Engineering: Theses and Dissertations

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 2972
  • ItemOpen Access
    Does Topology Provide Sufficient Structure for Non-Causal Explanations?
    (2023) Jha, Aditya
    There is a major debate as to whether there are non-causal mathematical explanations of physical facts that show how the facts under question arise from a degree of mathematical necessity considered stronger than that of contingent causal laws. Topology provides an ideal ground for such purported non-causal explanations since topological manifolds, on which the parameters of a dynamical system can be modelled, are typically associated with multiple invariants, which remain unaltered even if the manifold is bent, stretched or twisted, reflecting a change in the parameter modelled on the manifold. Understood in this sense, topological explanations seem to provide modal information about certain constraints on the system that may not be evident in detailed, and often, cumbersome causal explanations. This thesis examines some foundational issues in the applicability of topology to the natural world and their bearing on the debate on such purported non-causal (mathematical) explanations. More specifically, this thesis looks into various topological and geometrical formulations that essentially exploit the geometry of oscillating and complex systems, as an exercise in ‘geometric mechanics’, to provide a simple explanation of certain constraints imposed on their dynamics by the virtue of their geometry. The central question answered in this thesis is whether topology provide sufficient structure for such non-causal explanations. The answer, as the thesis demonstrates, is negative because topological explanations critically rely on idealisations, such as continuity and smoothness, which are realised only contingently in the natural world (or in mathematical approximations/models of the natural world); these idealisations impose some foundational limitations on the application of topology in modelling such systems ‘non-causally’. Consequently, purported topological explanations fail to fully circumvent the causal dependencies of such systems implying that these are not really ‘non-causal’ explanations. This thesis also extends the argument to mathematical explanations in general. It argues that purported mathematical explanations are essentially causal explanations in dis- guise and are no different from ordinary applications of mathematics to the natural world. This is because these explanations work not by appealing to what the world must be like as a matter of mathematical necessity, but by appealing to various contingent causal facts. These contingent facts, although assumed away in the why-question pertaining to a physical fact, still participate as causal facts in an explanation of why the fact obtains in the world. That is, the explained physical fact does not obtain because of a mathematical necessity but by appeal to the world’s network of causal relations.
  • ItemOpen Access
    Transverse velocities and shear stresses in river bends
    (2000) Plew, David Russell
    Expressions are presented for the transverse velocity profile and the transverse bed shear stress in a channel bend. These expressions are obtained from the momentum equations by assuming a logarithmic longitudinal velocity profile. The boundary conditions and expressions for eddy viscosity are consistent with the logarithmic longitudinal velocity profile. As a result, the transverse bed shear stress can be calculated without the need to use ratios of near-bed or slip velocities. The influence of non-zero depth-averaged velocities are included in the expressions. As a result, the expressions for the transverse velocity profile and transverse bed shear stress are not limited to situations of fully developed bend flow, as is the case with several previous profiles. The expression for the transverse velocity profile is verified by field experiments using acoustic doppler velocimeters to measure velocity profiles in two stream bends. A simple meandering stream bed topography model is developed illustrating how the expressions for transverse shear stress can be used. This topography model uses an analytical solution to the depth-averaged momentum and continuity equations to calculate depth_-averaged velocities. The shear stresses are calculated using the method presented in the thesis. Topography is calculated from sediment continuity. The use of a sine-generated curve to describe river meander geometry is briefly discussed. Regime equations are combined to give predictions of meander geometry from a known discharge and valley slope. Finally, results of experiments exploring behaviour of meandering channels in a sand tray are presented, illustrating various aspects of bed hydraulics and sediment movement.
  • ItemOpen Access
    Circular supply chains : enhancing consumer participation.
    (2023) Robertson, Juliet
    With increasing necessity for the world to respond to environmental crisis, environmental guardianship is essential and can be achieved through the circular economy (CE). The CE is defined by a rethinking of how resources are used, proposing a system of cascaded resource use and reuse, to fuel economic prosperity without depleting the earth’s resources. This is achieved through a circular consumption model that is in stark contrast to the traditional linear consumption model. To allow this change in society’s consumption, firms’ supply chains must also undergo change. The transformation of linear supply chains to circular closed-loop supply chains is on the horizon for many firms and is the topic of significant discussion within supply chain management research. An area that can be considered under-researched in this topic is that of consumers returning materials to the supply chain. As a result, this research addresses the question of how firms can invoke stronger consumer engagement in returning materials in closed-loop supply chains. To consider the consumer propensity to return materials, the extended theory of planned behaviour (ETPB) is employed as the theoretical lens. This study therefore focuses on the factors that influence consumers’ intentions to return materials. These include the traditional theory of planned behaviour factors as well as ‘habits’ and ‘moral norms’ that were expected to offer further explanation in this context. A survey, designed on Qualtrics, was created to collect data for quantitative analysis. The ‘intention to return’ dependent variable was investigated for three product categories that survey respondents used in their lives. Focus was given to product categories that are ‘low-involvement’, meaning little care is given to their purchase and their end-of-life. The three products were single-use coffee pods, printer cartridges and batteries. Data was collected from respondents who are users of these products from New Zealand and Australia, recruited through the Prolific platform. Research contributions are relevant to both supply chain practitioners moving towards a closed-loop system and for further research in the arena. The results showed a consistently significant influence from two variables across the three parallel studies: perceived behavioural control and moral norms. Given the minimal influence of other variables, this indicates a limitation to traditional consumer behaviour models on consumers engaging in low-involvement behaviours. Implications of these findings are in the theoretical, practical and policy realms. Potential applications of the findings are given throughout the discussion of results. A key contribution is the recommendation of robust policy to achieve the CE in New Zealand in regard to these unique goods.
