the effect of different speeds, feeds, and depths of cut as well as different edge preparations on the residual stress distribution beneath the machined surface. A reasonable agreement between the predicted and measured residual stress is obtained. The results obtained demonstrate the possibility of eliminating the existence of high tensile residual stresses in the workpiece surface by selecting the proper cutting conditions. The machined surfaces are examined using SEM to study the effect of indifferent process parameters and edge preparations on the quality of the machined surface. The phenomenon of material side flow is investigated to clarify the mechanism of this phenomenon. The effect of process parameters and edge preparations on sub-surface deformation is also investigated.</p> / Doctor of Philosophy (PhD)

<p>A flow configuration capable of exciting and interacting with the acoustic plane waves modes of a piping system is examined both experimentally and theoretically. The acoustic source is generated by placing two standard geometry orifice plates in the flow. Strong acoustic pressures exceeding 125 dB inside the pipe are generated with the orifice plate separation distance small (<2%) in comparison to the wavelength of the lowest frequency excited. The acoustic source is shown to excite those modes possessing an acoustic pressure node (acoustic velocity anti-node) at or near the source location. The Strouhal number based on mean orifice velocity and orifice plate separation ranges from 0.5 to 1.0. and is sensitive to cavity diameter. Flow visualization photographs examining the fluid mechanics of the phenomenon are provided. The photographs reveal the presence of an oscillating shear layer near the upstream orifice plate, and subsequent roll up into a large scale vortex and propagation to the downstream orifice plate. The effect of mean turbulence levels at the upstream separation plane on the fluid dynamics and coupled acoustic production is studied.</p> <p>An acoustic model of the piping system is developed using the 4 pole method. The acoustic model, as well as the modelling procedures, are examined and tested in detail. A theoretical model of the coupled fluid/acoustic oscillator is developed by combining published characteristics of separated inviscid sheared flows with the developed acoustic model. The theoretical model predictions compare favorably, both qualitatively, and quantitatively with the experimental results.</p> / Doctor of Philosophy (PhD)

<p>Alignment deformities at a knee joint having unicompartmental osteoarthritis increase the stress level of the affected compartment and the progression of the disease. The biomechanical integrity of the knee joint can be restored through osteotomy surgery of the proximal tibia, and is believed to retard the advancement of the disease.</p> <p>This research attempted to quantify the biomechanics of high tibial osteotomy surgery for correction of a varus knee deformity. Both elements of analytical and experimental modeling were completed. A two dimensional trigonometric model of the lower extremity was formulated that provided a more precise method of analyzing and forecasting the changes to the knee and limb orientations following osteotomy surgery.</p> <p>The experimental work resulted in two novel and original contributions: 1) the design of a direct measurement system for local dynamic contact pressures in the knee joint and 2) the retrofitting of an existing (non-operative) joint simulator. Measurement of the contact area and pressure patterns of the medial compartment was determined for three different knee alignment configurations (neutral, varus, and post-osteotomy). Alterations to the joint alignment increased loading along the intercondylar eminence, by redirecting the load inward, towards the joint centre.</p> / Doctor of Philosophy (PhD)

<p>A new approach for automated tool condition monitoring in machining by using fuzzy neural networks is proposed. The Multiple Principal Component (MPC) fuzzy neural networks are built based on three major components of soft computation, namely fuzzy logic, neural networks, and probability reasoning.</p> <p>The system architecture is a partially connected neural network with fuzzy classification at neurons and fuzzy membership grades for interconnections. Principal component analyses in multiple directions are implemented tor the feature extraction and the "maximum partition". The partitions of the learning samples are based on the distributions of the monitoring indices in the principal component directions. A fuzzy membership function is used to measure uncertainties in the sampled data and to form "soft boundaries" between the classes. A processing clement in the network is connected to others through the fuzzy membership grades and other information available. The partial connections make short training times and short routines in classifications.</p> <p>Three major issues in developing the MPC fuzzy neural networks are supervised classification, unsupervised classification and knowledge updating. The system obtains the knowledge about classifications by learning. The learning samples are obtained from cutting tests performed through a reasonable range of cutting conditions.</p> <p>Several sensors are used for monitoring feature extraction. The signals from different types of sensors at different locations insure that the most significant information about the changes in tool conditions is collected. Metal cutting mechanics are first considered for the sensor selection and the sensor allocation. The measured signals are further analyzed and the monitoring features are extracted. These indices are the inputs for the fuzzy neural networks. The tool conditions considered include sharp tool, tool breakage, and a few selected stages of tool wear. The experimental results in turning and drilling have shown good performance of the proposed monitoring system in these tests.</p> / Doctor of Philosophy (PhD)

<p>The study presented in this thesis aims at fulfilling an industrial need for solution methods suitable for the analysis of complex nonlinear engineering problems. A finite element contact algorithm is developed, implemented and verified. Furthermore, the algorithm is used for the numerical simulation of the hydraulic expansion of tube joints.</p> <p>A variational formulation of the general contact problem is presented where the virtual work principle is adopted to arrive at the discretized equilibrium equations for two generic bodies coming into contact under the effect of the applied loads. The discretized contact constraint equations are derived geometrically from the kinematics of the potential contact surface nodes. A direct engineering approach is used to develop a solution algorithm which is applicable for the analysis of contact problems in general. No assumptions are made regarding the geometry of the contacting bodies, the location and extent of contact and the nature of the external loading. A non-classical bi-Iinear friction model is introduced where micro tangential relative displacements are allowed even under very light normal traction. The bi-linear law may be reduced to the classical Coulomb's friction law as a special case. The algorithm is made capable of handling nonlinear continuum finite elements, i.e. elements with curved sides. The equilibrium equations are solved iteratively to calculate the contact nodal forces which remove any overlap without augmenting the contact constraint equations into the original system of equilibrium equations. Instead, the contact constraint equations are solved as an inner loop in the global nonlinear iteration loop which follows the full/modified Newton-Raphson iterative technique. The developed algorithm is implemented in the in-house general purpose non-linear finite element program INDAP¹ and verified through the solution of illustrative examples covering a wide range of contact problems of static, elastic, elasto-plastic, conformal and non-conformal contact interactions.</p> <p>The developed contact algorithm is used along with two finite element models to simulate the hydraulic expansion of tube joints. A 2-D plane stress model is used to investigate the effects of the different material mechanical properties on the joint strength. A 2-D axisymmetric model is adopted to overcome the limited scope of the plane stress model. The feasibility of ignoring the geometric nonlinearity and the friction interaction are investigated. A 2³ complete factorial numerical experiment is adopted to study the main and interaction effects of the expansion pressure, the initial radial clearance and the coefficient of friction on the residual contact pressure, representing the joint integrity, and the maximum residual tensile stress along the tube inner and outer surfaces, representing the joint quality. A new explanation to the break-off the joint strength when the joint is further expanded beyond a well-defined optimum is suggested. The adequacy of some strength measures is explored. The effects of the initial stresses and the cold-work surface layer along the tube outer surface are investigated.</p> / Doctor of Philosophy (PhD)