UNSW Canberra

Publication Search Results

Now showing 1 - 10 of 17
  • (2022) Xie, Yuekai
    Thesis
    The increase in waste disposal increases the demand for better design, construction, operation and management of the landfills. A common method to accelerate the decomposition, stabilization and settlement of municipal solid wastes (MSWs) is leachate recirculation. However, slope failure occurs due to the poor design and management of the recirculation and gas collection systems, together with the change in the geotechnical properties of MSWs due to decomposition. Based on the field compositions at a working cell in Mugga Lane Landfill, ACT, Australia, a series of experiments have been conducted to study the effects of leachate exposure on the geotechnical properties of MSWs with different compositions. The experimental programs include compaction, saturated hydraulic conductivity, direct shear, consolidated undrained triaxial shear and one-dimensional consolidation. Following this, fresh waste samples with different compositions were prepared in the simulators and recirculated with water or leachate to enhance the biodegradation. The samples with the same dry unit weights but different degrees of biodegradation (DOB) were tested to address the effects of decomposition on the geotechnical properties of the MSWs. Despite of the change in the composition and particle sizes, the variation in the leachate properties significantly affects the geotechnical properties of the MSWs. Three small laboratory scale and one larger laboratory scale, and four field scale bioreactors were set up in the laboratory or landfill site. The objectives are to investigate the effects of initial compaction, recirculation liquids and rates, loading levels and environmental temperatures on the long-term physical, mechanical and biochemical behaviours of the MSWs. The measured and analyzed properties include the settlement, initial creep, bio-compression and residual creep ratios, quantity and quality of leachate generation, rate and concentration of methane generation. A new equation was developed to capture the change in the methane generation rate with time. Further experiments were conducted to evaluate the water retention curves (WRCs) of MSWs under different degrees of biodegradation, dry unit weights and paper contents. Fresh MSW samples and soil-paper mixture were compacted to different initial dry unit weights and then decomposed for different periods to the same dry unit weight before testing. It was found that, in addition to dry unit weights of MSW, the effects of decomposition and paper contents on the unsaturated properties of the MSWs depend on the balance between the percentages of biodegradable and highly decomposed components. Two settlement profilers were installed in the landfill site to obtain the settlement of an MSW lift in a working cell of the landfill site. A global navigation satellite monitoring system was installed to monitor the settlement of three closed landfill cells. A settlement model was developed based on the change in the solid, liquid and gas phases of MSWs due to decomposition. The proposed settlement model was validated with the results from the laboratory and field scale bioreactors and in-situ monitoring data.

  • (2022) Nguyen, Tung
    Thesis
    Transparency ensures that the decision-making logic used by autonomous agents is available in a form comprehensible to human teammates. It enhances a system's performance, reliability and trustworthiness. Transparency in human-autonomy teaming is a key success factor in the communication between humans and autonomous agents. This thesis focuses on the problem of knowledge transfer between two neural network (NN) agents in a swarm-guidance task in the presence of a human observer. A novel three-module knowledge transfer framework is proposed to interpret non-symbolic knowledge of communicative autonomous agents into a transparent human-friendly form. The knowledge interpretation module transforms the NN into a rule-based knowledge representation. The relevant knowledge is then chosen by a knowledge selection module to transfer it to the other agent. Finally, the knowledge fusion module combines the newly incoming knowledge with the receiver agent's existing knowledge. Two algorithms are introduced to transform NNs into a rule-based knowledge representation. The first algorithm, the Exact-Convertible Decision Tree (EC-DT), rewrites the relationships between nodes and weights of the NN into a multivariate decision tree. The second algorithm, the Extended C-Net, leverages the training data to learn the association between the NN's nodes at the final hidden layer and the outputs and then uses recursive back-projections to derive the rules regulating input-output relationships. Performance is assessed using three measures: fidelity, compactness and transparency. EC-DT has higher fidelity, while Extended C-Net produces more compact rule sets. The fusion module then evaluates the effectiveness of transmitted rule-based knowledge in new environments. The rule-based representation is projected back into forms that can be integrated with the NN representation at the receiver agent's end. A retraining strategy, Priority on Weak State Areas (PoWSA), is introduced to help speed up the learning process in novel scenarios. Analyses of the proposed methodology in a swarm-guidance problem show higher training stability and chances of success in return for a slight increase in computational costs. The framework provides a more transparent knowledge representation that could be visualised to complement the verbal rule-based representation.