  • ItemOpen Access
    Acting aids in virtual production.
    (2023) Moisset, Sylvain
    Previous research has shown how using virtual production could solve some of the inherent problems that suffers the traditional film making workflow mainly by bringing more flexibility to the overall process, allowing more iterative work and more communication between all the departments involved. That being said, studies have also shown that for actors, working on a film that involves visual effects can put a lot of strain on their concentration and attention and lead to a frustrating experience for them and potentially less optimal results for the film. This project’s aim is to study if a virtual production setup can be used to provide actors, in a visual effects film making environment, with acting aids in the form of visual cues on a display. It also aims at studying which cues are more helpful for actors. A simple setup was designed to mimic the experience of acting in a professional virtual production volume. The system includes the possibility of displaying different cues within the virtual production environment. A user study was run on a sample of actors to identify if cuing using a virtual production setup leads to better results. Experts opinions were collected during the design phase of the project to assess the relevance of the project and get directions for the design of the prototype and the use study.
  • ItemOpen Access
    Visual-inertial odometry for windblown vegetated environments.
    (2023) Schofield, Sam
    Aerial manipulation (performing interaction tasks with unmanned aerial vehicles) can revolutionise the forestry and arboriculture industries by automating tree pruning and similar tasks. A major challenge currently preventing aerial manipulation from being used in these industries is the inability to accurately estimate the UAV's pose (position and orientation) in the target environments, specifi cally those with abundant wind- blown vegetation. Standard methods of UAV pose estimation, such as GPS, do not provide enough precision to allow aerial manipulation, while more precise methods, like visual odometry (VO), do not work reliably in environments dominated by such scene motion. This work takes several steps towards remedying the latter problem. First, a novel approach was proposed for aligning a motion-capture rigid body with the optical centre of a camera. The proposed method was more accurate than the existing approach and removed the need for a specially prepared calibration target. This work enabled more accurate ground truth trajectories to be captured and was also used when lab-testing a tree-pruning UAV that used a combination of camera and motion capture data for navigation. Next, a novel analysis of the state-estimation requirements of UAVs was performed to determine how existing VO algorithms would need to be improved to enable aerial manipulation in windblown vegetated environments. Preliminary experiments, completed in simulation and then verif ied using a real UAV, showed that VO accuracy was not a bottleneck to UAV position-hold performance in normal operating conditions. This result directed later research to focus on improving the robustness of VO in dynamic vegetated environments (such as forests) instead of trying to improve VO precision further. Several state-of-the-art visual-inertial odometry (VIO) algorithms were evaluated in scenes containing moving vegetation to determine whether existing algorithms met the desired robustness requirements. A preliminary analysis was performed using real data, followed by a more in-depth evaluation using a semi-synthetic dataset (consisting of real IMU data and synthetic images). These experiments con firmed the assumption that existing VIO algorithms were not suitable for use onboard a UAV performing manipulation tasks in a windblown forest. While generating the aforementioned semi-synthetic dataset, a major aw in existing methods for producing semi-synthetic data was discovered and addressed. A detailed comparison showed that VIO algorithms that were evaluated using data generated with the proposed approach performed signi ficantly (79%) better than when evaluated using data generated with existing methods. Finally, a novel outlier detection method was proposed that utilised the difference in nature between camera and scene-motion-induced optical ow. Experiments showed that, when used as a pre-processing step to RANSAC, the proposed method signi ficantly improved visual odometry accuracy (up to 20%) with minimal computational overhead in windblown forest environments.
