<p>In machining, the ability to automatically generate an optimum process plan is an essential step toward achieving automation, higher productivity, and better accuracy. These requirements are particularly emphasized in die and mold manufacturing, where complex tool and workpiece geometries involved make generation of the process plan a difficult task. High die production costs, narrow tolerance requirements, and the continuous demand for new components make process planning and NC code generation very complex and error-prone tasks. The current research need is to develop a system that is based on a simulation of the actual machining process, rather than simple geometric verification. Such a machining process simulator is needed to respond to the current need to enhance CAD/CAM technology with a machining process simulation that accounts for process mechanics and dynamics. A major impediment to implementing these systems is the lack of a general and accurate method for extracting the required geometric information. In this thesis, a novel approach to perform this task is presented. It uses general and accurate representations of the part shape, removed material, cutting edge and rake face. Solid models are used to represent the part and removed volume, B-spline curves are used for the cutting edge representation and B-spline surfaces for the rake face. Next, a generic solid modeler based milling process simulation system for 3-axis machining of complex parts using flat and ball end mills is implemented. It consists of geometric and physical simulators. In the geometric simulation, the tool swept volume is generated for every completed tool path (one NC block) and intersected with the part, yielding the corresponding removed material volume. The tool cutting edges are then intersected with that volume to produce the tool-part immersion geometry in the form of in-cut segments. These and an expression for the chip thickness are used to determine the chip load distribution along the cutting edge. The updated part can be used instead of the removed volume in the intersection step to generate the in-cut segments. The physical simulator models the mechanics and dynamics of the cutting process and uses the chip load to compute instantaneous cutting forces and predict other process parameters. The milling process simulation is demonstrated and verified experimentally for 2 1/2- and 3-axis ball end milling. In addition, it is shown that geometric modelling of 4- and 5-axis milling using different tool shapes as well as other machining processes such as turning and drilling can be performed using the same approach.</p> / Doctor of Philosophy (PhD)

<p>An experimental investigation is presented for boiling of water at atmospheric pressure on a single copper surface. The distribution of active site was investigated at different levels of heat flux and subcooling. The results obtained indicated that the active sites were distributed randomly on the heating surface, since the distribution of active sites followed the Poisson distribution.</p> <p>Changes in heat flux and subcooling did not affect the distribution of active sites, although additional active sites formed among the sites which had been already activated when heat flux was increased.</p> <p>The frequency of vapour bubble emission and the bubble flux density have been studied as well. The results obtained showed that the surface superheat had a great effect on the vapour bubble emission frequency, although the influence of subcooling was of lesser significance. The bubble flux density showed a marked tendency to cluster and was not uniformly distributed over the heating surface.</p> / Master of Engineering (ME)

<p>A study is carried out to investigate the cross-flow induced instabilIties In heat exchanger tube arrays. For this purpose. the shortcomings of the Lever and Weaver unsteady theoretical model for a single flexible tube are dealt with and the modified model is extended to a multiple flexible tube analysis. Among the significant modifications is the introduction of a decay function to take into account the decay of the perturbations. This model predicts both static and dynamic instabiIities in the transverse and longitudinal directions. It was found that a single flexible tube become tends towards divergence at high values of the mass-damping ratio. This phenomenon is associated with smaller vibration frequencies than the natural frequency of the heat exchanger tube and approaches zero (divergence) at very high mass-damping ratios. The single flexible tube model is extended to a multiple flexible tube model to investigate the effect of the motion of neighboring tubes. It was found that this effect is very important at high values of the mass-damping ratio where the instability is dominated by stiffness terms. The decay function is investigated experimentally. Velocity fluctuations are measured up to 4 tube rows upstream the Vibrating tube. The experimentally determined decay function is used to predict the critical velocities for the dynamic instability. Equations of the theoretical model are solved numerically and the agreement between the experimental data and the theoretical predictions is reasonable for all array configurations.</p> / Doctor of Philosophy (PhD)

