The contemporary product design and manufacturing environment requires a bidirectional and seamless data flow throughout all stages of data transactions. The establishment of STEP (STandard for Exchange of Product data) offers manufacturers a new method to exchange product data in the entire product life cycle. As an extension to STEP, STEP-NC provides the potential to finally close the gap between design and manufacturing in the drive for a complete, integrated product development environment. The STEP-NC data model is a long overdue improvement in the domain of computer numerical controls (CNC) where G-codes have been in use for more than half a century. STEP-NC brings richer information to CNCs presenting an opportunity for the development of more intelligent, interoperable and informative machining. The research work documented in this thesis introduces a fully STEP-compliant CNC system with an aim to solve the problems faced by the current CNC systems, e.g. disconnection with the upper-stream activities. The research is based on the premise that a STEP-NC program can document “generic” manufacturing information (whatto- do). This way, a STEP-NC program can be made machine-independent and has an advantage over the conventional G-code based NC program that is always generated for a specific CNC machine. The core of the STEP-compliant CNC system is the mapping mechanism which accepts STEP-NC data and translates it into the type of G-code that a specific controller can understand. The STEP-compliant CNC system has two principal elements, a STEP-NC Adapter and a STEP-NC Converter. The STEP-NC Adapter takes a generic STEP-NC program as its input and produces a more specific version of the STEP-NC program by “adapting” it to a local manufacturing environment and capabilities. This native STEP-NC file is a localised version of the generic STEP-NC file. It may be re-adapted to another manufacturing environment. The STEP-NC converter takes a native STEP-NC file and converts it to the required machine control data, i.e. G-codes that are ready to be loaded to the targeted CNC controller. The conversion process is also called a mapping or translating process that utilises the manufacturing environment databases developed for specific machine tools, or rather CNC controllers. The use of function block (IEC 61499) technology gives the system robustness and modularity. The STEP-NC adapter and converter are so developed that they can be embedded into a future CNC controller, making it a STEP-NC enabled controller. The system offers ample benefits to the CNC users by providing an interface for the STEP-NC data that carries rich information and the interface itself is also more userfriendly. It realises a long-awaited paradigm of common interface for CNC machine tools. With this system in place, one generic STEP-NC program can be made to drive different CNC machine tools, making NC programs portable and universally compatible. This system supports the scenario of “design anywhere and build anywhere”.

Continuous Fibre Reinforced Thermoplastics (CFRTPs) combine high strength, stiffness, impact and chemical resistance with possibilities for efficient part production by various thermoforming processes. Sheet forming of molten CFRTP laminates has generated much interest, but problems of buckling, wrinkling and predicting fibre distribution have meant a deeper understanding of these processes is needed. The first part of this study looks at the problem of gross buckling in homogeneous “trellis” flows of bidirectional laminates. Modelling the molten composite as a Newtonian fluid reinforced by inextensible fibres, linear stability analysis is used to determine the growth rate of small out-of-plane imperfections. Buckling is predicted when fibre tensions are negative, indicating that laminates must be kept in tension during forming to reduce such defects. A new approach to Grid Strain Analysis is presented, which uses surface fitting to determine the deformations occurring in sheet forming. The new method improves analysis of smooth, inhomogeneous deformations, and allows greater flexibility in viewing the results. The technique has been used to visualise deformations in a blister fairing made from cross-ply PLYTRON laminates. Arrow diagrams produced from the part demonstrate the tendency of bidirectional composites to deform by trellis flow, while transverse flow results from the action of the diaphragms used to form the part. The significance of inter-ply slip in CFRTP sheet forming provided the impetus to develop a finite element model for molten laminates, which treats each ply as a separate continuum. Contact between plies is modelled, with slip given a viscous response. Ply deformations are governed by a highly anisotropic elastic law, to handle the stiff fibres and as a first step towards a viscoelastic model of major intra-ply deformation modes. The finite element model parameters were adjusted to fit the part shape and load response of unidirectional PLYTRON laminates in bending. However, a perfect fit is unobtainable due to local transverse flow occurring at the bend in the real laminate. Nevertheless, the bending of the remainder of the ply is well described by the elastic model, using a fibre direction stiffness 25000 times that in the transverse direction. With the present model, a somewhat less anisotropic set of parameters gives the best overall fit and has been applied in several thermoforming simulations. As observed in experiments, matched-die bending simulations display ply buckling at high forming speeds. Hemispherical dome forming simulations exhibit out-of-plane buckling and near-inextensible fibre behaviour, with trellis-like deformation predominant in cross-ply laminates. In simulated double-diaphragm forming of bends and hemispherical domes, tension superimposed on the laminate from the stretching diaphragms is shown to eliminate buckling. However, high forming pressures and excessive transverse flow are a problem with current, stiff diaphragms. Final discussions look at improving contact modelling, reducing model sizes by adopting thin shell assumptions, and improving the ply model.