  • ItemOpen Access
    Non destructive evaluation of Nothofagus fusca (red beech) wood properties within the North West Coast of the South Island, New Zealan
    (2005) Silcock, Paul
    This study is comprised of two major research objectives: 1) to quantify the incidence and proportion of defect in beech forests and explore the efficacy of applying defect detecting technology to natural forest management; and 2) to identify and assess those physical and mechanical wood properties of red beech that influence wood quality and forest operations from two sites (Howard valley and Wooly river) managed by Forever Beech Ltd. Overall there is a significant incidence of defect (56 %) in standing red beech and a low stem-to-timber conversion rate (40 %). An acoustic tomographic tool, Fakopp 2D, can be used to detect the presence of internal defect in standing trees, although the software used in this study was not able to consistently predict the location or volume of defect, it was a cumbersome instrument to use in the forest and data collection was a slow process. No case could be made for incorporating this technology in beech forest management practice. Four wood quality properties were used to assess beech forest in the NW Coast of New Zealand. Basic density and hardness are measures of wood quality and quantity on which many end-use and customer specifications are based. Green density is a measure of wood quantity which is used in forest operations to determine log bucking patterns for heli-lifting. The green dynamic modulus of elasticity (MOE) is another useful measure of wood quality, assessed using the TOF instrument - TreeTap. Measuring stress wave velocity (SWV) in standing red beech using time of flight (TOF) instruments provides a means to evaluate wood quality prior to harvesting. Forests of the North West (NW) Coast of the South Island were found to have low average basic density in comparison to average national values. Basic density and associated wood properties exhibited greater variation between trees than between sites. Trees sampled in the Wooly river site had higher basic density, green density, MOE and hardness values than those growing in the Howard valley. The trees of Wooly river also exhibited higher growth rates and a lower incidence of defect. All assessed wood properties of red beech decreased with altitude. MOE increased with increasing growth rate. These data provided for customer specifications and the operations and management of Forever Beech Ltd. The advantages of using MOE as a measure of timber quality are discussed.
  • ItemUnknown
    Subclasses of Transversal Matroids
    (2023) Hogan, Emma
    Unlike the classes of graphic and representable matroids, the class of transversal matroids is not closed under minors. Well-known minorclosed subclasses of the class of transversal matroids include the classes of bicircular matroids, lattice path matroids, and multi-path matroids. In this thesis, we investigate several minor-closed subclasses of the transversal matroids. We establish the complete lists of excluded minors for the intersection of the classes of graphic and lattice path matroids, and the intersection of the classes of bicircular and lattice path matroids. We also consider the class of multi-path matroids, and its intersection with the classes of graphic and bicircular matroids.
  • ItemUnknown
    Identification of congestion network components and the timing of network improvements from transportation modelling analysis
    (2008) Cooney, Ryan John
    In New Zealand, transportation planning is currently carried out at a number of stages during a projects lifecycle. Transportation planning typically involves using a three or four step transportation model, the outputs from which are then analysed using the Level of Service Analysis and the Economic Analysis methods. This report describes how these two analysis methods are applied to the outputs of transportation models. Two case studies have been developed, a transportation model run, and the two analysis methods applied, to enable the comparison of the two methods. The case studies and transportation model have been developed from the Christchurch, Rolleston and Environs Transportation Study (CRETS). From these case studies it has been show that the two analysis methods are not always consistent with respect to justification of, or the timing of implementing, network upgrade options. The reasons why this is the case are then explored. Upon reviewing the current Government transport strategies, it has been determined that the current approach to transportation planning and the current analysis methods, are inconsistent with the strategies. The problem-orientated, decision orientated, and objective-orientated approaches to transportation planning are subsequently explored. It has been determined that the current Government transport strategies are generally consistent with the objective­ orientated approach to planning. It has also been determined that the current Government transport strategies require a much more holistic approach to planning. Transportation planning can no longer just focus on roading improvements. Packages of options or strategies must now include activities such as public transport, active modes, freight, land use activity, travel demand management and road operation. In order to assess these strategies a new analysis method is required. Multi­Criteria Analysis has been investigated as a strategy assessment method, and it has been determined that it is an appropriate method given the data sets to be assessed.
  • ItemUnknown
    A model of secondary growth for radiata pine : stem ring width and basic density variation in relation to crown structure
    (2003) Pont, David
    A data set comprising detailed crown architecture and stem wood properties measurements obtained from a single tree was used to investigate the relationships between crown and stem growth, with the objective of developing a model of stem growth in terms of crown structure. Four hypotheses relating crown structure to stem growth, based on the theories of Pressler (in Assmann, 1970), Larson (1962 b, and 1969) and the pipe model (Shinozaki et. al. 1964), were tested. Relationships obtained from analysis of crown architecture data were used to estimate past crown structure, in terms of foliage mass and average distance to foliage, at each age for the sample tree. The method of Duff and Nolan (1953 and 1957) was used to analyse ring area and basic density measurements taken from each annual ring at several positions along the stem. Analysis revealed three fundamental stem growth patterns, correlated with crown structure. The relationships identified between crown and stem growth were used to support the hypothesis that ring area is related to foliage mass, and to support the hypothesis that distance to foliage and foliage mass are related to basic density. Two hypotheses based on the pipe model, relating sapwood area to foliage mass, were found to apply within the crown only. Alternative hypotheses, including the length of the conductive path to foliage, accounted for changes in sapwood area within and below the crown. The relationships established formed the basis of a model of secondary growth, comprising four equations, which estimates ring area increment and ring basic density at any point on the stem from foliage mass and average distance to foliage. The model is economical in structure and mathematical formulation and evaluation showed it predicts stem ring area and basic density consistent with well-known patterns. Definition of the model in terms of crown structure results in realistic responses to factors such as between-tree competition and thinning. The chosen modelling approach, linking crown and stem growth, is suitable for future development using data sets representing different site, silvicultural and genetic factors.