<p>This thesis contains the details of an experimental and analytical study on the refilling and rewetting of hot horizontal tubes. The experiments were conducted to refill and rewet two 25.4 mm 1.0., 3 m long zircaloy tubes (1 mm and 2 mm wall thicknesses) under a wide range of initial and boundary conditions. In addition to obtaining the quenching rates and the various parametric effects, the data played an important role in understanding the nature of the quench front propagation and identifying the associated flow patterns. Moreover, the data were used to obtain quench curves at different axial and circumferential locations. These quench curves helped in investigating the parametric effects on the different modes of heat transfer associated with the rewetting process as well as giving valuable insight into the rewetting mechanisms involved in the process.</p> <p>A rewetting criterion based on vapour film collapse in the film boiling region is proposed. This criterion suggests that surface rewetting is initiated when the vapour film thickness reduces down to a value equal to the sum of the amplitude of the liquid-vapour interface fluctuations and the solid surface roughness. Accordingly, a theoretical model for predicting the dynamic behaviour of the liquid-vapour interface in the film boiling region near the bottom of a horizontal tube was developed. This study showed that the hydraulic disturbances represented by the system pressure fluctuations could induce interface fluctuations capable of rewetting the surface when the equilibrium vapour film thickness was close to the experimentally estimated critical value.</p> <p>A two-fluid model incorporating the proposed rewetting criterion, as well as derived models for the partition factor and the film boiling heat transfer coefficient, was developed to predict the transients of the refilling and rewetting of hot horizontal tubes. This model is capable of predicting the experimentally observed thermal-hydraulic transients of the refilling "and rewetting processes.</p> / Doctor of Philosophy (PhD)

<p>This thesis is primarily concerned with the energy required to hydraulically transport sand and lime slurries in a 5.1 cm.-diameter vertical pipeline for both steady and pulsating flows.</p> <p>In the steady flow experiments, three different grades of sand of average weighted diameter d = 0.66, 1.20 and 1.84 mm. were investigated. The lime flurry used had an average particle diameter of 1.8 x 10⁻³ mm. The delivered volumetric concentration C₅, the average mixture flow rate Q, and the pressure difference ΔP across the test section were measured within the Reynolds number range of 4 x 10⁴ to 2 x 10⁵.</p> <p>In the pulsating flow experiments, only one grade of sand of average weighted diameter d = 1.20 mm. was investigated at frequencies of 0.22 and 0.44 cycles per second.</p> <p>Two methods of predicting pressure gradient for slurry flow were attempted. The first was adapted from Newitt et. al. (1961). By comparing the measured pressure gradient i for various suspensions to those for water flowing under similar conditions iw, an excess dimensionless pressure parameter Φ was obtained. It was possible to correlate the steady flow data using Φ and another dimensionless parameter Ψ related to mixture velocity Vm, diameter ratio D/d and specific gravity of solid S. In the second method the excess pressure gradient (i-iw) is equated to the excess hydraulic head CT(S-1). The results are presented in graphical form and each method is appraised in the discussion.</p> <p>The energy dissipated per unit height of the test section and mass flow rate was determined for steady and pulsating flow of suspensions. The energy ratio JP/Js at the two frequencies was obtained for low concentration slurries (Cs < 8.0%). It was found that no energy savings were recorded for pulsatile flow under these conditions.</p> / Master of Engineering (ME)

<p>This thesis deals with the application of the Finite Element Method to the analysis of laminated fiber composites. Specifically, it addresses the problem of representing the variation of material properties through the element volume in order to reduce the number of degrees of freedom required to represent a laminated composite. This is accomplished by a modification in the evaluation of the element stiffness matrix, whereby the through thickness integration is evaluated separately for each ply of the laminate. This modification results in a significant reduction in the core memory requirements. The modification is implemented in a finite element code and used to investigate various aspects of laminate behaviour using a design philosophy based upon the laminae constituents rather than the more usual laminate properties. Problems involving lamina coupling, edge and surface deformations are solved for symmetric and unsymmetric laminates. The tensile behaviour of a particular laminate is predicted, and the ability to back calculate the specimen composition parameters is demonstrated. An explanation of the specimen size limitation for tensile tests is verified. Views of plate edge deformations are obtained that are not predicted by classical laminated plate theory, and that are currently unavailable in the literature. The modified element formulation is also used to implement a method whereby laminates of various lamination sequences may be ranked in terms of their energy absorption potential when subjected to quasistatic loading conditions by comparing the total energy absorbed before catastrophic failure. This is accomplished by the used of a damage analysis method that is based upon element integration point failure rather than the usual first ply failure criterion.</p> / Doctor of Philosophy (PhD)