  • (2022) As'ham, Khalil
    Thesis
    Light-matter interaction within a strong coupling regime has attracted much attention because of its potential applications in quantum manipulation, Bose−Einstein condensation, optical transistors, coherent emission or absorption, photovoltaics, ultrafast optical switching, sensing, low-threshold lasers, quantum fluid of light and all-optical logic devices. However, conventional materials used in strong coupling suffer from several challenges, as follows: low working temperature, small binding energy, low-quality factor and challenging fabrication. To overcome these limitations, transition metal dichalcogenides (TMDCs) and 2D lead halide perovskite materials are used to achieve a strong coupling regime between optical and polariton modes. This thesis initially reviews previous studies that used TMDCs or perovskite materials in a strong coupling regime. Owing to TMDCs’ optical properties, several studies reported the strong coupling between exciton and optical cavities. However, current structures face many challenges, including difficulties in fabrication and the associated cost increase. The strong coupling between exciton in TMDCs monolayer, plasmonic resonance in silver and anapole mode in Silicon nanodisk was analysed at room temperature and normal incident. This nanostructure provides large Rabi splitting accompanied by a giant field enhancement, thereby paving the way for the creation of exciton-polariton devices. Although previous studies used metallic nanocavities to excite plasmonic resonances that offered a small mode volume and strongly coupled them with exciton in TMDCs, damping losses in metals degraded the performance of exciton-polariton systems. Consequently, researchers used dielectric materials as an alternative to plasmonic materials to observe a strong coupling regime. However, exciton-polariton devices that use dielectric materials suffer from large mode volumes. This thesis explores the strong coupling between hybrid resonance and exciton in hybrid dielectric-metallic nanostructure and TMDC monolayer by using metallic and dielectric materials that include small mode volume low losses. Another material with potential for strong coupling is lead-halide perovskite. Lead halide perovskite materials have a larger binding energy at room temperature compared with conventional materials. Thus, the strong coupling between exciton of perovskite and photonic cavity has been demonstrated in different platforms, such as plasmonic nanocavity and distributed Bragg reflector-based microcavity. The former suffers from large intrinsic loss due to the nature of noble metals, and the latter may involve a complex fabrication process. In recent years, researchers have also used a guided-mode resonance supported by a photonic crystal slab to achieve exciton-polariton in perovskite metasurface. In this thesis, the strong coupling between Mie resonances and exciton is explored at room temperature without resorting to other photonic cavities. Finally, at the end of this thesis, the future perspective and a summary of this research's main results are presented.

  • (2023) Salim, Sara
    Thesis
    Web 3.0 represents the third generation of web technologies for incorporating decentralisation and agility in web applications. As it integrates Social Media (SM) in the Internet of Things (IoT), the endless synergies established promise consumers greater connectivity and interaction as well as more seamless movement between physical spaces. Despite these advantages, this integration also increases issues of cyber vulnerabilities and cyberattacks that cause financial, political and social damage in such data-rich environments. Although machine learning underpins this transition, the data used may contain sensitive information that could be compromised by privacy and security breaches. Therefore, techniques that maintain the utility of such valuable data while protecting individuals' privacy are increasingly being required, with federated learning-based privacy preservation ones the current standard. However, as federated learning approaches still require a central authority, they enable significant cyberattacks. In this thesis, a major contribution is the safeguarding of heterogeneous data, such as that in IoT-integrated SM networks (i.e., SM 3.0). It does so by introducing several new contributions, a novel privacy preservation IoT-integrated SM framework and extended and improved federated learning-based ones that intensively leverage the power of the privacy definition involved in differential privacy and the trustworthiness of blockchain modules to protect against privacy attacks. Apart from improving privacy preservation in conventional federated learning-based frameworks in future social platforms, in this work, accountability and trustworthiness are investigated, with blockchain-based systems integrated into the developed frameworks. Simultaneously, by applying privacy preservation techniques, a preserved and more statistical data version of IoT-integrated SM data is offered to machine learning-based applications to maintain high utility levels. The experimental results reveal the following two key properties; the proposed frameworks yield noticeable performance improvements with high privacy and comparable utility levels while also allowing enhanced recognition of user preferences in the SM 3.0 dataset developed as part of this work to address the unavailability of data sources as well as highly classifying a user's actions based on observations of the surrounding environment in the IoT datasets. Furthermore, since the proposed frameworks attain differential privacy on the training data, they are considered the same as standard federated learning but with a greater level of privacy preservation. In future, these frameworks will be extended by examining their integration with other distributed learning ones.