  • ItemOpen Access
    Performance and emission testing of wood-fired domestic heating appliances : a simplified method of testing
    (2001) Alcanzare, Robyn Joy C.
    Christchurch City often experiences smog during winter due to temperature inversion and this is highlighted by the use of domestic heating appliances. The particulate emissions from wood burning pose a health risk. The Canterbury Regional Council has drafted laws concerning the emission limits on woodburners and has adopted to use a standard method to test woodburners based on the wood burning rate. However, the current standard method has some limitations. A new standard test method for evaluating the efficiency, power output, and emission of a wood burner is proposed based on the analysis of the flue gas composition and measurement of the airflow. The method arose out of a study using a prototype, two-stage log burner, called the C3 Log Burner. Prior to developing the procedure, the air intakes and the water flowrate settings were optimised. Results showed that the optimised range for the water flowrate is 1.1 1pm and a 50% opening on the secondary air intake during steady state condition. Parameters such as moisture content, wood type and log size were studied to test the sensitivity of the test method. Larch with two sets of moisture content, (dry) 13.4% and (wet) 28.9% was burned but data were not gathered for the wet larch since it did not reach steady state condition. Softwoods, radiata pine and larch, were compared with a hardwood, eucalyptus gum. Results showed that hardwoods have lower emissions than softwoods with particulate levels of 0.059g/m3 , 0.065g/m3, and 0.046g/m3 for radiata pine, larch, and eucalyptus gum, respectively. The irregular-shaped radiata pine produced 0.047g/m3 compared to the regular-shaped of 0.059g/m3. The start-up, unsteady state condition, was also measured in terms of emissions and efficiency. The emissions were measured as 0.243g/m3 . The over-all efficiency of the C3 Log Burner lies between 75-79%, with an error of 7.0- 8.9%, and particulate emissions ranging between 0.046 to 0.065g/m3 (0.206 to 0.700g/kg) at steady state. Based on the test method used, the C3 Log Burner operates well under the 1.5g/kg limit set by the Canterbury Regional Council with an emission level 10 times lower than this limit. The proposed test method provides a simple and more convenient way of measuring the efficiency and the solid particulate emissions of the burner.
  • ItemOpen Access
    The application of dynamic mode decomposition to transient dynamics and non-autonomous dynamical systems.
    (2023) Peters, Nick
    Dynamic mode decomposition (DMD) is an equation-free method for computationally modelling dynamical systems and systems driven by partial differential equations. Based on measurements taken over time, DMD reconstructs a projection of the systems trajectory onto an observation space. It approximates the observed trajectory as a sum of spatial modes with simple temporal behaviours. A wellknown issue for DMD is that it has difficulty in reconstructing the trajectories of non-autonomous systems. These reconstructions tend not to resemble the original trajectory and may have large approximation errors associated with them. Here we present a modified DMD algorithm, called time varying dynamic mode decomposition (TVDMD), with the aim of improving reconstruction accuracy for trajectories of non-autonomous systems. In this thesis, both DMD and TVDMD are applied to a set of three systems of increasing complexity: the heat equation, the FitzHugh- Nagumo model with spatial diffusion, and the FitzHugh-Nagumo model with a parameter that changes over time. The two reconstructions are compared and it is shown that TVDMD outperforms DMD in terms of accuracy when modelling each of the three systems. The difference is small for the heat equation but is significant for the other two systems. However, this improvement in accuracy comes at the cost of increased complexity; the modes that make up the TVDMD reconstruction have more complicated temporal behaviours than those of the DMD reconstruction. Additionally, it is shown that sudden parameter changes or regime changes of a nonautonomous system can be detected by computing the DMD approximation error at each time step of the reconstruction.
  • ItemOpen Access
    The geometric wavefront sensor and atmospheric tomography for satellite imaging from ground-based telescopes.
