No description available.

No description available.

No description available.

This document describes research into the development of software to allow the visualisation of finite and discrete element simulations within a virtual environment. The particular requirements of rendering deep level mining simulations are given precedence, but the resulting techniques apply in general to all finite element problems. In particular, a new rendering method based upon incremental frame updating is presented, and this provides the basis for two new rendering optimisations: <I>Priority Ordering </I>predicts the visibility of objects within the scene from the current viewpoint. This mapping of objects to a scalar value allows the incremental rendering step to process geometry in order such that the objects contributing the largest amount to the scene quality are rendered first. This is shown to provide an improvement in the quality of the rendered image as the incremental update step proceeds, when compared to conventional algorithms; <I>Z Occlusion using Priority Ordering </I>uses the coverage information obtained from the generation of priority values to allow the lazy evaluation of Z buffer area for occlusion culling. This is shown to decrease the time taken to render an entire frame to completion where sufficient depth complexity is present to outweigh the Z buffer access overhead, and to provide little or no performance penalty where little depth complexity is present, or Z buffer access is disproportionately high. The design of a highly threaded rendering system which allows the asynchronous operation of multiple parts of the system is also documented.

Considerable attention has been given to the use of data from satellite sensors to estimate land-surface biophysical properties. In particular, data acquired at different sensor view angles and solar illumination angles with respect to the Earth's surface have been combined with mathematical models of radiation scattering (also known as Bidirectional Reflectance Distribution Function (BRDF) models) to estimate albedo, LAI and fAPAR. One of the most widely used sources of multi-angle image data has been the NOAA/AVHRR series of satellites. Most BRDF research has focused on the development of various classes of reflectance models. Ultimately, however, all such models rely on the provision of well calibrated, cloud-masked, geometrically registered and atmospherically corrected image data. For the specific case of the NOAA/AVHRR, this study examines the effects and implications of each of these data pre-processing stages for BRDF studies. A number of important conclusions are reached. Firstly, there is considerable ambiguity in the literature regarding the radiometric calibration of AVHRR data. To this end, a consistent methodology is developed and presented. Second, a range of algorithms exist which may be used to identify clouds and cloud shadow in AVHRR data. The APOLLO algorithm (Saunders and Kriebel, 1988) is adopted because it is the most rigorous cloud-mask and studies suggest that BRDF models are sensitive to residual cloud contamination. Third, geometric correction is achieved through a combination of satellite navigation and ground control points. This stage is believed to introduce the greatest uncertainty into the provision of reliable directional reflectance data. Finally, atmospheric correction is performed using a fast radiative transfer code (SMAC). A series of modifications to SMAC are developed to provide an enhanced performance, similar to that of a more computationally-intensive code (6S).

The objective of the research is to improve the solution capability of a finite element system by the use of parallel processing techniques. Initial research began with the transputer system at a time when the transputer was considered the ultimate parallel machine. The transputer based parallel program LINSUB was developed to solve linear static stress analysis problems using a modified Cholesky algorithm for substruturing. Further developments in this field were halted when it became clear that other parallel systems, namely, those concerned with workstation clusters, were moving rapidly ahead of the transputer in terms of performance and the transputer was rapidly becoming obsolete. Further work concentrated on parallel processing on workstation clusters in a local area network and the effectiveness of the <B>P</B>arallel <B>V</B>irtual <B>M</B>achine (PVM) used as the supportive parallel environment was investigated. This work was then extended onto a shared memory - MIMD system on the Silicon Graphics Power Challenge workstation. The program PARFEI (<B>PAR</B>allel <B>F</B>inite <B>E</B>lement <B>I</B>mplementation) was developed for large linear static stress analyses and nonlinear analyses of elasto-plastic problems in particular, using a multifrontal approach based on domain decomposition techniques. An implementation based on a specially designed share memory (SM) synchronization concept was incorporated into the same program, so that PARFEI can run sequentially or in parallel using either PVM or SM. Performance tests were carried out on the workstation clusters using PVM and on the shared memory machine using PVM and SM. The results revealed that, in general, PVM performed better for the linear system and SM was the more efficient of the two when an incremental/iterative process was involved.