In this thesis, efficient algorithms for fast shape approximation in the area of machine and robot vision are analysed. The core of the thesis consits of two main parts, an error analysis of chain coded silhouettes and a simplified, hence fast shape approximation scheme for practical applications. The area of shape description and approximation is introduced. General techniques for processing of boundary data are outlined followed by a brief description of the image processing system which was developed during the study. Processing procedures developed and their options are presented. Two applications realized by the author have emerged from the algorithm development, firstly a semi-automatic shape registration and measurement procedure and secondly, a low-cost robot vision system. The advantage of the experimental robot vision system is an effective communication between the robot and the object recognition system. The shortcomings of these applications, mainly inaccuracy of the shape description lead to the introduction of an error analysis for boundary descriptors. In contrast to existing methods, the error analysis can be applied to straight lines, circular arcs, and arbitrary shapes consisting of these two shape primitives. The analysis is comprehensive and covers all kinds of chain code sets representing various pixel shapes as well as different pixel configurations. The results of the analysis lead to the development of simple ways of overcoming the accuracy limitations of conventional methods. Two algorithms for improving the length estimation of object outlines are introduced. A comparison with existing methods showing the effectiveness of these algorithms is made. For practical applications of object recognition, a new, simple and hence fast and effective algorithm for polygon approximation, called the "arc operator" is developed. The combination of the arc operator and the length correcting algorithm directly improves the ability to identify partially visible shapes. The performance of the arc operator is compared to six other well known algorithms. Slight disadvantages in accuracy are outweighed by the enormous advantage in processing time, particularly with an integer implementation. An example of overlapping parts demonstrates the potential of the arc operator to extract characteristic shape descriptions.

The research detailed in this thesis investigates the practical application of Computational Fluid Dynamics (CFD) to downwind sail design. Simulations were performed using CFX-5, an unstructured commercial CFD package. The research focuses on the performance of the SST and k-ω turbulence models which were judged to be CFX-5's most appropriate turbulence models for downwind sail flows. Two-equation turbulence models are viewed as the most appropriate model type for sail simulations, they capture a significant amount of the flow physics whilst providing turnaround times for sail simulations of less than one day. CFD simulations were compared with experimental data for a flat plate at shallow angles of incidence. This test case holds particular relevance to sail flows since both flows are affected by leading edge separation bubbles which form due to knife-edge separation at sharp leading edges. The CFD captures this leading edge bubble well, with the SST model predicting the length of the bubble with 7% of the experimental value. Wind tunnel data was gathered for two-dimensional downwind sail sections for the purpose of CFD validation. A preliminary wind tunnel study was carried out using a low aspect ratio model. The tests were prone to three-dimensional effects and only three-dimensional CFD simulations were capable of successfully reproducing the flow. High aspect ratio wind tunnel test results were also conducted in an effort to obtain nominally two-dimensional wind tunnel data. Surface pressures were measured using Pressure Sensitive Paint (PSP), however due to the low dynamic pressure of the tests error appeared in the data and comparison with the CFD was poor Results show that CFD is capable of qualitatively reproducing downwind sail flows, the leading and trailing edge separation regions were captured and the CFD results compared well with wind tunnel flow visualization. Finally, CFD simulations were used to investigate the two-dimensional downwind sail design space through a parametric study of sail draft and camber. Results show that increasing camber increases both lift and drag a trend that also is evident in three-dimensional sail designs. It is also shown that gains can be made by using designs with draft values as far aft as 60% which helps reduce the extent of trailing edge separation. This parametric design study illustrates how CFD can be used successfully to analyse design trends and rank designs. The research presented illustrates how CFD can be used in the design process but also that care must be made in validating the method. Through this study the relative strengths and weaknesses of the turbulence models are better understood. Whilst CFD cannot yet be reliably used for downwind sail performance prediction, it is still a useful tool for investigating the flow structure which leads to better understanding of the design space.

An investigation of cyclic variability in a single-cylinder spark ignition engine fuelled by natural gas was carried out. The effects of combustion chamber shape and compression ratio were examined intensively. Four different combustion chambers were tested and compression ratio was varied in the range of 8.5:1 to 15:1. For each test condition, the in-cylinder pressure data were recorded at each degree of crank angle rotation for a total of two hundred consecutive engine cycles. These pressure-time histories were used for the detailed study of combustion characteristics. The generalized plot of maximum pressure and location at which maximum pressure occurs was demonstrated as an effective means of diagnosing the cyclic variability. It was shown that compression ratio had negligible effect on cyclic variability, but combustion chamber shape had a strong influence. The indicated mean effective pressure proved to be a good index for measuring cyclic variability for a wide range of operating conditions. However, for specific cases in which the variation of indicated mean effective pressure shows poor sensitivity, maximum pressure and maximum rate of pressure rise are useful measures. For high compression ratio engines, a compact combustion chamber arranged around the spark plug has been shown to be superior for reducing cyclic variability. In addition, a study of a model for the small-scale structure of turbulence using the data from these experiments showed good agreement with those from the literature. Alternative assumptions for estimating turbulence length scale were tested and a preferred method is recommended. The study showed that the fast burning combustion chamber is the solution to reduce cyclic variation. However, cyclic variations of maximum pressure and its location in the fast burning condition still exist and continued research to reduce these cyclic variations should concentrate on the variation of characteristics of flame kernel creation and growth from cycle to cycle.