    (2023) Hickman, Sierra
    The resolution of astronomical images captured from ground-based telescopes is partially limited by atmospheric turbulence, which causes non-uniform phase shifts to the wavefront of light from stars and other astronomical objects, and results in distorted images. Adaptive Optics (AO) and computer post-processing techniques – such as Deconvolution from Wavefront Sensing (DWFS), blind deconvolution, and Lucky Imaging (LI) – attempt to overcome the blurring effects of the atmospheric turbulence and aim to achieve diffraction-limited imaging resolution. A critical component in both AO and DWFS is the Wavefront Sensor (WFS), which estimates the phase aberrations introduced by atmospheric turbulence. The Geometric WFS uses geometric optics and ray tracing to measure the displacement of intensity fluctuations from two defocused images of the telescope aperture – the same defocused images utilised by the widely-established Curvature WFS. However, unlike the Curvature WFS, which measures the curvature of the wavefront, the Geometric WFS measures the slope of the wavefront. This thesis largely studies the Geometric WFS as part of the development of an astronomical imaging instrument for the University of Canterbury Mount John Observatory (UCMJO), which combines AO and computer post-processing techniques (including DWFS and LI) for astronomical observations and Space Situational Awareness (SSA) over a wide Field of View (FOV) by utilising the Geometric WFS and atmospheric tomography. The Geometric WFS is preferred over more popular WFSs, such as the Shack- Hartmann and Pyramid WFSs, for the hybrid astronomical imaging instrument due to the Geometric WFS’s potential to measure the respective wavefronts of multiple stars simultaneously, enabling atmospheric tomography, with fewer detectors. In laboratory demonstrations, a prototype Geometric WFS is shown to provide more accurate wavefront phase estimates than a similarly built Curvature WFS. A hybrid DWFS and LI computer post-processing system is demonstrated both in simulation and on-sky using both the Geometric and Curvature WFSs. Utilising the Full-Width at Half-Maximum (FWHM) imaging metric, DWFS is shown to provide the majority of the resolution improvement to the simulated and on-sky astronomical images, however the additional LI provides further improvement in the combined DWFS and LI system. The simulations of the hybrid system show that the Geometric WFS performs better than the Curvature WFS. However, a definitive assessment comparing the two WFSs in the on-sky hybrid system could not be made. When using the Geometric WFS, several parameters must be taken into consideration – such as the Signal-to-Noise Ratio (SNR), number of Radon angles, virtual propagation distance of the synthetic interaction matrix, and number of Zernike modes used to estimate the wavefront phase. In a laboratory experiment, these parameters are investigated using single Zernike mode wavefront phase aberrations. An inverse relationship is found between the wavefront phase error and the SNR, indicating greater accuracy in estimating wavefront phase aberrations from brighter astronomical objects. To accurately estimate low order Zernike modes (Z₄ − Z₁₅), the Geometric WFS requires a minimum of 5 Radon angles. The virtual propagation distance of a synthetic interaction matrix should be optimized by the physical defocus distance of the intra- and extra-focal detectors and/or the altitude of a strong atmospheric turbulence layer. Lastly, noise is amplified when estimating Zernike modes greater than the single Zernike mode wavefront phase aberration. Atmospheric tomography, which requires the wavefronts of multiple artificial guide stars to be measured simultaneously, is demonstrated on a laboratory optical test-bench using a single Geometric WFS in open-loop, and the wavefront phase aberration of a target region not directly measured by the Geometric WFS is estimated. A novel technique called the "cookie cutter" method, which changes the altitude of the wavefront phase layers (and thus the atmospheric tomography metapupil diameter) without physically altering the optical path by varying the amount of overlap between artificial guide stars, is also demonstrated to create atmospheric turbulence in the laboratory. Zernike mode gain factors are calculated from the Geometric WFS’s linearity response to individual Zernike mode wavefront phase aberrations, improving the calibration process. SSA requires imaging techniques – i.e. AO and computer post-processing techniques – to track Low Earth Orbital (LEO) space debris and satellites using groundbased telescopes. Existing imaging metrics, such as the Structural Similarity Metric (SSM) and regional properties, are re-purposed to quantify the quality of satellite and space debris images from ground-based telescopes. These imaging metrics are tested on observations of the International Space Station (ISS) from UCMJO, and satellite image deconvolution was demonstrated without a WFS, achieving moderate success by using only the spread of light from a background star as the Point Spread Function (PSF) estimate. The SSM helps identify the orientation of the ISS or other known satellites in relation to the observer. The regional properties method, which is easier to use and less subjective to human input, is ideal for lesser-known satellites with uncertain dimensions or characteristics.