  • (2023) Qasim, Muhammad
    Thesis
    In recent years, steel and polyvinyl-alcohol hybrid fibre reinforced engineered cementitious composites (SPVH-ECC) has emerged as a high-performance construction material due to its high strength and ductility. However, the application of SPVH-ECC for the strengthening of reinforced concrete (RC) beams has not yet been studied thoroughly. This thesis evaluated the interfacial bond behaviour between SPVH-ECC and concrete and analysed the flexural performance of strengthened RC beams using SPVH-ECC. The interfacial bond behaviour between SPVH-ECC and concrete was studied by conducting slant shear tests and splitting tensile tests. In order to study the effect of hybrid fibre reinforced ECC against mono fibre ECC, the interfacial bond strength between SPVH-ECC and concrete was compared to that of between polyvinyl-alcohol engineered cementitious composites (PVA-ECC) and concrete. The influence of interfacial surface roughness on bond strength between concrete and ECCs was also studied by considering as-cast and sandblasted interfaces. Interfacial surface roughness was measured using both quantitative and qualitative assessment methods. The obtained results of both methods were compared to access the performance of the two types of ECCs studied under different surface treatment conditions. A numerical study on interfacial bond behaviour between SPVH-ECC and concrete was also performed to analyse the stresses at the interface. A surface-to-surface cohesive contact model was defined. The numerical models were validated by comparing the computed results with those obtained from experiments. The application of SPVH-ECC for flexural strengthening of RC beams was investigated through four-point bending tests. The flexural behaviours of one unstrengthened 3500 mm long, 200 mm wide, and 325 mm deep RC beam and three RC beams strengthened with different configurations (SB-1, SB-2, and SB-3) of 50 mm thick SPVH-ECC layer(s) were studied. The numerical simulation of strengthened RC beams was also performed. The results of the numerical simulation were validated against the experimental results and a good agreement was observed. The validated numerical models were used to conduct a small-scale parametric study. Furthermore, simple analytical models were developed to predict the ultimate capacity of the strengthened RC beams using SPVH-ECC with embedded reinforcement bars.

  • (2023) Al-Ani, Ibrahim
    Thesis
    The study of light–matter interaction in semiconductors is a major topic in modern photonics and quantum research due to its great potential for creating high-performance photonic and optoelectronic devices, such as detectors, sensors, switches, modulators, and lasers. The recently emerging novel two-dimensional (2D) semiconductor materials, such as transition metal dichalcogenide (TMDC) monolayers and lead halide perovskite, have attracted huge research attention owing to their interesting physical and chemical properties. In this thesis, high Q-factor dielectric metasurfaces have been considered for the enhancement of light–matter interaction in these 2D materials, given their easy fabrication process, low dissipative losses, and strong field confinement. The thesis initially has provided an introduction and a review of recent advancements in light–matter interaction enhancement by TMDC monolayers and perovskite. Then, using a 1D grating made of silicon nitride, the enhancement of the weak coupling regime has been theoretically and experimentally demonstrated in TMDC monolayers. The emission of a WS2 monolayer has been enhanced by 32 times, whereas its absorption has been increased by up to 100%. Next, the enhancement of the strong coupling regime has been demonstrated by using an all-dielectric meta surface that supports a quasi-bound state in the continuum (Q-BIC) mode. The Rabi-splitting in WS2 monolayer has been increased to 65 meV by utilising the extreme electric field confinement enabled by the Q-BIC mode. Afterwards, using all-perovskite metagratings, a strong coupling enhancement has been demonstrated between the exciton of lead halide perovskite semiconductors and QBIC. The maximum Rabi splitting has been enhanced up to a record-high value of 400 meV. Finally, a double strong coupling regime in perovskite and WS2 monolayer, which has been obtained by using either single or dual Q-BIC mode, has been demonstrated in a single structure. Except for the supporting exciton resonance, the perovskite thin film has been patterned as a grating to support single or dual Q-BICs by harnessing its high index. This paper is the first demonstration of a double strong coupling regime using dual photonic modes coupled separately to the excitons of perovskite and WS2 monolayer in a single nanostructure. The findings of this thesis can pave the way for the development of photonic, optoelectronic, and novel polaritonic devices based on TMDC and/or perovskite semiconductor materials.