An original solution method ("Snippy") is developed for the rectangular stock cutting problem, and compared with the commercially available package "OREC" by applying both methods to industrial order books. Arguments are presented to show that the solutions found by Snippy are optimal for all except one of the order books. Snippy incorporates two innovations. The first is a new method for generating guillotinable cutting patterns, the Trickle Method. This is an incremental dynamic programming technique which is especially suited for interaction with the Simplex Method via column generation. The second development involves a class of non-guillotinable patterns that can be constructed using essentially guillotine techniques. In particular, a relatively minor modification to the Trickle Method allows the generation of good non-guillotinable patterns.

Whole document restricted, see Access Instructions file below for details of how to access the print copy. / An analysis of the industrial productivity of New Zealand heavy engineering companies is presented, and methods for improving the overall productivity of the heavy engineering industry and similar industries with job-shop type production are developed. The industry's productivity problems had been obvious for years. However, due to the lack of data and inadequacy of existing productivity improvement approaches, it had never been possible to quantify the extent of these problems, to analyse their causes and to develop methodologies for long-term improvement. The present investigation consists of two major aspects. The scientific element is concerned with the development of a methodology for productivity improvement appropriate to the situation of heavy engineering in New Zealand. This is supported by practice-oriented work in the industry, consisting of data acquisition activities in general, and of pilot studies in selected companies in order to assemble, analyse and evaluate specific data on productivity problems, and to apply and test the results thereof. The development of a methodology for productivity improvement is based on an extensive survey of literature on productivity measurement and improvement methods. The results of this survey, which was performed in parallel with the collection of industrial data, indicate that existing methods are not adequate to satisfy the requirements of productivity improvement in the local heavy engineering industry. On this basis, in-depth pilot studies in ten heavy engineering companies were performed. The objectives and methodology of these pilot studies are described in detail, as their results have a sizeable impact on the overall methodology chosen for this research. One of the most important conclusions drawn from the pilot studies is that productivity problems in the New Zealand heavy engineering industry cannot be solved by concentrating solely on workshop fabrication and technological factors. Generally these problems have complex cause-and-effect structures, and a multitude of non-technological factors from outside the workshop are involved. In order to account for these interrelated factors, a systems engineering approach was used, which offers a suitable basis for a productivity improvement methodology applicable to the situation as identified in the pilot studies. A main step in the system engineering approach is the development of a systems model which is used for structuring the complex inter-relationships found in practice. On the basis of this systems model of heavy engineering productivity a Productivity Assurance Programme is developed. This programme combines elements of quality assurance methods and the 'productivity cycle' principle of continuing improvement. The main elements of the Productivity Assurance Programme are matrices developed for the evaluation of the requirements of productive heavy engineering operation, and for the analysis of the productivity levels of the company where they are applied. The combination of these aspects provides a decision base on which organisational improvements can be founded. Due to its modular structure and the flexibility in defining specific productivity requirements, the applicability of the Productivity Assurance Programme is not limited to New Zealand heavy engineering companies, but also covers other job shop type industries with similar productivity problems. I case study illustrates the application of the Productivity Assurance Programme in practice.

This work presents a 50-industry input-output model of Northland's economy and demonstrates how input-output analysis could be used to enhance regional planning in New Zealand. As it is the first regional input-output model in this country to incorporate significant survey and secondary data the survey procedure and model construction are outlined. The input-output table is used to discuss important regional transactions and the purchase and sales patterns of industries. The model analyses industries' contributions to export receipts and import payments and calculates the impact of changes in export receipts on regional income and imports. A comprehensive multiplier analysis of Northland's economy covers output, income, employment and imports and confidence limits for the multipliers are developed using the Monte Carlo technique to simulate survey errors. The model explores the economic implications of forestry expansion in Northland and discusses the areas available for afforestation, planting rates, tree management, wood supply and wood processing options in the region. The modifications made to the model and data requirements for simulating forestry expansion are outlined and employment and income impacts given for three types of processing complexes and for forestry expansion as a whole. Finally an economic evaluation is made of the impacts of processing-plant construction and supporting services.