  • ItemOpen Access
    University of Canterbury traffic study : development of SATURN and TRACKS models
    (2000) Andjic, Zarko
    This study is concerned with the development and comparison of two models for the University of Canterbury and its surroundings. Models are developed using two modelling packages, TRACKS and SATURN, and comparison of standard outputs from the two models has been done. The comparison of the two models was focused on the assignment modules. The two models have also been compared with some of the outputs from the PARAMICS model of the University of Canterbury. After the comparison further analysis was continued using SATURN for the validation and the assessment of car parking policies, and TRACKS for the presentation of results and statistical analyses. Both AM and PM peaks are modelled, but more emphasis was laid on the morning peak, in view of the fact that problems with the demand for parking occur when students and staff arrive at the University. The report also describes the matrix derivation from the models developed previously and using the latest census data. The matrix has been segmented into 4 user classes, namely student, staff, residents and through traffic. The development of student and staff matrices was aided by the geocoding of all university student addresses and using these data for the derivation of trip origins. Matrices of the through traffic and residential traffic were improved using the SATME2 program from SATURN. Calibration and validation of the model are explained in detail. A new method for modelling car parking demand was suggested. Instead of representing on-street and on-campus car parks as zone centroids, as is common practise in modelling, this study used one-way links. By limiting capacities on these links, it was possible to simulate variations in car park occupancies in different scenarios. The model was than used in multiple-user-class elastic mode to test three policy measures, which combined parking pricing and a demand responsive public transport throughout the city. The outputs from the three scenarios showed that the majority of students and staff shifting from cars to public transpo1i live outside the modelled area, which means that a city wide model would be more appropriate for modal split analysis. Finally, roadside pollution levels were analysed for the three policy measures. The decrease in the number of students and staff coming to university by car has resulted in attracting more through traffic using roads encompassing the University. Therefore, the reduction in pollution levels was not as high as expected. More detailed analysis is required to address pollution problems around the university, including some field measurements.
  • ItemOpen Access
    Optimisation of freight transportation networks - supply chain modelling of modal mixes and network configurations for optimisation on measures of cost, lead time and sustainability.
    (2023) Lyu, Zichong
    The research is funded by Callaghan Innovation (grant number MAIN1901/PROP-69059-FELLOW-MAIN) and the Ministry of Transport New Zealand in partnership with Mainfreight Limited. Need – The freight industry is facing challenges related to climate change, including natural hazards and carbon emissions. These challenges impact the efficiency of freight networks, increase costs, and negatively affect delivery times. To address these challenges, freight logistics modelling should consider multiple variables, such as natural hazards, sustainability, and emission reduction strategies. Freight operations are complex, involving various factors that contribute to randomness, such as the volume of freight being transported, the location of customers, and truck routes. Conventional methods have limitations in simulating a large number of variables. Hence, there is a need to develop a method that can incorporate multiple variables and support freight sustainable development. Method - A minimal viable model (MVM) method was proposed to elicit tacit information from industrial clients for building a minimally sufficient simulation model at the early modelling stages. The discrete-event simulation (DES) method was applied using Arena® software to create simulation models for the Auckland and Christchurch corridor, including regional pick-up and delivery (PUD) models, Christchurch city delivery models, and linehaul models. Stochastic variables in freight operations such as consignment attributes, customer locations, and truck routes were incorporated in the simulation. The geographic information system (GIS) software ArcGIS Pro® was used to identify and analyse industrial data. The results obtained from the GIS software were applied to create DES models. Life cycle assessment (LCA) models were developed for both diesel and battery electric (BE) trucks to compare their life cycle greenhouse gas (GHG) emissions and total cost of ownership (TCO) and support GHG emissions reduction. The line-haul model also included natural hazards in several scenarios, and the simulation was used to forecast the stock level of Auckland and Christchurch depots in response to each corresponding scenario. Results – DES is a powerful technique that can be employed to simulate and evaluate freight operations that exhibit high levels of variability, such as regional pickup and delivery (PUD) and linehaul. Through DES, it becomes possible to analyse multiple factors within freight operations, including transportation modes, routes, scheduling, and processing times, thereby offering valuable insights into the performance, efficiency, and reliability of the system. In addition, GIS is a useful tool for analysing and visualizing spatial data in freight operations. This is exemplified by their ability to simulate the travelling salesman problem (TSP) and conduct cluster analysis. Consequently, the integration of GIS into DES modelling is essential for improving the accuracy and reliability of freight operations analysis. The outcomes of the simulation were utilised to evaluate the ecological impact of freight transport by performing emission calculations and generating low-carbon scenarios to identify approaches for reducing the carbon footprint. LCA models were developed based on simulation results. Results showed that battery-electric trucks (BE) produced more greenhouse gas (GHG) emissions in the cradle phase due to battery manufacturing but substantially less GHG emissions in the use phase because of New Zealand's mostly renewable energy sources. While the transition to BE could significantly reduce emissions, the financial aspect is not compelling, as the total cost of ownership (TCO) for the BE truck was about the same for ten years, despite a higher capital investment for the BE. Moreover, external incentives are necessary to justify a shift to BE trucks. By using simulation methods, the effectiveness of response plans for natural hazards can be evaluated, and the system's vulnerabilities can be identified and mitigated to minimize the risk of disruption. Simulation models can also be utilized to simulate adaptation plans to enhance the system's resilience to natural disasters. Novel contributions – The study employed a combination of DES and GIS methods to incorporate a large number of stochastic variables and driver’s decisions into freight logistics modelling. Various realistic operational scenarios were simulated, including customer clustering and PUD truck allocation. This showed that complex pickup and delivery routes with high daily variability can be represented using a model of roads and intersections. Geographic regions of high customer density, along with high daily variability could be represented by a two-tier architecture. The method could also identify delivery runs for a whole city, which has potential usefulness in market expansion to new territories. In addition, a model was developed to address carbon emissions and total cost of ownership of battery electric trucks. This showed that the transition was not straightforward because the economics were not compelling, and that policy interventions – a variety were suggested - could be necessary to encourage the transition to decarbonised freight transport. A model was developed to represent the effect of natural disasters – such as earthquake and climate change – on road travel and detour times in the line haul freight context for New Zealand. From this it was possible to predict the effects on stock levels for a variety of disruption scenarios (ferry interruption, road detours). Results indicated that some centres rather than others may face higher pressure and longer-term disturbance after the disaster subsided. Remedies including coastal shipping were modelled and shown to have the potential to limit the adverse effects. A philosophical contribution was the development of a methodology to adapt the agile method into the modelling process. This has the potential to improve the clarification of client objectives and the validity of the resulting model.