  • (2022) Reza, Wasim
    Thesis
    In recent years, the use of hybrid structures in the construction industry has increased significantly for various reasons, the most important of which are concerns about environmental sustainability and economic benefits. The present study is a part of a comprehensive research project associated with the development of hybrid structural systems. More precisely, this thesis investigates the structural performance of a hybrid timber-steel beams with web openings subjected to bending. The beam consists of an elongated thin galvanised steel web with press-formed stiffening ribs and flanges made of Laminated Veneer Lumber (LVL) fixed by steel nails to each side of the continuous steel web. A series of laboratory experiments were carried out to analyse the structural performance of this beam. First, a series of four-point bending tests were conducted in order to assess its flexural behaviour and load-carrying capacity, as well as to identify the failure mechanisms. Afterwards, a series of standardised tests were also performed to obtain the respective mechanical characteristics of the wood (LVL) and the steel used to make the hybrid beam. In the second phase, a three-dimensional finite element model of the beam was developed and verified with experimental results, with good agreement. It was developed with the aim of understanding all the stages of the failure modes and the interaction between different timber and steel parts. Furthermore, the FE model was used to carry out a parametric study to experiment with different configurations of hybrid beams. Finally, to facilitate the preliminary design of the beam, a closed-form analytical solution was formulated to obtain the ultimate load capacity and the effective bending stiffness with the load-displacement response of the beam. The closed-formed solution results were compared and verified with the experimental results as well as with the numerical results from the parametric study of different variations of the hybrid timber steel beam with good agreement. To conclude, it was found that in order to exhibit ductile response, the steel web must be relatively thin and elongated so that it fails first in a ductile manner before timber cracking, a behaviour similar to that of an under-reinforced concrete beam. On the other hand, a thin and extended steel web causes lateral instability and premature failure. Consequently, lateral supports are required to prevent lateral instability and obtain the true load-carrying capacity of the hybrid beam.

  • (2022) Dam, Tanmoy
    Thesis
    Generative adversarial networks (GAN) have attracted significant attention from the research community due to their superior clustering and classification abilities. However, data imbalance and continual learning settings cause poor GAN’s performance. In class imbalance problems, the thesis focuses on GAN-based novel data augmentation and oversampling strategies. In GAN-based augmentation, a three-player adversarial GAN game method called class-dependent mixture generator spectral GAN (MGSGAN) is introduced. MGSGAN plays an adversarial game with a discriminator and hyper-spectral classifier to improve its performance by considering the actual data and class-conditionals augmentation generated samples. Our experiments revealed that MGSGAN outperforms other leading approaches by a significant margin for hyperspectral image classification. A second method relies on oversampling techniques, where two adversarial oversampling strategies (adversarial oversampling and adversarial data space oversampling) are introduced with a mixtures of generator parameters updated through classifier parameters. A novel latent preserving single generator under the proposed three-player GAN game settings is introduced. We present four strategies covering all possible GAN games that can be played under three-player settings. The methods were competitive to the state-of-the-art GAN-based oversampling methods. GAN-based adversarial continual learning (ACL) is a recently developed iterative process where a shared feature network is played in an adversarial game with task-specific discriminators. Each task-specific network is introduced to handle the catastrophic forgetting of past tasks. As the job grows, the network structure of ACL also grows simultaneously. Therefore, ACL is not scalable. We propose an online Scalable Adversarial Continual Learning (SCALE) framework where the parameter generator transforms shared features into task-specific features. SCALE surpasses popular baseline methods on accuracy and execution time. Finally, in conjunction with a clustering inference network, a uniform prior-based conditional generative model, ClusterGAN, has recently achieved sub-optimal clustering performance. ClusterGAN may fail to generate all the modes due to disentanglement in the clustered inference network that can’t be fully achieved through only generative. SIMI-ClusterGAN is a new method where a prior network is introduced together with three additional losses to increase disentanglement in the clustering inference network. The method has been validated through benchmark datasets in which balanced and imbalanced scenarios have been considered.