<p>The sulphidation properties of Fe-10wt% Al alloys at 1173 K in H₂SH₂ atmospheres were studied in the range of PS₂ = 1 x 10⁻² atm to 1 x 10⁻⁶ atm with special reference to the morphological development in the sulphide scale. Nodular growth of Fe₁₋ᵪS has been observed at Ps₂ = 1 x 10⁻⁴ atm to 1 x 10⁻³ atm with extensive internal sulphidation beneath these nodules.</p> <p>Sulphidation kinetics of the Fe-10wt% Al alloys were approximated to a two-stage parabolic rate. Aduplex scale is observed with an outer layer of Fe₁₋ᵪS, an inner layer composed of Fe₁₋ᵪS + FeAl₂S₄ having a pearlitic appearance and Al₂S₃ platelets or needles at the alloy/scale interface. At long exposure, a film of FeAl₂S₄ + Al₂S₃ develops at the alloy/scale interface. Two models are advanced for the growth of this film. The presence of Al₂S₃ platelets or needles beneath the continuous Al₂S₃ film is interpreted to result from morphological breakdown of the Al₂S₃ alloy interface and internal sulphidation. Platinum markers were found between the outer and inner layers suggesting growth of the outer layer by outward iron migration and growth of the inner layer by inward sulphur migration.</p> <p>A schematic ternary Fe-Al-S isotherm is constructed using the alloy sulphidation results and available thermodynamic data with respect to the binary sulphide phases. Various virtual diffusion paths representative of the sulphidation kinetics and scale microstructures for Fe-10wt% Al and Fe-15wt% Al alloys are indicated on this schematic diagram.</p> / Master of Engineering (ME)

<p>Some of the advantages and problems involved in the conversion from manual to automatic control of foundry molding sand constituents are discussed. The principal problem is that a transducer capable of monitoring each additive content is currently not available. The results of a search for an analytical technique suitable for such a transducer are presented. Of the numerous analytical methods reviewed, the multiparameter electrical technique, appeared to be most advantageous at the time that the survey was made. A description of its experimental evaluation is given. While a suitable transducer is not positively identified, guidelines for further work are suggested.</p> / Master of Engineering (ME)

<p>Throughout this thesis the design of machine structures has been approached as a system comprised of three inseparable but distinct sub-systems; (1) computer hardware, (2) software, and (3) mathematical model. This work has evaluated the role of each sub-section and has developed a strategy to incorporate the capabilities of each sub-system into a cohesive approach to automated structural design.</p> <p>A governing criterion of this work has been the desirability of retaining the engineering in the design process and devising a system that effectively utilizes his decision making abilities while relieving him of the tedious tasks of data assemblage and analysis.</p> <p>Part One, the initial stage of this work, has been the development of a remote intelligent terminal system that has been designed to incorporate a selection of computer peripherals into a total system. This system can be fully utilized in a flexible manner by a user knowledgeable in FORKAN with a working knowledgeable of the timesharing system [missing text] the communication strategy developed. As well, a description is included of the generalized software applicable to the peripherals associated with the terminal and resident in the large scale computer.</p> <p>In addition, Part One involves a description of the development of an operating system for the in-terminal computer as well as a library of programs that can be used with the system while it is operating in a stand-alone mode.</p> <p>In Part Two of this work a software system has been developed that utilizes the terminal system and a large scale computer to demonstrate the application of this technology to a system for interactive design of machine structures. This software system is modular in approach allowing the user to enter the system at any point or terminate the program at any point - all files are automatically appended and stored by the system. Thus, no information is lost for partial runs.</p> <p>Part Three is a description of the sub-optimal algorithm used in the automated design section described in Part Two. Results of trial designs of a standardized milling machine structure (CIRP structure) are presented.</p> <p>In Part Four the terminal system and the associated software developed has been applied to an industrial design problem in the glass bottle industry. This application is presented to demonstrate how this type of intelligent terminal can be used as the central element in a hierachical computer system for Computer-Aided-Manufacturing (CAM).</p> / Doctor of Philosophy (PhD)

<p>In this work a new NC design, based on the existing APT system (processor and postprocessors) and consisting of an additional front-end interactive Basic preprocessor, is presented. Software was developed that enables a part programmer to generate a part program on a 16 K minicomputer (PDP11/34). In order to fully exploit the off-line capabilities of an intelligent terminal, routines were written that display the geometrical input. While verifying the latter the user is offered the opportunity to instantaneously modify the displayed geometry if required.</p> <p>Subroutines of the APT part programming language (macros) were developed to cover basic geometrical configurations of turned parts (bar stock and forgings) and 2½ axes milled pockets. This approach reduces part programming to inserting dimensions of the workpiece and the technology involved.</p> <p>As far as technology is concerned, a milling optimiza-routine was developed that can direct the user in the selection of a cutterradius, feed and rotational speed when machining pockets.</p> / Master of Engineering (ME)