  • ItemOpen Access
    The influence of wind on radiata pine tree shape and wood stiffness
    (2004) Bascunan Aldunate, Arturo
    This thesis explores the effects of wind on radiata pine (Pinus radiata) tree shape and wood stiffness, from the edge to the interior of a stand. The objective of the thesis is to detect, prior to felling, low stiffness trees or areas in a stand that due to wind will yield poor recovery of stiff lumber. The first three chapters review previous research. The first chapter focuses on the wood quality of radiata pine with special attention to stiffness, its causes and consequences. The second chapter introduces the principles of acoustic tools and how the forestry industry can use these to segregate raw material according to differences in wood quality. The third chapter describes the stress due to wind over a stand and how trees adapt internally and externally to wind. The next three chapter describe studies that were conducted in four narrow shelterbelts of radiata pine aligned perpendicular to the north-west winds. The chosen stands comprised three different age classes and two wind environments. At each stand, 30 transects were placed across the stand. At each transect, 20 trees were selected from the windward to the leeward edge of the stand. The outerwood stiffness of 2260 standing trees was assessed with the TreeTap acoustic tool. Results showed that stiffness and tree height increases with distance from the stand edge, while taper is reduced. The negative effect of wind on wood stiffness extends into the stand for a distance equivalent to one average tree height from both stand edges regardless of the age or location of the stand. The poorest trees were those located at the edges of the stand. High differences in outerwood stiffness between both sides of the stem were found, which increased with tree age and tree lean. A relationship could not be found between tree spacing to neighbours and tree outerwood stiffness. Knowledge of these low stiffness areas inside a stand will permit forest managers to develop different strategies to manage the variability in wood quality arising from wind. Future research might consider a sawmill study and further developing the methods utilized in this study in stands containing clonal trees.
  • ItemOpen Access
    Measuring VOCs in Christchurch ambient air : using SIFT-MS technique
    (2003) Chalermot, Warawan
    The project is aimed at monitoring the volatile organic compounds (VOCs) that may be regarded as undesirable in Christchurch's ambient air. Nine VOCs viz NH3 (ammonia), C2H4 (acetylene), C3H6OH (methanol), C2H5OH (ethanol), CH3CHO (acetaldehyde), C6H6 (benzene), C7H8 (toluene) and C8H10 (xylene) were-selected to be examined diurnally and seasonally. The samples were taken from various areas around the city. The selected ion flow tube mass spectrometry (SIFT-MS) technique was chosen as the analytical technique for the study as its fast analysis time and high sensitivity (10 ppb to 40 ppm) give it a significant advantage over more conventional methods. Among the results found were unexpected converse behaviour of the dominant species C3H6 (propene) and CH3OH (ethanol). In winter, C3H6 (propene) was the dominant pollutant of the nine selected VOCs whereas in summer it was C2H5OH (ethanol). Further, the concentration ratios of C6H6 (benzene) to C7H8 (toluene) are different from the value reported in other cities for urban atmospheric air. The presence of these species at the concentration levels found in the atmospheric conditions prevailing here are sufficiently high to be of concern and a monitoring program is advised.