  • (2022) Hassan, Md Kamrul
    Thesis
    Crumb Rubber Geopolymer Concrete (CRGPC) comprises geopolymer cement, water, coarse aggregates, and fine aggregates, which are partially or wholly replaced by crumb rubber made from scrap tyres. It is an environmentally friendly alternative to standard concrete and incorporates waste materials to promote a circular economy. The performance of the CRGPC under various structural loads needs to be well understood before it can be used as a replacement for standard concrete in structural applications. In this respect, this study aims to investigate the compressive behaviour of CRGPC under various passive confinement conditions. To achieve this objective, an experimental program was undertaken to measure the compressive strength, modulus of elasticity and stress-strain behaviour of various CRGPC under different confinement conditions. The CRGPC mixes were obtained by taking two different Geopolymer Concrete (GPC) mixes of different strengths and replacing 5%, 15% and 25% of the fine aggregates with the crumb rubber. Additionally, the effect of pre-treatment of the crumb rubber was investigated by comparing the behaviour of CRGPC with treated rubber with the one with un-treated rubber. In this way, a total of two GPC mixes and twelve CRGPC mixes were tested in the experimental program. Multiple batches of the same mix were used to make the test specimens for reliable test data. Two different passive confinement conditions were tested, one provided by the steel ties in a reinforced concrete column and the other of concrete-filled steel tubes (CFST). Tie confinement was varied by changing the tie spacing, whereas the tube confinement was varied by changing the tube thickness. Concrete behaviours under these confinements were compared with their unconfined compressive behaviour. The test results were then used to benchmark the performance of the existing models for standard concrete and geopolymer concrete on predicting the compressive behaviour of CRGPC under test conditions. Finally, several models are proposed to accurately predict the behaviour of CRGPC under unconfined and passively confined conditions.

  • (2023) Rawat, Sanket
    Thesis
    The brittleness, poor fire performance and durability issues associated with high strength concrete have restricted the resilience of concrete infrastructure to some extent. With rapidly increasing global demand of infrastructure development and repair, it becomes imperative to overcome these limitations. Engineered cementitious composite (ECC) is one of the alternate material types with superior tensile performance and durability due to its multiple fine cracking behaviour. However, behaviour of high strength ECC (HSECC) at elevated temperatures remains questionable so far. In general, the addition of polypropylene or polyvinyl alcohol fibres in ECC has proved to be an effective method to improve its thermal performance and spalling resistance. Nevertheless, both types of fibres may not be suitable for attaining the ultra-high performance. As building fires have relatively low occurrence, it is equally important to consider performance at room temperature while designing the material for fire resistance, an aspect which most of the studies overlook. Therefore, this doctoral research attempts to explore this viewpoint by systematically developing a new type of HSECC and investigating its mechanical performance under different thermal exposure conditions. Firstly, an experimental study was designed to find the optimum mix proportion of the control parameters. The key variables included volume of polyethylene and steel fibre, water-binder ratio, and weight of dolomite powder, ground granulated blast furnace slag, and fly ash. Taguchi method with grey relational analysis and utility concept was utilized to perform the multi-response optimization. An optimum mix design was proposed along with the most fitting optimization method which not only leads to the mechanical (compression and tensile) performance suitable for most of the structural applications but also simultaneously uses constituents with potentially better thermal performance. Thereafter, the residual mechanical properties of the developed mix were evaluated under temperature exposure ranging from 200 to 800°C. Subsequently, the effect of pre-drying, water quenching, specimen size, and heating rate on residual compressive strength was also investigated. In addition, different post-fire curing methods and the resulting extent of recovery were analysed. These observations were further related with microstructure and phase change analysis performed using scanning electron microscope, X-ray diffraction technique and thermo-gravimetric analysis.