The contemporary product design and manufacturing environment requires a bidirectional and seamless data flow throughout all stages of data transactions. The establishment of STEP (STandard for Exchange of Product data) offers manufacturers a new method to exchange product data in the entire product life cycle. As an extension to STEP, STEP-NC provides the potential to finally close the gap between design and manufacturing in the drive for a complete, integrated product development environment. The STEP-NC data model is a long overdue improvement in the domain of computer numerical controls (CNC) where G-codes have been in use for more than half a century. STEP-NC brings richer information to CNCs presenting an opportunity for the development of more intelligent, interoperable and informative machining. The research work documented in this thesis introduces a fully STEP-compliant CNC system with an aim to solve the problems faced by the current CNC systems, e.g. disconnection with the upper-stream activities. The research is based on the premise that a STEP-NC program can document “generic” manufacturing information (whatto- do). This way, a STEP-NC program can be made machine-independent and has an advantage over the conventional G-code based NC program that is always generated for a specific CNC machine. The core of the STEP-compliant CNC system is the mapping mechanism which accepts STEP-NC data and translates it into the type of G-code that a specific controller can understand. The STEP-compliant CNC system has two principal elements, a STEP-NC Adapter and a STEP-NC Converter. The STEP-NC Adapter takes a generic STEP-NC program as its input and produces a more specific version of the STEP-NC program by “adapting” it to a local manufacturing environment and capabilities. This native STEP-NC file is a localised version of the generic STEP-NC file. It may be re-adapted to another manufacturing environment. The STEP-NC converter takes a native STEP-NC file and converts it to the required machine control data, i.e. G-codes that are ready to be loaded to the targeted CNC controller. The conversion process is also called a mapping or translating process that utilises the manufacturing environment databases developed for specific machine tools, or rather CNC controllers. The use of function block (IEC 61499) technology gives the system robustness and modularity. The STEP-NC adapter and converter are so developed that they can be embedded into a future CNC controller, making it a STEP-NC enabled controller. The system offers ample benefits to the CNC users by providing an interface for the STEP-NC data that carries rich information and the interface itself is also more userfriendly. It realises a long-awaited paradigm of common interface for CNC machine tools. With this system in place, one generic STEP-NC program can be made to drive different CNC machine tools, making NC programs portable and universally compatible. This system supports the scenario of “design anywhere and build anywhere”.

Continuous Fibre Reinforced Thermoplastics (CFRTPs) combine high strength, stiffness, impact and chemical resistance with possibilities for efficient part production by various thermoforming processes. Sheet forming of molten CFRTP laminates has generated much interest, but problems of buckling, wrinkling and predicting fibre distribution have meant a deeper understanding of these processes is needed. The first part of this study looks at the problem of gross buckling in homogeneous “trellis” flows of bidirectional laminates. Modelling the molten composite as a Newtonian fluid reinforced by inextensible fibres, linear stability analysis is used to determine the growth rate of small out-of-plane imperfections. Buckling is predicted when fibre tensions are negative, indicating that laminates must be kept in tension during forming to reduce such defects. A new approach to Grid Strain Analysis is presented, which uses surface fitting to determine the deformations occurring in sheet forming. The new method improves analysis of smooth, inhomogeneous deformations, and allows greater flexibility in viewing the results. The technique has been used to visualise deformations in a blister fairing made from cross-ply PLYTRON laminates. Arrow diagrams produced from the part demonstrate the tendency of bidirectional composites to deform by trellis flow, while transverse flow results from the action of the diaphragms used to form the part. The significance of inter-ply slip in CFRTP sheet forming provided the impetus to develop a finite element model for molten laminates, which treats each ply as a separate continuum. Contact between plies is modelled, with slip given a viscous response. Ply deformations are governed by a highly anisotropic elastic law, to handle the stiff fibres and as a first step towards a viscoelastic model of major intra-ply deformation modes. The finite element model parameters were adjusted to fit the part shape and load response of unidirectional PLYTRON laminates in bending. However, a perfect fit is unobtainable due to local transverse flow occurring at the bend in the real laminate. Nevertheless, the bending of the remainder of the ply is well described by the elastic model, using a fibre direction stiffness 25000 times that in the transverse direction. With the present model, a somewhat less anisotropic set of parameters gives the best overall fit and has been applied in several thermoforming simulations. As observed in experiments, matched-die bending simulations display ply buckling at high forming speeds. Hemispherical dome forming simulations exhibit out-of-plane buckling and near-inextensible fibre behaviour, with trellis-like deformation predominant in cross-ply laminates. In simulated double-diaphragm forming of bends and hemispherical domes, tension superimposed on the laminate from the stretching diaphragms is shown to eliminate buckling. However, high forming pressures and excessive transverse flow are a problem with current, stiff diaphragms. Final discussions look at improving contact modelling, reducing model sizes by adopting thin shell assumptions, and improving the ply model.