  • ItemOpen Access
    Bus service reliability : bunching and arrival time predictability
    (2004) Kong, Meng Hui
    One of the aims of this research is to investigate the accuracy of predicted arrival times of buses provided by the Real-Time Tracking System (RTTS) currently being implemented in Christchurch. Surveys have shown an increase in bus patronage since its installation and the idea of improving bus services by providing accurate estimations to bus users is gradually rising in the government's priority list. This study will analyse the prediction errors of real-time information data collected manually from selected bus stops. Actual arrival times along with predicted times of arrival displayed at bus stops are gathered and compared. Comparisons between different routes and different directions for the Christchurch system are also studied. Real-time predictions are found to be reasonably accurate except for a few spots which can be rectified with improvements in the management system. Another purpose of this research is to improve an existing model for investigating bus bunching. Newell and Potts had done some research on the maintenance of a bus schedule in the 1960's and derived a formula to calculate bus departure times at each stop along the route, with or without delay to bus 1 occurring at stop 1. However, the model will break down once a bus is able to overtake the bus in front. This research tries to incorporate the overtaking feature into the existing model by setting up a numerical model in the Excel spreadsheet. Departure times are more realistic than those found using the formula derived by Newell and Potts. However, the extended model has its limitations. When buses deviate further and further away from schedule, it will eventually result in more than a pair of buses being bunched together. The extended model will fail because it is unable to cope with more than one pair of buses joining in a bunch. Therefore, more research should be done on this topic.
  • ItemOpen Access
    Quasinormal Modes of Black Holes
    (2023) Aimer, Jack William Alfred
    This thesis presents an expository account of the quasinormal modes of black holes in General Relativity, and the hyperboloidal approach to black hole perturbation theory, along with an extension of the hyperboloidal framework to the calculation of the quasinormal modes of the Kerr-Newman black hole. Perturbation equations are derived for the Schwarzschild, Kerr, and Kerr-Newman spacetimes, using axial metric perturbations for the former, and perturbations in the Newman-Penrose formalism for the latter two. For the Schwarzschild spacetime, quasinormal modes are calculated successfully using a direct numerical integration scheme. A novel hyperboloidal foliation of the Schwarzschild spacetime is then implemented, and quasinormal modes are calculated in this hyperboloidal framework using spectral methods. The two approaches are found to be in good agreement with each other, along with sources [41, 60] from the literature. For the Kerr spacetime, the direct numerical integration scheme is demonstrated not to generalise with sufficient accuracy. Following [66], the hyperboloidal framework is extended to the Kerr spacetime and the quasinormal modes are calculated using spectral methods. For comparison, we also calculate the quasinormal modes of the Kerr black hole using Leaver’s method [74]. The two methods are found to be in good agreement with each other. For the Kerr-Newman spacetime, the hyperboloidal framework is generalised using the I-fixing gauge of [51]. The perturbation equations of the Kerr-Newman black hole are recast in the corresponding compactified hyperboloidal coordinates, and asymptotic analysis in the Schwarzschild limit is found to justify our extension of the hyperboloidal framework.
  • ItemOpen Access
    Evaluation of the Paramics microsimulation model : central city area and urban arterial case studies
    (2001) Wilke, Axel Peter Carl
    This report describes the development of two Paramics microsimulation traffic models: a small area model for the inner city bus station in Christchurch and its immediate environs, and an area wide model for the Christchurch suburb of Riccarton. Microsimulation traffic models simulate individual vehicles, and their interactions, for their trips through a road network. Transport planners can undertake detailed assessment of current and future transport infrastructure and traffic management. The models have been used to research the application of Paramics for innovative and non­ standard designs, and to evaluate bus priority measures in congested road corridors. It has been concluded that Paramics microsimulation is suited for the task, and that the models are a valid description of the road network and its performance. Bus lanes have been found to be most effective in decreasing both bus travel times, and bus travel time variability, on the congested Riccarton Road. The bus station has since become operational, and it is evident that the inner city model was able to predict all operational problems of the road network and the bus station itself. It is considered that both models form a suitable means of teaching postgraduate students about traffic modelling, and many further research possibilities have been identified. The Christchurch City Council has expressed interest in extending the inner city model for testing of further traffic management changes, and taking advantage of the visual nature of microsimulation for subsequent consultation.
  • ItemOpen Access
    Transport energy reliance analysis for suburban shopping activities
    (2005) Saunders, Michael James
    Energy is a vital component of transport systems, which provides for access to activities. The finite nature of most transport energy sources may result in transport energy constraints as energy supplies diminish. Current transport models and transport planning rarely consider energy as an integral part of the transport system. The risk of future oil constraints and the possible impacts to the transport system is quantitatively unknown. New suburbs and developments are being designed that rely on automobiles and transport fuel consumption for access to fundamental activities. Access to these activities within these suburbs may be interrupted when energy constraints are realised. Of particular concern is access to basic necessities, such as food. A method was developed to assess transport energy reliance for food shopping activities within a suburb. When applied to a case study it was found that circa 60% reductions in energy consumption could be achieved with the addition of new food stores or through behaviour change. The addition of new food stores also allowed the suburban residents to feasibly access food stores without using transp01i energy. The method successfully incorporated energy into the simplified transport modelling of food shopping trips for a suburb. A strong relationship was also observed between the spatial pattern of the suburb and its energy consumption. This inferred that energy consumption for particular activities within a suburb could be considerably reduced through land-use planning.