This thesis is concerned with the development and application of reliable, creative and efficient computational tools for the analysis and structural optimization of 2D and 3D trusses under static and free vibration conditions and prismatic structures under static and free vibration conditions. The first part of the thesis deals with the analysis and optimization of 2D and 3D trusses under static and free vibration conditions. The second part of the thesis deals with prismatic structures under static and free vibration conditions which are idealised as effectively one-dimensional problems. The procedures used to define the structural shape and incorporating automatic mesh generation are described and the basic formulations of curved, Mindlin-Reissner finite elements and finite strips are presented for prismatic shells. The accuracy of the elements and strips is verified using several benchmark tests. The basic algorithm of structural shape optimization is then presented and several examples are carried out. In both parts the optimization analysis is carried out using two different algorithms: sequential quadratic programming (SQP) and genetic algorithms (GAs). The optimized structures are not checked for buckling. Six main computer FORTRAN programs are developed, documented and tested.

This thesis introduces a texture descriptor that is invariant to rotation. The new texture descriptor utilizes the property of the magnitudes of Fourier transform coefficients that do not change with spatial shift of input elements. Since rotating an image by an arbitrary angle does not change pixel intensities in an image but shifts them in circular motion, the notion of producing texture features invariant to rotation using 1-D Fourier transform coefficients can be realized if the relationship between circular motion and spatial shift can be established. By analyzing individual circular neighbourhoods centered at every pixel in an image, local and global texture attributes of the image can be described. Rotating the image has a similar effect as spatially shifting the pixels in the circular neighbourhood around without altering their intensities. A number of sequences can be formed by the intensities of pixels at various fixed distances from the center of the neighbourhood. Fourier transforming the sequences would generate coefficients that contain the texture information of the neighbourhood. From the magnitudes of these coefficients, several rotation invariant features are obtained. The capabilities of the new features are investigated in a number of classification and segmentation experiments. The experimental results compare favourably with those of prominent descriptors like the circular autoregressive model, the wavelet transform, the Gaussian Markov radom field and the co-occurrence matrix. In the majority of the instances, the new method shows superior performance.

This thesis introduces a new approach to the design of bench-top instrument systems, the Virtual Instrument Workstation (VIW). The VIW sets out to produce a much more flexible, modular instrumentation system than conventional designs, and one which is also cost effective. The main thrust of the work is to break down conventional instrument systems into clearly defined hardware and software modules that can then be shared concurrently among a number of Virtual Instruments (VIs). The use of a touch sensitive, window managed, software driven CRT display is central to the concept. This allows the high level of user interaction with the system and multiple VIs to be accomplished with an easy-to-understand front-panel. A special Instrument Bus system is defined to allow the easy production of instrument modules for use in the system. The communications between the instrument modules is unified, such that all data can be passed between all modules. VISYS, a real-time multi-tasking, operating system is developed to provide a software environment for the VIW. This system is UNIX based and, has many real-time extensions. The VIW described has the ability to perform the operation of several separate instruments simultaneously. The range presented in the work includes a Signal Generator, a Digital Multi-Meter, an Oscilloscope and a Component Meter. Both the hardware and software aspects of the VIW are presented in detail.

Mesh generation and visualization is a bottleneck problem in all finite element based numerical simulations. The research work reported in this thesis is a step towards integration of a 3D unstructured mesh generation toolkit. Both theoretical and implementational aspects are addressed. The general concepts of mesh generation have been presented. In the light of these concepts, the fundamental aspects of Delaunay triangulation for the unstructured mesh generator have been formulated. Steiner point creation algorithm has been examined by means of numerical experiments with respect to mesh control parameters. Concepts of point spacing tensor and point insertion criterion have been introduced aiming to deal with anisotropic meshes. Surface meshes are created based on 2D geometries by means of coordinate mapping. Triangular and quadrilateral patches in linear and quadratic forms and Non-Uniform Rational B-Spline (NURBS) patches have been used to define surface geometry. The mesh quality improvement has been discussed in terms of parametric plane stretching, diagonal swapping and smoothing procedures. Volume meshes are generated through 3D triangulation and interior point creation based on the surface meshes. Boundary surface conformity is gained <I>via</I> edge swapping, boundary edge and surface recovery, and the robustness of the algorithm has been discussed in a view of the accuracy of geometric judgements. Quality metrics for tetrahedral elements, mesh smoothing technique, visual quality assessment, and adaptivity extension have been discussed. Data conversion between this mesh generator and some existing CAD packages has been considered. Additionally, a scheme of interactive topology specification has been put forward for a 3D multiblock mesh generator. Finally, a finite element visualization facility FEView has been presented. This visualization tool is implemented based upon a object-oriented graphics library, and works as an external module to an interactive program Geomview. Within the FEView, a mesh can be considered as a collection of faces with edges, wire frame, point and element clouds, and the FE analysis results can be visualized <I>via</I> color shading and field icons. FEView provides functions such as domain control, animation control, and local analysis. Numerical results for 2D cases can be shown with 3D effects by using values of the scalar field.